JOURNAL OF INTERFERON RESEARCH 8:755-763 (1988)Mary Ann Liebert, Inc., Publishers

Elevation of 2',5'-01igoadenylate SynthetaseActivity and HLA-I Associated 02-Microglobulin

in Response to Recombinant Interferon- Administration in ChronicHBeAg-Positive Hepatitis

JUAN A. QUIROGA, IGNACIO MORA, JUAN C. PORRES,and VICENTE CARREÑO

ABSTRACT

We have analyzed the immunomodulatory effect of 5 and 2 MU of recombinant inter¬feron- (rIFN- ) administered to 8 carriers of HBsAg with histologically proven chronicactive liver disease. After the rIFN- administration, T8 lymphocyte subsets showed a

significant decrease (basal vs. 4 weeks, < 0.05) and T4/T8 ratios were higher than thebasal values in 6/8 patients. Serum levels of the HLA class I-associated /32-microglobulinincreased significantly in all patients within the first week of treatment, both with thehigh (p < 0.01) and the low (p < 0.05) rIFN- dose. Then, differences between the twodoses reached statistical significance (p < 0.03). Similar results (p < 0.05) were obtainedby measuring the 2,5-oligoadenylate (2-5A) synthetase activity, co-occurring with thedecreases in HBV-DNA polymerase and HBV-DNA, although no differences were foundbetween the two doses. In addition, levels of 2-5A synthetase correlated significantly withthose of ^2-microglobulin (r = 0.743, < 0.01). On the other hand, after the rIFN- ad¬ministration, all the patients had liver membrane antibodies (LMA) in their serum (p <0.05); only two patients (who were anti-HD positive) showed LMA at the end of thefollow-up. rIFN-7 has both antiviral and immunomodulatory effects in HBeAg carrierswith chronic liver disease.

INTRODUCTION

The major msTocoMPATTBnjTY complex class I antigens (HLA-I) are noncovalently associatedto a peptide, /32-microglobulin, that is necessary for the A, B, and C antigens to be displayed on

the cellular membrane. The expression of these antigens may be modified by the viral action, andplay an important role in the recognition and further lysis of infected or altered cells."·2' Similarly,it is known that interferon (IFN) is able to modify HLA-I expression on the hepatocellular mem-

Department of Gastroenterology, Fundación Jiménez Diaz, Universidad Autónoma, Madrid, Spain.

755

QUIROGA ET AL.

brane.'3' On the other hand, IFNs may be powerful antireplicative drugs, the antiviral effect ofwhich could be shown by activation of the 2',5'-oligoadenylate (2-5) synthetase activity.<4) This en¬

zyme has been used recently to monitor IFN- therapy in viral infections, <s·6' and may be an usefulclinical marker for IFN- administration.

In the last few years, human or recombinant (r) IFN- has been in widespread use in the treat¬ment of chronic hepatitis virus (HBV) infection.'" Recently, rIFN- has been introduced for thispurpose, alone or in combination with other drugs.'8·" This type of IFN may have a higher im-munomodulator potential, besides its own antiviral effect. In this study, we have analyzed the pos¬sible immunomodulator effect of rIFN-7 in chronic hepatitis type B. We have measured thechanges in peripheral blood lymphocytes (and in the helper to cytotoxic/suppressor cell ratiosas well), /J2-microglobulin, the presence or absence of nonorgan-specific autoantibodies and circu¬lating autoantibodies to the liver membrane (LMA), as well as the levels of 2-5A synthetase activityin sera from carriers of HBsAg with chronic liver disease treated with rIFN- .

MATERIALS AND METHODS

Patients: Eight male, Caucasian patients with chronic HBV infection, whose mean age was 25years (range 18-47), were studied; they showed a positive result to HBeAg, HBV-DNAp, and HBV-DNA. According to international criteria,"01 all of them had an histologically proven chronic ac¬tive hepatitis (CAH) with liver cirrhosis in one case, and had abnormal liver function tests. None ofthe patients were homosexuals and gave a negative result to the antibody against type 1 human im¬mune deficiency virus (anti-HIV-1). All the patients had been HBsAg carriers for at least 1 year(range 1-8). The onset of the carrier state usually was not known, although three patients sufferedfrom a transfusion and five were drug abusers (one with an acute hepatitis episode) in whom theabsence of drug addiction lasted for at least 6 months (range 6-36) prior to the study. Five patientswere positive to anti-HD, although the titer (1/5,000 in two cases and 1/100 in three) remained un¬

changed for one year before the study. Three of them had HDAg in the liver biopsy.Patients were randomly allocated to one of two groups. They received one daily dose of either 5

MU (group 1, = 4) or 2 MU (group 2, = 4) of rIFN- (Boehringer Ingelheim, FRG) per m2 ofbody surface by i.m. injection for 28 days. Serum samples were collected before the first dose,every week during the treatment, and monthly thereafter until the 12th month. This study was ap¬proved by the Hospital Ethics Committee, and informed consent was obtained from the patients.

