Correspondence Dr Giuseppe Tarantino, Cattedra di Gastroenterologia, Piazza delle Cliniche 2, 90127 Palermo, Italy Tel: +39 091 655 2280 Fax: +39 091 655 2156 e-mail: firstname.lastname@example.org
Recently several randomized trials involving exclusively HCV 2 and 3 patients have explored the possibility of reducing the duration of therapy with PEG IFNs and ribavirin to 12–16 weeks. Among these, the largest studies (ACCELERATE, NORTH-C and NORDynamIC) have failed to demonstrate, by intention-to-treat analysis, that short treatment is non-inferior to the standard duration of 24 weeks originated by phase 3 trials. Even though obtaining univocal conclusions from these studies are difficult to obtain due to some critical differences (trial design, genotypes 2/3 ratio, rate of cirrhosis at baseline, ribavirin dose, assays to detect HCV-RNA etc), all have proved that a rapid virological response (HCV-RNA negative at 4 weeks) is the strongest predictor of SVR. Therefore, excluding risk factors for virological relapse at baseline, and identifying in the early phase of treatment, features related to a sustained response, the decision to reduce the duration of treatment to less than 24 weeks in HCV-2 and 3 patients can be response-guided appropriately.
Ongoing studies will assess whether extended 48 week regimens can benefit non-RVR patients with HCV 2 or 3, especially those with more severe fibrosis.
Easy-to-treat genotypes: from phase III trials to studies of optimization
Since 2001, the results of two multicentric randomized-controlled trials (RCT) by Manns et al. (1) and Fried et al. (2) have led to the establishment of 48 weeks of combination therapy with pegylated interferons (PEG-IFN) and ribavirin as the standard of care for chronic hepatitis C (CHC). Given the significantly higher rate of sustained virological response (SVR) obtained in patients with genotype 2 and/or 3 as compared with those with genotype 1 or 4, the 48-week fixed pattern generated by these trials was soon shown to raise a significant issue of potential ‘over-treatment’ of virtually easy-to-treat patients.
In 2004, the study by Hadziyannis et al. (3) provided for the first time sound evidence to optimize treatment duration according to hepatitis C virus (HCV) genotype. In fact, this trial, despite having methodological limitations such as the lack of stratification for genotypes, showed a clearcut difference in SVR between 1 and non-1 genotypes (HCV 2 and 3). For the latter easy-to-treat genotypes, combination therapy with a flat low dose of ribavirin (800 mg/day) for 24 weeks was sufficient to achieve SVR rates comparable to those of more prolonged and dose-intensive schedules.
The two registration trials and subsequently those of optimization are the evidence base on which the American Gastroenterological Association has defined a guideline to drive the strategy for treatment of CHC (4).
More recently, other trials involving exclusively patients with HCV 2 and 3 have further explored the possibility of reducing the duration of therapy to 12–16 weeks. In this review, we shall focus on:
(a)the differences between HCV 2 and 3 in terms of SVR;
(b)an assessment of the major trials exploring strategies for reduction of the duration of therapy;
(c)factors reducing the likelihood of SVR;
(d)treatment of HCV 2 or 3 in patients with advanced cirrhosis; and
(e)the possibility of a response-guided therapy (RGT).
Hepatitis C virus genotypes 2 and 3 respond differently in terms of sustained virological response
In 2004, Zeuzem et al. (5) proved the efficacy and safety of PEG-IFN α-2b (1.5 μg/kg once per week) plus ribavirin (800–1400 mg/day) administered for 24 weeks in patients with HCV 2 or 3. The study enrolled 224 HCV 2 or 3 patients, and showed an overall virological response [intention to treat (ITT)] at the end of therapy in 211 (94%) reaching an SVR in 182 patients (81%). The rates of end-of-therapy response and SVR were higher in HCV 2 patients (100 and 93%) than in those with HCV 3 (93 and 79%). Virological relapse was more likely for HCV 3 (14%) than for HCV 2 (7%). A high viral load (HVL) at baseline (HCV-RNA levels >600 000 IU/ml) was important in determining SVR and relapse rate in HCV 3 patients [70 and 23% vs. 86 and 8% in those with a low viral load (LVL)]. Viral load instead did not have an influence in HCV 2 patients (91 and 9% in HVL vs. 95 and 5% in LVL). Furthermore, a rapid viral response (RVR: undetectable HCV-RNA at 4 weeks of therapy) emerged as a strong predictor of SVR. In fact, the rate of SVR in patients with easy-to-treat genotypes who achieved an RVR was very high (HCV 2: 94%, HCV 3: 85%), thus setting the stage to explore further reductions of treatment.
