JOURNAL OF INTERFERON AND CYTOKINE RESEARCH 22:709–717 (2002)© Mary Ann Liebert, Inc.

Safety and Systemic Absorption of Pulmonary DeliveredHuman IFN-b1a in the Nonhuman Primate:

Comparison with Subcutaneous Dosing

PAULINE L. MARTIN,1 SUJATA VAIDYANATHAN,2 JOAN LANE,1 MARK ROGGE,3

NANCY GILLETTE,4 BIRGIT NIGGEMANN,5 and JAMES GREEN1

ABSTRACT

Safety and bioavailability of pulmonary delivered interferon-b1a (IFN-b1a, AVONEX®, Biogen, Inc., Cam-bridge, MA) was evaluated in the nonhuman primate. Pulmonary bioavailability following intratracheal (i.t.)instillation of 50 mg/kg IFN-b1a to rhesus macaques was approximately 10%. To evaluate pulmonary safety,IFN-b1a was administered intrabronchially to rhesus and cynomolgus macaques at a dose of 60 mg/dose one,three, or seven times per week for 4 weeks. At scheduled termination, lungs were evaluated for gross and his-tomorphologic changes. IFN-b1a or vehicle (human serum albumin [HSA] in phosphate-buffered saline [PBS])treatment resulted in minimal to mild subchronic alveolitis, located primarily near the instillation sites. Theseresponses were considered nonspecific and consistent with either instillation of a foreign protein or minor in-jury associated with the instillation procedure. In one rhesus macaque treated every day for 4 weeks, IFN-b1a induced mild to moderate eosinophilic alveolitis, considered possibly an isolated type I hypersensitivityresponse to HSA or IFN-b1a. Partial resolution of pulmonary lesions was seen in all recovery animals killed2 weeks after cessation of treatment. In conclusion, this study shows that pulmonary administration of hu-man IFN-b1a is safe and that the pulmonary route of administration is a possible alternate route for the sys-temic delivery of IFN-b1a.

709

INTRODUCTION

INTERFERON-b1a (IFN-b1a, AVONEX®, Biogen, Inc., Cam-bridge, MA) is currently approved in the United States for

the treatment of multiple sclerosis (MS).(1) AVONEX is ad-ministered clinically as a once per week 30-mg (,0.4 mg/kg, 6MU) intramuscular (i.m.) injection. The mechanism of actionof IFN-b1a in slowing the progression of disabilities associatedwith MS is not fully understood. However, IFN-b is known tohave a number of immunomodulatory effects that could con-tribute to its clinical efficacy.(2) In addition to the im-munomodulatory effects, the type I IFNs (IFN-a and IFN-b)also possess antiviral and antiproliferative activities. IFN-a andIFN-b act by binding to the same cell surface receptors. Afterbinding to IFN receptors on the cell surface, cytoplasmic tran-scription factors translocate to the cell nucleus to enhance orinhibit expression of a variety of cellular gene products.

The preclinical safety evaluation of systemically adminis-tered IFN-b1a used the rhesus macaque, as this is one of thefew species in which human IFN-b1a has been demonstratedto be biologically active. Systemic administration of IFN-b1ato humans and rhesus macaques results in an increase in serumneopterin concentrations.(3) Neopterin, which is produced pre-dominantly by macrophages in response to stimulation by IFNs,can be used as a surrogate marker for IFN biologic activity.

In preclinical safety studies in rhesus macaques, IFN-b1a ad-ministered every other day by subcutaneous (s.c.) injection atdoses of up to 50 mg/kg (10 MU/kg) was not associated withany significant clinical or histopathologic toxicity. To date, over100,000 patients have been treated with i.m. AVONEX, and thesafety profile has been good. The most common side effects inpatients are flu-like symptoms.

Because IFN-b1a is a protein, the possible routes of admin-istration are limited by enzymatic degradation and possible pH

1Biogen, Inc., Cambridge, MA 02142.2Novartis Pharmaceutical Corporation, East Hanover, NJ 07936-1080.3Immunex Corporation, Seattle, WA 98101.4Sierra Biomedical, Sparks, NV 89431.5Covance Laboratories GmbH, 48163, Munster, Germany.

instability. Studies conducted in rats following intratracheal(i.t.) instillation of human IFN(4,5) have shown that the bioavail-ability is approximately 15–56%, whereas following aerosol in-halation, bioavailability of up to 70% has been achieved.(5) In-haled IFNs have been administered by standard liquid aerosolnebulizer to normal volunteers and to patients with thoracic car-cinomas.(6–12) In these studies, aerosol administration of type IIFNs (a and b) at single doses of up to 216 MU or at dailydoses of 20 MU for up to 67 weeks was generally safe and welltolerated.

The pulmonary absorption studies conducted in rats withIFN-a and the human safety studies have suggested that thepulmonary route of administration could be a feasible alternateroute for delivery of systemically efficacious doses of IFN. Thecurrent studies were designed to evaluate the pulmonarybioavailability and safety of human IFN-b1a in macaques,which are generally considered to be biologically responsive tohuman IFN and physiologically similar to humans. Preliminaryhistopathologic findings from this study were presented recentlyat the 2001 Annual Meeting of the Society of ToxicologicPathologists.(13)

MATERIALS AND METHODS

All studies were carried out in accordance with the declara-tion of Helsinki or with the Guide for the Care and Use of Lab-oratory Animals as promulgated by the U.S. National Institutesof Health.

