Sustained virological responses following standard anti-viral therapy in decompensated HCV-infected cirrhotic patients

Authors


Dr A. Andriulli, Division of Gastroenterology, “Casa Sollievo della Sofferenza” Hospital, IRCCS, viale Cappuccini 1, 71013 San Giovanni Rotondo, Italy.
E-mail: a.andriulli@operapadrepio.it

Summary

Background  Little data is available about predictors of sustained virological response (SVR) during anti-viral therapy of patients with decompensated HCV cirrhosis.

Aims  To determine whether rapid and early virological responses (RVR and EVR) could predict SVR and help optimize treatment in these patients.

Methods  A total of 94 cirrhotics underwent treatment with peg-interferon alfa-2b (1.5 μg/kg weekly) and ribavirin (800/1200 mg daily) for 48 or 24 weeks for genotypes 1/4 or genotypes 2/3, respectively.

Results  Overall, SVR was achieved in 33 patients (35.1%), 16% with genotype 1/4 and 56.8% with genotype 2/3 (P < 0.01). At treatment week 4, 34 patients had undetectable HCV-RNA, 10 with genotype 1/4 and 24 with genotype 2/3. Of RVR patients, 24 achieved SVR (70.5%), 6 and 18 with genotypes 1 and non-1. At the multivariate analysis, only EVR, genotypes 2 and 3, and adherence to full course and dosage of therapy retained their independent predictive power, with corresponding ORs of 25.5 (95% CI 3.0–217.3), 4.2 (95% CI 1.2–15.3) and 9.1 (95% CI 2.2–38.0), respectively.

Conclusion  In decompensated cirrhotic patients, anti-viral therapy with current regimens is feasible and associated with an overall SVR rate of 35.1%. Treatment ought to be pursued among patients who attain an EVR, and maintain a full course and dosage of therapy.

Introduction

In HCV-related cirrhotic patients with hepatic decompensation, safety and efficacy of anti-viral treatment are currently debated, with most concerns being related to the likelihood of attaining sustained virological response (SVR) whilst mitigating the increased risk of treatment-related adverse events.1, 2

Among patients with bridging fibrosis or compensated cirrhosis, use of pegylated interferon alpha (Peg-IFN) and ribavirin (RBV) has been reported with SVR rates of 13–41% for genotype 1 and 73–83% for genotypes 2 and 3.3–6 These figures are distinctly higher than the 0–30% overall eradication rate reported in cirrhotic patients with signs of liver decompensation.1, 2, 7–10 We have validated previously the feasibility of Peg-IFN based therapies for cirrhotics in Child-Pugh-Turcotte (CPT) classes A and B, who had experienced an episode of hepatic decompensation and have also provided data on the positive impact of therapy in improving hepatic function and survival in those who attained an SVR.1 Low tolerability of therapy, the increased risk of infectious events during treatment and the greater toxicity of treatment in such patients make it important to identify cirrhotic patients who possess a higher likelihood of achieving an SVR.

In non-cirrhotic patients, monitoring of viral kinetics during the initial months of therapy has been shown to predict SVR with high sensitivity, specificity, positive and negative predictive values.11–14 In genotype 1 patients and rapid virological response at treatment week 4 (RVR), SVR rates were considerably higher than in those still viremic at this time evaluation (89% vs. 19%, respectively). In addition, patients with a pre-treatment HCV RNA < 600.000 IU/mL were more likely to achieve an RVR than those with higher viremia levels.15, 16 In RVR patients with HCV genotype 2 and 3 infection, 24 weeks of treatment resulted in >87% SVR rates in both cirrhotics and non-cirrhotic patients.16–18 Whether such a predictor of SVR would be of help in the setting of cirrhotic patients undergoing anti-viral therapy after a decompensation episode has not been investigated yet. Our aim was to verify whether early monitoring of viral kinetics in this subset of patients might predict positively or negatively an SVR and facilitate development of rational and effective recommendations to discontinue prematurely medications in nonresponders, and keep on treatment in responders .

