July 3, 2013

Selection of Resistant–Associated Variants to the NS5A Inhibitor Daclatasvir: Revenge of the Hepatitis C Virus

Volume 145, Issue 1 , Pages 247-249, July 2013

Alessio Aghemo, Massimo Colombo

First Division of Gastroenterology, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Università degli Studi di Milano, Milan, Italy

published online 31 May 2013.

Philip S. Schoenfeld, Section Editor, John Y. Kao, Associate Section

Full Text

Karino Y, Toyota J, Ikeda K, et al. Characterization of virologic escape in hepatitis C virus genotype 1b patients treated with the direct-acting antivirals daclatasvir and asunaprevir. J Hepatol 2012 Nov 22 [Epub ahead of print].

Selection of resistant hepatitis C virus (HCV) strains to directly acting antivirals (DAA) was among the main concerns during the initial drug development process for telaprevir (TVR) and boceprevir (BOC; J Hepatol 2012;56 Suppl 1:S88–100). In phase I and II studies both first-generation NS3 protease inhibitors (PI) were associated with rapid development of HCV resistance in monotherapy studies, as well as in ribavirin (Rbv)-free treatment regimens (N Engl J Med 2009;360:1827–1838; Lancet 2010;376:705–716). These alarming figures were, however, rapidly counterbalanced by some key concepts that were learned from phase II and III studies. First, we understood that not only pegylated interferon (PegIFN) and Rbv were essential components of first-generation PI regimens, but also that selection of resistant HCV strains was the direct consequence of the inability of PegIFN/Rbv to suppress replication of minority resistant variants that are already present in most, if not all, HCV genotype 1 patients (Gastroenterology 2012;142:1369–1372). Second, through long-term follow-up studies of patients with a treatment failure to TVR/BOC, we recognized that a restoration to wild type HCV happens in ≤90% of patients after a 2-year off-treatment period (Gastroenterology 2012;142:1369–1372). From a clinical standpoint, this means that patients responding poorly to PegIFN/Rbv are those more prone to develop DAA resistance, and that these HCV resistant strains are unlikely to impair efficacy of future anti-HCV regimens that will be based on drugs targeting different and multiple key steps in HCV replication (Hepatology 2013 Mar 6 http://dx.doi.org/10.1002/hep.26371).

However, less is known about the clinical significance of HCV resistance to future treatment options. Indeed, treatment of HCV patients will be revolutionized by the advent of IFN-free regimens, that are currently in phase II–III development and that have been shown to be highly effective independent of HCV genotype (Hepatology 2013 Mar 6 http://dx.doi.org/10.1002/hep.26371). The rationale of IFN-free regimens is to combine drugs targeting different steps of HCV replication process, with the ultimate aim to suppress viral replication and allow clearance of infected hepatocytes. The first proof that this was possible came from 2 small studies that showed the combination of PI asunaprevir (ASV), and the NS5A inhibitor daclatasvir (DCV) resulted in high sustained virologic response (SVR) rates in Japanese patients and in US patients infected with HCV genotype 1b (N Engl J Med 2012;366:216–224; Hepatology 2012;55:742–748). In the study by Karino et al, the authors report on the characterization of virologic escape in Japanese HCV genotype 1b patients who previously failed (n = 21) or were ineligible/intolerant to PegIFN/Rbv (n = 22), who received treatment with a 24-week course of ASV plus DCV (J Hepatol 2012 http://dx.doi.org/10.1016/j.jhep.2012.11.012). In this phase IIa study, the authors performed population sequencing and clonal analysis sequencing of HCV at baseline, and during virologic breakthrough or posttreatment relapse. Interestingly, all cases of viral breakthrough (n = 3) and posttreatment relapse (n = 4) occurred in the ineligible/intolerant subpopulation. Noncompliance to treatment was not the apparent cause of these failures, because adherence by pill count was high in 6 of the 7 patients; in the lone nonadherent patient who relapsed posttreatment, the serum concentrations of ASV/DCV were considered optimal. All 3 patients with a virologic breakthrough had a moderate-to-high level of DCV resistant variants at baseline. However, the presence of DCV or ASV resistant-associated variants (RAVs) at baseline was not a strong predictor of treatment failure, because only 2 of the 11 patients with baseline RAVs experienced a posttreatment relapse. The most interesting and somewhat worrisome finding of the study, however, relies on the long-term virologic characterization of patients with a failure to ASV/DCV. The authors followed patients for 48 weeks after treatment completion and thus were able to report on HCV sequencing data performed for a prolonged off-treatment period. The 3 breakthroughs were characterized at the time of virologic failure by high level RAVs to both DCV (6467- to 65,000-fold) and ASV (120- to 280-fold), in all 3 patients the NS5A RAVs persisted through the 48 week follow-up period, whereas in 2 out of 3 patients the NS3 RAVS were replaced by wild-type virus. In the 4 patients with posttreatment relapse, a high level of RAVs to both ASV/DCV were found at the time of virologic relapse. However in the 48-week follow-up period, 3 out of 4 patients (75%) reverted to wild-type HCV at the NS5A region, whereas NS3 RAVs were not detectable in any patient at the week 36 follow-up visit. In summary, Karino et al demonstrate that virologic failure to ASV/DCV in HCV-1b patients is characterized by selection of RAVs to both DAA compounds. However, although NS3 RAVS failed to persist in most patients during follow-up, the NS5A RAVs were in all cases still detectable up to 48 weeks after treatment completion.


