LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology &...

143
Volume 6, Number 4, July 2019 Neurology.org/NN A peer-reviewed clinical and translational neurology open access journal ARTICLE α4-integrin deficiency in B cells does not affect disease in a T-cell–mediated EAE disease model e563 ARTICLE Harmful neutrophil subsets in patients with ischemic stroke: Association with disease severity e571 ARTICLE CSF parameters associated with early MRI activity in patients with MS e573 ARTICLE Pilot study of a ketogenic diet in relapsing-remiing MS e565

Transcript of LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology &...

Page 1: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Volume 6, Number 4, July 2019Neurology.org/NN

A peer-reviewed clinical and translational neurology open access journal

ARTICLE

α4-integrin defi ciency in B cells does not aff ect disease in a T-cell–mediated EAE disease model e563

ARTICLE

Harmful neutrophil subsets in patients with ischemic stroke: Association with disease severity e571

ARTICLE

CSF parameters associated with early MRI activity in patients with MS e573

ARTICLE

Pilot study of a ketogenic diet in relapsing-remitt ing MS e565

Page 2: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

TABLE OF CONTENTS Volume 6, Number 4, July 2019 Neurology.org/NN

Editor’s Corner

e578 Time is tissue in optic neuritisJ. Dalmau

Open Access

Editorial

e570 Contribution of polymorphonuclear neutrophils inthe blood periphery to ischemic brain injuryD.M. Hermann and M. Gunzer

Open Access Companion article, p. e571

Articles

e571 Harmful neutrophil subsets in patients with ischemicstroke: Association with disease severityD. Weisenburger-Lile, Y. Dong, M. Yger, G. Weisenburger,G.F. Polara, T. Chaigneau, R.Z. Ochoa, B. Marro, B. Lapergue,S. Alamowitch, and C. Elbim

Open Access Editorial, p. e570

e563 α4-integrin deficiency in B cells does not affectdisease in a T-cell–mediated EAE disease modelR.Z. Hussain, P.D. Cravens, W.A. Miller-Little, R. Doelger,V. Granados, E. Herndon, D.T. Okuda, T.N. Eagar, and O. Stuve

Open Access

e565 Pilot study of a ketogenic diet in relapsing-remittingMSJ.N. Brenton, B. Banwell, A.G.C. Bergqvist, D. Lehner-Gulotta,L. Gampper, E. Leytham, R. Coleman, and M.D. Goldman

Open Access Class of Evidence

e566 Different MRI patterns in MS worsening afterstopping fingolimodC. Lapucci, D. Baroncini, M. Cellerino, G. Boffa, I. Callegari,M. Pardini, G. Novi, M.P. Sormani, G.L. Mancardi, A. Ghezzi,M. Zaffaroni, A. Uccelli, M. Inglese, and L. Roccatagliata

Open Access

e572 Does time equal vision in the acute treatment ofa cohort of AQP4 and MOG optic neuritis?H. Stiebel-Kalish, M.A. Hellmann, M. Mimouni, F. Paul, O. Bialer,M. Bach, and I. Lotan

Open Access Class of Evidence

e573 CSF parameters associated with early MRI activityin patients with MSA. Klein, R.C. Selter, A. Hapfelmeier, A. Berthele, B. Muller-Myhsok,V. Pongratz, C. Gasperi, C. Zimmer, M. Muhlau, and B. Hemmer

Open Access

e574 Natalizumab treatment reduces microglial activationin the white matter of the MS brainM. Sucksdorff, J. Tuisku, M. Matilainen, A. Vuorimaa, S. Smith,J. Keitila, J. Rokka, R. Parkkola, M. Nylund, J. Rinne, E. Rissanen, andL. Airas

Open Access

e575 Guillain-Barre syndrome and related diseases afterinfluenza virus infectionM. Yamana, M. Kuwahara, Y. Fukumoto, K. Yoshikawa,K. Takada, and S. Kusunoki

Open Access

e576 Age matters: Impact of data-driven CSF proteinupper reference limits in Guillain-Barre syndromeP.R. Bourque, J. Brooks, C.R. McCudden, J. Warman-Chardon, andA. Breiner

Open Access

e580 Neuromyelitis optica spectrum disorder: Patientexperience and quality of lifeJ. Beekman, A. Keisler, O. Pedraza, M. Haramura,A. Gianella-Borradori, E. Katz, J.N. Ratchford, G. Barron, L.J. Cook,J.M. Behne, T.F. Blaschke, T.J. Smith, and M.R. Yeaman

Open Access

Continued

Copyright © 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

Page 3: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

e581 Trial of canakinumab, an IL-1β receptor antagonist,in patients with inclusion body myositisM.L. Kosmidis, D. Pikazis, P. Vlachoyiannopoulos, A.G. Tzioufas, andM.C. Dalakas

Open Access Class of Evidence

Clinical/Scientific Notes

e569 IgA autoantibodies against native myelin basicprotein in a patient with MSH. Schumacher, N.K. Wenke, J. Kreye, M. Holtje, K. Marcus,C. May, and H. Pruss

Open Access

e577 Immunotherapy improves sleep and cognitiveimpairment in anti-IgLON5 encephalopathyV. Brunetti, G. Della Marca, G. Spagni, and R. Iorio

Open Access Video

e579 Cerebellar ataxia as a presenting symptom ina patient with anti-NMDA receptor encephalitisM.H.F. Poorthuis, J.L.M. van Rooij, A.H. Koch, A.E.M. Verdonkschot,M.M. Leembruggen, and M.J. Titulaer

Open Access

Views & Reviews

e567 Strategies for treatment of childhood primaryangiitis of the central nervous systemJ. Beelen, S.M. Benseler, A. Dropol, B. Ghali, and M. Twilt

Open Access

e568 Immunoglobulin G4-related hypertrophicpachymeningitis: A case-oriented reviewM. Levraut, M. Cohen, S. Bresch, C. Giordana, F. Burel-Vandenbos,L. Mondot, J. Sedat, D. Fontaine, V. Bourg, N. Martis, andC. Lebrun-Frenay

Open Access

Diagnostic and Treatment Challenges

e582 Cataclysmically disseminating neurologicpresentation in an immunosuppressed lupuspatient: From the National Multiple SclerosisSociety Case Conference ProceedingsC.M. Perrone, R.P. Lisak, E.I.Meltzer, P. Sguigna, E. Tizazu, D. Jacobs,E. Melamed, A. Lucas, L. Freeman, G. Pardo, A. Goodman, E.J. Fox,K. Costello, M.S. Parsons, S.S. Zamvil, E.M. Frohman, andT.C. Frohman

Open Access

Correction

e585 α4-integrin deficiency in B cells does not affectdisease in a T-cell–mediated EAE disease model

Cover imageThe strong binding of a multiple sclerosis patient’s IgA antibodies toaxonal fiber tracts on a wildtype mouse brain section (left) wascompletely absent in shiverer MBP knockout littermate mice (right).See page e569

TABLE OF CONTENTS Volume 6, Number 4, July 2019 Neurology.org/NN

Copyright © 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

Page 4: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Academy OfficersJames C. Stevens, MD, FAAN, PresidentOrly Avitzur, MD, MBA, FAAN, President ElectAnn H. Tilton, MD, FAAN, Vice PresidentCarlayne E. Jackson, MD, FAAN, SecretaryJanis M. Miyasaki, MD, MEd, FRCPC, FAAN, TreasurerRalph L. Sacco, MD, MS, FAAN, Past President

Executive Office, American Academy of NeurologyCatherine M. Rydell, CAEChief Executive Officer201 Chicago AveMinneapolis, MN 55415Tel: 612-928-6100

Editorial OfficePatricia K. Baskin, MS, Executive EditorKathleen M. Pieper, Senior Managing Editor, NeurologySharon L. Quimby, Managing Editor, Neurology® Clinical PracticeMorgan S. Sorenson, Managing Editor, Neurology® Neuroimmunology& Neuroinflammation

Lee Ann Kleffman, Managing Editor, Neurology® GeneticsAndrea Rahkola, Production Editor, NeurologyRobert J. Witherow, Senior Editorial AssociateKaren Skaja, Senior Editorial AssociateKaitlyn Aman Ramm, Editorial AssistantKristen Swendsrud, Editorial AssistantJustin Daugherty, Editorial AssistantMadeleine Sendek, MPH, Editorial Assistant

PublisherWolters KluwerBaltimore, MD

Publishing StaffKim Jansen, Executive PublisherJessica Heise, Production Team Leader, Neurology JournalsMegen Miller, Production EditorStacy Drossner, Production Associate

Neurology® is a registered trademark of the American Academy of Neurology(registration valid in the United States).

Neurology® Neuroimmunology & Neuroinflammation (eISSN 2332-7812) is anopen access journal published online for the American Academy of Neurology,201 Chicago Ave, Minneapolis, MN 55415, by Wolters Kluwer Health, Inc. at14700 Citicorp Drive, Bldg 3, Hagerstown, MD 21742. Business offices arelocated at Two Commerce Square, 2001 Market St., Philadelphia, PA 19103.Production offices are located at 351 West Camden Street, Baltimore, MD21201-2436. © 2019 American Academy of Neurology.

Neurology® Neuroimmunology & Neuroinflammation is an official journal of theAmerican Academy of Neurology. Journal website: Neurology.org/NN, AANwebsite: AAN.com

Copyright and Permission Information: Please go to the journal website(www.neurology.org/nn) and click the Permissions tab for the relevant article.Alternatively, send an email to [email protected]. Generalinformation about permissions can be found here: https://www.lww.com/journal-permission.

Disclaimer: Opinions expressed by the authors and advertisers are notnecessarily those of the American Academy of Neurology, its affiliates, or ofthe Publisher. The American Academy of Neurology, its affiliates, and thePublisher disclaim any liability to any party for the accuracy, completeness,efficacy, or availability of the material contained in this publication (includingdrug dosages) or for any damages arising out of the use or non-use of any of thematerial contained in this publication.

Advertising Sales Representatives: Wolters Kluwer, 333 Seventh Avenue, NewYork, NY 10001. Contacts: Eileen Henry, tel: 732-778-2261, fax: 973-215-2485,[email protected] and in Europe: Craig Silver, tel: +44 7855062 550 or e-mail: [email protected].

Careers & Events: Monique McLaughlin, Wolters Kluwer, Two CommerceSquare, 2001 Market Street, Philadelphia, PA 19103, tel: (215)-521- 8468, fax:215-521-8801; [email protected].

Reprints: Meredith Edelman, Commercial Reprint Sales, Wolters Kluwer,Two Commerce Square, 2001 Market Street, Philadelphia, PA 19103, tel:215-356-2721; [email protected]; [email protected].

Special projects: US & Canada: Alan Moore, Wolters Kluwer, Two CommerceSquare, 2001 Market Street, Philadelphia, PA 19103, tel: 215-521-8638,[email protected]. International: Andrew Wible, Senior Manager,Rights, Licensing, and Partnerships, Wolters Kluwer, [email protected].

Copyright © 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

Page 5: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

A peer-reviewed clinical and translational neurology open access journal Neurology.org/NN

Neurology® Neuroimmunology & Neuroinflammation

EditorJosep O. Dalmau, MD, PhD

Deputy EditorScott S. Zamvil, MD, PhD, FAAN

Associate EditorsMichael Heneka, MDDennis Kolson, MD, PhDFriedmann Paul, MD

Editorial BoardRohit Aggarwal, MD, MSSergio E. Baranzini, PhDSusanne Benseler, MDJeffrey L. Bennett, MD, PhD, FAANSeverine Boillee, PhDMatthijs C. Brouwer, MD, PhDWolfgang Bruck, MDDavid B. Clifford, MD, FAANRussell C. Dale, MBChB, MRCP, PhDOscar H. Del Brutto, MD, FAANThomas G. Forsthuber, MDKazuo Fujihara, MDJeffrey M. Gelfand, MD, MAS, FAANJoseph El Khoury, MDMarkus P. Kummer, PhD

Jun Li, MD, PhDLin Mei, MD, PhDAndrew McKeon, MDChristopher Power, MDRichard M. Ransohoff, MDChristine Stadelmann, MDIsrael Steiner, MDLawrence Steinman, MDJesper Tegner, PhD, MScSilvia N. Tenembaum, MDMaarten J. Titulaer, MD, PhDAri Waisman, PhDHugh J. Willison, MBBS, PhDGregory F. Wu, MD, PhD

Neurology® Journals

Editor-in-ChiefRobert A. Gross, MD, PhD, FAAN

Deputy EditorBradford B. Worrall, MD, MSc, FAAN

Section Editors

BiostatisticsRichard J. Kryscio, PhDSue Leurgans, PhDV. Shane Pankratz, PhD

Classification of Evidence EvaluationsGary S. Gronseth, MD, FAAN

Equity, Diversity, and Inclusion (EDI)Roy H. Hamilton, MD, MS, FAANHolly E. Hinson, MD, MCR, FAAN

PodcastsStacey L. Clardy, MD, PhDJeffrey B. Ratliff, MD, Deputy Podcast Editor

OmbudsmanDavid S. Knopman, MD, FAAN

Scientific Integrity AdvisorRobert B. Daroff, MD, FAAN

Classification of Evidence Review TeamMelissa J. Armstrong, MDRichard L. Barbano, MD, PhD, FAANRichard M. Dubinsky, MD, MPH, FAANJeffrey J. Fletcher, MD, MScGary M. Franklin, MD, MPH, FAANDavid S. Gloss II, MD, MPH&TMJohn J. Halperin, MD, FAANJason Lazarou, MSc, MDSteven R. Messe, MD, FAANPushpa Narayanaswami, MBBS, DM, FAANAlex Rae-Grant, MD

Vision Neurology® Neuroimmunology &Neuroinflammation will be the premierpeer-reviewed journal in the fields ofneuroimmunology and neuroinflammation.

Mission Neurology: Neuroimmunology &Neuroinflammation will provide neurologistsand clinical research scientists withoutstanding peer-reviewed articles, editorials,and reviews to enhance patient care andfacilitate basic and translational research.

EditorialInquiries

Tel: 612-928-6400Toll-free: 800-957-3182 (US)Fax: [email protected]

StayConnected

facebook.com/NeurologyNN

twitter.com/greenjournal

youtube.com/user/NeurologyJournal

Copyright © 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

Page 6: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Editor’s Corner Volume 6, Number 4, July 2019Neurology.org/NN

Josep Dalmau, MD, PhD, Editor, Neurology® Neuroimmunology & Neuroinflammation

Time is tissue in optic neuritisNeurol Neuroimmunol Neuroinflamm July 2019 vol. 6 no. 4 e578. doi:10.1212/NXI.0000000000000578

Once again, this issue ofNeurology® Neuroimmunology & Neuroinflammation (N2) highlights thediversity of the field with studies on a range of clinical disorders and a variety of researchapproaches. Optic neuritis in patients with neuromyelitis optica spectrum disorders associatedwith aquaporin 4 antibodies (AQP4) or myelin oligodendrocyte glycoprotein (MOG) anti-bodies is considered steroid responsive, although patients can accumulate disability with re-current attacks. For these patients, Stiebel-Kalish et al.1 investigated whether the time fromsymptom onset to steroid treatment affected outcome. This was a retrospective study ofa consecutive cohort of patients after their first AQP4 or MOG antibody–related attack of opticneuritis. The results showed that patients from both cohorts who received treatment within 4days of symptom onset had improved outcomes compared with patients treated later and thateven a 7-day delay negatively affected outcome. The authors suggest that antibody-mediatedoptic neuritis should potentially be viewed with the same sense of urgency as heart attack orstroke and reminds the reader that for these patients “time is tissue.”

Evidence of an antecedent infection with Campylobacter jejuni, cytomegalovirus, orMycoplasmapneumoniae is frequently observed in patients developing Guillain-Barre syndrome (GBS),Fisher syndrome, or Bickerstaff brainstem encephalitis (BBE). In these patients, the neurologicdisorders share a similar pathogenic mechanism that includes elevated antiglycolipid antibodies.These syndromes rarely occur after infection with influenza virus, although influenza is verycommon across all age groups. In their study, Yamana et al.2 collected clinical information andacute-phase serum samples from consecutive patients with GBS, Fisher syndrome, or BBEoccurring after influenza infection and compared their findings with a cohort of patients withthese syndromes after C. jejuni infection. They found that the postinfluenza cohort had a sig-nificantly higher rate of Fisher syndrome and that postinfluenza GBS patients had a higherincidence of cranial nerve involvement, sensory disturbances, and ataxia. Antiglycolipid anti-bodies were detected more frequently in the postcampylobacter group, most commonly anti-GM1, whereas in the postinfluenza patients, anti-GQ1b and anti-GT1a were most frequent. Theauthors point out that GQ1b is densely localized in the paranodal regions of cranial nerves III, IV,and VI, suggesting a mechanistic relationship with the increased occurrence of Fisher syndromein the postinfluenza patients. Last, nerve conduction studies revealed acute inflammatory de-myelinating polyneuropathy in 60% of the postinfluenza group but only 25% of the post-campylobacter GBS group. It will be interesting to see how future studies elucidate the reasonswhy different infectious result in these varying clinical and serologic findings.

Moving from antibody-mediated demyelinating diseases to postinfectious disorders, the study ofBrunetti et al.3 brings us IgLON5 disease, an interesting and relatively newly described antibody-mediated autoimmune-neurodegenerative encephalopathy. IgLON5 disease is characterized bya sleep disorder, gait disturbance, bulbar symptoms, dysautonomia, and sometimes cognitivedecline. Both chronic and subacute clinical courses have been described as well as sudden deathlikely due to dysautonomia and respiratory distress. In most reports, patients did not improvewith immunotherapy, which may have been due to long delays between disease onset and

Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

Dr. Dalmau holds patents for the use of Ma2, NMDAR, GABABR, GABAAR, DPPX, and IgLON5 as autoantibody tests and receives royalties from the use of these tests.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

MORE ONLINE

Editor SummaryNPub.org/N2/edsum

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 7: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

diagnosis or treatment. Iorio et al. in a well-documented casereport describe a patient with a typical IgLON5 syndromewhose disease onset appears to have been 3–5 years beforediagnosis. After 1 year of combination immunotherapy(intravenous immunoglobulin prednisone, and azathioprine),the patient showed improvement in the sleep disorder andcognitive impairment. Although additional patients need to bestudied, this case suggests that delay in treatment initiationcannot fully explain the lack of response in some of thesepatients and gives hope to patients with long-standing disease.

Seeing the words “ketogenic diet” in the title of one of thestudies in this issue ofN2 supports the diversity of topics notedearlier. As is well known, the ketogenic diet was introduced asa treatment for epilepsy in the 1920s, and its popularity hasrisen and fallen over the years. In addition to the well-describedeffects of the ketogenic diet in shifting metabolism from usingglycolytic energy toward producing energy primarily from fattyacids, studies in the past few years have shown that the dietattenuates inflammatory biomarkers in the blood andCSF, andthese changes associated with benefit in models of experi-mental autoimmune encephalomyelitis. In light of theseresults, Brenton et al.4 conducted a pilot study of the safety andtolerability of the ketogenic diet in 20 patients with relapsingremitting multiple sclerosis in a 6-month, single-arm, un-controlled, open-label study. No subject experienced worsen-ing disease on the diet, and almost all adhered to the diet for theentire 6 months, which exceeds adherence rates reported inpatients with seizures. The patients showed reductions in body

mass index, total fat mass, and depression, along with im-provement in the expanded disability status scale scores. At 3months, there was a significant decrease in serum levels of theproinflammatory adipokine leptin. The results of this in-triguing pilot study should lead to prospective studies to definethe effect of the diet on disease course.

In addition to these studies, the July issue of N2 contains 2reviews on “Treatment strategies of childhood primary angiitisof the central nervous system” and “Immunoglobulin G4-related hypertrophic pachymeningitis” and other interestingarticles that I hope will catch your attention.

Study fundingNo targeted funding reported.

DisclosureDisclosure available: Neurology.org/NN.

References1. Stiebel-Kalish H, Hellman MA, Mimouni M, et al. Does time equal vision in the acute

treatment of a cohort of AQP4 and MOG optic neuritis? Neurol NeuroimmunolNeuroinflamm 2019;6:e572. doi: 10.1212/NXI.0000000000000572.

2. Yamana M, Kuwahara M, Yoshikawa K, Takeda K, Kusunoki S. Guillain-Barre syn-drome and related diseases after influenza virus infection. Neurol NeuroimmunolNeuroinflamm 2019;6:e575. doi: 10.1212/NXI.0000000000000575.

3. Brunetti V, Della Marca G, Spagni G, Iorio R. Immunotherapy improves sleep andcognitive impairment in anti-IgLON5 encephalopathy. Neurol NeuroimmunolNeuroinflamm 2019;6:e577. doi: 10.1212/NXI.0000000000000577.

4. Brenton NJ, Banwell BL, Bergqvist, et al. Pilot study of a ketogenic diet in relapsing-remitting multiple sclerosis. Neurol Neuroimmunol Neuroinflamm 2019;6:e565. doi:10.1212/NXI.0000000000000565.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 8: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

EDITORIAL OPEN ACCESS

Contribution of polymorphonuclear neutrophilsin the blood periphery to ischemic brain injuryDirk M. Hermann, MD, and Matthias Gunzer, PhD

Neurol Neuroimmunol Neuroinflamm 2019;6:e570. doi:10.1212/NXI.0000000000000570

Correspondence

Dr. Hermann

[email protected]

or Dr. Gunzer

[email protected]

Three major lines of argumentation presently support a role of polymorphonuclear neutrophils(PMNs) in ischemic brain injury:

1. PMNs abundantly accumulate in ischemic brain tissue in response to stroke both in rodentsand human patients, where they massively release enzymes, such as myeloperoxidase andelastase, and reactive oxygen species (ROS), which are known contributors to ischemicinjury.1–3

2. In mouse models of ischemic stroke induced by intraluminal middle cerebral arteryocclusion (MCAO), the selective depletion of PMNs by anti-Ly6G antibody or preventionof PMN brain entry by CXCR2 or VLA-4 blockade significantly reduces ischemic injuryand neurologic deficits.4,5

3. In patients with ischemic stroke, high neutrophil counts or high neutrophil to lymphocyteratios in the blood on admission are associated with poor neurologic outcome even whenadjustments for age, sex, vascular risk factors, and stroke severity are made.6,7

The combined evidence of these studies has prompted the idea that blood-derived PMNs areattracted into the ischemic brain, where they aggravate brain injury.8

In this issue of Neurology: Neuroimmunology & Neuroinflammation, Weisenburger-Lile et al.9

further strengthened and expanded this view, providing a detailed and comprehensive analysisof peripheral blood PMN characteristics in a cohort of 41 patients with acute ischemic stroke,which were compared with 22 healthy control subjects of the same age who were close relativesof patients. Blood samples were collected within 6 hours after stroke onset and, in a subgroup ofpatients, after 2 and 7 days. Flow cytometry studies revealed hyperactivation of circulatingPMNs in the acute stroke phase, that is, within 6 hours after stroke, indicated by decreasedCD62L and increased CD11b expression on PMNs, increased ROS production by unstimu-lated and stimulated PMNs, and increased circulating elastase levels in peripheral blood. Thenumber of necrotic PMNs was increased from 2 to 7 days after stroke, whereas the concen-tration of neutrophil extracellular trap components in the serum was decreased. An increasedpercentage of senescent, that is, CXCR4bright/CD62Ldim PMN was noted at all 3 time pointsexamined. PMNs with the capacity to reversely transmigrate from inflamed tissues back into theblood, defined as CD54high/CXCR1low, were increased in patients with stroke. States ofhyperactivation, senescence, and reverse migration were particularly pronounced in PMNsfrom patients exhibiting a high NIH Stroke Scale score (>12) on admission. The authorshypothesize that changes in PMN homeostasis may instrumentally contribute to ischemic braininjury, e.g., by promoting systemic inflammation, promoting blood-brain barrier breakdown, orinducing immunomodulation. The authors suggest that rebalancing PMN subsets or pre-venting reverse PMN transmigration might alleviate stroke consequences.

From the Department of Neurology (D.M.H.), University Hospital Essen, University of Duisburg-Essen; and Institute of Experimental Immunology and Imaging (M.G.), UniversityHospital Essen, University of Duisburg-Essen, Germany.

Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

RELATED ARTICLE

Harmful neutrophil subsetsin patients with ischemicstroke: Association withdisease severity

Page e571

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 9: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

The clear strength of this study is a meticulously conductedcharacterization of PMN subsets that provides a first in-depthcharacterization of PMN changes in ischemic stroke. Adownside is the choice of control subjects, which, represent-ing healthy humans, do not exhibit a similar vascular riskprofile. Differences in vascular risk factors and associateddiseases (coronary heart disease and large artery atheroscle-rosis) may at least partly account for the observations made.Hence, confirmation studies in an independent patient cohortmatched for risk factors and comorbidities will be required.Also, functional readouts of PMN activity such as migrationshould be studied in more detail,10 as they are further indi-cators of disease states.

What can we learn from these studies? In the past, we haveperhaps focused too exclusively on the PMN-associatedaggravation of ischemic injury inside the brain. By modu-lating immune responses in peripheral blood, reverselytransmigrated PMNs might potentially deregulate the neu-rovascular unit or exert bystander effects on additional im-mune cells such as T cells in the blood, which on brain entrymay aggravate ischemic damage.11 In line with this hy-pothesis, we have previously shown after intraluminalMCAO in mice that the delivery of antibodies inhibiting theintegrin VLA-4, which mediates the entry of T cells andPMNs into the brain, did not show any additional protectiveeffects when PMNs had been depleted.4 The depletion ofPMNs alone provided maximum protection. However, anti-VLA-4 acted synergistically with a selective T cell–depletingantibody to provide maximum protection.4 Hence, in thismodel, PMNs appeared to be a master switch that was re-sponsible for the injury-promoting capacity of both PMNsand T cells. The combined evidence of the now presentedstudy and our earlier study suggests that PMNs might act asan amplifier of detrimental immune responses that

compromise stroke outcome. Further studies on this issueare warranted.

Study fundingSupported by Deutsche Forschungsgemeinschaft (HE3173/11-1, GU769/10-1).

DisclosureD. Hermann has served on the scientific advisory board ofServier; has served on the editorial boards of Frontiers of CellularNeurosciences and Basic Research in Cardiology; and has receivedresearch support from the German Research Foundation. M.Gunzer has received speaker honoraria from Hexal and Mun-dipharma. Go to Neurology.org/NN for full disclosures.

References1. Gelderblom M, Leypoldt F, Steinbach K, et al. Temporal and spatial dynamics of

cerebral immune cell accumulation in stroke. Stroke 2009;40:1849–1857.2. Liesz A, Zhou W, Mracsko E, et al. Inhibition of lymphocyte trafficking shields

the brain against deleterious neuroinflammation after stroke. Brain 2011;134:704–720.

3. Perez-de-Puig I, Miro-Mur F, Ferrer-Ferrer M, et al.. Neutrophil recruitment tothe brain in mouse and human ischemic stroke. Acta Neuropathol 2015;129:239–257.

4. Neumann J, Riek-Burchardt M, Herz J, et al. Very-late-antigen-4 (VLA-4)-mediated braininvasion by neutrophils leads to interactions with microglia, increased ischemic injury andimpaired behavior in experimental stroke. Acta Neuropathol 2015;129:259–277.

5. Herz J, Sabellek P, Lane TE, Gunzer M, Hermann DM, Doeppner TR. Role ofneutrophils in exacerbation of brain injury after focal cerebral ischemia in hyper-lipidemic mice. Stroke 2015;46:2916–2925.

6. Maestrini I, Strbian D, Gautier S, et al. Higher neutrophil counts before thrombolysisfor cerebral ischemia predict worse outcomes. Neurology 2015;85:1408–1416.

7. Xue J, HuangW, Chen X, et al. Neutrophil-to-lymphocyte ratio is a prognostic markerin acute ischemic stroke. J Stroke Cerebrovasc Dis 2017;26:650–657.

8. Hermann DM, Gunzer M. Polymorphonuclear neutrophils play a decisive role forbrain injury and neurological recovery poststroke. Stroke 2019;50:e40–e41.

9. Weisenburger-Lile D, Dong Y, Yger M, et al. Harmful neutrophil subsets in patientswith ischemic stroke: Association with disease severity. Neurol NeuroimmunolNeuroinflamm 2019;6:e571. doi: 10.1212/NXI.0000000000000571.

10. Schuster M, Moeller M, Bornemann L, et al. Surveillance of myelodysplastic syn-drome via migration analyses of blood neutrophils: a potential prognostic tool.J Immunol 2018;201:3546–3557.

11. de Oliveira S, Rosowski EE, Huttenlocher A. Neutrophil migration in infection andwound repair: going forward in reverse. Nat Rev Immunol 2016;16:378–391.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 10: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

ARTICLE OPEN ACCESS

Harmful neutrophil subsets in patients withischemic strokeAssociation with disease severity

David Weisenburger-Lile, MD, Yuan Dong, PhD, Marion Yger, MD, Gaelle Weisenburger, MD,

Giulia Frasca Polara, MD, Thomas Chaigneau, MS, Riccardo Zapata Ochoa, MS, Beatrice Marro, MD,

Bertrand Lapergue, MD, PhD, Sonia Alamowitch, MD, and Carole Elbim, MD, PhD

Neurol Neuroimmunol Neuroinflamm 2019;6:e571. doi:10.1212/NXI.0000000000000571

Correspondence

Dr. Carole Elbim

[email protected]

AbstractObjectiveTo better understand the functional state of circulating neutrophils in patients with ischemicstroke (IS) for planning future clinical trials.

MethodsWe analyzed by flow cytometry activation state of circulating neutrophils and the distribution ofneutrophil peripheral subsets in 41 patients with acute IS less than 6 hours before admission andcompared them with 22 age-matched healthy controls.

ResultsOur results demonstrated continuous basal hyperactivation of circulating neutrophils duringacute IS, characterized by lower L-selectin expression and higher CD11b expression at the cellsurface, increased ROS production by neutrophils, and greater circulating levels of neutrophilelastase. Neutrophil hyperactivation was associated with deregulation of the equilibrium be-tween apoptotic and necrotic. Patients also had higher percentages than controls of theoveractive senescent (CXCR4bright/CD62Ldim) neutrophil subset and increased percentage ofneutrophils with a reverse transendothelial migration (CD54highCXCR1low) phenotype. Im-portantly, neutrophil alterations were associated with the clinical severity of the stroke, eval-uated by its NIH Stroke Scale score.

ConclusionAltogether, our results indicate that during acute IS, the inflammatory properties of circulatingneutrophils rise, associated with the expansion of harmful neutrophil subsets. These changes inneutrophil homeostasis, associated with disease severity, may play an instrumental role bycontributing to systemic inflammation and to the blood-brain barrier breakdown. Our findingshighlight new potential therapeutic approaches of stroke by rebalancing the ratio of senescentto immunosuppressive neutrophils or decreasing reverse neutrophil transmigration or both.

RELATED ARTICLE

EditorialContribution ofpolymorphonuclearneutrophils in the bloodperiphery to ischemic braininjury

Page e570

From the Sorbonne Universites (D.W.-L., Y.D., T.C., R.Z.O., S.A., C.E.), UPMC Univ Paris 06, UMRS 938, CdR Saint-Antoine, Hopital Saint-Antoine; INSERM (D.W.-L., Y.D., T.C., R.Z.O., S.A.,C.E.), UMRS 938, CdR Saint- Antoine, Team “Immune System, Neuroinflammation and Neurodegenerative Diseases,” Hopital St-Antoine; Service de Neurologie et d’UrgencesNeurovasculaires (D.W.-L., M.Y., S.A.), Assistance Publique-Hopitaux de Paris, Hopital Saint-Antoine; Division of Pneumology (G.W.), Foch Hospital, F-92150, Suresnes; Division ofNeurology (G.F.P.), Stroke Center, Assistance Publique-Hopitaux de Paris, Pitie-Salpetriere Hospital; Division of Radiology (B.M.), Assistance Publique-Hopitaux de Paris, Hopital Saint-Antoine; and Division of Neurology (B.L.), Stroke Center, Foch Hospital, F-92150, Suresnes.

Go to Neurology.org/NN for full disclosures. Funding information are provided at the end of the article.

The Article Processing Charge was funded by the authors.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 11: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Cerebral ischemia elicits a strong inflammatory response in-volving the transmigration of several subsets of leukocytes,including neutrophils.1 Murine models of ischemic stroke(IS) reveal that neutrophils are the first cells recruited into thebrain, minutes after stroke, accompanied by local productionof cytokines or chemokines and facilitated by a very earlyischemia-induced breakdown in the blood-brain barrier(BBB).2,3 Massive infiltration of neutrophils into ischemicbrain tissue has also been documented in human strokepatients.4 Although animal stroke models suggest that de-pleting circulating neutrophils before or at the onset of strokereduces infarct size, human drug trials aimed at preventingpotentially harmful subacute neutrophil adhesion and in-filtration into the brain parenchyma poststroke have not im-proved patient recovery.5,6 The failure of these trials might bepartially explained by the undiscriminating nature of theseapproaches: the anti-inflammatory potential of some circulatingneutrophil subsets means that globally blocking neutrophilaccess to the brain might not be a viable therapeutic approach.A better understanding of the functional state of circulatingneutrophils in patients with IS is therefore crucial for planningfuture clinical trials targeting postischemic inflammation.

Neutrophils, the most abundant circulating leukocytes, playa key role in the immune defense against bacterial and fungalpathogens. They are rapidly recruited from blood to sitesof infection, where they kill microorganisms via a number ofdifferent mechanisms including phagocytosis, production ofreactive oxygen species (ROS), release of granular proteins,neutrophil extracellular traps (NETs), a process calledNETosis. However, in case of inappropriate and/or excessiveactivation, neutrophils play also a key role in the induction orpromotion of inflammation-induced tissue injury.7 Recentstudies point to the existence of heterogenous populations ofmature and immature neutrophils. In fact, diverse subsets ofneutrophils differ markedly in their proinflammatory activity.Senescent neutrophils, an overly active subset, express ele-vated levels of CXCR4,8 whereas the immunosuppressiveCD16bright/CD62Ldim subset exhibits reduced proin-flammatory properties.9 Recent studies also demonstrate thatneutrophils can migrate back into the circulation, in a processreferred to as reverse transendothelial migration (rTEM).10

Contradictory results on the phenotype and functions of circu-lating neutrophils in patients with IS have been reported. Someauthors report fewer oxidative bursts and NET production,11

whereas others suggest increased degranulation and ROS pro-duction.12 Some analyzed neutrophil functions 24–48 hoursafter IS, although it is now clear that inflammatory pathways areinvolved in the first hours.1 None has investigated the distribu-tion of the senescent, immunosuppressive, and rTEMneutrophilsubsets or their consequences on the course of IS.

Our study, performed in whole blood to avoid artifacts due toisolation procedures,13 aimed to characterize fully the phe-notypes and functions of human peripheral neutrophils atdifferent stages of acute IS. We analyzed neutrophil activationstate and the distribution of 3 neutrophil subsets, the sen-escent subset, the immunosuppressive subset, and the rTEMsubset in the aim of addressing the right targets for planningfuture clinical trials. We investigated potential correlationsbetween neutrophil parameters and clinical markers of diseaseprogression.

MethodsStudy designThis study enrolled 41 patients with acute IS hospitalized inthe stroke unit of Saint-Antoine Hospital, Pitie-SalpetriereHospital (both Paris, France), and Foch Hospital (Suresnes,France) during the acute phase of brain ischemia. Patientswere included if they had had a stroke less than 6 hours beforeadmission, regardless of the NIH Stroke Scale (NIHSS) score,were aged ≥18 years, and showed no signs of infection at theadmission. All patients were examined by neurologists, ad-mitted to the stroke unit, and treated according to acceptedstroke guidelines.14 We prospectively collected the followingdata for each, with a structured questionnaire: vascular riskfactors (age, sex, presence of atrial fibrillation, hypertension,diabetes, hypercholesterolemia, previous stroke history, andtreatment); stroke severity assessed by the NIHSS at admis-sion (day 0), day 2, and day 7; and time from symptom onsetto brain imaging and sampling. The modified Ranking scale(mRs) was used for functional disability/dependence evalu-ations at 3 months. Standard stroke evaluation included brainimaging (MRI performed for 37 patients andCT performed for4 patients when MRI was contraindicated), blood sampling(especially neutrophil and lymphocyte counts), electrocardi-ography for at least 48 hours, transthoracic echocardiography,and vessel imaging by ultrasound, MR angiography, or angio-CT. We prospectively collected stroke information includinglocation of ischemia in the brain, presence and location of

GlossaryAAD = amino-actinomycin D; ANOVA = analysis of variance; APC = allophycocyanine; BBB = blood-brain barrier; DAMP =danger-associated molecular pattern; fMLP = N-formylmethionyl-leucyl-phenylalanine; HC = healthy control; HE =hydroethidine; HMGB = high-mobility group box; IS = ischemic stroke; LPS = lipopolysaccharide; MMP = matrixmetalloproteinase;MPO =myeloperoxidase;NET = neutrophil extracellular trap;NIHSS =NIH Stroke Scale;NLR =Nod-likereceptor; PBS = phosphate buffered saline; PMN = polymorphonuclear neutrophil; ROS = reactive oxygen species; rTEM =reverse transendothelial migration; sJAM-C = soluble JAM-C; TNF = tumor necrosis factor; TLR = Toll-like receptor.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 12: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

occlusion, and etiology according to the Trial of Org 10172 inAcute Stroke Treatment classification. IS volumes were mea-sured on diffusion sequences acquired on a 3-Tesla MRI in-strument at admission (and not measured for patients withcontraindications to brain MRI). A semiautomatic segmenta-tion procedure by Olea Sphere postprocessing software wasused for analysis.

The 22 age-matched recruited healthy controls (HCs), pre-viously presented,16 with blood biochemical and hematologicvalues within the normal range were recruited from the Unit ofNeurology of Memory and Language, Centre HospitalierSainte Anne (Paris, France). HCs were close members of thefamily of patients followed in the memory clinic. As previouslydescribed,16 HCs had no memory complaint or cognitive def-icit. In addition, we did not include HCs with a history of IS.

Exclusion criteria in the HC and patient groups included acuteor chronic inflammatory conditions and treatment with cor-ticosteroid or nonsteroidal anti-inflammatory drugs, both ofwhich could modulate neutrophil functions.

Whole blood was sampled, kept on ice, and transported im-mediately to the laboratory for neutrophil analysis. Impor-tantly, all patients were sampled at inclusion (day 0, D0),within the first 6 hours after IS symptom onset; 14 also hadsamples taken on D2 and D7.

Standard protocol approvals, registrations,and patient consentsThe Ethics Committee of Saint-Antoine Hospital approvedthe study, and all patients or their proxies provided writteninformed consent before participating began. The HCs wererecruited as part of their participation in the prospectivelongitudinal IMABio3 study (PHRC-0054-N 2010), whichwas approved by the Institutional Review Committee of Pitie-Salpetriere Hospital.

Determination of surface molecule expressionon resting neutrophilsHeparinized whole-blood samples (500 μL) were kept on iceand stained with allophycocyanine (APC)-anti-humanCD62L (BD Biosciences, San Jose, CA) and PE-anti-humanCD11b (clone ICRF44, BD Biosciences) antibodies, as pre-viously reported.15 The senescent CXCR4bright/CD62Ldim

neutrophil subset and the immunosuppressive CD16bright/CD62Ldim neutrophil subset were investigated as previouslydescribed,.16 To investigate the rTEM neutrophil subset,samples kept on ice were incubated for 45 minutes withFITC-anti-human CD81 (BD Biosciences) and PE-anti-CD54 (BD Biosciences) antibodies. The red blood cells werethen lysed with BD FACS lysing solution and resuspendedwith Cell Fix (BD Biosciences).16

Measurement of neutrophil oxidative burstSuperoxide anion (O2

−) production by neutrophils wasmeasured with a flow cytometry–based assay derived from the

hydroethidine (HE) oxidation technique, as previously de-scribed.16 After loading with HE for 15 minutes, whole-bloodsamples were incubated for 45 minutes at 37°C with phos-phate buffered saline (PBS) or either of the following Toll-likereceptor (TLR) agonists: Pam3CSK4 or lipopolysaccharide(LPS), or tumor necrosis factor (TNF)-α, as previouslyreported.22 Samples were then treated with PBS or 10−6 MN-formylmethionyl-leucyl-phenylalanine (fMLP) for 5minutes.22 Erythrocytes were lysed before analysis by flowcytometry.

Measurement of neutrophil apoptosisand necrosisSpontaneous neutrophil cell death in whole blood wasquantified with annexin V and the impermeant nuclear dye7-amino-actinomycin D (7-AAD). Whole-blood samples(500 μL) were incubated with APC-anti-CD15 antibody(clone HI98, BD Biosciences), FITC-annexin V, and 7-AAD(BD Biosciences), as previously described, and analyzed byflow cytometry.16

Flow cytometry analysisCells were analyzed with a Gallios flow cytometer and the datawith Kaluza software (Beckman Coulter). To determineneutrophil expression of surface molecules and ROS pro-duction, forward and side scatter were used to identify thegranulocyte population and to gate out other cells and debris.The purity of the gated cells was assessed using FITC- or PE-conjugated CD3, CD45, CD14, and CD15 antibodies. Tenthousand events were analyzed per sample, and the fluores-cence pulses were amplified by 4-decade logarithmic ampli-fiers. In all cases, unstained cells were run, and thephotomultiplier settings were adjusted so that the unstainedcell population appeared in the lower left-hand corner of thefluorescence display. In the multicolor analysis, single-cellcontrols were used to optimize signal compensation. All theresults were obtained with the use of a constant photo-multiplier gain value. To measure cell death in whole blood byflow cytometry, neutrophils were identified as CD15high cells,and 2 × 105 events were analyzed per sample.

Measurement of intravascular NETosisNETs are networks of DNA, histone, and proteolytic enzymesproduced by activated neutrophils, including myeloperox-idase (MPO), citrullinated histone 3, and elastase. SerumNET levels were quantified by detecting MPO-DNA com-plexes in serum samples. To prepare the DNA standard,NETs from a healthy subject were produced and isolated, andDNA was quantified in NET samples with PicoGreen. NET-associated MPO-DNA complexes were then quantified, alsoas previously described.17

Measurement of soluble pro- andanti-inflammatory mediatorsWhole-blood samples were centrifuged for 15 minutes at1,000gwithin 30minutes of collection. Soluble chemokine (IL-8) and cytokines (IL-1β, IL-6, TNF-α, G-CSF, IL-18, IL-22,

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 13: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

IL-10, and transforming growth factor-β) were detected withLuminex assays (Luminex Performance Assays, R&D Sys-tems).High-mobility group box-1 (HMGB1) and elastase weredetected from serum by ELISA.

StatisticsStatistical analyses were performed with Rstudio 1.0.143software. All tests were 2 tailed, with a significance level of α =0.05. When a parametric test was used, normality of distri-bution was tested with the Shapiro-Wilk test. Differencesbetween groups were assessed with the χ2 test or analysis ofvariance (ANOVA), followed by the Tukey post hoc test, asappropriate. ANOVA, adjusted for age, was used to compareneutrophil markers between the IS groups and controls (withage as a covariate). Bonferroni correction was used for mul-tiple comparisons. Linear partial correlation analysis, withadjustment for covariates, identified correlations.

Data availability statementData not provided will be available upon request.

ResultsParticipants’ characteristicsThe table summarizes the clinical characteristics of the 41patients with IS and 22 HCs. Patients were included in 2steps: the results from the first set of 27 patients led us toinclude a second set of 14 patients for complementary anal-yses. The 2 patient cohorts did not differ significantly forbaseline characteristics, including risk factors. Patients with ISand HCs did not differ significantly for age or sex, but riskfactors were significantly more frequent in the IS than theHCs. The proportion of etiologies for cerebral infarctionobserved in patients appears similar to recent data.18 Weobserved 4 brainstem strokes (9.8%), 13 cortical strokes(31.7%), 11 subcortical strokes (26.8%), and 13 cortical andsubcortical strokes (31.7%). Seven patients had hemorrhagictransformation. Five patients died within 7 days of onset. Asexpected,19 patients with IS had higher polymorphonuclearneutrophil (PMN) counts than HCs and higher PMN/lymphocyte ratios.

Highly activated circulating neutrophils inpatients with acute strokeWe analyzed neutrophil activation in the 27 patients with ISincluded in the first set of patients and in the 22 HCs. Neu-trophil activation is associated with the modulation of surfacemolecules, i.e., decreased L-selectin (CD62L) and increasedβ2-integrins (i.e., CD11b/CD18), through either stimulus-induced shedding or translocation from intracellular granules.As figure 1, A and B show, resting neutrophils from patientson D0 expressed less CD62L and more CD11b than HCs.Because neutrophil degranulation can also result in rapid re-lease of neutrophil serine proteases from azurophilic granulesinto the extracellular medium and circulation, we measuredthe level of circulating neutrophil elastase. Consistently with

previous data,20 serum neutrophil elastase levels were signif-icantly higher in patients with IS than HCs (figure 1C).

Measuring ROS production by neutrophils, we observed itwas higher in unstimulated neutrophils from whole blood(PBS-incubated sample) from patients than fromHCs (figure1D). We then assessed whether the higher constitutive ROSproduction by patient neutrophils was associated with higherproduction capacity in response to various activating stimuli.We have previously reported that neutrophils in whole bloodproduce minimal ROS in response to a single stimulus. 16Wetherefore studied neutrophil oxidative bursts in response tothe bacterial peptide fMLP after priming with TNF-α orvarious TLR agonists. ROS production by nonprimed neu-trophils (sample preincubated with PBS and stimulated withfMLP) was significantly higher in patients than in HCs (figure1E), and under the different priming conditions, the signifi-cantly higher ROS production in patients with IS (figure 1E)demonstrates their neutrophil hyperreactivity.

Among the patients included in our study, 14 were alsosampled on D2 and D7. Except for CD62L expression, whichreturned to normal levels on those days, no markers of neu-trophil activation were modulated from D0 during the courseof acute IS (figure 1, F–J). No significant differences amongany of the neutrophil activation markers were observed be-tween patients with or without subsequent stroke-associatedinfections or patients with or without IS risk factors (suchas hypertension, diabetes, and hyperlipidemia) (data notshown).

These results reflect basal hyperactivation of patients’ circu-lating neutrophils during acute IS, associated with their hy-perreactivity to various stimuli by ROS production.

Deregulation of neutrophil death in patientswith acute strokeNeutrophils constitutively undergo apoptosis, and this process iscritical for the resolution of inflammation. Neutrophil death bynecrosis, inversely, induces the release of cytotoxic mediators thatcan damage adjacent healthy tissue. Measuring the percentage ofneutrophils undergoing spontaneous apoptosis and necrosisshowed that the percentage of apoptotic neutrophils at D0 wassignificantly lower in patients with IS and tended to return tonormal values during the course of IS (figure 2, A and B). Thepercentage of necrotic neutrophils, however, was rising atD0, andthe increase became significant fromD2 (figure 2, C andD). Thisderegulation of the balance between apoptosis and necrosis,combined with basal neutrophil hyperactivation, may play a rolein establishing inflammatory processes in these patients.

Neutrophils can undergo NETosis, an alternative form ofprogrammed cell death associated with the release of decon-densed chromatin lined with granular components that cre-ates fibrous structures serving as NETs. These NETs canentangle pathogens but also contribute to both venous andarterial thrombosis.21 Patients with IS had lower levels of

4 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 14: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

circulating NETosis-derived products in their blood samplesat inclusion than HCs (figure 2E). This difference persisted atD2 and D7 (figure 2F).

Patients with or without either infection or IS risk factors didnot differ significantly for neutrophil apoptosis, necrosis, orintravascular NETosis (data not shown).

Expansion of the overactive senescentneutrophil subset in patientswith acute strokeThe phenotypes and functions of neutrophils change over timefrom their release into the blood (as fresh neutrophils) until theirdisappearance from the circulation (as aged neutrophils). Aged orsenescent CXCR4high neutrophils express lower levels of CD62Land higher levels ofβ2-integrins and produceROS at a higher rate

Table Characteristics of patients with ischemic stroke and HCs participating in the study

HC (n = 22)Patient set 1(n = 27)

Patient set 2(n = 14)

pValue, set 1vs HC

pValue, set 2vs HC

p Value, set 1vs set 2

Sex (male), (n) [%] 6 [35.3] 14 [51.8] 8 [57.14] 0.36 0.29 0.99

Age, yr, (mean) [min–max] 68.1 [22–89] 72.6 [18–96] 71.1 [40–92] 0.20 0.54 0.59

History of: (n) [%]

Diabetes 1 [4.54] 6 [22.2] 1 [5.89] 0.11 0.99 0.69

High blood pressure 2 [11.8] 21 [77.8] 10 [71.4] 0.001 0.001 0.71

Dyslipidemia 4 [23.5] 14 [51.9] 2 [14.3] 0.02 0.99 0.69

Myocardial infarction 0 9 [33.3] 4 [28.6] 0.002 0.02 0.99

Atrial fibrillation 0 8 [29.6] 4 [28.6] 0.006 0.02 0.99

Stroke 0 6 [22.2] 1 [7.14] 0.026 0.39 0.39

NIHSS D0 (mean) [min–max] — 9.15 [0–21] 8.64 [1–23] — — 0.64

Etiologies/TOAST (n) [%]

Thromboembolism — 6 [22.2] 2 [14.3] — — 0.69

Cardioembolism — 11 [40.7] 4 [28.6] — — 0.51

Lacunar stroke — 1 [3.7] 2 [14.3] — — 0.54

Other specific cause — 3 [7.4] 3 [21.4] — — 0.64

Undetermined cause — 6 [22.2] 3 [21.4] — — 0.99

Occlusion at inclusion (n) [%] — 16 [59.3] 6 [42.8] — — 0.51

Volume of cerebral infarct (mL) (mean)[min–max]

— 23.92[0.04–240.8]

16.17 [0–63.8] — — 0.99

Treatment at acute stage (n) [%] —

IV Thrombolysis — 17 [62.9] 5 [35.7] — — 0.115

Endovascular treatment — 9 [33.3] 4 [28.6] — — 0.99

Time from symptom onset to blood sample(min) (mean) [min–max]

— 129 [45–240] 186 [95–340] — — 0.09

Previous treatment by statins 2 [11.7] 9 [33.3] 3 [21.4] 0.158 0.636 0.494

Previous treatment by antithromboticagents

0 8 [29.6] 4 [28.6] 0.006 0.017 0.99

Neutrophil count (G/L) (mean) [min–max] 3.15[2.32–4.80]

6.1[2.37–14.40]

7.85[2.33–15.3]

0.007 0.002 0.22

Lymphocyte count (G/L) (mean) [min–max] 1.68[1.06–2.15]

1.48[0.53–3.15]

1,52[0.57–3.26]

0.43 0.55 0.91

Ratio PMN/lymphocyte (mean) [min–max] 2.09[1.37–4.36]

5.63[1.04–23.34]

7.70[1.04–26.84]

0.023 0.024 0.40

Abbreviations: HC = healthy control; NIHSS = NIH Stroke Scale; TOAST = Trial of Org 10172 in Acute Stroke Treatment.Sex, risk factors, and type of treatment were compared with the χ2 test. The Mann-Whitney test was used to compare quantitative variables.We considered all thrombolyses and all mechanical thrombectomies.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 5

Page 15: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

than the total circulating pool.8 Conversely, in contexts of acuteinflammation9 and healthy aging,22 an immunosuppressive subsetof CD16bright/CD62Ldim neutrophils show decreased adhesionproperties and less ROS production. Changes in the phenotypicand functional features of patients’ neutrophils might thus berelated to an imbalance between these 2 neutrophil subsets. Thesenescent CXCR4bright/CD62Ldim neutrophil subset expandedin patients with IS at D0 (figure 3, A and C), whereas the per-centage of neutrophils in the CD16bright/CD62Ldim

immunosuppressive subset did not differ between patientswith IS and HCs (figure 3, B and D). Accordingly, patientswith IS had a higher ratio of senescent to immunosuppressivesubsets than HCs (figure 3, E). Similar results were observedin patients with IS at D2 and D7 (figure 3, F–H). Moreover,PMN subsets did not differ significantly between patients withor without infection or IS risk factors (data not shown). Theseresults reflect the expansion of the senescent subset in patientswith IS may lead to greater PMN oxidative burst.

Figure 1 Hyperactivation of circulating neutrophils from patients with ischemic stroke

(A–E) Adhesion molecules expres-sion, polymorphonuclear neutrophil(PMN) elastase, and ROS productionwere studied in patients with ISat inclusion. Expression of CD62L(A) and CD11b (B) at the surfaceof resting PMNs was studied onwhole-blood samples kept on ice andincubeted with anti-CD62L and anti-CD11b monoclonal antibodies.Results are expressed as mean fluo-rescence intensity (MFI). (C) Circulat-ing levels of PMN elastase werequantified by ELISA; results areexpressed as ng/mL. (D and E) ROSproduction by unstimulated PMNswas studied after treatment of whole-blood samples for 50 minutes withPBS; results are expressed as MFI (D).ROS production by stimulated PMNswas measured after pretreatment ofwhole-blood samples for 45 minuteswith PBS, Pam3CSK4 (TLR1/2 agonist,1 μg/mL), or LPS (TLR4 agonist, 10 ng/mL), or TNF-α (TNF, 5 ng/mL) and in-cubation for 5 minutes with fMLP(10−6 M); results are expressed asMFI(E). (F–J) Follow-up of surface expres-sion of CD62L (F), surface expressionof CD11b (G), circulating levels ofPMN elastase (H), and ROS pro-duction by unstimulated (I) and stim-ulated PMNs (J) during the first weekafter IS symptoms. All samples camefrom age-matched HCs, (n = 22) andpatients with IS at day 0 (n = 27), day 2(n = 14), and day 7 (n = 14). Values aremean ± SEM. *Significantly differentfrom controls p < 0.05, **p < 0.01, ***p< 0.001, adjusted for age. D0 = day 0;D2 = day 2; D7 = day 7; fMLP = N-for-mylmethionyl-leucyl-phenylalanine;HC = healthy control; IS = ischemicstroke; LPS = lipopolysaccharide; ROS =reactive oxygen species; SEM = stan-dard error of mean; TLR = Toll-like re-ceptor; TNF = tumor necrosis factor.

6 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 16: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Increased levels of soluble JAM-C wereassociated with an increase in the rTEMneutrophil subset in patients withacute strokeJAM-C is an adhesion molecule expressed by endothelialcells that plays a role in tight junction formation, leukocyteadhesion, and transendothelial migration. Elastase-inducedproteolytic cleavage of endothelial JAM-C leading to sol-uble JAM-C (sJAM-C) was recently reported to be in-strumental in promoting neutrophil rTEM in vivo.23

Measuring the circulating levels of JAM-C in our first co-hort of 27 patients with IS showed a higher level at D0 thanin HCs (figure 4A); values returned to normal at D2 (figure4B). As expected,23 increased sJAM-C levels were

associated with an increased circulating level of neutrophilelastase (figure 4C).

We thus performed a complementary study in a second co-hort of 14 patients with IS to investigate the percentage ofcirculating rTEM neutrophils and its association with thesJAM-C level. The clinical characteristics of this second co-hort did not differ from that of the main cohort (table), nordid their sJAM-C level on D0 (421 ± 20 pg/mL and 463 ± 49pg/mL for the patient set 1 and set 2, respectively, p = 0.77),which was similarly significantly higher than in HCs. As pre-viously reported,23 a small population of neutrophils with anrTEM phenotype (CD54high, CXCR1low) was consistentlyfound in HCs, equivalent to 1% of all circulating neutrophils.In patients with IS, we could also identify a distinct populationof cells with an rTEM phenotype; its percentage was signifi-cantly higher in patients than in HCs (figure 4, D and E),and the increase of this percentage was associated with in-creased levels of sJAM-C (figure 4F). Taken together, theseresults demonstrate the expansion of the rTEM subset in ISmigration.

Neutrophil abnormalities are associated withthe severity of acute ISWe then investigated whether the neutrophil abnormalitiesobserved in patients with IS might be associated with strokeseverity. Patients were dichotomized according to the NIHSSscore at inclusion into minor-to-moderate (NHISS score≤12) and severe stroke (NHISS score >12) categories. Basalneutrophil activation at baseline was greater in severe than inminor-to-moderate strokes, as shown by patients’ lowerCD62L expression (figure 5A), higher circulating neutrophilelastase levels (figure 5C), and higher neutrophil productionof ROS (figure 5D), although CD11b expression did notdiffer significantly between the groups (figure 5B). Thesephenotypic and functional properties of neutrophils from thesevere stroke group were associated with a higher percentageof neutrophils in the senescent hyperactive subset (figure 5E),but not in the immunosuppressive subset (figure 5F). Finally,the sJAM-C level was higher in the severe stroke group (figure5G). In contrast, the percentage of apoptotic (figure 5H) andnecrotic neutrophils (data not shown) and the intravascularNETosis level (data not shown) did not differ between thepatient groups. Higher NIHSS scores at D0 were associatedwith decreased expression of CD62L (figure 5I), increasedROS production by both unstimulated (figure 5J) and LPS-primed neutrophils (figure 5K), and an increase in the per-centage of the senescent hyperactive subset (figure 5L), thelevel of sJAM-C (figure 5M), and the percentage of rTEMneutrophils (figure 5N). Increased NIHSS score at D2 was alsoassociated with increased ROS production by LPS-primedneutrophils (data not shown). The correlations were adjustedfor age. We found no correlation between neutrophil parame-ters and the infarct volume at inclusion and the presence ofmacroangiopathy, the occurrence of hemorrhagic trans-formation, the use of antithrombotic treatment, or the use ofthrombolysis or endovascular treatment (data not shown).

Figure 2 Deregulation of neutrophil death in patients withischemic stroke

(A–D) Polymorphonuclear neutrophil (PMN) death was measured immedi-ately after sampling by staining with annexin V and 7-AAD. Results areexpressed as the percentages of apoptotic (A and B) and necrotic (C and D)PMNs. (A and C) Values of PMN death in patients with IS at inclusion. (B andD) Follow-up of PMN death during the first week after IS symptoms. (E and F)Circulating NETs were quantified by MPO-DNA complex ELISA; results areexpressed as ng/μL DNA. Values of circulating NETs in patients with IS atinclusion (E). Follow-up of intravascular NETosis during the first week after ISsymptoms (F). All samples came from age-matched healthy controls (HCs, n= 22) and patients with IS at day 0 (n = 27), day 2 (n = 14), and day 7 (n = 14).Values aremean ± SEM. *Significantly different from controls p < 0.05, **p <0.01, ***p < 0.001, adjusted for age. AAD = amino-actinomycin D; D0 = day 0;D2 = day 2; D7 = day 7; HC = healthy control; IS = ischemic stroke; MPO =myeloperoxidase; NET = neutrophil extracellular trap; SEM = standard errorof mean.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 7

Page 17: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Considered together, these results strongly suggest that theimpaired homeostasis of circulating neutrophils associatedwith neutrophil hyperactivation is correlated with the clinicalseverity of acute stroke disease progression.

Proinflammatory environment in patientswith acute strokeIschemic cerebral tissue rapidly induced the release ofcytokines, chemokines, and danger-associated molecularpatterns (DAMPs), including HMGB1.24 Because cyto-kines25 and HMGB126 strongly regulate neutrophil activity,we investigated patients’ peripheral proinflammatory envi-ronment. Levels of classic proinflammatory (IL-1β, IL-6,IL-8, and TNFα) and anti-inflammatory (IL-10 and trans-forming growth factor-β) cytokines were similar in the IS

patient and HC groups (figure 6A). Serum levels of G-CSFand IL-22, a bifunctional cytokine with both proin-flammatory and protective functions, were significantlylower in the IS than the HC group. Inversely, serum levels ofthe proinflammatory cytokine IL-18 were higher in thepatients with IS (figure 6A), as were circulating HMGB1levels (figure 6, B and C), which were associated with theclinical NIHSS score on D0 (figure 6D). Taken together,these results demonstrate higher circulating levels of theproinflammatory cytokine IL-18 and of the DAMPHMGB-1in patients with IS.

We then investigated correlations between neutrophilparameters and proinflammatory mediator levels. Neutro-phil markers were not correlated with cytokine levels. In

Figure 3 Impaired homeostasis of the circulating neutrophils in patients with ischemic stroke

(A–E) Analysis of senescent and immunosuppressive polymorphonuclear neutrophil (PMN) subsets in patientswith IS at inclusion.Whole-blood sampleswereincubated for 45minutes at 4°C with Pe-Cy7-anti-human CXCR4, PE-anti-human CD11b, and APC-antihuman CD62L (A and B) or with FITC-anti-human CD16,PE-anti-human CD11c, Pe-Cy7-antihuman CD11b, and APC-anti-human CD62L (C and D) antibodies. (A) Representative dot plots of the PMN phenotypeaccording to CXCR4 and CD62L expression in aHC (left) and a patient with IS (right). (B) Representative dot plots of the PMNphenotype according to CD16 andCD62L expression in an HC (left) and a patient with IS (right). (C) Percentages of the CXCR4bright/CD62Ldim senescent PMN subset. (D) Percentages of theCD16bright/CD62Ldim immunosuppressive PMN subset. (E) Ratio between the senescent and the immunosuppressive PMN subsets. (F–H) Follow-up ofsenescent PMN subset (F), immunosuppressive PMN subset (G), and ratio between the senescent and the immunosuppressive PMN subsets (H) during thefirst week after IS symptoms. All samples came fromage-matchedHCs, (n = 22) andpatientswith IS at day 0 (n = 27), day 2 (n = 14), andday 7 (n = 14). Values aremean± SEM. *Significantly different from controls p < 0.05, **p < 0.01, ***p < 0.001, adjusted for age. D0 =day 0; D2 = day 2; D7 = day 7; HC = healthy control; IS= ischemic stroke; SEM = standard error of mean.

8 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 18: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

contrast, the increased serum HMGB1 level was associatedwith decreased CD62L expression on resting neutrophils(figure 6E), increased CD11b expression (figure 6F), andincreased ROS production by primed neutrophils (figure6G). These findings suggest that HMGB1 might preactivateneutrophils in vivo.

DiscussionIn the present study, phenotypes and functions of peripheralneutrophils from patients less than 6 hours after IS have beenextensively investigated. Our results demonstrated continu-ous basal hyperactivation of circulating neutrophils duringacute IS, characterized by lower L-selectin expression andhigher CD11b expression at the cell surface, increased ROSproduction by neutrophils, and greater circulating levels ofneutrophil elastase. The release of this soluble cytotoxicmolecule might induce vascular damage that aggravates is-chemia or diffuses into the brain parenchyma and causes

tissue injury. Neutrophil hyperactivation was associated withderegulation of the equilibrium between apoptotic and ne-crotic PMNs, which may be a supplemental factor promotingacute ischemic inflammation. Furthermore, the lack of dif-ference in neutrophil changes according to the presence orabsence of risk factors for stroke, such as type 2 diabetes andhypertension, suggests that these abnormalities might be di-rectly linked to IS.

Activated neutrophils may contribute to vascular inflammationduring IS through neutrophil-induced BBB impairment. TheBBB forms a protective barrier around the brain, with the es-sential function of maintaining brain homeostasis. During is-chemia and subsequent reperfusion, its integrity is disrupted,which results in the leakage of blood-borne proteins into braintissue. Loss of BBB integrity correlates with poorer long-termoutcomes in patients with IS.27 ROS, including those producedextracellularly by neutrophils,28 have been reported to induceBBB breakdown.29 In particular, ROS activate matrix

Figure 4 Patients with ischemic stroke have higher levels of soluble JAM-C and of PMN reverse transmigration

(A and B) sJAM-C was quantified byELISA; results are expressed as pg/mL.(A) Values of sJAM-C in age-matchedHCs, (n = 22) and patients with IS at in-clusion (n = 27) (patients included in set1). (B) Follow-up of sJAM-C levels duringthe first week after IS symptoms in 10patients. (C) Correlation between thelevels of sJAM-C and of circulating PMNelastase. (D and E) Quantification ofrTEM PMNs in patients with IS includedin set 2. Whole-blood samples wereincubated for 45 minutes at 4°Cwith FITCanti-human CD181 and PE-anti-human CD54 antibodies. (D)Representative dot plots of the PMNphenotype according to CXCR4 andCD62L expression in an HC (left) anda patient with IS (right). (E) Percentagesof the rTEM PMN subset (CD54high,CXCR1low) in age-matched HCs (n = 17)and patients with IS at inclusion (n = 14).(F) Correlation between the percentageof rTEM PMN subset and the level ofsJAM-C in patients with IS included inthe set 2. *Significantly different fromcontrols p < 0.05, **p < 0.01, ***p <0.001, adjusted for age. D0 = day 0; D2 =day 2; D7 = day 7; HC = healthy control;IS = ischemic stroke; PMN = poly-morphonuclear neutrophil; rTEM = re-verse transendothelialmigration; sJAM-C=soluble JAM-C.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 9

Page 19: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

metalloproteinase (MMP) activation, which in turn plays a keyrole in BBB breakdown by disrupting the basement membraneof the neurovascular unit.30 Among the variousMMPs,MMP-9(gelatinase B) is thought to be central to BBB impairment.Neutrophils are the major source of MMP-9 after IS, and

neutrophil-derivedMMP-9 has been shown to be important inpostischemic BBB disruption, leukocyte infiltration, and braindamage.29 BBB disruption triggers inflammatory response andinfiltration of immune cells including neutrophils, which are inturn associated with increased neuronal cell death through the

Figure 5 Neutrophil hyperactivation is associated with disease severity

(A andB) Expression of CD62L (A) and CD11b (B) at the surface of resting polymorphonuclear neutrophils (PMNs). Results are expressed asMFI. (C) Circulatinglevels of neutrophil elastase were quantified by ELISA. (D) ROS production by unstimulated PMNs; results are expressed in MFI. (E) Percentages of theCXCR4bright/CD62Ldim senescent PMN subset. (F) Percentages of the CD16bright/CD62Ldim immunosuppressive PMN subset. (G) sJAM-C was quantified byELISA; results are expressed as pg/mL (H) PMN apoptosis was measured immediately after sampling by staining with annexin V and 7-AAD. Results areexpressed as percentages of apoptotic PMNs. All measurements came from patients withminor-to-moderate stroke (IS NHISS ≤ 12, n = 17) and patients withsevere stroke (IS NHISS > 12, n = 10) at the inclusion. Values aremean ± SEM. Statistical significance as determined by the nonparametricMann-Whitney test isindicated. *Significantly different p < 0.05, **p < 0.01, ***p < 0.001. Correlation between the NIHSS score at inclusion and CD62L expression at the PMNsurface (I), ROS production by unstimulated PMNs (J), ROS production by LPS-stimulated PMNs (K), percentage of the CXCR4bright/CD62Ldim senescent PMNsubset (L), sJAM-C level (M), and percentage of rTEM PMNs (N). AAD = amino-actinomycin D; D0 = day 0; IS = ischemic stroke; LPS = lipopolysaccharide; MFi =mean fluorescence intensity; NIHSS =NIH Stroke Scale; PMN = polymorphonuclear neutrophil; ROS = reactive oxygen species; SEM = standard error ofmean;sJAM-C = soluble JAM-C.

10 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 20: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

release of neurotoxic substances and secondary injuries. Assenescent neutrophils in the circulation have been reported torapidlymigrate to the inflammatory site,31 we speculate that theexpansion of the senescent neutrophil subset observed inpatients with IS might result in enhanced migration of thesehyperactive cells in the brain. Both the percentage of senescentneutrophils and the parameters of neutrophil hyperactivationwere correlated with disease severity measured by the NIHSSscore.We did not, however, find any correlationwith the infarctvolume at inclusion. In fact, a limitation of our study is related

to the fact that themeasured volumematches the ischemic coreand does not include the volume of potentially reversible ischemia(penumbra). It has been suggested that inflammatory processestake place in the penumbral area and contribute to potentialenlargement of the infarct.32 Furthermore, themeasured volumeson diffusion do not match with the final volume, especiallyin treated patients.33,34 A potential association between neutro-phil abnormalities and the progression of damage in the pen-umbra and the final ischemic volume have to be explored infurther studies. In addition, we had a relatively limited number of

Figure 6 Proinflammatory mediators in patients with ischemic stroke

(A) Circulating levels of pro- and anti-inflammatory cytokines weremeasured with Luminex assays in samples from controls (HCs, n = 22) and patients with ISat day 0 (n = 27). (B and C) Circulating levels of HMGB1 were measured in serum by ELISA. (B) Comparative analysis of circulating levels of HMGB1 betweenage-matched HCs (n = 22) and patients with IS (IS) at day 0 (n = 27). (C) Follow-up of circulating levels of HMGB1 in patients with IS at day 0, 2, and 7. Values aremean ± SEM. *Significantly different from controls p < 0.05, **p < 0.01, ***p < 0.001, adjusted for age. Correlation between circulating levels of HMGB1 at day0 and the NIHSS score at day 0 (D). (E–G) Correlation between circulating levels of HMGB1 at day 0 and PMN activation markers at day 0: PMN surfaceexpression of CD62L (E) and of CD11b (F) and ROS production by LPS-stimulated PMNs (G). D0 = day 0; D2 = day 2; D7 = day 7; HC = healthy control; HMGB =high-mobility group box; IS = ischemic stroke; LPS = lipopolysaccharide; NIHSS = NIH Stroke Scale; PMN = polymorphonuclear neutrophil; ROS = reactiveoxygen species; SEM = standard error of mean.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 11

Page 21: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

patients, and our observations must be corroborated in largercohorts.