Pharmacokinetic and pharmacodynamic studies were con-ducted using the commercially available lyophilized AVONEXformulation. Each milliliter of reconstituted AVONEX contains30 mg recombinant human IFN-b1a (rHuIFN-b1a), 15 mg hu-man serum albumin (HSA) USP, 5.8 mg sodium chloride USP,5.7 mg dibasic sodium phosphate USP, and 1.2 mg monobasicsodium phosphate USP at a pH of approximately 7.3. For thepulmonary instillation studies, a formulation containing 60 mgAVONEX and 17 mg HSA in phosphate-buffered saline (PBS)was used. Vehicle controls contained HSA in PBS.

Pharmacokinetics and pharmacodynamics

Three female rhesus macaques were administered a single10 MU/kg (50 mg/kg) i.t. dose of IFN-b1a. One additional an-imal received an i.t. dose of a volume-matched (1 ml/kg) ve-hicle control. The i.t.-dosed animals were anesthetized with so-dium thiopental for injection (Pentothal, Abbott Laboratories,Abbott Park, IL) prior to dosing. IFN-b1a was administered tothe trachea as a fine spray using a 22-gauge lumbar punctureneedle. Blood was collected from a femoral vein prior to dos-ing and at 2, 4, 8, 12, 24, 48, and 72 h postdose. Serum washarvested and analyzed for IFN-b1a antiviral activity levels andneopterin concentrations. Animals were observed daily for anysign of overt toxicity.

The pharmacokinetic profile of IFN-b1a following s.c. dos-ing was evaluated at dose levels of 1 MU/kg (5 males, 6 fe-males) and 10 MU/kg (5 males, 5 females). Blood samples werecollected predose and at designated times through 24 h post-dose for evaluation of IFN-b antiviral activity levels. Theneopterin response following s.c. administration of IFN-b1a

was evaluated at dose levels of 0.25 MU/kg (5 males, 5 fe-males), 1 MU/kg (8 males, 6 females), and 10 MU/kg (11 males,11 females). Blood samples for analysis of neopterin concen-trations were collected prior to dosing and 24 h postdose. Pre-vious studies have shown that the peak neopterin response fol-lowing s.c. dosing occurs at 24 h postdose.(3)

Biologic assays

IFN-b serum levels. In the pharmacokinetic/bioavailabilitycomponent of this study, levels of IFN-b in serum were quan-tified as antiviral activity using a cytopathic effect (CPE) as-say. This assay has been described in detail elsewhere.(3) TheCPE assay detects the ability of IFN-b to protect human lungcarcinoma cells from cytotoxicity due to encephalomyelo-carditis virus (EMCV). The cells were preincubated with IFN-b1a to allow the synthesis of IFN-inducible proteins that mountan antiviral response.

In the pulmonary safety portion of this study, serum con-centrations of IFN-b were determined using a commerciallyavailable enzyme-linked immunosorbent assay (ELISA) (Fu-jirebio, Inc., Tokyo, Japan); 5 mg HuIFN-b1a is equal to 1 3

106 U (1 MU) of antiviral activity.Anti-IFN-b1a antibodies. Antibodies to IFN-b1a were mea-

sured using a sandwich-based ELISA assay that uses an anti-IFN-b monoclonal antibody (mAb) coated onto a microtiterwell to which IFN-b is bound. Bound IFN-b serves as the anti-gen for any anti-IFN-b antibodies present in the serum. The an-tibody-IFN-b antibody complex is detected using a secondarypolyclonal antihuman immunoglobulin G (IgG) conjugatedwith horseradish peroxidase (HRP). Samples that tested posi-tive for the presence of antibody in the ELISA assay were as-sayed for neutralizing activity. Neutralizing antibodies wereidentified by their ability to neutralize the protective effects ofIFN-b in the antiviral CPE assay.

Neopterin concentrations. Serum neopterin concentrationswere determined using a commercially available enzyme im-munoassay (EIA) kit (Immunochem™) (ICN Pharmaceuticals,Costa Mesa, CA) or by radioimmunoassay (RIA) using a 125Ikit (IBL, Hamburg, Germany). These assays were conducted inaccordance with the manufacturers’ instructions.

Pharmacokinetic and pharmacodynamic data analysis

Pharmacokinetic parameters were derived by model-inde-pendent analysis. Peak serum antiviral activity (Cmax) and thetime of peak activities (Tmax) were obtained by observation.Area under the serum activity curve (AUC) values were calcu-lated by the trapezoid algorithm. Pulmonary bioavailability was calculated by dividing the AUC over the first 24 h (AUC(0–24h)) following i.t. dosing at 10 MU/kg by theAUC(0–24h) following s.c. dosing at 10 MU/kg. For pharmaco-dynamic data analysis, the increases in neopterin concentrationsat 24 h postdose relative to the predose concentrations wereplotted vs. IFN-b1a dose. The dose-response data were fittedto a sigmoid curve using GraphPad Prism Software (San Diego,CA). The 50% maximally effective s.c. dose of IFN-b1a forneopterin induction was estimated for the fitted curve. Statisti-cal comparisons between treatment groups were conducted byone-way analysis of variance (ANOVA) or nonpaired t-tests.Data in tables and figures are expressed as the mean 6 SE.