Therefore, there is sound evidence that it is incorrect to consider patients with HCV 2 and 3 as a single group and the concept of ‘easy to treat’ applied to both these genotypes was an over-simplification.
Treating hepatitis C virus genotypes 2 and 3 with a combo for ;<24 weeks: is it effective?
Shortening the duration of combined antiviral treatment for CHC is a relevant matter in terms of costs and of patient's acceptability. Many subjects infected by HCV 2 and 3 become HCV-RNA negative after 4 weeks of combination therapy and they achieve higher rates of SVR with respect to those with a slower response. Small case series had suggested that SVR could be obtained in patients stopping combo-therapy after only 4–16 weeks. Most of these cases had an LVL and/or HCV 2 or 3.
Several trials, with important differences in the study design, have been devised to assess whether a shorter treatment duration is possible without compromising SVR. Seven studies have been published (Table 1), and others will soon be concluded.
Table 1. Differences in the design of studies of a short treatment for hepatitis virus 2 and 3
The first study that aimed to determine the efficacy of a treatment shorter than 24 weeks for HCV patients with genotypes 2 and 3 was a Norwegian non-randomized trial performed by Dalgard et al. (6) at 20 hospitals enrolling 122 treatment-naïve patients, mostly infected by HCV 3, who received 1.5 μg/kg PEG-IFN α-2b and ribavirin 800–1400 mg/day (weight-based). Patients with an early virological response (EVR) (by Amplicor™, Cobas Amplicor HCV monitor test, version 2.0, Roche Diagnostics, Indianapolis, IN, detection limit 50 IU/ml) at weeks 4 and 8 were treated for 14 weeks, while those without an EVR received 24 weeks of treatment. Altogether, 100 of 122 patients (82%) achieved an SVR. Ninety-five (78%) patients achieved an EVR, and of these, 85 (90%) achieved an SVR. Only 15 of the 27 patients without an EVR (56%), consequently assigned to the 24-week group, achieved an SVR. Among HCV 3 patients who achieved an EVR, SVR was achieved by 98% of the patients with a pretreatment viral load <600 000 IU/ml compared with 72% of those with ≥600 000 IU/ml.
The study of Mangia et al. (7), a randomized trial, involved 14 hospitals. Two hundred and eighty-three patients, the majority infected by HCV 2, were randomized in a 1:3 ratio to receive PEG-IFN α-2b (1.0 μg/kg once per week) plus ribavirin (1000–1200 mg/day) for 24 weeks (control group; patients n=70) or the same therapy for 12 weeks, if they achieved an RVR, or otherwise 24 weeks (study group of variable duration; patients n=213). Overall, RVR was achieved in about 63% of the patients and SVR in 76%. Owing to the low prevalence of patients with genotype 3 (25%) enrolled in the study, only 13 patients with this genotype (10 rapid responders and three non-rapid responders) were allocated to the control group of standard duration. Therefore, this study is not powered to provide evidence about these patients. Regarding HCV 2, 67% of patients without an RVR achieved an SVR. Among patients with an RVR, allocated to the 12-week treatment, 89% achieved an SVR with a relapse rate of 9% compared with the rapid responders treated for 24 weeks, out of which 89% achieved an SVR with a relapse rate of 2%.
