Journal of Medical Virology 38:20&206 (1992)

Serum Hepatitis C Virus (HCV)-RNA and Response to Alpha-Interferon in Anti-HCV Positive Chronic Hepatitis -

S. Magrin, A. Craxi, C. Fabiano, G. Fiorentino, L. Marino, P. Almasio, G.B. Pinzello, U. Palazzo, M. Vitale, A. Maggio, G. Bucca, F. Gianguzza, V. Shyamala, J.H. Han, and L. Pagliaro Divisione d i Medicina In.terna (S.M., C.F., G.F., L.M., G.B.P., U.P.) and Servizio d i Terapia e Prevenzione della Talassemia (M.V., A.M.), Ospedale V . Ceruello, and Clinica Medica (A.C., P.A., L.P.) and Dipartimento di Biologia Cellulare e dell0 Sviluppo (G.B., F.G.), University of Palernzo, Palerrno, Italy; Chiron Corporation, Emeryville, California (V.S., J.H.H.) Hepatitis C virus (HCV) replication was assessed before and during alpha-interferon (IFN) treatment in 22 anti-HCV positive patients with posttransfusion or sporadic chronic hepatitis (CH). Eleven patients were “responders” and 11 patients “non-responders” t o IFN. Thirteen antiHCV negative healthy subjects and five anti-HCV negative patients with autoimmune CH served as controls. Serum HCV-RNA was detected by the polymerase chain reaction (PCR) in all untreated antiHCV positive patients but in none of the anti-HCV negative subjects. PCR primers from the 5‘-noncoding (NC) region were more sensitive than primers from a non-structural (NS5) region in detecting HCV-RNA (21122, 95% vs. 7/22, 32%, respectively). Positive strand HCV-RNA titre and positivity rate for the negative strand were similar in responders and non-responders before IFN treatment, as well as anti-cl00-3 titre by enzymelinked immunosorbent assay (ELISA), and anti-51-1, anti-c33c, anti-c22 positivity rate by immunoblot assay (RIBA). HCV-RNA positivity by both NC and NS primers was more frequent before IFN among responders. During IFN treatment, serum HCV-RNA was detectable, mostly at l o w titres, in 1 (NC positive) of the 11 responders and in 9 (4 NS positive and 5 NC positive) of the 11 non-responders. Among the four non-responders who were NS positive during IFN, three were NC positive before IFN. Serum HCV-RNA was always found in our post-transfusion or sporadic anti-HCV positive patients with CH. Viraemia generally decreased during IFN treatment, but no available HCV markers clearly distinguished responders from non-responders before IFN treatment. Different NC and NS primers behaviour between responders and non-responders before and during treat0 1992 WILEY-LISS. INC.

ment with IFN suggests genomic heterogeneity and may explain non-responsiveness t o IFN. 0 1992 Wiley-Liss, Inc.

KEY WORDS: autoimmune hepatitis, PCR, IFN treatment

INTRODUCTION Hepatitis C virus (HCV) was cloned recently [Choo et al., 19891 and sequenced nearly completely IChoo et al., 1991; Takamizawa e t al., 19911. Serum antibodies to HCV (anti-HCV) can be found by using enzymelinked (ELISA) and immunoblot (RIBA) assays, while serum and liver HCV-RNA, both positive strand IWeiner e t al., 19901 and negative strand [Fong et al., 19911, can be detected by the polymerase chain reaction (PCR) amplification. Of all post-transfusion and sporadic chronic hepatitis (CHI cases, 70-80% are anti-c100-3 positive [Esteban et al., 19891, but virtually all have antibodies to c22 and/or c200 [Craxi e t al., 19911. HCV-RNA has been found in serum and in the liver of these patients [Weiner et al., 19901. Alpha-interferon (IFN) normalizes aminotransferases and reduces histological necroinflammation in about 50% of patients with post-transfusion or sporadic CH [Davis e t al., 1989; Realdi et al., 1990; Tine et al., 19911. A decrease of serum HCV-RNA after IFN treatment can be observed consistently [Kanai et al., 1990; Brillanti et al., 19911 by using PCR primers from the highly conserved 5’-terminal sequence [Okamoto et al., 1990; Han e t al., 19911. Since

Accepted for publication April 24, 1992. Address reprint requests t o Dr. Silvio Magrin, Divisione di Medicina Interna, Ospedale V. Cervello, Via Trabucco 180, 90146 Palerrno, Italy.

