Clinical trial: peg-interferon alfa-2b and ribavirin for the treatment of genotype-1 hepatitis C recurrence after liver transplantation

Authors


Prof P. Andreone, Dipartimento Malattie Apparato Digerente e Medicina Interna, Azienda Ospedaliero-Universitaria Policlinico Sant’Orsola-Malpighi, Via Massarenti, 9-40138 Bologna, Italy.
E-mail: pietro.andreone@unibo.it

Summary

Background  Treatment of hepatitis C virus (HCV) recurrence after liver transplantation (LT) is difficult with low response rates.

Aim  To assess the safety and efficacy of pegylated-interferon (PEG-IFN) alfa-2b + ribavirin (RBV) in patients with post-LT recurrent genotype-1 HCV and to establish stopping rules according to response.

Methods  Fifty-three patients with post-LT HCV recurrence were enrolled. Patients received PEG-IFN alfa-2b 1.0 μ/kg/week plus RBV 8–10 mg/kg/day for 24 weeks. Those with ‘early virological response at week 24’ (EVR24) continued treatment for 24 weeks (group A). Patients without EVR24 were randomized to continue (group B) or to discontinue (group C).

Results  Overall sustained virological response (SVR) was 26% (14/53). Alanine aminotransferase, rapid virological response, EVR12, EVR24, undetectable serum HCV-RNA at weeks 12 (cEVR12) and 24 (cEVR24) were related to SVR. cEVR12 and cEVR24 (OR: 14.7; 95% CI: 2.02–106.4) were independent predictors of SVR. All patients with SVR, had cEVR12. No patient in groups B and C achieved end-of-treatment response. One patient in group B had SVR.

Conclusions  Pegylated-interferon alfa-2b was effective in one of four of patients with HCV genotype 1 after LT. Treatment should be discontinued in patients with no virological response at week 12. Further studies are needed to evaluate whether a longer treatment period may be beneficial in patients with ≥2 log10 drop in HCV-RNA at week 24.

Introduction

Hepatitis C virus (HCV)-related end-stage liver disease is the leading indication for liver transplantation (LT) in Western countries.1, 2 Unfortunately, HCV reinfection after LT is almost universal, with histological evidence of chronic hepatitis in approximately 50% of patients within 1 year3 and cirrhosis in about 30% after 5 years.4 The course of post-LT HCV liver disease is accelerated, representing a significant cause of death or graft loss.5 Therefore, the need to optimize outcomes following LT for hepatitis C is one of the major goals of the transplant community.

Anti-viral therapy with standard interferon (IFN) plus ribavirin (RBV) has been used with poor results in the transplant setting, with sustained virological response (SVR) rates ranging from 5% to 30%.6–10 Currently available regimens with pegylated-IFNs (PEG-IFN) in association with RBV are more effective than standard IFN plus RBV in treating chronic hepatitis C in immunocompetent patients with overall SVR rates of 40–55%,11, 12 the lowest rates of response being represented by genotype-1 patients. However, in the LT setting, sustained response rates are markedly lower and highly variable.13–19 In addition to the low SVR rate, in the transplant setting standard IFNs and the newest PEG-IFNs plus RBV are poorly tolerated and are associated with a high incidence of adverse events requiring dose reduction or drug discontinuation.20 It is still unclear whether IFN-based treatment may increase the risk of acute and chronic cellular rejection13, 21, 22 and cases of de novo autoimmune hepatitis (AIH) have recently been reported in HCV transplanted patients during PEG-IFN treatment.23–25 Thus, anti-viral treatment of LT recipients with established recurrence of hepatitis C is much more uncertain than in immunocompetent patients and several questions such as dosage and duration of the treatment schedule and stopping rules are still a matter of debate.

The aim of our study was to assess the safety and efficacy of PEG-IFN alfa-2b and RBV in a population of transplanted patients with recurrent genotype-1 hepatitis C. The secondary endpoint was to establish stopping rules according to treatment response.

