This article is more than 5 years old. Information may no longer be current.

HCV Resistance: A Constellation of Considerations

As real-world data on direct-acting antiviral therapies emerge, signs indicate that the overwhelmingly positive sustained virologic response rates reported in clinical trials will hold firm.

As efficacy is continually shown, experts must look toward the ever-decreasing populations of difficult-to-treat patients. Here, HCV Next discusses with them whether resistance is an important factor determining outcome and if it should be accounted for prior to treatment. Small numbers of patients are inevitably going to fail, even with a robust armamentarium of effective therapies and combinations. Understanding all facets of those failures, including the resistance profiles, could help clinicians and researchers on the march toward eradication.

For Jordan J. Feld, MD, MPH, associate professor of medicine and Research Director of at the Toronto Centre for Liver Disease and HCV Next Editorial Board member, a critical component of this effort is paying close attention to individual patients. “To really understand the importance of these variants, they have to be reported correctly,” he said. “It is important to report drug-specific, rather than class-specific resistance-associated substitutions and then the patient population must be clarified as well — by genotype, subtype, treatment history and whether or not the patient is cirrhotic. Once you get into these small, well characterized subpopulations — ledipasvir-specific treatment-experienced genotype 1a cirrhotics with baseline NS5A variants, for example — you realize that it’s more important whether the virus is actually resistant to the drug. All of those specific data points matter. If we just look at the overall effect of NS5A variants in all patients as is often seen in the literature, the effect is diluted and gives the impression that resistance does not matter. Reporting in the literature needs to improve and we have to start thinking about it from the individual patient perspective.”

Jean-Michel Pawlotsky, MD, PhD, director of the National Reference Center for Viral Hepatitis B, C and D at Hopital Henri Mondor Université Paris-Est in Creteil, France, built on this point and addressed a point of semantics. “First, we should not say RAV, for resistance-associated variant, which is meaningless, but RAS, for resistance-associated substitutions,” he said. “With the drugs currently on the market, the presence of RASs at baseline remains an issue. They influence the response to therapy when they are present in large proportions at baseline with various regimens.”

Another key issue is the role of resistance testing, particularly as novel therapies and combinations gain approval. Ira M. Jacobson, MD, chair of the Department of Medicine and in the Department of Gastroenterology at Mount Sinai Beth Israel Medical Center, and HCV Next Co-Chief Medical Editor, weighed in. “It will be a while before we can give up resistance testing,” he said. “But emerging data suggest baseline RASs matter less and less with second generation regimens.”

Most experts, including Feld, believe that in the future, resistance testing may only be used before retreatment with salvage regimens. “We are starting to see strong data that retreatment regimens are working in a clinical trial setting,” he said. “Hopefully, resistance testing will be less and less important.”

Other issues at hand include the relative frequency of resistance to NS5A inhibitors compared with those observed for other drugs, the utility of various tools for sequencing, the role of guidelines and professional organizations and the importance of variables like genotype and cirrhosis.

Utility of Resistance Testing

Understanding all facets of the playing field is critical, according to Pawlotsky. He and Fourati wrote a paper outlining tools for the study of drug resistance and testing in clinical practice, including those at the drug development level and those that can be used in a clinical setting.

At the drug development stage, methods to characterize and understand resistance include HCV cell culture systems investigating phenotypic resistance, phenotypic assays in replicon systems, phenotypic assays in cell-culture derived HCV particles, assessment of cross-resistance, assessment of the replication capacity of the virus, cell-free in vitro assays assessing resistance and structural studies investigating resistance. In the clinical setting (clinical trials and practice), only sequencing tools, including population sequencing or next-generation sequencing, can be used.

“HCV resistance monitoring during drug development is fundamental to understanding the clinical impact of drug resistance,” they wrote. “In clinical practice, monitoring resistance for the persistence of RAVs will lead to better management of second-line therapy. In addition, resistance patterns derived from clinical samples are more representative of ‘real-life’ and could differ in complexity from those observed in clinical trials, therefore reinforcing the need to perform resistance testing in the context of virological failure in the clinical setting.”

Jacobson remains an advocate of baseline resistance testing in some patients. “I, for one, still feel it is useful, even when it’s not recommended,” he said. “As part of profiling my patient, I like to incorporate it into the constellation of considerations that go into my final decision of what regimen to use in patients with genotype 1a or genotype 3.”

