We read with interest the article by Hanouneh et al.1 on the prediction of pegylated interferon plus ribavirin (P/R) response in liver transplant (LT) recipients with histologically proven recurrent hepatitis C. As highlighted by the authors, the management of problematic patients such as graft recipients can be eased by early kinetics of hepatitis C virus (HCV), that is, HCV-RNA clearance at week 4 [rapid virological response (RVR)] and a 2-log drop of HCV-RNA at week 12 [early virological response (EVR)], which predict treatment outcome. However, although in LT patients EVR has an absolute negative predictive value (NPV) for P/R treatment outcome,1-4 the NPV and positive predictive value (PPV) of RVR are far from being absolute, even in non-LT patients. We were therefore surprised by Hanouneh et al.'s findings of RVR having 100% PPV and 88% NPV in LT patients receiving P/R. We think that the RVR accuracy discrepancies between this study and other studies4, 5 might reflect differences in the performance of the studies; for example, many patients (38%) in the present series were not tested for serum HCV-RNA at week 4 of treatment. We treated with P/R 46 patients transplanted for end-stage HCV with inclusion criteria similar to those investigated by Hanouneh et al.1 The treatment was intended for 48 weeks, being independent of the genotype or virological response to either peginterferon alfa-2a or peginterferon alfa-2b weekly, whereas the ribavirin dosing was less than that in Hanouneh et al.'s study (400–600 versus 1000–1200 mg/day). Serum HCV-RNA levels were assessed at baseline and at 4 weeks, 12 weeks, 48 weeks, and 6 months after the completion of therapy, and most patients received growth factors. The 2 cohorts were similar for clinicovirological features but differed in terms of the prevalence of non-Caucasians and body mass index (lower in our cohort). Cyclosporine A (CSA)–based immunosuppression was higher in our patients than in the American cohort (48% versus 17%). The rates of SVR were 41% and 35%, respectively, with a preference for genotypes 2 and 3 in both (76% versus 87%). The SVR rates in the 2 cohorts are in line with those reported by recent studies on P/R-treated HCV recipients ranging from 30% to 45%.6 In our patients, RVR had 86% PPV and 78% NPV (see Table 1), that is, midway between the data of Hanouneh et al.1 and other data.3 The results indicate that RVR is a good predictor of treatment outcome in LT patients with recurrent hepatitis C; however, it is not accurate enough to become a stopping rule of P/R therapy. In our cohort, CSA-based immunosuppression was independently associated with SVR in multivariate analysis (odds ratio = 6.65, 95% confidence interval = 1.79–24.7, P < 0.005), and this supports the synergic anti-HCV effect of CSA and interferon already demonstrated in vitro and in vivo.5 This was not investigated by Hanouneh et al. because the majority of their patients received tacrolimus-based immunosuppression. The higher dose of ribavirin in the American patients with respect to ours might have been beneficial in terms of SVR, too.6 Maria Francesca Donato*, Francesca Agnelli*, Cristina Rigamonti*, Eliana Arosio*, Massimo Colombo*, * First Division of Gastroenterology, Fondazione IRCCS (Istituto di Ricovero e Cura a Carattere Scientifico) Ospedale Maggiore, Policlinico, Mangiagalli e Regina Elena, Milan, Italy.
Read full abstract