Despite neutrophil hyperactivation, intravascular NETosisdecreased in patients with IS starting at D2 after IS. Thisdiscrepancy with previous data35 might be related to the dif-ferent techniques used to study NETosis. As chromatinconstitutes the backbone of NETs, and MPO is present in theNETs, we adopted complexes of nucleosomes and MPO asthe definition of NET remnants, as suggested,17 and quanti-fied serum NET levels by detecting the MPO-DNA com-plexes. The mixing of the nuclear components with thecontent of azurophilic granules is a critical step in NETosisand enables the formation of complexes between nucleo-somes andMPO. In contrast, in a previous study performed inpatients with IS,35 3 markers were evaluated: plasma cell-freeDNA, nucleosomes, and citrullination of histone 3, all onlyindirect markers of NETs. The decrease in intravascularNETosis observed in our study could be due to the decreasedcapacity of stimulated PMNs to produce NETs11 and to thepresence of NETs in thrombotic occlusions from patientswith IS.36

The occurrence of neutrophil reverse transmigration afterischemia/reperfusion injury has previously been reported inanimal models.23 Our findings describe an increased per-centage of neutrophils with an rTEM (CD54highCXCR1dim)phenotype in human IS. The recent report that rTEM neu-trophils show an activated morphology with unaltered anti-microbial activity suggests that they might be deleterious ifreinfiltrated into remote organs.37 Moreover, rTEM neu-trophils are relatively rigid, a physical characteristic that mightdelay their passage through the tissue microvasculature andprolong contact with the sinusoids. These cells could becomemechanically entrapped in the microvasculature of majororgans (most likely in the pulmonary circulation). We thuspostulate that a larger population of functionally primed,recirculating neutrophils in patients with IS, with a pre-dilection for mechanical or adhesive entrapment in the vas-culature, could contribute to distant organ damage andmultiorgan failure. As these rTEM neutrophils have beencharacterized by their prolonged survival and enhanced oxi-dative burst,38 their expansion in patients with IS might ex-plain, at least in part, the higher ROS production and thelower rate of neutrophil apoptosis in our cohort. Furthermore,because rTEM neutrophils are reported to inhibit T-cellproliferation,9 we can speculate that they might contribute tothe T lymphopenia observed during the stroke-inducedimmunodepression phase.

Consistent with data suggesting that sJAM-C may serve as anindirect marker of neutrophil rTEM,23,38 higher levels of thisadhesion molecule in patients with IS associated with thepercentage of rTEM neutrophils. On the other hand, sJAM-Calso acts as a potent proangiogenic molecule with chemotacticaction for endothelial cells, and it induces their formation ofvascular tubes in vitro.39 After ischemic injury, angiogenesis

increases the number of new collateral conduits, enhancingthe blood supply and improving ischemic brain function.40

Despite this potential beneficial action of sJAM-C, we clearlyobserved that the sJAM-C level was higher in the severe strokegroup.

As previously reported,41,42 we observed higher circulatinglevels of the proinflammatory cytokine IL-18 and of theDAMP HMGB-1 in patients with IS than in age-matchedHCs. IL-18 production requires activation of the inflamma-some, a group of multimeric protein complexes that sensestress signals, activate caspase-1, and thus cause its proteolyticcleavage. Notably, IS increases the expression and activationof the Nod-like receptor (NLR) pyrin domain containing 1and 3 (NLRP1 and NLRP3) inflammasome proteins andboth IL-1β and IL-18 in neurons.43 IL-1844 and HMGB126

enhance neutrophil functions, including priming neutrophilsfor oxidative burst and degranulation. In turn, activated neu-trophils produce IL-18. In addition, necrotic cells, includingnecrotic neutrophils, passively release HMGB1. Of interest,we found that the percentage of necrotic neutrophils washigher in patients with IS than HCs. These processes togethermight thus maintain a vicious cycle. We also observed thatincreased plasma levels of HMGB-1 were associated withmarkers of neutrophil activation.

The phenotypic and functional neutrophil changes that ac-company acute IS may also be related to the expansion of theoveractive aged neutrophil subpopulation. A recent studyshowed that neutrophil aging is regulated by the microbiota.8

Recent studies report gut dysbiosis in patients with IS.45,46

Commensal microbiota affects IS outcome in mice with is-chemic brain injury by impairing immune homeostasis in thesmall intestine, thus increasing IL-17+ γδ T cells and de-creasing regulatory T cells (Tregs).47 Of note, γδ T cells,a major lymphocyte population with innate immune features,can aggravate ischemic brain injury by secreting IL-17 andgenerating chemotactic signals for peripheral myeloid cellssuch as neutrophils.48 In contrast, Tregs are known to inhibitneutrophil functions49: in rodent IS models, adoptivelytransferred Tregs exert a protective effect by modulating pe-ripheral neutrophils and thus prevent proteolytic damage tothe BBB.50 Alterations in the microbiota could contributeboth to expansion of senescent neutrophils and changes inneutrophil activation in patients with IS. We can speculatethat modulation of the microbiome might affect diseaseprogression by limiting neutrophil aging and reduction pro-duction of neutrophil-derived cytotoxic molecules.

We demonstrated that homeostasis of circulating neutrophilsis impaired during acute IS. These findings need to be repli-cated in an independent cohort of patients in which a metic-ulous definition of neutrophil subsets will be performed. Theexpansion of aged neutrophils might play a role in poststrokesystemic inflammation and BBB disruption, and rTEM neu-trophils might further contribute to the propagation of sys-temic inflammation. Our findings highlight new potential

12 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 22: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

therapeutic approaches of stroke based on neutrophil mod-ulation, in particular, by rebalancing the ratio of senescent toimmunosuppressive neutrophils or decreasing reverse neu-trophil transmigration or both. Further studies are needed,however, to determine whether these changes might serve asa predictive indicator of IS.

Author contributionsC. Elbim conceived the study, designed the experiments, andsupervised the project. D. Weisenburger-Lile and Y. Dongconducted the experiments and analyzed the data. D.Weisenburger-Lile, M. Yger, G.F. Polara, and B. Laperguewere involved in patient recruitment and characterization andin the data analysis. G. Weisenburger-Lile, T. Chaigneau, andR. Zapata Ochoa participated in the data analysis. D.Weisenburger-Lile, S. Alamowitch, and C. Elbim wrote themanuscript. All authors read and approved the content of thefinal manuscript.

AcknowledgmentThe authors thank the nursing staff of the Stroke Unit ofSaint-Antoine Hospital Pitie Salpetriere Hospital and FochHospital for the patient management. The authors gratefullyacknowledge Annie Munier from the LUMIC flow cytometryfacility. DWL received a grant from the “Societe Francaise deNeurologie” for a part of this work.

Study fundingNo targeted funding reported.

DisclosureThe authors report no disclosures relevant to the manuscript.Disclosures available: Neurology.org/NN.

Publication historyReceived by Neurology: Neuroimmunology & NeuroinflammationDecember 19, 2018. Accepted in final form March 12, 2019.

References1. Ross AM, Hurn P, Perrin N, Wood L, Carlini W, Potempa K. Evidence of the

peripheral inflammatory response in patients with transient ischemic attack. J StrokeCerebrovasc Dis 2007;16:203–207.

2. Neumann J, Riek-Burchardt M, Herz J, et al. Very-late-antigen-4 (VLA-4)-mediatedbrain invasion by neutrophils leads to interactions with microglia, increased ischemicinjury and impaired behavior in experimental stroke. Acta Neuropathol 2015;129:259–277.

3. Shi Y, Zhang L, Pu H, et al. Rapid endothelial cytoskeletal reorganization enables earlyblood-brain barrier disruption and long-term ischaemic reperfusion brain injury. NatCommun 2016;7:10523.

4. Perez-de-Puig I, Miro-Mur F, Ferrer-Ferrer M, et al. Neutrophil recruitment to thebrain in mouse and human ischemic stroke. Acta Neuropathol 2015;129:239–257.

5. Enlimomab Acute Stroke Trial I. Use of anti-ICAM-1 therapy in ischemic stroke:results of the Enlimomab Acute Stroke Trial. Neurology 2001;57:1428–1434.

6. Krams M, Lees KR, Hacke W, et al. Acute stroke therapy by inhibition of neutrophils(ASTIN): an adaptive dose-response study of UK-279,276 in acute ischemic stroke.Stroke 2003;34:2543–2548.

7. Gunzer M. Traps and hyper inflammation—new ways that neutrophils promote orhinder survival. Br J Haematol 2014;164:189–199.

8. Zhang D, Chen G, Manwani D, et al. Neutrophil ageing is regulated by the micro-biome. Nature 2015;525:528–532.

9. Pillay J, Kamp VM, van Hoffen E, et al. A subset of neutrophils in human systemicinflammation inhibits T cell responses through Mac-1. J Clin Invest 2012;122:327–336.

10. de Oliveira S, Rosowski EE, Huttenlocher A. Neutrophil migration in infection andwound repair: going forward in reverse. Nat Rev Immunol 2016;16:378–391.

11. Ruhnau J, Schulze K, Gaida B, et al. Stroke alters respiratory burst in neutrophils andmonocytes. Stroke 2014;45:794–800.

12. Kelly PJ, Morrow JD, NingM, et al. Oxidative stress andmatrix metalloproteinase-9 inacute ischemic stroke: the biomarker evaluation for antioxidant therapies in stroke(BEAT-Stroke) study. Stroke 2008;39:100–104.

13. Pallister I, Bhatia R, Katpalli G, Allison D, Parker C, Topley N. Alteration of poly-morphonuclear neutrophil surface receptor expression and migratory activity afterisolation: comparison of whole blood and isolated PMN preparations from normaland postfracture trauma patients. J Trauma 2006;60:844–850.

14. Furie KL, Jayaraman MV. 2018 guidelines for the early management of patients withacute ischemic stroke. Stroke 2018;49:509–510.

15. Campillo-Gimenez L, Casulli S, Dudoit Y, et al. Neutrophils in antiretroviral therapy-controlled HIV demonstrate hyperactivation associated with a specific IL-17/IL-22environment. J Allergy Clin Immunol 2014;134:1142–1152.e5.

16. Dong Y, Lagarde J, Xicota L, et al. Neutrophil hyperactivation correlates with Alz-heimer’s disease progression. Ann Neurol 2018;83:387–405.

17. Kessenbrock K, Krumbholz M, Schonermarck U, et al. Netting neutrophils in auto-immune small-vessel vasculitis. Nat Med 2009;15:623–625.

18. Bejot Y, Bailly H, Durier J, GiroudM. Epidemiology of stroke in Europe and trends forthe 21st century. Presse Med 2016;45:e391–e398.

19. Xue J, Huang W, Chen X, et al. Neutrophil-to-Lymphocyte ratio is a prognosticmarker in acute ischemic stroke. J Stroke Cerebrovasc Dis 2017;26:650–657.

20. Vogelgesang A, Lange C, Blumke L, et al. Ischaemic stroke and the recanalization drugtissue plasminogen activator interfere with antibacterial phagocyte function.J Neuroinflammation 2017;14:140.

21. Martinod K, Wagner DD. Thrombosis: tangled up in NETs. Blood 2014;123:2768–2776.

22. Sauce D, Dong Y, Campillo-Gimenez L, et al. Reduced oxidative burst by primedneutrophils in the elderly individuals is associated with increased levels of theCD16bright/CD62Ldim immunosuppressive subset. J Gerontol A Biol Sci Med Sci2017;72:163–172.

23. Colom B, Bodkin JV, Beyrau M, et al. Leukotriene B4-neutrophil elastase Axis drivesneutrophil reverse transendothelial cell migration in vivo. Immunity 2015;42:1075–1086.

24. Shichita T, Ito M, Yoshimura A. Post-ischemic inflammation regulates neural damageand protection. Front Cell Neurosci 2014;8:319.

25. Elbim C, Bailly S, Chollet-Martin S, Hakim J, Gougerot-Pocidalo MA. Differentialpriming effects of proinflammatory cytokines on human neutrophil oxidative burst inresponse to bacterial N-formyl peptides. Infect Immun 1994;62:2195–2201.

26. Berthelot F, Fattoum L, Casulli S, Gozlan J, Marechal V, Elbim C. The effect ofHMGB1, a damage-associated molecular pattern molecule, on polymorphonuclearneutrophil migration depends on its concentration. J Innate Immun 2012;4:41–58.

27. Brouns R, Wauters A, De Surgeloose D, Marien P, De Deyn PP. Biochemical markersfor blood-brain barrier dysfunction in acute ischemic stroke correlate with evolutionand outcome. Eur Neurol 2011;65:23–31.

28. Ullen A, Singewald E, Konya V, et al. Myeloperoxidase-derived oxidants induce blood-brain barrier dysfunction in vitro and in vivo. PLoS One 2013;8:e64034.

29. Gidday JM, Gasche YG, Copin JC, et al. Leukocyte-derived matrix metalloproteinase-9 mediates blood-brain barrier breakdown and is proinflammatory after transient focalcerebral ischemia. Am J Physiol Heart Circ Physiol 2005;289:H558–H568.

30. Turner RJ, Sharp FR. Implications of MMP9 for blood brain barrier disruption andhemorrhagic transformation following ischemic stroke. Front Cell Neurosci 2016;10:56.

31. Uhl B, Vadlau Y, Zuchtriegel G, et al. Aged neutrophils contribute to the first line ofdefense in the acute inflammatory response. Blood 2016;128:2327–2337.

32. Vidale S, Consoli A, Arnaboldi M, Consoli D. Postischemic inflammation in acutestroke. J Clin Neurol 2017;13:1–9.

33. Luby M, Warach SJ, Nadareishvili Z, Merino JG. Immediate changes in stroke lesionvolumes post thrombolysis predict clinical outcome. Stroke 2014;45:3275–3279.

34. Kate MP, Riaz P, Gioia L, et al. Dynamic evolution of diffusion-weighted imaginglesions in patients with minor ischemic stroke. Stroke 2015;46:2318–2321.

35. Valles J, Lago A, Santos MT, et al. Neutrophil extracellular traps are increased inpatients with acute ischemic stroke: prognostic significance. Thromb Haemost 2017;117:1919–1929.

36. Ducroux C, Di Meglio L, Loyau S, et al. Thrombus neutrophil extracellular traps contentimpair tPA-induced thrombolysis in acute ischemic stroke. Stroke 2018;49:754–757.

37. Ellett F, Elks PM, Robertson AL, Ogryzko NV, Renshaw SA. Defining the phenotype ofneutrophils following reverse migration in zebrafish. J Leukoc Biol 2015;98:975–981.

38. Buckley CD, Ross EA, McGettrick HM, et al. Identification of a phenotypically andfunctionally distinct population of long-lived neutrophils in a model of reverse en-dothelial migration. J Leukoc Biol 2006;79:303–311.

39. Rabquer BJ, Amin MA, Teegala N, et al. Junctional adhesion molecule-C is a solublemediator of angiogenesis. J Immunol 2010;185:1777–1785.

40. Fan Y, Yang GY. Therapeutic angiogenesis for brain ischemia: a brief review. J Neu-roimmune Pharmacol 2007;2:284–289.

41. Ormstad H, Aass HC, Lund-Sorensen N, Amthor KF, Sandvik L. Serum levels ofcytokines and C-reactive protein in acute ischemic stroke patients, and their re-lationship to stroke lateralization, type, and infarct volume. J Neurol 2011;258:677–685.

42. Tsukagawa T, Katsumata R, Fujita M, et al. Elevated serum high-mobility group box-1protein level is associated with poor functional outcome in ischemic stroke. J StrokeCerebrovasc Dis 2017;26:2404–2411.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 13

Page 23: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

43. Fann DY, Lim YA, Cheng YL, et al. Evidence that NF-kappaB and MAPK signalingpromotes NLRP inflammasome activation in neurons following ischemic stroke. MolNeurobiol 2018;55:1082–1096.

44. Elbim C, Guichard C, Dang PM, et al. Interleukin-18 primes the oxidative burst ofneutrophils in response to formyl-peptides: role of cytochrome b558 translocation andN-formyl peptide receptor endocytosis. Clin Diagn Lab Immunol 2005;12:436–446.

45. Singh V, Roth S, Llovera G, et al. Microbiota dysbiosis controls the neuro-inflammatory response after stroke. J Neurosci 2016;36:7428–7440.

46. Yamashiro K, Tanaka R, Urabe T, et al. Gut dysbiosis is associated with metabolismand systemic inflammation in patients with ischemic stroke. PLoS One 2017;12:e0171521.

47. Benakis C, Brea D, Caballero S, et al. Commensal microbiota affects ischemicstroke outcome by regulating intestinal gammadelta T cells. Nat Med 2016;22:516–523.

48. Gelderblom M, Weymar A, Bernreuther C, et al. Neutralization of the IL-17 axisdiminishes neutrophil invasion and protects from ischemic stroke. Blood 2012;120:3793–3802.

49. Lewkowicz P, Lewkowicz N, Sasiak A, Tchorzewski H. Lipopolysaccharide-activatedCD4+CD25+ T regulatory cells inhibit neutrophil function and promote their apo-ptosis and death. J Immunol 2006;177:7155–7163.

50. Li P, Gan Y, Sun BL, et al. Adoptive regulatory T-cell therapy protects against cerebralischemia. Ann Neurol 2013;74:458–471.

14 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 24: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

ARTICLE OPEN ACCESS

α4-integrin deficiency in B cells does not affectdisease in a T-cell–mediated EAE disease modelRehana Z. Hussain, MSc, Petra D. Cravens, PhD, William A. Miller-Little, BA, Richard Doelger, BS, MSc,

Valerie Granados, PhD, Emily Herndon,MD, Darin T. Okuda,MD, ToddN. Eagar, PhD, andOlaf Stuve,MD, PhD

Neurol Neuroimmunol Neuroinflamm 2019;6:e563. doi:10.1212/NXI.0000000000000563

Correspondence

Dr. Stuve

[email protected]

AbstractObjectiveThe goal of this study was to investigate the role of CD 19+ B cells within the brain and spinalcord during CNS autoimmunity in a peptide-induced, primarily T-cell–mediated experimentalautoimmune encephalomyelitis (EAE) model of MS. We hypothesized that CD19+ B cellsoutside the CNS drive inflammation in EAE.

MethodsWe generated CD19.Cre+/− α4-integrinfl/flmice. EAEwas induced by active immunization withmyelin oligodendrocyte glycoprotein peptide (MOGp35-55). Multiparameter flow cytometrywas used to phenotype leukocyte subsets in primary and secondary lymphoid organs and theCNS. Serum cytokine levels and Ig levels were assessed by bead array. B-cell adoptive transferwas used to determine the compartment-specific pathogenic role of antigen-specific andnon–antigen-specific B cells.

ResultsA genetic ablation of α4-integrin in CD19+/− B cells significantly reduced the number of CD19+

B cells in the CNS but does not affect EAE disease activity in activeMOGp35-55-induced disease.The composition of B-cell subsets in the brain, primary lymphoid organs, and secondarylymphoid organs of CD19.Cre+/− α4-integrinfl/fl mice was unchanged during MOGp35-55-induced EAE. Adoptive transfer of purified CD19+ B cells from CD19.Cre+/− α4-integrinfl/fl

mice or C57BL/6 wild-type (WT) control mice immunized with recombinant rMOG1-125 orovalbumin323-339 into MOGp35-55-immunized CD19.Cre+/− α4-integrinfl/fl mice caused worseclinical EAE than was observed in MOGp35-55-immunized C57BL/6 WT control mice that didnot receive adoptively transferred CD19+ B cells.

ConclusionsObservations made in CD19.Cre+/− α4-integrinfl/fl mice in active MOGp35-55-induced EAEsuggest a compartment-specific pathogenic role of CD19+ B cells mostly outside of the CNSthat is not necessarily antigen specific.

From the Department of Neurology and Neurotherapeutics (R.Z.H., P.C.C., W.A.M.-L., R.D., V.G., D.T.O., O.S.) and Department of Pathology (E.H.), University of Texas SouthwesternMedical Center, Dallas; Department of Pathology and Genomic Medicine (T.N.E.), Houston Methodist Hospital; Neurology Section (O.S.), VA North Texas Health Care System, MedicalService; and Department of Neurology (O.S.), Klinikum Rechts der Isar, Technische Universitat Munchen, Germany.

Funding information and disclosures are provided at the end of the article. Full disclosure form information provided by the authors is available with the full text of this article atNeurology.org/NN.

The Article Processing Charge was funded by United States Department of Veterans Affairs I01BX001674.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 25: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Recent clinical trials with B-cell–depleting anti-CD20 thera-peutic monoclonal antibodies illustrated a pathogenic role forB lymphocytes in MS.1–4 Whether B-cell depletion outside ofthe CNS is sufficient to provide a detectable benefit in MS orwhether a reduction in the number of B lymphocytes withinthe CNS compartment is required to diminish inflammationremains incompletely understood.

In 1992, it was determined that the binding of leukocytes toinflamed CNS venules was inhibited by antibodies against α4-integrin.5 Natalizumab, a humanized recombinant mono-clonal antibody, was the first approved α4-integrin antagonistfor treatment of relapsing forms ofMS.6 Natalizumab is highlyeffective in decreasing the number of CD19+ B cells in CSF.7

The goal of this study was to investigate the role of α4-integrinablation in CD19+ B cells in a peptide-induced, primarilyT-cell–mediated experimental autoimmune encephalomyelitis(EAE)model and to identify compartment-specific contributionsof B cells to disease initiation and perpetuation. A T-cell–mediated EAEmodel was chosen to reflect the role of α4-integrinin B cells in patients with MS as closely as possible. Genetically,MS is most strongly associated with human leukcoyte antigen-DRB1*15:01,8,9 an association that implies a pathogenicinvolvement of an antigen-specific CD4+ T cell in MS.

Flow cytometry was used to phenotype leukocyte subsets inlymphoid organs and the CNS. Serum cytokine levels andimmunoglobulin (Ig) levels were assessed by ELISA. B-celladoptive transfer was used to determine the compartment-specific pathogenic role of antigen-specific B cells.

MethodsGeneration of CD19.Cre+/2 α4-integrin–deficient miceBecause α4-integrin is an absolute requirement for normalorgan development, α4-integrin–deficient (α−/−) mice areembryonic lethal.10 Thus, it is not possible to conduct EAEexperiments in animals that are completely devoid of α4-integrin. To examine how the deficiency of α4-integrin affectsthe migration of dendritic cells and B cells into the CNS andT-cell reactivation and retention in the CNS, we used cre-loxP–mediated recombination11 to create B-cell lineage–specific α4-integrin gene knockout mice. Specifically, wecrossed female mice that are homozygous for the α4-integrin–floxed allele (α4f/f)12 with commercially availableCD19.Cre+ males for the ablation of α4-integrin in B cells.Insertion of cre disrupts the CD19 coding sequence, leading to

a CD19 deficiency and a concomitant reduction in germinalcenters (GCs) in homozygous animals. Consequently,CD19.Cre+/+ mice behave functionally very similarly to B-cell–deficient mice. CD19.Cre+/+ mice on the C57BL/6 back-ground were used to generate CD19.Cre+/− α4-integrinfl/flmicethat appear developmentally normal and fertile. C57BL/6 micewere purchased from (The Jackson Laboratories, Bar Harbor,MN). α4-integrinfl/fl mice were used as controls. Male and fe-male mice were used for experiments. We observed no differ-ences regarding disease scores, cellular composition, or any ofthe biochemical and cellular outcomes between the 2 sexes.

PeptidesMouse myelin oligodendrocyte glycoprotein peptide(MOGp)35-55 (MEVGWYRSPFSRVVHLYRNGK) and oval-bumin (OVA)323-339 (ISQAVHAAHAEINEAGR) were syn-thesized by solid-phase Fmoc chemistry by QCB, Inc.(Hopkinton, MA) and CS Bio (Menlo Park, CA). RecombinantrMOG1-125 was as donation ofDr.Hans-Christian vonBudingenat the University of California, San Francisco (UCSF).

Experimental autoimmune encephalomyelitisTo induce active EAE, experimental mice were immunizedsubcutaneously with myelin MOGp35-55 (200 μg/100 μL/mouse), emulsified in an equal volume of complete Freund ad-juvant containing 4 mg/mL H37Ra Mycobacterium tuberculosis(Difco, BD, Franklin Lakes, NJ) in each flank as described.13

For B-cell adoptive transfer, spleens of donor mice immunizedwith MOG1-125 or OVA323-339 were removed at day 12, andsingle-cell suspensions were prepared as previously described.14

The Miltenyi kit 130-090-862 was used to purify a total of 10 ×106 CD19+ donor B cells (Miltenyi Biotec, San Diego, CA).Briefly, highly pure resting B cells were isolated by magneticlabeling and depleted of CD43-expressing B cells (activatedB cells, plasma cells [PCs], andCD5+B-1a cells) and non–B cells.Purified cells were subsequently transferred IV into recipientCD19.cre+/− α4-integrinfl/fl mice that were then immediatelyimmunized with MOGp35-55. For all experiments, individualanimals were observed and scored as described.13

Isolation of lymph node cells and splenocytesLymph node (LN) cells and splenocytes were isolated bypressing through a 70-μMnylonmesh cell strainer as described.13

Percoll PLUS density gradientIn all experiments in which tissue is referred to as CNS, andnot specifically as brains or spinal cords, CNS leukocytes wereisolated by Percoll PLUS (GE Healthcare Bio-Sciences,Pittsburgh, PA) gradient as previously described.13

GlossaryBM = bone marrow; DC = dendritic cell; EAE = experimental autoimmune encephalomyelitis; GC = germinal center; IFN =interferon gamma; IL = interleukin; LN = lymph node;MGZ = marginal zone;MOG = myelin oligodendrocyte glycoprotein;OVA = ovalbumin; PB = plasma blast; PC = plasma cell; WT = wild type.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 26: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Enzymatic CNS digestionsFor some experiments, brains and spinal cords were dissoci-ated enzymatically as described.13

Flow cytometryTo determine the absolute number of B cells and T cells indifferent compartments during EAE, multiparameter flowcytometry was used. Cells from bone marrows (BMs),spleens, LNs, brains, and spinal cords were brought intosingle-cell suspension as described.13 Cells were stained for 30minutes at 4°C with the following antibodies: CD45 PE-Cyannin-7 (30-F11; eBioscience, San Diego, CA) CD3 AlexaFlour 700 (17A2, eBioscience), CD19 PE-Texas Red (6D5,Invitrogen, Waltham, MA), CD19 PE (1D3 BD Bioscience,San Jose, CA), CD49d FITC (PS/2, Santa Cruz Bio-technology, Dallas, TX), CD4 APC (RM4-5, BD Bioscience),CD45R B220-APC (RA3-6B2, BioLegend, San Diego, CA),PE, PE Dazzle 594, PE-Texas Red, AF700, Pac Orange,PerCP, CD138 BV421(281-2, BioLegend), CD11b (APC,APC-Cy-7, PE, PerCP/Cy5.5, V450, B700 (M1/70, BioL-egend), CD23 PE-Cy7 (B3B4, BioLegend), CD5 PE-Cy5(53-7.3, BioLegend), CD1d PE (1B1, BioLegend), GL-7FITC (GL7, BioLegend), and CD21/CD35 BV510 (7G6,BD Bioscience, San Jose, CA). Isotype-matched mAbs wereused to set the gates. Next, 30,000–300,000 gated events wereacquired on an FACSAria II flow cytometer (BD Bio-sciences), equipped with Diva acquisition software (BDBiosciences). FlowJo (BD Biosciences) software was alsoused for some data analysis.

Intracellular cytokine stainingPhorbol 12-myristate 13-acetate (500 ng), ionomycin (500 ng),and GolgiPlug (1 μL) (BD Biosciences) were added to 1 × 106

cells and incubated for 3 hours at 37°C. Cells were then washedin phosphate-buffered saline and stained following the extra-cellular staining protocol for CD45-Alexa Fluor 700. After ex-tracellular staining, CD19+ cells were purified as described andfixed using Fixation Buffer (BioLegend) for 15 minutes in thedark at room temperature. Cells were then washed with Per-meabilization Buffer (BioLegend) twice and then stained usinginterferon gamma (IFNγ)-PE-Cy7 (BD Biosciences), in-terleukin (IL)-17A-FITC (BioLegend), tumor necrosis factorα-PE (eBioscience), and IL-6-APC (BioLegend). Cells were thenwashed, and acquired, and analyzed by flow cytometry.

ELISASerum samples were collected by submandibular bleeding atdays 13, 19, and 29 during early active EAE, maximum EAEdisease activity, and during chronic EAE. Quantitative ELISAfor IL-17, IL-10, IL-4, IL-5, and IFNγ was performed usingpaired mAb specific for corresponding cytokines as per themanufacturer’s recommendations (BD Biosciences or R&DSystems). IgM and IgG serum levels were also determined byELISA. The results of ELISA assays are expressed as an averageof triplicate wells ± SD. The EPOCH (BioTek, Winooski, VT)ELISA plate reader and software were used for data analysis(Molecular Devices Corporation, Sunnyvale, CA).

T-cell proliferation assayTo determine the capability of B cells to serve as APC to CD4+

T cells, flow cytometric proliferation assays using the greenfluorescent dye carboxyfluorescein succinimidyl ester or V450(BD Biosciences) were performed according to publishedmethods.15 On day 15 after active induction of EAE, spleens andLNs were harvested and processed into single-cell suspension.Splenocytes and lymphocytes were stained with carboxy-fluorescein succinimidyl ester and plated at 1 × 106 cells per well.The cells were then restimulated with 10 μg/mL of MOGp35-55

and incubated at 37°C for 5 days. Con A at a concentration of 2mg/mL and media were used as controls. On day 6, the cellswere collected, blockedwith Fc block antibody, stainedwithAPCCD4 antibody, and acquired and analyzed by flow cytometry.

HistologyFollowing fixation in 10% buffered formalin, LNs were pro-cessed and embedded in paraffin blocks. Four-micrometersections were cut, mounted on Fisherbrand Superfrost Plusglass slides (Fisher Scientific, Pittsburgh, PA), and stainedwith hematoxylin & eosin (Fisher Scientific) or anti-CD19.

Statistical analysisFor parametric tests, data were assessed for normality using theKolmogorov-Smirnov test. Normally distributed values werecompared using the unpaired 2-sided Student t test. Correla-tions between continuous and categorical variables wereassessed using the Mann-Whitney U-test. All statistical testswere 2 sided, and p < 0.05 indicated significance. All analyseswere performed with Prism 7 (Graphpad, La Jolla, CA).

Standard protocol approvals and registrationsAll experimental animals were maintained in a specificpathogen-free facility at the University of Texas (UT) South-western Medical Center. All protocols involving mice handlingwere approved by the UT Southwestern animal care facility.

Data availabilityData not shown will be shared by request.

ResultsThe percentage of CD49d+ CD19+ B cells isdiminished in primary and secondarylymphoid organs in naive CD19.Cre+/2 α4-integrinfl/fl miceTo characterize CD19.Cre+/− α4-integrinfl/fl mice, the ex-pression of α4-integrin (CD49d) was determined by multi-parameter flow cytometry. The percentage of CD49d+

CD19+ B cells in the BM, LNs, spleens, and Peyer patches issignificantly diminished in naive CD19.Cre+/− α4-integrinfl/fl

mice compared with C57BL/6 wild-type (WT) controlmice (figure 1A). Cre recombination efficiency is in-complete16 and determined mostly by the nucleotidesequence in the spacer region of the lox site. Also, there isa negative correlation between the Cre/lox recombinationefficiency and the length of DNA between the 2 lox sites. As

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 27: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

stated above, other investigators found a deletion efficiencybetween 75% and 80% in BM-derived pre–B cells and ap-proximately 90% in splenic B cells.17

CD19.Cre+/2 α4-integrinfl/fl mice immunizedwith MOGp35-55 display a regular clinicalEAE courseTo determine the effect of α4-integrin expression of B cells ina T-cell–mediated active model of EAE, CD19.Cre+/− α4-integrinfl/fl mice and C57BL/6 control mice were immunizedwithMOGp35-55. No significant differences regarding the EAEdisease incidence, severity, or phenotype were observed be-tween the mouse strains (figure 1B and table). α4-integrinfl/fl

mice were also used as controls and showed similar diseaseactivity (data not shown).

α4-integrin expressionon splenic B cells affectsantigen-specific CD4+ T-cell proliferationIt was previously demonstrated that α4-integrin can participatein costimulation of CD4+ T cells.18,19 Yet other investigatorsshowed that anti-CD49d plus anti-CD3/anti-CD28–coatedpolystyrene beads induced significantly greater T-cell pro-liferation than anti-CD3/anti-CD28 polystyrene beads alone.20

To test the effect of α4-integrin ablation on the capability ofCD19+ B cells to present antigen to antigen-reactive CD4+

T cells, flow cytometric proliferation assays were performed.CD4+ T-cell recall responses to MOGp35-55 in spleens weresignificantly diminished in CD19.Cre+/− α4-integrinfl/fl mice(figure 1C). However, in LNs, we did not observe any differ-ence in CD4+ T-cell proliferation between CD19.Cre+/− α4-integrinfl/fl mice and C57BL/6 WT control mice (figure 1D).

Figure 1 CD19.Cre+/− α4-integrinfl/fl mice are fully susceptible to actively MOGp35-55-induced experimental autoimmuneencephalomyelitis

(A) The percentage of CD49d+ CD19+ B cells in thebone marrow, lymph nodes, spleens, and Peyerpatches is significantly diminished in naiveCD19.Cre+/− α4-integrinfl/fl mice compared withC57BL/6 control mice. Cells were immunophe-notyped by multiparameter flow cytometry. (B)CD19.Cre+/− α4-integrinfl/fl mice behave similar toC57BL/6 wild-type (WT) mice regarding the dis-ease incidence, onset, and severity of active ex-perimental autoimmune encephalomyelitis (EAE).α4-integrinfl/fl mice were used as controls andshowed similar disease activity (data not shown).Active EAE was induced in CD19.Cre+/− α4-integ-rinfl/fl and C57BL/6 age-matched control mice bysubcutaneous immunizations with MOGp35-55 inincomplete Freund adjuvant (IFA) containing 4mg/mL mycobacteria. Mice received in-traperitoneal injection of 200 μL of pertussis toxin(Ptx) at 200 ng/mL on days 0 and 2. Mice wereobserved daily, and EAE severity was scored usinga 5-point scale. CD4+ T-cell recall responses toMOGp35-55 in (C) spleens were significantly di-minished in CD19.Cre+/− α4-integrinfl/fl mice, but(D) indistinguishable to those in C57BL/6 WTmicein lymphnodes. Cell proliferationwas determinedby a flow cytometric proliferation assay using thegreen fluorescent dye CFSE or V450. (E) At maxi-mum disease activity (days 13–15), the percent-age of CD3+ T cells in the CNS was similar inCD19.Cre+/− α4-integrinfl/fl and C57BL/6 WT mice.In contrast, the percentage of CD19+ B cells in theCNS was significantly diminished in CD19.Cre+/−

α4-integrinfl/fl mice. (F) The percentage of α4-integrin–positive (CD49d+) CD19+ B cells in theCNS during maximum EAE disease activity wascomparable between both mouse strains. Lym-phocytes were immunophenotyped by multipa-rameter flow cytometry. The number,appearance, and architecture of germinal centersin lymph nodes of (G and H) C57BL/6 WT controlmice were comparable to those in (I and J)CD19.Cre+/− α4-integrinfl/fl mice. Panels G and Iwere stained with hematoxylin & eosin, andpanels H and J were stained with anti-CD19.Magnification for G–J is ×4. *p < 0.05, **p < 0.01,***p < 0.001, ****p < 0.0001. CFSE = carboxy-fluorescein succinimidyl ester.

4 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 28: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

In both mouse strains, T-cell proliferation was substantial. Ourobservations appear to differ from those of other investigators,who showed that B cells do not presentMOGp35-55 to T cells.21

The Methods section of that article states that “…B cells weremagnetically activated cell sorting (MACS; Miltenyi Biotec,Bergisch Gladbach, Germany)-separated from lymph nodes orspleens.” Thus, a differential capability of B cells isolated fromeither LNs or spleens on MOG peptide presentation may nothave been fully tested.

The number of B cells in the CNS during activeMOGp35-55-induced EAE is diminished inCD19.Cre+/2 α4-integrinfl/fl miceTo examine the effect of α4-integrin deletion in CD19+ B cellson the ability of B lymphocytes to enter the CNS, cells wereisolated from brains and spinal cords by Percoll gradient andanalyzed by multiparameter flow cytometry. As MOGp35-55-induced EAE is a T-cell–mediated form of EAE, the number ofCD3+ T cells was also assessed by the same method. At maxi-mum disease activity (days 13–15), the percentage of CD3+

T cells in the CNS was similar in CD19.Cre+/− α4-integrinfl/fl

mice and C57Bl/6 WT mice (figure 1E). In contrast, the per-centage of CD19+ B cells in the CNS was significantly reducedin CD19.Cre+/− α4-integrinfl/fl mice (figure 1E). These obser-vations suggest that antigen-specific B cells within the CNS playa minor role in initiating and perpetuating EAE, given that thesemice showed the same disease susceptibility and disease courseas C57Bl/6 WT mice. Immunohistochemical studies to de-termine a differential anatomic distribution of B cells in theCNSbetween CD19.Cre+/− α4-integrinfl/fl mice and C57Bl/6 WTmice were not performed, as the absolute number of B cells inCD19.Cre+/− α4-integrinfl/fl mice was very low.

CD19+ B cells from CD19.Cre+/2 α4-integrinfl/fl

mice use CD49d to gain access to the CNSduring active MOGp35-55-induced EAETo determine whether CD19+ B cells from CD19.Cre+/− α4-integrinfl/fl mice that gain access to the CNS during active EAEuse α4-integrin, the percentage of α4-integrin–positive (CD49d+)CD19+ B cells was determined by multiparameter flowcytometry and found to be comparable between CD19.Cre+/−

α4-integrinfl/fl mice and C57Bl/6 WT controls (figure 1F).

The number and architecture of GCs in LNs ofCD19.Cre+/2 α4-integrinfl/fl mice are normalCD19.Cremice express cre under the transcriptional control ofthe B lineage–restricted CD19 gene.17 Insertion of cre disruptsthe CD19 coding sequence, leading to a CD19 deficiency anda concomitant reduction in GCs in CD19.Cre+/+ mice. At day15 during the acute phase of MOGp35-55-induced EAE, thenumber and architecture of GCs in LNs of CD19.Cre+/− α4-integrinfl/fl mice and C57BL/6 WT control mice were com-parable by hematoxylin & eosin (H&E) staining (figure 1, Gand H) and anti-CD19 staining (figure 1, I and J).

B cells of CD19.Cre+/2 α4-integrinfl/fl mice havenormal cytokine profiles during activeMOGp35-

55-induced EAETo test whether the absence of α4-integrin in CD19+ B cellsaffects their differentiation during inflammatory conditions,experimental animals were killed at day 13 during the acutephase of MOGp35-55-induced EAE, and B cells in disease-relevant compartments were immunophenotyped by multi-parameter flow cytometry and intracellular cytokine staining.Compared with B cells from C57BL/6 WT control mice, thepercentage of CD19+ B cells expressing IFNγ, IL-6, IL-10, orIL-17 in CD19.Cre+/− α4-integrinfl/fl mice was similar in theBM (figure 2A), spleen (figure 2B), LNs (figure 2C), andbrain (figure 2D) of CD19.Cre+/− α4-integrinfl/fl mice.

Outside the CNS, B cells of CD19.Cre+/2 α4-integrinfl/fl mice express normal maturationand activation markers during active MOGp35-

55-induced EAEAs stated above, α4-integrin is considered a costimulatorymolecule through which CD19+ B cells are capable ofinteracting with CD4+ T cells. To further elucidate theeffect of α4-integrin deficiency on CD19+ B cells on theirmaturation and activation in secondary lymphoid organs,experimental animals were killed at day 13 during theacute phase of MOGp35-55-induced EAE, and B cells inspleens and LNs were immunophenotyped by multipa-rameter flow cytometry. The percentages of CD19+ B220+

B cells expressing IgD, IgM, IgG, and major histocom-patibility complex (MHC) II in LNs (figure 2E) andspleen (figure 2F) were comparable in CD19.Cre+/− α4-integrinfl/fl mice and C57BL/6 WT control mice.

CNS B cells from CD19.Cre+/2 α4-integrinfl/fl

mice do not upregulate activation markersduring active MOGp35-55-induced EAEThe absolute number of CD19+ B cells is diminished in theCNS of CD19.Cre+/− α4-integrinfl/fl mice compared withC57BL/6 WT control mice during acute MOGp35-55-in-duced EAE (figure 1E). To evaluate the activation state ofthese cells, CNS B cells were isolated and assessed by mul-tiparameter flow cytometry during active EAE at day 13 afteractive induction or EAE with MOGp35-55. The surface ex-pression of MHC II was similar in CNS-derived B cells fromCD19.Cre+/− α4-integrinfl/fl and C57BL/6 WT controls

Table Disease incidence, mean maximum disease score,and mean day of active MOGp35-55-induced EAEdisease onset in CD19+/− α4-integrin mice andC57BL/6 control mice

Group IncidenceMean maximumdisease score

Mean day ofdisease onset

C57BL/6 72/80 2.89 10.57 ± 5.18

CD19+/2 α4-integrinfl/fl

62/81 2.45 9.26 ± 6.07

Abbreviations: EAE = experimental autoimmune encephalomyelitis; MOG =myelin oligodendrocyte glycoprotein.Results of 15 experiments are shown.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 5

Page 29: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

(figure 2G). Also, the expression of the costimulator mole-cules CD80 and CD86 and the first apoptosis signal receptorCD95 was comparable in brain-derived B cells fromCD19.Cre+/− α4-integrinfl/fl mice and C57BL/6 WT con-trols (figure 2H).

The composition of B-cell subsets in the brainof CD19.Cre+/2 α4-integrinfl/fl mice isunchanged during MOGp35-55-induced EAETo determine whether the deletion of α4-integrin on CD19+

B cells affects B-cell subsets differentially in their capability toenter the CNS, the percentage of innate and adaptiveB-lymphocyte subsets, including CD11b+CD23−B220− B1B cells, CD11b+CD23−B220lo/−CD5− B1b B cells,CD11b+CD23−B220lo/−CD5+ B1a B cells,CD19+CD5+CD1dhi B10 B cells, CD19+CD23+CD5−

naive follicular B cells, B220+CD138−CD95+GL7+ GCsB cells, B220+CD1d+CD5−CD23−CD21+ marginal zone

(MGZ) B cells, CD19low/–CD138+B220low/– plasma blast(PB), and CD19−CD138+ PCs, was assessed by multipa-rameter flow cytometry during acute MOGp35-55-inducedEAE at day 13 during early active EAE (figure 2I), day 19during maximum EAE disease activity (figure 2J), and day 29during chronic EAE (figure 2K) after active induction orEAE with MOGp35-55 and was found to be similar inCD19.Cre+/− α4-integrinfl/fl mice and C57BL/6 WT con-trols at all time points. A gating strategy used to identifysome of the B-cell subsets is shown in figure 3.

The composition of B-cell subsets in primaryand secondary lymphoid organs of CD19.Cre+/2

α4-integrinfl/fl mice is similar to those of WTcontrol mice during MOGp35-55-induced EAETo assess the effect of α4-integrin ablation on CD19+ B cellson B-cell phenotypes in lymphoid organs, innate and adaptiveB-lymphocyte subsets were characterized by flow cytometry.

Figure 2 CD19+ B cells in the CNS of CD19.Cre+/− α4-integrinfl/fl mice are phenotypically similar to wild-type (WT) B cellsduring myelin oligodendrocyte glycoprotein peptide (MOGp35-55)-induced experimental autoimmuneencephalomyelitis

Experimental animals were killed at day 13 during the acute phase of MOGp35-55-induced experimental autoimmune encephalomyelitis (EAE), and B cells indisease-relevant compartmentswere immunophenotyped bymultiparameter flow cytometry and intracellular cytokine staining. Comparedwith B cells fromC57BL/6 WT control mice, the percentage of CD19+ B cells expressing interferon gamma (IFNγ), interleukin (IL)-6, IL-10, or IL-17 in CD19.Cre+/− α4-integrinfl/fl

mice was similar in the (A) bone marrow (BM), (B) spleen, (C) lymph node (LNs), and (D) brain of CD19.Cre+/− α4-integrinfl/fl mice. The percentages of CD19+

B220+ B cells expressing IgD, IgM, IgG, and MHC II in (E) LNs and (F) spleen were comparable in both mouse strains. (G) In the CNS, the surface expression ofMHC II was similar in CD19.Cre+/− α4-integrinfl/fl mice and controls (H) In the brain, the expression of the costimulator molecules CD80 and CD86 and the firstapoptosis signal receptor CD95 was indistinguishable between CD19.Cre+/− α4-integrinfl/fl mice and C57BL/6 WT controls. The percentage ofCD11b+CD23B220lo B1 B cells, CD11b+CD23B220lo/−CD5+ B1a B cells, CD11b+CD23B220lo/−CD5− B1b B cells, B220+CD5+CD1dhi B10 B cells, CD19+CD23+CD5−

naive follicular B cells (FB), B220+CD138−CD95+GL7+ germinal center (GC) B cells, B220+CD1d+CD5−CD23−CD21+marginal zone (MGZ) B cells, CD138+B220low/–

plasma blast (PB), and B220−CD138+ plasma cells (PC) was similar in both mouse strains at (I) day 13 during early active EAE, (J) day 19 during maximum EAEdisease activity, and (K) day 29 during chronic EAE. Lymphocytes were immunophenotyped by multiparameter flow cytometry. A gating strategy used toidentify some of the B-cell subsets is shown in Figure 3. **p < 0.01.

6 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 30: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

We were unable to detect a change in composition ofCD19+CD11b+CD23−B220− B1 B cells,CD19+CD11b+CD23−B220−CD5+ B1a B cells,CD19+CD11b+CD23−B220−CD5− B1b B cells,CD19+CD5+CD1dhi B10 B cells,CD19+B220+CD1d+CD5−CD23−CD21+ MGZ B cells, andCD19low/–CD138+B220low/– PB CD19.Cre+/− α4-integrinfl/fl

mice and C57BL/6 WT controls at day 13 during early activeEAE in the BM (figure 4A), LNs (Figure 4B), and spleen(figure 4C).

Serum soluble inflammatory markers areunchanged in CD19.Cre+/2 α4-integrinfl/fl miceduring different stages of MOGp35-55-induced EAEDeletion of α4-integrin on CD19+ B cells diminishes theability of CD19+ B cells to enter the CNS during acuteMOGp35-55-induced EAE (figure 1E). It is conceivable thatinflammatory B cells accumulate in peripheral blood, alteringthe inflammatory milieu in that compartment. This wasobserved in patients withMS treated with natalizumab.22 Wefound that the serum expression of the cytokines IL-17A(figure 5A), IL-10 (Figure 5B), IL-4 (Figure 5C), IL-5

(figure 5D), and IFNγ (figure 5E), as measured by ELISA(ELISA) during acute and chronic MOGp35-55-induced EAEon days 13, 19, and 29, was comparable betweenCD19.Cre+/− α4-integrinfl/flmice and C57BL/6 WT controlmice.

Serum Ig levels are unchanged in CD19.Cre+/2

α4-integrinfl/fl mice during different stages ofMOGp35-55-induced EAETo further investigate whether the sequestration of CD19+

B cells through deletion of α4-integrin affects their acti-vation, proliferation, and antibody secretion, IgM and IgGlevels in serum were determined by ELISA. Serum levels ofIgM (figure 5F) and IgG (figure 5G) as measured byELISA were also similar between CD19.Cre+/− α4-integrinfl/fl mice and C57BL/6 WT control mice on days13, 19, and 29 of MOGp35-55-induced EAE. Given that theabsolute levels of IgM and IgG were similar between themouse strains, and that other investigators previouslyshowed that B cells from mice on the C57BL/6 back-ground do not readily recognize MOGp35-55,

21

we did not test the serum levels of MOGp35-55-specific IgMor IgG.

Figure 3 Immunophenotyping of murine B cells

B-lymphocyte subsets were immunophenotyped by multiparameter flow cytometry. A gating strategy that was used to identify some of the B-cell sub-populations is shown. (A) First, cellular area, height, and width measurements were obtained in a channel with linear scale to gate on singlets and excludedoublets. (B) Next, gates were set to include CD45+ leukocytes and exclude FSClow cell debris out. (C) CD138 and B220 were used to gate on plasmablasts(CD138+B220low/–), plasma cells (CD138+B220–), and B cells (B220+). Panels D–F show the gating for (E) B220+CD1d+CD5−CD23−CD21+IgM+ marginal zone Bcells, (E) B220+CD1d+CD5−CD23−CD21−IgM+ transitional 1 (T1) B cells, and (F) B220+CD1d+CD5−CD23+CD21+IgM+ T2 B cells. Panels G–I show gating for B1 B-cell populations, and the identification of (H) CD11b+CD23−B220lo/−CD5− B1b B cells and (I) CD11b+CD23−B220lo/−CD5+ B1a B cells. Panels J and K show theidentification of B220+CD138−CD95+GL7+ germinal center B cells. Panels L–O illustrate the gating for B220+CD19+CD5−GL7−CD23+CD21lo/int naive follicular Bcells. FSC = forward side scatter.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 7

Page 31: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Activated peripheral B cells contribute todisease severity in MOGp35-55-induced EAETo determine the compartment-specific pathogenic role ofB cells in the induction and perpetuation of the EAE animalmodel, CD19.Cre+/− α4-integrinfl/fl mice were immunizedwith MOGp35-55. As demonstrated, B cells in these mice havea diminished capacity to enter the CNS (figure 1E). Adoptivetransfer of purified CD19+ B cells from CD19.Cre+/− α4-integrinfl/fl mice and C57BL/6 WT mice immunized withrMOG1-125 or the control antigen OVA323-339 worsened the

clinical course of EAE in recipient mice immunized withMOGp35-55 (figure 5H), although these changes were notsignificant.

DiscussionThere is accumulating evidence to suggest that B cells andmyelin-specific antibodies play crucial roles in the pathogen-esis of MS. Whether B-cell depletion mediates its beneficial

Figure 4 CD19+ B cells in primary and secondary lymphoid organs of CD19.Cre+/− α4-integrinfl/fl mice are phenotypicallysimilar to wild-type B cells during MOGp35-55-induced experimental autoimmune encephalomyelitis

(A) In the bonemarrow, (B) lymph nodes, and (C) spleen, the percentage of CD19+CD11b+CD23−B220− B1 B cells, CD19+CD11b+CD23−B220−CD5+ B1a B cells,CD19+CD11b+CD23−B220−CD5− B1b B cells, CD19+CD5+CD1dhi B10 B cells, CD19+B220+CD1d+CD5−CD23−CD21+ marginal zone (MGZ) B cells, andCD19low/–CD138+B220low/– plasma blast (PB) was similar in CD19.Cre+/− α4-integrinfl/fl mice and C57BL/6 WT controls at day 13 during early activeEAE. Lymphocytes were immunophenotyped by multiparameter flow cytometry.

Figure 5ActivatedB cells outside of theCNSdrive disease activity inmyelin oligodendrocyte glycoprotein peptide (MOGp35-

55)-induced experimental autoimmune encephalomyelitis (EAE) in an antigen no-specific manner

The serum expression of the cytokines(A) IL-17A, (B) IL-10, (C) IL-4, (D) IL-5, and(E) IFNγ as measured by ELISA duringacute and chronic MOGp35-55-inducedEAE on days 13, 19, and 29 was com-parable between CD19.Cre+/− α4-integ-rinfl/fl mice and C57BL/6 wild-type (WT)control mice. Serum levels of (F) IgMand (G) IgG measured by ELSA at thesame time points were also similar be-tween mouse strains. (H) Adoptivetransfer of purified CD19+ B cells fromCD19.Cre+/− α4-integrinfl/fl mice orC57BL/6 WT control mice immunizedwith recombinant (r)MOG1-125 or oval-bumin (OVA)323-339 into CD19+/− α4-integrinfl/fl mice that were then imme-diately immunized with MOGp35-55

caused worse clinical EAE than wasobserved in MOGp35-55-immunizedC57BL/6 WT control mice (“C57BL/6”)that did not receive adoptively trans-ferred CD19+ B cells.

8 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 32: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

effects in patients with MS through B cells in the periphery, inthe CNS, or both is currently not completely understood.

Other investigators investigated the role of α4-integrin deficiencyin CD19+ B cells in EAE.23 These investigators showed thatB-cell very late activation antigen-4–deficient mice developedmore severe clinical disease than control mice after active in-duction of EAE with MOGp35-55 and that that this clinical effectis due to the requirement of regulatory B cells (Bregs) for α4-integrin to enter the CNS. Bregs were defined as CD19+IL-10+

or CD1dhiCD5+ B-cell subsets. Another group of investigatorsalso testedCD19Cre Itga4fl/flmice inMOGp35-55-induced activeEAE.24 CD19Cre Itga4fl/fl mice developed more severe EAEthan controlmice, and these investigators demonstrated a criticalrole for α4-integrin on the generation of Bregs in peripheralimmune organs.

Our own results diverge from these reports in that we couldnot detect a difference in active EAE disease severity betweenCD19.Cre+/− α4-integrinfl/fl mice and C57BL/6 WT controlmice. These experiments were conducted 15 times (table),and we are confident in our findings. Differences in EAEdisease severity and phenotype can be affected by conditionsunder which mouse colonies are maintained, as well as otherfactors that may affect the overall inflammatory milieu of ex-perimental conditions. This may explain the differences inresults. It is also conceivable that different mice were used bythe other investigators for their experiments. The inves-tigators of the 2 previous studies state that they usedCD19.Creα4fl/fl mice23 or that they used CD19Cre Itga4fl/fl

mice.25 It is not stated whether mice were heterozygous orhomozygous for CD19.cre or a mixture of both. As stated inthe Methods section, CD19.Cre+/+ mice behave functionallyvery similarly to B-cell–deficient mice, which develop moresevere EAE when immunized with MOGp35-55.

26

We found one potentially meaningful biological difference be-tween CD19.Cre+/− α4-integrinfl/fl mice and C57BL/6 controlmice. There was decreased MOGp35-55-specific CD4+ T-cellproliferation in spleens, which is an expected result, given thatα4-integrin can participate in costimulation of CD4+T cells.18,19

The observation that CD19+ B cells of CD19.Cre+/− α4-integrinfl/fl mice did deplete α4-integrin and that they weresignificantly less capable of entering the CNS after mice wereimmunized with MOGp35-55 suggests 3 interpretations: (1)Antigen-activated B cells contribute to disease burden in EAEin a seemingly dose-dependent manner, (2) they do not ab-solutely require recognition of a CNS autoantigen or a mimicthereof to do so, and (3) they do not require full α4-integrin–mediated access to the CNS to exacerbate disease.The first interpretation is supported by work of other inves-tigators who have previously demonstrated that B-cell–deficient C57BL/6 mice are resistant to active rMOG-inducedEAE26 and that the administration of a B-cell–depletingmonoclonal antibody prevents or reverses established rMOG-induced EAE in C57BL/6 mice.21 Weber and colleagues21

further showed that B-cell depletion results in a reduction ofMOGp35-55-specific Th1 and Th17 cells. Our own data suggestthat the differentiation ofMOGp35-55-specific T cells to becomeencephalitogenic does not absolutely require the presentationof autoantigen, but is likely driven by other B-cell factors, in-cluding soluble inflammatory mediators. The third in-terpretation of this experiment is supported by ourdemonstration that B cells from CD19.Cre+/− α4-integrinfl/fl

mice are significantly impaired in their ability to enter the CNS(figure 1E). Finally, it should be emphasized that our findingsdo not conclusively rule out a role of antigen-specific B cells inthe CNS on the course of CNS autoimmunity, as thresholdeffects may be difficult to detect with the EAE experiments thatwere conducted.

Our data suggest that the beneficial effects of anti-CD20therapies in patients with MS are mediated through theireffects on peripheral B cells. We also conclude that benefitsobserved in patients with MS during anti-α4-integrin therapyare likely not predominantly due to its effect of suppressingthe migration of B cells into the CNS. The interpretation ofour findings should be limited to early forms of MS, as theEAE model that was used and the observation period that wasused do not provide insight into late or progressive forms ofCNS autoimmunity. Magliozzi et al27 were the first inves-tigators to show that some patients with secondary-progressive MS develop lymphoid tissue in the meningesthat resemble B-cell follicles in secondary lymphoid organs.B cells in these structures may be susceptible to anti-CD20therapies with monoclonal antibodies or small molecules, andit is conceivable that their depletion may benefit patientsafflicted with that MS phenotype.

AcknowledgmentThe authors thank Dr. Thalia Papayannopoulou at theUniversity of Washington for providing them with α4-integrinflox/flox mice. Dr. Stuve was funded by a Merit Reviewgrant (federal award document number [FAIN]I01BX001674) from the United States (U.S.) Departmentof Veterans Affairs, Biomedical Laboratory Research andDevelopment.

Study fundingNo targeted funding reported.

DisclosureR.Z. Hussain, P.C. Cravens, W.A. Miller-Little, R. Doelger, V.Grandos, and E. Herndon report no disclosures. D.T. Okudareceived advisory and consulting fees from Celgene, Gen-entech, Genzyme, EMD Serono, and Novartis and researchsupport from Biogen; served on the scientific advisory boardof Osmotica; and served on the speakers’ bureau of Acorda,Genzyme, and Teva. T.N. Eagar serves as section editor ofArchives of Pathology and received research support from theNIH. O. Stuve served on the scientific advisory boards of TGTherapeutics and Genentech-Roche; served on the editorialboard of Therapeutic Advances in Neurological Disorders;

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 9

Page 33: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

consulted for Celgene, EMD Serono, and Novartis; andreceived research support from Sanofi and Genzyme.Disclosures available: Neurology.org/NN.

Publication historyReceived by Neurology: Neuroimmunology & NeuroinflammationNovember 2, 2018. Accepted in final form February 1, 2019.

References1. Hauser SL, Waubant E, Arnold DL, et al. B-cell depletion with rituximab in relapsing-

remitting multiple sclerosis. N Engl J Med 2008;358:676–688.2. Hawker K, O’Connor P, Freedman MS, et al. Rituximab in patients with primary

progressive multiple sclerosis: results of a randomized double-blind placebo-controlled multicenter trial. Ann Neurol 2009;66:460–471.

3. Hauser SL, Bar-Or A, Comi G, et al. Ocrelizumab versus interferon beta-1a in re-lapsing multiple sclerosis. N Engl J Med 2017;376:221–234.

4. Montalban X, Hauser SL, Kappos L, et al. Ocrelizumab versus placebo in primaryprogressive multiple sclerosis. N Engl J Med 2017;376:209–220.

5. Yednock TA, Cannon C, Fritz LC, Sanchez-Madrid F, Steinman L, Karin N. Pre-vention of experimental autoimmune encephalomyelitis by antibodies against alpha 4beta 1 integrin. Nature 1992;356:63–66.

6. Frohman EM, Racke MK, Raine CS. Multiple sclerosis—the plaque and its patho-genesis. N Engl J Med 2006;354:942–955.

7. Stuve O, Marra CM, Jerome KR, et al. Immune surveillance in multiple sclerosispatients treated with natalizumab. Ann Neurol 2006;59:743–747.

8. Haines JL, Ter Minassian M, Bazyk A, et al. A complete genomic screen for multiplesclerosis underscores a role for the major histocompatability complex. The MultipleSclerosis Genetics Group. Nat Genet 1996;13:469–471.

9. Sawcer S, Jones HB, Feakes R, et al. A genome screen in multiple sclerosis revealssusceptibility loci on chromosome 6p21 and 17q22. Nat Genet 1996;13:464–468.

10. Yang JT, Rayburn H, Hynes RO. Cell adhesion events mediated by alpha 4 integrinsare essential in placental and cardiac development. Development 1995;121:549–560.

11. Sternberg N, Hamilton D. Bacteriophage P1 site-specific recombination. I. Re-combination between loxP sites. J Mol Biol 1981;150:467–486.

12. Scott LM, Priestley GV, Papayannopoulou T. Deletion of alpha4 integrins from adulthematopoietic cells reveals roles in homeostasis, regeneration, and homing. Mol CellBiol 2003;23:9349–9360.

13. Hussain RZ, Miller-Little WA, Doelger R, et al. Defining standard enzymatic disso-ciation methods for individual brains and spinal cords in EAE. Neurol NeuroimmunolNeuroinflamm 2018;5:e437. doi: 10.1212/NXI.0000000000000437.

14. Cravens PD, Hussain RZ, Zacharias TE, et al. Lymph node-derived donor encepha-litogenic CD4+ T cells in C57BL/6 mice adoptive transfer experimental autoimmuneencephalomyelitis highly express GM-CSF and T-bet. J Neuroinflammation 2011;8:73.

15. Karandikar NJ, Crawford MP, Yan X, et al. Glatiramer acetate (Copaxone) therapyinduces CD8(+) T cell responses in patients with multiple sclerosis. J Clin Invest2002;109:641–649.

16. Gu H, Marth JD, Orban PC, Mossmann H, Rajewsky K. Deletion of a DNA poly-merase beta gene segment in T cells using cell type-specific gene targeting. Science1994;265:103–106.

17. Rickert RC, Roes J, Rajewsky K. B lymphocyte-specific, Cre-mediated mutagenesis inmice. Nucleic Acids Res 1997;25:1317–1318.

18. Davis LS, Oppenheimer-Marks N, Bednarczyk JL, McIntyre BW, Lipsky PE. Fibro-nectin promotes proliferation of naive and memory T cells by signaling through boththe VLA-4 and VLA-5 integrin molecules. J Immunol 1990;145:785–793.

19. Shimizu Y, van Seventer GA, Horgan KJ, Shaw S. Costimulation of proliferativeresponses of resting CD4+ T cells by the interaction of VLA-4 and VLA-5 withfibronectin or VLA-6 with laminin. J Immunol 1990;145:59–67.

20. Theien BE, Vanderlugt CL, Eagar TN, et al. Discordant effects of anti-VLA-4 treat-ment before and after onset of relapsing experimental autoimmune encephalomyelitis.J Clin Invest 2001;107:995–1006.

21. Weber MS, Prod’homme T, Patarroyo JC, et al. B-cell activation influences T-cellpolarization and outcome of anti-CD20 B-cell depletion in central nervous systemautoimmunity. Ann Neurol 2010;68:369–383.

22. Krumbholz M, Meinl I, Kumpfel T, Hohlfeld R, Meinl E. Natalizumab dispropor-tionately increases circulating pre-B and B cells in multiple sclerosis. Neurology 2008;71:1350–1354.

23. Lehmann-Horn K, Sagan SA, Winger RC, et al. CNS accumulation of regulatoryB cells is VLA-4-dependent. Neurol Neuroimmunol Neuroinflamm 2016;3:e212. doi:10.1212/NXI.0000000000000212.

24. Glatigny S, Wagner CA, Bettelli E. Correction: cutting edge: integrin alpha4 is re-quired for regulatory B cell control of experimental autoimmune encephalomyelitis.J Immunol 2017;198:3756.

25. Glatigny S, Duhen R, Oukka M, Bettelli E. Cutting edge: loss of alpha4 integrinexpression differentially affects the homing of Th1 and Th17 cells. J Immunol 2011;187:6176–6179.

26. Lyons JA, SanM, HappMP, Cross AH. B cells are critical to induction of experimentalallergic encephalomyelitis by protein but not by a short encephalitogenic peptide. EurJ Immunol 1999;29:3432–3439.

27. Magliozzi R, Howell O, Vora A, et al. Meningeal B-cell follicles in secondary pro-gressive multiple sclerosis associate with early onset of disease and severe corticalpathology. Brain 2007;130:1089–1104.

Appendix Authors

Name Location Role Contribution

Rehana Z.Hussain,MSc

University of TexasSouthwesternMedical Center,Dallas, TX

Author Designed andconceptualized thestudy; analyzed the data;and drafted themanuscript forintellectual content

Petra C.Cravens,PhD

University of TexasSouthwesternMedical Center,Dallas, TX

Author Designed andconceptualized thestudy; analyzed the data;and drafted themanuscript forintellectual content

William A.Miller-Little, BA

University of TexasSouthwesternMedical Center,Dallas, TX

Author Acquired the data;analyzed the data; andrevised the manuscript

RichardDoelger,MSc

University of TexasSouthwesternMedical Center,Dallas, TX

Author Acquired the data;analyzed the data; andrevised the manuscript

ValerieGranados,PhD

University of TexasSouthwesternMedical Center,Dallas, TX

Author Acquired the data andanalyzed the data

EmilyHerndon,MD

University of TexasSouthwesternMedical Center,Dallas, TX

Author Interpreted the data andrevised the manuscriptfor intellectual content

Darin T.Okuda,MD

University of TexasSouthwesternMedical Center,Dallas, TX

Author Interpreted the data andrevised the manuscriptfor intellectual content

Todd N.Eagar, PhD

Houston MethodistHospital, Houston,TX

Author Designed andconceptualized thestudy; analyzed the data;and drafted themanuscript forintellectual content

OlafStuve, MD,PhD

University of TexasSouthwesternMedical Center,Dallas, TX

Author Designed andconceptualized thestudy; analyzed the data;and drafted themanuscript forintellectual content

10 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 34: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

ARTICLE OPEN ACCESS CLASS OF EVIDENCE

Pilot study of a ketogenic diet in relapsing-remitting MSJ. Nicholas Brenton,MD, BrendaBanwell, MD, A.G. Christina Bergqvist, MD,Diana Lehner-Gulotta, RDN, CNSC,

Lauren Gampper, BA, Emily Leytham, BA, Rachael Coleman, MPH, and Myla D. Goldman, MD, MSc

Neurol Neuroimmunol Neuroinflamm 2019;6:e565. doi:10.1212/NXI.0000000000000565

Correspondence

Dr. Brenton

[email protected]

AbstractObjectiveTo assess the safety and tolerability of a modified Atkins diet (KDMAD), a type of ketogenic diet(KD), in subjects with relapsing MS while exploring potential benefits of KDs in MS.

MethodsTwenty subjects with relapsing MS enrolled into a 6-month, single-arm, open-label study of theKDMAD. Adherence to KDMAD was objectively monitored by daily urine ketone testing. Fatigueand depression scores and fasting adipokines were obtained at baseline and on diet. Brain MRIwas obtained at baseline and 6months. Intention to treat was used for primary data analysis, anda per-protocol approach was used for secondary analysis.

ResultsNo subject experienced worsening disease on diet. Nineteen subjects (95%) adhered to KDMAD

for 3 months and 15 (75%) adhered for 6 months. Anthropometric improvements were notedon KDMAD, with reductions in body mass index and total fat mass (p < 0.0001). Fatigue (p =0.002) and depression scores (p = 0.003) were improved. Serologic leptin was significantlylower at 3 months (p < 0.0001) on diet.

ConclusionsKDMAD is safe, feasible to study, and well tolerated in subjects with relapsing MS. KDMAD

improves fatigue and depression while also promoting weight loss and reducing serologicproinflammatory adipokines.

Classification of evidenceThe study is rated Class IV because of the absence of a non-KD control group.

MORE ONLINE

Class of EvidenceCriteria for ratingtherapeutic and diagnosticstudies

NPub.org/coe

From the Division of Pediatric Neurology (J.N.B., D.L.-G.), Department of Neurology, University of Virginia, Charlottesville; Department of Pediatric Neurology (B.B., A.G.C.B.),University of Pennsylvania/Children’s Hospital of Philadelphia; and Department of Neurology (L.G., E.L., R.C., M.D.G.), University of Virginia, Charlottesville.

Funding information and disclosures are provided at the end of the article. Full disclosure form information provided by the authors is available with the full text of this article atNeurology.org/NN.

The Article Processing Charge was funded by the ziMS Foundation.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 35: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Obesity is a recognized risk factor for MS,1 and dietary intakeis one potentially modifiable environmental contributor toMS that has garnered much interest from people livingwith MS.2 Furthermore, diet—as a modifier of diseaseprogression—is supported by recent work providing an in-dependent, epidemiologic association between dietary habitsand level of sustained disability.3 Moreover, diet quality andan active lifestyle were also associated with reduced levels offatigue, depression, pain, and cognitive impairment.3 Severaldiets have been popularized within the MS community, butformal studies with scientific evidence are lacking to supportthe use of a particular diet therapy.

Ketogenic diets (KDs) are high-fat, low-carbohydrate dietsthat mimic a fasting state. KDs create a metabolic shift fromglycolytic energy production toward oxidative phosphoryla-tion energetics by using fatty acids as a primary source ofenergy. As these fatty acids undergo beta-oxidation, ketonesare produced. This increase in oxidative phosphorylationcoupled with ketone production modifies the tricarboxylicacid cycle to limit reactive oxygen species generation. In ad-dition, ketone bodies transported across the blood-brainbarrier upregulate antioxidant pathway genes (particularly viathe Nrf2 pathway) and boost energy production in braintissue.4–7 Most importantly, KDs have been shown in multi-ple, recent animal and human studies to attenuate biomarkersof inflammation within the blood and CSF.8–12 One mecha-nism for its anti-inflammatory effects is via production of beta-hydroxybutyrate (BHB), a ketone body that blocks the NLRP3inflammasome and reduces proinflammatory cytokine pro-duction (notably, IL-1β) by human monocytes.8,9,13–15 Pre-clinical studies have assessed the benefits of the KD in theexperimental autoimmune encephalitis (EAE) mouse model ofMS. EAE mice fed a KD experienced several positive effectsincluding reversed motor disability, improved spatial learningand memory, increased hippocampal volumes, and remyelina-tion of periventricular lesions. These benefits were associatedwith suppressed production of inflammatory cytokines and en-hanced neuronal repair.10

Collectively, animal model data are supportive, but in-sufficient, to suggest that a KD may modify an individual’simmune response, thereby altering MS disease course withina human population. KDs, in particular, may alter key aspectsof MS pathogenesis via ketosis-induced upregulation of an-tioxidant pathways, reduced effector cell immunity, and en-hanced CNS bioenergetics, which may provide an alternativeenergy source to vulnerable neurons. In light of these prom-ising animal data, our group sought to assess the safety and

tolerability of a KD in a relapsing-remitting MS (RRMS)population.