“Resistance testing is currently not necessary before first-line therapy,” Pawlotsky said. “Treatment can hopefully be optimized by adding ribavirin or treating for a longer duration in certain subgroups of patients in whom RASs may have an impact. This is important because in most places around the world, HCV resistance testing is not available. If resistance testing is available, it can be useful to restrict reinforced therapy to patients with RASs detected at baseline.”

In contrast, in patients who failed a DAA-containing regimen, resistance testing is useful in assigning a retreatment regimen, according to Pawlotsky. “This should be done in specialized centers as most of the good retreatment options currently are off-label,” he said.

Feld agreed that the role of resistance testing will likely be relegated to second-line therapies. “However, resistance testing is still cheaper than even one extra pill, because these therapies are so expensive,” he said, but also offered some practical considerations. “I still would like to see comprehensive resistance information written in the report but, more importantly, I would like to see suggestions on how to use specific regimens given the presence of specific RASs. For example, rather than just saying ‘resistance likely’ to elbasvir, it would be more useful to recommend that therapy should be extended to 16 weeks with the use of ribavirin if elbasvir/grazoprevir (Zepatier, Merck) is going to be used in patients with genotype 1a. This information should be immediately available. The problem is labs don’t want to recommend how to treat patients.”

He added that while resistance testing is done centrally in Canada, with links in the report to Canadian treatment guidelines, the situation is different in other places. “Reports in the U.S. are not very user friendly,” he said.

Ahmed and Felmlee wrote about the importance of assessing for RASs in the pre-treatment setting. “In sofosbuvir (Sovaldi, Gilead Sciences) trials, while there were variants that emerged and were statistically significant with resistance, the majority of subjects that experienced relapse did not carry identifiable RAVs,” they wrote. “For other drug classes, the emergence of resistant variants may be derived from a very small proportion of the quasispecies that can only be detected with [next-generation sequencing], accompanied with costly analysis.” They suggested that resistance works in combination with other factors, including host response. This makes prediction of treatment response based on sequencing “untenable.”

They recommended, “At this time, data need to be collected on all of these classes of drugs and the next-generation of agents in various combinations for comprehensive understanding of how to minimize or ablate breakthrough mutations.”

Previous Treatment Failures

Pawlotsky recently published an editorial about treatment failures associated with HCV resistance. “This was an editorial commenting on a paper reporting in vitro results with entry inhibitors, showing that such molecules have no cross-resistance with current HCV DAAs and could thus be used for retreatment of failing patients,” he said. “This is a good idea, but these approaches are unlikely to be available for the clinic; they remain purely experimental. ... Having one more class of drugs with a high barrier to resistance such as entry inhibitors or host-targeted agents would help for rescue therapy, although this is unlikely to happen.”

Jacobson noted that triple regimens and second-generation regimens from the major pharmaceutical players are likely to diminish that small population even further, and that the clinical community awaits FDA approvals and package insert recommendations for patients who fail first-line therapy. “Retreatment strategies with second generation regimens might be expected to be tailored to the RAS information that we find in these patients,” he said. “Interestingly, however, data presented at recent meetings, particularly with the triple combination of sofosbuvir, velpatasvir and voxilaprevir suggest that RASs in patients who have failed previous DAA regimens do not influence retreatment outcomes.”

Feld suggested that continued examination of real-world data sets will help guide response to treatment failures. “We need to take a hard look at the real-world studies as opposed to data from clinical trials,” he said. “Patients in clinical trials are cherry picked. There are no adherence issues, no decompensated cirrhotics. We need to understand where resistance fits into all of that.”

Paul Y. Kwo, MD, professor of medicine and director of hepatology at Stanford University School of Medicine, said that even in his practice, a certain degree of selection happens. “We meet with patients and conduct a social and psychiatric assessment. We look at substance abuse and other factors and use a few clinical visits to determine whether they will be compliant with therapy,” he said. “We have selected certain patients out or deferred therapy if there are indications that they cannot comply with a course of DAA therapy. It is important to recognize that many of us haven’t marched into populations that will be non-compliant.”