IFN-free regimens are the future of anti-HCV treatment, because their advent will provide significant benefits from clinical and epidemiologic points of view (J Viral Hepat 2012;19:449–464). If these regimens are affordable for national healthcare systems, it is probable that large chunks of HCV patients that are currently excluded from treatment for contraindications or are unwilling to start an IFN-based regimen for tolerability reasons will finally receive an effective antiviral treatment. Moreover, the availability of effective and tolerable treatments should also increase the number of HCV patients who are diagnosed and eventually referred to a specialist, hence eliminating 1 of the key limiting factors that precludes access to therapy nowadays (J Hepatol 2012;57:1326–1335). Although some issues, especially the affordability of these regimens, might preclude the fulfillment of these scenarios, it is projected that in 2015 the first IFN-free regimens will become commercially available. Phase II studies have convincingly demonstrated that by combining 2 or 3 DAAs for 12–24 weeks, up to 80%–90% of HCV-1 genotype patients can achieve an SVR (Hepatology 2013 Mar 6 http://dx.doi.org/10.1002/hep.26371). These mesmerizing figures have been only partially confirmed by real-life experiences or by studies enrolling more difficult to cure patients, where SVR rates have been slightly lower. Whether this is directly caused by clinical or epidemiologic factors such as advanced liver disease, virus subtype, ethnicity, or adherence remains unknown. The obvious question that follows these results is this: Why do patients fail an IFN-free regimen? Is this owing to selection for RAVs? Until now, RAVs have not played a clinically meaningful role in the treatment of hepatitis C for several reasons. First, there is no stable genetic reservoir for HCV and hence RAVs are not archived. Second, most resistant variants to TVR/BOC are unfit in terms of their replication capacity and failure to persist. The study by Karino et al confirms this for the second generation of PI such as ASV, because NS3 RAVs reverted to wild type in 6 out of 7 cases (85%) during the 48-week follow-up period. However, this was not the case for NS5A RAVs that persisted in all patients with an on-treatment viral breakthrough. This shows that high level NS5A RAVS that are selected during treatment are relatively fit in terms of replication capacity and might persist for a long period of time after treatment discontinuation. The clinical implications of this finding are currently hard to foresee, because one would expect wild-type HCV, which has emerged as the most fit to replicate over a long evolutionary period, would slowly return to be the dominant viral strain in all patients over time (Best Pract Res Clin Gastroenterol 2012;26:487–503). Still, the selection of a replication fit RAV could be problematic in terms of spread of the disease, as in theory new HCV cases could circulate dominant RAVs that will impair efficacy of future treatment options. Moreover, these RAVs might be clinically relevant in cases where viral replication is the direct cause of the disease, such as in post-liver transplant patients who develop fibrosing cholestatic hepatitis C.