MARTIN ET AL.710

Safety evaluations

The pulmonary safety component of this study was con-ducted in cynomolgus and rhesus macaques to assess the localpulmonary tolerance and toxicity of IFN-b1a when adminis-tered intrabronchially, directed to a defined region within onelung lobe.

IFN-b1a was administered intrabronchially to a defined re-gion within one lobe of the right lung at a dose of 60 mg for 4weeks, with frequencies as described in Table 1. The total amountof IFN-b1a administered over the 4-week period was 0, 240, 720,or 1680 mg. In control animals, the IFN-b1a vehicle (HSA inPBS) was administered to the right lung, and an equivalent vol-ume of PBS was administered to the left lung. Animals were se-dated with ketamine (Ketavet®) (Pharmacia and Upjohn GmbH,Erlangen, Germany) and xylazine (Rompun®) (Bayer, Lever-kusen, Germany) prior to the pulmonary instillation procedure,and a local anesthesia (lidocaine spray, Astra GmbH, Wedel,Germany) was applied to the larynx.

Clinical signs were recorded predose, approximately 0.5, 1, and2 postdose on the day of dosing, and twice daily on the days with-out dosing. Physical examinations, including lung auscultation,were performed once predose, at the end of week 4, and at theend of the recovery period. Body weights, body temperature, andqualitative food consumption were monitored predose and at rou-tine intervals throughout the dosing and recovery periods. Bron-choscopy was performed on all animals at various times through-out the study in order to identify the precise location of thetreatments. Thoracic radiograms were taken once predose (base-line), during weeks 2 and 4, and at the end of the recovery pe-riod. Blood samples for hematology, clinical chemistry, IFN-bconcentrations, neopterin, and antibody investigations were takentwice prior to the first dose and 12 h after the first and last doses.One animal from each group was killed on day 31. The remain-ing rhesus macaques (1 per groups 1a, 2a, 3, and 4a) were killedon day 43 (4 weeks of treatment, followed by a 2-week recoveryperiod). Complete postmortem evaluations were performed, andlung weights were recorded. Lungs, tracheobronchial lymphnodes, submandibular lymph nodes, trachea, and all gross lesionsfrom each animal were fixed in 10% neutral buffered formalin.Tissues were embedded in paraffin, sectioned, stained with hema-toxylin and eosin (H & E), and examined microscopically. His-tomorphologic findings were scored visually using a 4-point scalein which 1 5 minimal, 2 5 mild, 3 5 moderate, and 4 5 severe.

RESULTS

Pulmonary safety and bioavailability following i.t. dosing

In the pharmacokinetic/pharmacodynamic component of thisstudy, animals administered IFN-b1a or vehicle by i.t. dosingshowed no overt signs of toxicity. There were no treatment-re-lated changes in physical condition, body weight, or food con-sumption. i.t. administration of IFN-b1a resulted in measurableserum IFN-b1a activity (Fig. 1). Peak serum activity was measured at 8–12 h postdose (Tmax). Comparison of the AUC(0–24 h) with that of the same dose of IFN-b1a adminis-tered s.c. indicated that the relative bioavailability by this routewas approximately 10% over the first 24 h (Table 2). When thes.c. dose was reduced by 10-fold to 1 MU/kg, the AUC valuesdecreased by 24-fold in males and by 28-fold in females (Table2). The AUC and Cmax values for the i.t. dose of 10 MU/kgwere, therefore, intermediate between those obtained followings.c. doses of 1 and 10 MU/kg (Fig. 1). There was no signifi-cant difference in the pharmacokinetic parameters obtainedfrom male and female macaques following s.c. dosing.

Serum neopterin concentrations were elevated in the 3 rhe-sus macaques dosed i.t. with 10 MU/kg (50 mg/kg) IFN-b1abut not in the 1 animal receiving vehicle i.t. (Fig. 2). Mean(6SE, n 5 3) neopterin levels increased from 1.6 6 0.21 ng/mlprior to treatment to a peak of 12.5 6 1.56 ng/ml 24 h post-dose. Following s.c. administration of IFN-b1a at doses rang-ing from 0.25 to 10 MU/kg, there was a dose-dependent in-crease in serum neopterin concentrations (Fig. 3). The ED50 forthe neopterin response was 1.3 MU/kg. The neopterin responsesin male and females following s.c. dosing and in females fol-lowing i.t. dosing are shown in Table 2. There was a tendencyfor the neopterin responses to be greater in females than inmales, although this difference did not attain statistical signif-icance. In addition, there was no statistically significant differ-ence in the neopterin response following s.c. vs. i.t. dosing at10 MU/kg.