Jacobson raised a practical question about the difference between real-world and clinical trial data sets. “I’m not sure I agree with the hypothesis that real-world data show lower SVR rates than what we’ve seen in clinical trial populations,” he said. “When it exists at all, the difference between trials and real world is very slight.” That said, Jacobson acknowledged one difference. “Real-world trials may have more decompensated cirrhotics, who were not included in pivotal trials,” he said. “If you mix in some of these patients, you are going to see lower SVR.”

Despite his concerns, Feld remains optimistic that many of these issues will be eliminated with emerging therapies. “The other thing that real-world studies have shown us is that resistance isn’t an insurmountable problem,” he said. “With so many drugs in the salvage armamentarium, most people will respond to something. We just have to throw the kitchen sink at them.”

C-ISLE

There is mounting evidence that this approach may work. Results of the C-ISLE study, presented at The Liver Meeting in 2016, indicated that 96% of treatment-naive and 97% of treatment experienced patients with genotype 3 disease and cirrhosis who were treated with grazoprevir/elbasvir plus sofosbuvir reached SVR12. There were 100 patients included in the analysis. Patients in the treatment-experienced arm received the triple regimen with or without ribavirin for 12 weeks or without ribavirin for 16 weeks. The SVR12 rate was 100% in the 12-week group without ribavirin, 94% for 12 weeks with ribavirin and 94% in the 16-week group.

This study also touched on a point about the way treatments are delivered, according to Feld. “An unfortunate aspect of the single tablet regimens is that you can’t combine, say, an AbbVie NS5A inhibitor with sofosbuvir or another company’s drug,” he said. “With the exception of this [grazoprevir/elbasvir plus sofosbuvir] regimen, you can’t mix and match. Again, as we see drug-specific, genotype-specific or subtype-specific resistance, it would be nice to use a [protease inhibitor (PI)] from one company and an NS5A inhibitor from another company.”

Kwo raised another concern with this approach. “The data show that this combination was highly effective, but the issue here is going to be the cost,” he said. “You can mix and match these therapies, but it will be difficult to get insurers to pay for it.”

That said, the SVR rates for this real-world study mirrored registration trial results, according to Kwo. “The places where lower SVR rates have been observed have largely been in genotype 3 and genotype 1 patients with decompensated cirrhosis,” he said. “In the real world, the issue has been that some decompensated cirrhotics don’t tolerate ribavirin particularly well. If you can’t tolerate ribavirin in your retreatment regimen, you may have a problem.”

POLARIS trials

Data indicate that resistance is most common with NS5A variants.

Feld explained, “You can either report a resistance substitute to NS5A as a class, or resistance to ledipasvir or another NS5A inhibitor. ... They are not necessarily the same.”

The FDA guidelines indicate that to prevent cross-resistance, patients who fail on an NS5A inhibitor should be retreated with sofosbuvir and simeprevir (Olysio, Janssen), but with next generation drugs coming into play, neither AASLD nor EASL maintain this recommendation in their latest guidelines.

The four POLARIS trials are key data sets in understanding NS5A resistance. POLARIS-1 included 415 patients with all genotypes who had failed an NS5A inhibitor-containing regimen, more than half of whom (55%) had undergone treatment with ledipasvir-containing regimens, while 23% of patients had previously been treated with daclatasvir (Daklinza, Bristol-Myers Squibb). POLARIS-2 included 941 patients with genotypes 1 through 6 who were DAA-naive with or without cirrhosis, except patients with genotype 3 and cirrhosis who were studied separately.

The patients in POLARIS-2 underwent treatment with either sofosbuvir/velpatasvir/voxilaprevir (SOF/VEL/VOX, Gilead) for 8 weeks or sofosbuvir/velpatasvir (Epclusa, Gilead) for 12 weeks. Twenty-three percent of patients had previously failed a peginterferon-based regimen and 18% had cirrhosis. POLARIS-3 included cirrhotics with genotype 3 HCV who were treated with 8 weeks of SOF/VEL/VOX or SOF/VEL for 12 weeks. There were 231 patients in the study, with 31% who had previously failed on interferon. The open-label POLARIS-3 study randomly assigned patients with genotype 3 HCV infection and cirrhosis to receive SOF/VEL/VOX daily for 8 weeks or SOF/VEL for 12 weeks. Of the 219 patients treated, 31% had previously failed treatment with an interferon-based regimen. In POLARIS-4, patients were treated with 12 weeks of both SOF/VEL/VOX or SOF/VEL. There were 333 patients with genotypes 1 through 4 disease who had DAA experience, but no experience with an NS5A inhibitor. Eighty-five percent of this cohort had sofosbuvir experience.

