Journal of Immunotoxicology, 1:131–139, 2004Copyright c© Taylor & Francis Inc.ISSN: 1547-691X print / 1547-6901 onlineDOI: 10.1080/15476910490894904

Reviews

Preclinical Safety and Immune-Modulating Effectsof Therapeutic Monoclonal Antibodies to Interleukin-6and Tumor Necrosis Factor-α in Cynomolgus Macaques

Pauline L. Martin, Joel Cornacoff, Uma Prabhakar, Thomas Lohr,and George TreacyCentocor, Inc., Malvern, Pennsylvania, USA

Jessica E. Sutherland, Sarah Hersey, and Elise MartinCharles River Laboratories Discovery and Development Services, Sparks, Nevada and Worcester,Massachusetts, USA

Inhibitors of tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6) have been shown to be efficacious in a number of autoim-mune diseases. In this study, the safety of long-term administrationof anti-TNF-α and anti-IL-6 monoclonal antibodies (mAbs) wasevaluated in cynomolgus macaques. Effects on the immune systemwere evaluated by analysis of lymphocyte subsets and histopathol-ogy of lymphoid tissues. To evaluate the functioning of the immunesystem, the ability of mAb-treated monkeys to mount a humoralimmune response (IgG and IgM) to keyhole limpet hemocyanin(KLH) was evaluated. Treatment with the anti-TNF-α mAb pro-duced no histopathological changes in any of the lymphoid tissuesexamined. There was a small (<2-fold) elevation in circulating T-and B-lymphocytes during anti-TNF-α mAb treatment that wasnot considered to be toxicologically significant. The antibody re-sponse to KLH was unaffected by anti-TNF-α mAb treatment.Treatment with anti-IL-6 mAb resulted in a decrease in the size ofgerminal centers in the spleens of a minority of the animals anda modest but significant decrease in the IgG antibody response toKLH. Weekly intravenous treatment with the anti-IL-6 mAb andtwice-weekly subcutaneous treatment with the anti-TNF-α mAbfor up to 6 months was not associated with any signs of toxicity,and no animal developed an infection throughout the study period.This study demonstrates that the anti-IL-6 and anti-TNF-α mAbsproduce specific modulating effects on the immune system withoutrendering the animals immune compromised.

Keywords nonhuman primate, interleukin-6, tumor necrosis factoralpha, immune modulation, monoclonal antibody.

Received 9 April 2004; accepted 23 August 2004.Address correspondence to Pauline L. Martin Ph.D., Department of

Toxicology and Investigational Pharmacology, Centocor, Inc., Malvern,Pennsylvania 19355, USA. E-mail: Pmarti27@cntus.jnj.com

INTRODUCTIONTumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6) are

proinflammatory cytokines that are synthesized and released bya number of cell types including macrophages. TNF-α is im-portant in host defense against foreign invaders such as bac-teria, whereas IL-6 is important in B-cell differentiation, anti-body production, and T-cell proliferation and activation (Akiraet al., 1990). IL-6 and TNF-α levels are elevated following ex-posure to bacterial endotoxins such as lipopolysaccharide (LPS)(Victoratos et al., 1997).

TNF-α and IL-6 have also been shown to be involved in thepathogenesis of various autoimmune diseases. Monoclonal anti-bodies (mAbs) directed against TNF-α or the IL-6 receptor havebeen shown in both animal disease models and clinically to re-duce disease severity in rheumatoid arthritis and Crohn’s disease(Takagi et al., 1998; Mihara et al., 2001; Choy et al., 2002; Itoet al., 2002; Shealy et al., 2002; Paleolog, 2003; Nishimoto et al.,2004). Inhibitors of IL-6 have also been shown to be effectiveboth preclinically and clinically at inhibiting the growth of avariety of human tumors (Tsunenari et al., 1996; Van Zaanenet al., 1998; Smith and Keller, 2001; Trikha et al., 2003).

Because inhibitors of TNFα and IL-6 produce their thera-peutic effects in autoimmune diseases by modulating immuneresponses, there is the potential that these inhibitors could in-crease susceptibility to infection, as has been shown with theTNFα inhibitors. Therefore, there is a need for improved im-mune functional testing in preclinical animal models in orderto more accurately predict human risk. Because anti-cytokinemonoclonal antibodies produce selective modulation of the im-mune system rather than overt immune suppression their effectscan be difficult to detect in normal healthy laboratory animals

131

Jour

nal o

f Im

mun

otox

icol

ogy

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f C

alif

orni

a Ir

vine

on

10/3

0/14

For

pers

onal

use

onl

y.

132 P. L. MARTIN ET AL.

using standard toxicology endpoints. In this study, the safetyand immune modulating effects of long-term treatment withan anti-TNF-α mAb and an anti-IL-6 mAb were evaluated incynomolgus macaques. Effects of treatments on the immune sys-tem were evaluated by analysis of lymphocyte subsets and im-munohistopathology of lymphoid tissues. In addition the func-tioning of the immune system was evaluated by the ability oftreated macaques to mount a humoral immune response to theT-cell dependent neoantigen keyhole limpet hemocyanin (KLH)was also evaluated.