Laboratory Tests: Liver function tests were performed using standard methods (SMAC-20 sys¬tem, Technicon, NY). Anti-HIV-1 antibodies were assayed by EIA (Abbott Laboratories, NorthChicago, IL).

Serum and Liver HBV and HDV Markers: HBV markers (HBsAg, anti-HBc, HBeAg, and anti-HBe) and HDV markers (anti-HD) were tested in serum by RIA (Abbott). HDAg was detected indeparaffinized liver sections by indirect immunoperoxidase staining. HBV-DNA polymerase was

determined according to Robinson."" HBV-DNA was assayed in serum by dot-blot hybridizationas previously reported."2'

2',5'-Oligoadenylate (2-5A) Synthetase: The levels of 2-5A synthetase activity were measured inserum by a sensitive radioimmunoassay (Eiken Chemical Co., Tokyo, Japan), using '"I-labeled2-5A as a probe and an antiserum to 2-5A."3' Briefly, 2-5A synthetase in human serum (50 µ ) isbound to and activated by poly(I):poly(C) agarose beads. The enzyme is measured by conversion ofATP into 2-5A, which competes with the labeled 2-5A for binding sites on the 2-5A antibody. Theamount of synthesized 2-5A (2-5A synthetase activity) is then determined by comparison with a

standard curve prepared by varying known amounts of 2-5A (0.10-8.1 pmoles/ml). All sampleswere analyzed in duplicate in the same run.

756

2-5A AND HLA-I IN RlFN- THERAPY OF CHRONIC HBV

Immunological Parameters: T3, T4 and T8 lymphocyte subsets were analyzed by indirect im-munofluorescence using monoclonal antibodies (OKT3, OKT4, and OKT8; Ortho Diagnostics,Raritan, NJ). Serum levels of rIFN- were measured by RIA (Centocor, Malvem, PA). Anti-rlFN-7 antibodies were detected by a biological assay based on the neutralization of IFN by its specificantibody on EMC virus-infected A549 cells.1"" The levels of the HLA class I-associated /32-micro¬globulin were measured in serum by RIA (Eiken); all samples were analyzed in duplicate in thesame run.

Autoantibodies: Nonorgan-specific autoantibodies [anti-nuclear (ANA), anti-smooth muscle(SMA), anti-mitochondrial (AMA), and anti-nDNA antibody] were assayed by indirect immuno-fluorescence (Bio-Dx, NJ).

Liver Membrane Antibody Assay: A detection test for LMA was performed in serum by indirectimmunofluorescence on rat perfused hepatocytes isolated by enzymatic treatment with collagenasefollowing the method of Seglen"s| with minor modifications."6' Samples showing a linear immu¬nofluorescence pattern on the hepatocytes were regarded as positive for LMA.

Statistical Analysis: Student's /, Fisher's exact, and signed rank tests were used for statisticalcomparisons, as was Pearson's correlation coefficient.

RESULTSAntiviral effect of rIFN-y: Changes in HBV markers and of 2-5A synthetase

A decrease in HBV-associated DNA polymerase activity and HBV-DNA concentrations was ob¬served in 6/8 patients during the rIFN- administration, as has been recently reported'9'; there¬after, a rebound in both HBV-DNAp and HBV-DNA was seen in five of them, irrespective of thedose given. Figure 1 shows the outcome of one of such patients. At the end of the follow-up, noneof the patients had lost HBeAg, and HBV-DNA remained detectable in all but one patient. Signifi¬cant increases were found in 2-5A synthetase levels (basal vs 1 week, < 0.05) co-occurring withthe decreases in HBV-DNA and HBV-DNAp, both in group 1 (5 MU rIFN- ) and in group 2 (2MU) (Fig. 2). The low rIFN- dose (2 MU) resulted in the higher and longer-lasting elevation in this

12 16 20 24 48

TIME (WEEKS)

FIG. 1. Evolution of HBV-DNAp, HBV-DNA, and ALT in a patient treated with 5 MU in whomboth HBV markers rose transiently during the rIFN-7 administration.