The third study focusing on this issue was a German multicentre, randomized trial conducted by Von Wagner et al. (8), involving six tertiary centres and enrolling 153 patients with HCV 2 and 3, 142 of whom (93%) achieved an RVR (defined as an HCV-RNA <600 IU/ml at the fourth week by Amplicor™) and were randomized at a 1:1 ratio to receive PEG-IFN α-2a (180 μg once per week) plus weight-based ribavirin (800–1200 mg/day) for 24 weeks (control group; patients n=70) or for 16 weeks (study group; patients n=72). Before randomization, patients were stratified according to HCV-RNA at baseline as LVL vs. HVL (≤800 000 vs. >800 000 IU/ml). RVR was achieved by 37 of 38 patients (97%) with HCV 2 and by 103 of 112 (92%) with HCV 3. Altogether, SVR was achieved by 119 of 153 patients (78%), and similar rates were observed among those randomized to the 16-week treatment (58 of 71, 82%) or the 24-week treatment (57 of 71, 80%). Only 39% of the patients without an RVR achieved an SVR after being treated for 24 weeks. All patients with HCV 2 and RVR had a 100% rate of SVR, regardless of the length of treatment and the baseline viral load. In fact, 93% of subjects with HCV 2 and HVL achieved an SVR. The SVR rate of HCV 3 patients with LVL was not reduced by the 16-week schedule as compared with 24 weeks, but HCV 3 patients with HVL had a better chance of SVR when treated for 24 weeks (67%) than with 16 weeks of therapy (55%), although because of the small size of the groups, this difference did not reach significance. The main suggestion that emerged from this work is that patients with HCV 2, regardless of baseline viral load, or HCV 3 and LVL who achieve an RVR can safely be treated for 16 weeks without compromising their chances of SVR.
In 2007, the results of a large, multicentre trial (ACCELERATE) were presented by Shiffman et al. (9). This trial, planned with a non-inferiority target of 6% between the two groups, enrolled 1469 patients infected by HCV 2 and 3 and randomized them upfront, regardless of RVR (by Amplicor™, detection limit 50 IU/ml) at 4 weeks, to receive 180 μg PEG-IFN α-2a (once per week) plus ribavirin at a flat dose of 800 mg/day. Overall, reducing the duration of treatment from 24 to 16 weeks both in HCV 2 and in HCV 3 patients reduced the chance of SVR (70 vs. 62%; P<0.001). This difference was still significant among 67% of patients who achieved an RVR (82% for 16 weeks vs. 90% for 24 weeks of treatment; P≤0.005). Among non-RVR patients, the 24-week schedule yielded a higher rate of SVR than that of the 16-week schedule (49 vs. 27%, P<0.001). In the ACCELERATE study, RVR was less frequent in comparison with the previous studies and, most importantly, ribavirin was given at a flat, lower dose. This may be specifically relevant, because a higher rate of SVR has been reported with higher average doses of PEG-IFN or ribavirin per kilogram of bodyweight by Manns et al. (1), while Reddy et al. (10) have shown that as long as patients achieve an RVR, subsequent ribavirin dose reductions have a modest impact on SVR.
Also in 2007, a randomized trial performed in Taiwan by Yu et al. (11) enrolled patients with CHC and HCV 2 (eligible n=150) and randomized them to either 16 or 24 weeks of treatment with PEG-IFN α-2a and weight-based ribavirin at a dose of 1000–1200 mg/day, regardless of RVR. The efficacy of therapy for 16 weeks was comparable with the standard 24 weeks of treatment in patients who achieved an RVR at 4 weeks (SVR rates: 94 vs. 95% respectively). Patients with an RVR had significantly higher SVR rates than those without an RVR in both treatment groups. The authors reported a lower incidence of adverse events such as alopecia and a lower cost of treatment for the short regimen. This study established a relationship between ribavirin dose and ribavirin adherence, with a higher relapse rate in patients without an RVR. Interestingly, the SVR rate of this study was higher than that observed in the previous studies, particularly with respect to the ACCELERATE trial. Several factors such as ethnic differences, mean body weight (significantly lower in the Taiwanese population) and mean ribavirin dose (15.3 mg/kg/day in this Taiwanese trial vs. 9.52 mg/kg/day in the ACCELERATE), besides possible geographical variability of HCV strains, could account for this discrepancy.
In 2008, another large RCT by Dalgard et al. (12), the NORTH-C, was published. This study was designed with a non-inferiority margin set at 10% between the two groups, and enrolled a total of 428 treatment-naïve patients with HCV 2 and 3 (80%), all receiving the same schedules reported in the original pilot study by the same group (6). Patients with an RVR, defined as HCV-RNA levels <50 IU/ml (Amplicor™) at 4 weeks of treatment, were randomized to 14 or 24 weeks of treatment while those without an RVR were treated for 24 weeks. Three hundred and two patients out of 428 (71%) achieved an RVR. The NORTH-C showed that 14 weeks were inferior to 24 weeks to obtain an SVR in the ITT population, the lower rate of SVR (81 vs. 91%) being mostly owing to a higher rate of relapse (19 vs. 9%). A ‘per protocol’ analysis reduced the difference (90 vs. 94%) slightly, leading the authors to support the short schedule because of cost savings and fewer side effects.