These studies demonstrate that treatment of cynomolgusmacaques with a monoclonal antibody to IL-6 resulted in a re-duction in the IgG humoral immune response to the neoantigenKLH but was not associated with any signs of toxicity follow-ing long-term treatment. Treatment with the anti-TNFα mAbresulted in no signs of immune suppression in this model.

METHODSAll in-life procedures were conducted in compliance with the

Animal Welfare Act, the Guide for the Care and Use of Labo-ratory Animals, and the Office of Laboratory Animal Welfare.Study protocols were reviewed and approved by the InstitutionalAnimal Care and Use Committee.

The anti-human IL-6 mAb used in these studies was a chimeric(human/murine) monoclonal (subtype IgG1) specific for IL-6.The anti-human TNF-α mAb was a fully human IgG1 mAb.Endotoxin levels in all mAbs used in these studies were lessthan 0.02 EU/mg of mAb. Both antibodies bind to human andcynomolgus cytokines with affinities of less than 1 nM and wereproduced by Centocor, Inc. (Malvern, PA). The dosing regimensused in these studies were chosen to mimic possible clinical dos-ing regimens and to produce sustained blood levels of mAbs inexcess of 100 µg/mL (i.e., >600-fold in excess of the affin-ity constants) in cynomolgus macaques throughout the multipledose treatment periods.

Monkeys of Mauritius origin weighing approximately 2.5 kgat the onset of the study were herpes B virus, simian retrovirus,simian immunodeficiency virus and tuberculosis negative. Allanimals had previously been vaccinated against measles.

TABLE 1Summary of study designs

mAb Tested mAb dosing regimen

mAb dosestested

(mg/kg)

Number ofanimals per

treatment groupDay of

first KLHDay of

second KLH

Anti-TNFαa Twice per week sc for 13 weeks 0, 10, 50 3M/3F 12 30Anti-TNFαa Twice per week sc for 26 weeks 0, 10, 50 5M/5F 12 30Anti-IL-6b Once per week iv for 4 weeks 0, 25, 50 3M/3F 10 24Anti-IL-6b Once per week iv for 13 weeks 0, 25, 50 5M/5F 10 24Anti-IL-6 Once per week iv for 26 weeks 0, 25, 50 5M/5F 16 30

M = Male, F = Female, a the 3-month and 6-month anti-TNFα studies were conducted as a single combined study. bThe1-month and 3-month anti-IL-6 studies were conducted as a single combined study. All mAb treatments were initiated onDay 1.

General Toxicity EvaluationsThe safety of repeated administration, for up to 6 months,

of the anti-TNF-α mAb and the anti-IL-6 mAb were evaluatedin young male and female, experimentally naı̈ve, cynomolgusmacaques. The anti-TNF-α mAb was administered twice perweek by subcutaneous injection and the anti-IL-6 mAb was ad-ministered once per week by 2-hour intravenous infusion. Thefirst day of dosing was designated day 1. Control animals re-ceived saline at a volume equivalent to the 50 mg/kg mAb doselevels.

Potential toxicity due to the repeated administration of anti-TNF-α and anti-IL-6 mAbs was assessed by evaluation of clin-ical signs, body weight, ophthalmic and physical examinations,food consumption, electrocardiograms, body temperature, bloodpressure, heart rate, respiratory rate, and clinical pathology (stan-dard panels of hematology, coagulation, serum chemistry, andurinalysis) prior to treatment and at various times throughout thecourse of the study. At the end of each of the treatment periodsshown in Table 1, 3 male and 3 female monkeys from each treat-ment group were euthanised and a complete anatomic pathologyexamination was performed including histopathological exami-nation of hematoxylin and eosin (H & E)-stained tissue sections.The remaining animals (2 per sex per group) were allowed a 4-to 12-week treatment-free period prior to necropsy to evaluatethe reversibility of any treatment related findings.

Immunotoxicity EvaluationsEffects of treatments on the immune system were assessed

by evaluation of white blood cell counts and differentials, phe-notyping of lymphocyte subsets, gross and histological exam-ination of lymphoid tissues, and functional immune challengefollowing immunization with KLH.

Lymphocytesubsetswereevaluated by fluorescence-activatedcell sorting (FACS) analysis (Charles River Laboratories, Sparks,NV). Heparin-anticoagulated peripheral blood samples werecollected and labeled with antibodies appropriate to identifyspecific lymphocyte and monocyte subsets. Cells evaluated wereCD2+CD20+ total B-cells, CD3+CD4+T-helper lymphocytes,CD3+CD8+ T-cytotoxic/suppressor lymphocytes, CD3-CD14+

Jour

nal o

f Im

mun

otox

icol

ogy

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f C

alif

orni

a Ir

vine

on

10/3

0/14

For

pers

onal

use

onl

y.