HCV-RNA in Chronic Hepatitis

201

TABLE I. Features of Patients at Presentation

R“ (n = 11) Age (yearsY Sex (MIF) AST (IUIl)‘

ALT (IUIl)“ Albumin (gldl)‘ Gamma-globulins (gill Bilirubin (mg/dl)“ CPHd CAH‘ Cirrhosis

39.1 (23-56) 912 81 (21-253) 155 (51-359) 4.15 (3.74.8) 1.64 (1.4-2) 0.73 (0.5-1.1) 1 8 2

tients were admitted to the hospital for the first time during the last 2 years. Patients were selected randomly among participants in ongoing IFN treatment trials (unpublished data). Recombinant interferon al47.6 (22-5 7) pha-2b (INTRON-A, Schering-Plough Corporation, Kenilworth, UK, 10 MU TIW for 2 months, then 5 MU 417 91 TIW for 4 months) was given to patients with sporadic (36-165) CH, while lymphoblastoid alpha-interferon (Wellferon, 134 (4G223) Wellcome, Beckenham, UK, 6 MU TIW for 2 months, then 3 MU TIW for 4 months) was given to patients 3.97 (3.54.4) with post-transfusion CH. 1.53 Patients with autoimmune CH underwent pred(1.2-2.2) nisone treatment (37.5 mg/day for 30 days, 25 mglday 0.66 (0.4-1.1) for 60 days, then 12.5 mg/day), with AST/ALT becoming normal within 3 months. 0 N R ~ (n = 11)

6 5

Methods Liver biopsy. All patients underwent percutane”R = responder to IFN. NR = non-responder to IFN. ous liver biopsy before treatment. Hepatic tissue speci‘Arithmetic mean (range).ASTiALT (normal limit): 50 glday) or of previous (5‘-TTCACGCAGAAAGCGTCTAG-3’;sense) and NC4 anti-sense). immunosuppressive or antiviral treatment. All pa- (5‘-GTTGATCCAAGAAAGGACCC-3’;

202

Magrin et al.

