Influence of the genotypes of hepatitis C virus on the severity of recurrent liver disease after liver transplantation

GASTROENTEROLOGY1995;108:1088-1096

LIVER, PANCREAS, AND BILIARY TRACT Influence of the Genotypes of Hepatitis C Virus on the Severity of Recurrent Liver Disease After Liver Transplantation CYRILLE FleRAY, * MICHELLE GIGOU,* DIDIER SAMUEL,* VALleRIE PARADIS, t SHUNJI MISHIRO, § GEERT MAERTENS, [I MICHEL REYNES, ~ HtROAKI OKAMOTO, § HENRI BISMUTH,* and CHRISTIAN BRI~CHOT ~'# • HepatobiliarySurgeryand LiverTransplantResearchUnit, H6pital Paul Brousse, Paris South University,Villejuif, France;*Service d'Anatomo-Pathologie,H6pital Paul Brousse, Villejuif, France;§JinchiMedical School, Tokyo, Japan; IIInnoGeneticsN.V., Gent, Belgium; ~INSERMUnit6 370, Centre HospitalierUniversitaireNecker, Paris, France;and #Hybridotest,Institut Pasteur, Paris, France

See editorial on page 1314. B a c k g r o u n d / A i m s : Several genotypes of hepatitis C vi-

rus (HCV) have been identified by phylogenetic analysis, but their clinical relevance remains elusive. Liver transplantation for HCV-related cirrhosis offers a unique opportunity for prospective studies of this issue. M e t h o d s : Sixty anti-HCV-positive liver recipients with precise virological and histological assessments were included. HCV genotype was determined with both typespecific capsid primers and a line probe genotyping assay. Results: HCV genotype l b was the predominant type before transplantation (40 of 60 patients); after liver transplantation, acute and chronic active hepatitis developed more frequently in these patients than in patients infected by other genotypes (31 of 40 and 24 of 40 vs. 8 of 20 and 4 of 20 patients). Actuarial rates of acute hepatitis and chronic active hepatitis were 77% and 59%, respectively, 3 years after transplantation in patients infected by type l b and 40% (P = 0.008) and 22% (P = 0.004) in those infected by other types. There was no statistical relation between the level of HCV viremia and HCV genotypes both before and after transplantation. In contrast, after transplantation, serum HCV RNA values were significantly increased in patients who developed hepatitis after transplantation. Conclusions: This study provides direct evidence that HCV l b is associated with more aggressive recurrent liver disease than other genotypes. epatitis C virus (HCV) is a positive-stranded, enveloped R N A virus that causes most cases of the parenterally transmitted non-A, non-B hepatitis. Since its discovery in 1988 and with the use of nucleotide

H

sequencing, it has been clear that there are markedly divergent nucleotide sequences among distinct isolates. Phylogenetic analysis has been used to classify HCV into at least six major genotypes on the basis of overall sequence similarity in both coding and noncoding parts of the viral genome. 1-5 It is now established that different genotypes can be identified in a given geographical area. Beyond taxonomic and epidemiological interest in these HCV genotypes, their clinical relevance remains to be determined. In particular, it has been suggested that type lb HCV (as defined in the classification of Simmonds et a l ) and referred to as types II and K1 by Cha et al. 3 and Nakao et al., 4 respectively) may be more resistant to interferon therapy. 7'8 It is also possible that some genotypes are more pathogenic for liver cells, but this is difficult to confirm in the absence of an HCV culture system. In addition, clinical studies aiming at comparing the course of infection by the different genotypes are hampered by the difficulty in determining the time since contamination. Orthotopic liver transplantation offers a unique opportunity to study prospectively the infection of normal livers by different types of HCV, with precise surveillance of clinical, laboratory, and histological parameters. Indeed, there is a high frequency of HCV recurrence after liver transplantation. #'l° Because of the immunosuppressive regimen, infection of the allograft by HCV is characterized by a high level of viral replication and developAbbreviations used in this paper: bDNA, branchedoligodeoxyribonucleotides; CID 50, chimpanzee infectious doses 50; LIPA, line probe assasy; PCR, polymerasechain reaction; RIBA, recombinant immunoblot assay. © 1995 by the American GastroenterologicalAssociation 0016-5085/95/$3.00

April 1995

m e n t of chronic active hepatitis (CAH). 11-.3 Therefore, we studied liver transplant recipients to assess the impact of the H C V genotype on viral pathogenicity and replication.