Materials and methods

Patients and study design

Between October 2001 and April 2005, 53 patients with post-LT genotype-1 HCV recurrence, were consecutively enrolled at the Bologna Liver Transplantation Center (Figure 1). Inclusion criteria for treatment were: LT caused by HCV-related cirrhosis, detectable HCV-RNA by PCR, HVC genotype-1 infection, elevated (>1.0×) serum alanine aminotransferase (ALT) levels and histological features of HCV hepatitis in the graft on liver biopsy. Patients in whom previous treatment with standard IFN and RBV after LT was not effective were also considered eligible for the therapy. Exclusion criteria were: infection by HCV genotypes other than 1, evidence of decompensated liver disease, histological evidence of rejection and drug-related injury, HBsAg positivity, HIV positivity, moderate-to-severe anaemia (Hb < 10 g/dL), leukopenia (WBC < 1500/μL), thrombocytopenia (PLT < 50 000), impaired renal function (creatinine clearance <50 mL/min), significant history of cardiovascular and psychiatric diseases and ongoing alcohol abuse and previous post-LT treatment with PEG-IFN. All patients gave written informed consent according to the Ethical Committee Procedures of our Hospital for drugs approved for a specific disease in the absence of a specific indication for a particular group of patients.

Figure 1.

 Flow chart of study design. EVR24, early virological response at week 24 = undetectable serum HCV-RNA or a >2 log10 drop from baseline obtained within the 24th week; ETR, end-of-treatment response; SVR, sustained virological response.

For the first 24 weeks, all patients received 1.0 μ/kg once weekly of PEG-IFN alfa-2b (Peg-Intron; Schering-Plough, Segrate, Italy; this dose was chosen to minimize the cytopenia typical of transplanted patients and the low tolerability of the drug20), plus RBV (Rebetol; Schering-Plough) at a dose of 8–10 mg/kg/day, lower than that recommended because of the known reduction in renal function caused by calcineurin inhibitors. After 24 weeks, patients with ‘early virological response (EVR24)’ continued treatment for an additional 24-week period (group A). EVR24 was defined as undetectable serum HCV-RNA or a ≥2 log10 drop from baseline obtained within the 24th week. Definition of EVR24 was based on the consideration that transplanted patients usually have high pre-treatment viral load and a 2 log10 fall after 24 months could be considered a satisfactory response suggesting to continue treatment until the end of the planned 12 months course. Moreover, there are no guidelines on response criteria in the transplant setting. Nevertheless, the different pattern of response was considered in the statistical analysis.

Patients without EVR24 were randomized with a 1:1 ratio to continue treatment for further 24 weeks (group B) or to discontinue (group C). All patients were followed up for 24 weeks after the end of treatment.

Complete blood count was carried out at baseline and weekly for the first 4 weeks using conventional tests and then monthly until the end of the study. Biochemical tests were conducted at baseline and monthly. Thyroid function tests and non-organ-specific autoantibody determination were carried out at baseline and then every 3 months; in the case of the thyroid dysfunction, anti-thyroid antibodies were searched.

Efficacy assessments

Quantitative HCV-RNA (Versant HCV-RNA 3.0 bDNA; Bayer Diagnostics (Berkeley, CA, USA) detection limit 3200 copies/mL) was measured at baseline. Qualitative HCV-RNA (Versant TMA; Bayer Diagnostics, limit of quantification of 5.3 IU/mL) was assessed at weeks 4, 12, 24, 36 and 48 during treatment and at weeks 4, 12 and 24 during post-treatment follow-up. If determination was positive, a quantitative test was performed. Rapid virological response (RVR) was defined as undetectable serum HCV-RNA after 4 weeks of treatment. EVR at week 12 (EVR12) and at week 24 (EVR24) were defined as undetectable serum HCV-RNA or a ≥2 log10 drop from baseline. Undetectable serum HCV-RNA at weeks 12 and 24 were defined as complete EVR12 (cEVR12) and complete EVR24 (cEVR24) respectively. Undetectable serum HCV-RNA at the end of the treatment was considered an end-of-treatment response (ETR). Undetectable serum HCV-RNA at week 24 after treatment cessation (SVR) was the primary endpoint.