When interpreting the findings by Karino et al, we need to remember that the study was conducted in a highly selected group of HCV-1b genotype patients without cirrhosis, who do not reflect the current epidemiology of HCV infection in most developed countries (J Hepatol 2011;55:245–264). Once these IFN regimens are moved into more difficult to cure subgroups of patients, like those with the HCV-1a subtype, advanced fibrosis or cirrhosis or HIV co-infection, the rate of RAVs selection might be further magnified. These subgroups of patients present a challenge not only as a consequence of impaired efficacy of some IFN-free regimens in the HCV-1a subtype, but also in terms of safety in patients with advanced fibrosis/cirrhosis and adherence in patients receiving multiple concomitant medications that might cause significant drug–drug interactions. Obviously, these are the worst-case scenarios, and to date we have no data to be sure that they will actually happen. Indeed the DAA combination used by Karino et al, a second wave PI plus an NS5A inhibitor, is now considered suboptimal as it features 2 drugs with low genetic barrier to resistance especially in patients with HCV-1a genotype. The optimal IFN-free regimen should in theory combine a drug with potent antiviral activity (PI or NS5A inhibitor) with a drug with high genetic barrier to resistance (NS5B Nucleoside inhibitor). However, some studies have shown that extremely high SVR rates can be obtained by combining a PI an NS5A inhibitor and an NS5B non-nucleoside inhibitor with or without Rbv (Hepatology 2013 Mar 6 http://dx.doi.org/10.1002/hep.26371). This demonstrates that matching an NS5A with a high genetic barrier DAA or with 2 low genetic barrier DAAs, should minimize the risk of selection of resistant viral strains. This seems especially true when an NS5A inhibitor is combined with an NS5B nucleoside inhibitor, which to date seems like the most well-tolerated IFN-free DAA regimen in terms of pill burden.

NS5B nucleoside inhibitors are characterized by a high barrier to resistance because the S282T mutation associated with decreased susceptibility to this class of compounds dramatically reduces HCV replication capacity (N Engl J Med 2013;368:34–44). This means that this mutation is very rarely found as a pretreatment naturally occurring variant and is also seldom found at the time of relapse (Hepatology 2012;56 Suppl 1:A551). In the context of this DAA combination, the finding by Karino et al, that viral breakthrough emerged only in intolerant/ineligible to PegIFN/Rbv patients could be clinically relevant. There is no clear-cut explanation for this, but the most reasonable is to attribute this finding to impaired adherence in this category of patients. The authors were quick to show that compliance with treatment, assessed through pill count, was high in the current study, but it is well known from the experience with chronic hepatitis B patients that compliance to simple and tolerable regimens is not as high as expected in real life (J Hepatol 2011;54:12–18). If these data were to be confirmed, they could suggest that in patients where optimal adherence can be problematic to achieve, a combination of only 2 DAAs that includes an NS5A inhibitor, could be suboptimal in terms of SVR rates and lead to selection of NS5A RAVs.

It is too early to draw sound clinical recommendations from the findings of the study by Karino et al, because many studies investigating IFN-free regimens are currently ongoing and we do not have solid data on resistance profiles yet; still, we think that overall Karino et al's data support a cautious approach towards future therapeutic regimens for HCV infection. Clearly, SVR rates will be increased by IFN-free regimens, and tolerability will be greatly improved compared with any PegIFN/Rbv-containing regimen, but the selection of durable resistant variant strains might be problematic, especially in some categories of patients. This piece of information not only needs to be taken into consideration when clinicians consider treatment deferral until the availability of IFN-free regimens for their patients, but should also probably be part of the informed deferral process that some consider a moral requirement for entry in the HCV treatment warehouse (Hepatology 2012;56:1591–1592).

PII: S0016-5085(13)00785-3


© 2013 AGA Institute. Published by Elsevier Inc. All rights reserved.


No comments:

Post a Comment