M a t e r i a l s and M e t h o d s Patients From October 1985 to September 1991, 100 patients underwent liver transplantation for decompensated hepatitis B surface antigen (HBsAg)-negative and anti-HCV-positive cirrhosis. Patients were excluded from the study if they (1) died within the first 12 months after transplantation (6 patients died of sepsis without liver graft dysfunction); (2) had no systematic histological assessment performed (n = 7); (3) had no biopsy performed within the month following abnormal liver test resuits (n = 6); (4) had no available serum samples (n = 10); or (5) had acquired infection by hepatitis B virus (HBV) after liver transplantation (n = 2). The remaining 69 patients were retrospectively analyzed for the detection of serum HCV RNA through nested polymerase chain reaction (PCR) using primers located in the 5' untranslated region; results were positive in 60 cases both before and after liver transplantation. The indication for liver transplantation was either end-stage cirrhosis (n = 49) or cirrhosis with hepatocellular carcinoma (n = 11). Small hepatocellular carcinomas ( < 2 0 mm) were also discovered after examination of explanted liver in 6 additional cases. HCV infection was clearly related to transfusions in only 4 patients, whereas the source of contamination was difficult to ascertain in the remaining 56 patients. However, 18 of 60 subjects had a previous history of surgery and, thus, possible transfusion. The patients originated from southern Italy (n = 38), northern Italy (n = 2), Spain (n = 1), Israel (n = 2), and France (n = 17). There were 51 men and 9 women, and the mean age at the time of transplantation was 47 years (range, 3 3 - 6 6 years). The liver graft histology at the time of transplantation was either normal or steatosic (2 cases), and the liver donors had been screened for anti-HCV since March 1990 (31 patients underwent grafting before this date). FKS06 and prednisolone were used for immunosuppression in 12 patients, and cyclosporin A, prednisolone, and azathioprine were administered to the other patients. All patients were HBsAg negative before and after transplantation. All subjects had systematic biopsies performed 1, 2, and 5 years after transplantation. Serum samples were obtained a mean of 45 days (range, 0 120 days) before transplantation and at the time of liver biopsies after transplantation. The end of the follow-up period w a s the day of the last available liver graft biopsy. None of the patients included in this series received antiviral therapy during follow-up. The mean follow-up was 42 months (range, 1 4 - 9 0 months). From March 1990 onwards, all blood donors were tested for anti-HCV antibodies and were excluded if the results were positive.

HCV GENOTYPES IN LIVER TRANSPLANT RECIPIENTS 1089

Methods Anti-HCV antibody detection. Anti-HCV antibodies were detected by using a recombinant immunoblot a s s a y (RIBA) (RIBA-2 Test System; Chiron Corp., Emeryville, CA) that detects antibodies against epitopes located in the capsid (c22) and nonstructural proteins (c33, c100.3, and c511). HCV RNA detection and genotyping with primers derived from the region coding for viral core. HCV typing by nested

PCR: Extraction of serum HCV RNA and complementary DNA synthesis. Sera (150 btL) were incubated in lysis buffer consisting of 100 mmol/L Tris-HCl (pH 8), 1% sodium dodecyl sulfate, and 50 btg/mL proteinase K for 1 hour at 37°C. Nucleic acids were extracted with phenol-chloroform (phenol was buffered to pH 4.5), precipitated with ethanol and sodium chloride (0.4 mol/L), redissolved in 30 btL sterile diethyl pyrocarbonate-treated water, aliquoted, and stored at -80°C. Ten microliters of this solution, corresponding to 50 ~tL of serum, was mixed with 10 pmol of downstream primer (primer 187), incubated for 5 minutes at 65°C, and then rapidly cooled on ice. Complementary DNA (cDNA) synthesis was performed at 37°C for 45 minutes in a mixture of 50 mmol/L Tris-HC1 (pH 7.5), 75 mmol/L potassimn chloride, 6 mmol/L magnesium chloride, 1 mmol/L deoxynucleoside triphosphate, a n d 100 IU of Moloney's murine leukemia virus (Bethesda Research Laboratories, Gaithersburg, MD) in a final volume of 20 ~L. cDNA was stored at -20°C. The primers used for the study have been described previously 14'15 (Table 1) and were designed to be either conserved in two or more HCV genotypes (primers 104, 256, i87, a n d 186) or specific for each genotype (primers 296, 256V, 104V, 132, 133, 134, 135, and 339). To diminish carryover, we modified the type-specific nested PCR described by Okamoto et al. 14 to perform the two steps of nested PCR sequentially in the same tube, as previously described) 6 We used two complementary approaches for genotyping, both based on capsid sequence detection. Type-specific nested PCR. Five microiiters of cDNA w a s deposited in a 200-btL Eppendorf tube. Then 20 btL of a PCR mix (deoxynucleoside triphosphate, 250 ~mol/L each, 50 mmol/L KC1, 1.5 mmol/L MgCI2, 0.01% gelatin, 5 pmol of each outer primer [187 and 104/104V], and 1.5 U of cloned Taq polymerase [Perkin Elmer Cetus, Norwalk, CT]) w a s added and covered with 25 btL of high-density oil (dimethylpolysiloxane AK 350; Waker-Chemie GmbH, Manchen, Germany). The second PCR mix (75 btL, with 60 pmol/L of each inner primer [296, 133, 134, 135, 339, and 104/104V] and no Taq polymerase) was injected to the surface of the oil, forming a droplet within it. The tubes were then subjected to a first amplification (25 cycles at 94°C for 30 seconds, 55°C for 30 seconds, and 72°C for 30 seconds) and centrifuged for 1 minute to mix the second PCR mixture with the first (the droplets are intact after the first amplification). The second amplification cycle (94°C for 30 seconds, 60°C for 30 seconds, and 72°C for 30 seconds) was thus performed in a final volume of 100 btL, with an inner/outer primer ratio of 12:1.