Safety assessments

Safety was assessed by clinical and laboratory testing, and by evaluating all the adverse events which were reported at each visit. According to the protocol, growth factors were recommended to allow the best compliance of patients in relation to predictable haematological side effects such as leukopenia and anaemia. Granulocyte colony-stimulating factor (G-CSF; Granulokine; Roche, Italy) at a dose ranging from 300 to 600 μg/week was used in the case of an absolute neutrophil count of <750/mm3. Erythropoietin (EPO; NeoRecormon; Roche) from 5000 to 30 000 IU/week was used to treat anaemia (Hb levels <10 g/dL). Peg-IFN was stopped, if significant side effects occurred and/or if cytopenia persisted (neutrophil count <750/mm3, platelet count <20 000/mm3). The RBV dose was reduced by 200 mg in the case of anaemia (Hb < 10 g/dL) unresponsive to EPO. Doses were further reduced at weekly intervals to maintain Hb levels >10 g/dL; for Hb level ≤9 g/dL, RBV was discontinued. A liver biopsy was performed before starting the therapy and repeated if liver function tests showed an unexplained abnormality occurring within the study period. Necroinflammatory lesions and fibrosis were graded according to the Knodell histological score26 by a single pathologist (DEA).

Statistical analysis

Data were analysed on an intention-to-treat basis and per protocol. Quantitative variables were presented as medians (ranges). Nonparametric tests were used to compare the differences between the groups (Mann–Whitney U-test for continuous variables; χ2 test or Fisher’s exact test for categorical variables). A two-sided P-value of <0.05 was considered statistically significant. Baseline predictors of SVR were evaluated by univariate and multivariate analyses. Variables achieving statistical significance or marginal significance (P < 0.10) on univariate analysis were entered into multivariate forward stepwise logistic regression analysis to identify significant variables independently associated with SVR. The odds ratios (ORs) and the associated 95% confidence intervals (CIs) in the presence of interactions were also estimated. All data analyses were conducted using the Statistical Package for Social Science (spss Version 11.5; SPSS, Inc., Chicago, IL, USA).

Results

Fifty-three patients (37 males and 16 females, median age 59 years, range 22–68) were enrolled in the study. Their baseline characteristics are reported in Table 1. At enrolment, seven (13%) patients showed histological, ultrasonographic and laboratory features of cirrhosis with or without signs of portal hypertension.

Table 1.   Baseline characteristics of patients at enrolment
  1. ALT, alanine aminotransferase; LT, liver transplantation.

Gender (M/F)37/16
Age59 (22–68)
BMI24.2 (15.5–40.5)
Months form LT14 (3–151)
Previous acute cellular rejection10 (19%)
Previous cytomegalovirus infection9 (17%)
Viral load (MEq/mL)10.6 (0.009–40)
ALT (IU/L)130 (52–945)
Neutrophil count (×103/mmc)2.9 (1.03–7.4)
Haemoglobin (g/dL)12.6 (10–16.1)
Cirrhosis7 (13%)
Pre-LT anti-viral treatment24 (45%)
Post-LT anti-viral treatment14 (26%)
Induction immunosuppression
 Monoclonal antibodies7 (13%)
 Steroids43 (81%)
 Other3 (6%)
Maintenance immunosuppression
 Ciclosporin26 (49%)
 Tacrolimus14 (26%)
 Ciclosporin + steroids5 (9%)
 Tacrolimus + steroids8 (15%)

Overall virological response

Nine patients (17%) achieved RVR and 21 (40%) EVR12. Among patients with EVR12, 16 (76%) had cEVR12. Twelve patients (23%) discontinued treatment between weeks 12 and 24 (six for disease decompensation, four for severe flu-like syndrome and fatigue, one for liver abscess and one for pericarditis). Notably, all patients who withdrew from treatment for decompensation had cirrhosis at enrolment. After 24 weeks of treatment, 23 patients (43%) had EVR24. Among them, 18 (34%) had cEVR24. According to the protocol, patients with EVR24 continued treatment up to 48 weeks (group A), while the 18 patients without EVR24 were randomly assigned to continue (group B, nine patients) or to stop (group C, nine patients) treatment. Overall, ETR was observed in 17 of 53 patients (32%) and the primary endpoint of SVR was obtained in 14 patients (26%). Demographic and clinical characteristics of the SVR patients are shown in Table 2.

