SHORT REPORT

Development of anti-interferon antibodies and breakthroughhepatitis during treatment for HCV infection in haemophiliacs

JOHN P. HANLEY,1 LISA M. JARVIS,2 PETER SIMMONDS2

AND CHRISTOPHER A. LUDLAM1 1Department of Haematology,

Royal Infirmary of Edinburgh, and 2Department of Medical Microbiology, University of Edinburgh

Received 28 February 1996; accepted for publication 16 May 1996

Summary. The development of anti-interferon antibodies maylead to treatment failure during interferon therapy. We havestudied the development of such antibodies in a group of 39haemophiliacs receiving interferon-®2a for chronic hepatitisC virus (HCV) infection.

Anti-interferon antibodies developed in five (13%) patientsand were associated with ‘breakthrough hepatitis’ in threecases. There was an association between the development ofanti-interferon antibodies and infection with HCV genotype3a (P�0 .01). This study suggests that the development of

anti-interferon antibodies may lead to treatment failure in aproportion of haemophiliacs with HCV infection. Theassociation with genotype 3a has not previously beenreported. Monitoring for the development of breakthroughhepatitis due to anti-interferon antibodies may provide theopportunity to develop strategies to overcome their effects.

Keywords: haemophilia, hepatitis C virus, genotype,interferon, anti-interferon antibodies.

Alpha interferon is used to treat hepatitis C virus (HCV)infection and response may be assessed both by serialmeasurement of serum alanine transaminase (ALT) or HCVRNA quantitation by polymerase chain reaction (PCR). Thosewho respond to interferon usually show a prompt normal-ization of ALT and clearance of serum HCV RNA within 8–12weeks of treatment (Hanley et al, 1996). In non-respondersALT and HCV RNA remain unchanged.

In some individuals ‘breakthrough hepatitis’ may occur, i.e.initial response with normalization of ALT followed by asustained rise in ALT, during interferon therapy. It has beensuggested that such episodes of ‘breakthrough hepatitis’may be caused by anti-interferon antibodies (Milella et al,1993). Therefore we investigated the development of anti-interferon antibodies in a group of 39 haemophiliacsreceiving interferon-�2a for chronic HCV infection.

PATIENTS AND METHODS

Patient characteristics. A total of 39 patients (37 male, twofemale) were studied (26 haemophilia A, nine haemophilia B,four von Willebrand’s disease). The median age was 35.5years (range 13–67 years). Six patients were anti-HIVpositive and all were HBSAg negative.

All patients were anti-HCV positive by Abbott second-generation enzyme immunoassay (A-EIA; Abbott, Weisbaden-Dalkenheim, Germany) and also positive on confirmatorytesting by second-generation recombinant immunoblot assay(RIBA-2, Chiron Corporation, Emeryville, Calif.). All hadpreviously received non-virus-inactivated coagulation factorconcentrates prior to 1985 and 38/39 had persistentlyelevated serum ALT levels (normal range 10–40 U/l).HCV RNA was detected in 38/39 patients by RT-PCR( Jarvis et al, 1994) and quantified by a limiting dilutiontechnique (Simmonds et al, 1990). The lower limit ofdetection using this assay was 80 HCV copies/ml. HCVgenotyping was performed as described previously (Davidsonet al, 1995).

Drug dosage and administration. Interferon-�2a (Roche) 3Mega Units three times per week was given subcutaneously.The intention was to treat for 24 weeks. Response wasassessed on the basis of serial monthly ALT levels and HCVRNA quantitation.

Measurement of anti-interferon antibodies. Serum samplesfor anti-interferon antibody assays were taken and storedfrom all patients at 0, 12 and 24 weeks. All sampleswere tested at the end of treatment. No response or geno-type data was provided to the laboratory performing theanti-interferon antibody assays. All samples were testedusing an enzyme immunoassay (EIA) which detects bindingantibodies (Hennes et al, 1987). Those samples whichwere positive by EIA were also assessed by an antiviral

British Journal of Haematology, 1996, 94, 551–556

551# 1996 Blackwell Science Ltd

Correspondence: Dr John P. Hanley, Department of Haematology,Royal Infirmary of Edinburgh, Lauriston Place, EdinburghEH3 9YW.

neutralization bioassay (ANB) which demonstrates theneutralizing properties of the antibody (Hennes et al, 1987).