The latest large-scale multicentre RCT fully reported (13) is the NORDynamIC, a study performed in four Scandinavian countries enrolling 382 patients with HCV 2 and 3 (72%) for treatment with PEG-IFN α-2a and a flat dose of ribavirin (800 mg/day) for 12 (n=148) or 24 (n=150) weeks. HCV-RNA was tested by AmpliPrep™ TagMan, detection limit 15 IU/ml. Twelve weeks of therapy was inferior to 24 weeks in the ITT population (SVR: 59 vs. 78%, P<0.0001), regardless of the HCV genotype (HCV 2: 56 vs. 82%, P=0.006: HCV 3: 58 vs. 78%, P=0.0015) and the fibrosis stage.
It is difficult to obtain univocal conclusions from all analysed studies because of some critical differences (Table 1). Among them, the most relevant are the population enrolled, and more specifically, the ratio between HCV 2 and 3 and the rate of cirrhosis at baseline; the study design (upfront randomization or randomization after an RVR); and the drugs tested, in particular, the dosage of ribavirin. Also, the assays used to detect HCV-RNA and the definition of an RVR were somewhat different. Three large trials (ACCELERATE, NORTH-C and NORDynamIC) have actually failed to demonstrate by ITT analysis that a short treatment with PEG-IFN and ribavirin is not inferior to the standard duration of 24 weeks. Nonetheless, in all the studies, the SVR rates of patients who achieved an RVR were also quite high for the short treatment groups, and various studies have reported anecdotally that patients who relapsed after a short treatment achieved an SVR when retreated for a longer period. In a recent report from an ongoing trial (Al-liver group) (14), 43 out of 67 HCV 2 or 3 patients who relapsed after receiving 12 weeks of PEG-IFN α-2b (1.5 μg/kg once per week) plus ribavirin (1000–1200 mg/day) were re-treated with a 24-week course: 31 (72%) achieved an SVR, 2 (5%) relapsed again and 10 (23%) showed no response. A short treatment, even if it is marginally less effective overall, would reduce the costs for both the patient and the society. As Dalgard et al. (12) argued, a treatment for 14 weeks instead of 24 weeks led to a 26% reduction in drug exposure and a reduction in costs, also considering retreatment of 10% of the patients who relapsed after achieving an end-of-treatment response. The duration of side effects was related to the duration of therapy; therefore, fewer side effects are expected for a shorter period.
Overall, all seven trials provided concordant evidence that:
•Patients with genotype 2 who achieve an RVR, given the high rate of SVR (90–100%), can be treated for <24 weeks. The exact duration of short therapy (12, 14 or 16 weeks) remains debatable.
•Patients with HCV 3 and an HVL are the most difficult to treat and may be in need of a prolonged therapy (24 or even 48 weeks). However, if they achieve an RVR, a short treatment is effective, because the gain by prolonging therapy up to 24 weeks is modest, if any.
•Patients without an RVR, both HCV 2 and 3, are not candidates for a short treatment, and even a 24-week course of treatment may yield unsatisfactory SVR rates (50–70%).
Studies are needed to assess whether the latter difficult-to-treat subgroup may benefit from prolongation of treatment beyond 24 weeks. Interestingly, a retrospective analysis of a pooled database of non-RVR HCV 2 and 3 patients from Fried's and Hadziyannis's studies (15), who had received both a flat and a semi-weight adjusted ribavirin dose and 24 or 48 weeks of therapy, suggest (Table 2) that the best results in these subjects are obtained by combining a higher dose of ribavirin and a more prolonged treatment period.