POLARIS-1, -3 and -4 met primary endpoints regarding SVR12 in patients treated with the triple regimen. In POLARIS-1, the triple regimen showed 96% SVR12 vs. 0% for placebo. POLARIS-3 produced the same SVR12 — 96% — for 8 weeks of the triple regimen and 12 weeks of SOF/VEL. POLARIS-4 treated patients for 12 weeks, either with SOF/VEL/VOX or SOF/VEL, resulting in 97% SVR12 for the triple regimen and 90% for the double. POLARIS-2 failed to reach a pre-specified 5% margin in terms of SVR12 for 8 weeks of the triple regimen vs. 12 weeks of SOF/VEL, with SVR12 rates of 95% and 98%, respectively.

“The POLARIS trials showed us a couple things,” Kwo said. “The first is that the combination of three DAAs — in this case sofosbuvir, velpatasvir and voxilaprevir — will be a highly effective salvage regimen in DAA-exposed individuals. The combination of three DAAs is highly effective, even in DAA-exposed individuals, and is associated with a very low rate of treatment emergent resistance-associated substitutions.”

He added that there is only a very small chance that patients will fail on three drugs with a RAS that could make future therapy more problematic. “These studies also showed that in DAA-exposed individuals, the three DAAs sofosbuvir/velpatasvir/voxilaprevir performed better than the two DAAs sofosbuvir/velpatasvir in the non-NS5A-exposed class,” Kwo said.

“We need more retreatment studies like the POLARIS trials,” Feld said. “However, to date, they have only been presented in oral format, which makes it a bit hard to dig deep and see if there were specific variants that still cause a problem. These data were presented in a very loose way. DAA- rather than class-specific resistance information was not included.”

Other NS5A Data

Hezode and colleagues suggested the FDA guidelines lack evidence to support retreatment with sofosbuvir/simeprevir in the event of failure on an NS5A inhibitor. Their study to determine its validity comprised 16 patients who had failed previous daclatasvir plus peginterferon and ribavirin, including 13 patients treated with asunaprevir (Sunvepra, Bristol-Myers Squibb) and three patients who were not treated with asunaprevir. They were assigned 12 weeks of retreatment with sofosbuvir/simeprevir without ribavirin. There were 11 genotype 1a patients, three with 1b disease and two with genotype 4 HCV. The cohort included nine patients with advanced fibrosis or compensated cirrhosis. All patients reached HCV RNA below the lower limit of quantification by the end of treatment.

A response by week 4 of therapy occurred in 10 of the 16 patients. Fourteen of 16 patients reached SVR12. Relapses by week 4 post-treatment occurred in two patients with genotype 1a cirrhotics who had failed in the regimen containing asunaprevir. R155K and Q80K polymorphisms failed to predict retreatment failure, according to the results.

Pawlotsky was the co-principal investigator on this trial. “We retreated patients who failed an NS5A-inhibitor containing regimen with drugs with no cross-resistance with the already administered drugs. This is why we picked sofosbuvir and simeprevir,” he said. “It was a small series and we had excellent SVR rates. Only two genotype 1a patients did not achieve SVR, most likely because we did not include ribavirin. We concluded that sofosbuvir/simeprevir is a good option to retreat patients who failed an NS5A inhibitor containing regimen, but better options with more DAAs and ribavirin are available and should be preferred, as indicated in the 2016 EASL guidelines.”

In another paper, Pawlotsky wrote that the presence of resistance to NS5A inhibitors at baseline had led to reduced SVR12 in certain groups, including those with genotype 1a or 3 disease, cirrhotics, and those who failed on peginterferon-based regimens. He added that resistant disease is frequently dominant in cases of virologic failure. “Viruses resistant to NS3-4A protease inhibitors disappear from peripheral blood in a few weeks to months, whereas NS5A inhibitor-resistant viruses persist for years,” he wrote. While retreatment options are emerging, optimizing first-line therapies to prevent failure in the first place is of critical importance, according to Pawlotsky.