Pulmonary safety following intrabronchial dosing

In the pulmonary safety component of this study conductedin rhesus and cynomolgus macaques, there were no deaths andno IFN-b1a-related effects on body weight, body temperature,

PULMONARY AND SUBCUTANEOUS HuIFN-b1a DOSING 711

TABLE 1. DESIGN FOR INTRABRONCHIAL SAFETY EVALUATION STUDY

Number of Number ofanimals/ Group Dose level doses per Total IFN-b1a

Group species designation (mg) week (mg) Terminala Recoveryb

1 1 cynomolgus Control 0 7 0 1 01a 2 rhesus Control 0 7 0 1 12 1 cynomolgus Low 60 1 240 1 02a 2 rhesus Low 60 1 240 1 13 2 rhesus Intermediate 60 3 720 1 14 1 cynomolgus High 60 7 1680 1 04a 2 rhesus High 60 7 1680 1 1

aAt least 72 h after the last dose.b2 weeks after the last dose.

Necropsy

hematology, clinical chemistry, thoracic radiographs, or organweights. The following clinical observations were noted: gasp-ing respiration after bronchoscopy, moist rales in the lungs, in-creased respiration, and epithelial lesions and excessive mucussecretion of the bronchi and larynx (identified by bron-choscopy). Based on the similarity of observations amongtreated and control animals, these observations were consideredto be related to the administration procedure not to the test ma-terial. Enlarged tracheobronchial lymph nodes were also de-tected in both treated and control animals. Thus, the instillationprocedure and HSA alone were considered sufficient to inducethis local lymph node enlargement. Local lymph node enlarge-ment has also been seen in rhesus macaques following s.c. ad-ministered HuIFN-b1a and HSA (results not shown).

Extensive histomorphologic evaluation of pulmonary tissueswas conducted on each animal. Evaluation of H & E-stainedpulmonary tissues did not reveal any IFN-b1a-specific effectsin the low-dose (60 mg once per week) or middose (60 mg threetimes per week) groups. Findings present in these two groups

were equivalent to findings in control groups (HSA or saline).Overall, the pulmonary changes in these animals were consis-tent in character and severity across the groups (Tables 3 and4). This is further illustrated in Figure 4, in which the meancomposite scores from groups 1a and 1 have been combinedand the mean composite scores from groups 2, 2a (terminal),and 3 (terminal) have been combined. Figure 4 shows that thelung histomorphologic findings seen in animals receiving IFN-b1a at doses of 60–180 mg/week were not significantly differ-ent from the findings in vehicle-treated animals.

Histomorphologic observations were characterized by lym-phohistiocytic interstitial infiltrates, alveolar histiocytosis, andhyperplasia of bronchus-associated lymphoid tissues (BALT).Responses in these animals were considered nonspecific andconsistent with instillation of a foreign protein (i.e., HSA orHuIFN-b1a) or minor injury associated with the liquid instilla-tion procedure. One rhesus macaque in the high-dose group (60mg daily for 4 weeks), terminated at 4 weeks, had minimal tomoderate eosinophilic alveolitis. The changes in this animal

MARTIN ET AL.712

FIG. 1. Comparison of serum IFN-b levels following s.c. and i.t. administration to rhesus macaques.

TABLE 2. PHARMACOKINETIC PARAMETERS FOLLOWING SUBCUTANEOUS AND INTRATRACHEAL ADMINISTRATION

OF 1 MU/KG OR 10 MU/KG HUMAN IFN-b1a IN RHESUS MACAQUES

Intratracheal dose

10 MU/kg

Gender Male Female Male Female FemaleAUC(0–24 h) 3682 3350 89218 94343 10308

(U*h/ml) 61033 6458 69660 616252 61939AUC(0–last) — — — — 14323

(U*h/ml) 62672Peak IFN-b level 240 239 6300 6109 666

(U/ml) 653 626 6933 61096 6131Peak neopterin response 1.57 3.24 4.57 6.83 10.90

(ng/ml) 60.81 61.14 60.55 62.39 61.50Tmax (observed) 5.0 4.8 7.6 6.1 10.7

(h) 61.0 60.8 60.6 60.7 61.3

Subcutaneous dose

1 MU/kg 10 MU/kg

were clearly more severe and were considered possibly an iso-lated type I hypersensitivity response to HSA or IFN-b1a. Par-tial resolution of the pulmonary infiltrates was seen in all rhe-sus macaques killed at 2 weeks after cessation of treatment(Table 4). Lesions in the group 4 (high-dose) rhesus macaquekilled after 2 weeks recovery were minimal and indistinguish-able from the other treatment and control groups, suggestingthat if this animal had an eosinophilic response to treatmentsimilar to that of the terminally sacrificed animal in this group,near complete resolution occurred during the recovery period.