Table 2. Retrospective analysis of treatment intensification in patients with hepatitis C virus genotype 2 or 3 (15)
However, these last assertions about the optimized ribavirin dose are in contrast with the results of a recent Austrian randomized trial by Ferenci et al. (16). This study has shown that 400 mg/day of ribavirin, given to HCV 3 patients in combination with a standard dose of PEG-IFN α-2a for 24 weeks, yielded essentially the same results in terms of SVR and relapse rates at a flat dose of 800 mg. The reasons for the discordance are unclear, and may lie in patient selection or in the absence of disease cofactors in this study population.
Factors reducing the likelihood of sustained virological response
Practically all predictors of SVR known to be effective for standard treatments have been re-validated for short therapies. In the original Dalgard study (6), a biopsy showing no or minimal fibrosis strongly predicted SVR: in fact, all the patients who relapsed after 14 weeks of treatment had bridging fibrosis/cirrhosis or an HVL at baseline. Zeuzem's study (5) confirmed that steatosis, a feature strongly related to high HCV-RNA levels and HCV 3, is also a strong predictor of non-response in these easy-to-treat patients. In Yu's study (11), enrolling only patients with HCV 2, achievement of RVR and patient's age were independent factors associated with SVR. In Dalgard's study (12), age ≤40 years, male gender and pretreatment viral load ≤400 000 IU/ml emerged as independent predictors of RVR. Finally, the NORDynamIC study (13) found that age and HCV-RNA levels on days 7 and 29 were independent predictors of SVR, and that a short treatment was useful in patients <40 years old, especially if HCV-RNA was undetectable on day 29, and also in patients ≥40 years old, provided that HCV-RNA was below 1000 IU/ml on day 7 in addition to being undetectable on day 29. If neither of these two criteria were met for patients ≥40 years old, 24 weeks of therapy were superior (P<0.0001). In a sub-analysis of 355 patients from this study (17), it was shown that the plasma ribavirin concentration at day 29 is a strong independent predictor of SVR. Unfortunately, ribavirin concentrations could not be predicted for an individual patient from the baseline parameters with sufficient precision. These data, however, support the concept that giving ribavirin at a flat dose of 800 mg/day, as done in many of these studies, may have caused under-dosing and consequent loss of SVR among the heavier patients.
Overall, these studies all concur that factors such as more advanced age, steatosis and severe fibrosis or cirrhosis affect the SVR rate. Thus, the identification of negative predictors of SVR should lead to preference of a standard or even a prolonged course of therapy.
Recently, Mangia et al. (14) presented data from an ongoing trial (Al-liver group) involving 11 Italian centres, which has enrolled up to now 718 patients with HCV 2 and 3 CHC, aiming to identify pretreatment predictors of relapse in RVR patients after a short course of therapy (12 weeks of PEG-IFN α-2b 1.5 μg/kg plus ribavirin 1000–1200 mg/day). Patients with an RVR were treated for 12 weeks; conversely, those without an RVR were treated for 24 weeks. In fact, 496 patients (69%) achieved an RVR, and after a 12-week therapy, 67 patients (14%) relapsed. There were no statistical differences in the relapse rate between genotypes 2 and 3. On performing multivariate analysis, factors associated with relapse were platelet count <140 000/mm3 [odds ratio (OR) 3.7] and body mass index (BMI) >30 kg/m2 (OR 3.0). On analysing the characteristics of patients with no SVR, at least one of the following unfavourable factors was present: cirrhosis, age >50 years or BMI>30 kg/m2. Thus, a short treatment cannot be recommended in older HCV 2 and 3 patients with BMI>30 kg/m2 and features of advanced liver disease.
Treatment of hepatitis C virus genotype 2 or 3 in patients with advanced cirrhosis
Treatment of patients with cirrhosis, although desirable because of the chance of improving the course of liver disease and reducing liver-related deaths after SVR, is often not attempted because of the disadvantage of the low tolerability of IFN-based regimens in subjects with advanced liver disease and the modest overall efficacy in the presence of advanced fibrosis. In patients with easy-to-eradicate HCV genotypes and cirrhosis, the expected rate of SVR is anyway higher, with 70% of HCV 2 and 46% of HCV 3 patients achieving viral clearance as compared with 28% HCV 1 in a post hoc analysis of the Hadzyiannis study (18). In these patients with cirrhosis, short regimens would hypothetically be better tolerated and more feasible.