IMMUNE-MODULATING EFFECTS OF IL-6 AND TNFα mAbs 133

monocytes, CD3-CD16+ NK cells, CD3+ CD44+ memoryT-lymphocytes and CD3+CD45RA+ naı̈ve lymphocytes. Bloodsamples were collected for FACS analysis prior to initiation oftreatments (3 time points), 1 day after the first dose and at approx-imately 1, 3, and 6 months during the treatment periods. FACSanalysis was not conducted in the 6-month anti-IL-6 mAb study.

Immunohistomorphologic analyses were conducted on sec-tions of spleen, thymus, Peyer’s patch, tonsil, and lymph node(axillary, mesenteric, mandibular, and inguinal). Tissues wereformalin-fixed and paraffin-embedded. Tissue sections werestained with hematoxylin and eosin and for CD3 (T-cells) andCD20 (B-cells) antigens using mouse anti-CD20 and mouse orrabbit anti-CD3 antibodies (Dako, CA). The negative control an-tibody for CD3 was mouse IgG (Cappel, CA) and the negativecontrol antibody for CD20 was mouse IgG2a (Calbiochem, CA).Tissues were stained using an Avidin-Biotin Complex (ABC) kit(Vector Laboratories, CA). Lymphocyte populations were eval-uated qualitatively.

Effects of Anti-TNF-α and Anti-IL-6 mAbs on the KLHAntibody Response

Immune system function was evaluated by the ability of themonkeys to mount a humoral immune response to KLH (ImjectMariculture KLH, Pierce Chemical Company). KLH (100 µgin 1.0 mL IFA) was administered intramuscularly to the rightor left quadriceps on 2 occasions 14–18 days apart (Table 1).Blood samples were collected from all animals prior to mAbtreatments and for up to 6 months during the mAb treatmentperiods and were analyzed for anti-KLH antibodies. The influ-ence of treatment with anti-TNF-α and anti-IL-6 mAbs on theanti-KLH antibody response was evaluated.

Serum samples collected prior to and during mAb treat-ments were analyzed for anti-KLH antibodies using an enzyme-linked immunosorbent assay (ELISA) method (Charles RiverLaboratories, Sparks, NV). Ninety-six-well microtiter plateswere coated with KLH antigen and stored overnight at 2–8◦C.Coated plates were washed to remove excess antigen and wereblocked. Triplicate samples of diluted sera from KLH treatedmonkeys and controls were added to the plates. Plates werewashed to remove unbound sample. A secondary antibody linkedto horseradish peroxidase was added to the plates. The sec-ondary antibody was goat anti-monkey IgG or goat anti-monkeyIgM. After washing out excess antibody, enzyme substrate 3,3′, 5, 5′-tetramethylbenzidine (TMB) was added. The reactionwas stopped by the addition of H2SO4.

The developed plates were read on a Spectramax microplatereader to determine the KLH antibody titer using the centerpoint method (defined as the dilution of serum that results inhalf-maximum absorbency).

Serum IL-6 ConcentrationsSerum samples from a subset of animals in the 3-month anti-

IL-6 mAb study were analyzed for total IL-6 concentration using

a commercially available high sensitivity ELISA kit (R&D Sys-tems, MN) according to the manufacturer’s instructions. Dupli-cate samples of diluted serum were added to wells of a 96-wellmicrotiter plate precoated with a mouse monoclonal antibodyagainst IL-6.

The developed plates were read on a Spectramax microplatereader and Softmax Pro-software version 3.1.2 was used for dataanalysis. The mean OD (490 –650 nm) for each standard concen-tration, after subtraction of the blank (0 standard) value, was plot-ted against their respective concentrations using log-log curvefit. Sample values were interpolated from the standard curve.

Statistical AnalysisAll data are expressed as the mean ± standard error of the

mean (SEM), which were calculated using Microsoft Excel Soft-ware Version 5.0. For the FACS analysis data, the mean of the3 pretreatment values was calculated. The cell populations fol-lowing treatment with mAb are expressed as a percent of thismean pretreatment value. These data manipulations were neces-sary in order to reduce the intra-and interanimal variability andto allow for any effect of the mAb treatments to be more clearlyidentified.

The area under the serum anti-KLH titer versus time curves(AUC) were calculated using the linear trapezoid rule usingGraphPad Prizm Software Version 3.03 (San Diego, CA). Themaximum measured anti-KLH antibody titers, Cmax, were de-termined by observation. Treatment-related differences in Cmax

and AUC values were analyzed by 1-way analysis of variance(ANOVA) followed by Dunnett’s multiple comparisons test inwhich the treatment groups were compared to the control us-ing GraphPad Prizm Software. Treatment related differences ofp < 0.05 were considered significant.