RESULTS

Binding anti-interferon antibodies were detected by EIA in five(13%) patients (Table Ia). Neutralizing properties weredemonstrated by ANB in 3/5 patients. The timing of antibodydevelopment and antibody titre varied (Fig 1). Breakthroughhepatitis occurred in three patients (nos. 1–3) who developedantibodies and this was accompanied by breakthroughviraemia in two of these (nos. 1 and 2). Breakthrough

552 Short Report

# 1996 Blackwell Science Ltd, British Journal of Haematology 94: 551–556

Fig 1. Timing of anti-interferon antibody development in relation to ALT (——) and HCV RNA (ÿ ÿ ÿ) levels during interferon treatment inpatients 1–5. EIA (IBU/ml): enzyme immunoassay (interferon binding units/ml); ANB (INU/ml): anti-viral neutralization bioassay (interferonneutralizing units/ml).

Table I(b). HCV genotype and anti-interferonantibody development showing a significantassociation with genotype 3a (P � 0:01).

Anti-interferonGenotype antibodies Total

1a 0 151b 1 92a 0 12b 0 23a 4 11

Table I(a). Characteristics of the five patients who developed anti-interferon antibodies.

Anti-interferon antibodyAge HCV ALT HCV RNA Duration of antibody

Patient (yr) genotype response response EIA ANB persistence (months)

1 16 3a Yes� BTH Yes� BTV Pos Pos 62 38 3a Yes� BTH Yes� BTV Pos Neg 43 13 1b Yes� BTH No Pos Pos 44 67 3a Yes Yes Pos Pos 55 33 3a No Yes Pos Neg 6

BTH: breakthrough hepatitis; BTV: breakthrough viraemia.

553Short Report

# 1996 Blackwell Science Ltd, British Journal of Haematology 94: 551–556

Fig 1 (continued). Patient 2.

Fig 1 (continued). Patient 3.

554 Short Report

# 1996 Blackwell Science Ltd, British Journal of Haematology 94: 551–556

Fig 1 (continued). Patient 4.

Fig 1 (continued). Patient 5.

hepatitis and viraemia occurred in only one additional patientwithout evidence of anti-interferon antibodies which wasthought to be due to intermittent compliance with interferontherapy (data not shown).

There was a significant association with HCV genotype3a (chi-square test: P � 0:01) (Table Ib). There were nochanges in HCV genotype detected in any patients duringinterferon treatment.

Follow-up information. After cessation of interferon therapy,follow-up assays were performed in those individuals whodeveloped antibodies. Within 4–6 months, both binding andneutralizing antibodies had either reduced to low titre orbecome undetectable (Table I).

DISCUSSION

Interferon therapy for HCV infection in non-haemophiliacsleads to a sustained response in approximately 25% ofindividuals. Recent studies have shown that the response tointerferon in haemophiliacs with HCV infection is inferior anda standard treatment schedule (3 MU three times per week for6 months) leads to few sustained responses (Telfer et al, 1995;Hanley et al, 1996). Factors which influence response tointerferon include age, duration of infection, severity of liverdisease, pre-treatment virus load and HCV genotype. Evalua-tion of treatment protocols using higher doses or moreprolonged courses of interferon in combination with otheranti-viral drugs have yet to be reported in haemophiliacs.

We have shown that the development of anti-interferonantibodies leading to breakthrough hepatitis may contributeto the poor response to interferon in haemophiliacs. Theimportance of timing of antibody development in relation tothe response to interferon has been noted by others (Bonetti etal, 1994). In our study, antibodies which appeared at the sametime as ALT normalization (nos. 1–3) or HCV RNA clearance(nos. 1 and 2) led to the development of breakthroughhepatitis or viraemia. On the other hand, antibodies whichappeared many weeks after the therapeutic response (nos. 4and 5) were not associated with breakthrough. It is possible,however, that the late appearance of antibody was a factor inthe early relapse seen in patient 4 after stopping interferon.

The association shown in this study between anti-interferon antibody development and infection with HCVgenotype 3a has not previously been reported. This observa-tion may be particularly important, because genotype 3a isusually associated with a favourable response to interferon. Itis also interesting to note that of the five individuals whodeveloped anti-interferon antibodies, four were aged <40years. Anti-interferon antibodies are unusual in youngerpatients (Porres et al, 1989) and haemophiliacs may be atincreased risk of antibody formation due to immunologicalabnormalities, irrespective of HIV infection (Watson &Ludlam, 1992).

It is important to recognize that breakthrough hepatitismay occur for reasons other than the development of anti-interferon antibodies, such as does reduction necessitated byside-effects or due to poor compliance. In addition, there issome evidence that each HCV genotype circulates as aheterogenous group of variants (terms ‘quasi-species’) which

evade host immune responses by continual evolution of newmutants (Enomoto et al, 1994). In theory, when interferon isgiven, initial response may be overcome by the rapidevolution of quasi-species (without a change in the majorcirculating genotype). It has also been reported that the majorcirculating genotype may change during interferon treatment(Devereux et al, 1995). We, however, found no evidence ofchanging genotypes in this study.