Serum IFN-b concentrations were below the limit of detec-tion of the assay (7.8 or 16 U/ml for rhesus, 15 U/ml forcynomolgus serum) in all animals prior to IFN-b1a treatment.IFN-b concentrations were detected in serum at 12 h after thefirst dose in all IFN-b1a-treated animals, indicating systemicexposure to IFN-b following pulmonary instillation (Table 5).Serum neopterin concentrations were below the limit of detec-tion of the assay (0.52 ng/ml for rhesus, 1.6 ng/ml for cynomol-gus) at all times in control animals and prior to IFN-b1a ad-ministration in all animals. Serum neopterin concentrationswere elevated at 12 h after the first dose in all animals receiv-ing IFN-b1a (Table 5). The majority of IFN-b1a-treated ani-mals developed low titers of neutralizing antibodies to IFN-b1aafter a 4-week dosing period. The appearance of neutralizingantibodies was associated with a reduction in the systemic con-centrations of both IFN-b and neopterin. However, at 12 h fol-lowing the last dose administration in week 4, serum neopterinconcentrations remained elevated above baseline in all treatedanimals. The presence of a low titer of neutralizing antibodiesin the serum did not appear to correlate with substantive dif-ferences in lung histopathology.

DISCUSSION

This study shows that IFN-b1a is systemically absorbed fol-lowing administration to the lungs and that no significant lungtoxicity attributable to administration of IFN-b1a was observed.

Intrabronchial delivery of local high concentrations of a liquidformulation of IFN-b1a once per week or three times per weekwas associated with mild alveolitis comparable in character andseverity with changes observed in vehicle control-treated ani-mals.

Pulmonary delivery of IFN-b1a (10–50 mg/kg) was associ-ated with detectable serum levels of IFN-b1a and increases inthe biologic response marker, neopterin. Serum neopterin con-centrations were not elevated following intratracheal or intra-bronchial delivery of saline or HSA, indicating that theneopterin elevations were induced by IFN-b1a and were not in-duced by the procedure itself. The increases in the neopterinconcentrations confirm that the rhesus macaque is biologicallyresponsive to HuIFN-b1a. One unexpected finding from thisstudy was that the cynomolgus macaque also appeared to be bi-ologically responsive to HuIFN-b1a. Previously, it was con-sidered that the rhesus macaque is the only species that respondsto HuIFN-b1a.

PULMONARY AND SUBCUTANEOUS HuIFN-b1a DOSING 713

FIG. 2. Serum neopterin concentrations following i.t. administration of 50 mg/kg of IFN-b1a or vehicle to rhesus macaques.

FIG. 3. Dose-response relationship for neopterin induction 24h following subcutaneous administration of human IFN-b1a torhesus macaques.

Comparison of the systemic exposure levels (AUC) follow-ing i.t. versus s.c. administration demonstrates that the relativebioavailability of IFN-b1a following pulmonary delivery is atleast 10%. This estimation is based on the assumption that 100%of the dose administered i.t. was deposited on the lung. Previ-ous studies have shown that the bioavailability of HuIFN fol-

lowing s.c. injection in rhesus macaques is 100%.(3) Thebioavailability of HuIFN-b1a in the nonhuman primate is sig-nificantly greater than that previously demonstrated in rats, inwhich bioavailability following i.t. dosing was only 1.4%.(3)

The overall pharmacokinetic profile of IFN-b was similarfrom both the s.c. and i.t. routes of administration. The serum

MARTIN ET AL.714

TABLE 3. HISTOMORPHOLOGIC EVALUATION OF LUNG FROM CYNOMOLGUS MACAQUES

FOLLOWING INTRABRONCHIAL INSTILLATION OF IFN-b1a

Selected findings: cynomolgus macaques

Group 2 Group 4

Treatment Saline HSA IFN-b1a IFN-b1a

Middle lobeMononuclear cell, septae (I)a 1b 1 1Alveolar macrophages (H) 1 2 2

Diaphragmatic lobeMononuclear cell, septae (I) 2 1 2BALTc (H) 2 1Alveolar macrophages (H) 1 1 2 1

Composite score 2 8 7 4

aI, infiltrate; H, hyperplasia.b1, minimal; 2, mild; 3, moderate; 4, severe.cBALT, bronchus-associated lymphoid tissues.

Group 1

TABLE 4. HISTOMORPHOLOGIC EVALUATION OF LUNG FROM RHESUS MACAQUES

FOLLOWING INTRABRONCHIAL INSTILLATION OF IFN-b1a

Selected findings: rhesus macaques

Terminal sacrifice groups Recovery sacrifice groups

Group 1a Group 2a Group 3 Group 4a Group 1a Group 2a Group 3 Group 4a

Treatment Saline HSA IFN-b1a IFN-b1a IFN-b1a Saline HSA IFN-b1a IFN-b1a IFN-b1a

Lymph nodesLymphoid hyperplasia 2a 2 1 1

TracheaInfiltrate, eosinophilis 1

Right second bifurcationInfiltrate, eosinophilis 1 1

Middle lobeEosinophils (I)b 1 2 1a 1 1Mononuclear cell, septae (I) 1 1 1a 1 2 1 2 1 1BALTc (H) 1a

Alveolar macrophages (H) 1 1 1a 1 1Alveolar epithelium (H) 1a

Diaphragmatic lobeEosinophils (I) 1 1a 1 3 1 1Mononuclear cell, septae (I) 1 1 2 1 2 2BALT (H) 1 1a 2 1Alveolar macrophages (H) 1 1 3 1Alveolar epithelium (H) 2