RESULTS

General Toxicity EvaluationsTreatment with anti-TNF-α and anti IL-6 mAbs for up to

6 months produced no treatment related effects on clinical signs,body weight, food consumption, ophthalmic and physicalexaminations, heart rate, blood pressure, electrocardiographicmeasurements, respiratory rate, body temperature, clinical chem-istry, coagulation parameters, urinalysis, or gross and micro-scopic anatomic pathology of nonlymphoid tissues. Anti-mAbantibodies (measured during the mAb treatment periods and dur-ing the treatment-free recovery period) were detected only in 1anti-IL-6 and 1 anti-TNF-α mAb treated animal (results notshown).

Immunotoxicity EvaluationsEffects of Anti-TNFα mAb and Anti-IL-6 mAb Treatmentson Circulating Leukocytes

Treatment of macaques with the anti-TNF-α mAb produced asignificant elevation in circulating T- and B-lymphocytes24 hours after the first dose administration (Figure 1). This

Jour

nal o

f Im

mun

otox

icol

ogy

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f C

alif

orni

a Ir

vine

on

10/3

0/14

For

pers

onal

use

onl

y.

134 P. L. MARTIN ET AL.

FIG. 1. Effect of anti-TNFα mAb treatment on T and B lymphocytes, monocytes and NK cells determined by flow cytometric analysis. Each bar represents themean ± SEM of cell counts from 10–16 monkeys. ∗ p < 0.05,∗∗ p < 0.01 significantly different from control group.

elevation in circulating lymphocytes persisted throughout the6-month dosing period. The magnitude of the elevation wasless than 2-fold for the T-lymphocytes and 2–3-fold for theB-lymphocytes. In the subset of animals (2 per sex per group)that were allowed a 3-month treatment-free period following6 months of anti-TNFα mAb treatment, T-cells returned to-ward control levels. B-cells remained elevated above baseline at12 weeks postdose, at which time anti-TNFα mAb concentra-tions were still detectable in the serum (results not shown).

The elevations in T-cells comprised elevations of T-helperlymphocytes, T-cytotoxic/suppressor lymphocytes, naı̈veT-lymphocytes and memory T-lymphocytes. With the possibleexception of a single time point (Figure 1), monocytes and NKcells were unaffected by anti-TNF-α mAb treatment.

Treatment with the anti-IL-6 mAb produced a small (<2-fold), but significant, increase in T and B-lymphocytes 24 hoursfollowing the first dose (Figure 2). All T-lymphocyte subsetsevaluated were elevated. However, by 1 and 3 months followingweekly dosing there were no treatment related changes in lym-phocytes. Monocytes and NK cells were unaffected by treatmentwith anti-IL-6 mAb.

Effects of Anti-TNF mAb and Anti-IL-6 mAb Treatmenton Lymphoid Morphology.

Immunohistopathologic evaluation of lymphoid tissues fromanimals treated with the anti-TNF-α mAb showed no qualitativechanges in T- and B-cell distribution relative to saline-controltreated animals.

Following 1 month of treatment with the IL-6 mAb, a minimalreduction in the size and number of germinal centers (follicularatrophy) was observed in the spleens of 3 of 6 animals in thehigh-dose treatment group (50 mg/kg). Following 3 months oftreatment follicular atrophy was seen in only 1 of 6 high-dosetreated animals and also in 1 of 6 saline-control animals. Fol-licular atrophy was not seen in any of the recovery animals andwas not seen in any of the animals treated with the anti-IL-6mAb for 6 months.

Effects of Anti-TNF-α mAb on the KLH Antibody ResponseThe effects of twice-weekly subcutaneous treatment with the

anti-TNF-α mAb on the IgG and IgM responses to KLH areillustrated in Figure 3. Comparison of the maximum anti-KLHtiters (Cmax) and area under the titers versus time curves (AUC)is shown in Tables 2 and 3. No significant differences weredetected between the treatment groups for Cmax or AUC foreither anti-KLH IgG or anti-KLH IgM antibodies.

Effects of Anti-IL-6 mAb on the KLH Antibody ResponseThe effects of once-weekly intravenous treatment with the

anti-IL-6 mAb on the IgG and IgM responses to KLH are illus-trated in Figure 4. Comparison of the Cmax and AUC values isshown in Tables 2 and 3.

The anti-KLH IgG response decreased in a dose-dependentmanner following treatment with the anti-IL-6 mAb and reachedstatistical significance at the 50 mg/kg dose level (Table 2).The magnitude of the reduction in AUC obtained at the

Jour

nal o

f Im

mun

otox

icol

ogy

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f C

alif

orni

a Ir

vine

on

10/3

0/14

For

pers

onal

use

onl

y.

IMMUNE-MODULATING EFFECTS OF IL-6 AND TNFα mAbs 135

FIG. 2. Effect of anti-IL-6 mAb treatment on T and B lymphocytes, monocytes and NK cells determined by flow cytometric analysis. Each bar represents themean ± SEM of cell counts from 10–16 monkeys. ∗ p < 0.05,∗∗ p < 0.01 significantly different from control group.

FIG. 3. Effects of anti TNF-α mAb treatment on anti-KLH IgG and IgMantibody responses in the cynomolgus macaque. Each value is the mean ± SEMof titers from 10–16 monkeys.