MethodsSubjects/recruitmentTwenty subjects with a diagnosis of RRMS per McDonald2010 criteria16 were enrolled into a 6-month, single-arm,uncontrolled, open-label pilot study examining the feasibility,safety, tolerability, and efficacy of a type of KD—the modifiedAtkins diet (KDMAD). This study provides Class IV evidenceof this primary research aim. To be eligible for inclusion,subjects had to be aged between 15 and 50 years, seen andtreated in the University of Virginia (UVA) MS Clinic, andmaintained on disease-modifying therapy. In addition, sub-jects had to exhibit disease stability, which was defined by both(1) the absence of clinical relapse and (2) absence of new/enlarging T2 or gadolinium-enhancing lesions on MRI for atleast 12 months before study enrollment. Subjects were ex-cluded from the study if they had an Expanded DisabilityStatus Scale (EDSS) score of >6.0, history of known cardio-vascular or renal disease, currently were pregnant or wereplanning pregnancy, or had a body mass index (BMI) of <20kg/m2 for adults or <10th percentile17 for those aged youngerthan 18 years.

Eligible study subjects were identified through our MS clinicaldatabase and contacted via a mailed recruiting advertise-ment. The mailing was sent to 150 subjects, meeting theeligibility criteria for this study. When a subject expressedinterest, they were provided our study objectives, param-eters, diet specifics, and study design. Those who remainedcommitted to the study completed a screening assessmentto confirm eligibility.

Standard protocol approvals, registrations, andsubject consentsThe study was approved by the University of Virginia In-stitutional Review Board for Health Sciences Research, and allsubjects provided informed consent (and assent when appli-cable) before conducting any study-related procedures.

Study proceduresAt the baseline visit, demographic information was collected,including age, race/ethnicity, sex, weight, height, and waistcircumference. Medical history and history of previous/current diet attempt(s) were reviewed and documented byboth subject report and medical records. MS-related datacollection included history of relapses, past/current disease-modifying therapies, and disease duration.

GlossaryBDI = Beck Depression Inventory; BHB = beta-hydroxybutyrate; BMI = body mass index; EAE = experimental autoimmuneencephalitis; EDSS = Expanded Disability Status Scale;KD = ketogenic diet;MIFS =Modified Fatigue Impact Scale; PASAT =Paced Auditory Serial Addition Test; RRMS = relapsing-remitting MS; SDMT = Symbol Digit Modalities Test.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 36: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Anthropometric measuresBody size and composition were used as indicators of growthand nutritional status. Weight was measured on an electronicScaleTronix digital scale (accurate to 0.1 kg), stature (accurateto 0.1 cm) measured on a stadiometer, and waist circumfer-ence measured with a nonstretchable plastic tape. All meas-urements were recorded at baseline and at each follow-upvisit. As a measure of body composition, subjects underwentbody mass assessment via air displacement plethysmography(BOD POD; COSMED, Rome, Italy) that uses whole-bodydensitometry to determine body composition (fat vs lean).BOD POD measurements were obtained at baseline and 6months on diet to determine body composition change sec-ondary to our diet intervention.

Diet interventionAt the baseline visit, subjects met with our study dietitian(D.L.G.), with specialized expertise in KDs, for nutritionpreassessment before diet start. The dietitian provideda personalized educational session on initiation and main-tenance of a KDMAD. The KDMAD is a type of KD that allowsmore flexibility and relatively high levels of ketosis byrestricting carbohydrates to <20 g per day and encouraginggreater fat intake, with virtually all subjects reaching somedegree of ketosis.18 On completion of baseline teaching, thedietitian provided contact information so that subjects coulduse her as a resource outside of study visits. The dietitian metwith study subjects at 1-month, 3-month, and 6-monthfollow-up study visits. At the end of the 6-month diet trial,the dietitian provided all subjects with a plan to wean safelyfrom KDMAD.

MRIAfter enrollment, all subjects underwent a protocolized3-Tesla clinical MRI of the brain with contrast to evaluate thenumber of FLAIR/T2 hyperintensities. Imaging was repeatedat 6 months on diet to assess for the number of new or newlyenlarged T2 lesions and/or contrast-enhancing lesions(CELs). Imaging studies for all subjects were performed onthe same MRI research scanner. Imaging studies werereviewed and interpreted by a trained neuroradiologist. Thesestudies were also independently reviewed by the study pri-mary investigator (J.N.B.).

Clinical and subject-reportedoutcome measuresAt the baseline visit, all subjects had a formal EDSS assess-ment completed by a single investigator (J.N.B.—aNeurostatus-certified examiner). In addition, all subjects un-derwent testing with the Multiple Sclerosis FunctionalComposite—a composite disability measure of ambulation(timed 25-foot walk), arm function (Nine-Hole Peg Test),and cognition (Paced Auditory Serial Addition Test[PASAT]). Subjects completed the Symbol Digit ModalitiesTest (SDMT), which serves as an objective measure ofvisuospatial processing speed. Finally, all subjects underwent

a 6-minute walk (6MW) assessment, which is a validatedoutcome measure of physical disability that measures the totaldistance walked in a timed 6 minutes.

Subject-reported fatigue was measured by the Modified Fa-tigue Impact Scale (MFIS), a 21-question survey that providesan assessment of the effects of fatigue in terms of physical,cognitive, and psychosocial functioning. Subject-reporteddepression was assessed by the Beck Depression Inventory(BDI) 1A, a 21-item inventory that measures attitudes andsymptoms of depression.

Blood studiesSubjects underwent fasting, early morning serum samplingbefore diet start, and at 3 and 6 months on diet. The labo-ratory evaluation included blood glucose, insulin, hemoglobinA1c, lipid profile (low-density lipoprotein [LDL], high-densitylipoprotein, triglycerides, and cholesterol), 25-hydroxyvitaminD, bicarbonate, and liver function. The majority of subjectswere already taking supplemental vitamin D at various dosesrecommended from their outpatient neurologist. Vitamin Dsupplementation was continued as advised, and doses werenot altered during the course of this study. Finally, fastingserologic adipokines (leptin and adiponectin) were assessedby ELISA, manufactured by Millipore.

Once enrolled, subjects were evaluated at baseline and 1, 3,and 6 months on diet. There was a final follow-up visit 3months after study completion to assess current dietaryhabits, subject-reported outcomes, and anthropometricmeasures. As an objective marker of diet compliance, subjectswere provided with urine ketone test strips for 6 months. Thesubject was advised to urinate into a cup and dip the test stripinto the urine. The test strip then changes color to reflect thelevel of ketosis (negative, trace, small, moderate, large, andvery large). All subjects were required to take daily, datedpictures of their test strips and email these pictures as ob-jective evidence of ketosis (i.e., “adherence”) throughout the6-month study period. Subjects were considered adherent tothe diet if they achieved some level of ketosis for that day (e.g.,trace ketones or higher). A negative ketone strip or a daywithout dated, photographic evidence of ketosis were con-sidered nonadherent days. Compliance with this diet wasdefined a priori as >85% compliant days out of the 6-monthstudy period. Any subject who was lost to follow-up before the6-month visit was counted as a nonadherent subject. Oncea subject was lost to follow-up, subsequent missing data wereexcluded in the clinical and laboratory data reporting.

Statistical analysisAll statistical analyses were conducted using SAS 9.4 software.This was a pilot study with no previous sample size calculationand thus no correction for multiple comparisons. For primarydata analysis, an intention-to-treat approach was used; thus,the data from all subjects completing the 3- and 6-month timepoints were included. As a secondary sensitivity analysis,a per-protocol approach was used to study the group of those

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 37: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

individuals who successfully met study compliance require-ments for the diet intervention. All descriptive statistics werecalculated and reported using standard appropriate statistics(e.g., means, frequencies, and t-statistic). Paired t tests and χ2

tests were used as appropriate for continuous and categoricalvariables to provide in-group comparisons from baseline to 3and 6 months on diet. Spearman correlation coefficients wereused to assess the relationship between change in leptin andBMI. A 2-sided p value of <0.05 was defined as statisticallysignificant.

Data availabilityAny data not published within the article are available, and theanonymized data will be shared by request from any qualifiedinvestigator.

ResultsStudy subject demographicsFrom a mailing sent to 150 eligible subjects, 51 subjectsresponded with interest. From these 51 responders, the first20 to respond were enrolled into the study. A single subjectdeclined enrollment on hearing the study details, and a singlesubject was excluded because of a BMI of <20 kg/m2 at thetime of enrollment. The remainder of interested subjects (n =29) were placed on a wait list for future study opportunities.

Baseline study subject demographics are detailed in table 1.The cohort’s median age was 38 years (range: 15–50 years)with an average disease duration of 9.5 years. Ninety-fivepercent of the cohort was either overweight or obese by theCenters for Disease Control and Prevention criteria,17 andmore than half of the subjects had attempted dietary changesbefore this study, specifically for the purposes of improv-ing MS.

Adherence to KDMAD

Among the 20 subjects, 2 were lost to follow-up. Before beinglost to follow-up, the first subject had complied with 1 monthof diet, and the second had complied with 3months of diet. Ofthe remaining 18 subjects, 15 (75%) met the study’s criteriafor “adherence” as defined by demonstrating urinary ketosis forat least 85% of total days from the 6-month diet study. Threesubjects did not meet the adherent criteria but were able toobjectively comply with a KDMAD for a median of 3.5 months.

Safety of KDMAD in relapsing MSA single subject reported sensory symptoms that raised con-cern for a clinical relapse; however, after obtaining a thoroughhistory and repeat imaging of the brain and spinal cord (thatdemonstrated no new or enlarging lesions), these symptomswere considered unlikely to represent an MS relapse by thetreating neurologist. One-third (n = 6) of subjects denied anyside effects on KDMAD. The most common side effectsreported by the remaining subjects included intermittentconstipation (n = 5), menstrual irregularities (n = 4), anddiarrhea (n = 3).

As an additional safety metric, all subjects had an MRI of thebrain with contrast performed before study start and thenagain at 6 months from diet start (n = 18). No subject wasfound to have new or enlarging FLAIR/T2 lesions at 6months. Likewise, no subject was found to have new CELs atthe 6-month scan.

Anthropometric outcomesAnthropometric outcome measures are reported in table 2.Using the intention-to-treat approach, subjects experienceda mean reduction of 3 points in BMI, which was maintainedat 3- and 6-month assessments (p < 0.0001). From the studycohort, 7 subjects (all of whom met the study compliancecriteria) changed Centers for Disease Control and Pre-vention BMI categories between baseline assessment and 6months on diet: obese to normal weight (n = 2), obese tooverweight (n = 3), and overweight to normal weight (n =2). Waist circumference also significantly decreased on dietat 3 (p ≤ 0.0001) and 6 months (p = 0.0004). Over 6

Table 1 Clinical characteristics of study subjects

Age, median years (range) 38 (15–50)

Sex

Female, n (%) 17 (85)

Race

White, n (%) 18 (90)

African American, n (%) 2 (10)

BMI category

Normal weight, n (%) 1 (5)

Overweight, n (%) 5 (25)

Obese, n (%) 14 (70)

MS disease duration, mean years (±SD) 9.5 ± 5.4

Time since last clinical relapse, mean years (±SD) 6.4 ± 3.3

Baseline EDSS score, median score (range) 2.0 (1.0–4.0)

No. of previous disease-modifying therapiesattempted, mean (±SD)

2.2 ± 1.1

Current disease-modifying therapy, n (%)

Interferon beta-1a 4 (20)

Glatiramer acetate 3 (15)

Teriflunomide 2 (10)

Dimethyl fumarate 4 (20)

Fingolimod 3 (15)

Natalizumab 3 (15)

Rituximab 1 (5)

Subjects with previous diet attempt(s), n (%) 13 (65)

BMI = body mass index; EDSS = Expanded Disability Status Scale.

4 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 38: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

months, the resting metabolic rate, as measured by BODPOD, was significantly reduced on diet (p = 0.005).

Subject-reported and clinicaloutcome measuresSignificant improvements in subject-reported total fatiguescores at 3 months (p = 0.0005) and 6 months (p = 0.002)were noted. Benefits were seen across all 3 MFIS subscales:physical, cognitive, and psychosocial. Before diet, 3 subjectsmet the criteria for mood disturbance or depression on theBDI. Only 1 subject met the BDI criteria for mood dis-turbance at 3 and 6 months on diet. Significant reductionsin raw BDI scores were noted at 3 (p = 0.0002) and 6months (p = 0.003). Change in fatigue and depressionscores from baseline to 3 and 6 months on diet is illustratedin figure 1.

EDSS scores significantly improved (p < 0.0001) betweenbaseline and 6 months on diet, primarily because of

improvements on bowel/bladder and sensory functionalsystems. In addition, there was no worsening on measures oflower limb functioning (6-minute walk, T25FW), upper limbfunctioning (9HPT), or cognition/attention (PASAT,SDMT). In fact, a significant improvement was noted on the9HPT with the nondominant hand (p = 0.006) at 6 months.

Laboratory outcomesLaboratory outcome data are presented in table 3. No subjecthad clinically relevant or significant changes in electrolytes,bicarbonate levels, liver function testing, or glucose levelsfrom baseline to 3 or 6 months on diet. Although a trendtoward higher 25-hydroxyvitamin D levels on diet was noted,this was not statistically significant. Fasting insulin and he-moglobin A1c levels were significantly decreased at both 3and 6 months on diet. Lipid profiles showed a significant in-crease in LDL and cholesterol at 3 months, although thisincrease declined at 6 months and was no longer significant. Asignificant decrease in triglyceride levels was noted at 6

Table 2 Intention-to-treat analysis of outcome measures pre- and post-KDMAD intervention

Baseline valuesΔ change at3 months (n = 19) p Value

Δ change at6 months (n = 18) p Value

Anthropometric measures

Body mass index (BMI) 34.1 ± 6.9 −3.0 ± 1.6 <0.0001 −3.0 ± 2.2 <0.0001

Waist circumference, cm 106.4 ± 14.0 −8.5 ± 4.8 <0.0001 −9.7 ± 8.9 0.0004

BOD POD assessment

Fat mass, kg 42.5 ± 16.6 — — −7.0 ± 5.3 <0.0001

Fat-free mass, kg 51.1 ± 10.8 — — −1.2 ± 2.4 0.048

Resting metabolic rate, kcal/d 1,510.3 ± 309.5 — — −57.8 ± 71.3 0.005

Subject-reported outcomes

Beck Depression Inventory 7.2 ± 5.8 −3.4 ± 3.2 0.0002 −2.9 ± 3.6 0.003

Total Modified Fatigue Impact Scale 34.1 ± 17.1 −12.9 ± 13.2 0.0005 −12.3 ± 14.4 0.002

Physical subscale 15.8 ± 8.9 −6.6 ± 6.5 0.0003 −5.8 ± 8.6 0.01

Cognitive subscale 15.8 ± 8.3 −5.3 ± 6.2 0.002 −5.6 ± 5.8 0.0008

Social subscale 2.4 ± 1.7 −1.0 ± 1.3 0.003 −0.9 ± 1.8 0.04

Clinical outcome measures

Expanded Disability Status Scale 2.2 ± 0.9 — — −0.8 ± 0.6 <0.0001

MS Functional Composite

Timed 25-foot walk, s 5.5 ± 2.3 — — +0.3 ± 0.8 0.15

9-Hole Peg Test dominant, s 20.6 ± 3.5 — — −0.4 ± 1.7 0.30

9-Hole Peg Test nondominant, s 22.9 ± 8.3 — — −1.0 ± 1.3 0.006

Paced Auditory Serial Addition Test (number correct) 45.5 ± 9.8 — — +1.2 ± 5.3 0.36

Symbol Digit Modalities Test (number correct) 59.5 ± 11.1 — — +1.1 ± 5.8 0.45

6-minute walk, feet 1,652 ± 397 — — +56 ± 123 0.07

All results within this table represent mean ± SDs. Δ change = mean 3- or 6-month value − mean baseline value. Bolded p values are statistically significant.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 5

Page 39: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

months on diet. Notably, a significant decrease in free carni-tine levels was seen at both 3 and 6 months on KDMAD.

Fasting serologic leptin was significantly decreased at 3 monthson KDMAD (p < 0.0001) and trended toward significance at 6months when using intention-to-treat analysis. The relative re-duction in leptin was not significantly correlated with the relativereduction in BMI (r = 0.46 n = 19; p = 0.06). There was

a nonsignificant trend for increasing serologic adiponectin levelson diet. The change in serologic leptin and adiponectin frombaseline to 3 and 6 months on diet is depicted in figure 2.

Per-protocol analysis for compliant subjectsData were secondarily analyzed using a per-protocol approach(table e-1, links.lww.com/NXI/A110 and table e-2, links.lww.com/NXI/A111). When using this approach, the measures

Figure 1 Subject-reported outcomes for subjects compliant to KDMAD at 3 (n = 19) and 6 months (n = 15) for depression

(A) Depression scores as a function of time adhering to a modified KD (as assessed by the Beck Depression Inventory) and (B) fatigue levels as a function oftime adhering to a modified KD fatigue (as assessed by the Modified Fatigue Impact Scale). Boxplots demonstrate the median and interquartile range.Whiskers represent the range. KD = ketogenic diet.

Table 3 Intention-to-treat analysis of laboratory measures pre- and post-KDMAD intervention

BaselineΔ change at3 months (n = 19) p Value

Δ change at6 months (n = 17) p Value

Insulin resistance

Insulin, uIU/mL 15.3 ± 10.6 −6.8 ± 9.1 0.005 −5.7 ± 9.1 0.02

Hemoglobin A1c, % 5.4 ± 0.6 −0.2 ± 0.3 0.005 −0.13 ± 0.26 0.049

Lipid profiles

Triglycerides, mg/dL 125 ± 47 −19 ± 46 0.09 −21.3 ± 38.5 0.03

Low-density lipoprotein, mg/dL 135 ± 43 +22 ± 28 0.003 +11.9 ± 31.4 0.13

High-density lipoprotein, mg/dL 54 ± 12 −1 ± 10 0.58 +2.7 ± 11.0 0.31

Cholesterol, mg/dL 210 ± 46 +17 ± 31 0.03 +11.1 ± 35.1 0.20

25-Hydroxyvitamin D, ng/mL 44 ± 20 +7 ± 16 0.06 +6.1 ± 12.6 0.06

Free carnitine, nmol/mL 36 ± 9 −5 ± 7 0.004 −8.9 ± 8.3 0.0003

Adipokines

Leptin, ng/mL 22.9 ± 11.8 −8.9 ± 6.8 <0.0001 −4.9 ± 9.6 0.06

Adiponectin, mcg/mL 10.1 ± 4.3 +0.7 ± 2.7 0.25 +1.4 ± 3.6 0.12

All results within this table represent mean ± SDs. Δ change = mean 3- or 6-month value − mean baseline value. Bolded p values are statistically significant.

6 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 40: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

demonstrating significance with intention-to-treat analysisremained strongly significant. Furthermore, serologic leptinwas significantly different from baseline at both 3 (p < 0.0001)and 6 months (p = 0.03).

Post-diet study outcomesFifteen subjects (75%) returned for the final 3-month visitfollowing completion of the diet study. The remaining 5subjects were lost to follow-up: 3 of these subjects met thecompliance criteria for the 6-month trial and 2 were non-compliant subjects. Of the 15 subjects who returned for follow-up, 2 did not meet the compliance criteria for the 6-month dietprotocol. From all subjects attending this post-study follow-upvisit, 4 (27%) continued to adhere to a strict KDMAD, and 9subjects (60%) reported eating a lower carbohydrate diet but notadhering as strictly to the tenets of the KDMAD. The remaining 2subjects (13%) reported returning to their pre-study dietaryhabits. At this follow-up visit, subjects continued to have signif-icant weight lost from baseline (prediet) in addition to persistentbenefits on fatigue (table e-3, links.lww.com/NXI/A112).

DiscussionDietary interventions are rapidly gaining popularity within theMS community,2 and recently, patterns of dietary intake havebeen directly linked with disability in MS.3 In particular, dietsthat mimic a fasting state (i.e., KDs) have potential for mul-timodal benefits in an MS population, given the documentedproperties of these diets in reducing serologic inflammation,providing a more efficient central energy source, and byupregulation of antioxidant pathways. Given these potentialbenefits on MS disease, our group took the first steps toward

elucidating the benefits of KDs within a human MS cohort bystudying the safety and tolerability of a KDMAD in disease-stable subjects with RRMS. Our data support the short-termsafety of these diets for patients with MS in terms of clinical,radiologic, and serologic measurements. Furthermore, ourdata provide evidence on the feasibility of studying these dietswithin an MS cohort, with 95% able to comply for 3 monthsand 3 quarters of all subjects able to comply for 6 months ofdiet intervention. Of note, compliance in our study exceedscompliance estimates from the current epilepsy literature.18–20

The majority of our study population maintained their weightloss and several continued to apply KDMAD principles towardfood selection in the 3 months following study completion.

From all safety laboratory parameters followed, there wasa noted transient increase in LDL and total cholesterol thatstabilized by 6 months on diet. Subjects benefitted froma reduction in triglyceride levels at 6 months on diet therapy.Free carnitine levels significantly declined on diet at 3 and 6months and are a known side effect of a KD. We recommendfractionated carnitine levels be followed on KDs and sup-plementation be initiated if deficiencies arise.

Based on our study in a human MS population, the KDMAD

provided significant benefits to subjects throughout thestudy with a significant reduction in several anthropometricmeasures—including BMI, fat mass, waist circumference, andresting metabolic rate. Subject-reported measures of fatigueand depression also improved on diet. Although the study wasnot designed nor powered to evaluate the effect of diet onMS,we did note a significant decrease in EDSS scores at 6 monthson diet—secondary to improved sensory and bowel/bladder

Figure 2 Change in baseline fasting serologic adipokine levels for compliant subjects at 3 (n = 19) and 6months (n = 15) onKDMAD

(A) Serologic leptin as function of time adhering to modified KD and (B) serologic adiponectin as a function of time adhering to modified KD. Boxplotsdemonstrate the median and interquartile range. Whiskers represent the range. KD = ketogenic diet.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 7

Page 41: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

symptoms. Our results confirm a significant relationship be-tween KD and declining serum leptin levels at 3 and 6 monthsfor those subjects adhering to the diet. Importantly, we didnot see a strong correlation between BMI and leptin levels,suggesting that the KD’s influence on leptin regulation islikely multimodal and not just secondary to its ability to in-duce a reduction in adiposity. Given leptin’s role as a proin-flammatory adipokine, this suggests one immunologicpathway in which these diets may benefit MS disease.

A limitation of this study includes the lack of a control group.In addition, our study likely recruited a motivated population(including a large proportion of overweight/obese subjects),which could affect the compliance rates reported. By includinga largely overweight/obese population, our study results maynot be generalizable to patients with MS of normal weight. Inaddition, our cohort likely exhibited heightened adherencesecondary to our compliance approach, which required sub-jects to provide daily photographic evidence of urinaryketones. On the other hand, this more objective complianceapproach represents a specific strength of this study—compared with previous diet studies that rely primarily ondietary recall. Although serologic BHB levels provide a moreaccurate reflection of ketosis,21 this method of compliancetesting is not feasible for daily assessment primarily secondaryto the high cost of blood ketone testing strips in addition tothe discomfort associated with daily fingersticks. Finally, ourinclusion of only subjects with relapsing, clinically and ra-diographically stable MS prohibits generalization to pro-gressive MS subtypes or to patients with activelyrelapsing MS.

Our results provide evidence supporting the safety and tol-erability of KDs in a relapsing MS population. These dietsappear to provide clinical benefits on common MS comor-bidities such as fatigue and depression. This early-phase studywas not designed to study the efficacy of a KDMAD onMS, andthus, future next steps include a prospective randomized, case-control study to define the effect of KDMAD on diseasecontrol.

Classification of evidenceThis study provides Class IV evidence that the modifiedAtkins diet is safe and well tolerated in subjects with relapsingMS, with benefits on fatigue, depression, and serologicproinflammatory adipokine levels.

AcknowledgmentThe authors thank Denise Bruen, NP, for her assistance in therecruitment of our study population.

Study fundingThis study was funded through private foundational fundingprovided by the ziMS Foundation.

DisclosureJ.N. Brenton has served as a consultant for Novartis. B.Banwell has served as a consultant and speaker for Novartis.She also serves as a nonremunerated advisor to BiogenIDEC, Sanofi, and Teva Neuroscience. She has givena lecture for Medscape. A.G.C. Bergqvist has served asa consultant for Nutricia North America, Dannon, Vitaflo,and the Charlie Foundation. D. Lehner-Gulotta, L.Gampper, E. Leytham, and R. Coleman report no dis-closures. M.D. Goldman has served as a consultant forADAMAS, Celgene, ENDECE, EMD Serono, NovartisPharmaceuticals, Sanofi Genzyme, and Teva Pharmaceut-icals. She has received research funding from Biogen Idec,Novartis Pharmaceuticals, National MS Society, andMedDay Pharmaceuticals. Full disclosure form informationprovided by the authors is available with the full text of thisarticle at Neurology.org/NN.

Publication historyReceived by Neurology: Neuroimmunology & NeuroinflammationDecember 9, 2018. Accepted in final form March 5, 2019.

Appendix Authors

Name Location Role Contribution

J. NicholasBrenton,MD

University ofVirginia,Charlottesville

Author Designed andconceptualized the study;analyzed the data; majorrole in the acquisition ofdata; and drafted themanuscript for intellectualcontent

BrendaBanwell,MD

Children’sHospital ofPhiladelphia,Philadelphia

Author Helped design andconceptualize the studyand revised themanuscript forintellectual content

A.G.ChristinaBergqvist,MD

Children’sHospital ofPhiladelphia,Philadelphia

Author Helped design andconceptualize the study andrevised the manuscript forintellectual content

DianaLehner-Gulotta,RDN

University ofVirginia,Charlottesville

Author Helped design andconceptualize the study;major role in the acquisitionof data; and revised themanuscript for intellectualcontent

LaurenGampper

University ofVirginia,Charlottesville

Author Major role in the acquisitionof data

EmilyLeytham

University ofVirginia,Charlottesville

Author Major role in the acquisitionof data

RachaelColeman

University ofVirginia,Charlottesville

Author Major role in the acquisitionof data

Myla DGoldman,MD

University ofVirginia,Charlottesville

Author Helped design andconceptualize thestudy and revised themanuscript forintellectual content

8 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 42: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

References1. Hedstrom AK, Olsson T, Alfredsson L. High body mass index before age 20 is

associated with increased risk for multiple sclerosis in both men and women. MultScler 2012;18:1334–1336.

2. Brenton JN, GoldmanMD. A study of dietary modification: perceptions and attitudesof patients with multiple sclerosis. Mult Scler Relat Disord 2016;8:54–57.

3. Fitzgerald KC, Tyry T, Salter A, et al. Diet quality is associated with disability andsymptom severity in multiple sclerosis. Neurology 2018;90:e1–e11.

4. Milder JB, Liang LP, Patel M. Acute oxidative stress and systemic Nrf2 activation bythe ketogenic diet. Neurobiol Dis 2010;40:238–244.

5. Sullivan PG, Rippy NA, Dorenbos K, Concepcion RC, Agarwal AK, Rho JM. Theketogenic diet increases mitochondrial uncoupling protein levels and activity. AnnNeurol 2004;55:576–580.

6. Bough KJ, Rho JM. Anticonvulsant mechanisms of the ketogenic diet. Epilepsia 2007;48:43–58.

7. Achanta LB, Rae CD. Beta-hydroxybutyrate in the brain: one molecule, multiplemechanisms. Neurochem Res 2017;42:35–49.

8. Ruskin DN, Kawamura M, Masino SA. Reduced pain and inflammation in juvenileand adult rats fed a ketogenic diet. PLoS One 2009;4:e8349.

9. Dupuis N, Curatolo N, Benoist JF, Auvin S. Ketogenic diet exhibits anti-inflammatoryproperties. Epilepsia 2015;56:e95–e98.

10. Kim DY, Hao J, Liu R, Turner G, Shi FD, Rho JM. Inflammation-mediated memorydysfunction and effects of a ketogenic diet in a murine model of multiple sclerosis.PLoS One 2012;7:e35476.

11. Ni FF, Li CR, Liao JX, et al. The effects of ketogenic diet on the Th17/Treg cellsimbalance in patients with intractable childhood epilepsy. Seizure 2016;38:17–22.

12. Shen Y, Kapfhamer D, Minnella AM, et al. Bioenergetic state regulates innate in-flammatory responses through the transcriptional co-repressor CtBP. Nat Commun2017;8:624.

13. Youm YH, Nguyen KY, Grant RW, et al. The ketone metabolite beta-hydroxybutyrateblocks NLRP3 inflammasome-mediated inflammatory disease. Nat Med 2015;21:263–269.

14. Yang X, Cheng B. Neuroprotective and anti-inflammatory activities of ketogenic dieton MPTP-induced neurotoxicity. J Mol Neurosci 2010;42:145–153.

15. Goldberg EL, Asher JL, Molony RD, et al. Beta-hydroxybutyrate deactivates neu-trophil NLRP3 inflammasome to relieve gout flares. Cell Rep 2017;18:2077–2087.

16. Polman CH, Reingold SC, Banwell B, et al. Diagnostic criteria for multiple sclerosis:2010 revisions to the McDonald criteria. Ann Neurol 2011;69:292–302.

17. Center for Disease Control. Healthy Weight. 2015. Available at: www.cdc.gov/healthyweight/assessing/bmi/childrens_bmi/about_childrens_bmi.html. AccessedMay 19, 2016.

18. Kossoff EH, McGrogan JR, Bluml RM, Pillas DJ, Rubenstein JE, Vining EP. Amodified Atkins diet is effective for the treatment of intractable pediatric epilepsy.Epilepsia 2006;47:421–424.

19. Cervenka MC, Henry BJ, Felton EA, Patton K, Kossoff EH. Establishing an adultepilepsy diet center: experience, efficacy and challenges. Epilepsy Behav 2016;58:61–68.

20. Chen W, Kossoff EH. Long-term follow-up of children treated with the modifiedAtkins diet. J Child Neurol 2012;27:754–758.

21. Gilbert DL, Pyzik PL, Freeman JM. The ketogenic diet: seizure control correlatesbetter with serum beta-hydroxybutyrate than with urine ketones. J Child Neurol 2000;15:787–790.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 9

Page 43: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

ARTICLE OPEN ACCESS

Different MRI patterns in MS worsening afterstopping fingolimodCaterina Lapucci, MD, Damiano Baroncini, MD, Maria Cellerino, MD, Giacomo Boffa, MD, Ilaria Callegari, MD,

Matteo Pardini, MD, PhD, Giovanni Novi, MD, Maria Pia Sormani, PhD, Giovanni Luigi Mancardi, MD,

Angelo Ghezzi, MD, Mauro Zaffaroni, MD, Antonio Uccelli, MD, Matilde Inglese, MD, PhD, and

Luca Roccatagliata, MD, PhD

Neurol Neuroimmunol Neuroinflamm 2019;6:e566. doi:10.1212/NXI.0000000000000566

Correspondence

Dr. Lapucci

[email protected]

AbstractObjectiveTo analyze MRI images in patients with MS who experienced worsening of neurologic status(WNS) after stopping fingolimod (FTY).

MethodsIn this retrospective study, demographic, clinical, and radiologic data of patients with MS whoexperienced WNS after stopping FTY were retrospectively collected. We introduced the“δExpanded Disability Status Scale (EDSS)-ratio” to identify patients who, after FTY with-drawal, showed an inflammatory flare-up exceeding the highest lifetime disease activity level.Patients with δEDSS-ratio > 1 were enrolled in the study.

ResultsEight patients were identified. The mean (SD) age of the 8 (7 female) patients was 35.3 (4.9)years. The mean FTY treatment duration was 3.1 (0.8) years. The mean FTYdiscontinuation–WNS interval was 4 (0.9) months. The 4 patients with δEDSS-ratio ≥ 2developed severe monophasic WNS (EDSS score above 8.5), characterized by clinical featuresand MRI findings not typical of MS, which we classified as “tumefactive demyelination pattern”(TDL) and “Punctuated pattern” (PL). Conversely, patients whose δEDSS-ratio was between 1and 2 had clinical features and brain MRI compatible with a more typical, even if aggressive, MSrelapse. In patients with TDL and PL, the flare-up of inflammatory activity led to severe tissuedamage resulting in T2 but also T1 lesion volume increase at 6-month follow-up.

ConclusionsPeculiar MRI features (TDL and PL), different from a typical MS flare-up, might occur in somepatients who experienced WNS after stopping FTY. Further studies, also involving immuno-logic biomarkers, are necessary to investigate TDL or PL pathophysiology.

M. Inglese and L. Roccatagliata are co-last authors.

From the Department of Neuroscience, Rehabilitation, Ophthalmology, Genetics (C.L., M.C., G.B., M.P., G.N., G.L.M., A.U., M.I.), Maternal and Child Health (DiNOGMI), University ofGenoa; Multiple Sclerosis Centre (D.B., A.G., M.Z.), Gallarate Hospital, ASST of Valle Olona, Gallarate; Department of Neurology (M.C., G.B., M.P., G.L.M., A.U.), Ospedale PoliclinicoSan Martino—Sistema Sanitario Regione; Liguria -Istituto di Ricovero e Cura a Carattere Scientifico per l’Oncologia; IRCCS Foundation C. Mondino National Neurological Institute (I.C.),Pavia; Department of Health Sciences (DISSAL) (M.P.S., L. R.), Ospedale Policlinico San Martino IRCCS, Genoa, Italy; Department of Radiology and Neuroscience (M.I.), Icahn School ofMedicine at Mount Sinai, New York; and Department of Neuroradiology (L.R.), Ospedale Policlinico San Martino IRCCS, Genoa, Italy.

Funding information and disclosures are provided at the end of the article. Full disclosure form information provided by the authors is available with the full text of this article atNeurology.org/NN.

The Article Processing Charge was funded by the authors.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 44: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Fingolimod (FTY) is an oral sphingosine-1-phosphate re-ceptor (S1P1) modulator approved for MS. In the past fewyears, worsening of neurologic status (WNS) has been de-scribed in a small series of patients after FTY discontinua-tion.1 This phenomenon remains controversial and regardedas MS “reactivation”2 or considered a distinct “rebound”phenomenon.3 Nevertheless, WNS after stopping FTY canlead to severe disability or can even be life-threatening, andthus the Food and Drug Administration recently issueda warning on this topic.4 We report a retrospective series of 8patients who developed WNS after FTY withdrawal focusingon the different MRI patterns in the acute phase. The aim ofthe study was to analyze MRI images in patients with MS whoexperienced WNS after stopping FTY.

MethodsStandard protocol approvals, registrations,and patient consentA written informed consent was obtained from all patients.

PatientsFrom the systematic revision of clinical records of patients, wecollected clinical-radiological data of patients with MS whodeveloped WNS after FTY withdrawal between November2013 and November 2017.

Clinical data analysisWe defined WNS by calculating the “post-FTY withdrawalδExpanded Disability Status Scale (EDSS)/pre-FTY with-drawal δEDSS ratio” (from now on called “δEDSS-ratio”),where

1. post-FTY-withdrawal δEDSS is the highest EDSS scorechange (δ) occurred after FTY withdrawal

2. pre-FTY-withdrawal δEDSS is the highest EDSS scorechange (δ) occurred during the whole previous MScourse (i.e., worst lifetime relapse).

Patients with δEDSS-ratio > 1 were enrolled in the study.

WNS after FTY discontinuation was distinguished as mono-phasic (one or more relapses, but with less than 1 monthbetween relapses), biphasic (2 relapses occurring at least 1month apart), and multiphasic (≥3 relapses).

MRI acquisition and analysisT2/FLAIR, TSE-T1 (before and after Gadolinium [Gd] ad-ministration), and DWI, performed before FTY withdrawal,during the WNS, and at 6-month follow-up (FU), were ana-lyzed to obtain T2 and T1 lesion volume (LV), number and

pattern of Gd-enhancing lesions and volume of tissue withrestricted diffusion.

Data are reported as mean ± SD.

Data availabilityRaw data are available upon appropriate request.

ResultsWe identified 8 patients. Seven patients were women. Themean age was 35.3 ± 4.9 years. The mean duration of FTYtreatment was 3.1 ± 0.8 years. The most common cause ofFTY discontinuation was the attempt to become pregnant(6 out of 7 female patients). The mean FTY discontinuation–WNS interval was 4 ± 0.9 months.

We stratified patients according to the δEDSS ratio:

1. δEDSS-ratio ≥ 2 (Pt.1–Pt.4)2. >1δEDSS-ratio < 2 (Pt.5–Pt.8).

Clinical featuresThe 4 patients with a δEDSS-ratio ≥ 2, developed a mono-phasic WNS and reached EDSS scores of up to 9, 8.5 and 9.5(Pt.1, Pt.3, and Pt.4); one patient (Pt.2) died. Pt.1, Pt.2, andPt.4 developed tetraplegia and multiple cranial nerve in-volvement with decrease in consciousness rapidly resulting incoma. Pt.3 presented a marked cognitive impairment associ-ated with motor disability.

The 4 patients with δEDSS-ratio between 1 and 2, showedmultiphasic (Pt.5), biphasic (Pt.6 and Pt.7), and monophasic(Pt.8) courses, characterized by multifocal neurologic deficits,without altered state of consciousness.

At 2-year FU, for patients with δEDSS-ratio ≥ 2, disabilityworsened when compared to patients with >1δEDSS-ratio < 2(increase in mean EDSS points: 3 ± 2.9 vs 1.1 ± 1.2) (figure 1).

Demographic and clinical features and treatment performedare detailed in the table.

MRI featuresThe 4 patients with a δEDSS-ratio ≥ 2 showed MRI featuresthat we defined as “tumefactive demyelination pattern” (TDL)(Pt.2–Pt.4) and “punctuated pattern” (PL) (Pt.3) (figure 2,figure e-1, links.lww.com/NXI/A109).

TDL was characterized by large T2/FLAIR hyperintenselesions surrounded by marked edema, mass effect on adjacent

GlossaryCL = classic MS pattern; EDSS = expanded disability status scale; FTY = fingolimod; FU = follow-up; LV = lesion volume; PL =punctuated pattern; TDL = tumefactive demyelination pattern; WNS = worsening of neurologic status.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 45: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

structures, and multiple enhancing lesions with an open ring(40%), nodular (30%), or closed ring (30%) enhancement. Alarge proportion of lesions had areas of restricted diffusion(up to 30%) (figure 2). PL presented innumerable small T2/FLAIR hyperintense lesions, mostly associated with contrastenhancement and restricted diffusion (15%) (figure 2).

The 4 patients with the δEDSS-ratio between 1 and 2 showedMRI features that we defined as “classic MS pattern” (figure 2,figure e-1, links.lww.com/NXI/A109). Brain MRIs showed

T2/FLAIR lesions with no edema nor mass effect, some withnodular/ring enhancement and only a small volume of tissuewith restricted diffusion.

At the 6-month FU, all patients exhibited a T2LV increasecompared to the pre-FTY suspensionMRI, although it decreasedaccording to the MRI scan at WNS, likely due to the partialresolution of T2 hyperintensity. This finding was more evidentfor patients with a δEDSS-ratio ≥ 2, particularly with TDL.

T1LV increased in all patients at the 6-month FU. Patientswho showed a higher volume of tissue characterized by re-stricted diffusion at WNS had a higher T1LV % increase at6-month FU (figure 1).

MRI data are detailed in the table.

DiscussionSevere WNS occurring in patients with MS after FTY with-drawal is a rare and not completely understood phenomenon.Although a recent post-hoc analysis of FREEDOMS-FREEDOMS II trials found no difference in the developmentof the so-called rebound between patients discontinuing FTYand the placebo group,5 a small series reported increases inclinical and radiologic disease activity after FTY cessation in10.9%–25.8% of patients.1,6 Furthermore, a recent study con-firmed that the “rebound” phenomenon after FTY suspensiondoes exist, with a risk estimated at 5%7 and recently FDA issueda warning about severe MS worsening after stopping FTY.4 Castaside the controversy regarding the frequency, it is relevant toconsider that the “rebound” phenomenon leads to permanentsevere disability, may be life-threatening or even fatal, as occurredto one patient of our cohort.

We retrospectively identified 8 patients who experienced,after FTY withdrawal, a WNS exceeding the highest lifetimelevel of MS activity, calculating the δEDSS-ratio to stratifyWNS severity. All patients were clinically stable during FTYtreatment, with 6 patients well enough to decide to planpregnancy.

Patients with δEDSS-ratio ≥ 2 developed severe monophasicWNS (EDSS score above 8.5), characterized by clinical fea-tures (i.e., tetraplegia resulting in coma) andMRI findings nottypical of MS. Large T2/FLAIR lesions with edema, ring/nodular enhancement, and restricted diffusion characterizedTDL, while PL showed innumerable millimetric enhancinglesions and a high rate of tissue with restricted diffusion.Conversely, patients with >1δEDSS-ratio < 2 had clinicalfeatures and brainMRI compatible with a more typical, even ifaggressive, MS relapse.

In patients with TDL and PL, the flare-up of inflammatoryactivity led to severe tissue damage resulting in T2 but alsoT1LV increase at 6 months; moreover, they demonstratedlarger areas of tissue with restricted diffusion, which may

Figure 1 EDSS score (A), T2LV (B) and T1LV (C) courses inpatients who experienced WNS after FTYwithdrawal

(A) EDSS scores were reported at FTY stop, WNS and 2-year follow-up.Patients with δEDSS-ratio ≥ 2 reached very high EDSS scores at WNS andaccumulatedmore disability at 2-year follow-up when compared to patientswith >1 δEDSS-ratio >2. (B) T2LV was calculated at FTY stop, WNS and 6-month FU. At the 6-month follow-up, all patients and particularly those withδEDSS-ratio ≥ 2, had a T2LV increase compared to the pre-FTY suspensionMRI, although decreased with respect to the MRI scan at WNS, likely due tothe resolution of a portion of T2 hyperintensity, probably reflecting oedema.(C) T1LVwas calculated at FTY stop and 6-month FU to consider chronic blackholes. At 6-month follow-up all patients, and particularly those with δEDSS-ratio ≥ 2, showed a T1LV increase with respect to the MRI scan at WNS. EDSS =expanded disability status scale; FTY = fingolimod; FU = follow-up; T1LV = T1lesion volume; T2LV = T2 lesion volume; WNS = worsening of neurologic status;δEDSS-ratio = post-FTY withdrawal δEDSS/pre-FTY withdrawal δEDSS ratio.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 46: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Table Clinical and MRI features of patients with MS who experienced WNS after FTY withdrawal

Pt.1 Pt.2 Pt.3 Pt.4 Pt.5 Pt.6 Pt.7 Pt.8

Demographics

Sex F F F F F F F M

Age 40 26 35 33 30 33 39 46

Clinical data

MS type at withdrawal SPMS RRMS RRMS RRMS RRMS RRMS RRMS SPMS

MS duration at FTY stop(y)

19 9 13 19 9 18 13 12

FTY duration (y) 4 2 4 3 3 3 3 4

Relapses and/or MRIactivity during FTY

No No No Yes No No No No

Cause of FTY withdrawal Progressive course Pregnancy attempt Pregnancyattempt

Pregnancy attempt Pregnancy attempt Pregnancy attempt Pregnancyattempt

Progressivecourse

FTY stop-WNS (m) 3 4 4 3 3 5 5 5

Pregnancy outcome NA Therapeutic abortionto treat WNS

Miscarriage 1 wkbefore WNS

Therapeutic abortion1 wk before WNS

Unsuccessful pregnancyattempt

Unsuccessfulpregnancy attempt

Unsuccessfulpregnancyattempt

NA

Gestational age attherapeutic abortion/miscarriage (wk)

NA 30 8 7 NA NA NA NA

Clinical features at WNS Tetraplegia cranial n.deficits (locked-in),coma

Tetraplegia cranial n.deficits (locked-in),coma

Severe cognitiveand motorimpairment

Tetraplegia cranial n.deficits (locked-in)coma

Left hypoesthesia (I),paraparesis (II),monoparesis (III)a

Cerebellar (I),sensori-motorsyndrome (II)a

Motor (I),cognitiveimpairment (II)a

Right sensori-motorsyndrome

Spine involvement Yes No Yes Yes Yes No No No

EDSS at WNS 9 10 8.5 9.5 6.5 6.5 5.5 6

δEDSS ratio 2 3.5 3.2 2.1 1.8 1.2 1.2 1.5

WNS course Monophasic Monophasic Monophasic Monophasic Multiphasic Biphasic Biphasic Monophasic

WNS therapy CTS, PEX, AHSCT CTS, CYC CTS, PEX, RTX CTS, PEX, RTX CTS, ALEM CTS, AHSCT CTS, RTX CTS, NAT

FTY stop-2 y FU EDSSincrease

1 7 0.5 3.5 3 0.5 0.5 0.5

MRI data

Pattern PL TDL TDL TDL CL CL CL CL

Continued

4Neurology:N

euroim

muno

logy&Neuroinflam

mation

|Vo

lume6,N

umber

4|

July2019

Neurology.org/N

N

Page 47: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Table Clinical and MRI features of patients with MS who experienced WNS after FTY withdrawal (continued)

Pt.1 Pt.2 Pt.3 Pt.4 Pt.5 Pt.6 Pt.7 Pt.8

T2 LVb (mL)

FTY stop 14 10 2 10 5 15 31 13

WNS 66 200 160 150 22c 34c 50c 25

6-mo FU 51 NA (exitus) 37 101 22 17 40 29

% Increase WNS-FTYstop

264 1900 7,900 1,400 340c 126c 61c 92

% Increase 6 mo FU-WNS

253 NA (exitus) 1750 910 340c 13c 29c 124

Gd + lesions at WNS (n) >50 >50 >50 35 3 (I), 3(II), 6(III)d 10 (I), 2(II)d 8 (I), 20 (II)d 6

Hypointense lesions atWNS (ADC) (% of Gdlesions)

Yes (≈15%)e Yes (≈25%)e Yes (≈30%)e Yes (15%) Yes (8%) Yes (2%) Yes (1%) Yes (1%)

LV restricted diffusionf

(mL)2.4 4.2 5.8 2.8 0.6 0.3 0.3 0.1

T1 LVb (mL)

FTY stop 1 7 2 0.3 1 6 8 1

6-months FU 18 NA (exitus) 13 6 6 9 9 1.4

% Increase 6 month FU-FTY stop

1700 NA (exitus) 550 1900 500 50 12 40

Abbreviations: ADC = apparent diffusion coefficient maps; AHSCT = autologous hematopoietic stem cells transplantation; ALEM = alemtuzumab; CL = Classic MS pattern; CTS = corticosteroids; CYC = cyclophosphamide; EDSS =expanded disability status scale; FTY = fingolimod; FU = follow-up; LV = lesion volume; NA = not applicable; NAT = natalizumab; PEX = plasma exchange; PL = punctuated pattern; Pt = patient; RRMS, relapsing-remittingMS; RTX =rituximab; SPMS, secondary progressive MS; T1LV = T1 lesion volume; T2LV = T2 lesion volume; TDL = tumefactive demyelinating pattern; WNS = worsening of neurologic status; δEDSS ratio = post-FTY withdrawal δEDSS/pre-FTY withdrawal δEDSS ratio.a For patients who experienced biphasic and multiphasic WNS, clinical features of each relapse are reported.b Obtained by using a manual segmentation technique on 3-mm-slice thickness (Analyze, version 12.0).c For Pt.5, Pt.6, and Pt.7, who experienced multiphasic and biphasic WNS, the highest T2LV was reported and used for the analysis.d For Pt.5, Pt.6, and Pt.7, who experienced multiphasic and biphasic WNS, Gd enhancing lesions number of each relapse (I, II, III) was reported.e For Pt.1, Pt.2, Pt.3, who had >50 Gd enhancing lesions, a maximum number of 50 Gd enhancing lesions was considered to calculate the rate of hypointense lesions on ADC maps.f Defined by ADC values <620 × 10−6 mm2/s.

Neurolo

gy.org/N

NNeu

rology:N

euroimmunology

&Neuroinflam

mation

|Volum

e6,N

umber

4|

July2019

5

Page 48: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

suggest that cytotoxic edema and/or high inflammatory cellsdensity within acute lesions resulted in more profound braintissue damage.8

The pathophysiology of the “rebound” phenomenon is stillunclear. Brain histologic examination of the patient who died(reported in a separate publication) revealed prominentastrocytic gliosis, with large hypertrophic reactive astrocytesshowing intense S1P1 expression.9 The role of astrocytes inmodulating the influx of leukocytes into the CNS has beendemonstrated in a model of experimental autoimmune en-cephalitis10 and warrants further investigation in the contextof FTY withdrawal. MRI features similar to the TDL and PLpatterns that we described have been reported in patients withevidence of B cell hyper-repopulation after alemtuzumabtreatment.11

The small size of our cohort represents a limitation of thestudy. We used the δEDSS-ratio as a method to stratify thepatients who worsened after FTY suspension and its role asa predictor of the outcome after WNS is unknown. In-terestingly, patients with worse clinical course and outcomewere those who showed peculiar MRI features that we defined

TDL and PL, different from a typical MS flare-up. Furtherstudies are necessary to investigate whether specific cellularsubsets play a role in patients who develop a severeWNS afterFTY cessation.

Study fundingThis study did not receive any funding support.

DisclosureC. Lapucci reports no disclosures. D. Baronicini receivedtravel grants from Genzyme, Novartis, and Merck for partic-ipation at national and international congresses; he receivedpersonal compensation fromAlmirall for scientific publicationand honoraria from Sanofy for participating in an advisoryboard. M. Cellerino, G. Boffa, and I. Callegari report no dis-closures. M. Pardini received research support from Novartisand personal fees from Teva and Merck. G. Novi reports nodisclosures. M.P. Sormani received consulting fees fromBiogen Idec, Merck Serono, Teva, Genzyme, Roche, Novartis,GeNeuro, and Medday; G.L. Mancardi received honoraria forlecturing, travel expenses for attending meetings, and financialsupport for research from Bayer Schering, Biogen Idec,Sanofi-Aventis, Merck Serono Pharmaceuticals, Novartis,

Figure 2 Brain MRI features of 3 representatives patients with MS who experienced WNS after FTY withdrawal

(A) Tumefactive demyelination pattern (TDL), Pt.3. (A.a) Axial T2/FLAIR images, showing large and edematous lesions, causing mass effect on adjacentstructures; (A.b) Axial TSE-T1 after Gd administration images, showing multiple ring and nodular enhancing lesions; (A.c) axial DWI (on the left) andcorrespondent ADC maps (on the right) images, showing hyperintense lesions on DWI with correspondent hypointense signal on ADC maps (red arrows),expression of restricted diffusion. (B) Punctuated pattern (PL), Pt.1, SPMS patient. (B.a) axial TSE-T2 images, medium in size (already detectable in previousscans) and new small hyperintense lesions, with no edema andmass effect; (B.b) axial TSE-T1 after gadolinium administration images, showing innumerablemillimetric enhancing lesions, both in infratentorial and supratentorial areas. (B.c) axial DWI (on the left) and correspondent ADCmaps (on the right) images,showing hyperintense lesions on DWI with correspondent hypointense signal on ADC maps (red arrows), expression of restricted diffusion. (C) Classic MSpattern (CL). (Pt.6) (C.a) Axial FLAIR images, showingmedium in size lesions, with no edema andmass effect; (B.b) axial TSE-T1 after gadoliniumadministrationimages, showing nodular and ring enhancing lesions. (C.c) Axial DWI (on the left) and correspondent ADC maps (on the right) images, showing hyperintenselesions on DWI with correspondent hyperintense signal on ADC maps (“T2-shine through effect”).

6 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 49: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Genzyme, and Teva. A. Ghezzi received honoraria forspeaking and consultancy by Novartis, Genzyme, Roche,Merck Serono, Teva, and Mylan. M. Zaffaroni received hon-oraria for consultancy and participation in advisory boards ortravel grants from Genzyme, Biogen Idec, Merck Serono,Sanofi-Aventis, Teva, and Novartis. A. Uccelli received grantsand contracts from FISM, Novartis, Fondazione Cariplo,Italian Ministry of Health; received honoraria and consulta-tion fees from Biogen, Roche, Teva, Merck, Genzyme, andNovartis. M. Inglese received research grants from NIH,DOD, NMSS, FISM, and Teva Neuroscience. L. Roccata-gliata received research grants from FISM. Disclosuresavailable: Neurology.org/NN.

Publication historyReceived by Neurology: Neuroimmunology & NeuroinflammationFebruary 26, 2019. Accepted in final form March 5, 2019.

References1. Hatcher SE, Waubant E, Nourbakhsh B, et al. Rebound syndrome in patients with

multiple sclerosis after cessation of fingolimod treatment. JAMA 2016;73:790–794.

2. Berger B, Baumgartner A, Rauer S, et al. Severe disease reactivation in four patients withrelapsing-remitting multiple sclerosis after fingolimod cessation. J Neuroimmunol 2015;282:118–122.

3. Havla JB, Pellkofer HL, Meinl I, et al. Rebound of disease activity after withdrawal offingolimod (FTY720) treatment. Arch Neurol 2012;69:262–264.

4. FDA. Gilenya (Fingolimod): Drug Safety Communication-Severe Worsening ofMultiple Sclerosis After Stopping the Medicine. 2018. Available at: fda.gov/Safety/MedWatch/SafetyInformation/SafetyAlertsforHumanMedicalProducts/ucm626264.htm. Accessed November 21, 2018.

Appendix Authors

Name Location Role Contribution

CaterinaLapucci, MD

DINOGMI,University ofGenoa

Author(correspondingauthor)

Designed andconceptualizedstudy; analyzedthe data; anddrafted themanuscript forintellectualcontent

D. Baroncini,MD

MultipleSclerosis Centre,GallarateHospital, ASST ofValle Olona,Gallarate, Italy

Author Acquisition ofdata; andrevised themanuscript forintellectualcontent

M. Cellerino,MD

DINOGMI,University ofGenoa

Author Acquisition ofdata; andrevised themanuscript forintellectualcontent

G. Boffa, MD DINOGMI,University ofGenoa

Author Acquisition ofdata; andrevised themanuscript forintellectualcontent

I.Callegari,MD

IRCCSFoundation C.MondinoNationalNeurologicalInstitute, Pavia,Italy

Author Acquisition ofdata; andrevised themanuscript forintellectualcontent

M. Pardini,MD, PhD

DINOGMI,University ofGenoa

Author Analyzed thedata; andrevised themanuscript forintellectualcontent

G. Novi, MD DINOGMI,University ofGenoa

Author Acquisition ofdata; andrevised themanuscript forintellectualcontent

Appendix (continued)

Name Location Role Contribution

M.P. Sormani,PhD

OspedalePoliclinico

Author Revised themanuscript forintellectualcontent

G.L.Mancardi,MD

DINOGMI,University ofGenoa

Author Revised themanuscript forintellectualcontent

A. Ghezzi, MD MultipleSclerosis Centre,GallarateHospital, ASST ofValle Olona,Gallarate, Italy

Author Revised themanuscript forintellectualcontent

M. Zaffaroni,MD

MultipleSclerosis Centre,GallarateHospital, ASST ofValle Olona,Gallarate, Italy

Author Revised themanuscript forintellectualcontent

A. Uccelli, MD Department ofNeurosciences,Rehabilitation,Ophthalmology,Genetics,Maternal andChild HealthUnit, Center ofExcellence forBiomedicalResearch,University ofGenoa, Genoa,Italy; OspedalePoliclinico SanMartino-IRCCS,Genoa, Italy

Author Revised themanuscript forintellectualcontent

M. Inglese,MD, PhD

DINOGMI,University ofGenoa; IcahnSchool ofMedicine, MountSinai, NY

Author Designed andconceptualizedthe study; andrevised themanuscript forintellectualcontent

L.Roccatagliata,MD, PhD

OspedalePoliclinico SanMartino IRCCS,Genoa

Author Designed andconceptualizedthe study;analyzed thedata; andrevised themanuscript forintellectualcontent

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 7

Page 50: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

5. Vermersch P, Radue EW, Putzki N, et al. A comparison of multiple sclerosis diseaseactivity after discontinuation of fingolimod and placebo. Mult Scler J Exp Transl Clin2017;3:2055217317730096.

6. Uygunoglu U, Tutuncu M, Altintas A, et al. Factors predictive of severe multiplesclerosis disease reactivation after fingolimod cessation. Neurologist 2018;23:12–16.

7. Frau J, Sormani MP, Signori A, et al. Clinical activity after fingolimod cessation:disease reactivation or rebound?. Eur J Neurol 2018;25:1270–1275.

8. Abdoli M, Chakraborty S, MacLean HJ, et al. The evaluation of MRI diffusion values ofactive demyelinating lesions in multiple sclerosis. Mult Scler Relat Disord 2016;10:97–102.

9. Giordana MT, Cavalla P, Uccelli A, et al. Overexpression of sphingosine-1-phosphate receptors on reactive astrocytes drives neuropathology of multiplesclerosis rebound after fingolimod discontinuation. Mult Scler 2018;24:1133–1137.

10. Voskuhl RR, Peterson RS, Song B, et al. Reactive astrocytes form scar-like perivascularbarriers to leukocytes during adaptive immuneinflammation of the CNS. J Neurosci2009;29:11511–11522.

11. Wehrum T, Beume LA, Stich O, et al. Activation of disease during therapywith alemtuzumab in 3 patients with multiple sclerosis. Neurology 2018;90:e601–e605.

8 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 51: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

ARTICLE OPEN ACCESS CLASS OF EVIDENCE

Does time equal vision in the acute treatmentof a cohort of AQP4 and MOG optic neuritis?Hadas Stiebel-Kalish, MD, Mark Andrew Hellmann, MD, Michael Mimouni, MD, Friedemann Paul, MD,

Omer Bialer, MD, Michael Bach, PhD, and Itay Lotan, MD

Neurol Neuroimmunol Neuroinflamm 2019;6:e572. doi:10.1212/NXI.0000000000000572

Correspondence

Dr. Stiebel-Kalish

[email protected]

AbstractObjectiveTo investigate whether visual disability which is known to accumulate by poor recovery fromoptic neuritis (ON) attacks can be lessened by early treatment, we investigated whether thetime from symptom onset to high-dose IV methylprednisolone (IVMP) affected visualrecovery.

MethodsA retrospective study was performed in a consecutive cohort of patients following their firstaquaporin-4 (AQP4)-IgG or myelin oligodendrocyte glycoprotein (MOG)-IgG-ON. Best-corrected visual acuity (BCVA) in ON eyes at 3 months (BCVA3mo) was correlated with timeto IVMP (days). In cases of bilateral ON, 1 eye was randomly selected.

ResultsA total of 29 of 37 patients had ON (27 AQP4-seropositive neuromyelitis optica spectrumdisorder [NMOSD] and 9 MOG-IgG-ON), 2 of whom refused treatment. Of the 27 patientsincluded, 10 presented later than 7 days from onset. The median BCVA3mo of patients treated>7 days was 20/100 (interquartile range 20/100–20/200). Patients treated >7 days had an ORof 5.50 (95% CI 0.88–34.46, p = 0.051) of failure to regain 0.0 logMAR vision (20/20) and anOR of 10.0 (95% CI 1.39–71.9) of failure to regain 0.2 logMAR vision (20/30) (p = 0.01)compared with patients treated within 7 days. ROC analysis revealed that the optimal criterionof delay in IVMP initiation was ≤4 days, with a sensitivity and specificity of 71.4% and 76.9%,respectively.

ConclusionsIn this retrospective study of ON with AQP4 and MOG-IgG, even a 7-day delay in IVMPinitiation was detrimental to vision. These results highlight the importance of early treatmentfor the long-term visual recovery in this group of patients. A prospective, multicenter study ofthe effects of timing of IVMP is currently underway.

Classification of evidenceThis study provides Class IV evidence that hyperacute treatment of AQP4 and MOG-ON withIVMP increases the chance for good visual recovery (20/20 vision) and that even a greater than7-day delay in treatment is associated with a higher risk for poor visual recovery.

MORE ONLINE

Class of EvidenceCriteria for ratingtherapeutic and diagnosticstudies

NPub.org/coe

From the Sackler School of Medicine (H.S.-K., M.A.H., O.B., I.L.), Tel Aviv University; Neuro-Ophthalmology Unit (H.S.-K., O.B.), Department of Ophthalmology, Rabin Medical Center;Neuro-Immunology Service and Department of Neurology (M.A.H., I.L.), Rabin Medical Center, Petah Tikva; Department of Ophthalmology (M.M.), Rambam Health Care Campus,and Ruth Rappaport Faculty of Medicine, Technion-Israel Institute of Technology, Haifa, Israel; NeuroCure Clinical Research Center and Experimental and Clinical Research Center(F.P.), Max Delbrueck Center for Molecular Medicine, Charite—Universitatsmedizin Berlin, Corporate Member of Freie Universitat Berlin, Humboldt-Universitat zu Berlin, and BerlinInstitute of Health; and Eye Center (M.B.), Medical Center, University of Freiburg and Faculty of Medicine, University of Freiburg, Germany.

Funding information and disclosures are provided at the end of the article. Full disclosure form information provided by the authors is available with the full text of this article at Neurology.org/NN.

The Article Processing Charge was funded by the authors.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 52: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Optic neuritis (ON) is a common inflammation of the opticnerve associated with numerous autoimmune conditions, in-cluding MS, neuromyelitis optica spectrum disorders(NMOSDs), chronic relapsing inflammatory optic neuritis(CRION), and autoimmune optic neuritis (AON).1–5

NMOSD is further subdivided into aquaporin-4 (AQP4)antibody–positive disease and a seronegative form.6 A subsetof patients with ON have serum IgG autoantibodies to myelinoligodendrocyte glycoprotein (MOG).7–11 The protein andcellular targets of these 2 antibodies are distinct in that AQP4is expressed on astrocytes and retinal Muller cells, whereasMOG is expressed by oligodendrocytes.12,13 Despite thesepathogenic differences, ON attacks in both conditions aretreated similarly with high-dose corticosteroids and/orplasma exchange (PE). Although some patients with MOGab disease meet the 2015 criteria for NMOSD, there is anongoing debate as to whether MOG ab-positive patientsshould receive a diagnosis of NMOSD.14 Although a signifi-cant number of MOG ab-positive patients have a relapsingcourse leading to accumulative disability, others do not

relapse; thus, their inclusion together with other AQP4-seronegative patients with NMOSD could compromise thestudy of therapeutic candidates in NMOSD.14

Acute treatment of ON in MS was shaped by the NorthAmerican Optic Neuritis Treatment Trial (ONTT), whichshowed that IV methylprednisolone (IVMP) accelerates re-covery but does not affect the final visual outcome.15,16

However, the clinical course of ON in NMOSD and in MOGab-positive patients differs from MS and is typically steroidresponsive or dependent. Disability fromboth AQP4 andMOG-ON is accumulated by poor recovery from attacks.17 The rec-ommended acute treatment options in antibody-mediated ONare high-dose IVMP, PE, and immunoadsorption.18,19

Historically, NMOSD-ONhas been associated with a poor visualoutcome.20 Studies have correlated the visual outcome of AQP4-ON attacks with the severity of visual loss at presentation, type ofantibody, and with the use of additional PE.21,22 Visual disability

GlossaryAON = autoimmune optic neuritis; AQP = aquaporin; AUC = area under the curve; BCVA = best-corrected visual acuity; IA =immunoadsorption; IVMP = IV methylprednisolone;MOG = myelin oligodendrocyte glycoprotein;OCT = optical coherencetomography; ON = optic neuritis; ONTT = Optic Neuritis Treatment Trial; PE = plasma exchange; RGC = retinal ganglioncell; VA = visual acuity.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 53: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

has been shown to be accrued with each attack, resulting in poorquality of life.13 Three previous studies focused on the effect oftiming of IVMP on visual outcome.23–25 These studies includedseveral subtypes of ON, with only a few patients with NMOSDand no MOG-positive patients.

In this study, we tested the hypothesis that timing of IVMPaffects visual outcome in a cohort of AQP4-IgG and MOG-IgG–positive patients with ON by analyzing the effect of thenumber of days until treatment commenced with the best-corrected visual acuity (BCVA) at 3 months.

MethodsPatientsWe conducted a retrospective case review of a cohort of allconsecutive patients presenting to a tertiary referral neuro-ophthalmology and neuroimmunology center at RabinMedical Center, Israel, with a first event of AQP4 or MOG-ON between January 2005 and June 2018.

Standard protocol approvals, registrations,and patient consentsThe study was performed following IRB approval in accor-dance with the World Medical Association Declaration ofHelsinki. The neuro-ophthalmology unit database wassearched for the diagnoses of NMOSD, AQP4, and MOG-associated ON.

Inclusion and exclusion criteriaON was diagnosed based on a combination of clinical history,objective findings as determined by clinical examination ofa neuro-ophthalmologist, and paraclinical tests. These includedpatients presenting with subacute onset vision loss, pain witheye movement, visual field defects consistent with an opticnerve injury, color defects, MRI evidence of optic nerve in-flammation (increasedT2 signal, gadolinium enhancement, andoptic nerve swelling),26 and neurophysiologic abnormalities(delayed visual evoked potential latencies).27 Exclusion criteriawere other ocular causes of poor visual acuity (VA) and treat-ment refusal. This retrospective cohort study focused on VA asa functional outcome and did not examine other functionalparameters such as visual field or structural-anatomic outcomemeasures such as optical coherence tomography (OCT) out-comes because for some patients, these were either missing (3patients) or performed by different machines (2 patients).

Patients had to have a diagnosis of NMOSD, AQP4, orMOG-associated ON based on established diagnostic criteria.6,28

AQP4 antibodies were tested using a commercial cell-basedkit (EUROIMMUN, Lubeck, Germany). In addition, AQP4IgG antibodies were tested at the Center for AutoimmuneNeurology in Barcelona, Spain, using tissue immunohisto-chemistry and cell-based assays.28,29 MOG-IgG antibodieswere tested by cell-based assays at the Center for Autoim-mune Neurology in Barcelona, Spain.30

TreatmentThe treatment received was IVMP at a daily dose of 1,000 mgfor 3–5 days, followed by oral prednisone (starting at 1 mg/kg/d). At the time of presentation, antibody status was notknown for the majority of patients, but oral prednisonetreatment was prolonged in patients with relapse of visual lossfollowing steroid cessation or in patients presenting withclinical or paraclinical findings suggestive of AQP4 or MOGantibody disease. Patients who refused acute treatment withIVMP for ON were excluded from this study (figure e-1,flowchart, links.lww.com/NXI/A116).

Clinical assessment and medical notesMedical notes had to include a detailed report of the timing ofpatient-reported onset of visual loss, timing of IVMP treat-ment, and documentation of high-contrast BCVA examina-tion in each eye at 3 months following the attack.

Main outcome measuresThemain outcomemeasure of this study was 3-month BCVA.Secondary outcomes were failure to regain 0.0 logMAR (20/20) and 0.2 logMAR vision (20/30) vision at the 3-monthfollow-up visit.

Level of evidenceThis is a level IV retrospective cohort study comparing the BCVAat 3 months of patients with AQP4 and MOG-ON presentingearly for IVMP treatment vs those patients presenting late.

Statistical analysisDescriptive statistics were calculated using SAS software (v9.4).Median logMAR BCVA at 3 months (“BCVA3mo”) andinterquartile ranges were documented at 3 months. Patientswere grouped according to BCVA3mo into those achieving 0.0logMAR (20/20) vision and those whose BCVA3mowas worsethan 0.0 logMAR. Outcome was correlated with time (in days)from symptom onset to IVMP (“time to IVMP”). For patientswith bilateral ON, 1 eye was randomly included in the analysis.A receiver operator curve (ROC) was used to analyze the bestsensitivity and specificity using the Youden index31 for the bestcutoff time to IVMP to achieve the best BCVA3mo. The relativerisk, OR, and confidence intervals for BCVA3moworse than 0.0logMAR (20/20) and 0.2 logMAR vision (20/30) were ana-lyzed for patients treated early (time to IVMP ≤ 6 days) com-pared with patients treated after day 7. Two-tailed tests wereused, and p < 0.05 was accepted as statistically significant.

Data availabilityData have been uploaded and will be made readily availableupon publication at the following Mendeley data repository:dx.doi.org/10.17632/ht5s9cc845.1.

ResultsThirty-seven patients were enrolled. Twenty-eight patientsfulfilled the 2015 diagnostic criteria6 for NMOSD (27 wereAQP4 positive, and 1 patient was seronegative), and another 9

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 54: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

had MOG-IgG positive ON. Included in this study were 27AQP4-positive and 9MOG-positive patients with ON. Figuree-1 (links.lww.com/NXI/A116) depicts the flowchart of thepatient files reviewed (n = 37) and those included in finalanalysis (n = 27). The mean age at presentation was 36.6 ±13.7 (range 8.3–68.1) years, and 85.2% (n = 23) were female.Themean age at presentation for patients withMOG-ONwas41.8 ± 11.1 (range 26.2–55.6) years, and 78% (n = 7) werefemale. BCVA at nadir revealed no trend toward worst BCVAnadir in the delayed treatment group (4 days as cutoff), witha mean of 1.55 ± 0.74 for those treated >4 days and 0.99 ± 0.85for those treated ≤4 days, p = 0.085. It is interesting to notethat this trend leveled off to no difference in BCVA nadirwhen comparing those treated <7 days and those treated ≥7days (BCVA nadir 1.17 ± 0.83 in those treated <7 days and1.47 ± 0.84 in those treated later, p = 0.41). Patients weretreated with IVMP on the same day they presented with ON.The median time to IVMP was 4 days for the whole cohort(range 1–65 days). Of those treated ≥7 days, the median timeto IVMP was 21 days (range 9–65 days). The median time toIVMP for those treated earlier than <7 days was 3 days and 2days for those treated within 4 days. Baseline demographicand clinical factors were similar in both the early treatmentgroup (<4 day treatment group) and those treated >4 days(percentage of MOG positive p = 0.59, male p = 0.94, age p =0.48, additional use of plasmapheresis p = 1).