Table 2.   Comparison of clinical variables associated with response between responders and nonresponders
VariablesSVR (14 patients)Nonresponders (39 patients)P
  1. SVR, sustained virological response; ALT, alanine aminotransferase; HCV, hepatitis C virus; RVR, rapid virological response; EVR12 and EVR24, early virological responses at weeks 12 and 24 respectively; LT, liver transplantation.

Gender (M/F)9/528/11N.S.
Age58.5 (45–67)59 (22–68)N.S.
Body mass index26 (21.4–31.5)24 (15.5–40.5)N.S.
Months form LT25.5 (1–148)11 (1–151)N.S.
Previous acute cellular rejection2 (14%)8 (20.5%)N.S.
Previous CMV infection2 (14%)7 (18%)N.S.
Cirrhosis1 (7%)6 (15%)N.S.
Baseline ALT (U/L)88 (52–364)146.5 (54–945)0.028
Baseline HCV-RNA (mEq/mL)8.06 (0.009–40)13 (1.1–40)N.S.
RVR6 (43%)3 (8%)0.007
EVR1212 (86%)9 (23%)0.000
EVR2413 (93%)10 (26%)0.000
cEVR1212 (86%)4 (10%)0.000
cEVR2413 (93%)5 (13%)0.0001
Pre-LT anti-viral treatment7 (50%)17 (43.5%)N.S.
Post-LT anti-viral treatment2 (14%)12 (31%)N.S.
Immunosuppression
 Ciclosporin10 (19%)16 (41%)N.S.
 Tacrolimus2 (14%)12 (31%)N.S.
 Ciclosporin + steroids1 (7%)4 (10%)N.S.
 Tacrolimus + steroids1 (7%)7 (18%)N.S.

The predictive values of RVR, EVR12, cEVR12, EVR24 and cEVR24 for achievement of SVR were evaluated. The positive predictive values were 67% (six of nine), 57% (12/21), 75% (12/16), 57% (13/23) and 72% (13/18) respectively, while the corresponding negative predictive values were 82% (36/44), 94% (30/32), 95% (35/37), 97% (29/30) and 97% (34/35) respectively.

At univariate analysis, the variables associated with SVR (Table 2) were: baseline ALT, RVR, EVR12, EVR24, cEVR12 and cEVR24. Nonresponse to a previous post-OLT anti-viral treatment was not predictive of lack of response; nevertheless, among the 14 patients with a previous nonresponse, SVR was obtained in two (14.3%). Similarly, cirrhosis was not a predictor of nonresponse.

At multivariate analysis, the independent predictors of SVR were cEVR12 and cEVR24 (OR: 14.7; 95% CI: 2.02–106.4 for both).

In a per-protocol analysis, SVR was 48% (14/29); in fact, 24 patients withdrew from treatment prematurely because of adverse events (see below); two of them were HCV-RNA negative at the time of discontinuation and maintained response during follow-up.

Subgroup analysis

Group A.  At week 24 of treatment, 23 out 53 of the enrolled patients had an EVR and were assigned to group A. Of them 18/23 (78%) had cEVR24.

At the end of treatment, 17/23 patients (74%) achieved ETR; of them 15 (88%) had cEVR24. During post-treatment follow-up, six of 17 (35%) relapsed; therefore, the primary endpoint of SVR was reached in 11 patients (48%). Interestingly, all patients achieving SVR, had cEVR12.

In group A, four (16%) dropped out before completing the 48 weeks of treatment because of adverse events: one because of severe asthenia and three because of graft dysfunction compatible with ‘probable AIH’;27 this kind of graft dysfunction has been extensively described by Berardi et al.24 These three patients were HCV-RNA negative at the time of discontinuation and two of them persistently cleared the virus during the follow-up achieving SVR. Therefore, considering the two dropout patients who cleared the virus, the overall SVR in group A was 56.5% (13/23).