757

QUIROGA ET AL.

2 3T 1 11 E (WEEKS)

FIG. 2. Changes in 2',5'-oligoadenylate (2-5A) synthetase activity during the rIFN- administra¬tion: ( ) group 1, 5 MU rIFN-7/m2 body surface; (A) group 2, 2 MU rIFN-7/m2 body surface.(Vertical bars represent standard deviation.)

activity, although no differences were found between the two doses. Just after concluding therIFN-7 administration, the 2-5A synthetase levels were slightly increased, but close to their basalvalues, and remained unchanged throughout the follow-up (data not shown).

Within the first week of treatment, a slight increase in ALT values was seen in 5/8 patients (Fig.3), followed by a significant decrease in ALT (p < 0.01) in the second week, which coincided withthe rebound in HBV-DNA and HBV-DNAp. Then ALT values rose up to the basal values.

Changes in immunological parametersDuring the therapy, an increase in circulating rIFN-7 levels was seen, especially in those patienta

treated with 5 MU, although differences did not reach statistical significance (data not provided).None of the patients developed anti-rIFN-7 antibodies. Fifteen days after starting the therapy, thenumber of peripheral blood lymphocytes increased in 5/8 patients, as well as the helper to cyto-toxic/suppressor T-cell (T4/T8) ratios, although no statistical differences were observed. Through¬out the rIFN-7 administration, this ratio was slightly higher than the basal value in 6/8 patients(Fig. 4A). This change was presumably made through a statistically significant decrease in the num¬

ber of T8 lymphocytes (basal vs. 4 weeks, < 0.05) (Fig. 4B). At the end of the follow-up, T-lym-phocyte subsets, as well as T4/T8 ratios, returned to the basal values (data not shown).

Serum /î2-microglobulin levels increased significantly in all patients within the first week of treat¬ment, both with the high (p < 0.01) and the low (p < 0.05) rIFN-7 dose (Fig. 5). At that time, dif¬ferences between the two doses reached statistical significance (p < 0.03). A plateau-like in 02-mi-croglobulin levels was observed during the next 2 weeks with the high rIFN-7 dose; in contrast, ß2-microglobulin values experienced a rise up to the third week with the low dose, which represented amaximum compared with the other dose. Thereafter, /32-microglobulin levels returned to the basalvalues in all patients and remained unchanged during the follow-up.

The values from the eight patients regarding 2-5A synthetase, /52-microglobulin, and T4/T8ratios were compared with ALT, HBV-DNA, and HBV-DNAp. A statistically significant linearcorrelation (r = 0.743, < 0.001) was only found between 2-5A synthetase and /32-microglobulinvalues.

758

2-5A AND HLA-I IN RIFN-7 THERAPY OF CHRONIC HBV

«00

350

300

= 250·

5 200

ISO·

100

50 -,1

TINE (KEEKS)

FIG. 3. Individual variations in ALT during the rIFN-7 administration: ( ) group 1 (patient no.1-4), 5 MU rIFN-7/m2 body surface; (A) group 2 (patient no. 5-8), 2 MU rIFN-7.

1.55

2 1.35

1.25.

1.15-

1.0514 0

TIME (WEEKS)

FIG. 4. Individual changes in the T4/T8 ratio (A) and in the T8 lymphocyte subset (B) before andafter the rIFN-7 administration: ( ) group 1 (patient no. 1-4), 5 MU rIFN-7/m2 body surface;(A) group 2 (patient no. 5-8), 2 MU rIFN-7.

12·_i

~ 11.§1 10

»CM

-1-1-12 3 15

TIME (WEEKS)

FIG. 5. Changes in /32-microglobulin during the rIFN-7 administration: ( ) group 1, 5 MU rlFN-7/m2 body surface; (A) group 2, 2 MU rIFN-7. (Vertical bars represent standard deviation.)

759

QUIROGA ET AL.

7-Globulin levels increased during therapy in patients treated with 5 MU rIFN-7, although no

significant differences were observed (mean g°7o ± SD of: basal, 1.97 ± 0.18; 2 weeks, 2.19 ±0.18). Three patients resulted positive to SMA and one to ANA in the basal sample. None of thepatients gave a positive result to AMA and anti-nDNA. The expression of those nonorgan-specificautoantibodies did not significantly change during, or after, the therapy.