Composite score 5 9 9a 5 19 0 6 6 2 2

a1, minimal; 2, mild; 3, moderate; 4, severe.bI, infiltrate; H, hyperplasia.cBALT, bronchus-associated lymphoid tissues.

concentrations of IFN-b were approximately 10-fold lower af-ter i.t. delivery of 10 MU/kg vs. s.c. delivery of 10 MU/kg.However, when compared with a 1 MU/kg s.c. dose, the serumlevels of IFN-b following i.t. administration of 10 MU/kg wereapproximately 3-fold greater. It can, therefore, be deduced thatsystemic exposure to IFN-b1a following an i.t. dose of 10MU/kg is equivalent to an s.c. dose of between 1 and 10 MU/kg.The apparent lack of dose proportionality in serum exposureslevels following s.c. dosing is possibly due to the assay method,which reports activity in 2-fold increments rather than as a con-

tinuous series. The dose-response curve for neopterin inductionfollowing s.c. dosing shows that at doses in excess of 3 MU/kg,the neopterin response would be expected to be .80% of themaximum obtainable response. Therefore, the observation thatthe neopterin response following i.t. administration of 10MU/kg is indistinguishable from the neopterin response fol-lowing s.c. administration of 10 MU/kg is not inconsistent withthe described dose-response relationship. The possibility thatsome additional neopterin may have been released from mastcells within the lung following local high IFN-b1a exposurecannot be excluded.

Overall, the pharmacokinetic component of this study dem-onstrated that HuIFN-b1a is absorbed systemically followingadministration to the primate lung. The CPE assay used to de-tect IFN in serum measures IFN antiviral activity and is, there-fore, a measure of biologically active IFN levels. The use ofthis assay demonstrated that at least 10% of the active IFN in-stilled into the lung was absorbed as active IFN. Although someIFN is likely inactivated within the lung, a sufficient amount ofactive IFN is absorbed to make this route of administration aviable alternative to s.c. or i.m. injections. Studies conductedin ventilated, blood-perfused, isolated rabbit lungs have shownthat HuIFN-a is rapidly catabolized when administered to thelung even though the lung does not contribute to the catabo-lism of systemic IFN.(14) Studies conducted in rats, however,have shown that the inclusion of a protease inhibitor did not af-fect the bioavailability of i.t. administered IFN-a.(5) Therefore,the pulmonary catabolism of human IFNs may depend on thespecies used. The extent of local catabolism of human IFNs inprimate lungs is unknown.

The i.t. administration used for the bioavailability compo-nent of this study used a spray to administer IFN to the lungs.In the absence of an optimized inhalation formulation, this tech-

PULMONARY AND SUBCUTANEOUS HuIFN-b1a DOSING 715

TABLE 5. INDIVIDUAL ANIMAL SERUM IFN-1b CONCENTRATIONS, ANTI-IFN-b ANTIBODY TITERS, AND

NEOPTERIN CONCENTRATIONS AFTER INTRABRONCHIAL ADMINISTRATION OF 60 mG IFN-b1a

Numberof Anti-IFN-bdoses neutralizingper 12 h after 12 h after antibody 12 h after 12 h afterweek Species Predose first dose last dose titer Predose first dose last dose

0 Rhesus ,16a ,16 ,16 NDb ,0.52 ,0.52 ,0.520 Rhesus ,16 ,16 ,16 ND ,0.52 ,0.52 ,0.520 Cynomolgus ,15 ,15 ,15 ND ,1.6 ,1.6 ,1.6

1 Rhesus ,16 44 ,16 16 ,0.52 1.8 0.981 Rhesus ,16 330 19 0 ,0.52 1.9 0.761 Cynomolgus ,15 68 230 2 ,1.6 2.6 6.1

3 Rhesus ,16 140 ,7.8 47 ,0.52 1.5 0.523 Rhesus ,16 95 ,7.8 2 ,0.52 1.7 1.3

7 Rhesusc ,16 110 ,7.8 2 ,0.52 2.8 1.17 Rhesus ,16 46 ,7.8 2 ,0.52 1.5 0.957 Cynomolgus ,15 90 ,15 15 ,1.6 4.6 2.6

aAll numbers preceded by , indicate values below the limit of detection of the assay.bND, not determined.cAnimal exhibiting eosinophilic alveolitis.

Serum IFN-b concentration (pg/ml)

Serum neopterin concentration (ng/ml)

FIG. 4. Mean composite score for histomorphologic findingsin the lungs of rhesus and cynomolgus macaques. For purposesof comparison, all HSA-treated control animals have been com-bined and all terminal sacrifice animals receiving doses of60–180 mg/week have been combined (n 5 3 per group).

nique was used in an attempt to distribute the IFN over a largearea of the lung surface. However, the liquid spray that wasused would be predicted to produce only large liquid dropletsrather than fine particles. Studies have shown that large liquiddroplets are deposited primarily within the upper respiratorytract.(15,16) For optimal pulmonary absorption, deposition withinthe alveoli is preferable because of the large alveolar surfacearea and the high degree of vascularization. Absorption acrossthe alveolar epithelium would be expected to result in greatersystemic exposure to IFN-b1a. For deposition within the deeplung, small aerosol particles are required. Therefore, with thedevelopment of an optimized aerosol formulation, the bioavail-ability of HuIFN-b1a potentially could be greater than that seenfollowing pulmonary instillation. Studies conducted in rats haveshown that the bioavailability of HuIFN-a is only about 15%when administered by instillation but is increased to about 70%when administered as an aerosol.(5)