Three-month VAThere was a significant inverse correlation betweenBCVA3mo (logMAR) and age (r = −0.41, p = 0.04) and daysto IVMP treatment (r = 0.43, p = 0.03), with a nearly sig-nificant correlation between BCVA3mo and logMAR VA atnadir (r = 0.38, p = 0.06). The distribution of BCVA3mo isdepicted in figure 1.

The BCVA3mo was similar between men and women (0.33 ±0.52 vs 0.17 ± 0.47, p = 0.61) and similar between AQP4-positive and MOG-positive patients (0.11 ± 0.09 vs 0.22 ±0.56, p = 0.38). Using multivariate analysis, with type of an-tibody (AQP4 vs MOG), age, days to IVMP treatment, log-MAR VA at nadir, and plasmapheresis treatment as theindependent variables, the 2 factors that remained significantin predicting BCVA3mo were days to IVMP treatment (r2 =15.5%, p = 0.03) and age (r2 = 16.5%, p = 0.04).

Failure to regain 0.0 logMAR (20/20) visionAn ROC analysis was performed with days to IVMP treat-ment as the predictor and failure to regain 0.0 logMAR (20/20) vision as the dependent variable. An area under the curve(AUC) of 0.71 was achieved (figure 2), and with a Youdenoptimal criterion of days to treatment >4 days, a sensitivityand specificity of 71.4% and 76.9%, respectively, were ach-ieved. Patients who were treated later than 4 days had an ORof 8.33 (95% CI 1.47–47.22) of failure to regain 0.0 logMARvision (p = 0.01). The individual AUCs of age and nadirBCVA as individual predictors of failure to regain 0.0 logMARvision at 3 months were lower (0.56 and 0.60, respectively),

and the addition of these 2 predictors to days to IVMPtreatment led to a minute improvement in the AUC (0.74)compared with days to treatment alone (0.71).

Failure to regain 0.2 logMAR vision(∼Snellen 20/30)A similar analysis with failure to regain 0.2 logMAR as thedependent variable revealed a Youden optimal criterion ofdays to treatment >7 days with an AUC of 0.84 (figure 3),sensitivity of 71.4%, and specificity of 80.0%. Patients treatedlater than 7 days had an OR of 10.0 (95% CI 1.39–71.86) offailure to regain 20/30 vision (p = 0.01). The individual AUCsof age and nadir BCVA as individual predictors of failure toregain 0.2 logMAR vision at 3 months were lower (0.59 and0.63, respectively), and the addition of these 2 predictors todays to IVMP treatment led to a reduced AUC (0.80) com-pared with days to treatment alone (0.84).

DiscussionIn our cohort, patients with AQP4- and MOG-positive ONresponded better to earlier IVMP. Of 27 patients with AQP4or MOG-ON (18 AQP4-IgG+ and 9 MOG-IgG+), and thosetreated later than 4 days had an OR of 8.33 of failure to regain20/20 0.0 logMAR vision (p = 0.01). Patients treated laterthan 7 days had an OR of 10.0 of failure to regain 20/30 0.2logMAR vision (p = 0.01). This finding corroborates a studyin patients with acute ON, demonstrating that retinal ganglioncell (RGC) layer loss starts within a few days of ON and maybe a predictor of visual loss.32

Figure 1 Distribution of BCVA3mo for patients treated withIVMP for AQP4 and MOG-ON

Note inverted logMAR scale: better acuity at top. Left boxplot: eyes ofpatients treated <7 days. Right boxplot: Eyes of patients treated ≥7 days.BCVA3mo = best-corrected visual acuity at 3 months after IVMP treatmentfor AQP4 and MOG-IgG-ON. Box plot details: thick horizontal bar: median;box: interquartile range (25%–75%). Dots: outliers. AQP = aquaporin; BCVA =best-corrected visual acuity; IVMP = IV methylprednisolone; MOG = myelinoligodendrocyte glycoprotein; ON = optic neuritis.

4 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 55: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

ON in patients with AQP4-IgG and MOG-IgG antibodies isfrequently steroid responsive or dependent, thus differingfrom MS-ON, in which IVMP does not affect visualoutcome.1,33 We tested our hypothesis that timing of acutetreatment affects visual outcome in AMDD-ON. Despite thesmall number of patients enrolled and investigated, we wereable to construct ROC curves to identify cut points that op-timize the balance between sensitivity and specificity in regardto the optimal time window for the administration of IVMPthat would also translate into greater improvement of theBCVA at 3 months. Administration of IVMP treatment at day4 or earlier was the identified cut point (71.4% sensitive;76.9% specific). Two additional variables affecting visualoutcome to a lesser degree were age and VA at nadir.

Offering a better visual outcome for AQP4-seropositive andMOG-seropositive patients with ON implies the need foraction in all forms of ON because at presentation, the etiologyis often unclear. InMS-related ON, the ONTT suggested thatfinal visual outcome is not affected by acute treatment withIVMP,15,16 leading to a sense of nonurgency in the acutephase of ON. A change of treatment paradigm, especially inacceleration of IVMP timing in acute ON treatment, may beneeded. For a significant number of patients who harborAQP4 ab orMOG ab at presentation, it may be crucial to startIVMP treatment for ON as soon as possible.

Perhaps most salient about this submission was the recogni-tion that as little as a 7-day delay in treatment inception (forNMOSD and anti–MOG-associated optic neuritides) wasfound to be detrimental in terms of the OR for improvingBCVA at 3 months after symptom onset. “Time is Tissue” isa core principal that is evolving in the field of neuro-immunology,34 making it imperative to potentially view anantibody-mediated ONwith a comparable sense of urgency in

terms of diagnosis and treatment akin to that of heart attackand stroke. Our findings are in good alignment with thefindings by Soelberg et al.,32 who reported that in ON, themajority of which were not antibody mediated, progressiveganglion cell layer loss at a rate of 0.2 μm/d can be observed asearly as 8 days after onset.

Corroborating the contention of Time is Tissue has been therecognition of inflammation as a fundamental antecedent ofthe cardinal hallmark of irreversible disability in those withinflammatory syndromes of the CNS; that being axonaltransection, the evolution of dying back and Wallerian de-generation. Among the most striking observations of thisproposed model of sequential steps in the pathobiology ofpostinflammatory neurodegeneration has been the de-generation of RGCs within a time epoch as short as 2 days ofthe onset of clinical symptoms ultimately designated as a de-rivative of such inflammation.

The results of our study strengthen 3 previous reports in otherforms of ON23–25 demonstrating a beneficial effect of hy-peracute IVMP. These studies23–25 did not focus on AQP4and MOG-ON; Osinga et al.23 described a cohort of 19patients with recurrent ON, 9 of whomwith relapsing isolatedON, 4 with MS-ON, 4 with chronic relapsing inflammatoryoptic neuropathy, and 2 with NMOSD-ON. These 19patients were analyzed for the effects of treatment within 2days (hyperacute treatment). The importance of hyperacutesteroids in ON treatment has experimental logic in animalmodels. In mice with experimental autoimmune encephalo-myelitis, the inflammatory process precedes axonal de-generation by 2 days.35 A goal of treatment within 2 days ofsymptom onset is difficult to achieve in clinical reality. An-other study by Zhu et al.36 showed that irreversible axonaldamage starts between days 5 and 7, supporting our clinical

Figure 2A receiver operating characteristic curve of days toIVMP as a predictor of failure to regain 0.0 logMAR(20/20) vision (AUC 0.71, p < 0.001)

AUC = area under the curve; IVMP = IV methylprednisolone.

Figure 3A receiver operating characteristic curve of days toIVMP as a predictor of failure to regain 0.2 logMAR(20/30) vision (AUC 0.84, p < 0.001)

AUC = area under the curve; IVMP = IV methylprednisolone.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 5

Page 56: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

finding that optimal treatment is by day 4, but that treatmentbefore day 7 still offers an opportunity for very good visualoutcome.

Previous MRI and OCT studies have demonstrated that thebulk of axonal loss and neuronal damage is sustained early inthe disease course for patients with MS.37,38 Although therates of ganglion cell–inner plexiform layer atrophy may beinfluenced by disease-modifying therapies in patients withMS, further studies, using the detailed structural OCT toolscurrently at our disposal, should re-examine the effect oftiming of IVMP on visual outcome in other forms of ON.

Few clinical studies on outcome of NMOSD-ON includedetails of accurate timing from symptom onset to acutetreatment, and there is much need for this detail to be ana-lyzed in larger cohorts. The results of this study show a trendindicating that even a 7-day delay in IVMP can be detri-mental to vision in AQP4 and MOG-IgG ON. Several lim-itations should be taken into consideration whenconsidering these results, including the study’s retrospectivedesign, the small sample size resulting in very large confi-dence intervals, the short follow-up duration, and the lack ofparaclinical data to confirm the functional results withstructural indices such as loss of retinal nerve fiber andganglion cell layers on OCT.

A prospective study in a larger cohort of patients withNMOSD examining the effects of timing of IVMP on addi-tional visual parameters such as OCT, visual fields BCVA, andon subsequent ON attacks seems warranted.

Study fundingNo targeted funding reported.

DisclosureH. Kalish received research support from the MaratierFoundation, Tel Aviv University, and Israeli Car AccidentPrevention Association. M.A. Hellmann and M. Mimounireport no disclosures. F. Paul served on the scientific advisoryboards of Novartis and MedImmune; received speaker hon-oraria and travel funding from Bayer, Novartis, Biogen, Teva,Sanofi-Aventis/Genzyme, Merck Serono, Alexion, Chugai,MedImmune, and Shire; serves as academic editor of PLoSONE and associate editor of Neurology: Neuroimmunology &Neuroinflammation; consulted for Sanofi-Genzyme, Biogen,MedImmune, Shire, and Alexion; and received research sup-port from Bayer, Novartis, Biogen, Teva, Sanofi-Aventis/Genzyme, Alexion, Merck Serono, German Research Council,Werth Stiftung of the City of Cologne, German Ministry ofEducation and Research, Arthur Arnstein Stiftung Berlin, EUFP7 Framework Program, Guthy Jackson Charitable Foun-dation, and NMSS. O. Bialer and M. Bach report no dis-closures. L. Lotan received travel funding from Teva, MerckSerono, Biogen, and Sanofi-Genzyme. Full disclosure forminformation provided by the authors is available with the fulltext of this article at Neurology.org/NN.

Publication historyReceived by Neurology: Neuroimmunology & NeuroinflammationDecember 10, 2018. Accepted in final form March 8, 2019.

References1. Balcer LJ. Clinical practice. Optic neuritis. N Engl J Med 2006;354:1273–1280.2. Galetta SL, Villoslada P, Levin N, et al. Acute optic neuritis: unmet clinical needs and

model for new therapies. Neurol Neuroimmunol Neuroinflamm 2015;2:e135. doi:10.1212/NXI.0000000000000135.

3. Soelberg K, Jarius S, Skejoe HPB, et al. A population-based prospective study of opticneuritis. Mult Scler J 2017;23:1893–1901.

4. Petzold A, Wattjes MP, Costello F, et al. The investigation of acute optic neuritis:a review and proposed protocol. Nat Rev Neurol 2014;10:447–458.

5. Atkins EJ. Optic neuritis. Encycl Neurol Sci 2014;13:681–686.6. Wingerchuk DM, Banwell B, Bennett JL, et al. International consensus diagnostic

criteria for neuromyelitis optica spectrum disorders. Neurology 2015;85:177–189.7. Jarius S, Kleiter I, Ruprecht K, et al. MOG-IgG in NMO and related disorders: a multicenter

study. Part 3: mOG-IgG-associated brainstem encephalitis. Mult Scler 2016; Conference:410-411. Available at: cochranelibrary.com/central/doi/10.1002/central/CN-01212258/full.

8. Biotti D, Bonneville F, Tournaire E, et al. Optic neuritis in patients with anti-MOGantibodies spectrum disorder: MRI and clinical features from a large multicentriccohort in France. J Neurol 2017;264:2173–2175.

9. Ramanathan S, Dale RC, Brilot F. Anti-MOG antibody: the history, clinical pheno-type, and pathogenicity of a serum biomarker for demyelination. Autoimmun Rev2016;15:307–324.

Appendix Authors

Name Location Role Contribution

HadasStiebel-Kalish, MD

Tel Aviv University &Rabin Medical Ctr.

Author Designed andconceptualized thestudy, analyzed thedata, and draftedthe manuscript forintellectual content

MarkAndrewHellmann,MD

Tel Aviv University &Rabin Medical Ctr.

Author Major role in theacquisition of dataand revised themanuscript forintellectual content

MichaelMimouni,MD

Rambam HCC, &Technion-IsraelInstitute ofTechnology

Author Interpreted the data,statistical analysis,and revised themanuscript forintellectual content

FriedemannPaul, MD

NeuroCure, ChariteUniversitatsmedizinBerlin, FreieUniversitat Berlin,Humboldt-Universitatzu Berlin, & BerlinInstitute of Health,Berlin, Germany

Author Interpreted the dataand revised themanuscript forintellectual content

Omer Bialer,MD

Tel Aviv University &Rabin Medical Ctr.

Author Major role in theacquisition of dataand revised themanuscript forintellectual content

MichaelBach, PhD

University of Freiburg,Germany & Faculty ofMedicine, Germany

Author Interpreted the data,graphical andstatistical analysis,and revised themanuscript forintellectual content

Itay Lotan,MD

Tel Aviv University &Rabin Medical Ctr.

Author Major role in theacquisition of dataand revised themanuscript forintellectual content

6 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 57: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

10. Akaishi T, Sato DK, Takahashi T, Nakashima I. Clinical spectrum of inflammatorycentral nervous system demyelinating disorders associated with antibodies againstmyelin oligodendrocyte glycoprotein. Neurochem Int Epub 2018 Oct 23.

11. Chalmoukou K, Alexopoulos H, Akrivou S, Stathopoulos P, Reindl M, Dalakas MC.Clinical/scientific notes. Neurol Neuroimmunol Neuroinflamm 2015;2:1–3.

12. Borisow N, Mori M, Kuwabara S, Scheel M, Paul F. Diagnosis and treatment of NMOspectrum disorder and MOG-encephalomyelitis. Front Neurol 2018;9:888.

13. Schmidt F, Zimmermann H, Mikolajczak J, et al. Severe structural and functionalvisual system damage leads to profound loss of vision-related quality of life in patientswith neuromyelitis optica spectrum disorders. Mult Scler Relat Disord 2017;11:45–50.

14. Zamvil SS, Slavin AJ. Does MOG Ig-positive AQP4-seronegative opticospinal in-flammatory disease justify a diagnosis of NMO spectrum disorder? Neurol Neuro-immunol Neuroinflamm 2015;2:e62. doi:10.1212/NXI.0000000000000062.

15. Beck RW. The optic neuritis treatment trial: three-year follow-up results. ArchOphthalmol 1995;113:136–137.

16. KupersmithMJ, Anderson S, Kardon R. Predictive value of 1 month retinal nerve fiberlayer thinning for deficits at 6 months after acute optic neuritis. Mult Scler J 2013;19:1743–1748.

17. Pittock SJ, Lucchinetti CF. Neuromyelitis optica and the evolving spectrum of au-toimmune aquaporin-4 channelopathies: a decade later. AnnNYAcad Sci 2016;1366:20–39.

18. Sato D, Callegaro D, Lana-Peixoto MA, Fujihara K; Brazilian Committee for Treat-ment and Research in Multiple Sclerosis. Treatment of neuromyelitis optica: anevidence based review. Arq Neuropsiquiatr 2012;70:59–66.

19. Kleiter I, Gahlen A, Borisow N, et al. Apheresis therapies for NMOSD attacks:a retrospective study of 207 therapeutic interventions criteria for rating therapeuticand diagnostic studies. Neurol Neuroimmunol Neuroinflamm 2018;5:e504. doi:10.1212/NXI.0000000000000504.

20. Wingerchuk DM,HogancampWF, O’Brien PC,Weinshenker BG. The clinical courseof neuromyelitis optica (Devic’s syndrome). Neurology 1999;53:1107.

21. Bonnan M, Valentino R, Debeugny S, et al. Short delay to initiate plasma exchange isthe strongest predictor of outcome in severe attacks of NMO spectrum disorders.J Neurol Neurosurg Psychiatry 2018;89:346–351.

22. Mori S, Kurimoto T, Ueda K, Nakamura M. Short-term effect of additional apheresison visual acuity changes in patients with steroid-resistant optic neuritis in neuro-myelitis optica spectrum disorders. Jpn J Ophthalmol 2018;62:525–530.

23. Osinga E, van Oosten B, de Vries-Knoppert W, Petzold A. Time is vision in recurrentoptic neuritis. Brain Res 2017;1673:95–101.

24. Plant GT, Sibtain NA, Thomas D. Hyperacute corticosteroid treatment of opticneuritis at the onset of pain may prevent visual loss: a case series. Mult Scler Int 2011;2011:815068.

25. Nakamura M, Nakazawa T, Doi H, et al. Early high-dose intravenous methylpred-nisolone is effective in preserving retinal nerve fiber layer thickness in patients withneuromyelitis optica. Graefe’s Arch Clin Exp Ophthalmol 2010;248:1777–1785.

26. Filippi M, Rocca MA, Ciccarelli O, et al. MRI criteria for the diagnosis of multiplesclerosis: MAGNIMS consensus guidelines. Lancet Neurol 2016;15:292–303.

27. Balcer LJ, Miller DH, Reingold SC, Cohen JA. Vision and vision-related outcomemeasures in multiple sclerosis. Brain 2015;138(pt 1):11–27.

28. Jarius S, Paul F, Aktas O, et al. MOG encephalomyelitis: international recommendationson diagnosis and antibody testing [in German]. Nervenarzt 2018;89:1388–1399.

29. Hoftberger R, Sabater L, Marignier R, et al. An optimized immunohistochemistrytechnique improves NMO-IgG detection: study comparison with cell-based assays.PLoS One 2013;8:6–11.

30. Hoftberger R, Sepulveda M, Armangue T, et al. Antibodies to MOG and AQP4 inadults with neuromyelitis optica and suspected limited forms of the disease. Mult Scler2015;21:866–874.

31. Youden WJ. Index for rating diagnostic tests. Cancer 1950;3:32–35.32. Soelberg K, Specovius S, Zimmermann HG, et al. Optical coherence tomography in

acute optic neuritis: a population-based study. Acta Neurol Scand 2018;138:566–573.33. Group RWB and the optic neuritis study; Beck RW, Cleary PA, Anderson MM Jr,

et al. A randomized, controlled trial of corticosteroids in the treatment of acute opticneuritis. The Optic Neuritis Study Group. N Engl J Med 1992;326:581–588.

34. Eshaghi A, Prados F, Brownlee WJ, et al. Deep gray matter volume loss drivesdisability worsening in multiple sclerosis. Ann Neurol 2018;83:210–222.

35. Shindler KS, Ventura E, Dutt M, Rostami A. Inflammatory demyelination inducesaxonal injury and retinal ganglion cell apoptosis in experimental optic neuritis. ExpEye Res 2008;87:208–213.

36. Zhu B, Moore GRW, Zwimpfer TJ, et al. Axonal cytoskeleton changes in experimentaloptic neuritis. Brain Res 1999;824:204–217.

37. Balk LJ, Cruz-Herranz A, Albrecht P, et al. Timing of retinal neuronal and axonal lossin MS: a longitudinal OCT study. J Neurol 2016;263:1323–1331.

38. GranbergT, FanQ,TreabaCA, et al. In vivo characterization of cortical andwhitematterneuroaxonal pathology in early multiple sclerosis. Brain 2017;140:2912–2926.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 7

Page 58: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

ARTICLE OPEN ACCESS

CSF parameters associated with early MRIactivity in patients with MSAna Klein, MD, Rebecca C. Selter, MD, Alexander Hapfelmeier, PhD, Achim Berthele, MD,

Bertram Muller-Myhsok, PhD, Viola Pongratz, MD, Christiane Gasperi, MD, Claus Zimmer, MD,

Mark Muhlau, MD, and Bernhard Hemmer, MD

Neurol Neuroimmunol Neuroinflamm 2019;6:e573. doi:10.1212/NXI.0000000000000573

Correspondence

Dr. Hemmer

[email protected]

AbstractObjectiveTo identify CSF parameters at diagnosis of clinically isolated syndrome (CIS) and MS that areassociated with early inflammatory disease activity as measured by standardized cerebral MRI(cMRI).

MethodsOne hundred forty-nine patients with newly diagnosed CIS and MS were included in theretrospective study. cMRI at onset and after 12 months was analyzed for T2 and gadolinium-enhancing lesions. CSF was tested for oligoclonal bands and intrathecal synthesis of im-munoglobulin G (IgG), A (IgA), and M (IgM) before initiation of disease-modifying therapy(DMT). In a subgroup of patients, CSF and serum samples were analyzed for sCD27,neurofilament light chain, and IgG subclasses 1 and 3. Association between CSF parametersand cMRI activity was investigated by univariable and multivariable regression analysis in allpatients, DMT-treated patients, and untreated patients.

ResultsIgG index, sCD27 levels in CSF, and to a lesser extent IgM index were associated with theoccurrence of new cMRI lesions. IgG index and sCD27 levels in CSF were highly correlated. Ina multivariable analysis, IgG index and to a lesser extent IgM index together with DMTtreatment status and gender were strongest predictors of future cMRI activity.

ConclusionsCSF parameters such as IgG and IgM index are independently associated with future MRIactivity and thus might be helpful to support early treatment decisions in patients newlydiagnosed with CIS and MS.

From the Department of Neurology (A.K., R.C.S., A.B., V.P., C.G., M.M., B.H.), Klinikum rechts der Isar, Medical Faculty, Technical University of Munich; Institute of Medical Informatics(A.H.), Statistics and Epidemiology, Medical Faculty, Technical University of Munich; Max Planck Institute of Psychiatry (B.M.-M.), Munich; Department of Neuroradiology (C.Z.),Klinikum rechts der Isar, Medical Faculty, Technical University of Munich; TUMNeuroimaging Center (M.M.), Klinikum rechts der Isar, Medical Faculty, Technical University of Munich;and Munich Cluster for Systems Neurology (SyNergy) (B.H.), Germany.

Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by the authors.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 59: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

MS is a chronic inflammatory demyelinating disease of theCNS, often leading to accumulation of severe disability overtime. In patients presenting with a first clinical manifestationof MS or clinically isolated syndrome (CIS), individual dis-ease course is still unpredictable.

Several cerebral MRI (cMRI) and CSF parameters are asso-ciated with early and possibly long-term outcome. A highlesion load on initial cMRI is associated with conversion fromCIS toMS1,2 and long-term outcome.3 CSF oligoclonal bands(OCBs) were shown to be associated with a higher risk ofa second relapse and hence clinical progression from CIS toMS.1,4,5 Other CSF parameters such as intrathecal IgM syn-thesis, albumin ratio between CSF and blood, IgM, sCD27,chitinase like protein 3, micro RNA, IgG subclasses, andneurofilament light chain (NFL) were shown to be associatedwith conversion of CIS to MS, relapses, accumulation of brainlesion, and disease progression.6–17

Based on these findings, the recently revised new diagnosticMS criteria contain OCBs as a marker to demonstrate dis-semination in time besides clinical and cMRI parameters. Thishas reinforced CSF examination during the diagnostic workupof patients with CIS and MS providing the opportunity to useCSF parameter for predicting the course of disease.

The aim of this study was to identify and confirm candidateCSF parameters potentially associated with early in-flammatory disease activity as determined by cMRI and sug-gest a predictive score for patients newly diagnosed with CISand MS. We decided to focus on parameters that are used instandard diagnostic workup such as intrathecal immunoglo-bulins and OCBs on the one hand and sCD27 as a marker forinflammation and B-cell differentiation and NFL as a markerfor neurodegeneration on the other hand. Because earlydisease-modifying therapy (DMT) treatment has a majorimpact on MRI parameters, we assessed the predictive valueof the CSF biomarker in patients who did or did not receivetreatment with DMT.

MethodsPatient cohort and inclusion criteriaA total number of 149 patients with a first manifestation ofMS/CIS from the Department of Neurology of the TechnicalUniversity of Munich were included in this retrospectivestudy. Patients eligible for analysis met the following criteria:(1) first clinical event suggestive of MS within the last 12months at time of CSF sampling, (2) no further history of

neurologic symptoms suggestive of an earlier disease mani-festation, (3) therapy naive regarding DMT at the time ofCSF sampling and baseline cMRI, (4) baseline cMRI withstandardized sequences 1–3 months after initial lumbarpuncture, and (5) follow-up cMRI with standardizedsequences 12 ± 3 months after baseline cMRI. One hundrednine of 149 patients did not receive IV or oral steroid treat-ment within 30 days before CSF sampling.

Treatment was recorded in all patients and used for stratifi-cation of subgroups. Fifty-two patients were not treated withDMT (DMT−), and 97 patients received DMT (DMT+)during the follow-up period. Of these patients, 63 receivedβ-interferons, 28 glatiramer acetate, 3 natalizumab, 2 dimethylfumarate, and 1 teriflunomide. Of the 52 patients in theDMT− group, 29 declined treatment, 11 discontinued earlydue to side effects, 2 received delayed treatment after 7months, and 2 became pregnant. In 8 patients, treatment wasnot recommended by the treating physician.

The study was approved by the ethics committee of theTechnical University of Munich. Written informed consentwas obtained from every participant.

CSF and serum analysisCSF and serum were collected during diagnostic workupbetween 2008 and 2014. In a subset of patients, CSF and serawere aliquoted and stored for later analyses at −80°C. Samplesunderwent routine diagnostic analysis, which included de-tection of OCBs through isoelectric focusing and silverstaining andmeasuring of IgG, IgM, IgA, albumin, and proteinconcentrations by nephelometry. Intrathecal synthesis ofimmunoglobulins (IgG, IgA, and IgM) was calculatedaccording to Reiber formula.18 Immunoglobulin indices werecalculated as follows: IgX-Index = [IgX (CSF)/IgX (serum)]/[Albumin (CSF)/Albumin (serum)]. Thirteen patients un-derwent lumbar puncture at another hospital. In this case,externally analyzed CSF data were acquired and included inthis study. As standard diagnostic CSF workup can vary fromhospital to hospital, some parameters are missing: 2 patientshave no information regarding IgM and 6 patients re-garding IgA.

From 84 patients, frozen CSF samples were available forfurther analysis, which included testing for levels of sCD27and NFL using commercially available ELISA kits (sCD27:Human sCD27 Instant ELISA; eBioscience, San Diego, CA;NFL: NF-light ELISA; UmanDiagnostics, Umea, Sweden) aswell as determining individual IgG subclass distribution by

GlossaryAUC = area under the curve; CIS = clinically isolated syndrome; cMRI = cerebral MRI; DMT = disease modifying therapy;NFL = neurofilament;OCB = oligoclonal band; ROC = receiver operating characteristic; TE = echo time; TI = inversion time;TR = repetition time.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 60: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

measuring levels of IgG1 and IgG3 by nephelometry (Sie-mens BN ProSpec, Germany). As CSF has a much lowerconcentration of sCD27 than serum and the applied ELISAkit was designed for serum analysis, we did not comply to themanufacturer’s instructions to dilute the samples 1:25 butused undiluted CSF.

Because of small sample volumes, experiments were per-formed in single wells. Of the 84 patients with frozen CSFsamples, 61 also had sufficient amounts of stored serum,which was subsequently analyzed together with matchedCSF samples in regard to IgG1 and IgG3 levels by nephe-lometry. IgG1 and IgG3 indices were then calculated asabove.

MRIThe first scan (baseline cMRI) was performed 1–3 monthsafter lumbar puncture; the second scan (follow-up cMRI) wasperformed 12 ± 3 months after baseline cMRI. Activity oncMRI was defined as the appearance of new T2 hyperintenseand/or new gadolinium enhancing lesions on follow-upcMRI. Brain images were all acquired on the same 3 T Philipsscanner.

The scanning protocol included 3D gradient echo (GRE) T1-weighted (w) sequence before and after gadolinium injection(orientation, 170 contiguous sagittal 1 mm slices; field ofview, 240 × 240 mm; voxel size, 1.0 × 1.0 mm; repetition time[TR], 9 ms; echo time [TE], 4 ms) and 3D fluid attenuatedinversion recovery (FLAIR) sequence (orientation, 144contiguous axial 1.5 mm slices; field of view, 230 × 185 mm;voxel size, 1.0 × 1.5 mm; TR, 10,000 ms; TE, 140 ms; in-version time, 2,750 ms).

Baseline cMRI data were analyzed in regard to lesion count,lesion load, and existence of gadolinium enhancing lesions.T2-hyperintense lesions were segmented from FLAIR andT1-weighted images by a lesion growth algorithm19 asimplemented in the lesion segmentation tool toolbox version2.0.15 (statistical-modelling.de/lst.html) for SPM12 (fil.ion.ucl.ac.uk/spm). Seven cMRIs (4.7%) did not follow ourstandardized protocol and were thus excluded.

Statistical analysisThe distribution of quantitative data is described by medianand range. Qualitative data are presented by absolute andrelative frequencies. Associations of potential risk factors tothe binary outcome cMRI activity are analyzed by univariableand multivariable logistic regression models and quantified byrespective ORs with 95% CIs. Univariable models were fit tothe whole study cohort while inclusion of an interaction termbetween the respective risk factor and a dummy variablecoding DMT+ and DMT− patients allowed to estimate effectsizes within and to test the difference between these sub-groups. The multivariable model made use of gender, DMT,OCBs, IgG, IgM, IgA (indices and synthesis), lesion volume,and lesion count in a stepwise forward variable selection that

was based on Akaike’s information criterion. The prognosticstrength of the resulting model was assessed by the area underthe curve (AUC) of the receiver operating characteristic(ROC) curve. Further multivariable models were fit ina separate analysis to assess the additional prognostic valueof CSF. Therefore a baseline model including gender, DMT,lesion volume, and lesion count in the baseline cMRI wascompared against another model that extended this list ofcovariates to include further CSF parameters such as OCBs,IgG, IgM, and IgA (indices and synthesis). The diagnosticstrength of the models was again quantified by the AUC. Thedifference was tested by a likelihood ratio test of these nestedmodels. To depict the correlation between sCD27 and IgGindex, we used Pearson correlation coefficient. All analyseswere performed using R 3.5.0 (The R Foundation for Sta-tistical Computing, Vienna, Austria). Hypothesis testing wasconducted on exploratory 2-sided 5% significance levels. Asthis was an exploratory study, we did not correct for multiplecomparisons.

Data availability statementThe data that support the findings of this study are availablefrom the corresponding author on reasonable request. Thedata are not publicly available to protect the privacy of theparticipating patients.

ResultsStudy cohortFrom the 149 patients analyzed in our retrospective study, 52patients did not start DMT within 6 months after baselinecMRI or discontinued medication again after less than 6months. The remaining 97 patients received DMT for at least6 months. DMT+ and DMT−patients were comparable inregard to age, sex distribution, and time between symptomonset and first MRI (table 1). In 81 patients, we observeddisease activity by cMRI between the baseline and follow-upscan after 12 months.

Intrathecal IgG, IgM, and IgA synthesisWe analyzed the association between intrathecal IgG syn-thesis determined by the IgG index and the occurrence of newMRI lesions by logistic regression. The IgG index was asso-ciated with new cMRI lesions after 12 months (OR 3.441,CI 1.482–9.206; p = 0.008, table 2). When we stratified thecohort according to DMT treatment, the effect was morepronounced in the DMT− patients (OR 32.021, CI1.574–651.251; p = 0.024) than DMT+ patients (OR 2.090,CI 0.787–5.553; p = 0.139, table 2). We also determinedintrathecal IgG synthesis using Reiber formula. IntrathecalIgG synthesis was observed in 50% of patients. DMT−patients with intrathecal IgG synthesis had a higher risk ofcMRI activity compared to those without intrathecal synthesis(OR 11.528, CI 2.237–59.399; p = 0.003) (table 3 and figure1). This effect was also seen in all patients (OR 1.517, CI0.776–2.995; p = 0.225). In DMT+ patients intrathecal IgG

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 61: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

synthesis was not associated with cMRI activity (OR 0.742,CI 0.332–1.661; p = 0.468, table 3).

Analysis of the relation of intrathecal IgM synthesis on cMRIactivity showed that a higher IgM index was associated withcMRI activity after 12months (OR 43.370, CI 1.177–2,914.982;p = 0.056) (table 2). The effect wasmore pronounced inDMT−patients (OR after correction by Firth method 5,970.034, CI0.508–70098464.401; p = 0.069) than DMT+ patients (OR5.092, CI 0.065–396.140; p = 0.464) (table 2). We stratifiedpatients into those with and without intrathecal IgM synthesisaccording to Reiber formula. Overall 14% of patients had in-trathecal IgM synthesis. Of the 20 patients with intrathecal IgMsynthesis, 15 showed cMRI activity in the follow-up scan (OR2.861, CI 0.993–9.572; p = 0.064) (table 3 and figure 1). All 9DMT− patients with intrathecal IgM showed cMRI activityduring the first 12 months (OR 11.630, CI 0.544–248.843; p =0.116) but only 6 out of 11 in DMT+ patients (table 3).

We found a negative association of the IgA index with cMRIactivity in DMT+ patients and only a marginal positive effect in

DMT− patients (table 2). Intrathecal IgA synthesis accordingto Reiber formula was detected in only 8% of all patients (n =11) and was negatively associated with a higher risk for cMRIactivity regardless of treatment status (table 3 and figure 1).

In a subgroup of patients, we analyzed the IgG subclasses 1and 3 in CSF as well as respective indices (table 2).

We found that elevated IgG1 indices were positively associ-ated with cMRI progression (OR 1.391, CI 0.645–3.314; p =0.416). This effect was more pronounced in DMT− patients(OR 9.876, CI 0.503–194.019; p = 0.132) but not in DMT+patients (OR 0.780, CI 0.279–2.179; p = 0.636). ElevatedIgG1 levels in CSF were associated with a higher risk of cMRIactivity (OR 1.023, CI 1.003–1.048; p = 0.039). The effect wasstronger in DMT− patients than DMT+ patients (table 2).

In all patients, IgG3 indices were not associated with cMRIprogression (OR 1.001, CI 0.086–13.063; p = 0.999). In-dependent analysis of DMT+ and DMT− patients showeda positive association in untreated patients (OR 2.558, CI

Table 1 Baseline characteristics of patient cohorts

MS/CIS all MS/CIS DMT2 MS/CIS DMT+

Demographic data

No. of patients [n] 149 52 97

Sex female [n (%)] 101 (68) 32 (62) 69 (71)

Age [median (range)] 33 (18–55) 33.5 (19–52) 33 (18–55)

Median time symptom onset to cMRI [d (range)] 67 (1–345) 64.5 (1–230) 67 (1–345)

Patients with MRI progression [n (%)] 81 (54) 37 (71) 44 (45)

cMRI data

Lesion volume [median (mL) (range)] 2.106 (0.093–35.872) 1.489 (0.093–26.762) 2.383 (0.140–35.872)

Lesion number > 9 [n (%)] 93 (62) 26 (50) 67 (69)

Impaired blood CSF barrier [n (%)] 40 (27) 12 (23) 28 (29)

T2 lesion increment [median (range)] 1 (0–15) 1 (0–8) 0 (1–15)

CSF data

OCB + [n (%)] 116 (78) 38 (67) 78 (80)

Intrathecal IgG [n (%)] 75 (50) 25 (48) 50 (52)

Intrathecal IgM [n (%)] 20 (14) 9 (17) 11 (12)

Intrathecal IgA [n (%)] 11 (8) 7 (14) 4 (4)

IgG index [median (range)] 0.74 (0.25–3.20) 0.74 (0.41–2.69) 0.74 (0.25–3.20)

IgM index [median (range)] 0.08 (0.00–0.67) 0.08 (0.00–0.67) 0.08 (0.00–0.51)

IgA index [median (range)] 0.27 (0.00–0.80) 0.28 (0.19–0.65) 0.26 (0.00–0.80)

Cell count [median (/μl) (range)] 7 (0–126) 10 (0–126) 6 (0–46)

Abbreviations: CIS = clinically isolated syndrome; cMRI = cerebral MRI; DMT = disease-modifying therapy; OCB = oligoclonal band.Patients with MS/CIS were divided into 3 groups: all patients, patients with, and patients without DMT. These were compared regarding demographic, MRI,and CSF data.

4 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 62: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Table 2 Comparison of cMRI progression risk in patients with MS/CIS in regard to CSF laboratory findings

MS/CIS

All patients DMT2 DMT+

MRI+/MRI2 MRI+/MRI2 MRI+/MRI2

IgG index

Median 0.81/0.69 0.81/0.53 0.74/0.73

p Value 0.008 0.024 0.139

OR 3.441 32.021 2.090

CI 1.482–9.206 1.574–651.251 0.787–5.553

p Value of interaction 0.091

IgM index

Median 0.09/0.07 0.1/0.05 0.08/0.08

p Value 0.056 0.069a 0.464a

OR 43.370 5,970.034a 5.092a

CI 1.177–2,914.982 0508–70098464.401a 0.065–396.140a

p Value of interaction 0.145

IgA index

Median 0.27/0.26 0.28/0.25 0.26/0.26

p Value 0.884 0.990 0.847

OR 0.781 1.035 0.662

CI 0.027–23.249 0.004–254.431 0.010–44.463

p Value of interaction 0.899

IgG1 index

Median 0.63/0.62 0.72/0.58 0.6/0.66

p Value 0.416 0.132 0.636

OR 1.391 9.876 0.780

CI 0.645–3.314 0.503–194.019 0.279–2.179

p Value of interaction 0.115

IgG1 (mg/L)

Median 27.85/23.2 35/19.55 25.3/24.4

p Value 0.039 0.052 0.338

OR 1.023 1.059 1.012

CI 1.003–1.048 1.000–1.122 0.988–1.036

p Value of interaction 0.149

IgG3 index

Median 0.34/0.34 0.39/0.33 0.34/0.39

p value 0.999 0.581 0.385

OR 1.001 2.558 0.141

CI 0.086–13.063 0.091–71.672 0.002–11.719

p Value of interaction 0.305

IgG3 (mg/L)

Continued

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 5

Page 63: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

0.091–71.672; p = 0.581) but not DMT+ patients (OR 0.141,CI 0.002–11.719; p = 0.385).

OCBsOCBs were positively associated with the occurrence of cMRIactivity after 12 months in the entire cohort (OR 1.855, CI0.824–4.279; p = 0.139) and in particular in DMT− patients(9.327, CI 2.296–37.890; p = 0.002) but not DMT+ patients(OR 0.714, CI 0.260–1.960; p = 0.513) (table 3). Of 38DMT− patients with OCBs, 32 patients (84%) showed newlesions in the follow-up cMRIs, whereas only 5 out of 14patients without OCBs (36%) (figure 1).

CSF, NFL, and sCD27 levelsWe compared patients with and without cMRI activity duringthe first 12 months after diagnosis in regard to their levels ofsCD27 in CSF and found that elevated sCD27 levels wereassociated with a higher risk of developing new lesions incMRI (table 2). Patients with higher levels of sCD27 in CSFhad a higher risk of new cMRI lesions during the follow-upperiod (OR 1.552, CI 1.087–2.430; p = 0.030). The same

trend was seen in DMT− patients (OR 2.619, CI0.875–7.842; p = 0.085) and to a lower extent in DMT+patients (table 2). Interestingly, sCD27 levels were highlycorrelated with IgG index (r = 0.884, p = 1.916 × 10−20). CSFNFL levels showed a marginal positive association in ourcohort with new cMRI activity after 12 months regardless oftreatment status (table 2).

Multivariable analysis of CSF laboratory inall patientsWe performed a multivariable logistic regression analysis toidentify parameters that were mutually associated with MRIprogression. Because of sample size, we decided to focus onthe entire cohort and did not analyze DMT+ andDMT−patients independently. We also did not includesCD27 because data were only available on a subgroup ofpatients and sCD27 strongly correlated with the IgG index.IgG index, DMT, gender distribution, IgM index, and in-trathecal IgA synthesis were selected into the model as themost important variables (table 4). Higher IgG indices werefound to be associated with a higher risk of cMRI progression

Table 2 Comparison of cMRI progression risk in patients with MS/CIS in regard to CSF laboratory findings (continued)

MS/CIS

All patients DMT2 DMT+

MRI+/MRI2 MRI+/MRI2 MRI+/MRI2

Median 0.77/0.68 0.68/0.65 0.8/0.7

p Value 0.683 0.397 0.673

OR 1.130 1.601 0.826

CI 0.643–2.193 0.539–4.758 0.339–2.012

p Value of interaction 0.356

sCD27 (U/mL)b

Median 11.72/9.26 14.55/5.36 11.5/10.12

p Value 0.030 0.085 0.200

OR 1.552 2.619 1.335

CI 1.087–2.430 0.875–7.842 0.858–2.075

p Value of interaction 0.262

NFL (pg/mL)c

Median 1620.84/644.12 1959.55/676.52 1524.98/634.4

p Value 0.324 0.425 0.541

OR 1.033 1.047 1.025

CI 0.971–1.107 0.936–1.170 0.947–1.109

p Value of interaction 0.765

Abbreviations: CIS = clinically isolated syndrome; cMRI = cerebral MRI; DMT = disease-modifying therapy; NFL = neurofilament.Patients with MS/CIS were divided into 3 groups: all patients, patients with DMT (DMT+) and patients without DMT (DMT−). Patients were compared in theirrisk for cMRI progression when regarding levels of immunoglobulin indices and concentrations of IgG subclasses 1 and 3, sCD27, and NFL by logisticregression.a Corrected by Firth method.b For a 10 unit change.c For a 500 unit change.

6 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 64: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

(OR 3.144, CI 1.156–8.554; p = 0.025). As expected, DMTwas also associated with an decreased risk of cMRI activity.Surprisingly in our cohort, female gender was associated witha higher risk of cMRI activity. Mean IgM indices were higherin patients with cMRI progression than in patients without(OR 115.834, CI 0.923–14,531.160; p = 0.054). IntrathecalIgA synthesis was associated with less cMRI activity whenusing a multivariable analysis (OR 0.274, CI 0.058–1.292; p =0.10172). The ROC analysis resulted in an AUC of 0.734 (p <0.001), which underlines the prognostic benefit of thesevariables (figure 2A).

Prognostic value of cMRI vs CSF findingsTo emphasize the additional prognostic value of CSFparameters in comparison to cMRI data, we computed 2additional multivariable models. Gender, DMT, lesion vol-ume, and lesion count in the baseline cMRI resulted ina model with an AUC of 0.679. Combined analysis with CSFparameters such as OCBs, intrathecal IgG, IgM, IgA synthesis,

and their respective indices increased the prognostic strengthof the model to an AUC of 0.751 (p = 0.025) (figure 2B).

Quantitative prediction of cMRI activityTo estimate a patient’s individual risk for future cMRI ac-tivity, we computed a numeric risk scale using the re-gression parameters of our multivariable model (figure 3).Values of each factor are assigned to a specific number ofpoints and addition of these leads to a cumulative score.This score then can be converted to predict the patient’sprobability of future cMRI activity.

DiscussionIn the present study, we investigated the predictive value ofCSF parameters at the time of diagnosis of CIS and MS forthe development of new lesions on a follow-up cMRI after12 months.

Table 3 Risk of cMRI progression in patients with and without intrathecal immunoglobulin synthesis and OCBs

MS/CIS

All patients DMT2 patients DMT+ patients

MRI+/2 MRI+/2 MRI+/2

Patients with IgG synthesis [n (%)] 44 (54%)/31(46%) 23 (62%)/2(13%) 21 (48%)/29(55%)

p Value 0.225 0.003 0.468

OR 1.517 11.528 0.742

CI 0.776–2.995 2.237–59.399 0.332–1.661

p Value of interaction 0.003

Patients with IgM synthesis [n (%)] 15(19%)/5(8%) 9(24%)/0(0%) 6(14%)/5(10%)

p Value 0.064 0.116a 0.511a

OR 2.861 11.630a 1.533a

CI 0.993–9.572 0.544–248.843a 0.428–5.489a

p Value of interaction 0.135

Patients with IgA synthesis [n (%)] 6(7%)/5(8%) 5(14%)/2(14%) 1(2%)/3(6%)

p Value 0.516 0.937 0.390

OR 0.648 0.931 0.364

CI 0.171–2.496 0.159–5.471 0.036–3.649

p Value of interaction 0.527

Patients with CSF OCBs [n (%)] 66(82%)/50(74%) 32(86%)/6(40%) 34(77%)/44(83%)

p Value 0.139 0.002 0.513

OR 1.855 9.327 0.714

CI 0.824–4.279 2.296–37.890 0.260–1.960

p Value of interaction 0.004

Abbreviations: CIS = clinically isolated syndrome; cMRI = cerebral MRI; DMT = disease-modifying therapy; OCB = oligoclonal band.Patients with MS/CIS were divided into 3 groups: all patients, patients with DMT (DMT+), and patients without DMT (DMT−). Patients were compared in theirrisk for cMRI progression when regarding intrathecal immunoglobulin G, M, and A synthesis and presence of OCBs by logistic regression.a Corrected by Firth method.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 7

Page 65: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Routine diagnostic workup of patients with CIS/MS usuallyentails the detection of IgG either in form of OCBs orquantitative intrathecal synthesis as shown by IgG index(usually > 0.7) or Reiber formula. Many labs also measureIgM and IgA levels in CSF. As mentioned above in particularintrathecal IgG synthesis seems to be associated with earlierinflammatory disease activity leading to our inclusion ofthese parameters in this study. sCD27 is the soluble versionof the surface molecule CD27 that belongs to the TNF-receptor family and can be found on B- and T-cells. It playsan important role in B-cell differentiation and immuno-globulin synthesis. Because patients with MS were found tohave higher levels of the soluble version of this molecule,20

we also included this parameter in our analysis. The sameapplies to NFL, which has been shown to be associated withneuronal damage and brain atrophy in patients withrelapsing-remitting MS.12,13,21–23

We demonstrate that soluble CD27 levels in CSF, IgG1 levels,IgG index, and possibly IgM index at the time of diagnosis areassociated with cMRI activity during the next 12 months inour retrospective study. In the subgroup of patients who hadnot been treated with DMT, sCD27, IgG1 levels, OCBs inCSF, IgG and possible IgM indices are associated with futurecMRI activity. In the multivariable analysis, IgG index, IgMindex, and intrathecal IgA synthesis, besides DMT treatmentstatus and gender, are the main variables predicting futurecMRI activity in our cohort.

Until now, CSF examination in patients presenting withsuspected CIS or MS mainly serves diagnostic purposes,which might be at least partially due to the lack of studiessystematically analyzing CSF parameters in patients present-ing with a first clinical event suggestive of MS. Most studiesconducted in the past dealing with a potentially prognosticvalue of CSF parameters did not discriminate betweenpatients who started DMT during follow-up and those whodid not. In our cohort, DMT has an influence on cMRI ac-tivity even within a short follow-up period of 12 months,which is in line with many previous studies.1,24–28 For this

Figure 1 Intrathecal immunoglobulin synthesis and OCBs in CSF

The number of patients with and without intrathecal IgG, IgM,and IgA synthesis and oligoclonal bands are shown and com-pared regarding cMRI progression during the first 12 months.White columns = patients with new lesions. Black columns =patients without new lesions. (A) All patients, (B) only patientswithout DMT, and (C) only patients receiving DMT. DMT = dis-ease-modifying therapy; OCB = oligoclonal band.

Table 4 Multivariable analysis of CSF laboratory findingsin all patients

p Value OR CI

IgG index 0.025 3.144 1.156–8.554

DMT 0.002 0.249 0.105–0.594

Gender 0.028 2.506 1.107–5.672

IgM index 0.054 115.834 0.923–14,531.16

Intrathecal IgA synthesis 0.102 0.274 0.058–1.292

Abbreviations: CIS = clinically isolated syndrome; DMT = disease-modifyingtherapy.Potentially prognostic biomarkers in patients with MS/CIS were analyzed bymultivariable logistic regression to obtain p-values, OR, and CI.

8 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 66: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

reason, we decided to discriminate between patients with andwithout DMT to exclude the effect of this confounding factor.

Moreover, in the studies conducted so far, progression has beenmainly defined by clinical outcome parameters, and follow-uptime periods varied largely even within patient cohorts. Asclinical relapses are sometimes challenging to discern preciselyfrom anamnestic information in clinical routine and calculationof Expanded Disability Status Scale (EDSS) scores (especiallyof lower values) varies from day to day depending on a patient’sphysical state as well as the individual examiner, we concludedthat MRI offers a more sensitive and objective parameter todetect disease activity, especially when MRI scanning is per-formed in a highly standardized fashion.

Overall our findings are in line with previous studies sug-gesting a predictive role of intrathecal IgG synthesis, IgMsynthesis, and the presence of OCBs for predicting futureclinical disease activity. The presence of OCBs in the CSF isa known risk factor for conversion to MS in patients withCIS.1,2,5 The same applies to quantitative measures of IgG andIgM synthesis in the CSF compartment.29 Little is knownabout how IgG1 and IgG3 levels in serum and CSF relate tofuture disease activity. Increased IgG1 levels in CSF are found inpatients with CIS and MS.30,31 We observed an association ofIgG1 levels in CSF and future MRI activity. However, thisanalysis was based on only a smaller subset of our cohort and theIgG1 index itself may have shown a positive association withcMRI activity but failed to reach the level of significance. A recent

study reported an association of higher serum IgG3 levels withfaster conversion to MS in patients with CIS.32 However, thisstudy was based on a small cohort of patients. We did observea slight positive association of IgG3 levels in CSF with MRIactivity; however, once again this trend failed to reach signifi-cance. Also sCD27 levels in CSF seem to be associated withconversion to MS and future relapse activity in adults and chil-drenwithCIS.9,33Our findings of an association between sCD27in CSFwithMRI activity are in line with these studies. However,IgG index and sCD27 levels in CSF are highly correlated sug-gesting that they are not independent biomarkers and ratherdepict markers of humoral inflammation in the CNS compart-ment. Most studies reported NFL in CSF or serum as a pre-dictive marker for future disease activity.16,34 Although weobserved a trend in the same direction, the effect was small in ourstudy and did not reach the level of significance, which might bedue to the low number of patients in the NFL analysis.

In summary, we provide evidence that IgG index and to a lesserextent IgM and IgA index are helpful markers to predict futureMRI activity in particular if they are combined with other vari-ables such as DMT status and gender distribution. Establishingreliable risk scores for patients with newly diagnosed CIS or MSmight be very helpful to guide early treatment decisions. Sucha risk scoremight include other biomarkers (e.g., microRNA andchitinase-like protein14,16,17) or imaging parameters (e.g., spinalMRI35). As our cohort only entails a small number of patients,further studies with larger cohorts are warranted to validate thesefindings.

Figure 2 Multivariable analysis of CSF parameters in regard to cMRI progression

(A) Analysis of potentially prognostic biomarkers inpatients with MS/CIS throughmultivariable logistic regression lead to selection of IgG index, DMT, genderdistribution, IgM index, and intrathecal IgA synthesis. A ROC analysis resulted in an AUC of 0.734 (p < 0.001). (B) The prognostic value of cMRI parametersincluding cMRI lesion load and lesion volume and baseline information such as gender distribution and DMT were compared to the additional prognosticvalue when including CSF parameters and shown in an ROC curve. AUC of MRI parameters alone was 0.679, AUC of MRI and CSF parameters combined was0.751 (p = 0.03). AUC = area under the curve; CIS = clinically isolated syndrome; cMRI = cerebral MRI; DMT = disease-modifying therapy; ROC = receiveroperating characteristic.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 9

Page 67: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

AcknowledgmentThe authors express their deep thankfulness to all contrib-utors of the study, especially the study nurses, for theirmotivated collaboration and recruitment efforts, all patientsfor their participation.

Study fundingThe study was supported by grants from the German FederalMinistry for Education and Research Competence networkMultiple Sclerosis (BMBF, 01GI1601D) and DiFuture (DataIntegration for Future Medicine, BMBF 01ZZ1804[A-F]).Biosamples were provided by the biobank of the Departmentof Neurology as part of the Joined Biobank Munich in theframework of the German Biobank node. Bernhard Hemmerreceived funding for the study from the Multiple MS EUConsortium and the SFB CRC 128.

DisclosureA. Klein and R. Selter reports no disclosures. A. Hapfel-meier received a honorarium from Biogen for consultingservices for the Biogen Symposium on Statistical Methodsin Real World Evidence 2017. A. Berthele reports grantsfrom Bayer HealthCare, personal fees from Biogen, MerckSerono, Teva, Novartis, and Genzyme, and compensationsfor clinical trials from Biogen, Novartis, Genzyme, Roche,Teva, and Alexion Pharmaceuticals. B. Muller-Myhsokreports no disclosures. V. Pongratz received researchfunding from Novartis (Oppenheim Forderpreis 2017). C.Gasperi reports no disclosures. C. Zimmer served on thescientific advisory board for Philips, received speakerhonoraria from Bayer Schering, Philips, serves as co-editor

on the advisory board of Clinical Neuroradiology. M.Muhlau received research support from Merck Serono andNovartis as well as travel expenses for attending meetingsfrom Merck Serono. B. Hemmer has served on scientificadvisory board for Novartis; he has served as DMSCmember for AllergyCare and TG Therapeutics; he or hisinstitution have received speaker honoraria from Desitin;holds part of 2 patents, one for the detection of antibodiesagainst KIR4.1 in a subpopulation of patients with MS andone for genetic determinants of neutralizing antibodies tointerferon β. All disclosures were outside the submittedwork. Go to Neurology.org/NN for full disclosures.

Publication historyReceived byNeurology: Neuroimmunology &Neuroinflammation June 28,2018. Accepted in final form March 21, 2019.

Figure 3 Risk calculator (nomogram) for future cMRI activity in patients with MS/CIS

Potential risk factors such as IgG index, DMT, gender, IgM index, and intrathecal IgA synthesis were converted to numeric values to calculate a combined riskscore to the evaluate individual risk of future cMRI activity. (A) Scoring of individual risk factors; (B) combined risk score and risk of future cMRI activity. CIS =clinically isolated syndrome; cMRI = cerebral MRI; DMT = disease-modifying therapy.

Appendix Authors

Authors Location Role Contribution

Ana Klein,MD

TechnischeUniversitatMunchen,Germany

Author Designed andconceptualized thestudy; performed dataacquisition, analysis,and interpretation;drafted the manuscriptfor intellectual content

Rebecca C.Selter, MD

TechnischeUniversitatMunchen,Germany

Author Performed dataacquisition, analysis,and interpretation;revised the manuscriptfor intellectual content

10 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 68: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

References1. Tintore M, Rovira A, Rio J, et al. Defining high, medium and low impact prognostic

factors for developing multiple sclerosis. Brain 2015;138:1863–1874.2. Kuhle J, Disanto G, Dobson R, et al. Conversion from clinically isolated syndrome to

multiple sclerosis: a large multicentre study. Mult Scler 2015;21:1013–1024.3. Fisniku LK, Brex PA, AltmannDR, et al. Disability and T2MRI lesions: a 20-year follow-

up of patients with relapse onset of multiple sclerosis. Brain 2008;131:808–817.4. Tintore M, Rovira A, Rio J, et al. Do oligoclonal bands add information to MRI in first

attacks of multiple sclerosis? Neurology 2008;70:1079–1083.5. Dobson R, Ramagopalan S, Davis A, Giovannoni G. Cerebrospinal fluid oligoclonal

bands in multiple sclerosis and clinically isolated syndromes: a meta-analysis of preva-lence, prognosis and effect of latitude. J Neurol Neurosurg Psychiatry 2013;84:909–914.

6. Uher T, Horakova D, Tyblova M, et al. Increased albumin quotient (QAlb) in patientsafter first clinical event suggestive of multiple sclerosis is associated with developmentof brain atrophy and greater disability 48 months later. Mult Scler 2016;22:770–781.

7. Ferraro D, Simone AM, Bedin R, et al. Cerebrospinal fluid oligoclonal IgM bandspredict early conversion to clinically definite multiple sclerosis in patients with clin-ically isolated syndrome. J Neuroimmunol 2013;257:76–81.

8. Durante L, Zaaraoui W, Rico A, et al. Intrathecal synthesis of IgM measured aftera first demyelinating event suggestive of multiple sclerosis is associated with sub-sequent MRI brain lesion accrual. Mult Scler 2012;18:587–591.

9. van der Vuurst de Vries RM, Mescheriakova JY, Runia TF, Jafari N, Siepman TA,Hintzen RQ. Soluble CD27 levels in cerebrospinal fluid as a prognostic biomarker inclinically isolated syndrome. JAMA Neurol 2017;74:286–292.

10. NorgrenN, Sundstrom P, Svenningsson A, Rosengren L, Stigbrand T, GunnarssonM.Neurofilament and glial fibrillary acidic protein in multiple sclerosis. Neurology 2004;63:1586–1590.

11. Disanto G, Adiutori R, Dobson R, et al. Serum neurofilament light chain levels areincreased in patients with a clinically isolated syndrome. J Neurol Neurosurg Psy-chiatry 2016;87:126–129.

12. Kuhle J, Barro C, Disanto G, et al. Serum neurofilament light chain in early relapsingremitting MS is increased and correlates with CSF levels and with MRI measures ofdisease severity. Mult Scler 2016;22:1550–1559.

13. Hakansson I, Tisell A, Cassel P, et al. Neurofilament light chain in cerebrospinal fluidand prediction of disease activity in clinically isolated syndrome and relapsing-remitting multiple sclerosis. Eur J Neurol 2017;24:703–712.

14. Bergman P, Piket E, Khademi M, et al. Circulating miR-150 in CSF is a novel can-didate biomarker for multiple sclerosis. Neurol Neuroimmunol Neuroinflamm 2016;3:e219. doi: 10.1212/NXI.0000000000000573.

15. Villar L, Garcia-Barragan N, EspinoM, et al. Influence of oligoclonal IgM specificity inmultiple sclerosis disease course. Mult Scler 2008;14:183–187.

16. Sellebjerg F, Royen L, Soelberg Sorensen P, Oturai AB, Jensen PEH. Prognostic valueof cerebrospinal fluid neurofilament light chain and chitinase-3-like-1 in newly di-agnosed patients with multiple sclerosis. Mult Scler 2018:1352458518794308.

17. Mollgaard M, Degn M, Sellebjerg F, Frederiksen JL, Modvig S. Cerebrospinal fluidchitinase-3-like 2 and chitotriosidase are potential prognostic biomarkers in earlymultiple sclerosis. Eur J Neurol 2016;23:898–905.

18. Reiber H, Felgenhauer K. Protein transfer at the blood cerebrospinal fluid barrier andthe quantitation of the humoral immune response within the central nervous system.Clinica Chim Acta Int J Clin Chem 1987;163:319–328.

19. Schmidt P, Gaser C, Arsic M, et al. An automated tool for detection of FLAIR-hyperintense white-matter lesions in Multiple Sclerosis. Neuroimage 2012;59:3774–3783.

20. Hintzen RQ, van Lier RA, Kuijpers KC, et al. Elevated levels of a soluble form of theT cell activation antigen CD27 in cerebrospinal fluid of multiple sclerosis patients.J Neuroimmunol 1991;35:211–217.

21. Varhaug KN, Barro C, Bjornevik K, et al. Neurofilament light chain predicts diseaseactivity in relapsing-remitting MS. Neurol Neuroimmunol Neuroinflamm 2018;5:e422. doi: 10.1212/NXI.0000000000000422.

22. Siller N, Kuhle J, Muthuraman M, et al. Serum neurofilament light chain is a bio-marker of acute and chronic neuronal damage in early multiple sclerosis. Mult Scler2018:1352458518765666.

23. Kuhle J, Nourbakhsh B, Grant D, et al. Serum neurofilament is associated withprogression of brain atrophy and disability in early MS. Neurology 2017;88:826–831.

24. Comi G, Filippi M, Barkhof F, et al. Effect of early interferon treatment on con-version to definite multiple sclerosis: a randomised study. Lancet 2001;357:1576–1582.

25. Comi G, Martinelli V, Rodegher M, et al. Effects of early treatment with glatirameracetate in patients with clinically isolated syndrome. Mult Scler 2013;19:1074–1083.

26. Radue EW, O’Connor P, Polman CH, et al. Impact of fingolimod therapy onmagneticresonance imaging outcomes in patients with multiple sclerosis. ArchNeurol 2012;69:1259–1269.

27. Miller DH, Soon D, Fernando KT, et al. MRI outcomes in a placebo-controlled trial ofnatalizumab in relapsing MS. Neurology 2007;68:1390–1401.

28. Wolinsky JS, Narayana PA, Nelson F, et al. Magnetic resonance imaging outcomesfrom a phase III trial of teriflunomide. Mult Scler 2013;19:1310–1319.

29. Huss A, Abdelhak A, Halbgebauer S, et al. Intrathecal immunoglobulin M production:a promising high-risk marker in clinically isolated syndrome patients. Ann Neurol2018;83:1032–1036.

30. Greve B, Magnusson CG, Melms A, Weissert R. Immunoglobulin isotypes reveala predominant role of type 1 immunity in multiple sclerosis. J Neuroimmunol 2001;121:120–125.

31. Di Pauli F, Gredler V, Kuenz B, et al. Features of intrathecal immunoglobulins inpatients with multiple sclerosis. J Neurol Sci 2010;288:147–150.

32. Trend S, Jones AP, Cha L, et al. Higher serum immunoglobulin G3 levels may predictthe development of multiple sclerosis in individuals with clinically isolated syndrome.Front Immunol 2018;9:1590.

33. Wong YYM, van der Vuurst de Vries RM, van Pelt ED, et al. T-cell activation markersCD27 is associated with clinically definite multiple sclerosis in childhood-acquireddemyelinating syndromes. Mult Scler 2018;24:1715–1724.

34. Arrambide G, Espejo C, Eixarch H, et al. Neurofilament light chain level is a weak riskfactor for the development of MS. Neurology 2016;87:1076–1084.

35. Zeydan B, Gu X, Atkinson EJ, et al. Cervical spinal cord atrophy: an early marker ofprogressive MS onset. Neurol Neuroimmunol Neuroinflamm 2018;5:e435. doi: 10.1212/NXI.0000000000000435.

Appendix (continued)

Authors Location Role Contribution

AlexanderHapfelmeier,PhD

TechnischeUniversitatMunchen,Germany

Author Performed data analysisand interpretation;revised the manuscriptfor intellectual content

AchimBerthele, MD

TechnischeUniversitatMunchen,Germany

Author Revised the manuscriptfor intellectual content

BertramMuller-Myhsok, PhD

Max-Planck-Institute ofPsychiatry,Munchen,Germany

Author Aided in data analysis;revised the manuscriptfor intellectual content

ViolaPongratz, MD

TechnischeUniversitatMunchen,Germany

Author Revised the manuscriptfor intellectual content;aided in data acquisition

ChristianeGasperi, MD

TechnischeUniversitatMunchen,Germany

Author Revised the manuscriptfor intellectual content

ClausZimmer, MD

TechnischeUniversitatMunchen,Germany

Author Revised the manuscriptfor intellectual content

MarkMuhlau, MD

TechnischeUniversitatMunchen,Germany

Author Revised the manuscriptfor intellectual content;aided in data acquisition

BernhardHemmer, MD

TechnischeUniversitatMunchen,Germany

Author Designed andconceptualized thestudy; performed dataanalysis andinterpretation; draftedthe manuscript forintellectual content

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 11

Page 69: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

ARTICLE OPEN ACCESS

Natalizumab treatment reduces microglialactivation in the white matter of the MS brainMarcus Sucksdorff, MD, Jouni Tuisku, MSc, Markus Matilainen, PhD, Anna Vuorimaa, BM, Sarah Smith, BSc,

Joonas Keitila, BM, Johanna Rokka, MSc, PhD, Riitta Parkkola, MD, PhD, Marjo Nylund, MSc,

Juha Rinne, MD, PhD, Eero Rissanen, MD, PhD, and Laura Airas, MD, PhD

Neurol Neuroimmunol Neuroinflamm 2019;6:e574. doi:10.1212/NXI.0000000000000574

Correspondence

Dr. Sucksdorff

[email protected]

AbstractObjectiveTo evaluate whether natalizumab treatment reduces microglial activation in MS.

MethodsWe measured microglial activation using the 18-kDa translocator protein (TSPO)-bindingradioligand [11C]PK11195 and PET imaging in 10 patients with MS before and after 1 yeartreatment with natalizumab.Microglial activation was evaluated as the distribution volume ratio(DVR) of the specifically bound radioligand in brain white and gray matter regions of interest.MRI and disability measurements were performed for comparison. Evaluation was performedidentically with 11 age- and sex-matched patients with MS who had no MS therapy.

ResultsNatalizumab treatment reduced microglial activation in the normal-appearing white matter(NAWM; baseline DVR vs DVR after 1 year of treatment 1.25 vs 1.22, p = 0.014, Wilcoxon)and at the rim of chronic lesions (baseline DVR vs DVR after 1 year of treatment 1.24 vs 1.18, p= 0.014). In patients withMSwith no treatment, there was an increase inmicroglial activation atthe rim of chronic lesions (1.23 vs 1.27, p = 0.045). No alteration was observed in microglialactivation in gray matter areas. In the untreated patient group, higher microglial activation atbaseline was associated with more rapid disability progression during an average of 4 years offollow-up.

ConclusionsTSPO-PET imaging can be used as a tool to assess longitudinal changes in microglial activationin the NAWM and in the perilesional areas in the MS brain in vivo. Natalizumab treatmentreduces the diffuse compartmentalized CNS inflammation related to brain resident innateimmune cells.

From the Turku PET Centre (M.S., J.T., M.M., A.V., S.S., J.K., J. Rokka, M.N., J. Rinne, E.R., L.A.), Turku University Hospital and University of Turku; Division of Clinical Neurosciences (M.S.,M.N., J. Rinne, E.R., L.A.), Turku University Hospital and University of Turku; and Department of Radiology (R.P.), University Hospital and University of Turku, Finland.

Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

Statistical analysis was conducted byMarkusMatilainen, PhD, Marcus Sucksdorff, MD, and Jouni Tuisku, MSc, from Turku PET Centre, Turku University Hospital and University of Turku.

The Article Processing Charge was funded by Turku University Hospital/Research Services.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 70: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Neuropathologic studies have demonstrated microglial acti-vation in the normal-appearing white matter (NAWM) and atthe rim of chronic active/smoldering lesions in brains ofpatients with MS.1 The more advanced the disease, the moremicroglial activation has been seen.1 This observation hasbeen confirmed in in vivo PET studies of patients with MSusing 18-kDa translocator protein (TSPO)-binding radio-ligands targeting activated microglia.2–6 In these cross-sectional studies, TSPO binding in the NAWM or at thechronic active lesion rim has been significantly strongeramong patients with secondary progressive MS (SPMS)compared with patients with relapsing-remitting MS (RRMS)or healthy controls.2–6

Transformation of relapsing MS into SPMS is clinically in-sidious and is known to be pathologically associated withCNS-resident innate immune cell activation.7 Microglia andmacrophages have protective and homeostatic functions, suchas clearing of debris. In acute MS lesions, they uptake thedegraded myelin, which promotes remyelination.8 In pro-gressive MS, microglia may instead acquire a harmful proin-flammatory phenotype and promote neurotoxicity bysecreting proinflammatory cytokines and free oxygen radi-cals.9 This may be one of the mechanisms driving chronictissue damage in areas surrounding chronic active/smolderinglesions, and this is believed to be one of the mechanismscontributing to clinical disease progression.8 It is presentlyunknown whether blocking inflammatory activity with effec-tive disease-modifying therapy alters microglia toward a lessactivated phenotype at sites related to progressive MS pa-thology. In this work, we addressed this question by per-forming an in vivo TSPO-PET imaging study in which wemeasured microglial activation longitudinally at 1-year in-terval in patients with MS with or without natalizumabtreatment.

MethodsStudy subjectsThe study subjects were initially recruited as part of aninvestigator-initiated substudy to the ASCEND study at theTurku University Hospital, Finland.10 ASCEND evaluatedthe effect of natalizumab on disease progression in SPMS.10

Because of the low number (n = 4) of study subjects par-ticipating in ASCEND at our center, patients with MS werealso recruited for PET imaging from the Turku neurologyoutpatient clinic. The ASCEND patients were treated andfollowed up in a blinded fashion. The rest of the patients wereevaluated clinically in an open-label fashion, with the image

analyses performed in a blinded fashion. The final composi-tion of the study cohort of 21 patients withMS is shown in thetable, top section. Exclusion criteria included corticosteroidtreatment within 30 days of evaluation, active neurologic orautoimmune disease other than MS or another comorbidityconsidered significant, inability to tolerate PET or MRI, anda current or desired pregnancy following study enrollment.None of the patients were JCV-Ab positive. TSPO-PETresults from 8 similarly imaged age- and sex-matched healthycontrols were used for comparison.

ProceduresTen patients were imaged using PET and MRI at baselinebefore initiation of natalizumab treatment and 1 year later.Eleven patients without any disease-modifying treatmentwere imaged for comparison using the same protocol. Ex-panded Disability Status Scale (EDSS) evaluation and re-cording of relapses were performed at baseline and after 1 year(table, middle section). Safety monitoring was performedregularly at 1-month intervals. Imaging was performed during2013–2017, and all patients were reevaluated for the EDSSscore in May 2018. Importantly, no adverse effects were ob-served during the follow-up period of patients.

MRI and data analysisFor the evaluation of MS pathology and for the acquisition ofanatomic reference for the PET images, conventional MRIwas performed with a 3-T Ingenuity TF PET/MR scanner(Philips). MRI of the healthy control group was performedwith a Gyroscan Intera 1.5-T Nova Dual scanner for anatomicreference for the PET images (Philips). The following MRIsequences were used for image acquisition: axial T2,3-dimensional (3D) fluid-attenuated inversion recovery, 3DT1, and 3D T1 with gadolinium enhancement.