Patients and methods

Patient selection

After disclosure in our previous work1 of the lack of safety of anti-viral treatment for CPT class C patients, we are currently enrolling decompensated patients in CPT classes A and B (with MELD ≤ 14) in an ongoing programme of therapy with Peg-IFN alfa-2b and ribavirin. The CPT class and MELD score were determined at the time of patient identification and confirmed at study entry. The present investigation refers to a new series of 94 cirrhotics included in this programme since May 2003.

All HCV-infected cirrhotics, admitted to the wards of our Institutions for a decompensated event, are offered anti-viral therapy, once the decompensated event is over. Treatment commonly starts 1 month after patients’ discharge from hospital. Patients are not eligible for the program if already treated with combination therapy of conventional interferon or Peg-IFN and ribavirin and if they have any of the following exclusion criteria: hepatitis B surface antigen positive, infection with the immunodeficiency virus, history of psychosis, rapid deterioration of liver and/renal function (such as deepening jaundice or hepatorenal syndrome), detection of hepatocellular carcinoma, current alcohol abuse, invalidating disease (severe chronic obstructive lung disease, history of cardiovascular disease and abnormal thyroid-stimulating hormone level), bacterial infections, platelet count less than 35.000 cells/μL, absolute neutrophil count less than 1000 cells/μL, haemoglobin level less than 10 g/dL, total bilirubin levels greater than 3 mg/dL and serum creatinine levels greater than 2.0 mg/dL.

Treatment schedule

Consenting genotype 2 and 3 patients received 1.5 μg/kg body weight of Peg-IFN alfa-2b (PEG-Intron, Schering Plough, Kenilworth, United States) weekly for 24 weeks in combination with oral ribavirin (Rebetol, Schering Plough) at a daily dosage of 800 or 1000 mg for patients weighing < or ≥75 kg. Genotypes 1 and 4 patients received the same dosage of Peg-IFN alfa-2b for 48 weeks in combination with daily dosage of 1000 or 1200 mg of ribavirin for patients’ body weight < or ≥75 kg. The study was approved by the Ethics Committee and conducted according to guidelines of the International Conference on Harmonization for Good Clinical Practice. The study was designed, conducted, and results analysed without financial support from producers of drugs used in the study.

Virological and safety evaluation

All patients were evaluated as out-patients for safety, tolerance and efficacy. On-treatment HCV-RNA levels were determined by qualitative PCR assay (Amplicor HCV test, Roche, Basel, Switzerland) at weeks 4 and 12, at the end of treatment and at week 24 of follow-up. The HCV-RNA level was quantified by PCR assay (Cobas Monitor 2.0, Roche) and basal HCV genotyping performed by reverse hybridization (Innolipa HCV, Innogenetics, Ghent, Belgium). Whenever neutrophil counts dropped <750 cells/μL or haemoglobin decreased >3 g/dL from baseline levels, granulocyte colony-stimulating factor (Myelostim 34, 10 mcg/kg, Italfarmaco, Italy) or erythropoietin (Eprex 40.000 units, Janssen-Cilag) was administered once weekly and continued until the end of therapy. Patients whose haematological parameters did not rise after 4 weeks of growth factors administration had to reduce stepwise dosages of Peg-IFN and/or ribavirin. Anti-viral therapy was discontinued in nonresponders (haemoglobin < 8.0 g/dL and neutrophil < 500 cells/μL) and blood transfused in the event of haemoglobin levels <8.5 g/dL.

Assessment

The primary efficacy end point was attaining an SVR, defined as undetectable serum HCV-RNA (<50 IU/mL) by qualitative PCR assay at the end of a 24-week follow-up. A further end point was to determine whether a rapid or early virological response (RVR or EVR, defined as undetectable HCV-RNA at weeks 4 or 12 after initiation of treatment) could be identified as a predictor of SVR.