As shown in Table 1, 4/8 (50%) patients exhibited LMA positivity in serum in the basal sample;two of them were anti-HD positive. At the end of the rIFN-7 administration, all the patients hadLMA in their serum, as well as after the first month of follow-up (p < 0.05). Later on, several pa¬tients lost LMA, and at the end of the follow-up only two patients (who were anti-HD positive) ex¬hibited LMA. Total anti-HD titer did not significantly vary throughout the study.

DISCUSSION

It has been reported that lymphoclastoid IFN is able to induce an increase in 2-5A synthetase in pa¬tients with HBeAg"7' together with a decrease in HBV-DNA polymerase,"8' and also in HBV-DNA."" In this report, we have observed increases in 2-5A synthetase while HBV-DNA and HBV-DNAp decreased during rIFN-7 administration. Furthermore, both the low (2 MU) and the high (5MU) rIFN-7 doses showed a similar elevation in this activity. This suggests that the administrationof 2 MU may be enough to induce an antiviral effect. In agreement with this, preliminary resultsobtained following administration of 2 MU/m2 rIFN-7 twice a week for 6 months show that amaintained loss of HBV-DNA may be achieved.'20'

As HBV itself does not have a direct cytopathic effect on hepatocytes, antibody- and cell-medi¬ated immune reactions are most likely responsible for liver injury.'2" It should be pointed out thatsimultaneous replication by both HBV and delta virus leads to more hepatic necrosis than that ob¬served when only one virus is present,'22' and that, although less likely, other viral factors such as

Table 1. Changes in LMA Expression during the Therapywith RIFN-7

Time (weeks after initiation of IFN treatment)*Patient no.3 0°¿ 2 4° 8¿ 12 24 36 48

1 + + +----

2 + + + + + + + +

3-

+ + + + +- +

4--

+ + + +--5

--

+ + + +--

6__

+ +____

7 + + + + ++ +-

8 + + + + + + +-

No. LMA-t- 44 8 8663 2(%) 50 50 100 100 75 75 37.5 25

"Group 1 (patients 1-4): 5 MU rIFN-7/m1 body surface; group 2 (patients 5-8): 2MU rIFN-7/m2 body surface.

brIFN-7 was administered for 4 weeks.cp < 0.05.dp < 0.05.

760

2-5A AND HLA-I IN RIFN-7 THERAPY OF CHRONIC HBV

non- , non-B hepatitis may not be excluded in our drug addicted, anti-HD-positive patients.Indeed, HBeAg-positive patients, who are actively replicating, may have altered the immunoregula-tory system.'23' IFN-7 has been demonstrated to be a factor implicated in the "in vitro" B-cell mat¬uration and later activation of antibody synthesis; thus, it could modulate the host's immune re¬

sponse in chronic viral liver disease.'24' In this way, we have found increased levels of 7-globulinhalf-way through the therapy in those patients treated with 5 MU. However, no significant effecton the expression of nonorgan-specific autoantibodies has been observed and none of the patientsdeveloped anti-rIFN-7 antibodies, as the consequence of rIFN-7 administration. Therefore, they donot explain the changes observed in the humoral response.

However, after finishing the rIFN-7 therapy, all the patients shared liver membrane-specificautoantibody (LMA). This situation prevailed for 1 more month, although a progressive decline inthe number of LMA-positive cases was further observed. At the end of the follow-up, only 2 pa¬tients remained LMA positive; indeed, they were anti-HD positive. In accordance with this, an as¬sociation between the presence of anti-HD and LMA has been reported recently."6' This suggeststhat rIFN-7 may induce a transient increase in the liver membrane antigen (LMAg) displayed on thehepatocellular surface.'25' However, further studies are needed during the rIFN-7 therapy to con¬firm this finding and exclude possible assay related fluctuations.

IFN-7 may increase HLA class I antigens expression on the cellular surface.'26' Our results showthat, after the daily rIFN-7 administration, 02-microglobulin shows a significant rise. This is inagreement with the changes on HLA-I reported recently following IFN therapy.127' The dose of 5MU seems to induce a very early effect on this HLA class I-associated protein (first week oftherapy); although, the 2 MU dose also produced a significant increase in /32-microglobulin levels,but this effect was delayed up to the third week. On the other hand, /32-microglobulin increased inparallel with the number of LMA-positive cases. It remains open whether a possible relationshipbetween HLA class-I antigens expression and LMA exists in these cases. Nevertheless, the possiblebenefit of rIFN-7 on the immune system requires further elucidation.