For each patient, the T1 image at the first time point wascoregistered in statistical parametric mapping (SPM8, version8; Wellcome Trust Centre for Neuroimaging) to the sumimage of realigned PET frames of the first session. All theother MRIs from the first session and the 1-year follow-upwere then coregistered to the T1 image of the first session. Foreach time point, the MS lesions were identified using theLesion Segmentation Tool (statistical-modelling.de/lst.html,a toolbox running in SPM8),11 as described previously.5 Theresulting lesion mask images were used to fill the corre-sponding T1 image with the lesion-filling tool in the LesionSegmentation Tool. The filled T1 was then used for seg-menting gray matter (GM) andWM volumes with FreeSurfer5.3 software (surfer.nmr.mgh.harvard.edu/).

Glossary3D = 3 dimensional; 9HPT = 9-Hole Peg Test; BPND = binding potential;DVR = distribution volume ratio; EDSS = ExpandedDisability Status Scale; GM = gray matter; IQR = interquartile range; NAWM = normal-appearing white matter; ROI = regionof interest; RRMS = relapsing-remitting MS; SPMS = secondary progressive MS; TSPO = translocator protein.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 71: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Table Demographic information and imaging parameters of the study cohort

Natalizumab (n = 10) No treatment (n = 11) p Value

Demographic information of the study patients

SPMS (n = 16, 76%) 6 (60%) 10 (91%) 0.097

RRMS (n = 5, 24%) 4 (40%) 1 (9%)

Age, y 49.15 (±9.33) 47.88 (±10.23) 0.700

Female 6 (60%) 7 (64%) 0.864

Male 4 (40%) 4 (36%)

Years since SPMS diagnose 3.58 (±1.82, n = 6) 3.44 (±2.29, n = 10) 0.830

No. of Gd+ lesions 0 0

Years since MS onset 13.82 (±8.63) 15.11 (±6.93) 0.600

EDSS score (median and IQR) 3.5 (2.62–3.88) 6.0 (3.75–6.0) 0.036a

No. of relapses 1 year before 0.70 (±0.82) 0.27 (±0.47) 0.220

Natalizumabbaseline

Natalizumabfollow-up

Natalizumabchangeb

Notreatmentbaseline

Notreatmentfollow-up

Notreatmentchangeb

Baseline: notreatment vsnatalizumabb

Follow-up: notreatment vsnatalizumabb

Results relatedto conventionalMRI parametersand clinicaloutcomemeasures

EDSS score 3.55 3.80 0.037a 5.09 5.32 0.089 0.036a 0.051

T2 lesionload (cm3)

14.74 13.31 0.262 28.87 27.38 0.351 0.018a 0.018a

T1 lesionload (cm3)

13.86 12.35 0.103 26.81 25.40 0.197 0.022a 0.010a

GM volume(cm3)

437.32 430.88 0.032a 433.09 425.05 0.023a 0.916 0.751

NAWMvolume(cm3)

443.68 443.22 0.683 425.69 423.72 0.689 0.504 0.418

Relapseswithin 1year

0.70 0.20 0.089 0.27 0.45 0.424 0.220 0.397

Natalizumab (n = 10) No treatment (n = 11) p Value

Microglial activation measured as [11C]PK11195 distributionvolume ratio at baseline

NAWM 1.25 1.25 1.000

GMcortex 1.25 1.22 0.477

Thalamus 1.36 1.38 0.699

Putamen 1.30 1.31 1.000

T1 1.19 1.19 0.751

T2 1.18 1.18 0.916

Lesion rim 3–6 mm 1.24 1.23 0.916

Abbreviations: DVR = distribution volume ratio; EDSS = Expanded Disability Status Scale; Gd+ = gadolinium enhancing positive; GM = gray matter; NAWM =normal-appearing white matter; RRMS = relapsing-remitting MS; SPMS = secondary progressive MS.Values are the mean (±SD where indicated) of the variables unless otherwise stated.a Statistically significant group difference at a level of p < 0.05.b p Value.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 72: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

The volumes of T2 lesion masks acquired with the LesionSegmentation Tool were used for the T2 lesion load evalua-tion. In addition, for each patient, the Lesion SegmentationTool masks at each time point were combined to a unifiedlesion region of interest (ROI). An average filled T1 imagewas also calculated from the filled T1 images of all MRI ses-sions, and it was used for ROI delineation for the cerebellum,striatum, thalamus, WM, and cortical GM with FreeSurfer.Furthermore, an NAWM ROI was created by removing thelesion ROI from theWMROI. Perilesional 3–6mmROI witha distance of 3–6mm to the lesionmask border was created bydilating the lesion mask image 3 mm and 6 mm and thenremoving the core from the resulting image. All ROIs werechecked and manually corrected slice by slice with both T1-hypointense and T2-hyperintense lesion mask images to ac-curately correspond to the MS lesions.

[11C]PK11195 radioligand production and PETThe radiochemical synthesis of [11C]PK11195 was per-formed as described previously.5 The mean injected dose was475.5 ± 53.1 MBq (mean ± SD) for natalizumab-treatedpatients with MS, 467.0 ± 73.1 MBq for untreated patientswith MS, and 478.6 ± 17.7 MBq for the healthy controls withno significant dose differences between the groups.

PET was performed with a brain-dedicated High-ResolutionResearch Tomograph scanner (Siemens/Control TechnologyIncorporated, Knoxville, TN) with an intrinsic spatial resolutionof approximately 2.5 mm.12 First, a 6-minute transmission scanfor attenuation correction was obtained using a 137Cs pointsource. Thereafter, 60-minute dynamic imaging was started si-multaneously with IV bolus injection of the radioligand. Headmovements were minimized using a thermoplastic mask.

PET analysisPET images were reconstructed using 17 time frames as de-scribed previously.5 The dynamic data were then smoothed usinga Gaussian 2.5-mm postreconstruction filter.5 Possible displace-ments between frames were corrected using mutual informationrealignment in SPM8. For each patient, the PET image from thesubsequent session was coregistered to the PET image of the firstsession using the sum images of each session. Finally, all imageswere resliced to match an MR voxel size of 1 mm.

Microglial activation was evaluated as specific binding of[11C]PK11195 using the distribution volume ratio (DVR).For the estimation of the [11C]PK11195 DVR, the time-activity curve corresponding to a reference region devoid ofspecific TSPO binding was acquired for each PET sessionusing a supervised cluster algorithm with 4 predefined kinetictissue classes (SuperPK software).13 The reference tissue-input Logan method, with a 20- to 60-minute time interval,was applied to the regional time-activity curves using the su-pervised cluster algorithm gray reference input.

In addition, voxel-wise parametric binding potential (BPND)maps were calculated using basis function implementation of

SRTM14 with 250 basis functions. Lower and upper boundsfor theta were set to 0.06 1/min and 0.8 1/min. The resultingparametric maps were normalized to MNI space (MontrealNeurological Institute database) in SPM8, after which thedifference BPND images between PET scans (1 year–baseline)were calculated for statistical parametric mapping of differ-ences between the 2 groups. The BPND images were trans-formed to DVR (DVR = BPND +1) for illustration purposes.

Statistical analysisStatistical analyses were performed using R (version 3.5.2).Variables are reported as median (interquartile range [IQR]).Mainly nonparametric tests were used because of the lownumber of subjects in both groups. The analysis of differencesin demographic variables in the 2 groups was conducted usingtheWilcoxon rank-sum test; for the differences in proportionsof patients with SPMS and females, a χ2 test was used. Thedifferences in DVR values and MRI parameters in the 2groups were compared using theWilcoxon rank-sum test. Thechanges of DVR values and MRI parameters between follow-up and baseline were analyzed separately in both treatmentgroups using the Wilcoxon signed-rank test. All correlationsare Spearman correlations, as the relationships between thevariables could not be seen as linear. The partial correlationsbetween the variables (EDSS change and DVR values) wereadjusted for the time between the EDSS measurements asfollows. The variables were regressed on the time difference,and the residuals of the model, i.e., the part in which the timedifference impact has been removed, were used to calculatethe partial correlations. A p value less than 0.05 from a 2-tailedtest was considered statistically significant for all analyses.

Standard protocol approvals, registrations,and patient consentsThe study was approved by the Ethics Committee of theHospital District of Southwest Finland. All participants pro-vided written informed consent according to the Declarationof Helsinki.

Data availability statementThe raw data used in preparation of the article will be sharedin anonymized format by request of a qualified investigator.

ResultsThe demographic characteristics, conventional MRI data, andclinical outcome measures of the patients are shown in thetable, top section. In total, 21 patients took part in the study,and 16 (76%) of them had SPMS. Of the 10 natalizumab-treated patients, 2 were treatment naive, and the rest wereswitching from other therapies, which included fingolimodand interferon beta-1a. The median time since previoustreatment was 2 months (IQR 1–3months). The second PETand MRI images were obtained 54.6 weeks (median, IQR53.0–60.1 weeks) after the baseline imaging amongnatalizumab-treated patients and 54.0 weeks (median, IQR

4 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 73: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

53.0–64.1 weeks) among the untreated patients. The reasonfor starting natalizumab was treatment failure with a previousdrug, based on either clinical relapses or high or increasingMRI lesion burden, or initial aggressive disease in case of thetreatment-naive patients.

In the untreated group, the median time since previoustreatment was 16 months (IQR 4.5–27.5 months). None ofthe patients had gadolinium-enhancing lesions in eitherMRI scan. The untreated group had higher T1 and T2lesion loads compared with the natalizumab-treated group(table, middle section). The natalizumab-treated group,which also included patients with RRMS, had a lowermedian EDSS score at baseline (table, top section). Com-pared with a group of age-matched healthy control persons(n = 8, historical data obtained and analyzed similarly to themethodology in this study), the patients with MS partici-pating in this study had significantly more microglial acti-vation in the NAWM (p = 0.026) and thalamus (p = 0.005)at baseline (figure 1A). In other ROIs, no significant dif-ferences were observed between the groups (data notshown). No significant differences were observed in thebaseline DVRs between the natalizumab-treated and un-treated groups in any brain area studied (table, bottomsection).

Natalizumab treatment reduced microglial activation in theNAWM. This was demonstrated as a significantly smallerDVR of the TSPO radioligand binding after treatment vsbefore treatment (1.22 vs 1.25; p = 0.014, Wilcoxon, figure1B). The same was true for the 3- to 6-mm perilesional area(1.18 vs 1.24; p = 0.014, figure 1B). Similarly, there wasa decrease in microglial activation following natalizumabtreatment in focal inflammatory lesions (figure 1B).

Contrary to the natalizumab-treated patients, we observedan increase in microglial activation during 1-year follow-upof patients with MS with no treatment. Here, the DVR of theTSPO radioligand binding increased from 1.23 to 1.27 (p =0.045) in the 3- to 6-mm perilesional area, and in theNAWM, there was a trend toward the increase, from 1.25 to1.28; p = 0.068 (figure 1B). No alteration was observed inmicroglial activation in the GM in either group (figure 1B).Differences in microglial activation in natalizumab-treated vsuntreated patients were confirmed using an independentvoxel-based whole brain analysis of parametric BPND dif-ference images (figure 2). Figure 3 demonstrates represen-tative PET and MRIs before and after untreated follow-up of1 year (A) and before and after natalizumab treatment (B),with surface plots visualizing the microglial activity aroundthe chosen lesions.

Both patient groups were evaluated for development of clin-ical disability using the EDSS score at baseline and after 1 year.In both groups the median EDSS score was unaltered after 1year (3.5 among natalizumab-treated cohort, figure 4A, and6.0 among untreated cohort, figure 4B).

Furthermore, in the untreated group, the EDSS score wasreevaluated after an average of 4 years (median 48.7 months,IQR 38.4–56.3 months) with a median value of 6.5. Higherbaseline TSPO radioligand binding in the NAWM and in the3–6 mm perilesional area was associated with a greater EDSSscore increase during the follow-up (p = 0.013 and p = 0.003,respectively; figure 4, C and D). Spearman partial correla-tions, where the effect on the time between the measure-ments has been removed, confirmed these results (p = 0.013and p = 0.007, respectively). Because some patients in thetreated group terminated the natalizumab medication afterthe 1-year follow-up, no later evaluation of the EDSS scorewas performed in this group.

DiscussionThe results from this longitudinal in vivo PET study of a totalof 21 patients with MS demonstrate that microglial activationis reduced following 1 year treatment with natalizumab. Ourstudy confirms reduction in microglial activation in focal MSlesions as shown before,15 but now indicates it also in areasrelevant for MS disease progression such as the NAWM andthe perimeter of chronic T1 lesions. It is expected thattreatments leading to reduced microglial activation in theseareas will help slow down the neurodegenerative process anddisease progression in MS.8

Microglial activation in the MS brain has been evaluatedin cross-sectional neuropathologic studies16–18 and in cross-sectional PET imaging studies among patients withMS.3–6,19–28 The results from these multiple studies areconcordant and demonstrate increasing microglial activationin the NAWM and in the perilesional areas with advancingdisease. Few longitudinal TSPO-PET studies have addressedtherapeutic effects on microglia.15,29–31 They indicate thatthe present MS therapies are effective in reducing activationof the innate immune system within the CNS, particularly inassociation with lesions. We demonstrate now for the firsttime a reduction in microglial activation both in the NAWMand at the rim of chronic active lesions. This confirms thatthe diffuse smoldering inflammation associated with MSdisease progression can be therapeutically targeted usingefficient anti-inflammatory medication. Monoclonal anti-bodies, such as natalizumab, are large proteins and likely donot enter the CNS in any significant amount through anintact blood-brain barrier.32 Hence, we interpret that thepositive study result is mostly due to efficient blocking ofinflammatory cell entry from the periphery into the CNSrather than direct effect of natalizumab on the resident in-nate immune cell population within the CNS. In fact, it hasbeen demonstrated earlier in a series of elegant animalexperiments that CNS-entering autoreactive T cells inducemicroglial activation and enhance TSPO expression.33

Our study was performed as an investigator-initiated substudyto the ASCEND study. ASCEND (sponsored by Biogen

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 5

Page 74: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Idec) was a phase 3 double-blind placebo-controlled ran-domized controlled trial in SPMS.10 In this study, natalizumabtreatment did not meet the primary efficacy end point ofreducing sustained disability progression unrelated to relap-ses. This was measured by a multicomponent end point,which included the EDSS, Timed 25-Foot Walk, and 9-HolePeg Test (9HPT). Natalizumab treatment did howeverreduce worsening of hand function of the patients in thisstudy, as measured by the 9HPT. This, along with ourfindings of reduced microglial activation following

natalizumab treatment, suggests that natalizumab treat-ment can be useful even in an advanced MS patient cohort,despite of the negative primary outcome in ASCEND.Problems of EDSS measurement as a clinical outcome instudies of progressive MS are well recognized, and theEDSS was slow to respond to treatment also in the presentstudy. Because preserving upper limb function is highlydesirable among patients, the 9HPT has been suggested asa primary outcome in some studies of progressive MS, witha final aim to maintain the usability of hands of patients.34,35

Figure 1 Boxplots of [11C]PK11195 DVR values at baseline and after 1 year

Evaluation ofmicroglial activation in different brain regions of interest performed using TSPO-PET imaging and [11C]PK11195 radioligand. (A) Boxplots of [11C]PK11195 DVR values in the NAWM, thalamus, and cortical GM of patients with MS (n = 21) and healthy controls (n = 8) at baseline. Microglial activation wasincreased in the NAWM and thalamus in patients with MS compared with controls, but not in the cortical GM. (B) Boxplots of [11C]PK11195 DVR values indifferent brain areas of untreated (n = 11) and natalizumab-treated (n = 10) patients with MS at baseline and at 1-year time points. The differences in DVRvalues between timepointswere compared using theWilcoxon signed-rank test. Asterisk denotes statistically significant groupdifference at a level ofp < 0.05.DVR = distribution volume ratio; GM = gray matter; NAWM = normal-appearing white matter; RRMS = relapsing-remitting MS; SPMS = secondary progressiveMS; TSPO = translocator protein.

6 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 75: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Ours is the first study to evaluate longitudinally the naturalevolution of MS disease pathology in terms of glial activation.We show that microglial activation increased during naturalevolution of MS disease in the NAWM and perilesional areasduring a period of 1 year. This was demonstrated in thecontrol group of untreated patients with MS who mostlypresented with secondary progressive disease type. Microglialactivation is thus a dynamic process associated with MS dis-ease evolution and can be measured using longitudinal in vivoTSPO-PET imaging. Of interest, higher degree of microglialactivation in the NAWM and in the perilesional area in thebaseline PET images of the untreated patients correlated witha greater increase in the EDSS score during follow-up ofa median duration of 4 years. This observation supports thehypothesis that microglial activation plays an eminent role incontributing to clinical disability and suggests that the level ofmicroglial activation could be used as a predictive imagingbiomarker for MS disease progression. To identify thosepatients who are more likely to progress would be of

significant value both for choosing treatments aiming to re-duce disease progression for patients withMS in the clinic andfor choosing suitable patients with MS for clinical trials ofprogressive MS.

The single previously published longitudinal TSPO-PETstudy evaluating the effect of natalizumab treatment detecteda significant decrease in microglial activation in both en-hancing and nonenhancing lesions. The investigators con-cluded that the NAWM had stable [11C]PK11195 uptakeover 6 months and thereafter deemed the NAWM as an idealreference tissue. The lack of treatment effect on the NAWMin the study by Kaunzner et al.15 may be due to the shortertreatment time compared with our study (6 months vs 12months). Similarly, 6-month fingolimod treatment did notreduce the microglial activation within the NAWM.30 Becauseof the heterogeneous nature of the natalizumab-treated pa-tient cohorts in these studies, it remains uncertain whethertreatment of a pure SPMS patient population would lead to

Figure 2 Demonstration of the natalizumab-induced reduction in [11C]PK11195 binding compared with the change in theuntreated group

Statistical parametric mapping wasused to demonstrate brain areaswhere radioligand binding is signifi-cantly different between natalizumab-treatedanduntreatedpatient cohorts.Two-sample t-testwasperformedwithnormalized parametric binding po-tential (BPND) difference images be-tween PET scans (1 year–baseline) inMNI space (Montreal Neurological In-stitute database), where the imageswere smoothed with 3D Gaussian 8-mm FWHM filter to ensure normalitybefore the statistical analysis. Multiplecomparisons were corrected usingthe family-wise error (FWE) rate, andthecritical significance level to reject thenull hypothesis was set to 0.05.

Figure 3 Visual demonstration of the change in periplaque microglial activation in the MS brain

Demonstration of alteration in microglial activation in in-dividual untreated SPMS and natalizumab-treated RRMSpatients. (A) Axial view of 3DT1 MRIs and respective DVRimages at baseline (top row) and after 1 year (bottom row) inan untreated patient with MS (a 53-year-old woman with 29years of MS and 2 years of SPMS and EDSS score of 6.0 atbaseline and 6.5 at follow-up). The radioligand bindingincreases at the edge of chronic black holes during the fol-low-up (white arrows). This can also be visualized usingsurface plots of the chosen lesions (on the right). (B) Axialview of 3DT1 MRIs and respective DVR images at baseline(top row) and after 1 year (bottom row) of natalizumabtreatment (a 46-year-old womanwith 7.6 years of RRMS andEDSS score of 2). The radioligand binding decreases at theedge of chronic black holes during treatment (white arrows).The surface plots on the right visualize the decreasedmicroglial activation around the chosen lesions. DVR = dis-tribution volume ratio; EDSS = Expanded Disability StatusScale; RRMS = relapsing-remitting MS; SPMS = secondaryprogressive MS.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 7

Page 76: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

similar reduction in microglial activation. This will need to beconfirmed in a separate study. No alteration was observed inmicroglial activation in the GM in the present study.

[11C]PK11195 is the first and most widely used TSPO-binding PET ligand in imaging of neuroinflammation.36 Wechose to use the [11C]PK11195 in this study because theanalysis methods for quantification of [11C]PK11195 bindinghave been carefully developed, and longitudinal image acqui-sition can be performed with ease without arterial cannulationand with good reproducibility.13,37 New second- and third-generation TSPO ligands with better binding characteristicshave been developed and tested also in imaging of MS.2,22–28

These, however, show heterogeneous binding to TSPO due togenetic polymorphism (rs6971), which complicates in-terpretation of the results. Moreover, the best quantificationmethods for these ligands are still being sought.38 TSPO ex-pression in the brain is not specific for activated microglia, butalso some astrocytes, macrophages, and endothelial cells ex-press the molecule. The specificity of TSPO binding regardingM1-like and M2-like microglia has been unclear, but recent

evidence from in vitro and animal studies suggests that theincreased expression of TSPO in neuroinflammation pre-dominates in proinflammatory M1-like microglia.39

Natalizumab, by blocking the entry of autoreactive lymphocytesand monocytes into the CNS, is one of the most effectivedisease-modifying treatments for RRMS.40 Our results demon-strate that natalizumab treatment leads to reduced activation ofmicroglia and macrophages behind an intact blood-brain barrierin brain areas relevant for disease progression. The study thussupports the usability of TSPO-PET imaging as a surrogateoutcome measure in studies of progressive MS and opens newvistas for designing future therapeutic studies in progressive MS.

Study fundingThis work was supported by Biogen Idec, Finnish Academy,Sigrid Juselius Foundation, The Finnish MS Foundation, TheFinnish Medical Foundation, The State Research Funding,and the European Union’s Seventh Framework Programme(FP7/2007-2013) under grant agreement no. HEALTH-F2-2011-278850 (INMiND).

Figure 4 EDSS score alteration during follow-up

(A) Boxplots of EDSS values atbaseline and at 1 year follow-up inthe natalizumab-treated group (n =10). Themedian EDSS score was notaltered during the year of natalizu-mab treatment. The median EDSSscore was 3.5 (IQR 2.62–3.88) atbaseline and 3.5 (IQR 3.12–4.38) af-ter 1 year. When EDSS values be-tween the time points werecompared using the Wilcoxonsigned-rank test, a slight differencewas observed between the values (p= 0.037). (B) Boxplots of EDSS valuesat baseline, 1 year, and at an aver-age of 4-year follow-up among un-treated patients (n = 11). Themedian EDSS score was not alteredduring the first follow-up year. Themedian EDSS score was 6.0 (IQR3.75–6.0) at baseline and 6.0 (IQR3.75–6.5) after 1 year. By 4 years, themedian EDSS score was increasedto 6.5 (IQR 4.75–7.25). The differ-ences in EDSS values between thetime points were compared usingthe Wilcoxon signed-rank test. Afteradjusting the p values for multiplecomparisons (Holm method), a sta-tistically significant difference wasobserved only between the baselineand 4 years (p = 0.041). (C and D)Higher baseline [11C]PK11195 DVRvalues in the NAWM and 3–6 mmperilesional area associated witha greater EDSS score increase dur-ing the 4-year follow-up (n = 11). Thecorrelation analysis was performedusing Spearman correlation. Aster-isk denotes statistically significantgroup difference at a level of p <0.05. DVR = distribution volume ra-tio; EDSS = Expanded Disability Sta-tus Scale; IQR = interquartile range;NAWM = normal-appearing whitematter; RRMS = relapsing-remittingMS; SPMS = secondary progressiveMS.

8 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 77: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

DisclosureM. Sucksdorff, J. Tuisku, M. Matilainen, A. Vuorimaa, S.Smith, J, Keitila, J. Rokka, R. Parkkola, M. Nylund, and J.Rinne report no disclosures. E. Rissanen has received speakerhonoraria from Teva, Biogen, and Roche, consultation feefrom Merck, and personal research grants from Turku Uni-versity Hospital and the Finnish MS Foundation. L. Airas:compensation for consulting—Merck and Roche; in-stitutional support for research—SanofiGenzyme and BiogenIdec. Go to Neurology.org/NN for full disclosures.

Publication historyReceived by Neurology: Neuroimmunology & NeuroinflammationDecember 19, 2018. Accepted in final form April 2, 2019.

References1. Frischer JM, Weigand SD, Guo Y, et al. Clinical and pathological insights into the

dynamic nature of the white matter multiple sclerosis plaque. Ann Neurol 2015;78:710–721.

2. Herranz E, GiannıC, Louapre C, et al. Neuroinflammatory component of gray matterpathology in multiple sclerosis. Ann Neurol 2016;80:776–790.

3. Debruyne JC, Versijpt J, Van Laere KJ, et al. PET visualization of microglia in multiplesclerosis patients using [11C]PK11195. Eur J Neurol 2003;10:257–264.

4. Versijpt J, Debruyne JC, Van Laere KJ, et al. Microglial imaging with positronemission tomography and atrophy measurements with magnetic resonance imagingin multiple sclerosis: a correlative study. Mult Scler 2005;11:127–134.

5. Rissanen E, Tuisku J, Vahlberg T, et al. Microglial activation, white matter tractdamage, and disability in MS. Neurol Neuroimmunol Neuroinflamm 2018;5:e443.doi: 10.1212/NXI.0000000000000443.

6. Rissanen E, Tuisku J, Rokka J, et al. In vivo detection of diffuse inflammation insecondary progressive multiple sclerosis using PET imaging and the radioligand 11C-PK11195. J Nucl Med 2014;55:939–944.

7. Mayo L, Quintana FJ, Weiner HL. The innate immune system in demyelinatingdisease. Immunol Rev 2012;248:170–187.

8. Gillen KM, Mubarak M, Nguyen TD, Pitt D. Significance and in vivo detection of Iron-Laden microglia in white matter multiple sclerosis lesions. Front Immunol 2018;9:255.

9. Correale J. The role of microglial activation in disease progression. Mult Scler 2014;20:1288–1295.

10. Kapoor R, Ho PR, Campbell N, et al. Effect of natalizumab on disease progression insecondary progressive multiple sclerosis (ASCEND): a phase 3, randomised, double-blind, placebo-controlled trial with an open-label extension. Lancet Neurol 2018;17:405–415.

11. Schmidt P, Gaser C, Arsic M, et al. An automated tool for detection of FLAIR-hyperintense white-matter lesions in multiple sclerosis. Neuroimage 2012;59:3774–3783.

12. de Jong HW, van Velden FH, Kloet RW, Buijs FL, Boellaard R, Lammertsma AA.Performance evaluation of the ECAT HRRT: an LSO-LYSO double layer high res-olution, high sensitivity scanner. Phys Med Biol 2007;52:1505–1526.

Appendix Authors

Name Location Role Contribution

MarcusSucksdorff,MD

Turku PET Centre,Turku UniversityHospital andUniversity ofTurku, Turku,Finland

Correspondingauthor andinvestigator

Study concept,data collection,manuscriptdrafting, andcritical revision ofthe manuscript

JouniTuisku, MSc

Turku PETCentre, TurkuUniversityHospital andUniversity ofTurku, Turku,Finland

Statistician andimage modeler

Data collectionand analysis andinterpretation ofdata

MarkusMatilainen,PhD

Turku PETCentre, TurkuUniversityHospital andUniversity ofTurku, Turku,Finland

Statistician Data collection,analysis andinterpretation ofdata, and criticalrevision of themanuscript

AnnaVuorimaa,BM

Turku PETCentre, TurkuUniversityHospital andUniversity ofTurku, Turku,Finland

Investigator Data collectionand analysis ofdata

SarahSmith, BSc

Turku PETCentre, TurkuUniversityHospital andUniversity ofTurku, Turku,Finland

Investigator Data collection,analysis of data

JoonasKeitila, BM

Turku PET Centre,Turku UniversityHospital andUniversity ofTurku, Turku,Finland

Investigator Data collectionand analysis ofdata

JohannaRokka, MSc,PhD

Turku PETCentre, TurkuUniversityHospital andUniversity ofTurku, Turku,Finland

Investigator Data collection

Appendix (continued)

Name Location Role Contribution

RiittaParkkola,MD, PhD

Department ofRadiology,UniversityHospital andUniversity ofTurku, Turku,Finland

Investigator Data collectionand analysis ofdata

MarjoNylund,MSc

Turku PETCentre, TurkuUniversityHospital andUniversity ofTurku, Turku,Finland

Investigator Data collectionand criticalrevision of themanuscript

Juha Rinne,MD, PhD

Turku PETCentre, TurkuUniversityHospital andUniversity ofTurku, Turku,Finland

Investigator Study conceptandinterpretation ofresults

EeroRissanen,MD, PhD

Turku PETCentre, TurkuUniversityHospital andUniversity ofTurku, Turku,Finland

Investigator Study concept,acquisition andinterpretation ofdata, and criticalrevision of themanuscript

Laura Airas,MD, PhD

Turku PETCentre, TurkuUniversityHospital andUniversity ofTurku, Turku,Finland

PI Study conceptand design,interpretation ofresults,manuscriptdrafting, andcritical revision ofthe manuscript

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 9

Page 78: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

13. Turkheimer FE, Edison P, Pavese N, et al. Reference and target region modeling of[11C]-(R)-PK11195 brain studies. J Nucl Med 2007;48:158–167.

14. Gunn RN, Lammertsma AA, Hume SP, Cunningham VJ. Parametric imaging ofligand-receptor binding in PET using a simplified reference region model. Neuro-image 1997;6:279–287.

15. Kaunzner UW, Kang Y, Monohan E, et al. Reduction of PK11195 uptake observed inmultiple sclerosis lesions after natalizumab initiation. Mult Scler Relat Disord 2017;15:27–33.

16. Frischer JM, Bramow S, Dal-Bianco A, et al. The relation between inflammation andneurodegeneration in multiple sclerosis brains. Brain 2009;132:1175–1189.

17. Lassmann H, van Horssen J, Mahad D. Progressive multiple sclerosis: pathology andpathogenesis. Nat Rev Neurol 2012;8:647–656.

18. Moll NM, Rietsch AM, Thomas S, et al. Multiple sclerosis normal-appearing whitematter: pathology-imaging correlations. Ann Neurol 2011;70:764–773.

19. Banati RB, Newcombe J, Gunn RN, et al. The peripheral benzodiazepine binding sitein the brain in multiple sclerosis: quantitative in vivo imaging of microglia as a measureof disease activity. Brain 2000;123(pt 11):2321–2337.

20. Politis M, Giannetti P, Su P, et al. Increased PK11195 PET binding in the cortex ofpatients with MS correlates with disability. Neurology 2012;79:523–530.

21. Giannetti P, Politis M, Su P, et al. Microglia activation in multiple sclerosis black holespredicts outcome in progressive patients: an in vivo [(11)C](R)-PK11195-PET pilotstudy. Neurobiol Dis 2014;65:203–210.

22. Datta G, Colasanti A, Kalk N, et al. 11C-PBR28 and 18F-PBR111 detect white matterinflammatory heterogeneity in multiple sclerosis. J Nucl Med 2017;58:1477–1482.

23. Vas A, Shchukin Y, Karrenbauer VD, et al. Functional neuroimaging in multiplesclerosis with radiolabelled glia markers: preliminary comparative PET studies with[11C]vinpocetine and [11C]PK11195 in patients. J Neurol Sci 2008;264:9–17.

24. Oh U, Fujita M, Ikonomidou VN, et al. Translocator protein PET imaging for glialactivation in multiple sclerosis. J Neuroimmune Pharmacol 2011;6:354–361.

25. Park E, Gallezot JD, Delgadillo A, et al. ((11)C-PBR28 imaging in multiple sclerosispatients and healthy controls: test-retest reproducibility and focal visualization ofactive white matter areas. Eur J Nucl Med Mol Imaging 2015;42:1081–1092.

26. Vomacka L, Albert NL, Lindner S, et al. TSPO imaging using the novel PET ligand[18F]GE-180: quantification approaches in patients with multiple sclerosis. EJNMMIRes 2017;7:89.

27. Takano A, Piehl F, Hillert J, et al. In vivo TSPO imaging in patients with multiplesclerosis: a brain PET study with [18F]FEDAA1106. EJNMMI Res 2013;3:30.

28. Colasanti A,GuoQ,MuhlertN, et al. In vivo assessment of brainwhitematter inflammationin multiple sclerosis with (18)F-PBR111 PET. J Nucl Med 2014;55:1112–1118.

29. Ratchford JN, Endres CJ, Hammoud DA, et al. Decreased microglial activation in MSpatients treated with glatiramer acetate. J Neurol 2012;259:1199–1205.

30. Sucksdorff M, Rissanen E, Tuisku J, et al. Evaluation of the effect of fingolimodtreatment on microglial activation using serial PET imaging in multiple sclerosis.J Nucl Med 2017;58:1646–1651.

31. Bunai T, Terada T, Kono S, et al. Neuroinflammation following disease modifyingtherapy in multiple sclerosis: a pilot positron emission tomography study. J Neurol Sci2018;385:30–33.

32. Hagens MH, Killestein J, Yaqub MM, et al. Cerebral rituximab uptake in multiplesclerosis: a (89)Zr-immunoPET pilot study. Mult Scler 2018;24:543–545.

33. Grebing M, Nielsen HH, Fenger CD, et al. Myelin-specific T cells induce interleukin-1beta expression in lesion-reactive microglial-like cells in zones of axonal de-generation. Glia 2016;64:407–424.

34. Giovannoni G, Cutter G, Sormani MP, et al. Is multiple sclerosis a length-dependentcentral axonopathy? The case for therapeutic lag and the asynchronous progressiveMS hypotheses. Mult Scler Relat Disord 2017;12:70–78.

35. Giovannoni G, Airas L, Bove R, et al. Ocrelizumab treatment effect on upper limbfunction in PPMS patients with disability: subgroup results of the ORATORIO studyto inform the ORATORIO-HAND study. Neurology 2019;92 (15 Suppl):P3.2-091.Available at: https://n.neurology.org/content/92/15_Supplement/P3.2-091.

36. Airas L, Rissanen E, Rinne JO. Imaging neuroinflammation in multiple sclerosis usingTSPO-PET. Clin Transl Imaging 2015;3:461–473.

37. Kang Y, He B, Schlyer D, Vallabhajosula S, Mozley P. Reproducibility of test-retest with11C -PK11195 using different input function approaches. J Nucl Med 2015;56:152.

38. Fujita M, Kobayashi M, Ikawa M, et al. Comparison of four (11)C-labeled PETligands to quantify translocator protein 18 kDa (TSPO) in human brain: (R)-PK11195, PBR28, DPA-713, and ER176-based on recent publications that measuredspecific-to-non-displaceable ratios. EJNMMI Res 2017;7:84.

39. Beckers L, Ory D, Geric I, et al. Increased expression of translocator protein (TSPO)marks pro-inflammatory microglia but does not predict neurodegeneration. MolImaging Biol 2018;20:94–102.

40. Rae-Grant A, Day GS, Marrie RA, et al. Comprehensive systematic review summary:disease-modifying therapies for adults with multiple sclerosis: report of the guidelinedevelopment, dissemination, and implementation subcommittee of the AmericanAcademy of neurology. Neurology 2018;90:789–800.

10 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 79: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

ARTICLE OPEN ACCESS

Guillain-Barre syndrome and related diseasesafter influenza virus infectionMasaki Yamana, MD, Motoi Kuwahara, MD, PhD, Yuta Fukumoto, MD, Keisuke Yoshikawa, MD,

Kazuo Takada, MD, PhD, and Susumu Kusunoki, MD, PhD

Neurol Neuroimmunol Neuroinflamm 2019;6:e575. doi:10.1212/NXI.0000000000000575

Correspondence

Dr. Kusunoki

[email protected]

AbstractObjectiveWe examined the clinical and serologic features of Guillain-Barre syndrome (GBS)-relateddiseases (GBSRDs), including GBS, Fisher syndrome (FS), and Bickerstaff brainstem en-cephalitis (BBE), after influenza virus infection (GBSRD-I) to reveal potential underlyingautoimmune mechanisms.

MethodsWe retrospectively investigated the presence of antiglycolipid antibodies against 11 glycolipidsand the clinical features of 63 patients with GBSRD-I. Autoantibody profiles and clinicalfeatures were compared with those of 82 patients with GBSRDs after Campylobacter jejuniinfection (GBSRD-C).

ResultsThe anti-GQ1b seropositivity rate was significantly higher, whereas the GM1 and GD1a se-ropositivity rates were significantly lower in GBSRD-I compared with GBSRD-C. Anti-GQ1band anti-GT1a were the most frequently detected antiglycolipid antibodies in GBSRD-I (both15/63, 24%). Consequently, FS was more frequent in GBSRD-I than GBSRD-C (22% vs 9%,p < 0.05). In addition, as for GBS, cranial nerve deficits, sensory disturbances, and ataxia weremore frequent in the cases after influenza infection (GBS-I) than in those after C. jejuniinfection (GBS-C) (46% vs 15%, 75% vs 46%, and 29% vs 4%, respectively; all p < 0.01). Nerveconduction studies revealed acute inflammatory demyelinating polyneuropathy (AIDP) in 60%of patients with GBS-I but only 25% of patients with GBS-C (p < 0.01).

ConclusionsAnti-GQ1b antibodies are the most frequently detected antibodies in GBSRD-I. Comparedwith GBS-C, GBS-I is characterized by AIDP predominance and frequent presence of cranialnerve involvement and ataxia.

From the Department of Neurology (M.Y., M.K., Y.F., K.Y., K.T., S.K.), Kindai University Faculty of Medicine; and Department of Neurology (K.T.), Shinjinkai Hospital, Osaka, Japan.

Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by the authors.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 80: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Guillain-Barre syndrome (GBS) is an acute acquired auto-immune disorder of the peripheral nerves that frequentlydevelops after infection. For instance, antecedent infectionsuch as Campylobacter jejuni, cytomegalovirus, orMycoplasmapneumoniae is observed in approximately 70% of patients withGBS. Alternatively, GBS following influenza virus infection(GBS-I) is relatively rare.1 However, influenza virus infectionis a common respiratory syndrome across all age groups. In-fluenza is also known to cause neurologic complications suchas encephalitis, encephalopathy, and Reye syndrome that re-quire differential diagnosis.2 In addition, several reports haveshown that influenza virus infection or influenza-like illnesscan also cause Fisher syndrome (FS) and Bickerstaff brain-stem encephalitis (BBE),3–5 which are caused by the patho-genetic mechanisms similar to those of GBS. Here, we callGBS, FS, and BBE as GBS-related diseases (GBSRDs).

Antiglycolipid antibodies are elevated in GBSRD and arestrongly implicated in the pathogenesis. Antibodies againstGM1 on the neuronal membrane ganglioside are oftendetected in GBS after C. jejuni infection, and antibodies togalactocerebroside (Gal-C) are often detected in neurologicdiseases following M. pneumoniae infection.6,7 The structuresof these carbohydrates are similar to carbohydrates expressedby the infectious agents, suggesting that a form of molecularmimicry is responsible for GBS-associated autoimmunity.8,9

In contrast to GBS associated with C. jejuni and M. pneumo-niae infection, the clinical and serologic features of GBSRDand GBS after influenza virus infection (GBSRD-I and GBS-I,respectively) have not been described in detail. The purposeof this study is to investigate the unique clinical and serologicfeatures of GBSRD-I and GBS-I.

MethodsPatients with GBSRD-IWe collected clinical information and acute-phase serumsamples from consecutive patients who are diagnosed withGBSRD-I. These serum samples were sent to our laboratoryfrom multiple hospitals in Japan between October 2009 andFebruary 2017 for the examination of antiglycolipid anti-bodies. GBS and BBE were diagnosed according to previouslypresented criteria,10,11 and FS was diagnosed according to theclinical triad of acute progressive ophthalmoplegia, ataxia, andareflexia without limb weakness or impairment of con-sciousness. Patients with the FS triad and limb weakness were

included in the GBS subgroup. In all patients with GBSRD,influenza virus infection was diagnosed by animmunochromatography-based rapid influenza diagnostictest (RIDT) within 4 weeks of the onset of symptoms.Immunochromatography-based RIDT detects nucleoprotein,which is one of the most abundant proteins in influenza virusand has fewer mutations than hemagglutinin (HA) andneuraminidase (NA). Although there are some differences inthe detection rate, RIDT can detect various subtypes of in-fluenza virus such as H1N1, H3N2, type B seasonal viruses,and pandemic H1N1 2009 viruses.12 The sensitivity of theRIDTs is approximately 60%, and the specificity is higher than95%.13 Patients with GBSRD-I with antecedent gastrointes-tinal infectious symptoms were excluded from the study.

Clinical and electrophysiological assessmentThe clinical and electrophysiological data of each patient withGBS were obtained retrospectively from the original attend-ing neurologist or pediatrician using a questionnaire.According to the Ho criteria, nerve conduction study (NCS)findings were used to classify cases as acute inflammatorydemyelinating polyneuropathy (AIDP), acute motor axonalneuropathy (AMAN), or unclassified.14 NCSs were

Table 1 Demographic and basic clinical features of theGBSRD-I and GBSRD-C groups

GBSRD-I(n = 63)

GBSRD-C(n = 82) p Values

Age, median [range] (y) 40 [3–83] 30.5 [3–87] NS

Sex (male/female) 36/27 41/41 NS

Type of influenzavirus (A/B/unknown)

37/17/10a NA

GBS/FS/BBE 48/14/1 74/7/1 <0.05b

Delay betweeninfection and GBSRDonset median [range] (d)

(n = 61)c

10 [2–28](n = 76)c

9 [2–28]NS

Abbreviations: BBE = Bickerstaff brainstem encephalitis; FS = Fisher syn-drome; GBS = Guillain-Barre syndrome; GBSRD = Guillain-Barresyndrome–related disease; GBSRD-C = Guillain-Barre syndrome–relateddisease after C. jejuni; GBSRD-I = Guillain-Barre syndrome–related diseaseafter influenza; NA = not applicable; NS = not significant.a One patient was infected with both A and B influenza viruses. The numberof days until onset was 8 after influenza A and 18 after influenza B.b FS was significantly more frequent in GBSRD-I than GBSRD-C.c In the GBSRD-I group, the patient with both influenza A and B and a patientwith unclear delay from influenza toGBS onsetwere excluded (n = 2of 63). Inthe GBSRD-C group, patients with unclear delay between influenza and GBSonset were excluded (n = 6 of 82).

GlossaryAIDP = acute inflammatory demyelinating polyneuropathy; AMAN = acute motor axonal neuropathy; BBE = Bickerstaffbrainstem encephalitis; FS = Fisher syndrome; GBS = Guillain-Barre syndrome; GBS-C = Guillain-Barre syndrome after C.jejuni infection; GBS-I = Guillain-Barre syndrome after influenza virus infection; GBSRD = Guillain-Barre syndrome–relateddisease; GBSRD-C = Guillain-Barre syndrome–related disease after C jejuni infection; GBSRD-I = Guillain-Barresyndrome–related disease after influenza virus infection;HA = hemagglutinin; NA = neuraminidase; NCS = nerve conductionstudy; RIDT = rapid influenza diagnostic test.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 81: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

performed at each participating hospital in a median of 8.5days (range [1–20] days) after GBS symptom onset.

Antiglycolipid antibodiesSerum IgG antibodies to 11 glycolipid antigens (GM1, GM2,GM3, GD1a, GD1b, GD3, GT1b, GQ1b, GT1a, Gal-C, andGalNAc-GD1a) were examined in all patients by ELISA, aspreviously described.15

Patients with GBSRD-CClinical and serologic features of GBSRD-I and GBS-I werecompared with those of GBSRD and GBS after C. jejuniinfection (GBSRD-C and GBS-C, respectively). The anti-glycolipid antibodies of these patients with GBSRD-C weretested in our laboratory between September 2012 and April2017. C. jejuni infection was diagnosed by fecal culture orC. jejuni antibody test.

Statistical analysisDifferences in proportions were evaluated using the χ2

test or Fisher exact test, and differences in the medianwere evaluated using the Mann-Whitney U test. p < 0.05(2 tailed) was considered significant for all tests. Statisticalcalculations were performed using SPSS version 2.0 (IBM,Japan).

Study approval and patient consentThis study was approved by the Internal Review Board ofKindai University Faculty of Medicine. All participants pro-vided written informed consent.

Data availabilityAnonymized data not published within the article will beshared by request from any qualified investigator.

ResultsComparison of GBS subtype and antiglycolipidantibody profile between GBSRD-I and GBSRD-CThe classic form of GBS (GBS-I or GBS-C according to theantecedent pathogen) was the most common disease subtypein both GBSRD-I and GBSRD-C groups (48/63 [76%] vs74/82 [90%]), while BBE was rare in both groups (1 patientper group). In contrast, FS was significantly more prevalent inthe GBSRD-I group compared with the GBSRD-C group(14/63 [22%] vs 7/82 [9%], p = 0.02).

Antiglycolipid antibodies were detected more frequentlyin patients with GBSRD-C than in patients with GBSRD-I(51/82 [62%] vs 25/63 [40%]; p < 0.01) (table 1). Therewere also significant differences in seropositivity rates forspecific antiglycolipid antibodies between groups. Anti-GM1antibody was substantially more common in GBSRD-Cthan in GBSRD-I (24/82 [29%] vs 3/63 [5%]; p < 0.01).Similarly, seropositivity for anti-GD1a was more frequentin GBSRD-C (18/62 [22%] vs 1/63 [2%]; p < 0.01). Incontrast, anti-GQ1b was significantly more frequent inGBSRD-I than in GBSRD-C (15/63 [24%] vs 7/82 [9%];p < 0.05). Anti-GT1a was also relatively common inGBSRD-I (24%), although the seropositivity rate did not

Table 2 Comparison of antiglycolipid antibody profile between patients with GBSRD-I and GBSRD-C

IgG antibody positive

GBSRD-I GBSRD-C

p Values(GBSRD-I vs GBSRD-C)

All(n = 63)

GBS-I(n = 48)

FS-I(n = 14)

BBE-I(n = 1)

All(n = 82)

GBS-C(n = 74)

FS-C(n = 7)

BBE-C(n = 1)

Overall, n (%) 25 (40%) 14 (29%) 10 (71%) 1 (100%) 51 (62%) 44 (59%) 6 (86%) 1 (100%) <0.01

GM1 3 (5%) 3 (6%) 0 (0%) 0 (0%) 24 (29%) 23 (31%) 1 (14%) 0 (0%) <0.01

GM2 0 (0%) 0 (0%) 0 (0%) 0 (0%) 5 (6%) 5 (7%) 0 (0%) 0 (0%) NS

GM3 1 (2%) 0 (0%) 1 (7%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) NS

GD1a 1 (2%) 1 (2%) 0 (0%) 0 (0%) 18 (22%) 16 (22%) 1 (14%) 1 (100%) <0.01

GD1b 9 (14%) 7 (15%) 2 (13%) 0 (0%) 12 (15%) 12 (16%) 0 (0%) 0 (0%) NS

GD3 2 (3%) 0 (0%) 1 (7%) 1 (100%) 1 (1%) 1 (1%) 0 (0%) 0 (0%) NS

GT1a 15 (24%) 4 (8%) 10 (71%) 1 (100%) 15 (18%) 9 (12%) 5 (71%) 1 (100%) NS

GT1b 0 (0%) 0 (0%) 0 (0%) 0 (0%) 3 (4%) 1 (1%) 1 (14%) 1 (100%) NS

GQ1b 15 (24%) 4 (8%) 10 (71%) 1 (100%) 7 (9%) 2 (3%) 4 (57%) 1 (100%) <0.05

GalNAc-GD1a 8 (13%) 6 (13%) 1 (7%) 1 (100%) 17 (21%) 16 (22%) 1 (14%) 0 (0%) NS

Gal-C 3 (5%) 3 (6%) 0 (0%) 0 (0%) 1 (1%) 0 (0%) 1 (14%) 0 (0%) NS

Abbreviations: BBE-C = Bickerstaff brainstemencephalitis after C. jejuni infection; BBE-I = Bickerstaff brainstemencephalitis after influenza virus infection; FS-C = Fisher syndrome after C. jejuni infection; FS-I = Fisher syndrome after influenza virus infection; GBS-C = Guillain-Barre syndrome after C. jejuni infection;GBS-I = Guillain-Barre syndrome after influenza virus infection; GBSRD-C = Guillain-Barre syndrome–related disease after C. jejuni infection; GBSRD-I =Guillain-Barre syndrome–related disease after influenza virus infection; NA = not applicable; NS = not significant.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 82: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

differ significantly from patients with GBSRD-C (18%).Among the 15 patients with GBSRD-I seropositive for anti-GQ1b, 10 were diagnosed with FS. In patients with GBS-I,anti-GD1b was the most frequently detected autoantibody(7/48, 15%) (table 2).

Clinical features of GBS-IThe clinical features of GBS-I (n = 48) and GBS-C aresummarized in table 3. Compared with patients with GBS-C,patients with GBS-I demonstrated a substantially lower anti-glycolipid antibody seropositivity rate (29% vs 59%, p < 0.01).Conversely, cranial nerve deficits, sensory disturbances, and

ataxia were significantly more frequent in GBS-I than inGBS-C (46% vs 15%, 75% vs 46%, and 29% vs 4%, re-spectively; p < 0.01). There were no significant differences inage, sex ratio, autonomic disturbance frequency, and the re-quirement for artificial ventilation between GBS-I and GBS-Csubgroups.

We also obtained the NCS findings from 30 patients withGBS-I. According to the Ho criteria, 18 cases were classified asAIDP and 12 were unclassified, whereas no case was classifiedas AMAN. The proportion with AIDP was significantly higherin the GBS-I subgroup than in the GBS-C subgroup (p <0.01). Among patients with GBS-I, there were no significantdifferences in demographic or clinical features between thoseinfected with influenza virus A or B.

DiscussionGBSRD is presented after not only gastrointestinal infectionbut also upper respiratory infection. However, most of theantecedent infectious agents of upper respiratory infection arenot identified. In the present study, we focused on influenzavirus infection and report the distinct clinical and serologicfeatures of GBSRD-I and GBS-I.

Anti-GQ1b and anti-GT1a were the most frequently detectedantiglycolipid antibodies in GBSRD-I. GQ1b is densely lo-calized in the paranodal regions of III, IV, and VI humancranial nerves. Therefore, anti-GQ1b antibody can causeophthalmoplegia, which is one of the major symptoms ofFS.16 In accord with this, FS was more frequent in GBSRD-Ithan GBSRD-C.

Compared with GBS-C, GBS-I is characterized by more fre-quent cranial nerve deficits, sensory disturbances, and ataxia.In addition, NCS findings revealed demyelinating neuropathyin a higher proportion of GBS-I cases than GBS-C cases.Furthermore, the frequency of AIDP was higher in GBS-Ithan in all GBS cases previously reported by a prospectivecohort study in Japan (60% vs 40%).17 This is compatible withthe results that GM1 and GD1a seropositivity rates weresignificantly lower in GBSRD-I compared with GBSRD-C.Besides, compared with cases with GBSRDs after M. pneu-moniae infection (GBSRD-M) reported recently,7 the clinicaland serologic features of GBSRD-I were somewhat differentfrom those of GBSRD-M, in which the anti-GQ1b antibodypositive rate and the frequency of FS cases were lower, and theanti-Gal-C antibody positive rate was higher than inGBSRD-I.

Previous reports (summarized in table 4) also found morefrequent cranial nerve deficits (8/19, 42%) and sensory dis-turbance (15/19, 79%) in GBS-I. Moreover, these NCSsrevealed demyelinating neuropathy in some patients withGBS-I (5/13, 38%), whereas axonal neuropathy was relativelyless common (2/13, 15%).1,18–23 Those are in accord with thecurrent findings.

Table 3 Comparison of clinical characteristic betweenpatients with GBS-I and GBS-C

FeaturesGBS-I(n = 48)

GBS-C(n = 74)

pValue

Age median [range] (y) 32[3–83]

35.5[3–87]

NS

Sex (male/female) 26/22 37/37 NS

Antiglycolipid antibody positive 14 (29%) 44 (59%) <0.01

Cranial nerve deficits 22 (46%) 11 (15%) <0.01

III, IV, and VI 8 2

V 3 1

VII 13 8

IX and X 9 8

XI and XII 7 5

Severe muscle weakness (MMT<3)

20 (42%) 33 (45%) NS

Sensory disturbances 36 (75%) 34 (46%) <0.01

Ataxia 14 (29%) 3 (4%) <0.01

Autonomic disturbance 10 (21%) 15 (20%) NS

Peak FG (n = 46) (n = 59)

FG1 4 10

FG2 11 12

FG3 11 5

FG4 13 26

FG5 7 6 NS

FG6 0 0

Electrophysiological examination (n = 30) (n = 55)

AMAN 0 (0%) 13 (24%) <0.01

AIDP 18 (60%) 14 (25%) <0.01

Unclassified 12 (40%) 28 (51%) NS

Abbreviations: AIDP = acute inflammatory demyelinating polyneuropathy;AMAN = acute motor axonal neuropathy; FG = functional grade; GBS-C =Guillain-Barre syndrome after C. jejuni infection; GBS-I = Guillain-Barresyndrome after influenza virus infection; MMT = manual muscle testing; NS= not significant.

4 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 83: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Influenza viruses A and B are wrapped by a lipid bilayer en-velope containing various glycoproteins, including HA andNA. Therefore, antiglycolipid antibodies may be produced byinfluenza virus infection because of possible molecular mim-icry between glycoproteins of influenza viruses and glycolipidslocalized in human peripheral nerves.

The main limitation of this study is that we could not com-pletely exclude the possible coexistence of other infections.Selection bias by attending physicians also could exist. Inaddition, we could not evaluate whether the severity and in-fection period of influenza virus infection are associated withthe severity of GBS-I because the study was retrospectivelyconducted. A future prospective study is needed to clarifythose issues. However, this study enrolled a relatively largenumber of patients, all with confirmed influenza infection.Therefore, clinical characteristics should reflect GBS-I andGBSRD-I.

In conclusion, antiglycolipid antibodies are detected in someGBSRD-I cases, and anti-GQ1b and anti-GT1a antibodies arethe most frequent. Compared with GBS-C, GBS-I is charac-terized by more frequent AIDP, cranial nerve deficits, sensorydisturbances, and ataxia. Further investigation is required toclarify the pathomechanisms of GBSRD-I.

Author contributionsM. Yamana has contributed to acquisition, analysis, and in-terpretation of data and drafted the manuscript. M. Kuwaharahas analyzed and interpreted data and also participated indrafting the manuscript and revising it. Y. Fukumoto, K.Yoshikawa, and K. Takada have contributed to acquisition ofdata. S. Kusunoki has made substantial contributions toconception and design of the study and also revised the

manuscript critically for important intellectual content. S.Kusunoki made final approval of the manuscript.

Study fundingThis work was supported by the Ministry of Education,Culture, Sports, Science and Technology of Japan (Grants-in-Aid for Scientific Research, 18H02745), the Ministry ofHealth, Labour and Welfare of Japan (Health and LabourSciences Research Grant on Rare and Intractable Diseases[Validation of Evidence-based Diagnosis and Guidelines, andImpact on QOL in Patients with Neuroimmunological Dis-eases]), and Intramural Research Grant (28-5) for Neuro-logical and Psychiatric Disorders of NCNP.

DisclosureNone of the authors report any disclosures relevant to themanuscript. Disclosures available: Neurology.org/NN.

Publication historyReceived byNeurology: Neuroimmunology & Neuroinflammation January4, 2019. Accepted in final form April 5, 2019.

References1. Jacobs BC, Rothbarth PH, van der Meche FGA, et al. The spectrum of antecedent

infections in Guilain-Barre syndrome: a case-control study. Neurology 1998;51:1110–1115.

2. Studahl M. Influenza virus and CNS manifestations. J Clin Virol 2003;28:225–232.3. Stowe J, Andrews N, Wise L, Miller E. Investigation of the temporal association of

Guillain-Barre syndrome with influenza vaccine and influenzalike illness using theUnited Kingdom General Practice Research Database. Am J Epidemiol 2009;169:382–388.

4. Matsumoto H, Kobayashi O, Tamura K, Ohkawa T, Sekine I. Miller Fisher syndromewith transient coma: comparison with Bickerstaff brainstem encephalitis. Brain Dev2002;24:98–101.

5. Hara M, Morita A, Ichihara K, et al. Miller Fisher syndrome associated with influenzaA infection. Intern Med 2012;51:2621–2623.

6. Hughes RA, Hadden RD, Gregson NA, Smith KJ. Pathogenesis of Guillain-Barresyndrome. J Neuroimmunol 1999;100:74–97.

7. Kuwahara M, Samukawa M, Ikeda T, et al. Characterization of the neurologicaldiseases associated with Mycoplasma pneumonia infection and anti-glycolipid anti-bodies. J Neurol 2017;264:467–475.

Table 4 Clinical summary of previously reported patients with GBS-I

No. ofPatients

Type of influenzavirus (no. ofpatients)

Duration untilonset frominfections

Cranialnervedeficits

Sensorydisturbances NCS (no. of patients)

Antigangliosideantibodies

Jacobs et al.1 3 A (2); B (1) ND ND ND ND ND

Sivadon-Tardy et al.,200918

14 A (10); B (4) 12 d [3–30] − (8); + (6) − (1); + (13) Demyelinating (4);normal (1); equivocal(3); ND (6)

Negative

Simpsonet al., 200919

1 A 7 d — + Equivocal GD1b

Chaari et al.,201020

1 A 15 d + + Reduction of MCV andprolonged DL

ND

Kutlesaet al., 201021

1 A 15 d — — Axonal GM1, GD1a, andGD1b

Corteseet al., 201222

1 A 1 d + — Normal Negative

Vasconceloset al., 201223

1 A 7 d — — Axonal Negative

Abbreviations: DL = distal latency; GBS =Guillain-Barre syndrome;MCV =motor nerve conduction velocity; NCS = nerve conduction study; ND =not described.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 5

Page 84: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

8. Ang CW, Laman JD, Willison HJ, et al. Structure of Campylobacter jejuni lip-opolysaccharides determines antiganglioside specificity and clinical features ofGuillain-Barre and Miller Fisher patients. Infect Immun 2002;70:1202–1208.

9. Kusunoki S, Shiina M, Kanazawa I. Anti-Gal-C antibodies in GBS subsequent tomycoplasma infection: evidence of molecular mimicry. Neurology 2001;57:736–738.

10. Asbury AK, Comblath DR. Assessment of current diagnostic criteria for Guillain-Barre syndrome. Ann Neurol 1990;27(suppl):S21–S24.

11. Koga M, Kusunoki S, Kaida K, et al. Nationwide survey of patients in Japan withBickerstaff brainstem encephalitis: epidemiological and clinical characteristics.J Neurol Neurosurg Psychiatry 2012;83:1210–1215.

12. Sakai-Tagawa Y, Ozawa M, Tamura D, et al. Sensitivity of influenza rapid diagnostictests to H5N1 and 2009 pandemic H1N1 viruses. J Clin Microbiol 2010;48:2872–2877.

13. Chartrand C, Leeflang MM, Minion J, Brewer T, Pai M. Accuracy of rapid influenzadiagnostic tests. Ann Intern Med 2012;156:500–511.

14. Ho TW, Mishu B, Li CY, et al. Guillain-Barre syndrome in northern China. Re-lationship to Campylobacter jejuni infection and anti-glycolipid antibodies. Brain1995;118(pt 3):597–605.

15. Kusunoki S, Chiba A, Kon K, et al. N-acetylgalactosaminyl GD1a is a target moleculefor serum antibody in Guillain-Barre syndrome. Ann Neurol 1994;35:570–576.

16. Chiba A, Kusunoki S, Obata H, Machinami R, Kanazawa I. Serum anti-GQ1b IgGantibody is associated with ophthalmoplegia in Miller Fisher syndrome and Guillain-Barre syndrome: clinical and immunohistochemical studies. Neurology 1993;43:1911–1917.

17. Mitsui Y, Kusunoki S, Arimura K, et al. A multicentre prospective study of Guillain-Barre Syndrome in Japan: a focus on the incidence of subtypes. J Neurol NeurosurgPsychiatry 2015;86:110–114.

18. Sivadon-Tardy V, Orlikowski D, Porcher R, et al. Guillain-Barre syndrome and in-fluenza virus infection. Clin Infect Dis 2009;48:48–56.

19. Simpson BS, Rajabally YA. Sensori-motor Guillain-Barre syndrome with anti-GD1bantibodies following influenza A infection. Eur J Neurol 2009;16:e81.

20. Chaari A, Bahloul M, Dammak H, et al. Guillain-Barre syndrome related to pandemicinfluenza A (H1N1) infection. Intensive Care Med 2010;36:1275.

21. Kutlesa M, Santini M, Krajinovic V, Raffanelli D, Barsic B. Acute motor axonalneuropathy associated with pandemic H1N1 influenza A infection. Neurocrit Care2010;13:98–100.

22. Cortese A, Baldanti F, Tavazzi E, et al. Guillain-Barre syndrome associated with theD222E variant of the 2009 pandemic influenza A (H1N1) virus: case report andreview of the literature. J Neurol Sci 2012;312:173–176.

23. Vasconcelos A, Abecasis F, Monteiro R, et al. A 3-month-old baby with H1N1 andGuillain-Barre syndrome. BMJ Case Rep 2012;2012:bcr1220115462.

6 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 85: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

ARTICLE OPEN ACCESS

Age mattersImpact of data-driven CSF protein upper reference limits in Guillain-Barresyndrome

Pierre R. Bourque, FRCP(C), John Brooks, FRCP(C), Christopher R. McCudden, FRCP(C),

Jodi Warman-Chardon, FRCP(C), and Ari Breiner, FRCP(C)

Neurol Neuroimmunol Neuroinflamm 2019;6:e576. doi:10.1212/NXI.0000000000000576

Correspondence

Dr. Bourque

[email protected]

AbstractObjectiveWe conducted a retrospective review of patients with a diagnosis of Guillain-Barre syndrome(GBS) to assess the diagnostic impact of applying age-adjusted upper limits for CSF totalprotein (CSF-TP) supported by a systematic literature review.

MethodsCases coded as GBS or inflammatory neuropathy for the period 2001–2016 at The OttawaHospital were reviewed. Cases were included if they met the Brighton criteria for GBS witha diagnostic certainty level 1 or 2 and had contemporaneous CSF-TP data. We excluded caseswith CSF pleocytosis >50 and cases with Miller-Fisher syndrome. Age-adjusted reference limitswere compared with conventional 0.45 and 0.6 g/L upper limits.

ResultsOne hundred thirty-eight cases met the study criteria, with a mean age of 47 years. The meaninterval from symptom onset to lumbar puncture was 7.9 days, andmean CSF-TPwas 1.23 g/L.There was a strong correlation between rising CSF-TP and time to lumbar puncture. Age-adjusted CSF-TP had a significantly lower sensitivity of only 45% in the first week (32% in thefirst 3 days) compared with 70% in the first week for the 0.45 g/L limit. All upper limits gainedhigh sensitivity after the first week.

ConclusionsThe low sensitivity of CSF-TP for the diagnosis of GBS is exacerbated by age-adjusted upperlimits. The main role of lumbar puncture in GBS in the first week may be to help exclude otherinflammatory or neoplastic etiologies of acute neuropathy. After the first week, the magnitudeof the CSF-TP rise reduces the effect of different upper reference limits.

From The Ottawa Hospital (P.R.B., J.B., J.W.-C., A.B.), University of Ottawa; The Ottawa Hospital Research Institute (P.R.B., C.R.M., J.W.-C., A.B.); and Department of Pathology andLaboratory Medicine (C.R.M.), The Ottawa Hospital, Ottawa, Canada.

Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by the authors.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 86: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

The presence of albuminocytologic dissociation is a well-validated diagnostic criterion for Guillain-Barre syndrome(GBS).1 Its severity is correlated with the timing of lumbarpuncture. CSF total protein (CSF-TP) may only be a crudesurrogate measure of disrupted blood-nerve barrier perme-ability and intrathecal antibody synthesis compared withserum/CSF ratios of albumin and immunoglobulin (Ig) G.2

Nonetheless, CSF-TP remains the most frequently performedCSF diagnostic test worldwide in the assessment of suspectedpolyradiculoneuropathy of acute onset, and it may significantlyinfluence the decision to initiate treatment with IVIg or plas-mapheresis. Electrodiagnostic studies also offer critical clues todetect and characterize peripheral nerve demyelination andaxonopathy but are less routinely available worldwide and arenormal in up to 20% of patients in the first week.3

Recent rigorous laboratory studies and a systematic review ofthe literature have validated significantly higher upper refer-ence limits (URLs) than the 0.45 g/L (45 mg%) limit com-monly used by hospital laboratories worldwide andcommonly referenced in the literature.4 This study exploresthe impact of newer data-driven age-adjusted upper referencevalues for CSF-TP for the diagnosis of GBS.

MethodsAll cases coded at the time of hospital discharge as either“Guillain-Barre syndrome” or “Inflammatory Neuropathy” forthe years 2001–2016 were reviewed by a neuromuscularspecialist (P.R.B). Cases were included in the analysis if theymet the Brighton criteria5 for GBS with a diagnostic certaintylevel 1 or 2 and had contemporaneous CSF-TP data. Allpatients had to present a mostly symmetric progressive pat-tern of limb and/or cranial nerve weakness with a monophasiccourse peaking within 28 days. Cases with a CSF white bloodcell count greater than 50 were excluded. Cases with Miller-Fisher syndrome, acute motor axonal neuropathy, and acutemotor and sensory axonal neuropathy were excluded, butfocal variants with bifacial or pharyngo-brachial onset wereincluded. Clinical onset of GBS was defined as the earliestdocumentation of sustained new paresthesia or sustainedsymmetrical muscular paresis.

For the correlation between CSF-TP and days-to-lumbarpuncture, an F-statistic test was performed, and a R2 co-efficient calculated. Days-to-lumbar puncture times weregrouped by weekly intervals. The diagnostic sensitivity ofCSF-TP was calculated as abnormal results/all cases of GBS.This was first calculated using our published institutionalupper reference values.6 This data set, based on 6,068 sam-ples, establishes URLs that are age adjusted, with

representative values of 0.54 g/L at age 30 years, 0.57 g/L atage 40 years, 0.60 g/L at age 50 years, 0.63 g/L at age 60 years,and 0.67 g/L at age 70 years. Diagnostic sensitivity was alsocalculated for the widely used fixed 0.45 g/L upper conven-tional limit and with a fixed 0.6 g/L upper limit. These 2 age-independent limits were compared with the age-adjustednorms using the McNemar test.

Standard protocol approvals, registrations,and patient consentsThe study was approved by the Ottawa Hospital ResearchInstitute Ethics Board (protocol 20160863).

Data availabilityAnonymized data will be shared by request from any qualifiedinvestigator.

ResultsOf 392 cases coded as GBS in the database, 152 were excludedbased on inconclusive documentation, coding error, or a morelikely non-GBS diagnosis. Forty-eight GBS cases were ex-cluded for lack of contemporaneous CSF-TP results, and 19cases were excluded because their subsequent clinical coursewas diagnostic of chronic inflammatory demyelinating poly-neuropathy. Fifteen of these patients presented as acute-onsetchronic inflammatory demyelinating polyneuropathy, initiallymimicking acute inflammatory demyelinating poly-neuropathy, with a mean CSF-TP of 1.18 g/L for the 13 casessubjected to lumbar puncture. Eighteen cases were excludedbecause they fulfilled the criteria for Miller Fisher syndrome.One hundred thirty-eight cases met all criteria (mean age 47years, 56%males). Electrodiagnostic studies were available forreview in all but 6 patients.

The mean number of days between clinical onset and lumbarpuncture was 7.9 days (range, 1–25 days). The mean CSFwhite blood cell count was 3.0/μL. The mean CSF-TP for allGBS cases was 1.23 g/L (range, 0.28–8.62 g/L). There wasa correlation of rising CSF-TP with increasing number of daysbetween clinical onset and the number of days to lumbarpuncture (p = 0.00048, figure). During the first week afteronset, applying data-driven reference intervals resulted ina sensitivity of only 45% compared with a sensitivity of 70%with the conventional fixed 0.45 g/L upper limit (table). Theage-adjusted URL had even lower sensitivity (32%) in the 0–3day interval. The age-adjusted upper limits did not performsignificantly differently compared with the fixed 0.6 g/L upperlimit. All 3 upper limits had statistically similar high sensitiv-ities in the range of 74%–87% in the second week and92%–100% subsequently.

GlossaryCSF-TP = CSF total protein; GBS = Guillain-Barre syndrome; Ig = immunoglobulin; URL = upper reference limit.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 87: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

DiscussionConsensus guidelines for the diagnosis of GBS allow the in-stitution where testing is performed to define their CSF-TPURL.5 A laboratory practice survey revealed that over 85% ofclinical centers worldwide presently use an antiquated arbitraryage-independent upper limit of 0.45 g/L, dating back to theseminal 1938 monograph of Houston Merritt.7 A recent sys-tematic review supports significantly higher URL that shouldbe stratified by age, as was used in the present study.4

Our study reaffirms the finding of time-dependent CSF-TPelevation in patients with GBS, as described in series poolingdata from observational and therapeutic studies8 or fromseveral Asian countries.9 Our study shows that within the firstweek, the sensitivity of CSF studies is quite low, in the order of45%. Using an age-independent CSF-TP URL of 0.45 g/Lmight artificially boost sensitivity to 70% within the first weekof GBS onset, but risks lowering specificity early in the diseasecourse, when costly treatment decisions are made, and accu-rate diagnosis is paramount.

The other commonly used value of 0.60 g/L, quoted in somerecent laboratory reference texts, more closely matched the

validated age-dependent reference values. It may constitutea reasonable better simple approximation for clinical practice.All URLs tested had similar high sensitivities for lumbar punc-tures performed more than 1 week after symptom onset, likelybecause of the large magnitude of the CSF-TP elevation thateventually occurs in GBS, with a mean value of 1.23 g/L in ourdata set. The diagnostic yield of CSF-TP in the first week mayalso be influenced by the criterion used to define the GBS timeof onset. We judged that prominent sustained new paresthesiaor cranial neuropathy would be considered by most clinicians tobe a reliable first index manifestations of GBS. In the study ofFokke et al,8 onset of weakness was the sole criterion, whichwould lead to lower values for the time to lumbar puncture.