Treatment goals were categorized as follows: full course, as patients who received >80% of the drugs for more than 80% of the standard duration of treatment; discontinuation, as withdrawal of therapy at any time before reaching the 80% expected period of treatment; low-adherent patients, as patients who reduced by more than 20% the planned dosages of anti-viral drugs. Adverse events were graded as mild (not requiring dose modification), moderate (requiring dose modification) or severe (requiring discontinuation of therapy) according to recommendations of the World Health Organization.

Statistical analysis

Data were analysed at the conclusion of the follow-up according to the intention-to-treat principle. Continuous variables were expressed as mean (±s.d.) and converted to dichotomous variables by identifying the best cut-off threshold on the basis of clinical standards. Categorical variables were reported as percentage values and compared using the chi-square test or Fisher’s exact test as needed. All P values were two-tailed. An SVR prediction model was developed including age ≤60 years, gender, genotype (1 and 4 vs. 2 and 3), basal viral load (≤600 000 vs. >600 000 IU/mL), HCV clearance at week 4 and 12, CPT points (6 and 7 vs. 8 and 9), MELD score (≤10 vs. >10), haemoglobin levels (≤12 vs. >12 g/dL), baseline neutrophil counts (≤1500 vs. >1500 cells/μL), platelet counts (≤75 000 vs. >75 000 cells/μL) and the number of patients who received full dosage and length of either peg-interferon and ribavirin vs. that of patients with low adherence or who discontinued treatment. Stepwise logistic regression analysis was performed to study the effect of prognostic factors either RVR or SVR. All considered variables were included in the model and a forward entry procedure was then applied. Positive and negative predictive values of clearing HCV at treatment week 4 (RVR) or at treatment week 12 (EVR) for predicting SVR were also calculated. Positive predictive value (PPV) was defined as the probability that the outcome of interest (SVR) occurred in patients fulfilling the prediction criterion of interest (RVR or EVR). Conversely, negative predictive value (NPV) was defined as the probability that the outcome of interest did not occur (no SVR) in patients not fulfilling the prediction criterion of interest. All the analyses were performed with Statistical Package for Social Science software (SPSS version 13, Inc., Chicago, IL, USA).

Results

Baseline characteristics

Starting from May 2003, 94 HCV-infected Caucasian patients entered the program and underwent therapy with Peg-IFN and ribavirin 1 month apart from a successful treatment of the decompensated event. Most patients were naïve to therapy (29% received prior interferon), infected by HCV genotype 1 (51.1%), had pre-treatment viral load ≤600.000 UI/mL (57.4%), scored CPT point ≥8 (41.6%), presented with ascites (72% at their first event) and 40% of patients had a MELD score ≥10. No patient had evidence of active decompensation at the beginning of anti-viral therapy. The flow-chart of patients throughout the study is given in Figure 1.

Figure 1.

 Viral clearance on and off therapy in 94 decompensated cirrhotics who underwent peginterferon and ribavirin therapy for 24 weeks (genotype 2 and 3) or 48 weeks (genotypes 1 and 4).

On-treatment viral clearance

Within the first 4 weeks of therapy, five patients dropped out and eight more patients reduced drug dosages: none of them achieved an SVR. At treatment week 4, 34 patients cleared the virus from the blood. Data of baseline patients and their virological features that might enable early identification of RVR are shown in Table 1. At the univariate analysis, age, gender, CPT points, MELD score, and counts of serum platelets, neutrophils and white blood cells did not differ between patients with and those without RVR. The only baseline predictors of RVR were a low pre-treatment viral load and genotypes 2 and 3. Overall, 54.5% of genotype 2 and 3 patients, but only 20% of those with genotypes 1 and 4 attained an RVR (P < 0.001).