Groups B and C.  No patient in either group B or C achieved ETR. However, 1 month after the end of treatment, one patient in group B developed a graft dysfunction compatible with ‘probable AIH’ (AIH score +10)27 and cleared serum HCV-RNA achieving an SVR. Therefore, the SVR rate in group B was 11% (1/9).

One patient in group B withdrew from treatment at week 36 because of severe asthenia.

Side effects

Adverse events were common during therapy and are listed in Table 3.

Table 3.   Adverse events
Adverse eventsRate
  1. Values in parentheses represent percentages.

  2. AIH, autoimmune hepatitis; SLE, systemic lupus erythematous.

Anaemia34/53 (64)
Neutropenia14/53 (26)
Asthenia21/53 (40)
Fever14/53 (27)
Flu-like symptoms13/53 (24)
Probable AIH7/53 (13)
Inappetence7/53 (12)
Death6/53 (11)
Liver decompensation6/53 (11)
Depression5/53 (9)
Severe infections4/53 (7.5)
Late hepatic artery stenosis4/53 (7.5)
Autoimmune gastritis1/53 (2)
Autoimmune hyperthyroidism1/53 (2)
SLE1/53 (2)

Haematopoietic side effects occurred in a large number of patients, requiring specific intervention after 4 weeks of therapy. Thirty-four (64%) patients developed anaemia and 24 (35%) required EPO treatment. Fourteen (26%) developed neutropenia requiring G-CSF. Severe infections occurred in four patients (7.5%): two pneumonia, one pericarditis and one liver abscess, the latter discontinued the treatment at week 12.

It is important to emphasize that six of seven (86%) cirrhotic patients developed liver decompensation (four ascites, two hepatic encephalopathy) before week 24 and discontinued treatment. At enrolment, five had a Child–Pugh score of A5 and one B7. Their Model for End Stage Liver Disease (MELD) score ranged from 11 to 16. Three of them were re-listed for LT, but died while on the waiting list 1, 6 and 7 months later.

Four patients (7.5%) developed late hepatic arterial stenosis without other hypothetical risk factors except for IFN treatment; three of these have previously been described.28

Interestingly, we did not observe any cases of rejection. However, we recorded a high rate of immune-mediated diseases: one case of autoimmune gastritis diagnosed by appearance of antibodies to gastric parietal cells and hypergastrinaemia,29 one case of systemic lupus erythematous (SLE) associated with ‘probable AIH’ (AIH score +11),30 one case of thyroid dysfunction with appearance of anti-thyroperoxidases and anti-thyroglobuline antibodies and seven cases (13%) of ‘probable AIH’.24 Of these, latter two died, one had a graft failure and was re-listed for LT and four had clinical and biochemical improvement with steroid treatment. Overall, in our study population, 19% had immune-mediated diseases during the treatment period.

PEG-IFN or RBV dose adjustments were necessary in three (6%) and 21 (40%) patients, respectively. RBV was discontinued in four patients (7.5%). Overall, 24 patients (45%) withdrew from treatment for adverse events. The overall mortality rate was 11% (six of 53): three cirrhotic patients died of decompensation; two of ‘probable AIH’ and one of liver failure caused by severe HCV recurrence.

Discussion

Anti-viral treatment of HCV recurrence after LT is difficult for several reasons: liver transplanted patients are predominantly genotype 1, have a high viral load and an accelerated course of liver disease with rapid and frequent progression to cirrhosis. Moreover, IFN plus RBV-based treatments are poorly tolerated in transplanted patients who often have baseline leukopenia and anaemia, and may require reduction in the dose for haematological side effects. Many studies have been conducted in this field,6–10, 13–19 but most of them have been small and not randomized. Therefore, it is impossible to formulate treatment guidelines in terms of optimal dose, duration and stopping rules. This means that, in the studies conducted to date, different approaches have been used. As a result, patients are often over-treated for long periods in the absence of response or under-treated with inadequate doses of IFN and RBV.