The main potential benefit of raising CSF-TP URL, usingdata-driven age-adjusted values supported by the recent lit-erature, is a predictable gain in specificity. A limitation of ouranalysis, similar to other series, is the lack of data necessary toaccurately measure specificity. This would require a system-atic hospital chart review to also capture all cases in whichGBS was a suspected diagnosis at the time of lumbar punctureand GBS was excluded as a final diagnosis. Nonetheless, thisstudy highlights that CSF-TP has low sensitivity in the first

Figure Relation of CSF-TP value and time between clinical onset and lumbar puncture

The Tukey method was used to remove outliers from the data setbefore computing the trend line. Although the F-statistic showsa positive correlation in the relationship between CSF-TP and timeto lumbar puncture (p = 0.00048), a large degree of variability inCSF-TP values remains unaccounted for (R2 = 0.071). CSF-TP = CSFtotal protein.

Table Diagnostic sensitivity at different time intervals and CSF-TP reference limits

0–6 days 7–13 days >13 days

Age-adjusted reference data 0.45 (30/66) 0.74 (34/46) 0.92 (24/26)

0.6 g/L 0.47 (31/66) p = 1 0.76 (35/46) p = 1 0.96 (25/26) p = 1

0.45 g/L 0.7 (46/66) p < 0.001a 0.87 (40/46) p = 0.041 1 (26/26) p = 0.480

Abbreviations: GBS = Guillain-Barre syndrome; CSF-TP = CSF total protein.Sensitivity is calculated as GBS cases with elevated protein/all GBS cases (numbers in brackets). It is expressed as a function of the time interval (clinical onsetto lumbar puncture) and 3 different upper limits: age adjusted (institutional reference study), 0.60, and 0.45 g/L. The p values are computed using theMcNemar test, comparing 0.6 and 0.45 g/L to the age-adjusted reference data. The threshold for significance using the Bonferroni correction is 0.008.a There is a significant difference for the 0.45 g/L limit at the 0–6 day interval only.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 88: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

week (when lumbar puncture mostly helps to rule out in-fectious or neoplastic disease) and only acquires a true highsensitivity for GBS by the second week.

Study fundingNo targeted funding reported.

DisclosureThe authors report no disclosures. Go to Neurology.org/NNfor full disclosures.

Publication historyReceived by Neurology: Neuroimmunology & Neuroinflammation March7, 2019. Accepted in final form March 28, 2019.

References1. Asbury AK, Cornblath DR. Assessment of current diagnostic criteria for Guillain-

Barre syndrome. Ann Neurol 1990;27(suppl):S21–S24.2. Deisenhammer F, Bartos A, Egg R, et al. Guidelines on routine cerebrospinal fluid

analysis: report from an EFNS task force. Eur J Neurol 2006;13:913–922.3. Uncini A, Kuwabara S. Electrodiagnostic criteria for Guillain-Barre syndrome:

a critical revision and the need for an update. Clin Neurophysiol 2012;123:1487–1495.

4. Breiner A, Moher D, Brooks J, et al. Adult CSF total protein upper reference limitsshould be age-partitioned and significantly higher than 0.45 g/L: a systematic review.J Neurol 2019;266:616–624.

5. Sejvar JJ, Kohl KS, Gidudu J, et al. Guillain-Barre syndrome and Fisher syndrome: casedefinitions and guidelines for collection, analysis, and presentation of immunizationsafety data. Vaccine 2011;29:599–612.

6. McCudden CR, Brooks J, Figurado P, Bourque PR. Cerebrospinal fluid total proteinreference intervals derived from 20 years of patient data. Clin Chem 2017;63:1856–1865.

7. Bourque PR, Breiner A, Moher D, et al. Adult CSF total protein: higher upperreference limits should be considered worldwide: a web-based survey. J Neurol Sci2019;396:48–51.

8. Fokke C, van den Berg B, Drenthen J, Walgaard C, van Doorn PA, Jacobs BC.Diagnosis of Guillain-Barre syndrome and validation of Brighton criteria. Brain 2014;137:33–43.

9. Wong AH, Umapathi T, Nishimoto Y, Wang YZ, Chan YC, Yuki N. Cytoalbumi-nologic dissociation in Asian patients with Guillain-Barre and Miller Fisher syn-dromes. J Peripher Nerv Syst 2015;20:47–51.

Appendix Authors

Author Location Role Contributions

Pierre R.Bourque,FRCP(C)

Ottawa,Canada

First andcorrespondingauthor

Study concept anddesign, acquisition ofdata, interpretation andanalysis of data, andstudy supervision

John Brooks,FRCP(C)

Ottawa,Canada

Author Statistical analysis ofdata, interpretation andanalysis, and criticalrevision of themanuscript

Christopher R.McCudden,FRCP(C)

Ottawa,Canada

Author Interpretation andanalysis of data andcritical revision of themanuscript forimportant intellectualcontent

Appendix (continued)

Author Location Role Contributions

Jodi Warman-Chardon,FRCP(C)

Ottawa,Canada

Author Critical revision of themanuscript forimportant intellectualcontent

Ari Breiner,FRCP(C)

Ottawa,Canada

Author Interpretation andanalysis of data andcritical revision ofthe manuscript forimportant intellectualcontent

4 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 89: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

ARTICLE OPEN ACCESS

Neuromyelitis optica spectrum disorderPatient experience and quality of life

Janine Beekman, PhD, Aysha Keisler, PhD, Omar Pedraza, MPH, Masayuki Haramura, PhD,

Athos Gianella-Borradori, MD, Eliezer Katz, MD, John N. Ratchford, MD, Gerard Barron, BSc,

Lawrence J. Cook, PhD, Jacinta M. Behne, MS, Terrence F. Blaschke, MD, Terry J. Smith, MD, and

Michael R. Yeaman, PhD

Neurol Neuroimmunol Neuroinflamm 2019;6:e580. doi:10.1212/NXI.0000000000000580

Correspondence

Dr. Beekman

[email protected]

or Dr. Yeaman

[email protected]

AbstractObjectiveTo gain insights into NMOSD disease impact, which may negatively affect QoL of patients,their families, and social network.

MethodsThe current study used validated instruments to assess physical, emotional, and socioeconomicburden of NMOSD on QoL among 193 patients.

ResultsA majority of patients reported an initial diagnosis of a disease other than NMOSD. Overall,two-thirds of patients reported NMOSD as having a strong negative impact on physical health(Short Form-36 [SF-36] score 27.1 ± 39.1), whereas emotional well-being was relativelyunimpaired on average (SF-36 score 54.0 ± 44.9). A subset of patients reported having thehighest category of emotional health despite worse physical health or financial burden, sug-gesting psychological resilience. Pain (r = 0.61) and bowel/bladder dysfunction (r = 0.41)imposed the greatest negative physical impact on overall QoL. In turn, ability to work correlatedinversely with worsened health (r = −0.68). Increased pain, reduced sexual function, inability towork, and reduced QoL had greatest negative impacts on emotional well-being. Dissatisfactionwith treatment options and economic burden correlated inversely with QoL.

ConclusionsCollectively, the current findings advance the understanding of physical, emotional, social, andfinancial tolls imposed by NMOSD. These insights offer potential ways to enhance QoL bymanaging pain, enhancing family and social networks, and facilitating active employment.

From the Ipsos Public Affairs (J.B., A.K., O.P.), Washington, DC; Chugai Pharmaceutical Co., Ltd. (M.H.), Chuo-ku, Tokyo, Japan; Chugai Pharma USA, Inc., (A.G.-B.), Berkeley Heights, NJ;Viela Bio (E.K., J.N.R.), 1 MedImmune Way, Gaithersburg, MD; MedImmune Ltd. Riverside Building (G.B.), Granta Park, Cambridge, UK; Department of Pediatrics (L.J.C.), University ofUtah, Salt Lake City, UT; The Guthy-Jackson Charitable Foundation (J.M.B.), Beverly Hills; Departments of Medicine and of Molecular Pharmacology (T.F.B.), Stanford UniversitySchool of Medicine, Stanford, CA; Department of Ophthalmology and Visual Sciences (T.J.S.), Kellogg Eye Center and Division of Metabolism, Endocrine and Diabetes, Department ofInternal Medicine, University of Michigan Medical School, Ann Arbor, MI; Department of Medicine (M.R.Y.), University of California, Los Angeles, Los Angeles; Divisions of MolecularMedicine and Infectious Diseases, Harbor-UCLA Medical Center; and Los Angeles Biomedical Research Institute at Harbor-UCLA Medical Center, Torrance, CA.

Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by The Guthy-Jackson Charitable Foundation.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 90: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Neuromyelitis optica spectrum disorder (NMOSD) is a po-tentially life-threatening neuroinflammatory disease targetingthe optic nerve, spinal cord, and brain.1–4 Relapses result incumulative neurologic disabilities, are unpredictable, and areinterspersed with remissions. Increased diagnostic accuracy andincreased health care provider awareness have resulted in in-creased prevalence up to 10/100,000 in some geographicregions.5–8 This estimate equates to >15,000 US patients and>100,000 cases worldwide. NMOSD disproportionately affectsfemales (up to 7:1).9,10 Positive anti–aquaporin-4 (AQP4)antibody neuromyelitis optica (NMO-IgG) is the most com-mon disease serotype; however, titers fail to predict diseasecourse.11,12 Recent evidence13,14 suggests that cases positive foranti–myelin oligodendrocyte glycoprotein (MOG) antibody(MOG-IgG) are pathogenically distinct from NMOSD.

Although studies suggest therapeutic benefit, no treatment ofNMOSDhas been found to be safe and effective in prospective,adequately powered clinical trials. However, recent results fromphase IIb and phase III trials are encouraging.15–17 In thesetrials, biologic therapeutics being evaluated include those tar-geting the complement C5 protein, the intereukin-6 receptor,and CD-19 protein on B cells. In addition, new and importantscientific insights have recently shed light on key mechanismsunderpinning NMOSD pathogenesis that may represent tar-gets for next-generation therapeutics.18–21

By comparison, few studies have systematically examined theimpact ofNMOSDon quality of life (QoL) in well-characterizedcohorts.22–25 Therefore, The Guthy-Jackson Charitable Foun-dation, Alexion Pharmaceuticals, Chugai Pharmaceutical Co.,MedImmune/Viela Bio, and Ipsos Public Affairs conducteda cooperative study of NMOSD patient experience and QoL.Through an interactive survey format, patient-reported clinical,demographic, and experiential data were systematically collectedfrom geographically dispersed patients with NMOSD acrossNorth America. The current analyses yielded novel insights thatmay afford potentially modifiable aspects of personal or clinicalcare to improve QoL in patients with NMOSD.

MethodsClinical research standards

Human subjects protectionThe study was conducted in accordance with 45 Code ofFederal Regulations Part 46 and the US Department of Healthand Human Services policies regarding conductance of HumanSubject Research. Protocols, survey instruments (figure e-1,links.lww.com/NXI/A120), and informed consent documents

were approved by a central institutional review board. Writtenand verbal consent/assent were obtained before enrollment.

Special population complianceThe online survey instrument was compliant with theAmericans with Disabilities Act. Participants were given theoption of completing the survey with assistance of a relative,friend, or caregiver if physical impairments precluded in-dependent participation.

Study goals and design

Study goalsGoals were (1) to gain understanding of the natural history ofNMOSD from a patient-reported perspective and (2) to as-sess NMOSD patient QoL using a rigorous survey method-ology comprising standardized and NMOSD-specific QoLmeasures. Both goals were intended to identify how NMOSDaffects patients and in so doing identify those aspects thatmight be modified.

Study designThis study used a cross-sectional survey design. Comparativedisease data were derived from published studies, which usedidentical standardized measures and for which parallel de-mographic and QoL data sets were available.

Survey themesThe survey instrument assessed multiple disease impacts us-ing quantitative Likert scales with dynamic ranges respectiveof each theme:

Health-related QoLThree validated scales were used to assess the impact ofNMOSD on health-related QoL: (1) select items from theRole-Physical and Role-Emotional subscales of the ShortForm-36 (SF-36) measured the impact of NMOSD onphysical and emotional health.26 Scores on the SF-36 sub-scales ranged from 0 to 100 (100 = highest functioning and0 the lowest). The scale is normalized to average US indi-viduals having a score of 50; (2) the MS QoL scale27 mea-sured effects of pain, bowel and bladder, and sexual function;(3) the Impact of Visual Impairment Scale assessed of visualimpairment affected perceived QoL.28

Perceived impact of NMOSD on daily livingParameters were measured by (1) overall QoL (distinct fromhealth-related QoL); (2) perceived impact on career; (3) sociallife; (4) personal relationships; (5) reproduction choices; (6)NMOSD-related pregnancy complications; and (7) degree towhich living situation was determined by necessity.

GlossaryANOVA = analysis of variance; AQP4 = aquaporin-4; GJCF = Guthy-Jackson Charitable Foundation; MOG = myelinoligodendrocyte glycoprotein; NMO = neuromyelitis optica; NMOSD = neuromyelitis optica spectrum disorder; QoL =quality of life; SF-36 = Short Form-36.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 91: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Diagnostic experienceMeasures included date of initial diagnosis (month/year); pre-senting symptoms; and diagnostic history, including time be-tween first symptoms, initial diagnosis, and NMOSD diagnosis.

Treatment experienceThe following data were collected: treatment history; reason fortreatment change; date of most recent treatment change; per-ceived effectiveness of current NMOSD treatment; concernsregarding current treatment; and outlook on future treatment.

Relapse experienceRelapses in the previous year were measured using severalaspects of impact and severity, including total number andfrequency of clinically-confirmed relapses; number requiringinpatient hospitalization; treatment regimens received forrelapses; average duration of relapses; and frequency ofemergency/urgent care for NMOSD.

Health care experienceEvolution of patient interactions with health care professionalswas assessed by first presentation to a health care provider withsymptoms consistent with NMOSD; initial referral to anNMOSD specialist; specialty of physician diagnosingNMOSD; factors influencing choice of current physician; fre-quency of scheduled clinical evaluations; and level of satisfac-tion with NMOSD physician/health care provider.

Economic burdenSpecific financial impact of NMOSD was estimated via timespent traveling to/from medical appointments; method oftransportation; need for in-home professional care; total costsand annual out-of-pocket expenses for care; financial supportreceived; burden of monthly out-of-pocket expenses; andperceived sufficiency of health care insurance.

Future uncertaintyFuture concerns of worsening of disease and unpredictabledevelopment of improved therapies were assessed.

Survey translationTranslation of the survey into Spanish involved a 2-stepprocess. First, a native-speaking translation linguist revieweddocuments against the source English file for consistency,terminology, and syntax. Next, a computer-aided translationtool (Translation Workspace XLiff Editor, v.2.49.1)29 wasapplied to review the instrument in contextual modules, re-solving semantic ambiguity.

Eligibility and enrollment

EligibilityParticipants were recruited from an opt-in digital mailing listof 2,000 individuals in the NMO advocacy community whorequested information. Those fulfilling inclusion criteria werestudy eligible: self-reported, established diagnosis of NMO orNMOSD30,31 and the ability to read textual content or hearquestions audibly and respond to questions.

Enrollment and implementationEligible subjects were consented and enrolled. The survey in-strument was implemented either by telephone or via an onlineinterface. Both modalities offered assistance through a clinicalstudy coordinator and provided options allowing completion in1 session or to complete over multiple sessions. Patient-reported clinical data assessed are summarized in figure e-1(links.lww.com/NXI/A120). Survey completion most com-monly occurred in the patient or caregiver residence. Care-givers sometimes assisted the patient and investigator in use ofthe computer interface or in patient historical recall.

Informatics and data securityStudy data were collected using a web-accessible electronicdata capture system with access limited to qualified studypersonnel. Each patient data set was curated for quality, in-ternal consistency, and completeness.

Statistical analysesDescriptive statistics (medians or interquartile ranges for nu-meric variables; counts or percentages for categorical variables)were evaluated to assess cohort demographic diversity. Pairwiseanalysis of variance (ANOVA), χ2 tests, and Pearson or Spear-man correlation analyses were used to assess magnitude andorientation of relationships between or among study variables.All analyses were performed in SPSS v. 25.0.32 Probability values(p) <0.05 and correlation values (r) > or < 0.5 were consideredsignificant.

Data availabilityDeidentified data obtained using the survey instrument usedin the current study (figure-e1, links.lww.com/NXI/A120)will be made available to qualified research personnel in ac-cordance with institutional review board policies and uponrequest approximately 6 months following the final publica-tion date.

ResultsCohort demographics

Sex, race, and ethnicityThe study population was predominantly female (N = 171;88.6%) and comprised diverse racial/ethnic backgrounds: 71.5%Caucasian/white; 16.1% African American/black; 6.7% Asian(AS); 6.7%Hispanic/Latina/o or Spanish American; 0.5%NativeAmerican, 1.0%Pacific Islander; and 2.6%Other. The distributionexceeds 100% because individuals could select multiple categories.The sample was predominantly English speaking; 2 participantsrequested the Spanish survey. Race and ethnicity distribution wasgenerally representative of the U.S. population, but reflecteda smaller proportion of HL participants than expected.

EducationTwenty-eight percent reported completing a primary or highschool education or general educational development (figure1A). Twenty-one percent hold an associates or technical

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 92: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

degree, 30.6% hold a bachelor’s degree, and 18.1% havea postgraduate education or professional degree. Typical ofonline survey research, the sample skewed slightly to a greaterproportion of subjects having a higher level of education thanthe general US population.

Employment statusApproximately 35% of the study cohort (N = 67) reportedcurrent employment (figure 1A), ranging from full-time(≥40 h/wk; 21.2%) to part-time work. Nine unemployedrespondents (4.6%) reported that they are actively seeking em-ployment. Of those unemployed, 18 are full-time homemakersor caregivers, 22 are retired, and 1 is a student.Most unemployedrespondents (63.7%; N = 79) reported being disabled.

IncomeHousehold annual income varied widely among the studysubjects (figure 1A). The study population comprised a smallerproportion of participants who earned less than $10,000 peryear compared with the broader US demographic.

Residential status and childrenStudy participants resided in one of 43 US states and theDistrict of Columbia, whereas 11 participants resided inCanada. The modal state of residence was California (N =27; 14.8%). The majority of participants (70.5%) livedwith their spouse/partner; 38.3% with their children;10.4% were living alone at the time of study. Nonereported living with domestic assistance or in an

Figure 1 (A) Demographic patterns among the present study cohort

(B) Relationship between physical and emotional healthfunctioning in NMO/SD. Criteria were based on SF-36 role-physical and role-emotional health measures. The relativesize of the circle represents the number of respondents witha given score. Most respondents fell within one of 3 catego-ries, as labeled. Note that the upper left category representsa particularly resilient group of patients with very poorphysical health but very robust emotional health.

4 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 93: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

institutional domicile/care facility. Most participants(73.1%) had children.

Survey assistance and future researchNineteen participants (<10%) received assistance in surveycompletion. One participant participated by telephone. Over92% of study subjects (N = 179) would consider participatingin a future study, whereas 8 (4.1%) declined consideringa future survey and 6 (3.1%) declined to answer this question.

Overall QoL

Physical and emotional healthRole-physical scores were relatively low but exhibited widevariability (median = 27.1 ± 36.1). Role-emotional

functioning was near-average, with broad variance(median = 54.0 ± 44.9). Data exhibited bimodal distribu-tion, with participants chiefly reporting either low or highfunctioning (figure 1B). Although physical and emotionalhealth were positively correlated (r = 0.513; p < 0.05), thedata also highlighted a complex health continuum(figure 1B).

Comparative QoLTo contextualize NMOSD QoL, SF-26 data were comparedwith data examining other autoimmune/inflammatory dis-orders (table 1). Where results are summarized physically,NMOSD impact on QoLwas rated similarly to systemic lupuserythematosus. Emotional impact of NMOSD was rated as

Table 1 Comparative impact and determinants of NMO/SD impact on QoL

Comparative

Sample size

Physicala Emotionala

SourceDisease cohort M SD M SD

Current study 193 27.1 39.1 54.0 44.9 —

Other NMOb 30 36.0 10.7 46.7 10.9 Zhao et al.14

MSc 368 18.0 NA 52.0 NA Riazi et al.35

Parkinson diseasec 227 19.0 NA 34.0 NA Riazi et al.35

Systemic lupus erythematosus 1,316 36.3 41.5 54.5 43.9 Wolfe et al.36

Amyotrophic lateral sclerosis 679 18.2 33.1 47.3 46.2 Jenkinson et al.37

Rheumatoid arthritis 13,722 39.9 42.0 63.5 42.4 Wolfe et al.36

NI rheumatic disorders 3,623 39.5 41.6 65.6 41.4 Wolfe et al.36

Antiphospholipid syndrome 270 43.5 49.6 56.8 49.4 Georgopoulou et al.33

Fibromyalgia 2,733 19.2 32.3 43.9 43.9 Wolfe et al.36

Determinant Range Mean SD

Overall QoL 1–6 4.58 1.41

Bodily pain 1–6 3.60 1.31

Impaired career 1–6 3.30 1.96

Ability to work at job 1–6 3.19 1.96

Affected choice whether to have children 1–6 2.11 1.88

NMO/SD-specific issue

Bowel/bladder function interfering with normal activities 1–5 2.26 1.28

Interfered with day-to-day work (inside or outside the home) 1–5 2.76 1.25

Satisfaction with sexual function 1–5 2.40 1.23

Social and personal relationships

Social life 1–10 5.40 3.05

Personal and family relationships 1–10 5.66 2.80

Abbreviations: MCS = Mental Health Component; NA = not available; NI = non-inflammatory; NMO = neuromyelitis optica; PCS = Physical ComponentSummary; QoL = quality of life; SF = Short Form.a SF-36 scores on role-physical (physical functioning) and role-emotional (emotional functioning) of respondents diagnosed with NMO or NMOSD vs othercomparison conditions of similar heterogeneity.b Zhao et al report the SF-36 PCS score andMCS Summary Scale, a broader scale, which contains the role-physical and role-emotional subscales but capturesa broader range of physical and mental/emotional functioning.c Riazi et al. did not report SDs for subscale means.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 5

Page 94: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

equivalent to MS, systemic lupus erythematosus, and anti-phospholipid syndrome.

NMOSD-specific experience

Impact of disease on QoLSpecific impacts of NMOSD on QoL are summarized in table1. On average, NMOSD imposed a significant negative effect(mean = 4.58 ± 1.41; scale 1–6 [1 = least impact, 6 = greatestimpact]); >70% reported QoL to be greatly affected. Deter-minants most associated with negative QoL were pain, impacton career, and ability to work. Other factors were painimpairing day-to-day tasks, impact on social activities, bowel/bladder dysfunction, and satisfaction with sexual function. Ofinterest, diagnosis of NMOSD failed to strongly influence thedecision to have children.

Initial presenting Symptom(s)The most common initial presenting symptoms (table 2)were numbness and/or tingling (68.4%), difficulty walking(54.4%), and visual disturbances (52.8%). Other pre-senting symptoms are as in figure-e1 (links.lww.com/NXI/A120).

Accuracy of initial diagnosisNearly two-thirds of the cohort (N = 125; 64.8%) reported aninitial diagnosis other than NMOSD. The most frequent wereMS (N = 80; 41.4%) or nonspecific optic neuritis (N = 44;22.7%) (table 2).

Demographics and serologic statusParticipants ranged in age from 19 to 76 years (mean = 49.2 ±12.8 years), whereas 13–73 years (mean = 44.7 ± 12.5 years)at diagnosis and 3months to 22 years (mean = 5.0 ± 3.8 years)from diagnosis to study enrollment. Eighty-two percent carrythe diagnosis of NMO (N = 158), whereas 18.1% (N = 35)were diagnosed with NMOSD. Among the entire study co-hort, 118 (61.1%) reported being anti–aquaporin 4 antibody(NMO-IgG) seropositive, 41 (21.2%) NMO-IgG seronega-tive, and 34 (17.6%) did not know.

Diagnostic or treatment delaysTime from initial symptoms to correct diagnosis ranged from0 (i.e., immediate NMO diagnosis) to 40 years (mean = 3.3 ±6.3 years). The time from correct diagnosis to treatmentinitiation ranged from 0 to 11 years (mean = 6 months ± 1.7years). The median timespan between first symptom andcorrect diagnosis was 6 months, and the median interval tospecific treatment initiation was 3 weeks.

Perceived efficacy of current treatmentThe mean rating of perceived effectiveness of current treat-ment across all participants was 8.2 ± 2.3 on the followingscale (1–10): 10 = treatment works very well; 1 = treatmentdoes not work well or at all. The most common medicationswere rituximab (60.6%), prednisone/corticosteroids (20.2%),and mycophenolate mofetil (17.1%). Of treatments beingprescribed for at least 10% of study subjects, those receiving

rituximab or mycophenolate mofetil reported highest per-ceived efficacy, whereas azathioprine was lowest (table 3).Four participants reported currently receiving no treatment,and 35 (18.1%) reported “other” treatments.

Concerns about treatment optionsMore than 50% of participants (51.8%) reported havingconcerns regarding their NMOSD treatment, mostly focusedon future effectiveness (table 3). Eighty-eight participants(45.6%) reported NMO medication changes over their dis-ease course. The majority of these patients (N = 48, 54.5%)reported changes because of poor efficacy, whereas 32

Table 2 Symptoms and diagnoses of initial diseaseepisode among patients with NMO/SD

Counta Percenta

Initial symptoms

Numbness/tingling 132 68.4

Difficulty walking 105 54.4

Vision problems 102 52.8

Pain 95 49.2

Fatigue 66 34.2

Bladder control problems 51 26.4

Paralysis 45 23.3

Spasticity (sudden involuntary contraction ofa muscle)

45 23.3

Bowel control problems 30 15.5

Protracted vomiting 25 13.0

Cognitive problems (such as memory, mood,and mental effectiveness)

27 14.0

Protracted hiccups 21 10.9

Excessive daytime sleepiness 22 11.4

Depression 20 10.4

Insomnia 17 8.8

Emotional symptoms 14 7.3

Sexual dysfunction 10 5.2

Initial diagnoses

MS 80 64.0

Optic neuritis 44 35.2

Transverse myelitis 37 29.6

Depression 12 9.6

Lupus 9 7.2

Stroke 5 4.0

Abbreviation: NMO = neuromyelitis optica.a Patient may have reported more than 1 diagnosis before NMO/SD.

6 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 95: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

(36.4%) reported intolerable side effects. Three patients(3.4%) changed medication during pregnancy, 4 (4.5%)participated in a clinical trial, and 10 (11.4%) changed therapybecause of cost (table 3).

Impact of relapsesTable 4 summarizes relapse frequency among study partic-ipants. Fifty-two patients (26.9%) reported no relapses.Among the remaining 141 patients, 115 (81.5%) reportedrelapses requiring hospitalization, whereas 26 (18.5%) hadrelapses managed as outpatients. Forty-five participants(23.3%) reported 6 or more relapses, with 2 patients having

≥6 in the previous year. One-hundred twenty-one patientshad not visited an emergency department because of relapsein the past year, whereas 6 others visited an emergency de-partment ≥6 times. Relapses were reported as lasting <4weeks by 95 participants (49.2%) (table 4); 8 patients (4.1%)reported relapses lasting >6 months. Most participants whoexperienced relapses in the past year were treated with eitherIV or (44%) or oral steroids (44%).

Health care experienceThe distribution of health care professionals initially soughtby patients for care is summarized in table 5. Primary care

Table 3 Perceived effectiveness, concerns, and history of treatment among patients with NMO/SD

Current treatment Count Percent Rating of current treatment SD Range

Rituximab 117 60.6 8.76 1.88 1–10

Prednisone/corticosteroid 39 20.2 7.69 2.53 2–10

Mycophenolate mofetil 33 17.1 8.30 2.11 2–10

Azathioprine (Imuran) 28 14.5 7.39 2.36 2–10

PLEX 12 6.2 9.00 1.48 5–10

Investigational drug/clinical trial 3 1.6 10.00 — —

Cyclophosphamide 1 0.5 8.00 — —

Tocilizumab 1 0.5 10.00 — —

Count Percent

Treatment concern

Future treatment effectiveness 56 56.0

Side effects 46 46.0

Ongoing significant disability 23 23.0

Ongoing relapses 19 19.0

Discomfort during administration 13 13.0

Inconvenience 11 11.0

Impact on pregnancy decisions 7 7.0

Treatment history

Azathioprine 47 53.4

Prednisone/corticosteroid 47 53.4

Mycophenolate mofetil 30 15.5

Rituximab 21 23.9

PLEX 16 18.2

IVIG 6 8.8

Cyclophosphamide 5 5.7

Investigational drug/clinical trial 4 4.5

Abbreviations: IVIG = intravenous immunoglobulin; NMO = neuromyelitis optica; PLEX = plasma exchange.Participants could be taking more than 1 medication; rating of current treatment captures the specific medication listed in the table and any othermedications or treatments they were currently undergoing.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 7

Page 96: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

physicians (N = 48; 24.9%) and general neurologists (N =47; 24.5) were most commonly consulted initially, followedby emergency department physicians (N = 41; 21.2%). Themost common initial referral was to a general neurologist(N = 77; 39.8%). Fifty-two patients (26.9%) were referred toNMOSD specialist neurologists, 40 (20.7%) to an ophthal-mologist, and 26 (13.4%) to a gastroenterologist. Generalneurologists were the most common specialists prescribingmedications (;90.6%).

Nearly one-half of the study cohort (N = 82; 42.4%) wereexamined by their doctor every 6 months, most commonlycoincident with rituximab infusion. Thirty-three percent (N =64) of patients saw their doctor at 3-month intervals; 16(8.3%) once per year, 2.6% (N = 5) every month, and 6(3.1%) only at the time of relapse. The remainder reportseeing their doctor “as needed” (N = 4) or not at all (N = 1).

Economic burden of disease

Costs attributable to diseaseStudy participants rated monthly out-of-pocket expenses dueto NMOSD as 5.71 ± 3.12 on a 10-point scale ranging from noburden (1) to significant burden (10). As summarized in table5, prescription medicines accounted for the greatest portionof NMOSDmedical costs for most patients (N = 88; 45.6%);many specified rituximab infusions as the single largest cost.The sole factor predictive of financial burden was receivingplasma exchange therapy during relapse. Twenty-five (13%)reported travel costs to/from health care providers as ac-counting for the greatest cost, whereas 18 (9%) stated

hospitalization accounted for greatest cost. Eleven percent(N = 21) reported that an in-home professional caregiverprovides service. Other significant costs reported were herbalsupplements, psychologist visits, or medical costs not coveredby Medicare.

Total costsBeyond subjective financial burden of disease, the total annualexpenses reported by the cohort as a whole was $1,109,357 oran average of $5,748 per respondent (table 5). The mostfrequently reported cost was prescription medication, with140 participants (73.5%) reporting an average out-of-pocketcost of $1,876 annually. Unprompted (not an original cate-gory of out-of-pocket costs), 3 participants reported lost in-come at an average of $65,000 annually. Although nota categorical option, 23 respondents reported paying forspecialists (including psychologists) out-of-pocket, at an av-erage of $1,554 annually. A large number of participants (N =124; 64.2%) reported travel costs to medical appointments, atan average of $468 per respondent. The largest total cost out-of-pocket was for hospitalization, accounting for $304,410annually in the sample, or an average of $7,248 per re-spondent (N = 42; 21.8%). Caregiver or support was anotherhigh cost, accounting for $70,580 annually in the sample, or anaverage of $3,361 per respondent (N = 21; 10.9%).

Financial support for careThe majority of study participants (N = 142; 73.6%) reportedthat health insurance sufficiently covered prescribed NMOSDmedicines. Among those with insufficient health insurance,expensive copayment and insurer denials were common

Table 4 Total and annual relapse profile of study participants

Relapse frequency 0 1 2 3–5 6+

Relapses ever experienced

Relapses experienced 52 33 22 38 45

Relapses requiring hospitalization 26 40 18 35 18

Relapses in the previous year

Relapses experienced 85 34 7 9 2

Relapses requiring hospitalization 23 17 6 6 0

No. of emergency department visits 121 30 18 18 6

Relapse duration Count Percent

1–7 d 30 22.7

1–2 wk 37 28.0

2–4 wk 28 21.2

1–2 mo 12 9.1

3–4 mo 11 8.3

5–6 mo 6 4.5

More than 6 mo 8 6.1

8 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 97: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Table 5 Health care professionals encountered and annual expenses due to NMO/SD

Health careprofessional

Physician first to evaluatesymptoms

Specialist first toreceive referral

Physician whodiagnosed NMO/SD

Physician currently prescribingmedications

Neurologist(nonspecialist)

47 77 99 175

Neurologist (NMOspecialist)

24 52 83 —

Ophthalmologist 21 40 11 0

Neuro-ophthalmologist 1 3 3 0

Rheumatologist 4 5 0 1

Gastroenterologist 1 26 16 0

Orthopedist 1 3 0 0

MS specialist 0 4 4 5

Primary care physician 48 1 0 11

Emergency departmentphysician

41 1 1 0

Hematologist 0 1 0 13

Physiatrist 0 0 0 1

Other 5 8 5 4

Out-of-pock expenses N Minimum Maximum Median

Prescription medicine(s) 140 $25 $30,000 $540

Travel to clinical care 124 $10 $10,000 $115

Emergency/urgent care 48 $50 $15,000 $275

Medical supplies 52 $50 $3,000 $330

Hospitalization 42 $100 $150,000 $1,950

Caregiver or service 21 $50 $14,000 $1,500

Support groups 7 $8 $360 $100

Other costs 58 $80 $125,000 $1700

Category of expenseaTotal cost to samplerespondents

No. of respondents reportingcost

Average annual cost perrespondent

Prescription medicine(s) (includinginfusions)

$262,598 140 $1,876

Emergency/urgent care $61,275 48 $1,277

Hospitalization $304,410 42 $7,248

Travel costs for medical care $58,003 124 $468

Caregiver or support services $70,580 21 $3,361

Medical supplies $34,173 52 $657

Support group $1,838 7 $263

Supplements $1,000 1 $1,000

Specialists $35,750 23 $1,554

Lost income $195,000 3 $65,000

Health insurance deductible $23,330 9 $2,592

Other costs (unspecified) $61,400 4 $15,350

Continued

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 9

Page 98: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

reasons. Twenty-four study participants (12.6%) reportedreceiving financial support for their NMOSD treatment,largely in the form of disability insurance, clinical trial par-ticipation, or support from friends and family.

Correlation analysisTo examine predictors of overall QoL, ANOVA was used todetect correlates among individual factors, including timesince diagnosis, total relapse number (a surrogate of diseaseseverity), and current treatments (overall and specifically forrelapses). Neither time since diagnosis nor current treat-ment regimen was predictive of overall QoL; however, thenumber of relapses correlated significantly with overall QoL(p = 0.001), with greater numbers of relapses diminish-ing QoL.

Other potential correlates affecting QoL in NMOSD wereexplored using a matrix Pearson or Spearman correlation

analysis of primary data elements. As shown in table 6, mul-tiple correlations were identified as trending to positively ornegatively affecting QoL.

DiscussionThe primary goal of this study was to determine the impactand correlates of NMO on patient QoL in a standardizedmanner using validated measures of physical and emotionalhealth impact on daily activities potentially affecting QoL. Byexamining specific tangible domains of QoL in parallel toperceived overall QoL, the patient experience regarding howthis rare disease affects daily life was revealed.

Several important themes were identified among the currentstudy cohort. First, NMOSD typically has strong negativeeffects on physical functioning. Physical functioning was

Table 5 Health care professionals encountered and annual expenses due to NMO/SD (continued)

Category of expenseaTotal cost to samplerespondents

No. of respondents reportingcost

Average annual cost perrespondent

Collective sample $1,109,357 193 $5,748

Abbreviations: NMO = neuromyelitis optica; QoL = quality of life.a For each cost category, respondents rated the largest burden on QoL and estimated annual expense for each. Costs were totaled by category and averagedfor cost-per-respondent and cost-in-sample estimates. Note: Other costs category included specialists, estimated lost income, health insurance deductibles,and unspecified costs.

Table 6 Exploratory correlation analyses among study data elements†

10 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 99: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

lower in the study cohort than in the general population andon par with individuals having MS or systemic lupuserythematosus.33–36 The predominant physical issues affect-ing QoLwere bodily pain, bowel and bladder dysfunction, andvisual impairment. These factors inversely correlated withability to work, the limitation of which negatively affectedQoL. Age was positively associated with many QoL measuressuch as pain, suggesting that disease increasingly negativelyaffctes QoL over time. Worse physical functioning also cor-related with greater uncertainty about the future. Notably,anti-AQP4 antibody serostatus reported as negative or un-known correlated with less impact on QoL than detectableanti-AQP4 antibody. This relationship is similar to that oftenobserved in AQP4 antibody–positive and MOG antibody–positive phenotypes.11–14 Conversely, seronegative statuscarried a significantly higher financial burden.

By comparison, emotional health was in general un-impaired, suggesting that poor physical health does notnecessarily correspond to diminished emotional health.Although some study participants exhibited congruentemotional and physical health, a subset of participantsreported the highest level of emotional health despite seri-ously impaired physical health. This inverse relationshipsuggests a degree of psychological resilience in somepatients despite physical impairment. Likewise, a portion ofparticipants reported that their disease had a positive effecton their social relationships. One possible explanation forsuch positive impact is that their disease provoked supportnetwork involvement. These findings are consistent with theconcept and impact of psychological resilience,38 which cantranslate to effective personal strategies of coping withhealth-related challenges.39

The constellation of presenting symptoms in many patientsresulted in an initial diagnosis of MS. Inaccurate diagnosiscombined with delay of appropriate therapy can negativelyaffect long-term outcomes in NMOSD.40 However, recentimplementation of international consensus criteria31 has in-creased the timeliness and accuracy of diagnosis and shouldimprove care in early disease.41 The number, duration, andseverity of relapses varied widely across the study cohort. Thisobservation corresponds to the absence of a standardizeddefinition and diagnostic algorithm for differentiating bonafide relapses from unrelated symptoms.

Not surprisingly, participants reporting higher treatmentratings also experienced higher physical and emotional func-tioning and higher QoL. Similarly, worse functioning wasassociated with larger financial burden. These themes areconcordant with those of previous studies.22,23,41 Of interest,patients receiving nonspecific immune-suppressing treat-ments tended to rate their regimens more negatively, whereasthose on target-specific treatments (e.g., biologics) rated theirtreatments more positively. Impact of NMO on QoL ex-tended beyond physical and emotional costs; respondentsreported a high financial burden, particularly for prescription

medicines, travel costs, hospitalization, and specialist care.Furthermore, the per-respondent cost and total cost estimatesin this study provide a useful estimate of personal and healthcare costs of NMOSD to society.

Results of the current study emphasize the significant negativeimpact NMOSD can have on patient QoL, particularly inrelation to physical disability, pain, bowel and bladder dys-function, or visual impairment.42–46 These manifestationscorrespond to reduced ability to work at a job or perform dailyactivities, and a decreased QoL, which also reconcile withnegative impacts of anxiety, disability, or depression inNMOSD.25,47 Factors contributing to these adverse out-comes may include (1) delayed or inappropriate treatmentdue to initial misdiagnosis; (2) real or perceived efficacy orlack of efficacy of current treatment options; (3) lack ofa standard definition of relapse; and (4) disease-specificeconomic burden. These issues underscore the importance ofrecent advances in diagnostic timeliness and accuracy, as wellas ongoing clinical trials intended to establish the first ap-proved therapies for NMOSD. Prospectively, global collabo-ration aimed at implementation of a standard relapsedefinition and severity score should contribute to improvedclinical care. Likewise, the pursuit of predictive biomarkers ofrelapse to allow mitigating interventions and the initiation ofstudies aimed to durably restore immune tolerance as a cura-tive therapy hold promise for increasingly effective medicalsolutions for NMOSD patients. Synergistic and prospectiveapproaches such as these aimed at addressing disease causesand effects hold great promise to significantly add to the QoLfor patients with NMOSD, other patients with rare disease,and beyond.48,49

AcknowledgmentThe authors are deeply grateful to the patients whovolunteered to participate in this research. Alexion Pharma-ceuticals provided input into the study design and courtesyreview of the manuscript. Special appreciation is expressed toMs. Megan Weber for analytic support. This collaborativeproject was supported in-part by The Guthy-JacksonCharitable Foundation, Alexion Pharmaceuticals, Inc., ChugaiPharmaceuticals Co., Ltd., Viela Bio, and MedImmune, Ltd.

Study fundingThis study was sponsored in part by The Guthy-JacksonCharitable Foundation, Alexion Pharmaceuticals, Inc., ChugaiPharmaceutical Co., Ltd., Viela Bio, and MedImmune Ltd.

DisclosureJ. Beekman is an employee of Ipsos Public Affairs, a researchfirm paid to conduct this research study. A. Keisler was anemployee of Ipsos Public Affairs, a research firm paid toconduct this research study. O. Pedraza is an employee ofIpsos Public Affairs, a research firm paid to conduct this re-search study. M. Haramura is an employee of Chugai Phar-maceutical Co., Ltd., which is conducting clinical trialsfocused on NMOSD and a sponsor of the current study.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 11

Page 100: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

A. Gianella-Borradori was an employee of Chugai Pharma-ceutical Co., Ltd., which is conducting clinical trials focusedon NMOSD and a sponsor of the current study. E. Katz andJ.N. Ratchford are employees of Viela Bio, which is con-ducting a clinical trial focused on NMOSD and a sponsor ofthe current study. G. Barron is an employee of MedImmune,which is conducting a clinical trial focused on NMOSD anda sponsor of the current study. L.J. Cook and J.M. Behne aresupported in part by The Guthy-Jackson Charitable Foun-dation, which is a sponsor of the current study. T.F. Blaschke,T.J. Smith, and M.R. Yeaman are advisors to The Guthy-Jackson Charitable Foundation, which is a sponsor of thecurrent study. Go to Neurology.org/NN for full disclosures.

Publication historyReceived by Neurology: Neuroimmunology & NeuroinflammationJanuary 11, 2019. Accepted in final form April 8, 2019.

Appendix Authors

Name Location Role Contributions Disclosure(s)

JennineBeekman,PhD

Ipsos PublicAffairs,Washington,DC

Author Designed/conceptualizedthe study;acquired the data;analyzed the data;interpreted thedata; and draftedthe manuscriptfor intellectualcontent.

Dr. Beekman isan employee ofIpsos PublicAffairs, a researchfirm paid toparticipate in thisresearch study.

AyshaKeisler,PhD

Ipsos PublicAffairs,Washington,DC

Author Designed/conceptualizedthe study;acquired thedata; analyzedthe data;interpreted thedata; and draftedthe manuscriptfor intellectualcontent.

Dr. Keisler was anemployee ofIpsos PublicAffairs, a researchfirm paid toparticipate in thisresearch study.

OmarPedraza,MPH

Ipsos PublicAffairs,Washington,DC

Author Designed/conceptualizedthe study;acquired the data;analyzed the data;interpreted thedata; and draftedthe manuscriptfor intellectualcontent.

Dr. Pedraza is anemployee ofIpsos PublicAffairs, a researchfirm paid toparticipate in thisresearch study.

MasayukiHaramura,PhD

ChugaiPharmaceuticalCo., Ltd.

Author Designed/conceptualized thestudy; interpretedthe data; andreviewed andrevised themanuscript forintellectualcontent.

Dr. Haramura isan employee ofChugaiPharmaceuticalCo., Ltd., which isconductingclinical trialsfocused onNMOSD.

AthosGianella-Borradori,MD

Chugai PharmaUSA, Inc.

Author Designed/conceptualizedthe study;interpreted thedata; andreviewed andrevised themanuscript forintellectualcontent.

Dr. Gianella-Borradori was anemployee ofChugaiPharmaceuticalCo., Ltd., which isconducting clinicaltrials focused onNMOSD anda sponsor of thisstudy.

Appendix (continued)

Name Location Role Contributions Disclosure(s)

EliezerKatz, MD

Viela Bio Author Designed/conceptualizedthe study;interpreted thedata; andreviewed andrevised themanuscript forintellectualcontent.

Dr. Katz is anemployee of VielaBio, which isconductinga clinical trialfocused onNMOSD anda sponsor of thisstudy.

John N.Ratchford,MD

Viela Bio Author Designed/conceptualizedthe study;interpreted thedata; andreviewed andrevised themanuscript forintellectualcontent.

Dr. Ratchford isan employee ofViela Bio, which isconductinga clinical trialfocused onNMOSD anda sponsor of thisstudy.

GerardBarron,BSc (Hons)

Viela Bio Author Designed/conceptualizedthe study;interpreted thedata; andreviewed andrevised themanuscript forintellectualcontent.

Mr. Barron is anemployee ofMedImmune,which isconductinga clinical trialfocused onNMOSD anda sponsor of thisstudy.

LawrenceJ. Cook,PhD,MStat

University ofUtah, Salt LakeCity, UT

Author Designed/conceptualizedthe study;acquired thedata; analyzedthe data;interpreted thedata; andreviewed/revisedthe manuscriptfor intellectualcontent.

Dr. Cook issupported in-partby The Guthy-JacksonCharitableFoundation,which isa sponsor of thisresearch.

Jacinta M.Behne, MS

Guthy-JacksonCharitableFoundation,BeverlyHills, CA

Author Designed/conceptualizedthe study;interpreted thedata; and revisedthe manuscriptfor intellectualcontent.

Ms. Behne issupported in-partby The Guthy-JacksonCharitableFoundation,which isa sponsor of thisresearch.

Terrence F.Blaschke,MD

StanfordUniversity, PaloAlto, CA

Author Designed/conceptualizedthe study;interpreted thedata; and revisedthe manuscriptfor intellectualcontent.

Dr. Blaschke is anAdvisor to TheGuthy-JacksonCharitableFoundation,which isa sponsor of thisresearch.

Terry J.Smith, MD

University ofMichigan, AnnArbor, MI

Author Designed/conceptualizedthe study;analyzed thedata; interpretedthe data; anddrafted themanuscript forintellectualcontent.

Dr. Smith is anAdvisor to TheGuthy-JacksonCharitableFoundation,which isa sponsor of thisresearch.

Michael R.Yeaman,PhD

University ofCalifornia, losAngeles, CA

Author Designed/conceptualizedthe study;analyzed thedata; interpretedthe data; anddrafted themanuscript forintellectualcontent.

Dr. Yeaman is anAdvisor to TheGuthy-JacksonCharitableFoundation,which isa sponsor of thisresearch.

12 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 101: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

References1. Weinshenker BG, Wingerchuk DM. Neuromyelitis spectrum disorders. Mayo Clin

Proc 2017;92:663–679.2. Akaishi T, Nakashima I, Sato DK, et al. Neuromyelitis optica spectrum disorders.

Neuroimaging Clin N Am 2017;27:251–265.3. Marignier R, Cobo Calvo A, Vukusic S. Neuromyelitis optica and neuromyelitis optica

spectrum disorders. Curr Opin Neurol 2017;30:208–215.4. Whittam D, Wilson M, Hamid S, et al. What’s new in neuromyelitis optica? A short

review for the clinical neurologist. J Neurol 2017;264:2330–2344.5. Flanagan EP, Cabre P, Weinshenker BG, et al. Epidemiology of aquaporin-4 auto-

immunity and neuromyelitis optica spectrum. Ann Neurol 2016;79:775–783.6. Houzen H, Kondo K, Niino M, et al. Prevalence and clinical features of neuromyelitis

optica spectrum disorders in northern Japan. Neurology 2017;89:1995–2001.7. Sepulveda M, Aldea M, Escudero D, et al. Epidemiology of NMOSD in Catalonia:

influence of the new 2015 criteria in incidence and prevalence estimates. Mult SclerEpub 2017 Oct 1.

8. Pandit L, Asgari N, Apiwattanakul M, et al. Demographic and clinical features ofneuromyelitis optica: a review. Mult Scler 2015;21:845–853.

9. Bukhari W, Prain KM, Waters P, et al. Incidence and prevalence of NMOSD inAustralia and New Zealand. J Neurosurg Psych 2017;88:632–638.

10. Papp V, Illes Z, Magyari M, et al. Nationwide prevalence and incidence study ofneuromyelitis optica spectrum disorder in Denmark. Neurology 2018;10:1212.

11. Pittock SJ, Lucchinett CF. Neuromyelitis optica and the evolving spectrum of auto-immune aquaporin-4 channelopathies: a decade later. Ann NY Acad Sci;1366:20–39.

12. Kessler RA, Mealy MA, Jimenez-Arango JA, et al. Anti-aquaporin-4 titer is not pre-dictive of disease course in neuromyelitis optica spectrum disorder: a multicentercohort study. Mult Scler Relat Disord 2017;17:198–201.

13. Zamvil SS, Slavin AJ. Does MOG Ig-positive AQP4-seronegative opticospinal in-flammatory disease justify a diagnosis of NMO spectrum disorder ? Neurol Neuro-immunol Neuroinflamm 2015;2:e62. doi:10.1212/NXI0000000000000062.

14. Jarius S, Ruprecht K, Kleiter I, et al. MOG-IgG in NMO and related disorders:a multicenter study of 50 patients. Part 1: frequency, syndrome specificity, influence ofdisease activity, long-term course, association with AQP4-IgG, and origin.J Neuroinflamm 2016;13:279.

15. Alexion Pharmaceuticals. Alexion announces successful phase 3 PREVENT study ofSoiris (Eculizumab) in patients with neuromyelitis optica spectrum disorders(NMOSD). Available at: alexionpharma.com/press-release. Accessed March 25,2019. Also see ClinicalTrials.gov NCT01892345.

16. Chugai Pharmaceutical Co., LTD. Chugai presents results from phase III study ofsatralizumab in NMOSD at ECTRIMS 2018. Available at: chugai-pharm.co.jp.Accessed March 25, 2019. Also see ClinicalTrials.gov NCT02028884.

17. VielaBio Inc. A double-masked, placebo-controlled study with open label period toevaluate MEDI-551 in neuromyelitis optica and neuromyelitis optica spectrum dis-order. Available at: ClinicalTrials.gov. NCT02200770. Accessed March 25, 2019.

18. Bradl M, Reindl M, Lassmann H. Mechanisms for lesion localization in neuromyeltisoptica spectrum disorders. Curr Op Neurol 2018;31:325–333.

19. Papadopoulos MC, Bennett JL, Verkman AS. Treatment of neuromyelitis optica:state-of-the-art and emerging therapies. Nat Rev Neurol 2014;10:493–506.

20. Sato DK, Callegaro D, Lana-Peixoto MA, et al. Seronegative neuromyelitis opticaspectrum—the challenges on disease definition and pathogensis. Arq Neuropsiq2014;72:445–450.

21. Kim SH, KimHJ. A step forward towards personalized immunosuppressive therapy inneuromyelitis optica spectrum disorder. J Neurol Neurosurg Psych 2017;88:619.

22. Moore P, Jackson C, Mutch K, et al. Patient-reported outcome measure for neuro-myelitis optica: pretesting of preliminary instrument and protocol for further de-velopment in accordance with international guidelines. BMJ Open 2016;6:e011142.

23. Eaneff S, Wang V, Hanger M, et al. Patient perspectives on neuromyelitis opticaspectrum disorders: data from the PatientsLikeMe online community. Mult SclerRelat Disord 2017;17:116–122.

24. Schmidt F, Zimmermann H, Mikolajczak J, et al. Severe structural and functionalvisual system damage leads to profound loss of vision-related quality of life in patientswith neuromyelitis optica spectrum disorders. Mult Scler Relat Disord 2017;11:45–50.

25. Shi Z, Chen H, Lian Z, et al. Factors that impact health-related quality of life inneuromyelitis optica spectrum disorder: anxiety, disability, fatigue and depression.J Neuroimmunol 2016;293:54–58.

26. Ware JE, Sherbourne CD. The MOS 36-item Short-Form Health Survey (SF-36): I.Conceptual framework and item selection. Med Care 1992;30:473–483.

27. Vickrey BG, Hays RD, Harooni R, et al. A health-related quality of life measure formultiple sclerosis. Qual Life Res 1995;4:187–206.

28. Weih LM,Hassell JB, Keeffe JE. Assessment of the impact of vision impairment. InvestOphthal Vis Sci 2002;43:927–935.

29. Lionbridge Technologies, Inc. Translation Workspace XLiff editor, version 2.49.1.Waltham:Lionbridge; 2018. Available at: Lionbridge.com. Accessed March 25, 2019.

30. Wingerchuk DM, Lennon VA, Pittock SJ, et al. Revised diagnostic criteria for neu-romyelitis optica. Neurology 2006;66:1485–1489.

31. Wingerchuk DM, Banwell B, Bennett JL, et al. International consensus diagnosticcriteria for neuromyelitis optica spectrum disorders. Neurology 2015;85:177–189.

32. IBM SPSS Statistics Software, Version 25.0. Armonk: IBM Corporation; 2017.Available at: IBM SPSS Software. Accessed March 25, 2019.

33. Georgopoulou S, Efraimidou S, MacLennan SJ, et al. Antiphospholipid (Hughes)syndrome: description of population and health-related quality of life (HRQoL) usingthe SF-36. Lupus 2015;24:174–179.

34. Jenkinson C, Fitzpatrick R, Swash M, et al. The ALS Health Profile Study: quality oflife of amyotrophic lateral sclerosis patients and careers in Europe. J Neurol 2000;247:835–840.

35. Riazi A, Hobart JC, Lamping DL, et al. Using the SF-36 measure to compare thehealth impact of multiple sclerosis and Parkinson’s disease with normal populationhealth profiles. J Neurol Neurosurg Psych 2003;74:710–714.

36. Wolfe F, Michaud K, Li T, et al. EQ-5D and SF-36 quality of life measures in systemiclupus erythematosus: comparisons with rheumatoid arthritis, non-inflammatoryrheumatic disorders and fibromyalgia. J Rheum 2010;37:296–304.

37. Jenkinson C, Hobar J, Chandola T, et al. Use of the short form health survey (SF-36)in patients with amyotrophic lateral sclerosis: tests of data quality, score reliability,response rate and scaling assumptions. J Neurol 2002;249:178–183.

38. Ong AD, Bergeman CS, Bisconti TL, et al. Psychological resilience, positiveemotions, and successful adaptation to stress in later life. J Personal Soc Psych2006;91:730–749.

39. Tugade MM, Fredrickson BL, Feldman-Barrett L. Psychological resilience and pos-itive emotional granularity: examining the benefits of positive emotions on coping andhealth. J Pers 2004;72:1161–1190.

40. Borisow N, Mori M, Kuwabara S, et al. Diagnosis and treatment of NMO spectrumdisorder and MOG-encephalitis. Front Neurol 2018;9:888.

41. Hyun JW, Jeong IH, Joung A, et al. Evaluation of the 2015 diagnostic criteria forneuromyelitis optica spectrum disorder. Neurology 2016;86:1772–1779.

42. Zhao S, Mutch K, Elsone L, et al. Neuropathic pain in neuromyelitis optica affectsactivities of daily living and quality of life. Mult Scler J 2014;20:1658–1661.

43. Sakakibara R. Neurogenic lower urinary tract dysfunction in multiple sclerosis, neu-romyelitis optica and related disorders. Clin Auton Res Epub 2018 Aug 3.

44. Methley AM, Mutch K, Moore P, et al. Development of a patient-centered conceptualframework of health-related quality of life in neuromyelitis optica: a qualitative study.Health Expect 2017;20:47–58.

45. Mutch K, Zhao S, Hamid S, et al. Bladder and bowel dysfunction affect quality oflife: a cross sectional study of 60 patients with aquaporin-4 antibody-positiveneuromyelitis optica spectrum disorder. Mult Scler Relat Disord 2015;4:614–618.

46. Chanson J-B, Zephir H, Collongues N, et al. Evaluation of health-related quality of life,fatigue and depression in neuromyelitis optica. Eur J Neurol 2011;18:836–841.

47. Chavarro VS, Mealy MA, Simpson A, et al. Insufficient treatment of severe depressionin neuromyelitis optica spectrum disorder. Neurol Neuroimmunol Neuroinflamm2016;3:e286. doi:10.1212/NXI.0000000000000286.

48. Yeaman MR, Jackson V. Rare to the Rescue. The Scientist. 2018. Available at: the-scientist.com. Accessed March 25, 2019.

49. Jackson V, Yeaman MR. The Power of Rare. New York: Simon & Schuster; 2017.Available at: simonandschuster.com/The-Power-of-Rare. Accessed March 25, 2019.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 13

Page 102: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

ARTICLE OPEN ACCESS CLASS OF EVIDENCE

Trial of canakinumab, an IL-1β receptorantagonist, in patients with inclusion bodymyositisMichalis L. Kosmidis, MD, Dimitris Pikazis, MD, Panayotis Vlachoyiannopoulos, MD,

Athanasios G. Tzioufas, MD, and Marinos C. Dalakas, MD, FAAN

Neurol Neuroimmunol Neuroinflamm 2019;6:e581. doi:10.1212/NXI.0000000000000581

Correspondence

Dr. Dalakas

[email protected]

AbstractObjectiveTo assess whether canakinumab, a monoclonal antibody against IL-1β approved for auto-inflammatory diseases, is effective as target-specific therapy in patients with sporadic inclusionbody myositis (sIBM).

MethodsBecause in sIBM IL-1β colocalizes with amyloid precursor protein and upregulates amyloidaggregates enhancing degeneration, targeting IL-1β with canakinumab may arrest diseaseprogression. On this basis, 5 ambulatory patients with sIBM participated in an institutionalreview board–approved open-labeled study with 150 mg canakinumab [4 bimonthly, thenmonthly subcutaneous injections] for a mean period of 15.8 months. Patients were assessedbimonthly with a manual dynamometer in 12 proximal and distal muscles and with grip force(GF) in both hands. Total muscle strength (TMS) was expressed in kilograms. Efficacy wasdefined as >15% increased strength after 12 months.

ResultsPatient 1 stopped at month 5 because of 23% loss in TMS and 32.35% in GF; patient 2 showed37.1% increase in TMS and 13% inGF bymonth 9; patient 3 exhibited 26.7% reduction in TMSand 10% in GF at month 33; patient 4 showed 6.5% reduction in TMS and 1.6% in GF after 15months, denoting relative stability; and patient 5 showed 30.4% loss in TMS and 20.8% in GFafter 18 months. In patients 2 and 4, in whom 3-year longitudinal data were available, no effecton disease progression was noted.

ConclusionsIn this long-term, open-label study, canakinumab showed small, but not clinically appreciable,stabilizing benefits in 2 of 5 patients with sIBM over 1 year, was ineffective in 2 others, andmight have worsened one. No patient improved.

Classification of evidenceThis study provides Class IV evidence that canakinumab was ineffective for patients with sIBM.

MORE ONLINE

Class of EvidenceCriteria for ratingtherapeutic and diagnosticstudies

NPub.org/coe

From the Neuroimmunology Unit (M.L.K., D.P., P.V., A.G.T., M.C.D.), Department of Pathophysiology, National and Kapodistrian University of Athens Medical School, Athens, Greece;and Thomas Jefferson University (M.C.D.) Philadelphia, PA.

Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by the authors.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 103: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Sporadic inclusion body myositis (sIBM) is the most com-mon inflammatory myopathy above age 50 years with earlyinvolvement of quadriceps femoris, long finger flexors, biceps,foot dorsiflexors, and mild facial weakness.1 IBM is slowlyprogressive, over years, with an estimated mean annual de-cline in strength by 5.4% based on quantitative musclestrength testing2 or by a 3.87 modified medical researchcouncil score per year.3 The cause is unclear, but bothinflammatory/autoimmune and degenerative changes coexistand may enhance each other.1 The autoimmune componentis mainly characterized by activated and clonally expandedCD8+ cytotoxic T cells invading healthy-appearing and majorhistocompatibility complex-I–expressing muscle fibers, over-expression of costimulatory and adhesion molecules, upre-gulation of chemokines and cytokines, such as interleukin(IL)-1α, IL-1β, IL-2, tumor necrosis factor alpha and trans-forming growth factor, and activation of dendritic cells andB cells.1,4–9 IL-1β, a potent proinflammatory cytokine se-creted by monocytes and macrophages, seems to be in theinterface between inflammation and degeneration because itcolocalizes with amyloid precursor protein on muscle fibersand leads to overexpression of amyloid precursor protein withsubsequent accumulation of β-amyloid, upregulation of in-ducible nitric oxide synthase, production of nitric oxide, andcell death of human myotubes exposed to IL-1β.1,7–9 On thisbasis, we hypothesized that targeting IL-1βmay arrest diseaseprogression. Toward this goal, we first conducted a pilot studywith anakinra, a nonselective IL1-receptor antagonist thatblocks the biologic activity of IL-1, in 5 patients with sIBM.10

Anakinra, approved for rheumatoid arthritis, was ineffective insIBM.10 We have now tested the efficacy of canakinumab, anIgGκ monoclonal antibody, that offers a target-specificblockade of IL-1β, as a more promising approach.

Study design and patientsCanakinumab (Ilaris)Canakinumab binds with high affinity and specificity to hu-man IL-1β, blocking its interaction with IL-1 receptors. In thisproof-of-principle, open-label study, we treated 5 patientswith sIBM with 150 mg subcutaneous canakinumab every 8weeks, following the doses and schedules approved for theother autoinflammatory conditions, for a mean period of 15.8months (range 5–33 months). In 3 patients (3, 4, and 5), theschedule was empirically increased to monthly injectionshoping for a better effect.

PatientsAll patients were ambulatory at study entry and fulfilled thetypical clinical and histologic criteria of sIBM.1,5 Two of 5patients had also participated in our previous negative study

with anakinra, and their available quantitative muscle strengthdata were used to assess whether canakinumab had any effecton disease progression. The patients’ clinical characteristicsare shown in the table.

Muscle strength, expressed in kilograms, was measured witha muscle dynamometer by the same examiner (M.L.K.) every1 or 2 months assessing (1) the mean values of bilateral gripforce (GF) and (2) total muscle strength (TMS) based on thesum score in 6 muscle groups bilaterally (12 groups total):arm abduction, elbow flexion, wrist extension, hip flexion,knee extension, and ankle dorsiflexion. All biopsies wereperformed and read by us (Prof. Dalakas’s laboratory).

Primary research questionClass IV was defined as >15% increased strength after 12months.

Standard protocol approvals, registrations,and patient consentsAll participants signed informed consent before the diagnosticmuscle biopsy was performed and before entering the study.The study protocol was approved by the Institutional ReviewBoard of the Laikon Hospital, Athens University MedicalSchool. The study was registered at EudraCT (number: 2014-004602-15).

Data availabilityAnonymized data will be shared by request from any qualifiedinvestigator.

ResultsTreatment effect

Patient 1A 54-year-old man had progressive muscle weakness and at-rophy for the last 12 years initially treated as polymyositis. Thediagnosis of IBM was confirmed with a muscle biopsy per-formed 2 years before enrollment. Because of rapid pro-gression, he was treated with intravenous immunoglobulin,methotrexate, and mycophenolate without improvement.He received canakinumab for 5 months, from November2014 to April 2015. A 23% loss in TMS and 32.35% in GFwas noted over a 5-month period (figure). He continued toworsen with more generalized weakness including facialmuscles and dysphagia. The accelerated worsening, per-ceived as possibly related to the study drug, prompted earlytreatment discontinuation.

Patient 2A 45-year-old woman had slowly progressive proximal anddistal muscle weakness and typical clinicohistological features

GlossaryGF = grip force; sIBM = sporadic inclusion body myositis; TMS = total muscle strength.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 104: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

of sIBM. She had participated in the anakinra study from 2010to 2011 without benefit but steady progression. She receivedcanakinumab from March 2016 to November 2016. A 37.1%increase in TMS and 13% increase in GF were noted over an8-month treatment (figure). Of interest, when she receivedanakinra, she exhibited a 7.4% increase in TMS and a 30%decrease in GF10 suggesting a more beneficial effect of canaki-numab in disease progression. Despite thesemild gains howeverdemonstrable in our measurements, the patient did not ap-preciate any positive functional changes in her daily activitiesand requested not to continue with additional treatments.

Patient 3An 80-year-old woman had progressive muscle weaknessfor the last 29 years with typical clinicohistologic features ofsIBM. She had received mycophenolate and steroidswithout benefit. At study entry, she was ambulatory withaids. She was treated with canakinumab for 33 months,from September 2015 until June 2018. A 26.7% loss in TMSand 10% loss in GF were noted during the study period(figure). She reported progressive weakness with moredifficulty in her daily activities and felt that the drug wasineffective.

Figure (A and B) Effect of canakinumab in muscle strength measured in each patient as mean bilateral GF (A) and TMS(B) during the mean study period of 15.8 months (range 5–33 months)

No significant or clinically meaningful changes were noted in improving strength or stabilizing disease progression. GF = grip force; TMS = total musclestrength.

Table Main clinical characteristic of patients with sIBM participating in the canakinumab study

Age atsymptomonset

Diseaseduration (y) tillstudyparticipation Initial symptoms

Muscle groups involvedduring the course of thedisease and studyparticipation Dysphagia

Musclebiopsyfindings

Previoustherapies

Patient1

42 12 Lower extremityweakness, followedby upper extremities

Facial and bulbar muscles;upper and lower extremityproximal and distal muscles

Yes Typicalfor IBM

IVIg,methotrexate,andmyocophenolate

Patient2

36 10 Upper and lowerproximal and distalmuscles

Upper and lower proximal anddistal muscles

No Typicalfor IBM

Anakinra

Patient3

51 29 Upper and lowerproximal and distalmuscles

Upper and lower proximal anddistal muscles

No Typicalfor IBM

Mycophenolateand steroids

Patient4

54 14 Upper and lowerproximal and distalmuscles

Upper and lower proximal anddistal muscles

No Typicalfor IBM

Mycophenolate,prednisolone,and anakinra

Patient5

65 5 Upper and lowerproximal and distalmuscles

Upper and lower proximal anddistal muscles

Yes Typicalfor IBM

No previoustreatments

Abbreviations: IBM = inclusion body myositis; IVIg = intravenous immunoglobulin; sIBM = sporadic inclusion body myositis.Clinical characteristics of patients with sIBM participated in the canakinumab study. Two of five patients developed overt dysphagia during the diseasecourse. Patients 2 and 4 had also participated in the previous study with anakinra.10

None of the patients was receiving any other immunotherapy during the study period with canakinumab.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 105: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Patient 4A 68-year-old man had progressive weakness for the last 14years and typical sIBM. Treatment with mycophenolate andprednisolone was ineffective. He received canakinumab for15 months, from April 2014 until July 2015. A 6.5% re-duction in TMS with 1.6% reduction in GF were notedduring the study period, a sign we interpreted as diseasestability (figure). He also reported functional improvementin the use of his hands after the first few months, but thisbenefit did not persist or captured in follow-up visits. Thepatient had participated in the anakinra study, from January2010 until October 2010, and exhibited a 3.5% reduction inTMS and 5% reduction in GF.

Patient 5A 70-year-old man had a 5-year history of progressive weak-ness and typical clinicohistologic features of sIBM. He re-ceived canakinumab for 18 months, from May 2015 toNovember 2016. Three months after therapy initiation, heunderwent cholecystectomy for cholelithiasis. Following theoperation, he reported worsening of his weakness and pro-gressive dysphagia. A reduction in his muscle strength, by30.4% in TMS and 20.8% in GF, was observed during the 18-month treatment period (figure).

DiscussionCanakinumab was well tolerated but had a variable effect inthe 5 treated patients with sIBM. None of the patients im-proved. Two patients showed objective signs of disease sta-bility, which were not however perceived by them as positivelyaffecting their daily activities; 2 others continued to progresswith the same pace as before therapy, even during a 33-monthextended study period; and still another with accelerateddisease progression might have worsened.

The ineffectiveness of the drug may be due to various reasons:(1) the patient’s advanced disease at study entry that mighthave precluded clinically perceptive changes; (2) IL-1βblockade may not be meaningful after the disease is alreadyadvanced, although in vitro has a myotoxic effect andenhances amyloid accumulation; (3) the dose used might beinsufficient to inhibit the inflammatory and neurodegenera-tive effect of IL-1β; or (4) IL-1β might not be one of theprimary players that drives progression and neuro-inflammation in sIBM.

Whether canakinumab will have an effect in disease pro-gression in a subset of patients with sIBM with early diseaseremains uncertain. Such a consideration is entertained be-cause of the relative stability noted in 2 patients; this effectwas, however, minor, and the rationale to conduct a futurecontrol trial in more patients with early disease using such anexpensive drug may not be warranted.

Study fundingNo targeted funding reported.

DisclosureM.L. Kosmidis, D. Pikazis, P. Vlachoyiannopoulos, and A.G.Tzioufas report no disclosures. M.C. Dalakas served on theCIDP steering committee of Novartis and currently serves onDSMB for Baxalta and Octapharma; received travel fundingand/or speaker honoraria from Baxter, CSL, and Merck/Serono; serves on the editorial board of Neurology®, ActaMyologica, Acta Neurologica Scandinavica, and Annals of Neu-rology and as associate editor of Therapeutic Advances inNeurology; consulted for Therapath, Novartis, Baxter, Octa-pharma, CSL, Pfizer AG, Dysimmune Diseases Foundation,and Elsevier; and received institutional support to ThomasJefferson University and/or University of Athens MedicalSchool from Merck Serono, Genzyme, Novartis, Guillain-Barre/CIDP Foundation, Dysimmune Diseases Foundation,Biogen, and Newfactor. Go to Neurology.org/NN for fulldisclosures.

Publication historyReceived by Neurology: Neuroimmunology & NeuroinflammationFebruary 12, 2019. Accepted in final form April 24, 2019.

References1. Dalakas MC. Inflammatory muscle diseases. N Engl J Med 2015;372:1734–1747.2. Cox FM, Titulaer MJ, Sont JK, Wintzen AR, Verschuuren JJ, Badrising UA. A 12-year

follow-up in sporadic inclusion body myositis: an end stage with major disabilities.Brain 2011;134:3167–3175.