Table 1.   Univariate analysis of baseline patient and virological features as predictors of Rapid Virological Response (RVR)
Baseline parametersRVR n = 34No RVR n = 60 P value
n % n %
Age (years)
 >601955.884371.670.12
 ≤601544.121728.33
Gender
 Male2058.823050.000.41
 Female1441.183050.00
Baseline Viremia (IU/mL)
 >600 000720.593355.000.001
 ≤600 0002779.412745.00
Genotype
 1 and 41029.414066.67<0.001
 2 and 32470.592033.33
Child-Pugh-Turcotte
 Points: 8 and 91647.062338.330.40
 6 and 71852.943761.67
MELD score
 Points: 10 to 141441.182338.330.79
 7 to 92058.823761.67
Platelets (cells/μL)
 >75000 2264.713660.000.65
 ≤750001235.292440.00
Neutrophils (cells/μL)
 >15002985.294676.670.31
 ≤1500514.711423.33
WBC (cells/μL)
 >30002882.354575.000.41
 ≤3000617.651525.00
Haemoglobin (g/dL)
 >121955.884066.670.29
 ≤121544.122033.33
Full dose administration (within week 4)
 Yes3294.124981.670.124
 No (reduced/  discontinued)25.881118.33

The RVR status was attained in 79.4% and in 20.6% of low and high viremic patients at baseline (P = 0.001). In patients with genotypes 2 and 3 infection, the likelihood of achieving an RVR was higher in patients with low viremia than in those with high viremia: 20 of 32 (62.5%) vs. 4 of 12 (33.3%). Corresponding figure for genotypes 1 and 4 were: 6 of 22 (27.3%) vs. 4 of 28 (14.3%). At the stepwise logistic regression analysis, both genotypes 2 and 3 and low viremia retained an independent prediction power for RVR, with respective odds ratios of 3.75 (CI 1.44–9.71) and 3.58 (CI 1.29–9.48).

By treatment week 12, 20 more patients tested HCV-RNA negative, seven with genotypes 1 and 4 and 13 with genotypes 2 and 3 (Figure 1). Two RVR patients dropped out at treatment weeks 10 and 11 respectively, but were considered as EVR being HCV-RNA negative at the end of follow-up off therapy (Figure 1). By treatment week 24, one more genotype 1 patient tested negative. Overall, 51 patients (54.2%) cleared the virus by the end-of-treatment, 16 of them (31.3%) harbouring genotype 1 and 4 infections and the remaining patients with genotype non-1 infection (68.7%) (P = 0.02).

Off-therapy viral clearance

At the end of follow-up, 18 of 51 patients (43.1%) with end-of-treatment response relapsed, 10 of 35 patients (28.5%) with genotypes 2 and 3 and 8 of 16 patients (50.0%) with genotype 1 and 4; the difference was significant (P = 0.01). Overall, 33 of 94 decompensated cirrhotics (35.1%) attained an SVR, 25 patients with genotypes 2 and 3 and eight with genotypes 1 and 4. At the univariate analysis, SVR was more commonly predicted by genotypes non-1, low viremia at baseline, attaining an RVR and EVR status, and by being capable of tolerating full course and dosage of each of the two anti-viral drugs (Table 2). Other baseline features of patients, specifically either the CPT or the MELD scores, were not predictors of an SVR. At the stepwise logistic regression analysis, only EVR, genotypes 2 and 3, and adherence to full course and dosage of the two anti-viral drugs retained their independent predictive power, with corresponding ORs of 25.5 (95% CI 3.0–217.3), 4.2 (95% CI 1.2–15.3) and 9.1 (95% CI 2.2–38.0) respectively. Achieving an RVR had a PPV of 70.6% (24 of 34 patients) for an SVR and an NPV of 85.0% (51 of 60 patients). Corresponding figures for EVR were 59.3% (32 of 54 patients) and 97.5% (39 of 40 patients). A single patient with genotype 2 and low pre-treatment viral load achieved RVR and SVR after being treated for only 45 days when therapy was discontinued due to profound platelet reduction.