ACKNOWLEDGMENTS

We thank C. Compernolle from Boehringer Ingelheim (Barcelona, Spain) for her kind coopera¬tion. This work was supported in part by a grant from the "Fondo de Investigaciones Sanitarias^ dela Seguridad Social (FISss)," Spain, 1162/85.

REFERENCES

1. EDDLESTON, A.L.W.F., and WILLIAMS, R. (1979). HLA system and liver disease, in: Progress inLiver Disease, vol. 6. H. Popper and F. Schaffner (eds.). New York: Grune and Stratton, pp. 285-299.

2. NAGAFUCHI, Y., and SCHEUER, P.J. (1986). Expression of/32-microglobulin on hepatocytes in acuteand chronic type hepatitis. Hepatology 6, 20-23.

3. LENGYEL, P. (1982). Biochemistry of interferons and their actions. Annu. Rev. Biochem. 51, 251-282.4. WILLIAMS, B.R.G., BROWN, R.E., GILBERT, C.S., GOLGHER, R.R., WRESCHNER, D.H.,

ROBERTS, W.K., SILVERMAN, R.H., and KERR, I.M. (1981). Assay of (2'-5')-oligo(A) synthetized "invitro" and the analysis of naturally occurring (2'-5')-oligo(A) from intact cells. Methods Enzymol. 79,199-208.

5. SCHATTNER, ., MERLIN, G., WALLACH, D., ROSENBERG, H., BINO, T., HAHN, T., LEVIN,

761

QUIROGA ET AL.

S., and REVEL, M. (1981). Monitoring of interferon therapy by assay of (2'-5')oligo-isoadenylate synthe¬tase in human peripheral white blood cells. J. Interferon Res. 1, 587-594.

6. MERRITT, J.A., BORDEN, E.C, and BALL, L.A. (1985). Measurement of 2',5'-oligoadenylate synthe¬tase in patients receiving interferon-alpha. J. Interferon Res. 5, 621-628.

7. SHERLOCK, S., and THOMAS, H.C. (1985). Treatment of chronic hepatitis due to hepatitis virus.Lancet ii, 1343-1346.

8. MÜLLER, R., KLEIN, H., VIDO, I., NIEHOFF, G., LAUTZ, H.U., GEBEL, M., and SCHMIDT,F.W. (1986). Antiviral treatment in chronic hepatitis B. Data of 5 prospectively controlled randomizedtrials. J. Hepatol. 3 (suppl. 2), S217-S223.

9. PORRES, J.C., MORA, I., GUTIEZ, J., BARTOLOMÉ, J., QUIROGA, J.A., BAS, C, COMPER-NOLLE, C, ORDÌ, J., CHOCARRO, ., and CARREÑO, V. (1988). Antiviral effect of recombinantgamma interferon in chronic hepatitis virus infection: A pilot study. Hepato-gastroenterol. 35, 5-9.

10. BIANCHI, L., DE GROÓTE, J., DESMET, V.J., GEDIGK, P., KORB, G., POPPER, H., POULSEN,H., SCHEUER, P.F., SCHMID, M., THALER, H., and WEPLER, W. (1977). Acute and chronic hepa¬titis revisited. Review by an international group. Lancet ii, 914-919.

11. ROBINSON, W.S. (1975). DNA and DNA polymerase in the core of the Dane particle of hepatitis B. Am.J. Med. Sci. 270, 151-159.

12. WELLER, I.V.D., FOWLER, M.J.F., MONJARDINO, J., and THOMAS, H.C. (1982). The detectionof HBV-DNA in serum by molecular hybridization: A more sensitive method for the detection of com¬

plete HBV particles. J. Med. Virol. 9, 273-280.13. SAWAI, H., ISHIBASHI, K., ITOH, M., and WATANABE, S. (1985). Radioimmunoassay for 2-5A

using a novel 12SI-labeled 2-5A analog, and analysis of 2-5A and 2-5A synthetase activity in humanplasma, in: The 2-5A System: Molecular and Clinical Aspects of the Interferon-Regulated Pathway. NewYork: Alan R. Liss, Inc., pp. 97-104.