Feld remains optimistic. “We are soon very likely to have triple combinations that will overcome these specifics,” he said.

Other Issues at Hand

For Kwo, it is worth noting that the discussion of RASs and treatment failures pertains to a very small patient population. “Failure rates for first-line therapies are very low,” he said. “If you keep slicing it, you’re going to end up with a small number of people that will be difficult to treat.”

He offered a few options for moving forward for the very rare patient who might not be able to achieve SVR. “One possibility in a decompensated patient who cannot achieve SVR is to offer them a liver transplant,” he said. “Post-transplant, the patient will have a new liver with no decompensation, so we will be able to offer a protease inhibitor, which is not recommended in transplant candidates. As we have seen, the triple therapies work well as salvage regimens.”

The other option is to increase the duration of therapy in compensated and decompensated patients. “We must pay attention to the role of ribavirin in decompensated patients, but using a sofosbuvir-based regimen with ribavirin for 24 weeks may be sufficient for retreatment,” he said.

Based on the ASTRAL-4 study, Jacobson echoed that ribavirin is still considered necessary in these patients unless contraindicated.

“The available data actually suggest that ribavirin supplants the need for extension of therapy to 24 weeks,” Jacobson said.

Another consideration pertains to HCV genotypes seen infrequently in the U.S., according to Jacobson. “Resistance to genotypes 4 and 6 is not a major concern in the U.S.,” he said. “They are not often included in clinical trials, so we are not quite sure what to make of those genotypes. I do know that we are relatively lacking in data and recommendations with baseline resistance for those genotypes.”

Kwo added: “It is important right now that we correctly characterize the people who fail therapy,” he said. “The good news is that this population won’t be large.” — by Rob Volansky

• References:
• Ahmed A, Felmlee DJ. Viruses. 2015;doi:10.3390/v7122968.
• Foster GR, et al. Abstract #74. Presented at: The Liver Meeting; Nov. 11-16, 2016; Boston.
• Fourati S, Pawlotsky JM. Viruses. 2015;doi:10.3390/v7122941.
• Hezode C, et al. Hepatology. 2016;doi:10.1002/hep.28491.
• Pawlotsky JM. Gastroenterology. 2016;doi:10.1053/j.gastro.2016.04.003.
• Pawlotsky JM. Nat Rev Gastroenterol Hepatol. 2015;doi:10.1038/nrgastro.2015.184.
• Perales C. Viruses. 2015;doi:10.3390/v7112902.
• Xiao F, et al. PLoS Pathog. 2014;doi:10.1371/journal.ppat.1004128.

• For more information:
• Jordan Feld, MD, MPH, can be reached at the Toronto Centre for Liver Disease Sandra Rotman Centre for Global Health, 200 Elizabeth Street 9EB 240, University of Toronto, Toronto, ON M5G 2C4 Canada; email: Jordan.Feld@uhn.ca.
• Ira M. Jacobson, MD, can be reached at 10 Union Square E #2G, New York, NY 10003; email: ijacobson@chpnet.org.
• Paul Y. Kwo, MD, can be reached at Stanford University School of Medicine, 750 Welch Road #210, Palo Alto, CA, 94304; email: pkwo@stanford.edu.
• Jean-Michel Pawlotsky, MD, PhD, can be reached at Hopital Henri Mondor, Department of Virology, 51 Avenue du Maréchal de Lattre de Tassigny, 94010 Creteil, France; email: jean-michel.pawlotsky@aphp.fr.

Disclosures: Feld reports receiving support for research and/or honoraria for scientific consulting for AbbVie, Abbott, Bristol-Myer Squibb, Gilead, Janssen, Merck, Theravance and Trek. Jacobson reports consulting for AbbVie, Achillion, Bristol- Myers Squibb, Gilead, Intercept, Janssen, Merck and Trek; receiving grant or research support from AbbVie, Bristol-Myers Squibb, Gilead, Janssen and Merck; and speaking and teaching for AbbVie, Bristol-Myers Squibb, Gilead and Janssen. Kwo reports financial relationships with AbbVie, Bristol-Myers Squibb, Gilead Sciences, Janssen and Merck. Pawlotsky reports being an advisor and speaker for AbbVie, Bristol-Myers Squibb, Gilead, Janssen and Merck.