50 mg/kg dose level was 26–35% and the reduction in Cmax was28–39%. The AUC and Cmax for the anti-KLH IgM responseswere not significantly affected by anti-IL-6 mAb treatment(Table 3).

Effect of Anti-IL-6 mAb on Serum IL-6 ConcentrationsSerum IL-6 concentrations were measured by ELISA from

a subset of animals in the 3-month IL-6 mAb study (3 animalsper dose level). Prior to the initiation of treatments, serum con-centrations of IL-6 were below 1 pg/mL. During the first 2-hoursaline infusion (2.5 mL/kg) in the control animals serum IL-6concentration increased to 22 ± 12.9 pg/mL and subsequentlyreturned to pretreatment levels. IL-6 levels were not elevatedwith subsequent saline dosing. Administration of KLH on Days10 and 24 in saline-treated animals was not associated with adetectable elevation in serum IL-6 concentrations.

Serum IL-6 concentrations increased during the infusion ofthe IL-6 mAb and continued to rise after the end of the infusion(Figure 5). Following weekly administration of IL-6 mAb, peakIL-6 concentrations of about 1,600 pg/mL were obtained at aboutDay 30–56. The IL-6 levels in the anti-IL-6 mAb-treated animalsremained elevated throughout the anti-IL-6 treatment period anddecreased slowly after the end of the mAb treatment period.

DISCUSSIONThis study demonstrates that treatment of cynomolgus

macaques with anti-IL-6 and anti-TNF-α mAbs for up to 6

Jour

nal o

f Im

mun

otox

icol

ogy

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f C

alif

orni

a Ir

vine

on

10/3

0/14

For

pers

onal

use

onl

y.

136 P. L. MARTIN ET AL.

TABLE 2Summary of anti-KLH IgG responses (Mean ± SEM) in cynomolgus macaques following two administrations of KLH/adjuvant

Cmax (IgG Titer) AUC(0−3 months) (IgG Titer/Day) AUC(0−6 months) (IgG Titer/Day)

Anti-TNF-α (6 months treatment)Control 2500 ± 193 145214 ± 8153 265453 ± 15806

(n = 16) (n = 16) (n = 10)25 mg/kg 3182 ± 594 176428 ± 26642 272121 ± 37082

(n = 16) (n = 16) (n = 10)50 mg/kg 2974 ± 321 153089 ± 12872 249339 ± 22898

(n = 16) (n = 16) (n = 10)

Anti-IL-6 (3-month treatment)Control 1552 ± 106 90572 ± 4680 —

(n = 10) (n = 10)10 mg/kg 1342 ± 103 78154 ± 5663 —

(n = 10) (n = 10)50 mg/kg 1118 ± 122∗ 66782 ± 5216∗∗ —

(n = 10) (n = 10)

Anti-IL-6 (6-months treatment)Control 3807 ± 457 193144 ± 20583 443498 ± 34824

(n = 10) (n = 10) (n = 10)10 mg/kg 2976 ± 211 152804 ± 9588 376084 ± 29014

(n = 10) (n = 10) (n = 10)50 mg/kg 2308 ± 247∗∗ 125162 ± 13259∗∗ 297887 ± 33254∗∗

(n = 10) (n = 10) (n = 10)

∗ p < 0.05, ∗∗ p < 0.01, significantly different from control.

months was safe and that only minor modulating effects on theimmune system were observed.

Weekly treatment with the anti-IL-6 mAb produced a smalldecrease in the size of germinal centers in the spleen of 3 of6 high-dose treated animals after 1 month of treatment. However,this effect was not evident with more prolonged dosing. Treat-ment with the IL-6 mAb also produced a small reduction in theIgG antibody response to KLH without affecting the anti-KLHIgM antibody response. The anti-TNF-α mAb had no effect onlymphoid morphology and had no effect on either the anti-KLHIgG or IgM responses.

The inhibition of the IgG response by anti-IL-6 mAb treat-ment in cynomolgus macaques is similar to that previously pub-lished for mice (Mihara et al., 2002; Okazaki et al., 2002) andconstitutes a validation that the method used in the macaqueswas sufficiently sensitive to detect the immune modulating ef-fects of therapeutic agents. The clinical significance of the smallinhibition in IgG antibody production in macaques is not ap-parent from this study. The monkeys did not experience a de-crease in T- and B-cell numbers, and there were minimal changesin splenic architecture in a minority of the animals. None ofthe animals developed an infection during the study and nosigns of systemic toxicity were apparent. Therefore, althoughtreatment with the IL-6 antibody did produce an inhibition ofthe anti-KLH IgG antibody production, animals were still ableto mount a substantial antibody response to the T-cell depen-

dent neoantigen indicating that the animals were not immune-compromised.