3. Peng A, Koffman BM, Malley JD, Dalakas MC. Disease progression in sporadicinclusion body myositis: observations in 78 patients. Neurology 2000;55:296–298.

4. Dalakas MC. Polymyositis, dermatomyositis and inclusion-body myositis. N Engl JMed 1991;325:1487–1498.

5. Dalakas MC. Sporadic inclusion body myositis—diagnosis, pathogenesis and thera-peutic strategies. Nat Clin Pract Neurol 2006;2:437–447.

6. Keller CW, Schmidt J, Lunemann JD. Immune and myodegenerative patho-mechanisms in inclusion body myositis. Ann Clin Transl Neurol 2017;4:422–445.

7. Dalakas M. Molecular immunology and genetics of inflammatory muscle diseases.Arch Neurol 1998;55:1509–1512.

Appendix Authors

Name Location Role Contributions

Kosmidis Michalis,MD, PhD

AthensMedicalSchool

Author Analyzed the data,drafted the manuscriptfor intellectual content,and major role in theacquisition of data

Pikazis Dimitrios,MD

AthensMedicalSchool

Author Major role in theacquisition of data andrevised the manuscriptfor intellectual content

VlachoyiannopoulosPanayiotis, MD

AthensMedicalSchool

Author Revised the manuscriptfor intellectual content

Tzioufas Athanasios,MD, PhD

AthensMedicalSchool

Author Revised the manuscriptfor intellectual content

Dalakas Marinos,MD, FAAN

AthensMedicalSchool

Author Designed andconceptualized thestudy, interpreted thedata, and revisedthe manuscript forintellectual content

4 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 106: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

8. Schmidt J, Barthel K, Wrede A, Salajegheh M, Bahr M, Dalakas MC. Interrelation ofinflammation and APP in sIBM: IL-1 beta induces accumulation of beta-amyloid inskeletal muscle. Brain 2008;131(pt 5):1228–1240.

9. Schmidt J, Barthel K, Zschuntzsch J, et al. Nitric oxide stress in sporadic inclusionbody myositis muscle fibres: inhibition of inducible nitric oxide synthase prevents

interleukin-1beta-induced accumulation of beta-amyloid and cell death. Brain 2012;135(pt 4):1102–1114.

10. Kosmidis ML, Alexopoulos H, Tzioufas AG, Dalakas MC. The effect of anakinra,an IL1 receptor antagonist, in patients with sporadic inclusion body myositis(sIBM): a small pilot study. J Neurol Sci 2013;334:123–125.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 5

Page 107: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

CLINICAL/SCIENTIFIC NOTES OPEN ACCESS

IgA autoantibodies against native myelin basicprotein in a patient with MSHeike Schumacher, Nina K. Wenke, PhD, Jakob Kreye, MD, Markus Holtje, PhD, Katrin Marcus, PhD,

Caroline May, PhD,* and Harald Pruss, MD*

Neurol Neuroimmunol Neuroinflamm 2019;6:e569. doi:10.1212/NXI.0000000000000569

Correspondence

Dr. Pruss

[email protected]

Myelin basic protein (MBP) is one of the most abundant proteins in the human brain. Activeimmunization with MBP induces experimental autoimmune encephalomyelitis, and anti-MBPantibodies have been repeatedly described in MS.1 However, its role in MS pathogenesis orprediction of disease progression is still unclear.2,3 Previous studies utilized enzyme-linked im-munosorbent assay or immunoblot assays with linear epitopes of MBP, thus potentially over-looking autoantibodies that bind to MBP’s natural conformation. These initial studies alsoincluded antibodies against anothermyelin protein, myelin oligodendrocyte glycoprotein (MOG).As happened for MBP, conflicting results stimulated the discussion of whether MOG antibodiescontribute to MS pathogenesis.2,3 More recent work demonstrated that there are presumablypathogenic MOG antibodies defining the new entity of MOG antibody-associated disease;4

however, they bind to conformational MOG only.

Here we report on a patient with MS with immunotherapy-responsive severe cognitive im-pairment having high-level immunoglobulin A (IgA) autoantibodies against conformationalMBP, suggesting the possibility of myelin-directed humoral autoimmunity beyond MOG.

Case reportA 54-year-old woman with a 20-year history of relapsing–remitting MS (Expanded DisabilityStatus Scale 3.5) was admitted for a suspected relapse with subacute-onset rapidly progressingcognitive decline, presenting with dementia and echolalia. Apart from unsteady gait, doublevision, and lack of coordination, cerebellar and motor signs were relatively spared, and the MRIshowed new lesions (figure, A and B). Previous treatments included mitoxantrone (19 cycles,cumulative dose 137 mg/m2) and beta-1a interferon (3 years of 44 μg 3 times per week). Giventhe unusual predominance of cognitive symptoms with rapid deterioration from 18 to 14/30points in Mini-Mental State Examination, secondary autoimmune encephalitis was considered.Indirect immunofluorescence revealed high titers of brain-reactive IgA antibodies (serum 1:3,200, CSF 1:32, antibody index 6.1 indicating intrathecal synthesis; immunoglobulin M/Gnegative) labeling axonal fibers throughout the unfixed brain, particularly in cerebellum (figure,C), corpus callosum, and hippocampus. The fine parallel fiber staining suggested binding tomyelin epitopes (figure, C, insert). MOG antibodies were excluded (Prof. Hoftberger, Vienna,Austria). Immunotherapy, including plasma exchange (10 sessions every other day) and rit-uximab (1,000mg every 6months for 2 years), resulted in the disappearance ofMBP antibodiesafter 6 months and improvement of cognitive symptoms (Mini-Mental State Examination 16/30), which remained stable for 3 years until the last follow-up, antibodies remained negative.

To identify the antigen, immunoprecipitation and mass spectrometry were performed. Onehundred micrograms of IgA purified from the plasma exchange eluate were incubated overnight

*These authors contributed equally to this work.

From the German Center for Neurodegenerative Diseases (DZNE) Berlin (H.S., N.K.W., J.K., H.P.); Institute of Integrative Neuroanatomy (M.H.), Charite—Universitatsmedizin Berlin;Medizinisches Proteom-Center (K.M., C.M.), Ruhr-University Bochum; Department of Neurology and Experimental Neurology (H.P.), Charite—Universitatsmedizin Berlin; and Centerfor Autoimmune Encephalitis and Paraneoplastic Neurological Syndromes (H.P.), Berlin, Germany.

Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by the authors.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 108: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

with rat brain lysate and samples run on sodium dodecyl sulfate(SDS) gels. Bands were analyzed with mass spectrometry,5 anddata were analyzed as described,6 matching MBP only. Doubleimmunolabeling showed exact co-localization of patient anti-body with a commercial anti-MBP antibody (figure, D–F). Incontrast to the commercial antibody, the patient’s IgA did notbind to rat brain lysate in denaturing Western blots (notshown), suggesting that they recognize the natural epitopeconformation. Direct proof for the target antigen was obtainedusing MBP knockout mice in which the antibody binding wascompletely lost (figure, G–J).

DiscussionWe report the case of a patient with MS with rapidly pro-gressing cognitive decline having high-level autoantibodies

against conformational MBP. Previous studies using denaturedepitopes in enzyme-linked immunosorbent assay and Westernblots could not establish a clear link between MBP antibodiesand disease,2,3 potentially because they overlooked specificbinding to the conformational epitope. This first report ofMBPautoantibodies against native MBP revives the discussion ofwhether such antibodies might be related to a subgroup of MSpatients, convey pathology, or serve as a biomarker for pro-gression or cognitive symptoms.

Similar to MBP, the pathogenic role of MOG antibodies hasbeen debated intensely. Only more recent studies focusing onantibody interactions with native MOG could convincinglydemonstrate their pathogenic potential.7 For example, suchMOG antibodies were present in a subgroup of patients withsevere MS.

Figure Myelin binding of high-level MBP IgA antibodies from a patient with MS

(A) Cerebral MRI shows atrophy, widespreadpostinflammatory changes and (B) new contrast-enhancing lesions (arrow). (C) Using 20μmunfixedrat brain sections, patient IgA (4.25 mg/mL, di-lution 1:10) labels fine axonal fibers (green, goatanti-human IgA, Dianova, Hamburg, Germany, di-lution 1:200) throughout the brain, in particular inthe cerebellar cortex (colabeling with a GABAAreceptor antibody [red; Santa Cruz Biotechnology,Dallas, TX, USA, dilution 1:200] for better ana-tomical visualization of the cerebellar cortex). (C,inset) Highermagnification shows parallel stainingof fibers, indicative of myelin antigens. (D–F)Double-labeling of patient IgA (green) with a com-mercial anti-MBP antibody [red, Santa Cruz Bio-technology, Dallas, TX, USA, dilution 1:200]demonstrates complete overlap in rat cerebellarcortex (merged in [F]). The characteristic immu-nofluorescencewith strong binding to axonal fibertracts on a 20 μmparaformaldehyde-fixed mousebrain section (G, red) was completely absent inshiverer MBP knockout (mbpshi) littermate mice(H), exemplarily shown at higher magnification inthewhitematter of the cerebellum (arrowheads inG.b and H.b; double-labeling with DAPI for cellnuclei in blue) or the anterior commissure (I, J;double-labelingwith theneuronalmarkerNeuN ingreen). Bars represent 50 μm in C–F, 1 mm in G–Hand 50 μm in I, J. CA = anterior commissure; CB =cerebellum; CC = corpus callosum;CTX = cortex; FX= fornix; GCL = granule cell layer; HPF = hippo-campal formation;MB =midbrain; ML =molecularlayer; PCL = Purkinje cell layer; WM=whitematter;and wt = wild-type.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 109: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

The clinical improvement with immunotherapy in our pa-tient paralleled the disappearance of antibody titers, sug-gesting that the antibodies may have contributed to thedisease. IgA antibody transfer into animals should bean important future experimental step to confirmpathogenicity.

It is unclear at present whether autoantibodies against na-tive MBP are also detected in a subgroup of patients withMS. We did not find a similar immunofluorescence patternusing the serum and CSF of 352 consecutive patients withsuspected encephalitis (including 46 patients with MS)and serum of 82 healthy controls, suggesting that—similarto MOG antibodies—the specific myelin staining ob-served in the present patient is rare. Prospective studiesusing established cohorts with MS and clinically isolatedsyndrome (CIS) should therefore be screened system-atically to determine the frequency of native MBP-targeting autoantibodies, the association with clinicalphenotypes, CIS conversion to MS, relapses, and diseaseprogression.

Author contributionsH. Schumacher: major role in the acquisition of data,analysis and interpretation of data, and drafting the man-uscript. N.K. Wenke: acquisition of data. J. Kreye: acqui-sition of data. M. Holtje: acquisition of data. K. Marcus:acquisition of data. C. May: acquisition, analysis and in-terpretation of data, and critical revision of manuscript.H. Pruss: acquisition, analysis and interpretation of data,and critical revision of manuscript.

AcknowledgmentThe authors are grateful to Professor Brian Popko, De-partment of Neurology, University of Chicago, for providingMBP knockout mouse brains.

Study fundingThis work was supported by the HUPO Brain ProteomeProject (HBPP) and PURE, a project of North Rhine-Westfalia, a federal German state.

DisclosureThe authors report no disclosures. Disclosures available:Neurology.org/NN.

Publication historyReceived byNeurology: Neuroimmunology &NeuroinflammationMay 19,2018. Accepted in final form March 14, 2019.

References1. Olsson T, Baig S, Hojeberg B, Link H. Antimyelin basic protein and antimyelin antibody-

producing cells in multiple sclerosis. Ann Neurol 1990;27:132–136.2. Berger T, Rubner P, Schautzer F, et al. Antimyelin antibodies as a predictor of

clinically definite multiple sclerosis after a first demyelinating event. N Engl J Med2003;349:139–145.

3. Kuhle J, Pohl C, Mehling M, et al. Lack of association between antimyelin antibodiesand progression to multiple sclerosis. N Engl J Med 2007;356:371–378.

4. Probstel AK, Dornmair K, Bittner R, et al. Antibodies to MOG are transientin childhood acute disseminated encephalomyelitis. Neurology 2011;77:580–588.

5. Plum S, Helling S, Theiss C, et al. Combined enrichment of neuromelanin granulesand synaptosomes from human substantia nigra pars compacta tissue for proteomicanalysis. J Proteomics 2013;94:202–206.

6. Molina M, Steinbach S, Park YM, et al. Enrichment of single neurons and definedbrain regions from human brain tissue samples for subsequent proteome analysis.J Neural Transm (Vienna) 2015;122:993–1005.

7. Reindl M, Rostasy K. MOG antibody-associated diseases. Neurol NeuroimmunolNeuroinflamm 2015;2:e60. doi: 10.1212/NXI.0000000000000060

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 110: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

CLINICAL/SCIENTIFIC NOTES OPEN ACCESS

Immunotherapy improves sleep and cognitiveimpairment in anti-IgLON5 encephalopathyValerio Brunetti, MD, Giacomo Della Marca, MD, PhD, Gregorio Spagni, MD, and Raffaele Iorio, MD, PhD

Neurol Neuroimmunol Neuroinflamm 2019;6:e577. doi:10.1212/NXI.0000000000000577

Correspondence

Dr. Iorio

[email protected]

Autoimmune encephalopathy associated with immunoglobulin G (IgG) autoantibodies bindingto IgLON5 is characterized by a sleep disorder that includes sleep-disordered breathing andparasomnia, accompanied by gait disturbance, bulbar symptoms, dysautonomia, and cognitivedecline.1 The effect of immunotherapy in patients with anti-IgLON5 disease remains unclear.

We herein report the polysomnographic findings and the clinical and immunologic charac-teristics of a patient with IgLON5-IgG before and after 1 year of immunotherapy.

Case reportA 69-year-old man was admitted to our institution for excessive daytime sleepiness associated withcomplex stereotypedmovements. Three years before, the patient was admitted to the intensive careunit for respiratory failure due to vocal cords palsy in adduction and underwent tracheotomy;a pacemaker was placed after 2 episodes of severe bradycardia. Five years before, he experiencederectile dysfunction and constipation; then, his wife reported complex movements during sleep.

Video-polysomnography revealed finalistic movements mimicking daily activities (oneiricstupor), such as eating, at sleep onset and in the second half of the night (video 1). Electro-encephalographic activity during those periods was characterized by a double peak of frequencyat 5 and 10 Hz; moreover, K-complex and sleep spindles were absent. This undifferentiatedNREM sleep was followed by a poorly structured N2 sleep, with rare K-complex and sleepspindles and then by normal N2 sleep and few epochs of N3 sleep. REM sleep was undetectable(figure, A). Neurologic examination revealed mild gait ataxia. The patient had an open tra-cheostomy, and he was on assisted ventilation during the night. He had cognitive impairmentand needed help for most activities of daily living (ADL). Neuropsychological examinationshowed a Mini Mental Status Examination (MMSE) score of 24/30, a prevalent impairment ofvisuospatial functions: Multiple Features Target Cancellation (MFTC) accuracy was 0.89,number of false recognition was 4, and time of execution was 240 seconds; the accuracy of theRey Word Recognition Test (RWRT) was 0.78, and forward spatial span was 3. Anti-IgLON5encephalopathy was suspected, and IgLON5-IgG was tested by indirect immunofluorescenceon mouse brain and cell-based assays, as previously described.2 IgLON5-IgG was detected in bothpatient’s serum (titer: 1:5,000) and CSF (1:100). The patient’s human leukocyte antigen (HLA)typing revealed HLA-DQB1*05:01 and HLA-DRB1*10:01. Brain MRI and 99mTc-hexamethyl-propyleneamine oxime (HMPAO)-single photon emission computed tomography were normal.The patient was treated with monthly IV immunoglobulins (0.4 g/kg/d for 5 days), prednisone(0.8 mg/kg), and azathioprine (2 mg/kg). Subsequently, he progressively improved, showinga reduction of daytime sleepiness and of motor activation during the sleep, as reported by thespouse. After 1 year of immunotherapy, home-based unattended PSG showed a partial im-provement of sleep architecture: REM sleep was present and characterized by physiological muscleatonia. Undifferentiated NREM sleep was present exclusively in the second half of the night, butwith less sustained muscular activation (figure, B). Despite the improvement of the sleep

From the Institute of Neurology (V.B., G.D.M., G.S., R.I.), Universita Cattolica del Sacro Cuore; and Fondazione Policlinico Universitario “A. Gemelli” IRCCS (G.D.M., R.I.), Roma, Italy.

Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by the authors.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

MORE ONLINE

Video

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 111: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

organization and the reappearing of the REM sleep, we did notobserve a significant reduction in the percentage of un-differentiated NREM sleep (e-table, links.lww.com/NXI/A117). The neurologic examination showed a resolution of gaitataxia. An improvement of cognitive functions was observed: theMMSE score was 30/30; MFTC accuracy was 0.96, number offalse recognition was 5, and time of execution was 91 seconds;the accuracy of RWRT was 0.85, and forward spatial span was 6.These findings were consistent with an improvement in thedomains of visual spatial attention, working memory, and epi-sodic memory. The patient regained independence in the ADL.IgLON5 antibodies were still detected in both serum (1:1,000)and CSF (1:10), but with a lower titer.

DiscussionThe effect of 1 year of immunotherapy in a patient withanti-IgLON5 encephalopathy is herein documented by pol-ysomnography and neuropsychological examination. Anti-IgLON5 disease involves prevalently the brain stem andhypothalamus, 2 structures that are crucial for wake-NREM-REM sleep alternation.3 In fact, wake is promoted by theascending reticular activating system located in the brainstem

and basal forebrain. Conversely, median and ventrolateralpreoptic nuclei of the hypothalamus are considered the gen-erators of slow-wave sleep; REM-generator neurons are sitedin pedunculopontine and laterodorsal tegmental nuclei.4 Thedisease is characterized by autoantibodies binding IgLON5,a neuronal cell adhesion molecule, and by the neuropatho-logic evidence of Tau aggregates, suggesting a combined au-toimmune and neurodegenerative pathogenesis. Evidence iscompelling that IgLON5-IgG has pathogenic potential.5 Theeffect of immunotherapy in anti-IgLON5 encephalopathy isstill debated, with some patients being poorly responsive toimmunotherapy1,6 and others showing a good response.7

Our observation suggests that sustained immunotherapy canimprove sleep disorder and cognitive impairment in anti-IgLON5 disease. The improvement of neurologic symptomswas associated with a decrease in IgLON5-IgG titer. Furtherstudies are needed to develop a standardized therapy regimenfor this severe disease.

Author contributionsV. Brunetti: acquisition, analysis, and interpretation of dataand drafting of the manuscript. G.D. Marca and G. Spagni:

Figure Hypnogram and density spectral array images

(A) Hypnogram andDSA before the treatment. Hypnogram: undifferentiated NREM sleep is prevalent, particularly at the sleep onset and in the second half ofthe night (black arrows); REM sleep is not detectable. DSA showing the power spectrum of electroencephalographic frequencies (0–64 Hz) in bipolar C3-O1derivation: undifferentiated NREM sleep is characterized by a continuous activity with a double peak of frequency at 5 and 10Hz. (B) HypnogramandDSA after 1year of immunotherapy. Hypnogram: an NREM-REM cycle is present. Undifferentiated NREM sleep is still present in the second half of the night (black arrow). DSA:undifferentiated NREM sleep is still characterized by a continuous activity with a double peak of frequency at 5 and 10 Hz. DSA = density spectral array.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 112: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

data acquisition and analysis. R. Iorio: study concept and design;acquisition, analysis, and interpretation of data; and critical re-vision of the manuscript for important intellectual content.

Study fundingNo targeted funding reported.

DisclosureThe authors report no relevant disclosures. Go to Neurology.org/NN for full disclosures.

Publication historyReceived byNeurology: Neuroimmunology & Neuroinflammation January13, 2019. Accepted in final form March 21, 2019.

References1. Sabater L, Gaig C, Gelpi E, et al. A novel non-rapid-eye movement and rapid-eye-

movement parasomnia with sleep breathing disorder associated with antibodies toIgLON5: a case series, characterisation of the antigen, and post-mortem study. LancetNeurol 2014;13:575–586.

2. Iorio R, Damato V, Evoli A, et al. Clinical and immunological characteristics of thespectrum of GFAP autoimmunity: a case series of 22 patients. J Neurol NeurosurgPsychiatry 2018;89:138–146.

3. Gelpi E, Hoftberger R, Graus F, et al. Neuropathological criteria of anti-IgLON5-related tauopathy. Acta Neuropathol 2016;132:531–543.

4. Scammell TE, Arrigoni E, Lipton JO. Neural circuitry of wakefulness and sleep.Neuron 2017;93:747–765.

5. Sabater L, Planaguma J, Dalmau J, Graus F. Cellular investigations with humanantibodies associated with the anti-IgLON5 syndrome. J Neuroinflammation 2016;13:226.

6. Gaig C, Graus F, Compta Y, et al. Clinical manifestations of the anti-IgLON5 disease.Neurology 2017;88:1736–1743.

7. Honorat JA, Komorowski L, Josephs KA, et al. IgLON5 antibody: neurologicalaccompaniments and outcomes in 20 patients. Neurol Neuroimmunol Neuroinflamm2017;4:e385. doi: 10.1212/NXI.0000000000000385.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 113: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

CLINICAL/SCIENTIFIC NOTES OPEN ACCESS

Cerebellar ataxia as a presenting symptom ina patient with anti-NMDA receptor encephalitisMichiel H.F. Poorthuis, MD, Josephus L.M. van Rooij, MD, Anna H. Koch, MD, Annelies E.M. Verdonkschot, MD,

Machteld M. Leembruggen, MD, PhD, and Maarten J. Titulaer, MD, PhD

Neurol Neuroimmunol Neuroinflamm 2019;6:e579. doi:10.1212/NXI.0000000000000579

Correspondence

Dr. Titulaer

[email protected]

Anti-NMDA receptor (anti-NMDAR) encephalitis is a treatment-responsive autoimmune en-cephalitis, first described in 2007.1 Ovarian teratomas are found in one-third of the patients.2 Theclinical features of this disorder vary between patients and age groups and usually includeabnormal (psychiatric) behavior or cognitive dysfunction, speech dysfunction (pressured speech,verbal reduction, and mutism), seizures, movement disorders, dyskinesias, or rigidity/abnormalpostures, decreased level of consciousness, autonomic dysfunction, or central hypoventilation.2

Cerebellar ataxia has been described as a symptom during the first months of the disease,especially in young children, in combination with other symptoms.2,3 It is extremely rare as theinitial symptom, especially in adults. We report a case of a female adult with anti-NMDARencephalitis presenting with cerebellar ataxia associated with recurrent mature ovarian teratomas.

Case reportA 32-year-old woman, born in South Korea and adopted at age 4 months, presented with vertigo,nausea, and vomiting for 4 days. Her medical history consisted of bilateral cystectomy revealingmature teratomas, discovered by ultrasound examination after a missed abortion at age 26 years.During cesarean sections afterward (ages 29 and 31 years), no macroscopic abnormalities wereseen. Furthermore, she had had depressive symptoms, treated with venlafaxine for years.

Neurologic examination showed a horizontal gaze-evoked nystagmus to the right without otherneurologic signs or symptoms. Laboratory investigations on admission were normal, and brainCT showed no abnormalities.

Initially, she improved after treatment with antiemetic drugs, but after 3 days, she deterioratedquickly, also complaining of headache. Neurologic examination showed nystagmus in alldirections and dysarthric speech (cerebellar) that further worsened to impaired speech re-stricted to one-word sentences. She showed bilateral dysmetria of the lower and especially theupper limbs, truncal ataxia, and inability to stand and walk. Psychiatric evaluation showed rapidprogression of depressive symptoms with suicidal ideation and labile affect.

BrainMRI andMRVwere normal. CSF analysis and extensive laboratory investigations showedpleocytosis (table). Anti-NMDAR antibodies were negative in serum, but positive in CSF,4

confirming the diagnosis of definite anti-NMDAR encephalitis.3

The patient was treated with IV methylprednisolone 1,000 mg (day 13, 5 days) and IV immu-noglobulins 0.4 g/kg (day 16, 5 days). Thorax/abdomenCT and transvaginal ultrasound revealed2 lesions in the pelvic area with fat tissue and calcifications, suspect for teratomas. Bilaterallaparotomic ovariectomy was performed (day 19). Pathologic examination showed mature cystic

From the Department of Neurology (M.H.F.P., J.L.M.R., M.M.L.) and Department of Gynaecology (A.H.K., A.E.M.V.), Tergooi, Blaricum; and Department of Neurology (M.J.T.), ErasmusMC, Rotterdam, The Netherlands.

Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

Informed consent: The patient gave informed consent.

The Article Processing Charge was funded by the authors.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 114: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

teratomas, without immature components, containing nervoustissue. Hormone replacement therapy was started.

Her neurologic condition improved within a week, but thedepressive mood remained. Recovery was hampered by uro-sepsis, treated with cefuroxime/amoxicillin. She was treatedwith a second course of methylprednisolone 4 weeks after theinitial treatment and immunoglobulins at 8 weeks forremaining speech impairments and severe depression. Thisresulted in further improvement of both. After 6 weeks, thepatient was transferred to a rehabilitation unit.

After 6 months, the patient returned home. She was able toperform activities of daily living independently, but neededwalking aids outside due to residual ataxia and had notreturned to work (yet).

DiscussionThis case with cerebellar ataxia as an initial symptom high-lights an unusual presentation of anti-NMDAR encephalitis. Ifcerebellar ataxia is present in patients with anti-NMDAR

encephalitis, it is almost exclusively found in (young) chil-dren, and most frequently, it appears later in the disease incombination with other symptoms.2 Different brainstem-cerebellar symptoms have been described, such asopsoclonus-myoclonus syndrome, ocular movement abnor-malities, and low cranial nerve involvement in patients withovarian teratomas, but these symptoms have more frequentlybeen described in whom no NMDAR antibodies could beidentified.5 Although 2 simultaneously occurring paraneo-plastic neurologic syndromes, due to an ovarian teratoma,cannot be fully excluded, this is considered unlikely. Thedevelopment of multiple symptoms quickly into diseasescompatible with anti-NMDAR encephalitis (psychiatricsymptoms and mutism), the confirmation of NMDAR anti-bodies by different tests,4 and the identification of an ovarianteratoma are suitable with a diagnosis of “definite anti-NMDAR encephalitis.”3

Although it is known that anti-NMDAR IgG antibodies bindto granular cells in the cerebellum (but not to Purkinje cells),6

it is unknown why only approximately 5% of patients showcerebellar complaints. MRI abnormalities of the cerebellumhave been described in 6% of patients.7 A small study showed

Table Overview of investigations

CSF analysis

White blood cells 30 × 106/L (100% mononuclear cells)[ref: <5 × 106/L]

Protein 31 [ref: 24–49] mg/dL

Glucose 3.6 [serum: 6.4] mmol/L

Erythrocytes <5,000 × 106/L

Oligoclonal bands Present

IgG index 0.59

IgG quotient 3.1 [ref: <2.8]

PCR in liquor

HSV-1, HSV-2, VZV, EBV, enteroviruses, parechoviruses, and Borrelia Negative

Extensive laboratory investigations, among these:

TSH, anti-TPO, and vitamin E Normal

Ceruloplasmin, anti-tissue transglutaminase, anti-endomysium, and anti-GQ1b Negative

Antibodies in serum and CSF

Anti-NMDAR in serum Negative

Anti-NMDAR in CSF (cell-based assay and immunohistochemistry) Positive

Anti-Hu, anti-Yo, anti-Ri, anti-Tr, anti-amphiphysine, anti-CV2 (CRMP5), anti-Ma1, anti-Ma2, anti-GAD65,anti-LGII, anti-Caspr2, anti-GABABR, anti-GABAAR, anti-AMPAR, and anti-DPPX

Negative

Urine analysis

Dipstick test and pregnancy test Negative

Abbreviations: anti-NMDAR = anti-NMDA receptor; anti-TPO = anti–thyroid peroxidase; EBV = Epstein-Barr virus; HSV = herpes simplex virus; TSH = thyroid-stimulating hormone; VZV = varicella zoster virus.Abnormal values are shown in italic font.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 115: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

progressive cerebellar atrophy by follow-up MRI in 2 of 15patients with anti-NMDAR encephalitis.6

In conclusion, cerebellar ataxia is unusual in adult patients andan extremely rare presenting symptom of anti-NMDAR en-cephalitis. This case shows that anti-NMDAR encephalitisshould be considered in the differential diagnosis of cerebellarataxia, especially in patients with previous teratomas and thosedeveloping other symptoms shortly afterward.

AcknowledgmentThe authors thank J.J. Oudejans, pathologist, Tergooi, Blaricum,The Netherlands, and C.E. de Boer, physiatrist, Tergooi,Blaricum, The Netherlands, for their advice on this case.

Study fundingNo targeted funding reported.

DisclosureM.J. Titulaer has filed a patent for methods for typing neuro-logical disorders and cancer, and devices for use therein, and hasreceived research funds for serving on a scientific advisory boardof MedImmune LLC, for consultation at Guidepoint GlobalLLC, for teaching courses by Novartis, and an unrestricted re-search grant from Euroimmun AG. M.J. Titulaer has receivedfunding from the Netherlands Organization for Scientific Re-search (NWO, Veni incentive), from the Dutch EpilepsyFoundation (NEF, project 14-19), and from ZonMw (Memo-rabel program). The other authors report no conflicts of in-terest. Go to Neurology.org/NN for full disclosures.

Publication historyReceived by Neurology: Neuroimmunology & NeuroinflammationJanuary 3, 2019. Accepted in final form April 21, 2019.

References1. Dalmau J, Tuzun E, Wu HY, et al. Paraneoplastic anti-N-methyl-D-aspartate receptor

encephalitis associated with ovarian teratoma. Ann Neurol 2007;61:25–36.2. Titulaer MJ, McCracken L, Gabilondo I, et al. Treatment and prognostic factors for

long-term outcome in patients with anti-NMDA receptor encephalitis: an observa-tional cohort study. Lancet Neurol 2013;12:157–165.

3. Graus F, Titulaer MJ, Balu R, et al. A clinical approach to diagnosis of autoimmuneencephalitis. Lancet Neurol 2016;15:391–404.

4. Gresa-Arribas N, Titulaer MJ, Torrents A, et al. Antibody titres at diagnosis and duringfollow-up of anti-NMDA receptor encephalitis: a retrospective study. Lancet Neurol2014;13:167–177.

5. Armangue T, Titulaer MJ, Sabater L, et al. A novel treatment-responsive encephalitiswith frequent opsoclonus and teratoma. Ann Neurol 2014;75:435–441.

6. Iizuka T, Kaneko J, Tominaga N, et al. Association of progressive cerebellar atrophywith long-term outcome in patients with anti-N-Methyl-d-Aspartate receptor en-cephalitis. JAMA Neurol 2016;73:706–713.

7. Dalmau J, Gleichman AJ, Hughes EG, et al. Anti-NMDA-receptor encephalitis: caseseries and analysis of the effects of antibodies. Lancet Neurol 2008;7:1091–1098.

Appendix Authors

Name Location Role Contribution

Michiel H.F.Poorthuis, MD

Department ofNeurology,Tergooi, Blaricum,The Netherlands

Author Conception anddesign, acquisition ofdata or analysis andinterpretation ofdata, wrote the firstdraft of themanuscript, andapproved the finalversion forpublication

Josephus L.M.van Rooij, MD

Department ofNeurology,Tergooi, Blaricum,The Netherlands

Author Conception anddesign, acquisition ofdata or analysis andinterpretation ofdata, wrote the firstdraft of themanuscript, andapproved the finalversion forpublication

Appendix (continued)

Name Location Role Contribution

Anna H. Koch,MD

Department ofGynaecology,Tergooi, Blaricum,The Netherlands.

Author Conception anddesign, acquisition ofdata or analysis andinterpretation ofdata, wrote the firstdraft of themanuscript, andapproved the finalversion forpublication

Annelies E.M.Verdonkschot,MD

Department ofGynaecology,Tergooi, Blaricum,The Netherlands.

Author Conception anddesign, acquisition ofdata or analysis andinterpretation ofdata, revised themanuscript criticallyfor importantintellectual content,and approved thefinal version forpublication

Machteld M.Leembruggen,MD, PhD

Department ofNeurology,Tergooi, Blaricum,The Netherlands

Author Conception anddesign, acquisition ofdata or analysis andinterpretation ofdata, revised themanuscript criticallyfor importantintellectual content,and approved thefinal version forpublication

Maarten J.Titulaer, MD,PhD

Department ofNeurology,Erasmus MC,Rotterdam, TheNetherlands

Author Conception anddesign, acquisition ofdata or analysis andinterpretation ofdata, revised themanuscript criticallyfor importantintellectual content,and approved thefinal version forpublication

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 116: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

VIEWS & REVIEWS OPEN ACCESS

Strategies for treatment of childhood primaryangiitis of the central nervous systemJocelyne Beelen, BMSc, Susanne M. Benseler, MD, PhD, Anastasia Dropol, HBSc, Brianna Ghali, BSc, and

Marinka Twilt, MD, PhD

Neurol Neuroimmunol Neuroinflamm 2019;6:e567. doi:10.1212/NXI.0000000000000567

Correspondence

Dr. Twilt

marinka.twilt@

albertahealthservices.ca

AbstractObjectiveChildhood primary angiitis of the CNS (cPACNS) is a devastating neurologic disease. Nostandardized treatment protocols exist, and evidence is limited to open-label cohort studies andcase reports. The aim of this review is to summarize the literature and provide informedtreatment recommendations.

MethodsA scoping review of cPACNS literature from January 2000 to December 2018 was conductedusing Ovid, MEDLINE, PubMed, Embase, Cochrane Database of Systematic Reviews,Cochrane Central Register of Controlled Trials, ClinicalTrials.gov, Vasculitis Foundation,European Vasculitis Society, CanVasc, Google Scholar, and Web of Science. Potentially rele-vant articles were selected for full-text review using the STROBE checklist if they met thefollowing inclusion criteria: (1) reported treatment, (2) addressed pediatrics, (3) focused onthe disease of interest, (4) included ≥5 patients, (5) original research, and (6) full-lengtharticles. Reviews, expert opinions, editorials, case reports with <5 patients, articles lackingtreatment information, or non-English articles were excluded. A standardized assessment toolmeasured study quality. Treatment and outcomes were summarized.

ResultsOf 2,597 articles screened, 7 studies were deemed high quality. No trials were available so nometa-analysis was possible. Overall, treatment strategies recommended are induction withacute antithrombotic therapy subsequently followed by high-dose oral prednisone taper over3–12 months and long-term platelet therapy. In angiography-positive progressive–cPACNSand angiography-negative–cPACNS, we also recommend 6 months of IV cyclophosphamidetherapy, with trimethoprim/sulfamethoxazole as part of induction, and maintenance therapywith mycophenolate mofetil/mycophenolic acid.

ConclusionNo grade-A evidence exists; however, this review provides recommendations for treatment ofcPACNS.

From the Cumming School of Medicine (J.B., S.M.B., A.D.), University of Calgary, Alberta, Canada; Section of Rheumatology (S.M.B., M.T.), Department of Pediatrics, Alberta Children’sHospital, Alberta Children’s Hospital Research Institute, Calgary, Alberta, Canada; and University of Calgary (B.G.), Alberta, Canada.

Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by the authors.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 117: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Childhood primary angiitis of the CNS (cPACNS) is an in-creasingly recognized inflammatory brain disease. Previouslyhealthy children present with severe neurologic deficits that,left untreated, could lead to devastating neurologic insult andeven death. Early recognition and intervention with targetedtherapy has led to better survival. The term PACNS was firstcoined by Calabrese for adults in 19881 and has been adjustedfor the pediatric population: patients <18 years of age pre-senting with a newly acquired focal or diffuse neurologic orpsychiatric deficit without an underlying systemic disorder andevidence of vasculitis on angiography and/or histopathology.2

Classification of cPACNS is based on vessel size, with 3 subtypesrecognized: angiography-positive nonprogressive (APNP) andangiography-positive progressive (APP) disease, affecting thelarge/medium-sized vessels, and angiography-negative (AN)disease, affecting small cerebral vessels.2,3 Children with APNP-cPACNS typically present with a monophasic event consisting offocal neurologic deficits and evidence of ischemic stroke on MRIwith corroborating angiography.2,4,5 Patients with APP-cPACNSpresent with both focal and diffuse neurologic deficits4,5 andprogressive vessel narrowing on angiography beyond 3months ofdisease.2 In contrast, patients with AN-cPACNS exhibit normalangiography because of the difficulty in observing small vessels ondigital subtraction angiography (DSA) or magnetic resonanceangiography (MRA).6,7 However, MRI is frequently abnormal,characterized by multifocal inflammatory lesions not specific toany vessel distribution or territory.8,9 Thus, elective brain biopsy ismandated for diagnostic confirmation of AN-cPACNS.4,7,10

Children typically present with both focal and diffuse neurologicand psychiatric deficits together with seizures.3,4,7,9 The overallincidence of cPACNS remains unknown.

Similar to many other pediatric diseases, treatment is pre-dominantly derived from adult PACNS literature.10,11 Whiletherapeutic strategies for cPACNS have been described, theycome largely from observational and open-label cohort studiesbecause of the absence of randomized control trials and are notstandardized across centers. To date, no scoping review of thepediatric literature on treatment of cPACNShas been performed.Therefore, in this review, we aim to describe the evidence andefficacy of treatment regimens reported in the pediatric literature.

MethodsSearch strategy and selection criteriaA scoping review of the literature was conducted for thetimeframe January 2000 to December 2018. Databases

utilized in the search were Ovid, MEDLINE, PubMed,Embase, Cochrane Database of Systematic Reviews,Cochrane Central Register of Controlled Trials, Clin-icalTrials.gov, Vasculitis Foundation, EUVAS (EuropeanVasculitis Society), CanVasc, Google Scholar, and Web ofScience. Searches were conducted with MeSH terms, in-cluding vasculitis, vasculitides, angiitis, angiitides, arteritis,central nervous system, CNS, brain, cerebral PACNS, pedi-atric, adolescent, child, and infant.

Studies were eligible if they met the following inclusion cri-teria: (1) treatment reported, (2) pediatric populationaddressed, (3) focused on disease of interest, cPACNS, (4)included ≥5 patients, (5) original research, and (6) full-lengtharticles. Only articles written in English were included.Reviews, expert opinions, editorials, case reports, and studieswithout treatment information were excluded. Studies werescreened by perusing titles and abstracts for relative content.Articles deemed relevant were selected for full-text review andindividually assessed for inclusion criteria using the STROBEchecklist for cohort, case-control, and cross-sectional studiesby 2 of 4 reviewers: J.B., A.D., B.G., and M.T. In case ofconflicting evaluations, a third reviewer, S.M.B., was asked tomake the final decision. Selected articles were assessed forquality. Studies reporting data on the same population wereincluded individually and discussed together.

Data analysisQuality was assessed through utilization of a modified versionof Pasma et al.12 Quality Assessment Tool. Questionsaddressed patient recruitment through sampling method,participation, treatment and outcome measurements, andconflict declaration. Nonrelevant questions were droppedfrom the tool. Three questions were deemed essential: >80%participation, reproducible treatment strategy, and re-producible outcome measure. A score of 1 was given to eachquestion on satisfaction of both reviewers’ assessment, witha maximum total score of 6. A study with a total score of 4 orhigher and at least 2 of 3 essential questions was consideredhigh quality.

Studies were evaluated for design, location, sample size, andpatient demographics. This information together with treat-ment strategies and outcomes was aggregated. Studies werecompared by diagnostic subtype, treatment regimen, andoutcomes. Acceptable study outcomes included mortality andneurologic outcome, preferably using the Pediatric StrokeOutcome Measure (PSOM),13 a determinant of neurologic

GlossaryAN = angiography-negative; APNP = angiography-positive nonprogressive; APP = angiography-positive progressive; ASA =aspirin; CA = conventional angiography; cPACNS = childhood primary angiitis of the CNS; DSA = digital subtractionangiography; IVMP = IV methylprednisolone; MMF = mycophenolate mofetil; MRA = magnetic resonance angiography;NASCET = North American Symptomatic Carotid Endarterectomy Trial; PedsQL = Pediatric Quality of Life; PGA =Physicians Global Assessment; PSOM = Pediatric Stroke Outcome Measure; vWF = von Willebrand factor.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 118: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

dysfunction across 4 domains: sensorimotor, language pro-duction, language comprehension, and cognition/behavior.

Role of funding sourceThere was no funding source for this study. The corre-sponding authors had full access to all the data in the studyand final responsibility for the decision to submit forpublication.

Data availabilityData not provided in this article including comprehensivetreatment and outcome summaries and search strategies areavailable to be shared by request.

ResultsThe search strategy identified a total of 2,596 articles, with 1additional study identified by author MT (figure 1). Of 2,597articles screened for title and abstract, 110 articles had full-textretrieval for assessment of inclusion criteria and 9 originalarticles were included for detailed analysis2,4,9,14–19 (figure 1).Reasons for article exclusion included: treatment descriptionmissing (n = 24), exclusion of pediatric population (n = 8),not PACNS (n = 5), case reports <5 children (n = 20), book

chapters/reviews/editorials (n = 39), and abstract only (n =5). Seven studies were deemed high quality after quality as-sessment was conducted. Study characteristics are summa-rized in table. Of 9 studies reviewed, 4 described the samestudy population from Lahore, Pakistan16–19; 4 studies de-scribed the same population from Toronto, Canada2,4,9,14;and 1 reported a case series from Los Angeles, US.15

The Pakistani group16–19 identified children <16 years of agesubsequently diagnosed with cPACNS at their center betweenJanuary 2009 and December 2010. Diagnostic categorizationwas two-fold: based on stroke characteristics and cPACNSsubtype. A total of 68 patients were identified: 50 presentedwith ischemic stroke, 10 with hemorrhagic stroke, and 8 withboth ischemic and hemorrhagic lesions. Alternatively, 51patients were classified as APNP-cPACNS and 17 as APP-cPACNS. Diagnoses were based on conventional angiography(CA) and/or MRA in all patients.

Induction therapy, 3 days of IV methylprednisolone (IVMP),and/or IV immunoglobulin for 5 days with subsequent oralprednisone taper over 30 days, was completed by 56 patients.Patients with ischemic stroke also received IV heparin andsubsequent oral anticoagulation therapy. Supplementary

Figure 1 Prisma 2009 flow diagram

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 119: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Table Summarized description of included studies

Study Diagnosis of cPACNS Age (median y) Treatment Outcome

Maliket al.19

aAPP (n = 17) 7·4 (1·5–16) Induction therapy 5–10 D (n = 56) Maintenance therapy 24 M (n = 56) Mortality atdischarge (n = 68)

Neurologic examination at discharge (n = 56)

APNP (n = 51) IVMP ± IVIG + oral prednisone + IVheparin for ischemic infarct

ASA: 41 pts Deceased: 12 pts Normal: 11 pts

APP and APNP diagnosesbased on CA and/or MRAabnormalities (n = 68)

ASA + AZA: 51 pts Survived: 56 pts Minor disabilities: 14 pts

Moderate disabilities: 11 pts

Severe disabilities: 20 pts

Maliket al.17

aAPP (n = 17) 7·4 (1·5–16) Induction therapy 5–10 D (n = 56) Maintenance therapy 24 M (n = 56) Mortality atdischarge (n = 68)

PSOM at discharge (n = 56)

APNP (n = 51) IVMP ± IVIG + oral prednisone + IVheparin for ischemic infarct

ASA: 40 pts See Malik et al.19 Normal: 11 pts

APP and APNP diagnosesbased on CA and/or MRAabnormalities (n = 68)

ASA + AZA: 16 pts Mortality at follow-up (n = 56)

Minor disabilities: 14 pts

Deceased: 20pts

Moderate disabilities: 11 pts

ASA: 15 pts Severe disabilities: 20 pts

ASA + AZA: 5pts

PSOM at follow-up (median 32 M) (n = 32)

Relapseb at follow-up (n = 56)

Normal: 8 pts

20 pts Minor disabilities: 10 pts

ASA: 18/40 pts Moderate disabilities: 10 pts

ASA + AZA: 2/16 pts

Severe disabilities: 4 pts

Maliket al.18

aAPP (n = 17) 7·4 (1·5–16) Acute anticoagulation therapy (n = 58) Long-term therapy (n = 54) Mortality atdischarge (n = 68)

Neurologic examination at discharge (n = 56)

APNP (n = 51) IV heparin for ischemic infarct ASA: 40 pts See Malik et al.19 See Malik et al.19

Continued

4Neurology:N

euroim

muno

logy&Neuroinflam

mation

|Vo

lume6,N

umber

4|

July2019

Neurology.org/N

N

Page 120: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Table Summarized description of included studies (continued)

Study Diagnosis of cPACNS Age (median y) Treatment Outcome

APP and APNP diagnosesbased on CA and/or MRAabnormalities (n = 68)

ASA + AZA: 14 pts

Alhaboobet al.16

aAPP (n = 17) 7·4 (1·5–16) Induction therapy 5-10 D (n = 56) Maintenance therapy 24 M (n = 56) Mortality atdischarge (n = 68)

Neurologic examination at discharge (n = 56)

APNP (n = 51) IVMP ± IVIG + oral prednisone + IVheparin for ischemic infarct

ASA therapy: 41 pts See Malik et al.19 See Malik et al.19

APP and APNP diagnosesbased on CA and/or MRAabnormalities (n = 68)

ASA + AZA: 15 pts

Benseleret al.2

APP (n = 20) 7·17 (0·7–17·6) Acute antithrombotic therapy (n = 62) Long-term antithrombotic therapy +steroids ± IV cyclo (n = 22)

Mortality (n = 62) Complete neurologic recovery con PSOM atfollow-up (mean 20 M) (n = 22)

APNP (n = 42) Long-term antithrombotic therapy (n =40)

APP-cPACNS: 13 pts 0 pts APP-cPACNS: 9 pts

APP and APNP diagnosesbased on MRA abnormalities(n = 62) and/or CAabnormalities (n = 51)

APP-cPACNS: 7 pts APNP-cPACNS: 9 pts APNP-cPACNS: 13 pts

APNP-cPACNS: 33 pts

Hutchinsonet al.9

AN (n = 19) 9·8 (5·5–17·8) Induction therapy 6 M (n = 14) Maintenance therapy 18 M (n = 13) Mortality (n = 19) Good neurologic outcomed on PSOM:

Confirmatory biopsy (n = 19) IV cyclo + oral prednisone MMF or AZA + oral prednisone 0 pts 12 M: 8/16 pts

Angiography methodsincluded: MRA, venography,and CA

7/9 pts AZA switched to MMF 24 M: 9/13 pts

Confirmatory biopsy includedlesional/nonlesionalspecimen and lymphocyticvasculitis histology

Cellucciet al.4

AP (n = 14) 9·8 (3·3–17·8) ASA therapy (n = 14) IV cyclo x 6 M ± MMF or AZA (n = 27) Mortality (n = 39) Good neurologic outcomed on PSOM:

AN (n = 25) AP-cPACNS AP-cPACNS: 7 pts 0 pts 52% of patients at 12 mo

AP diagnosis based on CAand/or MRA abnormalities(n = 14)

Oral prednisone 2-6 M or 12 M (n = 38) AN-cPACNS: 23 pts Flare duringfollow-up (n = 39)

65% of patients at 24 mo

Continued

Neurolo

gy.org/N

NNeu

rology:N

euroimmunology

&Neuroinflam

mation

|Volum

e6,N

umber

4|

July2019

5

Page 121: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Table Summarized description of included studies (continued)

Study Diagnosis of cPACNS Age (median y) Treatment Outcome

AN diagnosis based on brainbiopsy showing lymphocyticvasculitis (n = 22)

APNP-cPACNS: 13 pts 3 pts did not receive IV cyclo 6 pts Summary scores significantly decreasedover time (p < 0.001)

Those who did not undergobrain biopsy had clinicalfeatures, blood work, andlumbar puncture consistentwith AN-cPACNS, and negativemicrobiologic andantineuronal investigations

AN-cPACNS: 25 pts No maintenance therapy (n = 2)

Elberset al.14

AP (n = 27) 7·5 (2–15) Unfractionated heparin (n = 27) Acute IVMP + oral steroid (n = 12) Mortality Stroke recurrence (n = 39)

Abnormal MRA (n = 25) Transitioned to warfarin: 5 pts Long-term immunosuppression (n =6)

0 pts 4 pts

Abnormal DSA (n = 24) Transitioned to ASA: 17 pts ± ASA: 4 pts

Abnormal DSA was found inthe 2 patients with normalMRA

Transitioned off heparin: 1 pt MMF: 2 pts

Gallagheret al.15

AP (n = 5) 8 (5–11) Induction therapy (n = 5) Maintenance therapy (n = 3) Mortality (n = 5) Clinical/radiologic evaluation at last follow-up (n = 5)

AP diagnosis based on MRAabnormalities (n = 3) and/orCA abnormalities (n = 4)

Oral prednisone ± IVMP ±anticoagulation + IV cyclo

Methotrexate or AZA ± ASA 0 pts Asymptomatic: 3 pts

Lost to follow-up: 1 pt

Residual deficits: 1 pt

Abbreviations: AN = angiography-negative; AP = angiography-positive; ASA = aspirin; AZA = azathioprine; CA = conventional angiography; cPACNS = childhood primary angiitis of the CNS; D = days; DSA = digital subtractionangiography; IV cyclo = IV cyclophosphamide; IVIG = IV immunoglobulin; IVMP = IV methylprednisolone; M = months; MRA = magnetic resonance angiography; NP = nonprogressive; P = progressive; PSOM = pediatric strokeoutcome measure; pts = patients.a Also described as ischemic stroke (n = 50), hemorrhagic stroke (n = 10), and both (n = 8).b Relapse defined as emergence of signs and symptoms of stroke confirmed with neuroimaging of brain after remission.c Complete neurologic recovery defined as neurologic deficit severity score of 0.d Good neurologic outcome defined as ≤0.5 across any domain.

6Neurology:N

euroim

muno

logy&Neuroinflam

mation

|Vo

lume6,N

umber

4|

July2019

Neurology.org/N

N

Page 122: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

calcium and vitamin D were prescribed with anticonvulsants,antipsychotics, antibiotics, antivirals, and antacids as needed.A total of 12 patients died before completing inductiontherapy from complications involving cerebral artery and/orparenchymal bleeding. The remaining 41 patients were allo-cated to 24 months of maintenance therapy consisting ofeither aspirin (ASA) daily for ischemic stroke or together withdaily azathioprine for progressive arteriopathies. Despitea clearly delineated therapeutic regimen, the Pakistani groupreported conflicting data in regard to the number of patientsassigned to each therapy across the 4 studies. Both the initialstudy by Malik et al.19 and Alhaboob et al.16 identified 41patients assigned to ASA alone and 15 patients to ASA to-gether with azathioprine for 24 months. However, Maliket al.17 indicated that 40 patients were assigned to the 24months ASA only group, and 16 patients to ASA and aza-thioprine, with ASA duration increased to 60 months in thisstudy. Malik et al.18 did not provide details of inductiontherapy, only that ischemic infarcts were initially treated withIV heparin. From this group, 40 patients were discharged onASA for 24 months and 14 received adjunctive azathioprinetherapy.

The Pakistani group reportedmortality and morbidity acrossdifferent intervals. Remission was described as a completeabsence of disease activity in clinical symptoms, exam find-ings, lab markers, and imaging for at least 3 months, andrelapse was defined as an emergence of signs/symptoms ofstroke confirmed by neuroimaging (CA and/or MRA) afterremission. All 4 studies documented mortality at dischargein 12 patients (17·6%). Malik et al.17–19 reported clinicalstate at time of discharge for all survivors, defined as neu-rologic assessment for motor, visual, and/or speech diffi-culties. Normal examination was reported in 11 (20%),minor disability in 14 (25%), moderate disability in 11(20%), and severe disability in 20 (35%) patients. Maliket al.15 described relapse in a total of 30 patients (54%)during maintenance therapy. From the ASA only group, 18patients (45%) relapsed within the first 24 months, whichresulted in 10 deaths, and of the remaining 22 patients(55%) who completed ASA therapy, 5 relapsed and 5 died.Overall mortality in the ASA group was 15/40 (38·5%). Ofthe 16 patients on maintenance with ASA and azathioprine,2 patients (13%) relapsed within 24 months and both sur-vived. The number of deaths reported was 5 (31%): 3patients died from massive cerebral hemorrhage, and 2 of 3patients died after relapse within 6 months of successfullycompleting therapy. Therefore, the overall mortality duringmaintenance reported by Malik et al.17 at a medium follow-up of 34 months was 20 (35·7%) of the initial 56 survivors,either as a direct result or as a complication of first or secondrelapse, with 5 deaths reportedly due to non-cPACNS cau-ses. Malik et al.17 also reported the outcome at last follow-upusing the PSOM.13 From the 32 patients assessed: 8 (25%)a normal outcome, 10 (31%) minor disabilities, 10 (31%)moderate disabilities, 4 (13%) severe disabilities, and 4patients were lost to follow-up.

The Canadian studies2,4,9,14 similarly identified patients fromthe same center, though each study varied in sample size anddiagnostic subtype. Benseler et al.2 identified 62 children <18years of age between January 1990 to December 2002 di-agnosed with AP-cPACNS: 42 APNP-cPACNS and 20 APP-cPACNS. Diagnoses were based on MRA and/or CAabnormalities.

All patients received antithrombotic therapy with ASA, hep-arin, or warfarin. A total of 41 patients (65%) continued withthis treatment approach long-term: 33 patients with APNP-cPACNS and 7 patients with APP-cPACNS. In contrast, 22patients (35%) received immunosuppressive interventionwith steroids alone or a combination of steroids and IV cy-clophosphamide: 9 (41%) APNP-cPACNS and 13 (59%)APP-cPACNS. Benseler et al.2 measured outcome at lastevaluation (mean 20 months) as complete neurologic re-covery, defined as neurologic deficit severity score of 0 onPSOM; any other deficit severity score was defined as in-complete recovery. A total of 22 patients (35%) madea complete neurologic recovery: 13 (31%) of APNP-cPACNSand 9 (45%) of APP-cPACNS. No deaths were reported.

Hutchinson et al.9 identified 19 patients <18 years of age withAN-cPACNS diagnosis between January 2002 and December2009. MRA, venography, and CA were used. Diagnosis in allpatients was based on a confirmatory biopsy. Patients receivedinduction and maintenance therapy for 24 months. Inductiontherapy was defined as 6 months of IV cyclophosphamidemonthly, together with Pneumocystis jiroveci pneumonia pro-phylaxis, oral prednisone daily, supplementary calcium, vita-min D, and anticonvulsants/antipsychotics as needed.Thirteen patients (68%) completed induction and went on tomaintenance therapy. Maintenance consisted of an additional18 months of therapy, with 5 patients (35%) assigned tomycophenolate mofetil (MMF) and 9 (65%) assigned toazathioprine with anticonvulsants/antipsychotics as needed.Seven (78%) of 9 patients on azathioprine switched to MMFbecause of treatment failure or intolerance. Only 5 patients(36%) completed maintenance therapy and were taken offmedication. Primary outcome was assessed on PSOM at 24months, with good outcome defined as a score of ≤0·5 acrosseach of the 4 domains. Secondary outcomes included treat-ment efficacy and safety, determined by PSOM at 12 monthsand final follow-up, disease flare, and discontinuation ofimmunosuppressants. Good neurologic outcome on PSOMat 12 months was reported in 8 (50%) of 16 patients and 9(70%) of 13 at final follow-up. Disease flare during treatmentwas described in 8 patients (42%): 2/19 (11%) during in-duction, 5/14 (36%) during maintenance, and 1/5 (20%) offmedication. No patients relapsed onMMF. Of 5 patients whocompleted maintenance therapy, 4 achieved remission, de-fined as complete absence of disease activity in clinicalsymptoms, examination findings, lab markers, and imaging forat least 3 months. Safety was determined by mortality, seriousinfection, and each of cataracts, avascular necrosis, vertebralfractures, or type II diabetes mellitus. No patients in this study

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 7

Page 123: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

died; however, 2 were reported to have had a serious infectionrequiring hospital admission; 1 developed cataracts, 1 avas-cular necrosis, and 3 with vertebral fractures. No diabetesmellitus was reported. Pediatric Quality of Life (PedsQL)20

was reported for both child and parent at final follow-up,together with cognitive outcome testing by the Weschler In-telligence Scale for Children IV and Weschler Adult In-telligence Scale III at 14 months postdiagnosis. Abnormalscores were defined as any index scores that were ≥1 SDsbelow the mean of the normative sample. The PedsQLquestionnaire was completed by 17 patients and 15 parents:median physical score was 78 (range = 0–100) for childrenand 66 (range = 9–100) for parents, and median psychosocialscore 73 (range = 27–100) for children and 63 (range =37–100) for parents. Cognitive outcome was assessed in 10children: Full Scale Intelligence Quotient was reported asabnormal in 8 (80%), working memory in all 10 (100%),verbal comprehension in 7 (70%), perceptual reasoning in 6(60%), and processing speed in 5 (50%).

Cellucci et al.4 identified 39 patients <18 years of age betweenJune 2001 and October 2010 diagnosed with cPACNS: 14 AP-cPACNS, not further subdivided, and 25 AN-cPACNS. AP-cPACNS diagnosis was based on CA and/or MRA (n = 14).AN-cPACNS diagnosis was based on confirmatory biopsy in 22patients. Patients without biopsy had features, blood work, andlumbar puncture consistent with AN-cPACNS. Some patientswith AP-cPACNS were also represented in the Benseler et al.2

study and in the Hutchinson et al.9 study.

All 14 patients with AP-cPACNS received ASA, and 9 (64%)received additional anticoagulation with unfractionated/low-molecular weight heparin for 2 months. High-dose predni-sone was prescribed in all but 1 patient with AP-cPACNS,with a varied tapering schedule: 2–6 months in 6 patients(46%) and 12 months in 7 (54%). IV cyclophosphamide over6 months followed by MMF or azathioprine was given toa further 7 patients (54%) with AP-cPACNS; 1 patient withAP-cPACNS received anticoagulation therapy only. High-dose prednisone with 12-months taper was prescribed in allpatients with AN-cPACNS. IV cyclophosphamide for 6months in combination with azathioprine or MMF wasreported in 20 patients (80%) with AN-cPACNS, while aza-thioprine or MMF alone as induction/maintenance therapywas described in 3 patients (12%) with AN-cPACNS. Twopatients (8%) with AN-cPACNS did not receive therapy.Outcomes were reported serially across 24 months and in-cluded disease activity as measured by the Physicians GlobalAssessment (PGA), a visual analog scale with a score of zeroindicating no disease activity and 10 indicating severe activity;von Willebrand factor (vWF) antigen levels where highnumbers indicate high disease activity with >1·40 IU/mLbeing considered abnormal; and neurologic outcome onPSOM with good outcome defined as a score of ≤0·5 acrosseach domain. Disease activity was elevated at time of di-agnosis, but significantly decreased over time in all patients(p < 0·001). vWF antigen levels decreased over time (p <

0·001). PSOM summary scores decreased significantly overtime (p < 0·001), with 52% of patients having a good neu-rologic outcome at 12 months, and 65% at 24 months. A totalof 6 (15%) developed disease flare during follow-up defined asan increase in PGA by at least 1 cm in presence of recurrentsymptoms, laboratory changes, and/or MRI findings. Nopatients died.

Elbers et al.14 described 27 patients <18 years of age with AP-cPACNS diagnosed between January 1998 and December2013. Diagnosis was based on MRA and DSA abnormalities.Patients might overlap with the Benseler et al.2 and Cellucciet al.4 studies. Acute treatment with unfractionated heparinwas described in 23 patients (85%): 5 subsequently transi-tioned to warfarin, 17 to ASA, and none in 1. ASA alone,or with immunosuppression, was described in the remaining4 patients. Two patients received chronic immunosuppres-sion with MMF and 1 received plasmapheresis. IVMP to-gether with a 3-month oral steroid taper was described in 12patients, with no long-term immunosuppression. An addi-tional 7 patients were treated with acyclovir. Outcomes in-cluded vascular imaging at 12 months, assessed using theNorth American Symptomatic Carotid Endarterectomy Trial(NASCET) criteria and described as improved (normal an-giography, improved flow, or fewer abnormal vessels), stable(no change in flow abnormality or number of involved ves-sels), or worsened (involvement of new vessels or worseningof an existing flow abnormality by at least 1 point on follow-upimaging), and possibly discordant (worsened patients wherea new or worsening arterial abnormality coexisted with animproved or normalized vessel), as well as stroke recurrence.The NASCET criteria scores vessel narrowing from 1 to 4progressing from normal (0%–9%), to mild (10%–29%),moderate (30%–69%), and severe stenosis (70%–99%) orcomplete occlusion (no flow detected). A total of 10 (37%)patients were described as improved, 4 of which receivedsteroids as treatment; 6 (22%) as stable, 2 of which receivedsteroids; and 11 (41%) worsened with 7 described as dis-cordant, 6 of which received steroids. Stroke recurrence wasreported in 4 patients (15%). No patients died.

Gallagher et al.15 described 5 cases of AP-cPACNS at theircenter with no fixed treatment regimen. Diagnosis was basedonMRA and/or CA abnormalities. Two patients were treatedwith IVMP, either at their first or subsequent disease pre-sentation. All patients received IV cyclophosphamide; how-ever, 1 discontinued after the first infusion due to intolerance.All patients received oral prednisone of varying doses; 2 re-ceived short-term (<6 months) steroid therapy, 2 remainedon oral steroids long-term (>6 months), and 1 was lost tofollow-up after 6 months. One patient received long-termtreatment with azathioprine and 2 with methotrexate. Four(80%) of the 5 patients received anticoagulation therapy witheither ASA or warfarin during their illness. No validatedoutcome measures were used to assess outcomes for thesepatients; however, at last follow-up, 3 patients were reportedas neurologically asymptomatic, 1 had mild residual deficits,

8 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 124: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

and 1 was lost to follow-up but continued on treatment attheir last documented visit.

DiscussionThis is the first scoping review of treatment in cPACNS.Despite the paucity of randomized clinical trials, evidence in

the literature offers support for treatment strategies. ForAPNP-cPACNS, the authors recommend treatment withlong-term antiplatelet therapy to reduce the risk of strokerelapse and mortality as reported in the Pakistani cohort.16–19

The authors concur with CNS vasculitis expert opinion inrecommending short-term immunosuppressive with IVMPand acute antithrombotic therapy, subsequently followed by

Figure 2 Recommended treatment protocol

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 9

Page 125: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

high-dose oral prednisone (figure 2A).3,10,21–23 In APP-cPACNS, the authors support a combination of anti-coagulation and induction therapy, with both IV steroids andIV cyclophosphamide with steroid taper (figure 2B)3,4,15–19

to increase the prospect of neurologic recovery.3,10,15,21–26

Induction therapy for AN-cPACNS should similarly consist ofIV steroids and IV cyclophosphamide (figure 2C). Evidencein the literature9,10,21–24 for long-termmaintenance therapy inAPP-cPACNS and AN-cPACNS supports daily MMF/mycophenolic acid in preference to azathioprine to avoidthe possibility of treatment failure or intolerance, as reportedby Hutchinson et al.9 and the Pakistani group (figures 2B,2C).16–19

In refractory disease, Batthish et al.25 reported successful useof infliximab therapy in treating 2 cases of AN-cPACNS withgood control of inflammation and subsequent prevention ofbrain damage. Rosati et al.26 recently published a case series of4 patients with AP-cPACNS successfully treated with long-term MMF. All revealed subsequent stability or improvementon MRI/MRA with no progression of arterial disease, and norelapses were reported in the follow-up period (range 10–42months).26 Sen et al.24 also report successful maintenancetreatment with MMF in 3 cases of cPACNS that had failedmethotrexate or azathioprine and steroid treatment alone. Nopatients were reported to have had recurring symptoms, sideeffects, or new lesions on MRI.24

While this review provided a comprehensive review of treat-ment literature in cPACNS, there were several limitations.Based on our quality assessment definition for Q4, a mini-mum of 1 reproducible reported outcome was sufficient tosatisfy outcome reproducibility. Many studies satisfied thiscriterion by reporting mortality, despite a lack of or irrepro-ducibility of supplementary outcomes. Furthermore, usinga validated measure like the PSOMwas sufficient to satisfy theQ5 requirement. While the Pakistani group utilized thePSOM in outcome assessment, the lack of outcome defi-nitions led to difficulty in result generalizability. Finally, whilethe same patient cohort was included in all 4 Pakistani groupstudies and treatment regimens were well described, thenumber of patients reported to each treatment arm wasgrossly inconsistent and calculated incorrectly. Despite thesecontradictions, the studies sufficiently satisfied the re-producible treatment criterion and were consequently ratedhigh quality. Therefore, the authors believe that the qualityassessment for a number of studies reviewed is overstated;were these studies to be rescored retrospectively, they wouldbe deemed low quality. Finally, the case series by Gallagheret al.,15 despite meeting inclusion criteria, only described casereports with inconsistent treatment regimens and lackedvalidated outcome measures.

In conclusion, available literature on treatment in cPACNS werethoroughly reviewed and findings summarized. Based on theevidence and current expert opinion, the authors provide rec-ommendations for cPACNS treatment strategies. Rapid initiation

with the recommended therapeutic interventions would serve tooptimize survival and prevent permanent brain injury in patientswith cPACNS to achieve the best possible outcome.

AcknowledgmentThe authors would like to thank librarian Rachel Zhao for hercontribution to the development and application of the searchstrategy.

Study fundingNo targeted funding.

DisclosureJ. Beelen reports no disclosures; S. Benseler serves on theeditorial board for Neurology: Neuroimmunology & Neuro-inflammation; A. Dropol, B. Ghali, and M. Twilt reports nodisclosures. Go to Neurology.org/NN for full disclosures.

Publication historyReceived byNeurology: Neuroimmunology & Neuroinflammation January24, 2019. Accepted in final form March 4, 2019.

References1. Calabrese LH, Mallek JA. Primary angiitis of the central nervous system. Report of 8

new cases, review of the literature, and proposal for diagnostic criteria. Medicine(Baltimore) 1988;67:20–39.

2. Benseler SM, Silverman E, Aviv RI, et al. Primary central nervous system vasculitis inchildren. Arthritis Rheum 2006;54:1291–1297.

3. Benseler SM, deVeber G, Hawkins C, et al. Angiography-negative primary centralnervous system vasculitis in children: a newly recognized inflammatory central ner-vous system disease. Arthritis Rheum 2005;52:2159–2167.

4. Cellucci T, Tyrrell PN, Sheikh S, Benseler SM. Childhood primary angiitis of thecentral nervous system: identifying disease trajectories and early risk factors forpersistently higher disease activity. Arthritis Rheum 2012;64:1665–1672.

Appendix Authors

Name Location Role Contribution

JocelyneBeelen,BMSc

Cumming School ofMedicine, Universityof Calgary, Calgary

Author Evaluated literature;compiled the data;drafted the manuscript;completed manuscriptrevisions

SusanneBenseler,MD, PhD

Alberta Children’sHospital, Universityof Calgary, Calgary

Author Conceptualized thepaper; revised the papercritically for intellectualcontent

AnastasiaDropol,HBSc

Cumming School ofMedicine, Universityof Calgary, Calgary

Author Conceptualized thepaper; evaluated theliterature; compiledthe data; drafted themanuscript

BriannaGhali, Bsc

University ofCalgary, Calgary

Author Completed the literaturesearch; evaluated theliterature; revised thepaper critically forintellectual content

MarinkaTwilt, MD,PhD

Alberta Children’sHospital, Universityof Calgary, Calgary

Author Conceptualized thepaper; completed theliterature search;evaluated the literature;revised the papercritically for intellectualcontent

10 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 126: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

5. Cellucci T, Tyrrell PN, Twilt M, Sheikh S, Benseler SM. Distinct phenotype clustersin childhood inflammatory brain diseases: implications for diagnostic evaluation.Arthritis Rheumatol 2014;66:750–756.

6. Splendiani A, Catalucci A, Limbucci N, Turner M, Krings T, Gallucci M. Pediatricinflammatory diseases: part III: small vessels vasculitis. Neuroradiol J 2012;25:715–724.

7. Elbers J, Halliday W, Hawkins C, Hutchinson C, Benseler SM. Brain biopsy inchildren with primary small-vessel central nervous system vasculitis. Ann Neurol2010;68:602–610.

8. Aviv RI, Benseler SM, DeVeber G, et al. Angiography of primary central nervoussystem angiitis of childhood: conventional angiography versus magnetic resonanceangiography at presentation. AJNR Am J Neuroradiol 2007;28:9–15.

9. Hutchinson C, Elbers J, Halliday W, et al. Treatment of small vessel primary CNSvasculitis in children: an open-label cohort study. Lancet Neurol 2010;9:1078–1084.

10. Twilt M, Benseler SM. The spectrum of CNS vasculitis in children and adults. Nat RevRheumatol 2011;8:97–107.

11. Salvarani C, Brown RD Jr, Calamia KT, et al. Primary central nervous systemvasculitis: analysis of 101 patients. Ann Neurol 2007;62:442–451.

12. Pasma A, van’t Spijker A, Hazes JM, Busschbach JJ, Luime JJ. Factors associated withadherence to pharmaceutical treatment for rheumatoid arthritis patients: a systematicreview. Semin Arthritis Rheum 2013;43:18–28.

13. Kitchen L, Westmacott R, Friefeld S, et al. The pediatric stroke outcome measure:a validation and reliability study. Stroke 2012;43:1602–1608.

14. Elbers J, Armstrong D, Yau I, Benseler S. Vascular imaging outcomes of childhoodprimary angiitis of the central nervous system. Pediatr Neurol 2016;63:53–59.

15. Gallagher KT, Shaham B, Reiff A, et al. Primary angiitis of the central nervous systemin children: 5 cases. J Rheumatol 2001;28:616–623.

16. Alhaboob AA, Hasan GM, Malik MA, Rehman MZ. Therapeutic benefits and sideeffects of Azathioprine and Aspirin in treatment of childhood primary arterial stroke.Ann Neurosci 2014;21:10–13.