1090

FERAY ET AL.

GASTROENTEROLOGY Vol. 108, No. 4

Table 1. Nucleotide Sequence of Primers Used for HCV Genotyping Name 256(4) 256 V 5 (4) 187(~) 104 (4) 104 V(~) 186(~) 296(~) 132(~) 133(~) 134(~) 135(~) 339(~)

Sequence 5'CGCGCGACGAGGAAGACTTC 3' 5' CGCGCGACGCGTAAAACTTC 3' 5' GACGAGCGGTATGTACCCCA 3' 5' AGGAAGACTTCCGAGCGGTC 3' 5' CGTAAAACTTCTGAACGGTC 3' 5' ATGTACCCCATGAGGTCGGC 3' 5' GGATAGGCTGACGTCTACCT 3' 5' TGCCTTGGGGATAGGCTGAG 3' 5' GAGCCATCCTGCCCACCCCA 3' 5' CCAAGAGGGACGGGAACCTC 3' 5' ACCCTCGTTTCCGTACAGAG 3' 5'GCTGAGCCCAGGACCGGTCT 3'

Position

Type

139-158 139-158 420-401 148-167 148-167 410-391 196-177 204-185 291-272 321-302 270-251 236-216

l, ll, lll, l V ( l a , lb, 2a, 2b) V(3a) All I, II, lll, IV V All I I II III IV V

NOTE. All primers were located in the viral capsid domain of HCV. -% sense; *-, antisense.

Conserved nested PCR and runoff reaction, To improve the sensitivity of the assay and to provide an internal control, each serum sample was also tested by using a conserved nested PCR (permitting amplification of all the HCV genotypes), followed by a runoff reaction using type-specific primers. The nested PCR permitting amplification of all HCV types was similar to the type-specific nested PCR described earlier, except that primers 104/104V and 186 were used for the inner amplification. Typing was then performed in a runoff reaction as follows: 5 btL of the conserved nested PCR was added to a mixture of PCR buffer, 1 U of Taq polymerase, and 20 pmol/L of each specific primer (primers 296, 133, 134, 135, and 339) in a final volume of 20 ~L. This type-specific amplification was performed for five cycles (94°C for 30 seconds; 60°C for 30 seconds; and 72°C for 30 seconds). Gel analysis. All PCR products and runoff reactions were analyzed after electrophoresis in a 12% nondenaturing acrylamide gel. After ethidium bromide staining, expected molecular weights were checked by means of video capture and digitalization of gels (The Imager; Applig~ne, Strasbourg, France). HCV RNA detection and genotyping with line probe

assay. A nested PCR assay using universal primers in the 5' untranslated region was performed using sets of biotinylated primers in which sequences' degeneracy was included to allow annealing to all HCV genotypes, iv Nested PCR was performed in a one-tube assay, and products were subjected to electrophoresis on agarose gel. One tenth of the biotinylated PCR fragments was hybridized to oligonucleotides directed against the variable region of 5 ' untranslated region, immobilized as parallel lines on membrane strips (line probe assay, [LIPA]; Innogenetics, Gent, Belgium). After hybridization according manufacturer's instructions, streptavidin labeled with alkaline phosphatase was added, and incubation with a chromogen allowed the development of the colorimetric reaction. The reactivity of the amplified fragments with one or more lines in the strip allows recognition of the genotype. The five major

types (1-5) and their subtypes (la, lb, 2a, 2b, 3a, 3b, 4, and 3a) can be identified. Quantitative detection of serum NOV RNA. W e used

a signal amplification method based on branched oligodeoxyribonucleotides (bDNA) developed by the Chiron Corp. Briefly, these bDNAs have a unique primary segment and a set of identical secondary fragments covalently attached to the primary sequence through branch points. The primary sequence is designed to hybridize to HCV RNA, The secondary fragments are used to direct the binding of multiple copies of a small oligonucleotide labeled with alkaline phosphatase. This latter is detected using an enzyme-triggerable dioxetane substrate, and visible light output is recorded on a luminometer, thus permitting quantitation of HCV RNA. This test can detect as few as 350,000 HCV genomes per milliliter. Sensitivity of nested PCR and specificity of genotyping. The sensitivity of our assay was assessed by performing