In the safety component of this study, macaques received 60mg IFN-b1a into a defined region within one lobe of the lung.When adjusted for body weight, this dose corresponds to 10mg/kg. The current clinical dose of IFN-b1a is 30 mg i.m., whichcorrespond to 0.43 mg/kg for a 70-kg human. If the pulmonarybioavailability of IFN-b1a in humans is 10%, it can be assumedthat a lung deposition of 4.3 mg/kg IFN-b1a would be requiredto produce serum concentrations equivalent to a 30-mg i.m.dose. The once per week dose administered to a single sitewithin the macaque lung (10 mg/kg) is, therefore, at least 2-foldgreater than the projected human lung exposure (4.3 mg/kg),which for an optimized aerosol formulation, would be assumedto be distributed evenly across the entire lung surface. The threetimes per week dosing and the seven times per week dosing ex-pose a single region of the macaque lungs to weekly doses ofIFN-b1a that are in excess of 6-fold and 14-fold the anticipatedhuman total lung exposure, respectively.

In control, low-dose, and middose rhesus macaques, andcynomologus macaques in all groups, pulmonary instillation ofPBS, HSA, or IFN-b1a caused minimal to mild subchronicalveolitis characterized by lymphohistiocyte interstitial infil-trates, alveolar histiocytosis, and hyperplasia of BALT.Changes were located primarily near the sites of instillation.Responses in these animals were thus considered nonspecificand consistent with instillation of a foreign protein (i.e., HSAor IFN-b1a) or minor injury associated with the liquid instilla-tion procedure. In one high-dose rhesus macaque (60 mg dailyfor 4 weeks), IFN-b1a induced mild to moderate eosinophilicalveolitis. Changes in this animal were considered possibly anisolated type 1 hypersensitivity response to HSA or IFN-b1a.Partial resolution of the pulmonary lesions was seen in all re-covery animals killed 2 weeks following cessation of treatment.The majority of IFN-b1a treated animals developed low titersof neutralizing antibodies to IFN-b following a 4-week dosingperiod. However, the serum titers had no impact on the ani-mals’ health and did not complicate anatomic pathology inter-pretation. It is anticipated that humans would develop less ofan immune response to HuIFN-b1a administration than do non-human primates. There is a very low incidence of neutralizingantibody production in MS patients treated chronically withAVONEX, whereas in nonhuman primates, neutralizing anti-bodies are evident within 4 weeks of systemic dosing. In addi-tion, any aerosol formulation designed for human use would beunlikely to contain HSA.

In conclusion, these studies show that HuIFN-b1a is sys-temically bioavailable when administered to the lungs via i.t.or intrabronchial liquid instillation. When IFN-b1a was ad-ministered to the lungs at frequencies of one to three times perweek, no significant lung toxicity attributable to IFN-b1a wasevident. Only when the dosing frequency was increased toeveryday was one instance of a possible type I hypersensitiv-ity response observed (eosinophilic alveolitis). All histopatho-logic alterations appeared to be largely reversible within 2weeks of cessation of treatment. Overall, these studies suggestthat the pulmonary route of administration could be a suitablealternate route to the i.m. route of administration for the sys-temic delivery of HuIFN-b1a.

ACKNOWLEDGMENTS

We thank the following people for their assistance in the con-duct of these studies: Deborah Novicki, Curtis Chan, SharonBungo, Kaela Porter, Steve Vanech, Bruce Burnett, and mem-bers of the Biogen Bioassay group. These studies were spon-sored by Biogen Inc.

REFERENCES

1. JACOBS, L.D., COOKFAIR, D.L., RUDICK, R.A., HERNDON,R.M., RICHERT, J.R., SALAZAR, A.M., FISCHER, J.S., GOOD-KIN, D.E., GRANGER, C.V., and SIMON, J.H. (1996). Intra-muscular interferon-beta-1a for disease progression in relapsingmultiple sclerosis. Ann. Neurol. 39, 285–294.

2. WEINSTOCK-GUTTMAN, B., RANSOHOFF, R.M., KINKEL,R.P., and RUDICK, R.A. (1995). The interferons: biological ef-fects, mechanisms of action, and use in multiple scleroses. Ann.Neurol. 37, 7–15.

3. PEPINSKY, R.B., LEPAGE, D.J., GILL, A., CHAKRABORTY,A., VAIDYNANTHAN, S., GREEN, M., BAKER, D.P., WHAL-LEY, E., HOCHMAN, P.S., and MARTIN, P. (2001). Improvedpharmacokinetic properties of a polyethylene glycol-modified formof interferon-b1a with preserved in vitro bioactivity. J. Pharmacol.Exp. Ther. 297, 1059–1066.

4. PATTON, J.S., TRINCHERO, P., and PLATZ, R.M. (1994).Bioavailability of pulmonary delivered peptides and proteins: a-interferon, calcitonins and parathyroid hormones. J. Controlled Re-lease 28, 79–85.