Previously published studies conducted in mice have shownthat IL-6 is not a requirement for antibody production but thatIL-6 enhances the antibody response. IL-6-deficient mice showa reduced IgG antibody response and have germinal centerssmaller than those of wild-type mice (Kopf et al., 1998). Themonkey studies additionally demonstrated that the IL-6 mAbinhibited the IgG responses to a neoantigen to a greater extentthan the IgM responses. This effect is similar to that seen inmice, suggesting that IL-6 has similar functions in macaquesand mice. IL-6 has been proposed to be involved in the terminaldifferentiation of committed B-cells (Kawano et al., 1995; Kopfet al., 1998). These studies conducted in cynomolgus macaqueswith the anti-IL-6 mAb provide a validation that the KLH immu-nization procedure is sufficiently sensitive to detect the selectiveimmune modulating effects of anti-cytokine mAbs.

Administration of KLH in IFA to saline treated macaques wasnot associated with an elevation of circulating IL-6. Adminis-tration of the anti-IL-6 mAb was, however, associated with anelevation in the amount of circulating IL-6 detected by ELISA.The commercially available ELISA assay used to detect IL-6 inserum does not differentiate between free IL-6 and IL-6 that isbound to the anti-IL-6 mAb. Preliminary data (not shown) indi-cate that the increase in detectable IL-6 in the mAb-treated ani-mals is due to bound IL-6 and not due to free IL-6 and is therefore

Jour

nal o

f Im

mun

otox

icol

ogy

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f C

alif

orni

a Ir

vine

on

10/3

0/14

For

pers

onal

use

onl

y.

IMMUNE-MODULATING EFFECTS OF IL-6 AND TNFα mAbs 137

TABLE 3Summary of anti-KLH IgM responses (Mean ± SEM) in cynomolgus macaques following two administrations of KLH/adjuvant

Cmax (IgM Titer) AUC(0−3 months) (IgM Titer/Day) AUC(0−6 months) (IgM Titer/Day)

Anti-TNF-α (6 month treatment)Control 353 ± 27 25447 ± 1752 47407 ± 2022

(n = 16) (n = 16) (n = 6)25 mg/kg 429 ± 34 28462 ± 1526 50840 ± 1945

(n = 16) (n = 16) (n = 7)50 mg/kg 358 ± 17 25361 ± 1276 49256 ± 2930

(n = 16) (n = 16) (n = 8)

Anti-IL-6 (3 month Treatment)Control 340 ± 17 22318 ± 1328 —

(n = 10) (n = 10)10 mg/kg 312 ± 28 20374 ± 1000 —

(n = 10) (n = 10)50 mg/kg 305 ± 11 20513 ± 730 —

(n = 10) (n = 10)

Anti-IL-6 (6 month treatment)Control 353 ± 16 18748 ± 2395 —a

(n = 10) (n = 10)10 mg/kg 324 ± 23 18130 ± 2293 —a

(n = 10) (n = 5)50 mg/kg 327 ± 24 16393 ± 2474 —a

(n = 10) (n = 10)

aSerum titers fell below the limit of detection of the assay by 6-month, therefore partial AUC values only could be calculated over the first 3months of treatment.

FIG. 4. Effects of anti IL-6 mAb treatment on anti-KLH IgG and IgM antibody responses in the cynomolgus macaque. Top panels are from the combined1-month/3-month study. Bottom panels are from the 6-month study. Each value is the mean ± SEM of titers from 10–16 monkeys.

Jour

nal o

f Im

mun

otox

icol

ogy

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f C

alif

orni

a Ir

vine

on

10/3

0/14

For

pers

onal

use

onl

y.

138 P. L. MARTIN ET AL.

FIG. 5. Effects of anti IL-6 mAb treatment on total serum IL-6 concentrationsmeasured by ELISA. Data are the Mean ± SEM of 3 values conducted induplicate.

not biologically active. Studies conducted with human, mouse,and baboon blood have shown that circulating IL-6 can existas a large molecular mass complex of IL-6 bound to a solubleIL-6 receptor (May et al., 1992, 1993; Ndubuisi et al., 1998). Inthis complexed form, IL-6 immunoreactivity is masked. How-ever, when an anti-IL-6 mAb is administered, the IL-6 binds tothe mAb and is now detectable by an ELISA assay. Therefore, theresults seen in the cynomolgus macaque are consistent with theobservations in other species and suggest that normal cynomol-gus macaques have a large amount of bound IL-6 in the bodythat is in a biologically inactive form.

Treatment of macaques with the anti-TNF-α mAb producedno inhibition of the antibody response to KLH and lymphoid tis-sues were unaffected by treatment. Unlike anti-IL-6 mAb treat-ment in monkeys that showed effects similar to those seen inIL-6 deficient mice, the effects of TNF-α mAb treatment inmacaques does not mirror the effects observed in TNF-α de-ficient mice. TNF-α deficient mice exhibit impaired humoralimmune responses and an absence of splenic germinal centers(Pasarakis et al., 1996). Clearly, effects exhibited by mice ge-netically altered to lack specific cytokines do not always mimicthe effects of long-term inhibition of cytokines in animals witha fully developed immune system.