PCR with primers located in untranslated region on dilutions of chimpanzee infectious doses 50 (CID 50). We were able to detect HCV RNA in a 105 dilution of CID 50 (a gift from Dr. P. N. Lelie and Dr. H. W. Reesink, Red Cross Blood Transfusion Service, Amsterdam, The Netherlands), corresponding to approximately 50 particles per milliliter of serum. To check the specificity of our modified assays, a panel of 20 sera was blindly tested by one of us (H.O.) using the original method. Furthermore, amplified products generated from cDNAs corresponding to the different HCV genotypes obtained in Japan were used as positive controls. Histology. Liver graft biopsies were performed when liver test results became abnormal and at 1, 2, and 5 years after transplantation. Hepatitis was defined by the association of hepatocyte necrosis with portal and lobular infiltration by mononuclear cells. Resolutive acute hepatitis was defined by a normal histology for at least 6 months after the initial diagnosis of acute hepatitis. Chronic lobular hepatitis was defined by the persistence of lobular hepatitis without piecemeal necrosis for more than 6 months. CAH was defined by the existence

April 1995

of piecemeal necrosis and assessed using the histological activity score of Knodell et al) 8 Acute rejection was defined by the association of venous endothelitis, portal infiltration, and cholangitis, and chronic rejection was defined as the disappearance of more than 50% of bile ducts. All patients underwent at least two posttransplantation biopsies (mean + SD, 5 + 3). Liver histology was reviewed in every case in a blinded fashion by two pathologists (V.P. and M.R.). Statistical analysis. Actuarial rates of hepatitis were calculated using the Kaplan-Meier method. Comparisons of actuarial rates were made with the log rank test. The incidence of acute hepatitis and CAH relative to the HCV type was analyzed by using the Z2 test. A paired t test was used to compare quantitative values of HCV RNA before and after transplantation.

Results Amplification of Serum HCV RNA The 60 patients of this series were included on the basis of detection of serum HCV RNA before and after transplantation by using primers located in the 5' untranslated part of HCV RNA, thus permitting application of the LIPA genotyping assay. Sera from 49 of these 60 patients also showed amplification with primers located in the capsid encoding sequences. Thus, our approach permitted application of two complementary genotyping assays.

HCV Genotyping Genotyping based on 5' noncoding sequence. Before transplantation, type la, lb, 2a, and 3a were detected in 12, 41, 5, and 1 of the 60 patients, respectively. Other types or mixed coinfection were not detected. One subject could not be classified in spite of a positive amplification of the 5' untranslated part (Figure 1). Pretransplantation and posttransplantation types were similar in 57 of 60 patients using LIPA. One patient infected by type la and 2 infected by type l b before transplantation were classified as having type lb, la, and 2a 1 year after transplantation, respectively. These 3 patients with a changing genotype after transplantation had underwent transplantation before March 1990. After transplantation, types la, lb, 2a, and 3a were detected in 12, 40, 6, and 1 patients, respectively; 1 patient remained unclassified. Comparison of genotyping in the capsid and in the 5' noncoding sequences. Using primers derived from the viral capsid, HCV RNA was detected in the sera before and after transplantation in 49 cases, thus permitting comparison of results obtained using the two methods on 98 samples. As shown in Table 2, 84 of 98 samples showed concordant results, as defined by the

HCV GENOTYPESIN LIVER TRANSPLANTRECIPIENTS 1091

identification of the same genotypes by the two assays. Total concordance was observed for genotype Ib in 65 relevant samples. Discordances between specific nested PCR and LIPA were thus observed in n o n - l b genotypes: 4 of 20 samples classified as la were found to be type 2a (n = 2) or 3a (n = 2) and 5 samples classified as 2a (n = 1), 2b (n = 2), and 3a (n = 2) were classified as la using LIPA. Among 7 samples that remained unclassified using specific nested PCR (in spite of positive amplification of capsid region), 5 were classified as la (n = 1) and 2a (n = 4), and 2 remained unclassified using LIPA. Altogether, this comparison showed a complete correlation between specific nested PCR and LIPA for the identification of type l b but some discordances for other types, in particular for genotype 2a. In another concomitant recent study, 19 we showed, using sequence analysis, that LIPA correctly identified HCV type in such discordant cases and that the genetic heterogeneity of core region (in particular for type 2) explained the failure of type-specific PCR procedure in European patients. Therefore, in all subsequent analyses, we considered genotyping provided by the LIPA classification.

Liver Graft Histology According to HCV Genotype Of the 60 patients, 41 had evidence of recurrent liver disease (e.g., hepatitis) after transplantation. Four of these 41 had resolutive hepatitis, 11 had chronic lobular hepatitis, and 26 had CAH. One case of cirrhosis was diagnosed 14 months after transplantation. Thirteen patients developed acute rejection. O f the 6 patients who underwent retransptantation, 4 had chronic rejection, and 2 had arterial thrombosis (Table 3).

Infection by HCV Type l b Thirty-one of the 40 patients infected by HCV type lb after transplantation developed acute hepatitis after transplantation. O f these 31 patients, 20 had CAH, 10 had chronic lobular hepatitis, and 1 had resolutive acute hepatitis. Two additional cases of CAH were diagnosed in the absence of a recognized episode of acute hepatitis. One case of cirrhosis was identified 14 months after transplantation. The mean Knodell score for the 24 patients with CAH was 8 (range, 4 - 1 3 ) . Eight of the 40 patients had acute rejection. Four patients underwent retransplantation: 2 because of chronic rejection and 2 for chronic arterial thrombosis and biliary complications.

Infection by Other Types Among the 60 patients tested using LIPA, 12 patients were reinfected by type la after LT; 6 of these

1092

F#RAY ET AL

1

GASTROENTEROLOGY Vol. 108, No. 4

2

L

3

4

5

L

6

7

[,

g

9

10

11 12

Figure 1. Representative results of HCV genotype determination. Lanes 1-3, genotyping performed with nested PCR using conserved and type-specific primers. Bands with sizes corresponding to type II (lanes 1 and 2) and I (lane 3) were shown. In addition, because of the high levels of serum HCV RNA in these patients, bands corresponding to conserved primers were also shown. Lanes 4-12, genotyping performed with nested PCR using only conserved primers followed by a primer extension (runoff) using type-specific primers. This approach provided an internal positive control (i.e., the amplification of the capsid with conserved primers). Primer extension with type-specific primers showed bands corresponding to type I (lane 6), type II (lanes 4, 5, 7, and 10), type II] (lane 9), and type IV (lane 8 and 11). On lane 12, genotype could not be obtained, but the amplification of the capsid excluded false negative results. Arrows indicate the size of the capsid amplified with conserved primers. L, l O 0 - b a s e pair ladder.

12 developed acute hepatitis during follow-up. Three of these 6 subjects had disease progression to CAH (Knodell score, 4-8), 1 had chronic lobular hepatitis, and 2 had acute hepatitis that resolved. Of the 6 patients who were infected by type 2a after transplantation, 2 had acute

2a

2b

3a

UCc

0 1

1 1

1 1

0 1

hepatitis that resolved in 1 patient and progressed to CAH in another. The 2 patients infected either by type 3a or by unclassified type did not develop hepatitis during follow-up. Of the 49 patients tested though type-specific nested PCR, 9 patients were reinfected by type la after transplantation; 3 of these 9 developed acute hepatitis during follow-up. One of these 3 subjects had disease progression to CAH (Knodell score, 4), and 2 had acute hepatitis that resolved. The 2 patients who were infected by type 2a after transplantation and the 2 patients infected either by type 2b or type 3a both before and after transplantation did not develop hepatitis during follow-up. Of the 4 subjects with unclassified HCV infection after trans-

2 2

Table 3, HCV Types According to LIPA Before and After Liver Transplantation and Histological Evolution of the Liver Graft

Table 2. Comparison of Type-Specific Nested PCR and LIPA Before and After Liver Transplantation in 9 8 S a m p l e s From 4 9 Patients Nested PCR LIPA la a lab

la

9 7 16

lb a lb b 2a a 2a b

lb

33 32 65 1 1

0 1 1

Transplantation

2b a 2b ~ 3a a 3a b UCa UCb Total

20 16/20

65 65/65

2 1/2

UC, unclassified. aBefore liver transplantation. bArter liver transplantation. °Unclassified despite positive amplification.

2 0

2 0

1 1 2 7 2/7

Type

Before

After

CLH

rAH

CAH

Others ~

la lb 2a 3a UC Total

12 41 5 1 1 60

12 ~ 40 c 6~ 1 1

1 10

2 1 1

3 22 1

11

4

26

6 7 4 1 1 19

CLH, chronic Iobular hepatitis; rAH, resolutive acute hepatitis; UC, unclassified. aNo evidence of hepatitis: rejection or subnormal histology. ~One patient was type l b before transplantation. °One patient was type l a before transplantation.

April 1995

HCV GENOTYPES IN LIVER TRANSPLANTRECIPIENTS 1093

plantation, 1 had CAH (Knodell score, 6) and 1 had resolutive acute hepatitis. Two patients had acute rejection and 2 chronic rejection 18 and 60 months after transplantation and subsequently underwent retransplantation. Comparisons

between

type

lb

and

other

types. As shown in Figure 2, the liver graft histology

was significantly more severe in patients infected by HCV type lb than by other genotypes. Indeed, acute hepatitis was diagnosed in 31 of 40 patients infected by type lb and in 8 of 20 of patients infected by other types (Z2 = 8.24; P = 0.004). Actuarial rates of acute hepatitis and CAH were 77% and 59%, respectively, 3 years after transplantation in patients infected by type lb and 41% (log rank test, 2.37; P = 0.008) and 22% (log rank test, 2.68; P = 0.004) in patients infected by other types. Altogether, identical conclusions in comparison between lb and "non-lb" were obtained in the 60 subjects tested with LIPA and in the 49 patients tested with specific nested PCR.

Quantitation of Serum HCV RNA in the bDNA Assay Enough serum sample was available for the assay in 41 and 39 subjects before and with a mean of 13 months (range, 8-13 months) after transplantation, respectively (Figure 3). Before transplantation. The bDNA assay was positive (>0.33 × 106 virions/mL) in 16 of 25 patients infected by type lb (mean, 3.5; SD, 3.8 X 106 virions/ mL) and in 7 of 16 patients infected by other types (mean, 5.6; SD, 5.1 × 106 virions/mL). There was no statistically significant difference in the level of HCV RNA according to the genotypes. In addition, no statistical relation was found between the level of pretransplant serum HCV RNA in the bDNA assay and the incidence of recurrent hepatitis. After transplantation, bDNA assay was positive in 17 of 18 and in 10 of 21 patients with and without hepatitis. The mean HCV RNA levels were 17.3 106 virions/mL (SD, 15) and 4.6 106 virions/mL (SD, 3.8), respectively. Thus, the HCV viremia was significantly elevated in patients with recurrent hepatitis (P < 0.01). The results were then compared with the genotype. Type lb. bDNA was detected in the 11 of 12 subjects tested at the time of acute hepatitis or CAH (mean, 16.7 × 106 virions/mL; range, 2.3 to 56 × 106 virions/mL) and in 6 of 12 patients tested at the time of normal histological findings (mean, 4 × 106 virions/mL; range, 0.8 and 6.2 X 106 virions/mL).

Other types, bDNA was detected in the 6 of 6 subjects tested at the time of acute hepatitis or CAH (mean, 19 × 106 virions/mL; range, 3.8-30 × 106 virions/mL) and in 4 of 9 subjects tested at the time of normal histological findings (mean, 5.6 × 106 virions/ mL; range, 1.7 and 11 × 106 virions/mL). Altogether, there was no statistical relation between level of serum HCV RNA and genotypes.

Discussion This study provides direct evidence for a more aggressive course of recurrent infection by HCV type lb than by other types of HCV after liver transplantation. Indeed, liver graft recipients infected by type lb developed acute hepatitis more frequently than patients infected by other types (31 of 40 vs. 8 of 20). Furthermore, acute hepatitis resolved in only 1 of 31 patients infected by type lb compared with 3 of 8 patients infected by other types. After transplantation, progression to CAH occurred more frequently in patients infected by type lb (22 of 40) than in those infected by other types (4 of 20). Finally, serum levels of HCV RNA correlated with the development of recurrent hepatitis but not with the different types of HCV. Type lb was the predominant type of HCV (41 of 60 patients; 68%) in the European cirrhotic patients requiring liver transplantation either for end-stage liver disease or for hepatocellular carcinoma. This is in agreement with results of recent reports from Italy,2° and France 21 showing a higher incidence of HCV type lb in subjects with cirrhosis and hepatocellular carcinoma than in those without severe liver damage. It is important to note that our conclusions were validated by two different genotyping assays. However, a potential bias in these investigations is the difficulty of determining the duration of HCV infection, which may influence the severity of associated liver diseases. A major advantage of our present investigation is that the exact duration of HCV infection was known in each case. Indeed, the patients had HCV viremia detectable both before and after liver transplantation. Furthermore, because they were also selected for the availability of systematic histological follow-up and histological monitoring of liver graft guided by monthly laboratory tests, the onset of acute hepatitis and CAH could be dated precisely. However, it can be argued that, because of all these severe criteria for selection, the prevalence of genotypes and the natural history in the studied patients may be different from those in the general population undergoing transplantation for end-stage liver diseases due to HCV. Although this cannot be definitely excluded, it should be emphasized that

1094

FERAY ET AL.

GASTROENTEROLOGY Vol. 108, No. 4

A

B

100

'°°t

75 4

t

/

type lb

75

type lb

50

0

i

0

other types

I

i

25

%

j 12

,

~

,

24

~ months

0

%

36

i

0

i 12

,

other types

i 24

,

i months 36

N=49

C

100

#

I

25

N=60

D

100

75,

75 F type lb

50.

I

# 25.

%

?

50

]

Ij f,

0

,

0

~ 12

,

I

50

I

j 24

I I

25

other types

,

~ months 36

N=60

1-- type lb

%

other types

0

months 0

12

24

36

N=49

Figure 2. Actuarial rate of (A and B) acute hepatitis and (C and D) CAH in 60 patients tested using LIPA (A and C) and in 49 patients tested using type-specific nested PCR (B and D).

it would not change the main conclusion of our study showing the influence of HCV genotypes on severity of liver disease after liver transplantation. Another potential problem in these patients is that associated lesions such as rejection and other causes of hepatitis can theoretically hinder the interpretation of recurrent liver disease caused by HCV. However, rejection with venous endothelitis and cytomegalovirus-related hepatitis (which never progresses to CAH) were clearly different from rejection with hepatitis caused by HCV. It is noteworthy that the incidence of acute or chronic rejection did not depend on the HCV type. Another major feature of the transplantation in this setting is the very high rate of recurrent hepatitis: the rapid progression to CAH occurs in a context of intense viral replication, thus reflecting the specific cytopathogenicity

of the different HCV genotypes. By using b D N A technology for the quantitation of serum HCV RNA, we confirmed previous results suggesting significantly enhanced viral replication in these patients undergoing immunosuppressive therapy. W e also found that the amount of serum HCV RNA was positively correlated with the occurrence of hepatitis, suggesting that the severe histological course (contrary to the generally milder course in immunocompetent subjects 22) was caused by intense HCV replication. The results of this study showing that infection by HCV type l b is associated with more severe recurrent liver disease after transplantation than by the other HCV genotypes raise important issues with regard to the mechanism involved in this pathogenicity. Several hypotheses can be envisaged. HCV type lb may have a

April 1995

HCV GENOTYPES IN LIVER TRANSPLANT RECIPIENTS

HCVeq/mlserum

others have only HVR-1. This may lead to suggestions that HCV type lb is more versatile, enabling it to escape from host immunity. In conclusion, this study shows that the relative pathogenicity of different HCV genotypes can differ markedly, a finding that may have implications for the management of liver transplantation.

I08 o o ooD o

10 7

o o go []

o

oo

o []

o o o

o []

1095

o

References 1 06

Detection limit of bDNA

H

H*

type II

N

other types

Figure 3. Serum HCV RNA levels in patients according to HCV types

and histological diagnosis after liver transplantation. H, acute hepatitis or CAH; N, no hepatitis. *One patient was tested both during hepatitis and at the time of CAH.

higher replication rate and thus lead to earlier recurrence and, therefore, more severe liver disease after liver transplantation. However, we failed to detect any significant difference in serum HCV RNA levels before transplantation among patients infected by the different types of HCV. After transplantation, serum HCV RNA levels were higher in patients with hepatitis than in those with a normal liver. However, when patients with a comparable severity of liver histology were analyzed, we did not find a significant difference in the replication levels of the HCV genotypes. This observation is consistent with our recent results obtained in immunocompetent patients with CAH .2 but should be now validated in a larger series of transplant patients. In addition, it should be emphasized that, because of the very high prevalence of genotype lb in our patients with end-stage liver disease, we could not separately analyze the various n o n - l b genotypes. Other properties o f H C V l b should also be considered. The viral genome may encode for protein(s) with an enhanced cytopathic effect on liver cells. It has been reported that modifications in the processing of some pestivirus-encoded nonstructural proteins can generate an enhanced cytopathic effect. 23 It is also plausible that different HCV-encoded epitopes interact differently with the host-immune response. Indeed, a number of studies have shown significant genetic variability in the HCV genome during the course of HCV infection in a given subject. It has been postulated that amino acid changes in HCV proteins are caused by immune selection. Furthermore, HCV type l b has two hypervariable regions in the envelope domain (HVR-1 and HVR-2), whereas

1, McOmish F, Chan S, Dow B, Gillon J, Frame W, Crawford R, Yap P, Follett E, Simmonds P. Detection of three types of hepatitis C virus in blood donors: investigation of type-specific differences in serologic reactivity and rate of alanine aminotransferase abnormalities. Transfusion 1993; 33:7-13. 2. Okamoto H, Kueai K, Okada S, Yamamoto K, lizuka H, Fukuda S, Tsuda F, Mishir OS. Full-length sequence of a hepatitis C virus genome having poor homology to reported isolates: comparative study of four distinct genotypes. Virology 1992; 188:331-341. 3. Cha T, Beall E, Irvine B, Kolberg J, Chien D, Kuo G, Urdea M. At least five related, but distinct, hepatitis C viral genotypes exist. Proc Natl Acad Sci USA 1992;89:7144-7148. 4. Nakao T, Enomoto N, Takada N, Takada A, Date T. Typing of hepatitis C virus genomes by restriction fragment length polymorphism. J Gen Virol 1991; 72:2105-2112. 5. Simmonds P, McOmish F, Yap O, Chan S, Lin C, Dubleisko G, Saeed A, Holmes E. Sequence variability in the 5' non coding region of hepatitis C virus: identification of a new virus type and restrictions on sequence diversity. J Gen Virol 1993;74:661668. 6. Simmonds P, Alberti A, Alter H, Bonino F, Bradley D, Br6chot C, Brouwer J, Chan S, Chayama K, Chen D, Choo Q, Colombo M, et al. A proposed system for the nomenclature of hepatitis C viral genotypes. Hepatology 1994;19:1321-1324. 7. Finkelstein S, Sayegh R, Uchman S, Christensen S, Swalsky P. HCV undergoes extensive mutational change in NS5 region in association with relapse/breakthrough following alpha-interferon therapy. Hepatology 1992; 16:132-136. 8. Takada N, Takase S, Enomoto N, Takada A, Date T. Clinical backgrounds of the patients having different types of hepatitis C virus genomes. J Hepatol 1992; 14:35-40. 9. F6ray C, Samuel D, Thiers V, Gigou M, Pichon F, A Bismuth M, Reynes, Maisonneuve P, Bismuth H, C Br6chot. Reinfection of liver graft by hepatitis C virus after liver transplantation. J Clin Invest 1992;89:1361-1365. 10. Wright TL, Donegan E, Hsu H, Fermi L, Lake J, Kim M, Combs C, Fennessy S, Roberts JP, Asher NL, Greenberg HB. Recurrent and acquired hepatitis C viral infection in liver transplant recipients. Gastroenterology 1992; 103:317-322. 11. Ferrel L, Wright T, Roberts J, Ascher N, Lake J. Hepatitis C viral infection in liver transplant recipients. Hepatology 1992; 16:865-876. 12. F6ray C, Gigou M, Samuel D, Paradis V, David M, Reynes M, Br6chot C, Bismuth H. The course of hepatitis C infection after liver transplantation. Hepatology 1994;20:1137-1143. 13. Chazouill~res O, Kim M, Fermi L, Robert J, Ascher N, Wilber J, Neuwald P, Pescador RS, Urdea M, Wright T. Quantitation of hepatitis C virus RNA in liver transplant recipients. Gastroenterology 1994; 106:994-999. 14. Okamoto H, Sugiyama Y, Okada S, Kurai K, Akahane Y, Sugai Y, Tanaka T, Sato K, Miyakawa Y, Mayumi M. Typing hepatitis C virus by poiymerase chain reaction with type-specific primers: application to clinical surveys and tracing infectious sources. J Gen Virol 1992; 73:673-679.

1096

FERAY ET AL.

16. F6ray C, Gigou M, Samuel D, Bernuau J, Reyes G, Reynes M, Bismuth H, Br~chot C. Hepatitis C virus RNA and hepatitis B virus DNA in fulminant hepatitis. Gastroenterology 1993;104: 549-555. 17. Stuyver L, Rossau R, Wyseur A, Duhamel M, Vanderborght B, Maertens HHG. Typing of hepatitis C virus isolates and characterization of new subtypes using a line probe assay. J Gen Virol 1993; 74:1093-1102. 18. Knodell RG, Ishak RG, Black WC, Chen TS, Craig R, Kaplowitz N, Kiernan TW, WoNman J. Formulation and application of a numerical scoring system for assessing histological activity in asymptomatic chronic active hepatitis. Hepatology 1981;1:431-435. 19. Giannini C, Thiers V, Nousbaum J, Stuyvers L, Maertens G, Br6chot C. Comparative evaluation of type specific probes based assays for hepatitis C virus genotyping. J Hepatol (in press). 20. Pozzato G, Kaneko S, Morretti M, Croce L, Franzin F, Unoura M, Bercich L, Tiribelli C, Crovatto M. Different genotypes of hepatitis C virus are associated with different severity of liver disease. J Med Virol 1994;43:291-296. 21. Nousbaum JB, Pol S, Gigeu M, Okamoto H, F~ray C, Poussin K,

GASTROENTEROLOGY Vo1.108, No. 4

Paterlini P, Berthelot P, Mishiro S, Brechot C. Evidence for take over of HCV type II on type I among European patients with cirrhosis with and without HCC. Ann Intern Med 1995;122:161168. 22. Alter M, Margolis H, Krawczynski K, Judson F, Mares A, Alexander J, Hu P, Miller J, Gerber M, Sampliner R, Meeks EL, Beach M. The natural history of community-acquired hepatitis C in United States. N Engl J Med 1992;327:1899-1905. 23. Collet MS, Moennig V, Horzinek MC. Recent advances in pestivirus research. J Gen Virol 1989;70:253-266. Received May 28, 1993. Accepted December 7, 1994. Address requests for reprints to: Cyrille F~ray, M.D., Centre Hepato-biliaire, H6pital Paul Brousse, 14 Avenue Paul Vaillant Couturier, Villejuif 94800, France. Fax: (33) 1-45-59-38-57. Supported by grants from the Institut National de la Sant~ et de la Recherche M6dicale (INSERM Unit6 370 and CRE no. 931402), the Assistance Publique-HSpitaux de Paris, European Community, Caisse Nationale Assurance Maladie, Association pour la Recherche Centre le Cancer, and Diagnostic Pasteur-Sanofi.