A post hoc analysis of the ACCELERATE data (19) found that the presence of bridging fibrosis or cirrhosis (Metavir F3 and F4) was equally predictive of less SVR for the standard 24-week treatment (76% of 629 <F3 vs. 60% of 189 ≥F3, P=0.0001) and for the short 16-week treatment (67% of 538 <F3 vs. 48% of 191 ≥F3, P=0.0001). It must be stressed, however, that at least two-thirds of these patients had precirrhotic F3 fibrosis, and thus the average results are not truly representative of advanced liver disease.
In an RCT in patients with HCV cirrhosis and portal hypertension by our group (20), PEG-IFN α-2b (1.0 μg/kg/week) with ribavirin (800 mg/day) for 52 weeks led to an SVR in 83% of HCV 2/3, a significantly higher rate than that obtained in HCV-1 patients (20%). Moreover, all HCV 2/3 patients achieved an RVR at 4 weeks of therapy. An earlier study in patients with advanced fibrosis or cirrhosis treated with PEG-IFN monotherapy had also shown a significant difference in SVR rates between HCV 2/3 patients (51%) and those with HCV 1 (12%) (21).
Unpublished data from our group (Fig. 1) suggest that in this setting, RVR cannot be considered as a reliable positive predictor of SVR. In fact, 25% of HCV 2/3 patients with cirrhosis achieved an EVR although didn't achieve a RVR at 4 weeks, and half of them went on to achieve an SVR up to 52 weeks of treatment.
Taken together, these findings suggest that HCV 2 or 3 patients with cirrhosis should be treated regardless of the stage of liver disease, owing to the reasonable chance of SVR. No data are available that address specifically the issue of short treatments, but the higher efficacy of the 24-week regimen in the ACCELERATE study and the high rates of SVR obtained in our experience point towards the need for longer treatment regimens.
Optimizing treatment duration with a response-guided therapy
Growing evidence supports the importance of RVR to individualize the treatment for CHC by optimizing the exposure to PEG-IFN and to ribavirin. Recent studies (22–24) in fact have confirmed that 24 weeks of treatment are effective and sufficient to achieve an SVR in patients with HCV 1 or 4 who achieve an RVR at 4 weeks. This strategy is indicated as ‘RGT’, and requires a precise definition of on-treatment responses. Thus, an RVR is defined by HCV-RNA negativity (<50 IU/ml) at week 4 of therapy. An EVR can either be complete EVR when HCV-RNA is positive at week 4 but negative at week 12, and partial or slow EVR when HCV-RNA is positive at weeks 4 and 12, but displays a ≥2 log10 drop from baseline at week 12. Lastly, non-EVR is defined by a <2 log10 drop of HCV-RNA from baseline at week 12.
In a post hoc analysis of the data from the ACCELERATE study (25), RVR emerged as the strongest predictor for SVR (Fig. 2), while baseline viral load, at a cut-off of 800 000 IU/ml, was not significant. In fact, patients with RVR and LVL achieved an SVR in 94% of cases, while those with RVR and HVL achieved an SVR in 88% of cases. Only 49% of non-RVR subjects achieved an SVR, accounting ultimately for most treatment failures in the study.
Consequently, the occurrence of an RVR should be considered the main decisive factor for RGT in patients with CHC owing to HCV 2 or 3. Ongoing studies such as the NCORE 2/3, a trial of optimization with PEG-IFN α-2a and ribavirin, will indicate whether extended 48-week regimens can benefit non-RVR patients with HCV 2 or 3.
Reducing the duration of treatment to <24 weeks in patients with CHC owing to HCV 2 or 3 should be a balanced decision taking into account:
(a)the exclusion at baseline of factors that are likely determinants of relapse or non-response;
(b)the early identification during treatment of features determining a high likelihood of SVR;
(c)the likelihood of responding to retreatment in case of failure to eradicate HCV during the first cycle of therapy; and
(d)the patient's preference and attitudes towards treatment patterns.
Bearing all the above in mind, one may suppose that a patient with CHC owing to HCV 2 or 3 (LVL only), without an advanced stage of fibrosis, below 50 years, with a BMI<30 and no metabolic syndrome and with a projected good adherence to the schedule, who has an RVR at 4 weeks could be more difficult to find in real-life practice than in clinical trials.
Conflicts of interest
The authors have not declared any conflicts of interest.