The clinical importance of anti-interferon antibodies is notyet fully understood; however, the development of break-through hepatitis in individuals with factors usually asso-ciated with a favourable response to interferon (e.g. youngage, genotype 3a, low pre-treatment virus load, and absenceof cirrhosis) is particularly important to detect. Anti-interferon antibody assays should be performed when break-through hepatitis develops during interferon treatment. Therecognition of antibody development may lead to theavoidance of treatment failure either by dose escalation orswitching to lymphoblastoid interferon which may not crossreact with the antibody. The effectiveness of such approachesis yet to be evaluated.

ACKNOWLEDGMENTS

We are grateful to Mrs U. Hennes (Roche) who performed theanti-interferon antibody assays and to Dr Henry Watson fordiscussion of the manuscript. This study was supported by theMedical Research Council and Roche Products Ltd.

REFERENCES

Bonetti, P., Diodati, G., Drago, C., Casarin, C., Scaccabarozzi, S., Realdi,G., Ruol, A. & Albertim, A. (1994) Interferon antibodies in patientswith chronic hepatitis C virus infection treated with recombinantinterferon alpha-2�. Journal of Hepatology, 20, 416–420.

Davidson, F., Simmonds, P., Ferguson, J.C., Jarvis, L.M., Dow, B.C.,Follet, E.A.C., Seed, C.R.G., Krusius, T., Lin, C., Medgyesi, G.A.,Kiyokawa, H., Olim, G., Duraisamy, G., Cuypers, T., Saeed, A.A.,Teo, D., Conradie, J., Kew, M.C., Nuchaprayoon, C., Ndimbie, O.K. &Yap, P.L. (1995) Survey of major genotypes and subtypes ofhepatitis C virus using restriction fragment length polymorphism ofsequences amplified from the 50 non-coding region. Journal ofGeneral Virology, 76, 1197–1204.

Devereux, H., Telfer, P., Dusheiko, G. & Lee, C. (1995) Hepatitis Cgenotypes in haemophilic patients treated with alpha-interferon.Journal of Medical Virology, 45, 284–287.

Enomoto, N., Kurosaki, M., Tanaka, Y., Marumo, F. & Sato, C. (1994)Fluctuation of hepatitis C virus quasispecies in persistent infectionand interferon treatment revealed by single-strand conformationpolymorphism analysis. Journal of General Virology, 75, 1361–1369.

Hanley, J.P., Jarvis, L.M., Andrews, J., Dennis, R., Hayes, P.C., Piris, J.,Lee, R., Simmonds, P. & Ludlam, C.A. (1996) Interferon for chronichepatitis C infection in hemophiliacs: influence of virus load,genotype and liver pathology on response. Blood, 87, 1704–1709.

Hennes, U., Jucker, W., Fischer, E.A., Krummenacher, T., Palleroni,A.V., Trown, P.W., Linder-Ciccolunghi, S. & Rainisio, M. (1987) Thedetection of antibodies to recombinant interferon alfa-2a in humanserum. Journal of Biological Standardization, 15, 231–244.

Jarvis, L.M., Watson, H.G., McOmish, F., Peutherer, J.F., Ludlam, C.A.& Simmonds, P. (1994) Frequent reinfection and reactivation of

555Short Report

# 1996 Blackwell Science Ltd, British Journal of Haematology 94: 551–556

hepatitis C virus genotypes in multitransfused haemophiliacs.Journal of Infectious Diseases, 170, 1018–1022.

Milella, M., Antonelli, G., Santantonio, T., Currenti, M., Monno, L.,Mariano, N., Angarano, G., Dianzani, F. & Pastore, G. (1993)Neutralizing antibodies to recombinant alpha-interferon and responseto therapy in chronic hepatitis C virus infection. Liver, 13, 146–150.

Porres, J.C., Carreno, V., Ruiz, M., Marron, J.A. & Bartolome, J. (1989)Interferon antibodies in patients with chronic HBV infectiontreated with recombinant interferon. Journal of Hepatology, 8,351–357.

Simmonds, P., Zhang, L.Q., Watson, H.G., Rebus, S., Ferguson, E.D.,Balfe, P., Leadbetter, G.H., Yap, P.L., Peutherer, J.F. & Ludlam, C.A.(1990) Hepatitis C quantification and sequencing in bloodproducts, haemophiliacs and intravenous drug users. Lancet, 336,1469–1472.

Telfer, P., Devereux, H., Colvin, B., Hayden, S., Dusheiko, G. & Lee, C.(1995) Alpha interferon for hepatitis C virus infection inhaemophilic patients. Haemophilia, 1, 54–58.

Watson, H.G. & Ludlam, C.A. (1992) Immunological abnormalities inhemophiliacs. Blood Reviews, 6, 26–33.

556 Short Report

# 1996 Blackwell Science Ltd, British Journal of Haematology 94: 551–556