Table 2.   Univariate analysis of prognostic factors for sustained virological response (SVR) outcome
Baseline parametersSVR n = 33No SVR n = 61 P value
n % n %
Age (years)
 >601854.554472.130.08
 ≤601545.451727.87
Gender
 Male2163.642947.540.13
 Female1236.363252.46
RVR
 RVR2472.731016.39<0.01
 No RVR927.275183.61
EVR
 EVR3296.92236.0<0.01
 No EVR13.13964.0
Baseline Viremia (IU/mL)
 >600 000618.183455.74<0.01
 ≤600 0002781.822744.26
Genotype
 1 and 4824.244268.85<0.01
 2 and 32575.761931.15
Child-Pugh-Turcotte
 8 and 91339.392642.620.76
 6 and 72060.613557.38
MELD score
 Points: 10 to 141236.362540.980.66
 7 to 92163.643659.02
Platelets (cells/μL)
 >750002266.673659.020.46
 ≤750001133.332540.98
Neutrophils (cells/μL)
 >15002678.794980.330.85
 ≤1500721.211219.67
WBC (cells/μL)
 >30002472.734980.330.39
 ≤3000927.271219.67
Haemoglobin (g/dL)
 >121751.524268.850.09
 ≤121648.481931.15
Full dose administration
 Yes2987.882744.26<0.001
 No (reduced/  discontinued)412.123455.74

Of the 94 cirrhotics enrolled in the present investigation, only nine patients were listed for liver transplantation. An SVR was achieved in four patients, three of them with genotype 2; none of them was transplanted 6 months off therapy, whereas a single patient developed a hepatocarcinoma during the follow-up and was successfully transplanted within 3 months from diagnosis.

Adherence

Overall, 56 of 94 (59.6%) patients tolerated full dosage and duration of treatment, whilst 18 (19.1%) patients discontinued treatment earlier for lower haematological counts [platelet count < 20 000 cells/μL and/or neutrophils < 500 cells/μL (n = 6), haemoglobin < 8.0 g/dL (n = 3)], intolerance and depression (n = 5), severe infections (n = 2), haematemesis (n = 1) and hepatocellular carcinoma (n = 1).

Those patients with severe infections were administered antibiotics with quick resolution of the infective event. All treatment withdrawal occurred during the initial 24 weeks of therapy and most of the infective events were in patients with CPT score ≥8.

Low adherent patients who required dose reduction of Peg-IFN and/or ribavirin were 20 out of 94 (22.3%). In these patients, dosages of both drugs were reduced for neutrophils <750 cells/μL (n = 6), haemoglobin <10 g/dL (n = 5), concurrent decrease of neutrophils and haemoglobin (n = 4), moderate infection (n = 3) and flu-like symptoms and depression (n = 2). Erythropoietin and granulocyte growth factors were administered respectively in eight and five individuals. All treatment timing and dose reduction occurred during the initial 20 weeks of therapy. Only three among low adherent patients achieved SVR. Although a trend to better tolerance of therapy was observed in patients with less advanced disease, no statistically significant correlation was found in a comparison of adherence and severity of liver disease by different CPT scores. During treatment and according to the time of the first event, 32 decompensated events were reported.

Discussion

In HCV-infected patients with a decompensated cirrhosis who received anti-viral therapy, achieving an SVR has been correlated with improved liver function, as apparent from significant reductions in CPT and MELD scores after treatment.1 Major benefits after HCV eradication and partial recovery of liver metabolic activity were no allograft failure secondary to recurrence of viral infection8 and the potential removal of those patients who cleared HCV-RNA from the liver transplant waiting list. These benefits have to be counterbalanced by the increased risk of severe infections, particularly relevant in CPT class C patients, and by the low tolerability of anti-viral treatment in advanced disease.