Although clinical experience has indicated that treatmentwith TNF-α inhibitors is associated with an increased risk ofinfections, including tuberculosis, there is no evidence from thenonhuman primate studies that would predict this clinical risk.There was no decrease in the numbers of lymphocytes, mono-cytes or NK cells. In fact, within 24 hours of initiation of treat-ment, there was a significant 2- to 3-fold increase in the numberof circulating lymphocytes. An increase in circulating lympho-cytes has also been observed in rheumatoid arthritis patientstreated with the anti-TNF-α mAb infliximab (Paleolog et al.,1996). The increase in circulating lymphocytes was shown tobe accompanied by a decrease in endothelial cell adhesion mark-ers and a reduced migration of lymphocytes into the synovialtissues (Tak et al., 1996). TNF-α enhances the adhesion of lym-phocytes to endothelial cells through upregulation of adhesion

molecules (Thornhill et al., 1991). Inhibition of this up-regulationby the anti-TNF-α mAb would be predicted to cause a reduc-tion in adhesion of lymphocytes to the endothelium and is likelythe mechanism causing reduced migration of lymphocytes toinflamed joints, and therapeutic efficacy, in rheumatoid arthritispatients. Whether TNF-α plays a tonic role in regulating lym-phocyte adhesion in nondiseased animals is not known at thistime but could explain the small but apparently toxicologicallyinsignificant increase in circulating lymphocytes in monkeys fol-lowing anti-TNF-α mAb treatment and may represent a usefulsurrogate marker of efficacy.

Overall, these studies have shown that long-term inhibitionof IL-6 or TNFα with blocking mAbs produced no toxicolog-ical effects in normal cynomolgus macaques and only minormodulation of the immune system under the conditions of thesestudies.

ACKNOWLEDGMENTSThe authors would like to thank William Hall for immuno-

histopathology evaluation, Patrick Lappin for flow cytometryanalysis and Tammy Pawling and Patricia Barr for study coor-dination. All in-life portions of these studies were conducted atCharles River Laboratories Discovery and DevelopmentServices and were sponsored by Centocor Inc, Malvern,Pennsylvania.

REFERENCESAkira, S., Hirano, T., Taga, T., and Kishimoto, T. 1990. Biology of multi-

functional cytokines: IL-6 and related molecules (IL-1 and TNF). FASEBJ. 4:2860–2867.

Choy, E. H., Isenberg, D. A., Garrood, T., Farrow, S., Ioannou, Y., Bird, H.,Cheung, N., Williams, B., Hazleman, B., Price, R., Yoshizaki, K., Nishimoto,N., Kishimoto, T., and Panayi, G. S., 2002. Therapeutic benefit of blockinginterleukin-6 activity with an anti-interleukin-6 receptor monoclonal antibodyin rheumatoid arthritis: a randomized, double blind, placebo-controlled, dose-escalation trial. Arthritis Rheum. 46:3143–3150.

Ito, H., Hirotani, T., Yamamoto, M., Ogawa, H., and Kishimoto, T. 2002. Anti-IL-6 receptor monoclonal antibody inhibits leukocyte recruitment and pro-motes T-cell apoptosis in a murine model of Crohn’s disease. J. Gastroenterol.37:56–61.

Kawano, Y., Noma, T., Kou, K., Yoshizawa, I., and Yata, J. 1995. Regulation ofhuman IgG subclass production by cytokines: human IgG subclass productionenhanced differentially by interleukin-6. Immunology 84:278–284.

Kopf, M., Herren, S., Wiles, M. V., Pepys, M. B., and Kosco-Vilbois, M. H.1998. Interleukin 6 influences germinal center development and antibodyproduction via a contribution of C3 complement component. J. Exp. Med.188:1895–1906.

May, L. T., Neta, R., Moldawer, L. L., Kenney, J. S., Patel, K., and Sehgal,P. B. 1993. Antibodies chaperone circulating IL-6. J. Immunol. 151:3225–3236.

May, L. T., Viguet, H., Kenney, J. S., Ida, N., Allison, A. C., and Sehgal, P. B.1992. High levels of “complexed” interleukin-6 in human blood. J. Biol.Chem. 267:19698–19704.

Mihara, M., Kotoh, M., Nishimoto, N., Oda, Y., Kumagai, E., Takagi, N.,Tsunemi, K., Ohsugi, Y., Kishimoto, T., Yoshizaki, K., and Takeda, Y. 2001.Humanized antibody to human interleukin-6 receptor inhibits the develop-ment of collagen arthritis in cynomolgus monkeys. Clin. Immunol. 98:319–326.

Jour

nal o

f Im

mun

otox

icol

ogy

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f C

alif

orni

a Ir

vine

on

10/3

0/14

For

pers

onal

use

onl

y.

IMMUNE-MODULATING EFFECTS OF IL-6 AND TNFα mAbs 139

Mihara, M., Nishimoto, N., Yoshizaki, K., and Suzuki, T. 2002. Influences ofanti-mouse interleukin-6 receptor antibody on immune responses in mice.Immunol. Lett. 84:223–229.