17. Malik MA, Ahmad N, Malik H. Maintenance treatment of childhood primary angiitisof central nervous system with spirin and azathioprine. Pediatr Ther 2013;3:174–179.

18. Malik MA, Choudry GR, Malik H. Recurrence prevention of childhood primaryangiitis of central nervous system by combination of azathioprine and aspirin. Am JMed Stud 2013;1:22–27.

19. Malik MA, Zia-ur-Rehman M, Nadeem MM, et al. Childhood primary angiitis of thecentral nervous system. J Coll Physicians Surg Pak 2012;22:570–574.

20. Varni JW, Burwinkle TM, Seid M, Skarr D. The PedsQL 4.0 as a pediatric populationhealth measure: feasibility, reliability, and validity. Ambul Pediatr 2003;3:329–341.

21. Twilt M, Benseler SM. Childhood inflammatory brain diseases: pathogenesis, di-agnosis and therapy. Rheumatology (Oxford) 2014;53:1359–1368.

22. Gowdie P, Twilt M, Benseler SM. Primary and secondary central nervous systemvasculitis. J Child Neurol 2012;27:1448–1459.

23. Moharir M, Shroff M, Benseler SM. Childhood central nervous system vasculitis.Neuroimaging Clin N Am 2013;23:293–308.

24. Sen ES, Leone V, AbinunM, et al. Treatment of primary angiitis of the central nervoussystem in childhood with mycophenolate mofetil. Rheumatology (Oxford) 2010;49:806–811.

25. Batthish M, Banwell B, Laughlin S, et al. Refractory primary central nervous systemvasculitis of childhood: successful treatment with infliximab. J Rheumatol 2012;39:2227–2229.

26. Rosati A, Cosi A, Basile M, et al. Mycophenolate mofetil as induction and long-termmaintaining treatment in childhood: primary angiitis of the central nervous system.Joint Bone Spine 2017;84:353–356.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 11

Page 127: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

VIEWS & REVIEWS OPEN ACCESS

Immunoglobulin G4-related hypertrophicpachymeningitisA case-oriented review

Michael Levraut, MD, Mikael Cohen, MD, Saskia Bresch, MD, Caroline Giordana, MD,

Fanny Burel-Vandenbos, MD, PhD, Lydiane Mondot, MD, Jacques Sedat, MD, Denys Fontaine, MD, PhD,

Veronique Bourg, MD, Nihal Martis, MD, and Christine Lebrun-Frenay, MD, PhD

Neurol Neuroimmunol Neuroinflamm 2019;6:e568. doi:10.1212/NXI.0000000000000568

Correspondence

Dr. Levraut

[email protected]

AbstractObjectiveMeningeal involvement in Immunoglobulin G (IgG)-4-related disease is rare and only de-scribed in case reports and series. Because a review into the disease is lacking, we present 2 casesfollowed by a literature review of IgG4-related hypertrophic pachymeningitis (IgG4-HP).

MethodsTwo IgG4-HP cases were reported, one involving the spinal cord and responding to surgicalmanagement and a second involving the brain and responding to Rituximab therapy. We thenreview clinical cases and case-series of histologically proven IgG4-HP that were published in thePubMed-NCBI database.

ResultsForty-two case reports and 5 case-series were studied (60 patients, 20 women). Themedian agewas 53. Eighteen patients had systemic involvement and 24 had single-organ IgG4-HP. Fifty-five percent of patients had an elevated serum IgG4. Treatment was surgical in 20/53 cases.Steroid therapy and immunosuppressors were effective in 85% and more than 90% of the cases,respectively. The rate of disease relapse was 42.1% after steroid therapy was discontinued.

Discussion/conclusionIgG4-HP is characterized by the lack of extra-neurologic organ-involvement and systemic signs.Histopathologic studies should be performed as it is crucial for diagnosis because serummarkers are rarely informative. 18F-FDG positon tomography can be useful to characterizesystemic forms. There is no specific CSF marker for IgG4-HP and the diagnostic value of CSFIgG4 levels needs to be studied with larger samples.We provide a treatment algorithm for IgG4-HP. Such treatment strategies rely on early surgery, steroids, and early immunosuppressivetherapy to prevent neurologic complications.

From the Service de Medecine Interne (M.L., N.M.), Hopital l’Archet 1, Centre Hospitalier Universitaire de Nice, Universite Cote d’Azur; Service de Cytologie Pathologique (F.B.-V.),Hopital l’Archet 1, Centre Hospitalier Universitaire de Nice, Universite Cote d’Azur; and Service de Neurologie (M.C., S.B., C.G., V.B., C.L.-F.), Service de Radiologie (L.M.), Service deRadiologie Interventionnelle (J.S.), and Service de Neurochirurgie (D.F.), Hopital Pasteur 2, Centre Hospitalier Universitaire de Nice, Universite Cote d’Azur, France.

Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by the authors.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 128: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Immunoglobulin G (IgG)-4-related disease (IgG4-RD) isa polyclonal lymphoproliferative disorder affecting manyorgans. Diagnosis is histologic and shows lymphoplasmocyticinfiltration with IgG4+ plasma cell proliferation, storiformfibrosis, and obliterative phlebitis.1,2 The pancreas, salivaryglands, retroperitoneum, and lymph nodes are the mostcommonly affected.3–7 Central neurologic manifestations arerare with reports of mostly hypertrophic pachymeningitis(HP) and hypophysitis. Previous studies have found thatretrospective analyses of idiopathic HPwere able to ultimatelyidentify IgG4-RD in several cases.4 Specific diagnostic criteriahave been defined for HP associated with IgG4-RD (IgG4-HP), which rely on histopathologic analysis.7 Unfortunately,because of its scarcity, only case reports and a few case seriesof IgG4-HP are available in the medical literature. No treat-ment algorithm is known for this uncommon location ofIgG4-RD where relapsing occurs frequently. Because a betterunderstanding of the disease is needed, we report 2 cases anda literature review of the clinical, biological, and treatmentspecificities of IgG4-HP.

Case reportsCase 1A 55-year-old Caucasian male with a history of genetic he-mochromatosis, diabetes mellitus, and high blood pressurepresented with a progressive walking difficulty over a 2-weekperiod. Clinical examination reported T2-T3 dorsalgia, severeparesis with pyramidal syndrome, decreased vibratory sensa-tion of the lower limbs, and bladder dysfunction. There wasno weight loss, asthenia, fever, or extra-neurologic abnor-malities. Spinal cordMRI revealed many cervico-thoracic T2/fluid attenuated inversion recovery hyper intense signals fromC3 to T3 that were all enhanced by gadolinium injectionrevealing active inflammatory myelitis. Spinal venous di-latation of the cervico-thoracic junction suggestive of duralfistula was also reported (figure 1, A). The brain MRI wasnormal as were visual evoked potentials. Blood tests did notreveal an inflammatory syndrome (the C-Reactive proteinlevel was below 5 mg/L and fibrinogenemia was 3.15 g/L).Immunologic and infectious assessment did not help the di-agnosis. Anti-aquaporin 4 and myelin oligodendrocyte gly-coprotein auto-antibodies were negative. Analysis of the CSFrevealed elevated protein levels (1.96 g/L) and lymphocyticmeningitis (69 white blood cells [WBCs]/mm3 with 95%of lymphocytes). Viral PCR testing and CSF cytology werenot informative. The CSF IgG index was 0.53 despite a highlevel of CSF IgG (18.7 mg/dL) and there was no oligoclonalband. CSF levels of Interleukin-6 were high at 1923 pg/mL.

Treatment with IV corticosteroids (1 g per day for 5 days)resulted in partial regression of neurologic signs. Spinal cordarteriography confirmed the diagnosis of a dural fistula thatwas treated by embolization (figure 1, A). Three months later,the MRI showed complete regression of spinal cord lesions.The patient was able to walk without limitation to the walkingperimeter. Only mild proximal motor weakness and pro-prioceptive ataxia of the lower limbs persisted.

The same patient presented to our department 3 years laterwith similar neurologic features. The MRI of the spine founda posterior T1 isointense signal intensively enhanced withgadolinium and a T2 hypointense signal related to an intra-dural, extra-medullary, posterior, and compressive tumor-likelesion regarding T2-T3 surrounded with edema (figure 1, B).Surgical subtotal resection of the lesion was performed andconfirmed its extramedullary location, originating from thedura and highly adherent to the spinal cord pia mater. His-topathologic examination of the surgical specimen identifieda fibro-inflammatory lesion with dural thickening due toa lymphoplasmocytic infiltrate, storiform fibrosis, and oblit-erative vasculitis. Immunohistochemical analysis showeda polyclonal lymphocytic infiltration (expression of bothkappa and lambda light chains) including 53 IgG4+ plasmacells per high-power field (HPF) and an IgG4+/IgG ratiogreater than 40% (figure e-1a, links.lww.com/NXI/A115).There was no evidence of malignant, meningiomatous, orhistiocytic disease (negativity of epithelial markers cytokeratinand Epithelial Membrane Antigen, progesterone receptors,protein S100, CD1a, langherin, and Anaplastic LymphomaKinase). Staining for microbial pathogens (Gram, Gomori-Grocott and Ziehl) was negative. Plasma IgG4 levels were,however, normal (59 mg/dL). There was neither eosinophilianor complement consumption. [18F]-fluorodeoxyglucosepositron emission tomography (FDG PET) (PET-CT) didnot document other lesions. IgG4-HP was diagnosed basedon histopathologic criteria.

Near-complete clinical recovery was achieved 3 months aftersurgery. Bladder dysfunction had subsided and walking did notrequire help. Nevertheless, proprioceptive disorders of thelower limbs persisted. The MRI did not find residual lesions.Given the patient’s history of diabetes mellitus, steroids werenot initiated, and a close follow-up attitude was adopted.

Case 2A 66-year-old Caucasian male with a history of high bloodpressure, dyslipidemia, and smoking presented to the emer-gency department for subacute onset of non-painful bilateralloss of vision. He had been suffering from chronic pulsatile

GlossaryCYC = cyclophosphamide; GPA = granulomatosis with polyangeitis; HPF = high-power field; IgG4 = immunoglobulin G4;IgG4-HP = IgG 4-related hypertrophic pachymeningitis; IgG4-RD = IgG 4-related disease; RI = Responder Index; RTX =Rituximab; WBC = white blood cell.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 129: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

headaches for many months. There was no fever or weight loss.Physical examination revealed ptosis and left reactive mydriasiswith a decrease in bilateral visual acuity at 5/10 in the right eyeand less than 1/20 (light perception) in the left eye. Neurologicand general somatic examinations were normal. Visual evokedpotentials revealed a bilateral severe axomyelenic defect in botheyes. Funduscopic examination showed left temporal pale ret-ina. A brain MRI found a gadolinium-enhanced, bi-frontalmeningeal thickening invading the optic nerve sheaths andcavernous sinuses (figure 1, C). Microbial analyses and theimmunologic bioassay were not informative. There was no in-flammatory syndrome, eosinophilia, or complement consump-tion. Hypogammaglobulinemia was found on electrophoresis ofplasma proteins. Plasma IgG4 levels were normal (40.1mg/dL).

The CSF opening pressure was normal and its analysis showedprotein levels of 0.99 g/L without pleocytosis (2 WBC/mm3)or inflammatory markers (IgG index of 0.63 and CSF IgG-levelsat 6.29 mg/dL). Interleukins-6 and -10 levels were also normal.Meningeal biopsy identified an inflammatory infiltrate rich inIgG4+ plasma cells with perivascular storiform fibrosis andobliterative phlebitis. Immunostaining showed more than 10IgG4-positive plasma cells per HPF. The IgG4+/IgG ratio wasgreater than 40% (figure e-1b, links.lww.com/NXI/A115).PET-CT did not show pathologic hypermetabolism. The di-agnosis of IgG4-HP was made.

Corticosteroid therapy (1 mg/kg/d) after pulse IV methyl-prednisolone for 5 days at 1 g q.d. was introduced, but it was

Figure 1 MRI of both patients with IgG4-HP and spinal cord arteriography of the first patient

(A) Multiple T2 hyperintense signals (white arrows)(A.a) all enhanced after gadolinium injection (A.b)of the posterior half of the cervico-thoracic spinalcord. T2 hypointense signal of the posterior sub-arachnoid space from C5 to D3 suggesting a duralfistula (red arrow) (A.c), confirmed by medullaryarteriography (A.d) with opacification of a venouspeloton opposite the meninge of the left lateralpart of the dural sheath corresponding to themedullary veins visible on MRI. (B) Intra-ductal, in-tra-dural, postero-median, polylobulated tumorlesion, well-defined, in T1 isointense signal (B.a),intensely and homogeneously enhanced after in-jection of gadolinium (B.b), T2 hypointense signal(B.c), next to D2-D3 realizing a mass effect on thespinal cord. There is an extensive intramedullaryedema supra and under-lesion (B.d). (C) Bi-frontalmeningeal thickening enhanced by gadolinium in-vading the sheaths of optic nerves and cavernoussinuses (white arrows) (C.a and c). T2/FLAIR hy-perintense signal of the left optic nerve testifyingto radiologic optic neuritis, white arrows (C.b d).FLAIR = fluid attenuated inversion recovery.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 130: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

ineffective. IV Rituximab (RTX) was started 12 months later(1 g day 1 and day 15) and was followed by maintenancetreatment (M8). Fifteen months later, clinical and radiologicimprovement was observed with partial regression of head-aches and recovery of visual acuity at 10/10 in the right eyeand less than 1/20 (movements of the hand perceived at 50cm) in the left eye. At follow-up 1 year later, a brain MRIshowed the complete regression of meningeal thickening.

A systematic reviewMethods and definitionsWe reviewed clinical cases and case-series of cerebral and spinalcordHPwith histologically proven IgG4-RD that were publishedbetween January 2008 and September 2017 in the PubMed-NCBI database. The key words used in the search were“Pachymeningitis,” “IgG4-related disease,” “HP,” “IgG4-HP,”“IgG4 neurologic impairment,” and “IgG4CNS.”Demographics,clinical and biological data, pathology features, and treatmentstrategies were collected for each patient, when available.

Diagnostic criteria are those of the Japan College ofRheumatology8:

1. A clinical study showed characteristic diffuse/localizedswelling or masses in single or multiple organs

2. A hematologic study showed elevated levels of serumIgG4 (135 mg/dL or higher)

3. A histopathologic study showed the following 2 findings:c Histologic findings: marked lymphocyte and plasmo-

cytic infiltration and fibrosisc IgG4-positive plasma cell infiltration: ratio of IgG4/

IgG positive cell >40% and IgG4-positive plasma cells/HPF >10

When (1) + (2) + (3) are fulfilled, it is definite.

When (1) + (3) are fulfilled, it is probable.

When (1) + (2) are fulfilled, it is possible.

Isolated IgG4-HP was characterized by the absence of extra-meningeal organ-involvement (including the pituitary gland).Mutually exclusive lesions of the brain or spinal cord defined“single-organ pachymeningitis.” When extra-meningeal in-volvement was reported, IgG4-HP was defined as systemicIgG4-RD. Non-specified IgG4-HP was retained when datawas lacking for classification.

Data availabilityThe authors confirm that the data supporting the findings of thisstudy are available within the article and its supplementarymaterials (e-references e1-e47, links.lww.com/NXI/A115).

FindingsWe report findings from 60 patients that were taken from 42case reports and 5 case-series. The complete list of

bibliographical references is provided in the supplementaldata online document.

Patients’ characteristicsBrain involvement was mostly reported (49/60 cases). Spinalcord lesions were described in 16 cases, of which 9 werecervical. Thoracic and lumbar spinal cord lesions were de-scribed respectively in 8 and 5 cases. The disease affected boththe brain and the spine in 5 patients. Systemic IgG4-RD wasnoted in 18 cases and mostly accompanied cranial manifes-tations of IgG4-HP. The extra-meningeal manifestations arereported in figure 2. Isolated meningeal damage was identifiedin 25 patients of whom 24 had single-organ IgG4-HP. Sev-enteen patients could not be classified because of a lack ofdata.

The demographics, clinical, pathology, and blood work-upsare shown in table 1. Patients with IgG4-HP were mostlymales with a male-to-female ratio of 2. Median age was 53years (range: 19–82). Headache was common in cases ofcerebral IgG4-HP. Visual impairment and cranial nerve palsywere each present in less than one-third of cases.

Plasma IgG4 levels of more than 135 mg/dL were found in 20of the 36 patients for whom this information was available.The median level of plasma IgG4 levels was higher in patientswith systemic IgG4-RD, whereas meningeal IgG4 infiltrationwas higher in isolated IgG4-HP. A biological inflammatorysyndrome was inconsistently present (10/27 cases) and wasmore frequently found with systemic involvement. Eosino-philia was only reported in the case of a patient who alsopresented with eosinophilic granulomatosis with polyangiitis.

CSF analysisCSF findings are presented in table 2. Aseptic lymphocytemeningitis was identified in more than half of the specimenswith median CSF protein levels of 0.91 g/L and a median CSFwhite blood count of 19.5/mm3. The CSF profile was gen-erally oligoclonal (5 cases). Intrathecal IgG synthesis wasfound in 11/13 patients. CSF IgG4 levels were only availablefor 4 patients, but they were significantly increased witha median value of 5.225 mg/dL—which is more than 10 timesthe upper limit of the normal range (0.32 mg/dL).

TreatmentTreatment strategies are presented for 53 patients in table 3.Nearly all patients received systemic steroids (48/53), ofwhich 20 received pulse steroid therapy. Relapse occurredin 16 cases and was less frequent in the “pulse therapy”group (13/23 patients vs 3/15 patients). Relapse occurredless frequently when the steroid dose was greater than 49mg/d (relapse rate was 52.6%, whereas it was 100% withlower dosages). In the same way, the occurrence of relapsewas lower when steroid therapy was followed over a longerperiod of time (100% relapse for less than 6 months ofsteroid therapy vs 42.9% when steroids were continued overa year). All the steroid treatment specific data are availablein table e-1, links.lww.com/NXI/A115.

4 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 131: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Twenty patients were treated surgically, mostly in cases of cere-bral involvement. Only 3 patients with systemic IgG4-RD re-quired surgical treatment. Relapse occurred less often when

surgery was done (33% vs 44% without surgery). Surgical treat-ment was in itself sufficient to achieve disease remission within 6months (range: 3–12 months) in 4 cases of isolated IgG4-HP.

Table 1 Demographic, clinical, histologic, and blood work-up data

Total IgG4-HP Systemic IgG4-HP Isolated IgG4-HP Non-specified IgG4-HP

n 60 n 18 n 25 n 17

Median age, ya 60 53 (19–82) 18 48 (32–70) 25 52 (35–71) 17 55 (19–82)

Sex ratio (M/F) 60 2 (40/20) 18 1.6 (11/7) 25 2.1 (17/8) 17 2.4 (12/5)

Single brain involvement 60 44 (73.3%) 18 13 (72.2%) 25 20 (80%) 17 11 (64.7%)

Single spine involvement 11 (18.3%) 1 (5.6%) 4 (16%) 6 (35.3%)

Both brain and spine involvement 5 (8.3%) 4 (22.2%) 1 (4%) 0

Headaches 55 36 (65.4%) 17 15 (88.2%) 25 15 (60%) 13 6 (46.2%)

Visual acuity lost 55 16 (29.1%) 17 6 (35.3%) 25 6 (24%) 13 4 (30.8%)

Cranial nerve palsy 55 16 (29.1%) 17 9 (52.9%) 25 6 (24%) 13 1 (7.7%)

Epilepsy 55 7 (12.7%) 17 2 (11.8%) 25 5 (20%) 13 0

Motor weakness (limbs) 55 16 (29.1%) 17 3 (17.6%) 25 6 (24%) 13 7 (53.8%)

Sensory loss (limbs) 51 9 (16.4%) 17 2 (11.8%) 24 3 (12.5%) 10 4 (40%)

Serum IgG4 > 135 mg/L 36 20 (55.6%) 12 9 (75%) 19 7 (36.8%) 5 4 (80%)

Median serum IgG4 (mg/L)a 29 155 (40.1–939) 11 165 (54–260) 15 97 (47–939) 3 162 (155–182)

Blood inflammatory syndrome 27 10 (37%) 7 5 (71.4%) 12 3 (25%) 8 2 (25%)

Eosinophils >500/mm3 22 1 (4.5%) 4 1 (25%) 11 0 7 0

Median IgG4+/IgG ratioa 35 50.5 (40–94) 9 49.5 (40–70) 15 58 (40–94) 11 50 (40–73.3)

Median IgG4+/HPFa 39 50 (10–310) 7 51.7 (10–310) 18 63 (10–200) 14 40.05 (10–140)

Abbreviations: HPF = high-power field; IgG4 = immunoglobulin G4; IgG4-HP = IgG 4-related hypertrophic pachymeningitis; n = number of patients withavailable data.a Range.

Figure 2 Classification of IgG4-HP

a–cOther locations of IgG4-RD involvement by patients: a1. Orbital pseudotumor, kidney and lung. 2. Pancreas. 3. Lymph nodes, submandibular glands andlung. 4. Retroperitoneal fibrosis. 5. Mastoid. 6. Retroperitoneal fibrosis. 7. Aortitis. 8. Episcleritis. 9. Hypophysitis. 10. Lacrymal and submandibular glands. 11. Orbitalpseudotumor. 12. Maxillary pseudotumor. 13. Episcleritis. b1. Orbitory pseudotumor. 2. Pulmonary pseudo-tumor and hypophysitis. 3. Orbitary pseudotumor. 4.Retroperitoneal fibrosis. c1. Submandibular gland. IgG4-HP = IgG 4-related hypertrophic pachymeningitis; IgG4-RD = IgG 4-related disease; NS = non-specified.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 5

Page 132: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Twenty-five patients received an immunosuppressive drug,either as a steroid sparing agent (11/25) or after failure toachieve disease control (19/25). RTX was the most usedimmunosuppressant as first-line (10/25 patients) and second-line therapy (6/7 patients). The clinical response was repor-ted for all patients with the first immunosuppressive therapy.The clinical response data for patients who received morethan 1 immunosuppressant were available for 7 of the 8patients. The overall clinical remission was obtained for 24patients; 11 had isolated IgG4-HP, 11 had systemic IgG4-RD,and 2 had non-specified IgG4-HP. RTX (14 patients) andcyclophosphamide (CYC) (5 patients) were associated withthe best response rate in comparison to other drugs. Onepatient died of an infection with RTX therapy.

DiscussionIgG4-RD has been described as a multisystemic, fibro-inflammatory, and lymphoproliferative disease that is char-acterized as a polyclonal IgG4-infiltration, leading to storiformfibrosis and oblitering phlebitis.1 Histopathologic studies areessential for diagnosis and allow infectious, neoplastic, and/orother inflammatory diseases to be eliminated.

The 2011 Japan College of Rheumatology classification criteriafor IgG4-RD are the most commonly used.8 Serum IgG4 levelsare not specific markers for IgG4-RD, as they also increase in

cancerous, allergic, and autoimmune disorders.6,9,10 The speci-ficity and positive predictive value of the serum IgG4 for IgG4-RD are 60 and 34% respectively.11

Meningeal involvement of IgG4-RD is rare and was first de-scribed in 2009.12 Its prevalence is estimated to be slightlyabove 2% of overall clinical manifestations, based on datafrom recent series.5,6,13–15

To the best of our knowledge, we have presented the first caseof an arteriovenous dural fistula in IgG4-HP, suggesting itcould develop after a local precipitating factor, if everidentified.

Synchronous pachymeningitis of the brain and spinal cord isless common. Clinical manifestations of the disease are sub-ject to localization or complications due to compression and/or infiltration of neurologic and/or vascular adjacentstructures.

IgG4-HP is isolated in more than half of the overall casescontrasting with the preferential multisystemic involvementreported by other authors.13,14,16

Serum IgG4 levels were increased in 55.5% of the studiedpopulation. Stone et al. and Zhang et al. identified a preva-lence rate ranging from 70% to 94.1%, respectively.1,5 How-ever, using the serum IgG4 level as a diagnostic criteria for

Table 2 Cerebrospinal fluid data

Total IgG4-HP Systemic IgG4-HP Isolated IgG4-HP Non-specified IgG4-HP

n 60 n 18 n 25 n 17

Protein >0,4 G/L 28 21 (75%) 9 7 (77.8%) 15 11 (73.3%) 4 3 (75%)

Median protein level (g/L)a 20 0.925 (0.22–4.17) 7 0.83 (0.44–1.68) 11 0.99 (0.22–4.17) 2 0.72 (0.44–1)

WBC >5/mm3 30 18 (60%) 11 7 (63.6%) 15 7 (46.7%) 4 4 (100%)

Median WBC (/mm3)a 20 19.5 (0–150) 8 23.5 (1–112) 9 14 (0–150) 3 17 (10–21)

Median lymphocytes count (%)a 18 91.5 (0–100) 8 91.5 (85–100) 8 94.5 (0–100) 2 65.5 (51–80)

IgG >8 mg/dL 13 11 (84.6%) 3 3 (100%) 8 6 (75%) 2 2 (100%)

Median IgG count (mg/dL)a 10 14.3 (6.29–38.1) 3 14.3 (8.7–18.8) 5 13.03 (6.29–38.1) 2 23.8 (13.3–34.3)

Median IgG indexa 6 0.965 (0.53–1.84) 2 1.82 (1.8–1.84) 3 0.63 (0.53–0.84) 1 1.09

Profile 8 Oligoclonal (5) 2 Oligoclonal (2) 5 Oligoclonal (2) 1 Oligoclonal (1)

Transudate (2) Transudate (2)

Normal (1) Normal (1)

Oligoclonal bands 10 7 (70%) 4 3 (75%) 6 4 (66.7%) 0 —

IgG4 > 0.32 mg/dL 4 4 (100%) 2 2 (100%) 2 2 (100%) 0 —

Median IgG4 (mg/dL)a 4 5.225 (4.94–8.5) 2 5.05 (4.94–5.1) 2 6.925 (5.35–8.5) 0 —

Abbreviations: IgG4 = immunoglobulin G4; IgG4-HP = IgG 4-related hypertrophic pachymeningitis; n = number of patients with available data; WBC: whiteblood cells.a Range.

6 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 133: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Table 3 Treatment and relapse characteristics

Total IgG4-HP Systemic IgG4-HP Isolated IgG4-HP Non-specified IgG4-HP

n 60 n 18 n 25 n 17

Surgical excision 53 20 (37.7%) 16 3 (18.8%) 25 10 (40%) 12 7 (58.3%)

Steroid therapy 53 48 (90.6%) 16 16 (100%) 25 21 (84%) 12 11 (91.7%)

Steroid efficacy 53 41 (77.4%) 15 14 (93.3%) 22 17 (77.3%) 10 9 (90%)

IV pulse steroid therapy 39 20 (41.7%) 12 8 (66.7%) 23 10 (43.5%) 4 2 (50%)

IV pulse steroid efficacy 20 15 (75%) 8 7 (87.5%) 10 6 (60%) 4 2 (50%)

Relapse after steroids 38 16 (42.1%) 11 5 (45.4%) 19 8 (42.1%) 8 3 (37.5%)

Relapse after IV pulse steroids 15 3 (20%) 7 2 (28.6%) 6 1 (16.7%) 2 0

IS 49 25 (51%) 15 11 (73.3%) 22 12 (54.5%) 12 2 (16.7%)

Efficacy with first line IS 25 17 (68%) 11 (4 RTX, 3 AZA, 2 MTX, 1 CYC, 1 CSA) (5 RTX, 4 MTX, 2 CYC, 1 MMF) (1 RTX, 1 CYC)

9 (81.8%) 6 (50%)

(4 RTX, 2 AZA, 1 MTX, 1 CYC, 1 CSA) 12 (3 RTX, 2 CYC, 1 MTX) 2 2 (100%)

Efficacy with further line IS 8 7 (87.5%) 2 2 (100%) 5 5 (100%) — —

(2 RTX) (3 RTX, 1 itRTX, 1 CYC)

Abbreviations: CSA = cyclosporine A; CYC = cyclophosphamide; IgG4 = immunoglobulin G4; IgG4-HP = IgG 4-related hypertrophic pachymeningitis; IS = immunosuppressants; itRTX = intrathecal rituximab; MMF =mycophenolate mofetil; MTX = methotrexate; n = number of patients with available data; RTX = rituximab.

Neurolo

gy.org/N

NNeu

rology:N

euroimmunology

&Neuroinflam

mation

|Volum

e6,N

umber

4|

July2019

7

Page 134: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

IgG4-RD remains debatable because IgG4 levels are closelyrelated to disease activity and the extent of organdamage.6,10,17 It appears that IgG4-HP is infrequently asso-ciated with systemic disease and explains normal-to-low se-rum IgG4 levels and the lack of inflammatory markers. This isfurther illustrated by a recent study showing that an increasein serum IgG4 is usually associated with lymph node in-volvement, pancreatic, biliary, and/or ENT features.18 On theother hand, skin, prostate, and/or retroperitoneum in-volvement is associated with lower levels of serum IgG4.18 Assupported by our review, IgG4-HP is also associated withlower IgG4 plasmatic levels, inflammatory syndrome, eosin-ophilia, and complement consumption. It is thus essential toascertain extra-neurologic organ-involvement in patients withIgG4-HP presenting with high serum IgG4 levels and elevatedplasma inflammatory markers. In such circumstances, PET-CT has been shown to be of diagnostic value.19–21

CSF alone is not a reliable parameter for the diagnosis ofIgG4-HP. It is, however, common to find lymphocytic

meningitis and a moderate increase in CSF proteins withintrathecal synthesis of IgG with oligoclonal bands. Further-more, Della-Corte et al.22 reported that an increase in in-trathecal IgG4 levels could be a surrogate marker forneurologic IgG4-RD when tissue biopsies could not beperformed.

IgG4-HP closely follows granulomatosis with polyangeitis(GPA) as a leading cause of inflammatory HP.23 Other di-agnoses to consider are lymphoma, sarcoidosis, tuberculosis,rheumatoid arthritis and Langerhans-cell histiocytosis.24,25

Interestingly, increases in plasma IgG4 levels have also beenobserved in cases of GPA. Therefore, histopathologic studieshave to be an integral part of the diagnostic workup. Tocomplicate matters, overlapping syndromes of IgG4-RD andantineutrophil cytoplasmic antibodies-related vasculitideshave been described in a French cohort of 18 cases.26

There is no specific treatment protocol for meningeal forms ofIgG4-RD. Steroids remain the first-line treatment,27 though

Figure 3 Treatment algorithm for IgG4-HP

* Oral steroid therapy should be reduced over a period of 1-year minimum. AZA = azathioprine; CYC = cyclophosphamide; IgG4-HP = IgG 4-relatedhypertrophic pachymeningitis; MMF = mycofenolate mofetil; MTX = methotrexate.

8 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 135: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

several other studies also report the efficacy of immunosup-pressive agents such as RTX.28,29 Short-term response tosurgery in single-organ IgG4-HP is good and, in some cases,may be sufficient to achieve remission. Systemic steroids areprescribed for most patients and appear to reduce relapse rate.Patients with early relapse—most often than not—had notreceived steroid pulse therapy. In addition, relapse is lessfrequent when oral steroids are prescribed at a minimal doseof 50 mg/d and are maintained for a minimum of one yearperiod. Because steroids should be continued over a year,immunosuppressive drug use as steroid sparing agent shouldbe considered early on. Though used in half of the patients,immunosuppressive therapy is best suited to multisystemorgan-involvement. The clinical response rates are particu-larly good in patients treated with RTX or CYC. One reporteven described the efficacy of intrathecal RTX in the event oftreatment failure from other immunosuppressors.30 Car-ruthers et al.31 published in 2015, a prospective clinical trial of30 patients, studying the efficacy of RTX in IgG4-RD. Re-mission, obtained in 77% of patients, was calculated froma combined score, the Responding Index. The data of ourreview concord with the literature designating RTX as theimmunosuppressive drug of choice for IgG4-RD andIgG4-HP.

The disease-activity score or Responder Index (RI) has beenproposed by some authors.31 Patients were categorized intohigh- or low-risk groups for relapse based on serum IgG4levels, the extent of organ involvement, and organ failure.IgG4-HP is rarely associated with other organ-involvement,but it is known to relapse despite steroid therapy in nearly halfof cases, regardless of the type of meningeal involvement. TheRI does not seem to reflect this risk of relapse in cases ofisolated IgG4-HP. Because systemic involvement is less fre-quent in IgG4-HP, serum markers such as CD19+, CD27+,CD20−,and CD38(hi) cells, IgG4 and IgE levels,32–34 may beless reliable than in IgG4-RD with multi-organ involvementfor predicting relapse. On the other hand, CSF IgG4 levelsmay better reflect disease activity in IgG4-HP but would alsorequire regular spinal taps—limiting its use to specific sit-uations (i.e., treatment resistance, initiation of immunosup-pressive therapy).

Based on our experience and the available data from previousstudies, a treatment algorithm for IgG4-HP is presented infigure 3.

ConclusionIgG4-HP is an uncommon form of IgG4-RD and is one of theleading causes of meningeal inflammatory disease. It is mainlycharacterized by the lack of extra-neurologic organ-involvement and systemic signs. Histopathologic studiesshould, when possible, be performed as it is crucial for di-agnosis because serum markers are rarely informative.Treatment strategies rely on early surgery, steroids, and early

immunosuppressive therapy (such as RTX) to prevent neu-rologic complications.

Study fundingNo targeted funding.

DisclosureM. Levraut reports no disclosures;M. Cohen received honorariafor advisory board participation from Biogen, Novartis, Rocheand Ad Scientam; S. Bresch reports no disclosures; C. Giordanareceived honoraria fromAbbvie; F. Burel-Vandenbos reports nodisclosures; J. Sedat reports no disclosures; D. Fontaine isa consultant for Medtronic, St Jude-Abott, Autonomic Tech-nologies, Renishaw; V. Bourg reports no disclosures; N. Martisreports no disclosures; C. Lebrun-Frenay received honoraria forexpertise and presentations at meetings from Teva, Biogen,Merck, Roche, Novartis. Go to Neurology.org/NN for fulldisclosures.

Publication historyReceived byNeurology: Neuroimmunology & Neuroinflammation January24, 2019. Accepted in final form March 12, 2019.

Appendix Authors

Name Location Role Contribution

MichaelLevraut, MD

Universite Coted’Azur, Hopitall’Archet 1

Author Major role in acquisition ofdata; drafted themanuscript

MikaelCohen, MD

Universite Coted’Azur, HopitalPasteur 2

Author Revised themanuscript forintellectual content

SaskiaBresch, MD

Universite Coted’Azur, HopitalPasteur 2

Author Revised themanuscript forintellectual content

CarolineGiordana,MD

Universite Coted’Azur, HopitalPasteur 2

Author Revised themanuscript forintellectual content

Fanny Burel-Vandenbos,MD, PhD

Universite Coted’Azur, HopitalPasteur 1

Author Interpreted the data,revised the manuscript forintellectual content

LydianeMondot, MD

Universite Coted’Azur, HopitalPasteur 2

Author Revised themanuscript forintellectual content

JacquesSedat, MD

Universite Coted’Azur, HopitalPasteur 2

Author Revised themanuscript forintellectual content

DenysFontaine,MD, PhD

Universite Coted’Azur, HopitalPasteur 2

Author Revised themanuscript forintellectual content

VeroniqueBourg, MD

Universite Coted’Azur, HopitalPasteur 2

Author Revised themanuscript forintellectual content

Nihal Martis,MD

Universite Coted’Azur, Hopitall’Archet 1

Author Designed andconceptualized the study;revised the manuscript forintellectual content

Continued

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 9

Page 136: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

References1. Stone JH, Zen Y, Deshpande V. IgG4-related disease. N Engl J Med 2012;366:

539–551.2. Deshpande V, Zen Y, Chan JK, et al. Consensus statement on the pathology of IgG4-

related disease. Mod Pathol 2012;25:1181–1192.3. Kim EC, Lee SJ, Hwang HS, Kim J, Kim MS. Bilateral diffuse scleritis as a first

manifestation of immunoglobulin G4–related sclerosing pachymeningitis. Can JOphthalmol 2013;48:e31–e33.

4. Wallace ZS, Carruthers MN, Khosroshahi A, et al. IgG4-related disease and hyper-trophic pachymeningitis. Medicine (Baltimore) 2013;92:206–216.

5. Zhang PP, Zhao JZ, Wang M, et al. The clinical characteristics of 346 patientswith IgG4-related disease [in Chinese]. Zhonghua Nei Ke Za Zhi 2017;56:644–649.

6. Wallace ZS, Deshpande V, Mattoo H, et al. IgG4-Related disease: clinical and labo-ratory features in one hundred twenty-five patients. Arthritis Rheumatol 2015;67:2466–2475.

7. Lindstrom KM, Cousar JB, Lopes MBS. IgG4-related meningeal disease: clinico-pathological features and proposal for diagnostic criteria. Acta Neuropathol 2010;120:765–776.

8. Umehara H, Okazaki K, Masaki Y, et al. Comprehensive diagnostic criteria for IgG4-related disease (IgG4-RD), 2011. Mod Rheumatol 2012;22:21–30.

9. Su Y, Sun W, Wang C, et al. Detection of serum IgG4 levels in patients with IgG4-related disease and other disorders. PLoS One 2015;10:e0124233.

10. Culver EL, Sadler R, Simpson D, et al. Elevated serum IgG4 levels in diagnosis,treatment response, organ involvement, and relapse in a prospective IgG4-relateddisease UK cohort. Am J Gastroenterol 2016;111:733–743.

11. Carruthers MN, Khosroshahi A, Augustin T, Deshpande V, Stone JH. The diagnosticutility of serum IgG4 concentrations in IgG4-related disease. Ann Rheum Dis 2015;74:14–18.

12. Chan S-K, Cheuk W, Chan K-T, Chan JKC. IgG4-related sclerosing pachymeningitis.Am J Surg Pathol 2009;33:1249–1252.

13. Sekiguchi H, Horie R, Kanai M, Suzuki R, Yi ES, Ryu JH. IgG4-Related disease:retrospective analysis of one hundred sixty-six patients. Arthritis Rheumatol 2016;68:2290–2299.

14. Inoue D, Yoshida K, Yoneda N, et al. IgG4-related disease: dataset of 235 consecutivepatients. Medicine (Baltimore) 2015;94:e680.

15. Lin W, Lu S, Chen H, et al. Clinical characteristics of immunoglobulin G4-related disease:a prospective study of 118 Chinese patients. Rheumatology (Oxford) 2015;54:1982–1990.

16. Chen Y, Zhao JZ, Feng RE, et al. Types of organ involvement in patients withimmunoglobulin G4-related disease. Chin Med J 2016;129:1525–1532.

17. Tabata T, Kamisawa T, Takuma K, et al. Serial changes of elevated serum IgG4 levelsin IgG4-related systemic disease. Intern Med 2011;50:69–75.

18. Umehara H, Okazaki K, Nakamura T, et al. Current approach to the diagnosis ofIgG4-related disease–combination of comprehensive diagnostic and organ-specificcriteria. Mod Rheumatol 2017;27:381–391.

19. Takahashi H, Yamashita H, Morooka M, et al. The utility of FDG-PET/CT and otherimaging techniques in the evaluation of IgG4-related disease. Joint Bone Spine 2014;81:331–336.

20. Lee J, Hyun SH, Kim S, et al. Utility of FDG PET/CT for differential diagnosis ofpatients clinically suspected of IgG4-related disease. Clin NuclMed 2016;41:e237–e243.

21. Zhang J, Chen H, Ma Y, et al. Characterizing IgG4-related disease with 18F-FDGPET/CT: a prospective cohort study. Eur J Nucl Med Mol Imaging 2014;41:1624–1634.

22. Della-Torre E, Galli L, Franciotta D, et al. Diagnostic value of IgG4 indices in IgG4-related hypertrophic pachymeningitis. J Neuroimmunol 2014;266:82–86.

23. Yonekawa T, Murai H, Utsuki S, et al. A nationwide survey of hypertrophic pachy-meningitis in Japan. J Neurol Neurosurg Psychiatry 2014;85:732–739.

24. AbdelRazek MA, Venna N, Stone JH. IgG4-related disease of the central and pe-ripheral nervous systems. Lancet Neurol 2018;17:183–192.

25. Kamisawa T, Zen Y, Pillai S, Stone JH. IgG4-related disease. Lancet 2015;385:1460–1471.

26. Danlos F-X, Rossi GM, Blockmans D, et al. Antineutrophil cytoplasmic antibody-associated vasculitides and IgG4-related disease: a new overlap syndrome. Auto-immun Rev 2017;16:1036–1043.

27. Khosroshahi A, Wallace ZS, Crowe JL, et al. International consensus guidancestatement on the management and treatment of IgG4-related disease. ArthritisRheumatol 2015;67:1688–1699.

28. Ebbo M, Grados A, Samson M, et al. Long-term efficacy and safety of rituximab inIgG4-related disease: data from a French nationwide study of thirty-three patients.PLoS One 2017;12:e0183844.

29. Carruthers MN, Topazian MD, Khosroshahi A, et al. Rituximab for IgG4-relateddisease: a prospective, open-label trial. Ann Rheum Dis 2015;74:1171–1177.

30. Della-Torre E, Campochiaro C, Cassione EB, et al. Intrathecal rituximab for IgG4-related hypertrophic pachymeningitis. J Neurol Neurosurg Psychiatry 2018;89:441–444.

31. Carruthers MN, Stone JH, Deshpande V, Khosroshahi A. Development of an IgG4-RD responder index. Int J Rheumatol 2012;2012:259408.

32. Wallace ZS, Mattoo H, Carruthers M, et al. Plasmablasts as a biomarker for IgG4-related disease, independent of serum IgG4 concentrations. Ann RheumDis 2015;74:190–195.

33. Mattoo H, Mahajan VS, Della-Torre E, et al. De novo oligoclonal expansions ofcirculating plasmablasts in active and relapsing IgG4-related disease. J Allergy ClinImmunol 2014;134:679–687.

34. Wallace ZS, Mattoo H,Mahajan VS, et al. Predictors of disease relapse in IgG4-relateddisease following rituximab. Rheumatology (Oxford) 2016;55:1000–1008.

Appendix (continued)

Name Location Role Contribution

ChristineLebrun-Frenay, MD,PhD

Universite Coted’Azur, HopitalPasteur 2

Author Designed andconceptualized the study;revised the manuscript

10 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 137: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

DIAGNOSTIC AND TREATMENT CHALLENGES OPEN ACCESS

Cataclysmically disseminating neurologicpresentation in an immunosuppressed lupuspatientFrom the National Multiple Sclerosis Society Case Conference Proceedings

Christopher M. Perrone, MD, Robert P. Lisak, MD,* Ethan I. Meltzer, MD,*‡ Peter Sguigna, MD,

Etsegenet Tizazu, MD, MS, Dina Jacobs, MD, Esther Melamed, MD, PhD,* Ashlea Lucas, PA-C,*

Leorah Freeman, MD, PhD,* Gabriel Pardo, MD,* Andrew Goodman, MD,* Edward J. Fox, MD, PhD,*

Kathleen Costello, MS, ANP-BC,*§ Matthew S. Parsons, PhD, Scott S. Zamvil, MD, PhD,*†

Elliot M. Frohman, MD, PhD,*† and Teresa C. Frohman, MPAS, PA-C*†§

Neurol Neuroimmunol Neuroinflamm 2019;6:e582. doi:10.1212/NXI.0000000000000582

Correspondence

Dr. Frohman

[email protected].

Case presentationA 54-year-old woman presented with complaints of myalgias, fatigue, and progressive legweakness of several weeks duration, followed by acute loss of vision in her left eye. Her medicalhistory was notable for long-standing systemic lupus erythematosus (SLE) that was activelytreated with prednisone, mycophenolate mofetil, and quinacrine.

Initially, it was thought she was exhibiting evidence of an exacerbation in her SLE and wastreated with IV methylprednisolone in conjunction with plasma exchange. Despite a week ofcontinuous treatment, she developed acute vision loss in her right eye, hearing changes, anddifficulty swallowing. Given the progression of symptoms, she was transferred to the Hospital ofthe University of Pennsylvania for further assessment.

On admission, her neurologic examination demonstrated normal mental status. The cranialnerve examination was most notable for absent light perception bilaterally, nonreactive pupils,severe diffuse ophthalmoparesis, loss of facial sensation in the left V1 and V2 trigeminaldistribution, and diminished hearing to finger rubbing on the right and right palate deviation.Motor deficits included mild bilateral upper extremity weakness and paraplegia. Deep tendonreflexes were pathologically exaggerated in the upper extremities, but absent in the lowerextremities. Vibratory sensation was decreased throughout all 4 limbs, whereas pinprick sen-sation was reduced in the lower limbs in association with a sensory level at T10.

Blood studies were notable for lymphopenia (absolute lymphocyte count 200), positive double-stranded DNA antibody, and mildly reduced C3 and C4 complement levels. MRI of the brain(figure 1, A–D) demonstrated multifocal nonenhancing parenchymal lesions and patchy,noncontinuous enhancements of multiple cranial nerves (optic chiasm, left III, bilateral V, and

*National Multiple Sclerosis Society Case Conference Proceedings Faculty.

†Guest Editors-in-Chief: Teresa C. Frohman, MPAS, PA-C, Elliot M. Frohman, MD, PhD, and Scott S. Zamvil, MD, PhD.

‡Guest Assistant Editor: Ethan I. Meltzer, MD.

§Guest Managing Editors: Kathy Costello, MS, ANP-BC and Teresa Frohman, MPAS, PA-C.

From the Hospital of the University of Pennsylvania (C.M.P., E.T., D.J., ), Philadelphia, PA; Department of Neurology (R.P.L.), Wayne State University, Detroit, MI; Department ofNeurology (E. I. Meltzer, E. Melamed, A.L., L.F., E.J.F.), Dell Medical School at the University of Texas at Austin, TX; Department of Neurology (P.S.), MS Fellowship Training Program, UTSouthwestern School of Medicine, Dallas, TX; Oklahoma Medical Research Foundation (G.P.), Oklahoma City, OK; Department of Neurology (A.G.), University of Rochester, NY;Central Texas Neurology Consultants, and Department of Neurology (E.J.F.), Dell Medical School at the University of Texas at Austin, TX; The National Multiple Sclerosis Society (K.C.),New York, NY; Yerkes National Primate Research Center (M.S.P.), Emory University, Atlanta, GA; Department of Neurology and Program in Immunology (S.S.Z.), University ofCalifornia San Francisco, San Francisco, CA; and Departments of Neurology and Ophthalmology (E.M.F., T.C.F.), Dell Medical School at the University of Texas at Austin, TX.

Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by the National Multiple Sclerosis Society.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloadingand sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1

Page 138: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

right VIII), in conjunction with leptomeningeal enhancement.MRI of the spine (figure 1, E–H) was notable for T2 signalprolongation, spanning the level from T2 to the conusmedullaris. The enhancement pattern was restricted to thecaudal aspect of the spinal cord.

CSF showed xanthochromia, 44 white blood cell count/mm3

(61% lymphocytes, 34% neutrophils, and 5% monocytes),elevated total protein (115 mg/dL), and 6 CSF-restrictedoligoclonal bands.

Ophthalmologic consultation revealed bilateral disseminatedretinal nerve fiber layer lesions, most consistent with cottonwool spots, in conjunction with multifocal retinal hemor-rhages. Taken together, the evidence was most consistentwith severe, bilateral, outer retinal necrosis (figure 2).1

Differential diagnosisThis patient presented with a disseminated and pathologiccataclysmic syndrome that included subacute paraplegia (thederivative of a centrally predominant and longitudinally ex-tensive, edematous, transverse myelitis [LETM] with multi-ple areas of patchy focal enhancement on MRI), bilateralcranial neuropathies, and bilateral optic nerve involvementwith severe retinal necrosis.

The combination of these findings listed in the precedingparagraph is suggestive of a disease process capable of rapidlymoving across the rostral-caudal extent of the neuraxis; witha meningitic infectious process being the most likely etiology(table). Notwithstanding this likelihood, in a patient such asours, SLE is often posited as a chief diagnostic considerationwhen characteristics of a diffuse vasculitis are present. Nev-ertheless, an important and highly salient “pitfall principle” is

that a bona-fide lupus-related vasculitis is truly exceptional.2

Alternately, SLE can certainly account for a diffuse vascul-opathy, with semiologic features typically including head-aches, seizures, psychiatric manifestations, encephalopathy,and stroke like symptoms; all of which were lacking in ourpatient.3 Furthermore, spinal cord and root involvement inSLE is extraordinarily rare.4,5

As our patient already had 1 autoimmune disorder, disordersof immune dysregulation such as neuromyelitis optica(NMO) spectrum disorder, Sjogren syndrome, neuro-sarcoidosis, and Susac syndrome were considered in the dif-ferential (see below).6,7

NMO can be a legitimate overlap disorder and can accountfor a constellation of neurologic deficits, such as opticneuropathy, retinal vascular disturbances, multifocalbrainstem processes, and the LETM, which was docu-mented in our patient. However, the extra-axial patchyenhancement of multiple cranial nerves and the multiplicityof them would be highly exceptional disease characteristics,thereby making NMO a highly untenable diagnosticconsideration.

Sjogren syndrome can produce myelitis, even the longitudinalextensive type, which is characteristic of NMO. Alternately,the diffuse inflammatory constellation of targeted structuresacross the rostral-caudal extent would be highly unusual, andour patient had none of the features of keratoconjunctivalsicca that is typical for Sjogren syndrome.

Neurosarcoidosis must be a prime consideration in a patientwith evidence of CSF pleocytosis and a clinical course, whichlocalizes to, even if only in part, a basilar meningitis.

Figure 1 MRIs

(A and B) Axial FLAIR images of the brain demon-strate multifocal parenchymal lesions includingthe right hippocampus, right midbrain, left tem-poral-occipital and left anterior temporal areas,and the left thalamus (arrow heads). There is alsoevidence of intraventricular debris with a fluidlevel in the bilateral occipital horns (C, arrowheads). Although there were some confluentchanges to the parenchyma in a periventriculardistribution, other lesions appeared more lep-tomeningeal (C, arrow head pointing toward a lin-ear hyperintensity, which on closer inspection islocalized to the leptomeningeal zone of a left pa-rietal gyrus lesion). (D) T1 post-gadolinium imagesdemonstrated enhancement of the optic chiasm,several other cranial nerves, and the tentorium(arrow heads). (E and F) Sagittal and axial T2-weighted images of the thoracic spine reveala longitudinally extensive, centrally predominanthyperintensity (arrow heads, seen in NMO), al-though enhancement was not observed. We fur-ther illustrate the caudal extent of the lesion,which clearly involves the conus and the exitingnerve roots of the cauda equine, both of whichdemonstrate enhancement on T1 post-gadoli-nium images (G andH, respectively; arrow heads).FLAIR = fluid attenuation inversion recovery.

2 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 139: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Susac syndrome could affect both retina and hearing;however, the retinal pathology is distinctive in producingbranch retinal artery occlusions with “cloud-like” abnor-malities, often in conjunction with intravascular Gass pla-ques, whereas the pathology identified in our patient wasinstead most consistent with a severe retinal necrotic process,most likely secondary to an infectious etiology.7,8 Furthermore,the presence of a concomitant and longitudinally extensivemyelitis and the absence of callosal lesions make this diagnosisunlikely.7

Although the diffuse nature of the disease process raises con-sideration of malignant disorders, the rapid evolution ofsymptoms could be considered atypical.9 The rapid andbroadly disseminated involvement of the rostral to caudal ex-tent of the neuroaxis should arouse suspicion for an infectiousdisorder, which includes a basilar meningitis, particularly giventhe multifocal and bilateral involvement of cranial nerves.

Pathoetiologic infectious agents that can cause a basilarmeningitic process with a propensity for spread via CSF andleptomeningeal surfaces include tuberculosis, listeriosis, lep-tospirosis, mycoplasma, enteroviruses, syphilis, HIV, humanT cell lymphoma virus I/II, members of the herpes family ofviruses, and a number of unusual viral agents such as WestNile and Zika.10

Given our patient’s known immune suppression, infiltrationof the subarachnoid space by a viral infection could produce

the rapid, diffuse, and widespread involvement of multipleareas of the CNS. Several specific microbial infectious agentsincludingMycoplasma, enterovirus, and herpes viruses such ascytomegalovirus (CMV), varicella zoster virus (VZV), andherpes simplex virus can all cause cranial neuropathies andmyelitis.11,12

Narrowing the differential further, progressive outer retinalnecrosis is most commonly seen as a reactivation of a latentherpetic infection in immunocompromised patients.8

These findings made investigation for a basilar meningitisa priority.

Final diagnosisVZV infection resulting in oculo-meningo-encephalo-myelo-radiculitis was confirmed by positive VZV PCR testing in theCSF. Treatment was initiated urgently with 14 days of IVacyclovir.

After 3 months, the patient regained the ability to swallow,recovered full strength in both upper extremities, and hadpartial motor improvement of both lower extremities. Brainand spinal cord imaging abnormalities nearly resolved, but sheremains blind bilaterally. The presence of fulminant outerzonal retinal necrosis, accompanied by disseminated cottonwool spots and retinal nerve fiber hemorrhages at pre-sentation, reflects the massive degree of destruction targetingthe retinal ganglion cell neurons and the axonal fiber pathwaysof the optic nerve.

The pathology affecting the eyes in this patient could notseriously be characterized as an optic neuropathy, but ratherwas a fulminant and rapidly evolving infectious processeventually leading to vascular occlusions, ischemia, hemor-rhage, and ultimately necrosis. Such cases require promptrecognition and treatment with antimicrobial therapy, po-tentially in conjunction with steroids, to mitigate the oftenassociated inflammatory burden that can contribute to tissuedamage in conjunction with the infectious process itself.8,13

Furthermore, severe consequences can follow from the acutechanges in pressure dynamics within an anatomically re-stricted compartment such as the eye.

In the context of severe edema, optic nerve sheath de-compression needs to be considered urgently with the objectiveof preventing or mitigating a compartment syndrome, wherethe rapid intensification of pressure can lead to both vascularstasis and thereby attenuation or even abolishment of retinalperfusion and interrupted axonal transport in the retinal gan-glion cell fibers that become the optic nerve. By analogy, ful-minant spinal cord edema can result in extensive vasculardeficits across the spinal cord when edema intensifies andexceeds the space it resides in, or it interrupts the circum-ferentially organized vaso corona, a collateral vascular supply tothe spinal cord, which can be compressed and thereby fail toprovide rescue perfusion in the setting of severe swelling, es-pecially when there is restricted spinal canal compliance.

Figure 2 Fundus of a patient with progressive outer retinalnecrosis from VZV infection

The virus replicates rapidly and spreads through the outer layer of theretina, producing necrotic (white) coalescing lesions and intraretinalhemorrhages. The vitreous typically does not reflect signs of intraocularinflammation, likely due to an immunosuppressed state. Relative pres-ervation of the retinal vasculature also helps to differentiate progressiveouter retinal necrosis from retinal artery occlusions. VZV = varicella zostervirus.

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 3

Page 140: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

DiscussionEstimating infection riskVZV is a ubiquitous, neurotropic virus. Although primaryinfection has declined since the introduction of the live-attenuated vaccine (Zostavax), in 1995, the majority ofadults (and the patient in this case) acquired the infection inchildhood.14,15 Following the primary infection, VZVbecomes latent in the dorsal roots, cranial nerves, and au-tonomic ganglia throughout the neuroaxis. The virus is pri-marily contained by T-cell–mediated immunity for much ofadulthood.15,16

Certain risk factors are associated with virus reactivation.Advancing age (>50 years) is associated with immune se-nescence, which either primarily or predominantly impactsmemory T cells, and results in diminished immune pro-tection.17 While B-cell–derived antibodies to primary in-fection or vaccination may exist as surrogates for immunity,these do not necessarily convey protection as T-cell functiondeclines.17 In fact, it is estimated that 50% of individuals willexperience VZV reactivation by age 85 years. As a means toboost immunity in the setting of immune senescence, theCDC recommends recombinant shingles vaccine (Shingrix)for all immunocompetent patients aged 50 years and older.15

Our patient is older than 50 years and had not yet received theshingles vaccine. Moreover, immunosuppressive conditionsfrom disease states such as HIV or from medications (evi-denced by lymphopenia in this case) also increase the risk ofVZV reactivation.

Manifestations of VZV reactivationThe most common result of reactivation of VZV is herpeszoster, a dermatomal vesicular eruption that is often painful. Atrigeminal nerve ganglionitis (the majority affecting V1) ordorsal root ganglionitis in the cervical or thoracic regions ismost common. Importantly, a rash is not always a heraldingsign for viral reactivation, and there was no evidence of rashduring or preceding this patient’s hospitalization.14–16,18

VZV reactivation can result in a number of cranial neuropathies,producing optic neuritis, ophthalmoplegias, facial palsy, andpharyngeal and hypoglossal dysfunction. Within the peripheralnervous system, VZV reactivation can affect anterior and pos-terior spinal roots resulting in a radiculitis and paresis.12–16,18

Notwithstanding these observations, the majority ofpatients experience a meningoencephalitis, whereas VZVmyelitis is a rare complication.15 Most cases of VZVmyelitisoccur as a postinfectious, self-limited paraparesis. However,in the setting of immunosuppression, VZV can invade thespinal cord, causing progressive myelitis and longitudinallyextensive cord lesions, as observed in our patient.18 Ifwidespread parenchymal lesions are present in the brain, anassociation with VZV vasculopathy must also be consid-ered, particularly given the destructive nature of infection-related vasculitis.

Table Differential diagnosis for basilar meningitis

Classification Subtype Details

Bacterial Tuberculosis

Listeriosis

Brucellosis

Syphilis

Lyme

Nocardia

Actinomyces

Leptospirosis,aseptic

Aseptic CSF pattern

Mycoplasma,aseptic

Aseptic CSF pattern

Viral Enteroviruses

HIV

HTLV-I/II

Herpes viruses HSV-I/II, CMV, EBV, and VZV

West Nile

Zika

Fungal Cryptococcus

Coccidioidomycosis

Blastomycosis

Histoplasmosis

Aspergillosis

Mucormycosis Immunosuppressed anddiabetics most commonlyaffected

Noncaseatinggranulomatousinflammation

Sarcoidosis

Inflammatorydemyelinating

MS and NMO AQP-4 antibodies in NMO

Bickerstaffencephalitis

Fisher syndrome GQ1b antibodies

Paraneoplastic Classic onconeural antibodieswhich target intracellularepitopes include antibodiesagainst Hu, Yo, CRMP-5, Ma2, Ri,and amphiphysin

Non–onco-neogenic cellsurface antigen Ab

Limbic encephalitis Anti-NMDA/AMPA/Caspr-2/Gly

Leptomeningealcarcinomatosis

High volume and multiple CSFcytologies may be required toconfirm this diagnosis

Abbreviations: AMPA = alpha-amino-3-hydroxy-5-methyl-4-isoxazoleproprionicacid; AQP-4 = aquaporin protein-4; Caspr-2 = contactin-associatedprotein-like 2; CMV = cytomegalovirus; EBV = Epstein-Barr virus; Gly =glycine; HSV = herpes simplex virus; HTLV-I/II = human T cell lymphomavirus I/II; NMDA = N methyl D aspartate; NMO = neuromyelitis optica;VZV = varicella zoster virus.

4 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 141: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Following reactivation in the trigeminal and other cranialnerve ganglia, the virus spreads transaxonally via afferentfibers to the adventitia of intracranial blood vessels (andacross the spectrum from arterioles to large arteries) poten-tially resulting in (or causing in lieu of resulting) stroke.19

Ocular disease can be devastating in the immunosuppressedpatient, with VZV representing the most common cause ofsevere progressive outer retinal necrosis, whereby rapid rep-lication of the virus leads to focal areas of necrosis andintraretinal hemorrhage (figure 2).1 The retinal vasculatureand vitreous are often normal. In contrast to the more com-mon CMV retinitis, VZV-related lesions accumulate and co-alesce rapidly leading to blindness; making early recognitionimperative, as vision loss is often extensive and irreversible.8

Detection and treatment of VZV infectionAny concern for VZV affecting the CNS should prompt CSFanalysis, the current gold standard over more invasive brainbiopsy. In the first 7 days of symptoms, the best test fordiagnosis is VZV PCR. After several days, the sensitivity ofthe PCR starts to decrease, so the measurement of in-trathecal synthesis of VZV antibody (both immunoglobulinG and immunoglobulinM) in the CSF is recommended after7 days.20

Early treatment is paramount. If clinical suspicion is high(known immunosuppression and evidence on examination ofCNS involvement), empiric treatment is warranted. Thetreatment of choice is IV acyclovir. The duration of treatmentis 7 days in immunocompetent patients and 10–14 days inimmunocompromised patients.13,20 As in this case, there canbe significant recovery with treatment.

ConclusionThis case underscores the importance of consideringa broad differential diagnosis in patients with autoimmuneconditions that present with acute, multifocal and rapidlydisseminating neurologic manifestations, particularly whentraversing the rostral to caudal extent of the entire neuraxis.Such patients are truly at potential risk of cataclysmic in-fectious complications as a consequence of prolonged im-munosuppressive therapy. The rapid evolution of thispatient’s symptoms and several rare manifestations of VZVreactivation in the absence of rash highlight the importanceof early consideration, detection, and urgent treatmentintervention.

Author contributionsC. Perrone: concept, manuscript drafting, clinical and im-aging review, literature review, and critical revision of themanuscript. R. Lisak: conception and critical revision ofthe manuscript for intellectual content. E. I. Meltzer,P. Sguigna, E. Tizazu, D. Jacobs, E. Melamed, A. Lucas,

L. Freeman, G. Pardo, A. Goodman, E. J. Fox, K. Costello,and M. S. Parsons: critical revision of the manuscript forintellectual content. S. Zamvil and E. M. Frohman: con-ception and critical revision of the manuscript for in-tellectual content. T. C. Frohman: critical revision of themanuscript for intellectual content.

AcknowledgmentThe National Multiple Sclerosis Society Case ConferenceProceedings are both a live educational Webinar, in conjunc-tion with a peer review publishing initiative from the monthly“Diagnostic and Treatment Challenges in MS and Neuro-immunology Webinars,” sponsored by the National MultipleSclerosis Society Fellowship Training Program andDirected byElliot M. Frohman, MD, PhD, and Teresa C. Frohman, MPAS,MSCS, PA-C, from theDellMedical School at theUniversity ofTexas at Austin.

Study fundingDiagnostic and Treatment Challenges in MS and Neuro-immunology Webinars is sponsored by the National MultipleSclerosis Fellowship Training Program.

DisclosureC. Perrone received a clinical fellowship funded in part by theNMSS Institutional Clinical Training Award. R.P. Lisak hasbeen funded for research support by the NIH, NationalMultiple Sclerosis Society (USA), Mallinckrodt Pharma-ceuticals, Genentech, Teva Pharmaceuticals, Novartis, Medi-mmune, Ra Pharmaceuticals, Alexion, Argenx, and Chugai;served as a consultant to Novartis, Argenx, Syntimmune,Alexion, GLG Consulting, Alpha Sites, Insights Consulting,Informa Pharma Consulting, and Slingshot Consulting; andserved on the speaker’s bureau of Teva Pharmaceuticals(nonbranded talks only). E. I. Meltzer received speaker feesfrom Sanofi Genzyme. P. Sguigna and E. Tizazu report nodisclosures. D.A. Jacobs received consulting fees from Biogenand Sanofi Genzyme and lecture fees from EMD Serono. E.Melamed served as consultant for EMD Serono. A. Lucasreports no disclosures. L. Freeman served as consultant forCelgene and received honorarium from McGraw-Hill Edu-cation, speaker fees from the National Multiple SclerosisSociety, research funding from Race to Erase MS and PCORI,and program sponsorship from Biogen. G. Pardo serves asa member of the National Medical Advisory Council of theNational Multiple Sclerosis Society and has served as a con-sultant and/or been in the speaker bureau of Biogen, Celgene,Genentech, Sanofi Genzyme, EMD Serono, Novartis, andTeva Pharmaceuticals. A.D. Goodman received consultingfees during the past 2 years from Acorda, Adamas, AbbVie,EMD Serono, and Teva; his employer has received researchsupport for clinical trials from the following sponsors: Atara,Biogen, Roche, SanofiGenzyme, Novartis, Sun Pharma, andTeva. E. J. Fox received compensation for research, con-sulting, speakers’ bureau, and/or advisory work fromAcorda, Bayer, Biogen, Celgene, Chugai, EMD Serono,

Neurology.org/NN Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 5

Page 142: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

Genentech/Roche, Mallinckrodt, MedDay, Novartis,Sanofi Genzyme, Teva, and TG Therapeutics. K. Costelloand M. parson report no disclosures. S.S. Zamvil is DeputyEditor of Neurology: Neuroimmunology & Neuro-inflammation; is a member of the advisory board for theInternational Society of Neuroimmunology; has served asa consultant and received honoraria from Biogen Idec,EMD Serono, Genzyme, Novartis, Roche/Genentech, andTeva Pharmaceuticals, Inc.; has served or serves on DataSafety Monitoring Boards for Lilly, BioMS, Teva, andOpexa Therapeutics; and receives research grant supportfrom the NIH, the NMSS, The Maisin Foundation, Bio-gen, and Celgene. E. M. Frohman received speaker hon-oraria from Sanofi Genzyme, Novartis, Alexion, andAcorda. T. C. Frohman received consulting fees fromGenzyme and Acorda. Go to Neurology.org/NN for fulldisclosures.

Publication historyReceived by Neurology: Neuroimmunology & NeuroinflammationMay 5, 2019. Accepted in final form May 8, 2019.

References1. Bryan RG, Myers FL. Progressive outer retinal necrosis in a patient with rheumatoid

arthritis. Arch Ophthalmol 1998;116:1249.2. Clancy R, Marder G, Martin V, Belmont H, Abramson SB, Buon J. Circulating

activated endothelial cells in systemic lupus erythematous: further evidence for diffusevasculopathy. Arthritis Rheum 2001;44:1203–1208.

3. van Dam AP. Diagnosis and pathogenesis of CNS lupus. Rheumatol Int 1991;11:1–11.

4. Johnson RT, Richardson EP. The neurological manifestations of systemic lupuserythematosus. Medicine (Baltimore) 1968;47:337–369.

5. The American College of Rheumatology nomenclature and case definitions forneuropsychiatric lupus syndromes. Arthritis Rheum 1999;42:599–608.

6. Asgari N, Jarius S, Laustrup H, et al. Aquaporin-4-autoimmunity in patients withsystemic lupus erythematosus: a predominantly population-based study. Mult Scler2018;24:331–339.

7. Kleffner I, Dorr J, Ringelstein M, et al. Diagnostic criteria for Susac syndrome.J Neurol Neurosurg Psychiatry 2016;87:1287–1295.

8. Holland GN. The progressive outer retinal necrosis syndrome. Int Ophthalmol 1994;18:163–165.

9. Balbuena-Merle RI, Sante-Prez M, Perez-Berenguer J, Velez-Rosario R, Correa-RivasM, Jimenez A. Leptomeningeal carcinomatosis as the initial manifestation of meta-static disease diagnosed in post-mortem examination: a case series. PR Health Sci J2019;38:64–67.

10. Smith JE, Aksamit AJ, Jr. Outcome of chronic idiopathic meningitis. Mayo Clin Proc1994;69:548–556.

11. Jacob A, Weinshenker BG. An approach to the diagnosis of acute transverse myelitis.Semin Neurol 2008;28:105–120.

12. Brizzi KT, Lyons JL. Peripheral nervous system manifestations of infectious disease.Neurohospitalist 2014;4:230–240.

13. Gilden D. Varicella zoster virus and central nervous system syndromes. Herpes 2004;11(suppl 2):89A–94A.

14. Pergam SA, Limaye AP. Varicella zoster virus (VZV). Am J Transpl 2009;9:S108–S115.

15. Cohen JI. Clinical practice: herpes zoster. N Engl J Med 2013;369:255–263.16. Gilden D, Nagel MA, Cohrs RJ, Mahalingam R. The variegate neurological mani-

festations of varicella zoster virus infection. Curr Neurol Neurosci Rep 2013;13:374.17. Wenjuan T, Rao S. Mechanisms underlying T cell immunosenescence: aging and

cytomegalovirus infection. Front Microbiol 2016;7:2111.18. Skripuletz T, Pars K, Schulte A, et al. Varicella zoster virus infections in neurological

patients: a clinical study. BMC Infect Dis 2018;18:238.19. Nagel MA, Gilden D. Update on varicella zoster vasculopathy. Curr Infect Dis Rep

2014;16:407.20. Grahn A, Studahl M. Varicella-zoster virus infections of the central nervous system:

prognosis, diagnostics and treatment. J Infect 2015;71:281–293.

6 Neurology: Neuroimmunology & Neuroinflammation | Volume 6, Number 4 | July 2019 Neurology.org/NN

Page 143: LWW NEURIMMINFL NEURIMMINFLcover 6 4...Alex Rae-Grant, MD Vision Neurology® Neuroimmunology & Neuroinflammation will be the premier peer-reviewed journal in the fields of neuroimmunology

CORRECTION

α4-integrin deficiency in B cells does not affect disease in aT-cell–mediated EAE disease modelNeurol Neuroimmunol Neuroinflamm 2019;6:e585. doi:10.1212/NXI.0000000000000585

In the article “α4-integrin deficiency in B cells does not affect disease in a T-cell–mediated EAEdisease model" by Hussain et al.,1 first published online April 16, 2019, funding for the papershould have been listed in the Study Funding section. The authors regret the error.

Reference1. Hussain RZ, Cravens PD, Miller-Little WA, et al. α4-integrin deficiency in B cells does not affect disease in a T-cell–mediated EAE

disease model. Neurol Neuroimmunol Neuroinflamm 2019;6:e563.

Copyright © 2019 American Academy of Neurology 1

Copyright © 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.