Several reports have disclosed the feasibility of anti-viral therapy with Peg-IFN and ribavirin for HCV-infected cirrhotics who had experienced an event of hepatic decompensation that could be managed successfully by appropriate therapy,1, 2, 7–10 but the reported SVR rates were extremely discouraging with values as low as 0%10 and as high as 20%.1, 7–9

By virtue of using neither standard dosages of each of the two anti-viral drugs nor standard treatment duration, the suboptimal treatment schedules adopted in previous trials may have been responsible for low virological response rates. Indeed, the attempt to prevent recurrence of hepatitis C after liver transplantation was managed by treating patients during the waiting time with shorter treatment duration (median duration of 12 weeks),8 daily administration of high dose of interferon-α2b in monotherapy9 and low dose of interferon with or without low dose of ribavirin.10 Also in our previous study, the selection of a short treatment regimen (24 weeks for all genotypes) with a low dose (1.0 μg/kg) of Peg-IFN-α2b plus standard dose of ribavirin might have underestimated the expected outcome.1

In the present report, we used standard dosages and duration of anti-viral drugs for treating cirrhotic patients in CPT classes A and B, who had recovered from a previous episode of hepatic decompensation. By using the recommended schedule of therapy, we were able to achieve an SVR in 33 of 94 cirrhotic patients (35.1%) who agreed to undergo treatment. In addition, in patients infected with favourable HCV genotypes, the SVR attained a rate of 56.8% after 24 weeks of therapy, a value that could be considerable and promising for cirrhotics with genotypes 2 and 3, but that is substantially lower than the 80–90% figures registered in those with a less advanced liver disease.16–18 We acknowledge the present experience being heavily weighted towards favourable genotypes and that the overall SVR rate might not be applicable to clinics treating HCV-infected cirrhotics with a higher predominance of genotype 1. Indeed, with the standard schedule of treatment lasting 48 weeks, only 20 of 50 cirrhotics infected with genotypes 1 and 4 will have an on-treatment response, with most of responders (60%) relapsing when therapy had to be withdrawn and only a discouraging minority (16%) experiencing an SVR.

With the previously highlighted deficiencies of current schedule of anti-viral therapy for the cirrhotic population infected with HCV, it seems relevant to delineate baseline features that might be able to select those most likely to clear the virus. To shed some lights on this problem, we have shown that the most robust predictors of SVR were attaining the EVR, genotypes 2 and 3 and adherence to full course and dosage of therapy with corresponding ORs of 25.5 (95% CI 3.0–217.3), 4.2 (95% CI 1.2–15.3) and 9.1 (95% CI 2.2–38.0), respectively.

These findings could serve as relevant criteria to assist clinical counselling by considering that the strongest positive predictive factor for an SVR was attaining the EVR, all patients who clear the virus at treatment week 12 should be encouraged to keep on therapy for the assigned treatment duration. On the contrary, failing to clear the virus by treatment weeks 12 was highly predictive of treatment failure (NPV of 85.0% and 97.6% respectively) and would suggest stopping treatment. Although the treatment was limited to those patients with a Child score ≤9, more than 40% experienced withdrawals or low-adherence to anti-viral therapy. The need for very close clinical surveillance of these patients by experienced physicians cannot be overemphasized. Proper assessment and management of infective events and ensuing liver decompensation is undoubtedly life-saving.

In conclusion, the present investigation on decompensated cirrhotic patients has shown the feasibility of anti-viral therapy with the same schedule that is currently recommended for compensated cirrhotics, with an overall SVR rate of 35.1% in the whole population, of 56.8% in the subgroup infected with genotypes 2 and 3, and of 16.0% in those infected with genotypes 1 and 4.

Treatment appears worth pursuing only in the setting of patients with good prognostic characteristics, such as low viral load and EVR and in those adhering to full dosages of either peg-interferon and ribavirin.

Acknowledgement

Declaration of personal and funding interests: None.

Ancillary