Ours is the first study with the aim of giving indications on what to do in HCV genotype-1 patients with post-LT hepatitis who are not responders after 24 weeks of treatment with PEG-IFN plus RBV. We are aware that our definition of virological response at week 24 is different from the classical definition used not only in immunocompetent patients, but also in transplanted patients.31 However, our decision was based on the consideration that transplanted patients usually have a high pre-treatment viral load, concomitant immunosuppressive treatment, which may interfere with viral kinetic and low tolerability with difficulties in achieving full-dose treatment. Thus, a drop in HCV-RNA level ≥2 log10 after 24 weeks of treatment (EVR24) was defined as an inclusion criterion for randomization.

According to an intention-to-treat analysis, SVR was observed in 26% of the patients. This rate is comparable to that reported in previous studies in genotype-1 patients.13–17, 32 As expected, the highest response rate was observed in Group A comprising patients who achieved EVR24. However, only patients with undetectable serum HCV-RNA at weeks 12 (cEVR12) and 24 (cEVR24) obtained SVR. Among the patients who were HCV-RNA positive at week 24 but with a ≥2 log10 drop, two had ETR, but none obtained SVR. On the basis of these results, we can suggest that 48 weeks of combination treatment for genotype-1-transplanted patients can be proposed only for patients showing an undetectable serum HCV-RNA at week 12 (cEVR).

Recently, a longer treatment period has been suggested for immunocompetent patients with a slow virological response.33 Prolongation of treatment has been shown to induce a higher rate of SVR compared with standard treatment, especially in genotype-1 patients. This approach might be indicated also in transplanted patients because in our study ETR patients without a cEVR24 relapsed after treatment cessation. The continuation of treatment could have a rationale in terms of slowing disease progression as has recently been suggested;19, 32 however, for these patients, a careful evaluation of the cost–benefit ratio is recommended especially when considering the high rate of adverse events observed.

None of the patients without EVR24 who continued PEG-IFN plus RBV until the end of the 48 week course (group B) achieved ETR, but one cleared the virus after stopping anti-viral treatment in concomitance with the onset of a ‘probable AIH’. Similarly, two patients of group A achieved SVR despite premature discontinuation because of ‘probable AIH’. As extensively discussed in another paper,24 we believe that this viral clearance is most likely related to the change of the immunological setting of the patients.

When considering SVR in a per-protocol analysis, response rates peaked to 48%; the dropout rate was quite high in our study (45%), although we used growth factors when needed and we tried to continue treatment as long as possible in all patients. This confirms that safety and tolerability are the main factors limiting PEG-IFN plus RBV treatment in transplanted patients.

In our study, PEG-IFN and RBV doses were lower than those recommended for immunocompetent patients. Nonetheless, the safety profile of our schedule was poor with a high incidence of side effects, premature discontinuation and deaths. Furthermore, we also observed an unexpectedly high rate of autoimmune disorders, despite the absence of rejection episodes. The risk of development of ‘de novo’ AIH in HCV liver-transplanted patients during PEG-IFN plus RBV treatment has important clinical implications (two patients died and one had a graft failure) and should be considered in the differential diagnosis together with acute and chronic rejection. Periodic testing of autoantibodies is advisable during treatment.

With regard to cirrhotic patients, it is important to note that most of them developed early decompensation during treatment, and three died while on the waiting list for re-transplantation. Moreover, only one cirrhotic patient achieved SVR, suggesting that treatment at early stages of hepatitis C recurrence is the best strategy in terms of both safety and efficacy. A careful pre-treatment assessment of patients, not only in terms of virological response but also in terms of tolerability, is highly recommended.

In conclusion, our study shows that in spite of the low tolerability profile, treatment with PEG-IFN alfa-2b plus RBV in genotype-1 patients with recurrent post-LT hepatitis C induces SVR in more than one-fourth of cases. Patients with mild to moderate liver disease are the best candidates for treatment. Undetectable serum HCV-RNA at week 12 (cEVR12) is the best predictor of SVR after a standard course of 48 weeks of treatment. In patients with a slow viral response, a longer treatment period should be considered in designing further clinical studies.

Acknowledgements

Declaration of personal interests: None. Declaration of funding interests: F. Lodato received research funding from Roche S.p.A. M. Biselli and S. Lorenzini received research funding from Associazione per la Ricerca sulle Malattie Epatiche (ARME), Bologna, Italy.

Ancillary