14. KAWADE, Y., and WATANABE, Y. (1984). Neutralization of interferon by antibody: Appraisals ofmethods of determining and expressing the neutralization titer. J. Interferon Res. 4, 571-584.

15. SEGLEN, O.P. (1973). Preparation of rat liver cells. Exp. Cell. Res. 82, 391-398.16. DE TORRES, M., CARREÑO, V., BARTOLOMÉ, J., PORRES, J.C., BAS, C, QUIROGA, J.A., and

MORA, I. (1987). Liver membrane antibodies in chronic delta hepatitis. Digestion 38, 171-177.17. IKEDA, T., PIGNATELLI, M., LEVER, A.M.L., and THOMAS, H.C. (1986). Relationship of HLA

protein display to activation of 2-5A synthetase in HBe antigen or anti-HBe positive chronic HBV infec¬tion. Gut 27, 1498-1501.

18. FURUTA, M., AKASHI, K., NAKAMURA, Y., MATSUMOTO, K., YAMAGUCHI, H., TAKA-MATSU, S., and SHIMIZU, T. (1987). 2',5'-oligoadenylate synthetase activity in peripheral blood lym¬phocytes as a clinical marker in interferon therapy for chronic hepatitis-B. J. Interferon Res. 7, 111-119.

19. CARREÑO, V., PORRES, J.C., MORA, I., GUTIEZ, J., QUIROGA, J.A., RAMON Y CAJAL, S.,OLIVA, H., COMPERNOLLE, C, and BARTOLOMÉ, J. (1987). A controlled study of treatment withrecombinant interferon alpha in chronic hepatitis virus infection: Induction and maintenance schedule».Antiviral Res. 8, 125-137.

20. CARREÑO, V., and MORA, I. (1987). Combination of recombinant interferons alpha and gamma intreatment of chronic hepatitis B. Lancet ii, 1086.

21. MONDELLI, M., and EDDLESTONE, A.L.W.F. (1984). Mechanisms of liver cell injury in acute andchronic hepatitis B. Sem. Liver Dis. 4, 47-58.

22. RIZZETTO, M., VERME, G., RECCHIA, S., BONINO, F., FARCI, P., ARICO, S., CALZIA, R.,PICCIOTTO, ., COLOMBO, ., and POPPER, H. (1983). Chronic hepatitis in carriers of hepatitis surface antigen with intrahepatic expression of the delta antigen. An active and progressive disease unre¬

sponsive to immunosuppressive treatment. Ann. Int. Med. 98, 437-441.23. EGGINK, H.F., HONTHOFF, H.J., HUITEMA, S., GIPS, C.H., and POPPEMA, S. (1982). Cellular

and humoral immune reactions in chronic active liver disease. I. Lymphocyte subsets in liver biopsies ofpatients with untreated idiopathic autoimmune hepatitis, chronic active hepatitis and primary biliary cir¬rhosis. Clin. Exp. Immunol. 50, 17-24.

24. BILLIAU, A. (1986). The mode of action of interferons in viral infections and their possible role in thecontrol of hepatitis B. J. Hepatol. 3 (suppl. 2), S171-S179.

25. MEYER ZUM BÜSCHENFELDE, K.H., HÜTTEROTH, T.H., MANNS, M., and MOLLER, . (1980).The role of liver membrane antigens as targets in autoimmune-type liver disease. Springer. Sem. Immuno-pathol. 3, 297-315.

26. WALLACH, D., FELLOW, M., and REVEL, M. (1982). Preferential effect of interferon on the syn¬thesis of HLA antigens and their own mRNA in human cells. Nature 299, 833-836.

762

2-5A AND HLA-I IN RIRN-7 THERAPY OF CHRONIC HBV

27. PIGNATELLI, M., WATERS, J., BROWN, D., LEVER, ., IWARSON, S., SCHAFF, ., GERETY,E., and THOMAS, H.C. (1986). HLA class I antigens on the hepatocytes membrane during recovery fromacute hepatitis virus infection and during interferon therapy in chronic hepatitis virus infection. Hepa-tology 6, 349-353.

Address reprint requests to:Dr. V. Carreño

Department of GastroenterologyFundación Jiménez Diaz

Universidad AutónomaAvenida Reyes Católicos, 2

28040 MadridSpain

Received 1 April 1988/Accepted 1 July 1988

763