Ndubuisi, M. I., Patel, K., Rayanade, R. J., Mittelman, A., May, L. T., and Sehgal,P. B. 1998. Distinct classes of chaperoned IL-6 in human blood: differentialimmunological and biological availability. J. Immunol. 160:494–501.

Nishimoto, N., Yoshizaki, K., Miyasaka, N., Yamamoto, K., Kawai, S.,Takeuchi, T., Hashimoto, J., Azuma, J., and Kishimoto, T. 2004. Treatmentof rheumatoid arthritis with humanized anti-interleukin-6 receptor antibody.Arthritis Rheum. 50:1761–1769.

Okazaki, M., Yamada, Y., Nishimoto, N., Yoshizaki, K., and Mihara, M. 2002.Characterization of anti-mouse interleukin-6 receptor antibody. Immunol.Lett. 84:231–240.

Paleolog, E. 2003. The therapeutic potential of TNF-alpha blockade in rheuma-toid arthritis. Expert Opin. Investig. Drugs 12:1087–1095.

Paleolog, E. M., Hunt, M., Elliot, M. J., Feldman, M., Maini, R. N., and Woody,J. N. 1996. Deactivation of vascular endothelium by monoclonal anti-tumornecrosis factor alpha antibody in rheunmatoid arthritis. Arth. Rheum. 39:1082-1091.

Pasarakis, M., Alexopoulou, L., Episkopou, V., and Kollia, G. 1996. Immuneand inflammatory responses in TNFα-deficient mice: a critical requirementfor TNFα in the formation of primary B cell follicles, follicular dendriticcell networks and germinal centers, in the maturation of the humoral immuneresponse. J. Exp. Med. 184:1397–1411.

Shealy, D. J., Wooley, P. H., Emmell, E., Volk, A., Rosenberg, A., Treacy, G.,Wagner, C. L., Mayton, L., Griswold, D. E., and Song, X. Y. Anti-TNF-alphaantibody allows healing of joint damage in polyarthritic transgenic mice.Arthritis Res. 2002; 4:R7.

Smith, P. C., Keller, E. T. 2001. Anti-interleukin-6 monolcolanal antibody in-duces regression of human prostate cancer xenografts in nude mice. Prostate48:47–53.

Tak, P. P., Taylor, P. C., Breedveld, F. C., Smeets, T. J. M., Daha, M. R., Kluin,P. M., Meinders, A. E., and Maini, R. N. 1996. Decreases in cellularity andexpression of adhesion molecules by anti-tumor necrosis factor alpha mono-clonal antibody treatment in patients with rheumatoid arthritis. Arth. Rheum.39:1077–1081.

Takagi, N., Mihara, M., Moriya, Y., Nishimoto, N., Yoshizaki, K., Kishimmoto,T., Takeda, Y., and Ohsugi, Y. 1998. Interleukin-6 receptor blockage ame-liorates joint disease in murine collagen-induced arthritis. Arthritis Rheum.41:2117–2121.

Thornhill, M. H., Wellicome, S. M., Mahiouz, D. L., Lanchbury, J. S.,Kyan-Aung, U., and Haskard, D. O. 1991. Tumor necrosis factor com-bines with IL-4 or IFN-gamma to selectively enhance endothelial cell ad-hesiveness for T-cells. The contribution of vascular cell adhesion molecule-1-dependent and independent binding mechanisms. J. Immunol. 146: 592–598.

Trikha, M., Corringham, R., Klein, B., and Rossi, J.-F. 2003. Targeted anti-interleukin-6 monoclonal antibody therapy for cancer: a review of the rationaleand clinical evidence. Clin. Cancer Res. 9:4653–4665.

Tsunenari, T., Akamatsu, K., Kaiho, S., Sato, K., Tsuchiya, M.,Koishihara, Y., Kishimoto, T., and Ohsugi, Y. 1996. Therapeutic poten-tial of humanized anti-interleukin-6 receptor antibody: antitumor activ-ity in xenograft model of multiple myeloma. Anticancer Res. 16:2537–2544.

Van Zaanen, H. C. T., Lokhorst, H. M., Aarden, L. A., Rensink, H. J. A. M.,Warnaar, S. O., van der Lelie, J., and van der Oers, M. H. J. 1998. Chimericanti-interleukin 6 monoclonal antibodies in the treatment of advanced mul-tiple myeloma: a phase I dose-escalation study. Br. J. Haematol. 102:783–790.

Victoratos, P., Yiangou, M., Avramidis, N., and Hadjipetrou, L. 1997. Regula-tion of cytokine gene expression by adjuvants in vivo. Clin. Exp. Immunol.109:569–578.

Jour

nal o

f Im

mun

otox

icol

ogy

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f C

alif

orni

a Ir

vine

on

10/3

0/14

For

pers

onal

use

onl

y.