TABLE 11. Serum HCV-RNA (Positive Strand) in RNA was extracted from 150 p1 of serum by adding Untreated Patients* 450 pl of buffer containing 80 mmolil “Tris”-HC1 pH 7.5, 0.8 mmol/l EDTA, 160 mmol/l NaCl, 0.6% (wiv) R NR (n = 11) (n = 11) SDS, and 1 mg/ml proteinase K, mixed and incubated at 50°C for 60 min. Proteins were removed by two ex- NS 0 1 tractions with phenollchloroform and one with chloro- NC 6 9 5 1 form alone. Twenty micrograms of glycogen was added NC + NS to the aqueous phase and the RNA precipitated by addi- *R = responders to IFN; NR = non-responders to IFN; NS = nontion of 3 volumes of ice-cold absolute ethanol and 1/10 structural primers; NC = non-coding primers. (viv) 3 M Na acetate. For cDNA synthesis the RNA was pelleted and dissolved in 10 pl of sterile distilled water, heated a t 65°C for 5 min, and cooled rapidly on ice. cDNA synthesis was carried out after adjustment of the primers, or not added, or only sense primer was used, mixture to contain 50 mmol1l “Tris”-HC1 pH 8.3, 75 were carried out a s suggested [Fong e t al., 19911 and mmolil potassium chloride, 3 mmol1l magnesium chlo- found to be negative. The results were considered valid ride, 10 mmolil dithiothreitol, 0.2 mmol/l of each dNTP, only if they were consistent in repeated experiments; 40 units of ribonuclease inhibitor (BRL, Bethesda, all samples were tested at least twice. To assess the sensitivity of the PCR by NS primers, MD), 200 units of cloned Moloney murine leukemia virus reverse transcriptase (BRL), and 1 pmol/l of the we analyzed tenfold serial dilution of pDXlOO plasmid. appropriate primer in a final volume of 30 pl, which A single molecular of the pDXlOO plasmid was detectwas incubated at 37°C for 60 min. The reverse tran- able by ethidium bromide staining without the use of a scriptase was then inactivated by heating at 95°C for 20 radioactive probe, as previously reported [Garson et al., 1990al. min and then quickly chilled on ice. To assess the sensitivity of the PCR by NC primers Round 1 of the PCR was carried out in a 100 pl mixture containing 10 mmol/l “Tris”-HC1pH 8.3,50 mmol/l we tested serial dilution of a n in vitro synthesized conpotassium chloride, 1.5 mmol magnesium chloride, trol HCV-RNA (positive strand), which was identical to 0.01% (weightivolume) gelatin, 2.5 units recombinant part of the NC region [Shyamala and Han, 19911. As Taq DNA polymerase (Perkin Elmer, Cetus), 200 few a s ten copies of HCV genome were detectable. The titer of HCV positive strand RNA was estimated pmol/l of each dNTP. and 1 pmol/l of each “outer” primer. After a n initial 5 min denaturation at 94”C, 35 by testing tenfold serial dilutions of cDNA from PCRcycles of 94°C for 60 sec, 50°C for 60 sec, 72°C for 60 sec positive samples. The specificity of the PCR assays was confirmed by were carried out, followed by a 7 min extension at 72°C. Ten microliters of the round 1 PCR products was trans- partial sequencing of randomly selected PCR products ferred to the round 2 100 pl reaction mix. Twenty-five from both NS and NC regions. Sequencing was carried cycles of 94°C for 60 sec, 50°C for 60 sec, 72°C for 60 sec out with Sequenase (United States Biochemical, Clevewere carried out, followed by a 7 min extension a t 72°C. land, OH) a s previously described [Casanova et al., Bands were visualized by ethidium bromide staining, 19901, and with AmpliTaq Sequencing Kit (Perkin photographed a t 302 nm, and identified by size [Kanai Elmer, Cetus), according to the manufacturer’s instrucet al., 19901 and by hybridization to a phosphorus-32- tions. labelled HCV-cDNA insert. RESULTS The positive strand of serum HCV-RNA was also detected by using primers from a non-structural (putative Serum HCV-RNA (positive strand) was detected in NS5) region in a “nested” PCR [Garson et al., 1990al. all untreated anti-HCV positive subjects and in none of “Outer” primers were d94 (5’-ATGGGGCAAAGGAC- the anti-HCV negative subjects. The NC primers were GTCCG-3’; sense) and d95 (5’-TACCTAGTCATAGC- more sensitive (21/22,95%)than the NS primers (7122, CTCCGTGAAG-3’; anti-sense). “Inner” primers were 32%) (Table 11). N1 (5’-GAGGTTTTCTGCGTCCA-3‘;sense) and N2 5’Before starting IFN treatment, responders and nonGCGATAGCCGCAGTTCT-3’; anti-sense). PCR was responders were similar for serum HCV-RNA positivcarried out as for the NC region primers. Bands were ity rate, both positive and negative strands (Table 1111, visualized by ethidium bromide staining, photographed for HCV viraemia (Table 111, Fig. l ) , a s well a s for a t 302 nm, and identified by size [Garson et al., 1990al anti-c100-3 antibody titre by ELISA (median 1/32 in and by hybridization to a phosphorus-32-labelled cDNA responders vs. 1/64 in non-responders, range 111 to clone (designed pDX100, kindly provided by Dr. David 11512 in both groups), and for anti-5-1-1 (8/11vs. 7/11), Parker, Wellcome Diagnostics, UK). anti-c33c (11/11 vs. lO/ll), anti-c22 (10111 vs. 10111) The recommended precautions were taken to reduce positivity rate by RIBA. In the six patients in whom the risk of contamination of samples with PCR products HCV-RNA negative strand was not detected (two re[Kwok and Higuchi, 19891. All serum samples from sponders and four non-responders), positive strand titre patients were extracted in parallel with samples from was lower (median titre: than in patients where healthy subjects. PCR negative controls in which re- negative strand was detected (median titre: verse transcriptase was destroyed before the addition of (Table 111). HCV-RNA positivity by both NC and NS

HCV-RNA in Chronic Hepatitis

203

TABLE 111. HCV Replication and Histological Features in Patients Treated With IFN*

No.

Agelsex

Source

Alt

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

371F 42lM 58lM 23/M 52JM 39/M 55/F 44lM 441M 56lM 38lM 40lF 54JF 52JF 55lM 55/F 57lF 45lM 47lM 53lF 22/F 441M

PT PT spor PT

110 51 221 264 120 184 117 100 85 359 98 110 129 76 159 183 125 223 217 46 136 70

spor spor spor spor spor

PT spor

PT

PT spor spor spor

PT spor spor spor spor spor

Histology seCAH miCAH seCAH moCAH miCAH miCAH moCAH CPH acCIR acCIR miCAH miCAH acCIR acCIR acCIR seCAH moCAH acCIR moCAH acCIR moCAH acCIR

IFN R R R R R R R R R R R NR NR NR NR NR

NR NR NR NR NR NR

HCV-RNA strand Positive (primer) titre Negative 10’ (NC)

lo-’ (NC) 10-1 (NC) lo-‘ (NC +NS)

10-1(NC +NS) (NC) 10’ (NC) 10’ (NC +NS) (NC +NS) lo-* (NC +NS) 10’ (NC) 10- (NC) lo-’ (NC) lo-’ (NC) lo-’ (NC) 10-1(NC) 10-1 (NC +NS)

lop1 (NS)

10’ (NC) 10-1 (NC) (NC) lo-’ (NC)

+ + + + + ++ + + + + + + + -

+ +

*ALT = alanine aminotransferase (normal limit: