HCV Eradication in Primary or Secondary Prevention Optimizes Hepatocellular Carcinoma Curative Management

The outcomes of patients with hepatitis C virus (HCV)-related cirrhosis who benefit from curative interventions for hepatocellular carcinoma (HCC) remain poorly defined. In this situation, the prognosis is not only driven by cancer recurrence, which is expected to occur in 50% patients at 2–3 years (1), but also progression toward end-stage liver disease (ESLD; ref. 2). Prospective studies have established that HCV clearance decreases the risk of HCC and ESLD, whether this is obtained using IFN (3, 4) or direct-acting antiviral agents (DAA; refs. 5, 6). However, the impact of HCV eradication in patients with HCC remains controversial.

Most of the available data have been obtained by meta-analyses of heterogeneous small sized cohorts conducted in the IFN era, with conflicting results (7, 8). The results are conflicting because of (i) the aggregation of heterogeneous and monocentric small sized cohorts involving major selection biases, and (ii) the low rates of sustained virologic responses (SVR) in patients with cirrhosis, reaching around 50% following IFN-based therapy. Furthermore, although there had previously been speculation regarding the potential benefits of IFN on HCC recurrence (9) randomized controlled trials failed to demonstrate the efficacy of IFN-based adjuvant therapy in patients undergoing curative HCC procedures (10, 11). Finally, the controversy surrounding the potential deleterious impact of DAAs on the risk of HCC recurrence (and potential higher tumoral aggressiveness) has not only probably limited prescriptions in these nevertheless priority candidates for antiviral treatment, but also prevented any definite conclusions based on rigorous prospective data from being drawn (12).

The ANRS CO12 CirVir cohort is a prospective cohort of French patients with HCV-related cirrhosis included since 2006 in HCC surveillance programs. It offers long follow-up periods during both IFN- and DAA-based regimens (3, 13). In this setting, a significant proportion of patients developed HCC, and their SVR status has been recorded in a protocol-driven framework and previously analyzed as a time-dependent covariate (6, 14, 15). The aim of this analysis was to assess the benefits of HCV eradication following a first-line curative procedure.

This study was sponsored and funded by the ANRS. The protocol was approved by an Ethics Committee (Comité de Protection des Personnes, Aulnay-sous-Bois, France) and complied with the ethical guidelines of the 1975 Declaration of Helsinki. All patients gave their written informed consent to participate in the cohort (http://anrs.fr).

This work constituted an ancillary study derived from the CirVir cohort (13) with specific goals and objectives redefined according to the STROBE statement (16). Patients were recruited in 35 French clinical centers between 2006 and 2012. Their selection criteria were: (i) aged 18 years and over; (ii) histologically proven cirrhosis, whatever the timing of the biopsy; (iii) HCV antibody-positive; (iv) absence of previous complications of cirrhosis; (v) patients belonging to Child-Pugh class A. Doppler US examinations were performed every 6 months. In the event of any focal lesions being detected by US, a diagnostic procedure using contrast-enhanced imaging (CT scan or MRI) and/or guided biopsy was performed according to the 2005 American Association for the Study of Liver Diseases guidelines (17) updated in 2011 (18). For this analysis, only patients treated using a first-line curative approach consisting of percutaneous ablation (PA) or hepatic resection (HepR) were considered (17–19). The index date (baseline) was the day the procedure was performed. The treatment was considered to be successful if an absence of viable tumor was confirmed by a three-phase contrast CT or MR examination within 3 months of the procedure.

Regular endoscopic surveillance was ensured. In the event of esophageal varices, preventive therapy was recommended using either beta-blockers or endoscopic ligation (20). All events that occurred during follow-up, whether they were liver-related or not, were recorded on the basis of information obtained from the medical records of patients in each center. In particular, all episodes of liver decompensation encompassing ascites, hepatic encephalopathy, and gastrointestinal bleeding were described, as well as their severity, management according to international recommendations and outcomes. All extrahepatic events occurring during follow-up were also recorded (21). Likely causes of deaths were established.

A total of 1,323 patients with HCV-related compensated cirrhosis were included in the CirVir cohort (Fig. 1). During a median follow-up of 68.3 months, 218 patients developed HCC and were considered for this analysis. Ninety-two of the latter were excluded, either because of incomplete data regarding HCC management (n = 7) or because PA or HepR was not used as first-line HCC therapy and/or did not enable a radiologic remission within 3 months (n = 85). Among the latter, 11 had undergone PA or HepR without remission and the remaining 74 received first-line noncurative procedures including embolization techniques (alone or combination with other treatments), systemic therapy, or best supportive care. Following this selection process, 126 patients (95 ablations and 31 resections leading to HCC remission within three months) constituted the population under study.

Before February 2014 (22), all the commercially available antiviral therapies initiated during follow-up were IFN-based (except for patients included in clinical trials testing DAAs). Patients with HCV genotype 1 or 4 infection received Peg-IFN plus a standard dose of ribavirin (RBV, 1,000 mg/day if body weight <75 kg or 1,200 mg/day if body weight >75 kg) for 48 weeks. Patients with HCV genotype 2 or 3 infection received Peg-IFN plus low-dose RBV (800 mg/day) for 16 or 24 weeks. After 2011, genotype 1 patients could also receive either 12 weeks of telaprevir (TVR, 750 mg every 8 hours) in combination with Peg-IFN and RBV, then 36 weeks of Peg-IFN/RBV, or 4 weeks (lead-in phase) of Peg-IFN and RBV and then 44 weeks of Peg-IFN/RBV and boceprevir (BOC, 800 mg every 8 hours) according to the European label. Since February 2014 (23), IFN-free regimens have gradually become available. All patients included in the CirVir cohort had received at least one course of IFN-based therapy before receiving DAAs.

An SVR was defined as an undetectable HCV RNA level determined using a qualitative PCR assay (<50 IU/mL) at the end of a 12-week untreated follow-up period (24). For all analyses, patients were classified according to the timing and status of their SVR as follows: patients who never achieved an SVR during follow-up, before or after an HCC diagnosis [non-SVR group, considering characteristics at the index date (i.e., date of HCC treatment) of the study], patients who achieved an SVR between inclusion in the cohort and HCC diagnosis (SVR before HCC group, considering characteristics at the index date of the study) and patients who achieved an SVR after HCC management (SVR after HCC group, considering characteristics at the date of SVR achievement). The non-SVR group included the few patients without data regarding their virologic response.

Time to HCC recurrence and overall survival (OS) were defined as the primary endpoints for this analysis, while time to 1- (very early) and 2-year (early) HCC recurrence, recurrence-free survival (RFS) and time to liver decompensation were considered as secondary endpoints. OS was defined as time from HCC treatment until patient death from any cause. RFS was defined as time from HCC treatment until the first event of either recurrence or death from any cause. Patients were censored if their outcome had not been determined by December 31, 2016, or by the last follow-up date before the end of the study. Patients who underwent liver transplantation were censored for analysis at the date of transplantation.

The characteristics of patients at the date of HCC treatment (non-SVR and SVR before HCC) or SVR achievement (SVR after HCC), at the diagnosis of HCC de novo and HCC recurrence were compared using one-way ANOVA or the Kruskal–Wallis rank-sum test for continuous variables, and the χ test or Fisher exact test for categorical variables. Time-dependent multivariate Cox proportional hazards regression was performed to determine the association between virologic response and time to HCC recurrence, OS, RFS, and time to liver decompensation, conditional on covariates at HCC treatment for all patients and on covariates at the time of SVR for those patients achieving SVR after HCC treatment.

In addition to the SVR variable, other time-dependent covariates included clinical and biological features measured at baseline and updated at SVR achievement [namely, aspartate aminotransferase, alanine aminotransferase, gamma-glutamyl transferase, α-fetoprotein, prothrombin time, albumin, bilirubin, platelet count, body mass index [BMI], dyslipidemia, diabetes, arterial hypertension, ongoing alcohol/tobacco/drug consumption]. The association between virologic status and the hazard of HCC recurrence was assessed by multivariate analysis, entering all variables associated with HCC risk at the P < 0.20 level in univariate analysis, then applying a backward stepwise approach to retain significant factors at the P < 0.10 level in the final model (model 1). DAA intake status was then further entered as a time-dependent binary variable to examine its effect on HCC recurrence risk independently of SVR status (model 2). SVR status and DAA treatment were systematically retained in the models as the main predictors of interest.

Variables for which missing data rate was lower than 20% were imputed using the multiple imputation by chained equation procedure in all multivariate Cox regression analyses. Univariate analysis was performed by computing unadjusted HRs along with their 95% confidence intervals (95% CI) using Cox proportional hazard regression models performed on the raw (unimputed) data.

Because conventional Kaplan–Meier curves are inadequate to depict survival curves according to time-dependent exposures, a clock reset approach (4) was used to build cumulative incidence curves as a function of the three different patient groups of interest. By definition, patients belonging to the “SVR after HCC” group were non-SVR at baseline and achieved SVR during follow-up before the onset of the studied outcome. Therefore, in view of the previously described three-group categorization, patients in the “SVR after HCC” group, switching to SVR from the non-SVR status were censored at the time of SVR. Follow-up subsequent to the switch was then reset as time zero for those patients in the SVR after HCC group. Survival curves using this approach were provided for illustrative purposes, along with unadjusted P values formally estimated using time-dependent Cox proportional hazard modeling, as previously detailed. Survival curves according to DAA intake were constructed using the same procedure, along with the P values estimated by a time-dependent Cox proportional hazard regression model.

Statistical analyses were performed using Stata 13.0 (StataCorp). P values <0.05 were considered to be statistically significant.

For descriptive purposes, the 126 patients with HCC were divided as described above into the non-SVR group (n = 74), SVR before HCC group (n = 28), and SVR after HCC group (n = 24). Follow-up was recorded up to 31 December 2016, at which date the median duration of follow-up since HCC treatment was 26.0 months (IQR: 14.1–42.7). Table 1 shows the patient and HCC characteristics of the three groups, considering the time of enrolment for all patients as the date of HCC treatment. Except for higher AFP levels in non-SVR patients, the tumor burden was similar in all groups, as was the proportion of PA and HepR, with PA representing the most frequently applied procedure.

During a mean follow-up of 26.0 months after implementation of a curative procedure, 59 patients (43%) experienced an overall recurrence (1- and 3-year incidences of 19.5% and 57.4%, respectively). These recurrences tended to be distributed at the same proportions in each group [41/74 (55.4%) vs. 11/28 (39.3%) vs. 7/24 (29.2%) for non-SVR, SVR before or after HCC, respectively; P = 0.054), Supplementary Table S1]. Most recurrences [n = 33 (73.3%)] were classified within the Milan criteria (Supplementary Table S1). Although this distribution was comparable in all groups, patients who achieved an SVR were more frequently retreated with sequential PA than non-SVR patients (83.3% vs. 53.9%; P = 0.032).

SVR achieved before or after HCC did not influence the risk of overall recurrence when compared with the non-SVR group [HR = 1.01 (0.52–1.92); P = 0.98 and HR = 0.51 (0.21–1.27), P = 0.15, respectively; Fig. 2A] or when these two SVR subgroups were combined [HR = 0.77 (0.43–1.39); P = 0.39]. Table 2 shows the results from univariate and multivariate analysis of the factors associated with time to HCC recurrence. SVR did not influence the probability of a relapse, whether it was achieved before or after HCC occurrence (Model 1) and adjusted for DAA intake (Model 2). The same observation applied when using raw unimputed data (Supplementary Table S2) and restricting the analysis to two-year (Supplementary Table S3) or one-year recurrences (Supplementary Table S4). Nor was recurrence-free survival impacted by SVR status (Supplementary Fig. 1), adjusted or not for DAA intake (Supplementary Tables S5 and S6).

During follow-up, 46 patients (36.5%) died and 18 were transplanted (including eight without HCC). Death rates were higher in the non-SVR group than in the SVR before or after HCC groups [40/68 (58.8%) vs. 3/28 (10.7%) vs. 3/30 (10.0%), respectively; P < 0.001]. Patients who achieved an SVR, whether before or after HCC, had an improved survival compared with patients without an SVR [HR = 0.24 (0.08–0.80), and HR = 0.27 (0.08–0.89); global P = 0.009; Fig. 2B]. When combined in a single group, the protective effect of SVR irrespective of time to its achievement was 0.26 (0.11–0.61); P = 0.002. SVR was confirmed by multivariate analysis as an independent predictor of survival (Table 3), whether it was not adjusted (Model 1) or adjusted to DAA intake (Model 2), while results using unimputed data yielded HRs of similar magnitude with a trend for statistical significance (global test P = 0.069).

The principal cause of death in the entire cohort was ESLD (n = 18; 41.9%), followed by HCC progression (n = 16; 37.2%) and extrahepatic causes (n = 9, 20.9%; missing data = 3). During follow-up, 41 patients (32.5%) experienced a first episode of liver decompensation following HCC treatment (ascites = 37, PHT-related bleeding = 12, encephalopathy = 24). These events were mostly seen in non-SVR patients (n = 35; 49.3%) than in patients belonging to the SVR before HCC (n = 3; 10.7%) or after HCC (n = 3; 27%) groups (P < 0.001). Although an SVR achieved after treatment for HCC did not impact this outcome [HR = 0.54 (0.15–1.91); P = 0.54], SVR achieved before HCC occurrence was associated with a lower risk of liver decompensation after HCC treatment [HR = 0.29 (0.09–0.93); P = 0.038; Fig. 2C]. When combined in a single group, patients who achieved HCV clearance before or after HCC had an overall lower risk of decompensation [HR = 0.37 (0.15–0.89); P = 0.028].

Among the 52 patients who achieved SVR, HCV eradication was seen in 27 after the implementation of DAAs (six before HCC occurrence, 21 achieving SVR after HCC treatment, including two patients starting DAAs before HCC management). To assess the impact of DAAs on HCC recurrence, we only considered patients who received DAAs following a curative antitumor procedure, irrespective of their SVR status, and compared their outcomes with those of other patients. These analyses involved 114 patients among whom 25 received DAAs after HCC treatment and before any recurrence (seven additional patients received DAAs after a first recurrence). SVR rate was thus 21/25 (85%) in this subgroup. Cases of HCC and the characteristics of patients at the time of HCC treatment as a function of DAA intake are shown in Supplementary Table S6. The characteristics of HCC recurrences were similar in the two groups [in-Milan tumors = 3 (50.0%) vs. 28 (75.7%) in the DAA group vs. others, respectively; P = 0.33]. Rates of HCC developing as a function of DAA intake were similar, and DAA intake was not associated with a higher risk of HCC recurrence over time [HR = 0.53 (0.23–1.21); P = 0.13; Fig. 3A]. This result was confirmed by multivariate analysis (Supplementary Table S7), irrespective of SVR status, and confirmed using unimputed data (Supplementary Table S8).

Patients who received DAAs following curative HCC management had a higher overall survival rate [HR = 0.19 (0.04–0.79); P = 0.022; Fig. 3B].

This analyses are consistent with a globally beneficial effect of HCV eradication on outcome following the curative management of HCC that has developed on cirrhosis. Although the analyses failed to show a lower probability of tumor recurrence, SVR status, whatever its time of achievement and the regimen used, was associated with improved overall survival. This observation can be explained by the long-term preservation of liver function, leading to a decrease in episodes of hepatic decompensation and increased rates of sequential HCC curative procedures in the event of a relapse of cancer. Finally, these analyses also confirm previously reported data from the CirVir cohort (25) suggesting the lack of a deleterious effect of DAAs on HCC recurrence.

Most studies published to date retrospectively assessed the outcomes of aggregated cohorts and introduced major selection bias that prevented any definite conclusions being drawn regarding HCC recurrence (26, 27). As a general rule, overall survival is able to capture the complexity of interactions between the recurrence of an oncologic process, the progression of liver failure and the intervention of comorbidities as extrahepatic causes of death accounting for 20.9% of cases. In this context, the prospective design of the CirVir cohort has enabled us to conclude that the beneficial effect of viral clearance on survival is mainly related to a reduction in non-HCC liver-related deaths and events (Fig. 2A and B), which is supported by complex multivariate analyses taking account of numerous confounders (Table 3).

The lack of a direct effect of SVR on HCC recurrence is unambiguously illustrated through the rigorous prospective design of the CirVir cohort (Fig. 2A) and the performance of complex analyses (Table 2). This observation can be explained by several facts. First, the effectiveness of viral clearance on the carcinogenic process may differ over time after curative HCC therapy and this efficacy may be greater two years after HCC therapy. In this setting, it is not surprising that the analyses focused on the risk of early recurrences at one or two years (Supplementary Tables S2 and S3) did not show any differences in relapse rates. Second, in most cases, the curative management of HCC in the CirVir cohort took the form of ablation procedures, which tend to be applied in patients with more advanced liver disease and/or comorbidities. In this setting, the proportion of progression to ESLD, as well as the high rates of non-HCC–related deaths representing 62.8% of cases as competing risks, may decrease an observational effect of SVR on tumor relapse. The relative contribution of these tumor- and liver-related factors to the impact of SVR could therefore be clarified by longer follow-up of the cohort. This assumption was tested by analyzing recurrence-free survival (Supplementary Fig. S1; Supplementary Table S4), which confirmed the lack of benefit of HCV clearance regarding a relapse of liver cancer, irrespective of non-HCC–related deaths.

Following the breakthrough of DAAs to treat HCV, a wave of papers suggested a potential increase in HCC recurrence after DAA implementation (28, 29). Mostly, based on retrospective single-center reports, it rapidly became apparent that they were hampered by various biases in patients' selection and interpretation of results (30). These alarming reports were, however, counterbalanced by the findings of several studies, including preliminary analyses performed on the CirVir cohort in 2016 (25). In this analysis, we only considered patients whose treatment for HCC had involved a first-line curative procedure. DAA implementation following tumor management was not associated with a higher risk of relapse (Fig. 3A), which was confirmed by multivariate analyses taking account of confounders and SVR status (Supplementary Table S5). This lack of a deleterious effect was furthermore supported by the increased survival of patients who received DAAs before or after HCC development (Table 4 and Fig. 3B). These data provide further support to the suggestion that DAAs are safe in patients who have achieved an effective HCC remission (31). In addition, our findings do not confirm previous reports suggesting lower SVR rates following DAAs implementation in patients with HCC (32), an observation which must be interpreted with caution given the small-size sample (n = 21) of highly selected patients in our study, and in whom first-generation regimen were sometimes prescribed.

Apart from its observational design, the main limitation of this study concerned its relatively small sample size. This was related to the conduct of analyses in carefully selected patients only, so as to limit measurement biases as much as possible. Patients indeed presented incidental cases of HCC diagnosed during surveillance programs performed in a larger cohort of 1,671 patients belonging to the CirVir cohort, which allowed us to perform accurate analyses that took account of SVR as a time-dependent covariate. This approach also ensured the quality of follow-up and a recording of events that was standardized according to the protocol in the 35 different centers including patients. In addition, we carefully reviewed all 218 incidental cases of patients managed for HCC and retrospectively excluded those who were finally not considered in a curative setting following the performance of resection or ablation (see Fig. 1); the absence of such a rigorous approach can be considered as a major ascertainment bias that was common to most previous studies in the field, in particular, those which suggested a deleterious effect of DAAs on HCC recurrence (28, 29). Nevertheless, because randomized studies comparing the outcomes of patients with HCC treated or not with antivirals will most likely never be performed, these analyses, with all their acknowledged limitations, are probably as close as we can get to scientific reasons to recommend the achievement of HCV clearance in the global strategy for the management of patients with HCC. Finally, it is tempting to speculate that a longer follow-up of patients might be required to display any differences in long-term HCC recurrence (beyond 2–3 years).

In summary, the achievement of SVR in patients with cirrhosis before or after HCC development is not associated with a modified risk of short- or medium-term tumor recurrence following the implementation of a curative procedure. Nevertheless, HCV eradication enables optimal HCC management by preventing a deterioration of liver function and increasing rates of HCC recurrence re-treatment by means of percutaneous ablation, leading to an improvement in overall survival whatever the antiviral regimen.

P. Nahon reports personal fees from Abbvie, AstraZeneca, Bayer, BMS, EISAI, Gilead, Roche outside the submitted work. T. Asselah reports personal fees from Abbvie, Gilead, Janssen, and personal fees from MERCK outside the submitted work. S. Pol reports grants and personal fees from Gilead, Abbvie, personal fees from Janssen, MSD; personal fees from Shinogui, Biotest, Vivv, and personal fees from Novo Nordisk outside the submitted work. G. Pageaux reports personal fees from Gilead and personal fees from Abbvie outside the submitted work. E. Audureau reports personal fees from GBT and personal fees from Hemanext outside the submitted work. No disclosures were reported by the other authors.

P. Nahon: Conceptualization, resources, supervision, writing–original draft. R. Layese: Data curation, methodology. C. Cagnot: Project administration. T. Asselah: Investigation. D. Guyader: Investigation. S. Pol: Investigation. G.-P. Pageaux: Investigation. V. De Lédinghen: Investigation. D. Ouzan: Investigation. F. Zoulim: Investigation. E. Audureau: Formal analysis, writing–review and editing.

This study was sponsored by the ANRS (France REcherche Nord & sud Sida-hiv Hépatites: FRENSH). This work was also supported, in part, by a grant from AbbVie. This work is dedicated to the memory of Professor Jean-Claude Trinchet.

ANRS CO12 CirVir group:

Pierre Nahon¹, Tarik Asselah², Dominique Guyader³, Stanislas Pol⁴, Hélène Fontaine⁴, Dominique Larrey⁵, Victor De Lédinghen⁶, Denis Ouzan⁷, Fabien Zoulim⁸, Dominique Roulot⁹, Albert Tran¹⁰, Jean-Pierre Bronowicki¹¹, Jean-Pierre Zarski¹², Vincent Leroy¹², Ghassan Riachi¹³, Paul Calès¹⁴, Jean-Marie Péron¹⁵, Laurent Alric¹⁶, Marc Bourlière¹⁷, Philippe Mathurin¹⁸, Sebastien Dharancy¹⁸, Jean-Frédéric Blanc¹⁹, Armand Abergel²⁰, Lawrence Serfaty²¹, Ariane Mallat²², Jean-Didier Grangé²³, Pierre Attali²⁴, Yannick Bacq²⁵, Claire Wartelle²⁶, Thông Dao²⁷, Dominique Thabut²⁸, Christophe Pilette²⁹, Christine Silvain³⁰, Christos Christidis³¹, Eric Nguyen-Khac³², Brigitte Bernard-Chabert³³, Sophie Hillaire³⁴, Vincent Di Martino³⁵.

¹AP-HP, Hôpital Jean Verdier, Service d'Hépatologie, Bondy, Université Paris 13, Bobigny et INSERM U1162, Université Paris 5, Paris; ²AP-HP, Hôpital Beaujon, Service d'Hépatologie, and University Paris Diderot, Sorbonne Paris Cité, CRI, UMR 1149; ³CHU Pontchaillou, Service d'Hépatologie, Rennes; ⁴AP-HP, Hôpital Cochin, Département d'Hépatologie et INSERM UMS20 et U1223, Institut Pasteur, Université Paris Descartes, Paris; ⁵Hôpital Saint Eloi, Service d'Hépatologie, Montpellier; ⁶Hôpital Haut-Lévêque, Service d'Hépatologie, Bordeaux; ⁷Institut Arnaud Tzanck, Service d'Hépatologie, St Laurent du Var; ⁸Hôpital Hôtel Dieu, Service d'Hépatologie, Lyon; ⁹AP-HP, Hôpital Avicenne, Service d'Hépatologie, Bobigny; ¹⁰CHU de Nice, Service d'Hépatologie, et INSERM U1065, Université de Nice-Sophia-Antipolis, Nice; ¹¹Hôpital Brabois, Service d'Hépatologie, Vandoeuvre-les-Nancy; ¹²Hôpital Michallon, Service d'Hépatologie, Grenoble; ¹³Hôpital Charles-Nicolle, Service d'Hépatologie, Rouen; ¹⁴CHU d'Angers, Service d'Hépatologie, Angers; ¹⁵Hôpital Purpan, Service d'Hépatologie, Toulouse; ¹⁶CHU Toulouse, Service de Médecine Interne-Pôle Digestif UMR 152, Toulouse; ¹⁷Hôpital Saint Joseph, Service d'Hépatologie, Marseille; ¹⁸Hôpital Claude Huriez, Service d'Hépatologie, Lille; ¹⁹Hôpital St André, Service d'Hépatologie, Bordeaux; ²⁰Hôpital Hôtel Dieu, Service d'Hépatologie, Clermont-Ferrand; ²¹AP-HP, Hôpital Saint-Antoine, Service d'Hépatologie, Paris; ²²AP-HP, Hôpital Henri Mondor, Service d'Hépatologie, Créteil; ²³AP-HP, Hôpital Tenon, Service d'Hépatologie, Paris; ²⁴AP-HP, Hôpital Paul Brousse, Service d'Hépatologie, Villejuif; ²⁵Hôpital Trousseau, Unité d'Hépatologie, CHRU de Tours; ²⁶Hôpital d'Aix-En-Provence, Service d'Hépatologie, Aix-En-Provence; ²⁷Hôpital de la Côte de Nacre, Service d'Hépatologie, Caen; ²⁸AP-HP, Groupe Hospitalier de La Pitié-Salpêtrière, Service d'Hépatologie, Paris; ²⁹CHU Le Mans, Service d'Hépatologie, Le Mans; ³⁰CHU de Poitiers, Service d'Hépatologie, Poitiers; ³¹Institut Mutualiste Montsouris, Service d'Hépatologie, Paris; ³²Hôpital Amiens Nord, Service d'Hépatologie, Amiens; ³³Hôpital Robert Debré, Service d'Hépatologie, Reims; ³⁴Hôpital Foch, Service d'Hépatologie, Suresnes; ³⁵Hôpital Jean Minjoz, Service d'Hépatologie, Besançon. FRANCE.

Scientific Committee:

P. Nahon (principal investigator), R. Layese and F. Roudot-Thoraval (data management), P. Bedossa, M. Bonjour, V. Bourcier, S. Dharancy, I. Durand-Zaleski, H. Fontaine, D. Guyader, A. Laurent, V. Leroy, P. Marche, D. Salmon, V. Thibault, V. Vilgrain, J. Zucman-Rossi, C. Cagnot (ANRS), V. Petrov-Sanchez (ANRS).

Clinical centers (unit/participating physicians):

CHU Jean Verdier, Bondy (P. Nahon, V. Bourcier); CHU Cochin, Paris (S. Pol, H. Fontaine); CHU Pitié-Salpétrière, Paris (D. Thabut); CHU Saint-Antoine, Paris (L. Serfaty); CHU Avicenne, Bobigny (D. Roulot); CHU Beaujon, Clichy (P. Marcellin); CHU Henri Mondor (A. Mallat); CHU Paul Brousse (P. Attali); CHU Tenon, Paris (J.D. Grangé); CHRU Hôpital Nord, Amiens (D. Capron); CHU Angers (P. Calès); Hôpital Saint-Joseph, Marseille (M. Bourlière); CHU Brabois, Nancy (J.P. Bronowicki); Hôpital Archet, Nice (A. Tran); Institut Mutualiste Montsouris, Paris (F. Mal, C. Christidis); CHU Poitiers (C. Silvain); CHU Pontchaillou, Rennes (D. Guyader); CH Pays d'Aix, Aix-en-Provence (C. Wartelle); CHU Jean Minjoz, Besancon (V. Di Martino); CHU Bordeaux - Hôpital Haut-Leveque, Pessac (V. de Ledinghen); CHU Bordeaux - Hôpital Saint-André, Bordeaux (J.F. Blanc); CHU Hôtel Dieu, Lyon (C. Trepo, F. Zoulim); CHU Clermont-Ferrand (A. Abergel); Hôpital Foch, Suresnes (S. Hillaire); CHU Caen (T. Dao); CHU Lille (P. Mathurin); CH Le Mans (C. Pilette); CHU Michallon, Grenoble (J.P. Zarski); CHU St Eloi, Montpellier (D. Larrey); CHU Reims (B. Bernard-Chabert); CHU Rouen (O. Goria, G. Riachi); Institut Arnaud Tzanck, St Laurent-du-Var (D. Ouzan); CHU Purpan, Toulouse (J.M. Péron, L. Alric); CHU Tours (Y. Bacq).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.