5. NIVEN, R.W., WHITCOMB, K.L., WOODWARD, M., LIU, J.,and JORNACION, C. (1995). Systemic absorption and activity ofrecombinant consensus interferons after intratracheal instillationand aerosol administration. Pharmacol. Res. 12, 1889–1895.

6. KINNULA, V., MATTSON, K., and CANTELL, K. (1989). Phar-macokinetics and toxicity of inhaled human interferon-a in patientswith lung cancer. J. Interferon Res. 9, 419–423.

7. KINNULA, V., CANTELL, K., and MATTSON, K. (1990). Ef-fect of inhaled natural interferon-alpha on diffuse bronchioalveo-lar carcinoma. Eur. J. Cancer 26, 740–741.

8. MAASILTA, P., HALME, M., MATTSON, K., and CANTELL,K. (1991). Pharmacokinetics of inhaled recombinant and naturalalpha interferon. Lancet 337, 371.

9. HALME, M., MAASILTA, P.K., MATTSON, K., and CANTELL,K. (1994). Pharmacokinetics and toxicity of inhaled human naturalinterferon-beta in patients with lung cancer. Respiration 61, 105–107.

10. VAN ZANWIJK, N., JASSEM, E., DUBBELMANN, R., BRAAT,M.C., and RUMKE, P. (1990). Aerosol application of interferon-alpha in the treatment of bronchioloalveolar carcinoma. Eur. J. Can-cer 26, 738–740.

MARTIN ET AL.716

11. GIOSUE, S., CASARINI, M., AMEGLIO, F., ALEMANNO, L.,SALTINI, C., and BISETTI, A. (1996). Minimal dose of aeroso-lized interferon-alpha in human subjects: biological consequencesand side effects. Eur. Respir. J. 9, 42–46.

12. MARTIN, R.J., BOGUNIEWICZ, M., HENSON, J.E., CEL-NIKER, A.C., WILLIAMS, M., GIORNO, R.C., and LEUNG,D.Y. (1993). The effects of inhaled interferon gamma in normalhuman airways. Am. Rev. Respir. Dis. 148, 1677–1682.

13. VAIDYANATHAN, S., GREEN, J., MARTIN, P., LANE, J.,BURNETT, B., GILLETT, N., and NIGGEMANN, B. (2001). Pul-monary toxicity study of interferon-b1a when administered as aliquid instillation to rhesus and cynomolgus macaques. Toxicol.Pathol. 29, 714.

14. BOCCI, V., PESSINA, G.P., PACINI, A., PAULESU, L.,MUSCETTOLA, M., and MORGENSEN, K.E. (1984). Pulmonarycatabolism of interferons: alveolar absorption of 125I-labeled hu-man interferon alpha is accompanied by partial loss of biologicalactivity. Antiviral Res. 4, 211–220.

15. COLTHORPE, P., FARR, S.J., TAYLOR, G., SMITH, I.J., andWYATT, D. (1992). The pharmacokinetics of pulmonary-deliv-

ered insulin: a comparison of intratracheal and aerosol administra-tion to the rabbit. Pharmacol. Res. 9, 764–768.

16. SCHLESINGER, R.B. (1985). Comparative deposition of inhaledaerosols in experimental animals and humans: a review. J. Toxi-col. Environ. Health 15, 197–214.

Address reprint requests to:Dr. Pauline L. Martin

Department of Preclinical DevelopmentBiogen, Inc.

14 Cambridge CenterCambridge, MA 02142

Tel: (617) 679-2975Fax: (617) 679-3463

E-mail: pauline_martin@biogen.com

Received 4 February 2002/Accepted 19 March 2002

PULMONARY AND SUBCUTANEOUS HuIFN-b1a DOSING 717

This article has been cited by:

1. Sebastien Vallee, Swapnil Rakhe, Thomas Reidy, Sandra Walker, Qi Lu, Paul Sakorafas, Susan Low,Alan Bitonti. 2012. Pulmonary Administration of Interferon Beta-1a-Fc Fusion Protein in Non-HumanPrimates Using an Immunoglobulin Transport Pathway. Journal of Interferon & Cytokine Research 32:4,178-184. [Abstract] [Full Text HTML] [Full Text PDF] [Full Text PDF with Links]

2. Fernanda Andrade, Mafalda Videira, Domingos Ferreira, Bruno Sarmento. 2011. Nanocarriers forpulmonary administration of peptides and therapeutic proteins. Nanomedicine 6:1, 123-141. [CrossRef]

3. Mónica Lahoz, Marcelo Andrés Kauffman, Julio Carfagnini, Alejandro Vidal, Mariana Papouchado, AídaSterin-Prync, Roberto Alejandro Diez, Carlos Nagle. 2008. Pharmacokinetics and pharmacodynamics ofinterferon beta 1a in Cebus apella. Journal of Medical Primatology . [CrossRef]

4. John Thipphawong. 2006. Inhaled cytokines and cytokine antagonists. Advanced Drug Delivery Reviews58:9-10, 1089-1105. [CrossRef]