• Treating Hepatitis C: Criticisms of sofosbuvir

    The following is part of a series of guest posts on hepatitis C by Allan Joseph, a medical student at the Warren Alpert Medical School of Brown University and TIE research assistant. You can follow Allan on Twitter: @allanmjoseph. Links to all posts in the series to which this post belongs are in the introductory post. That post includes a glossary of terms as well. 

    The National Association of Medicaid Directors (NAMD) has recently been pushing a report by The Center for Evidence-Based Policy (CEBP) at Oregon Health and Science University arguing that the evidence for sofosbuvir is far flimsier than Gilead Sciences claims. Let’s go through some of the major claims the document makes. This isn’t a comprehensive summary of the CEBP report, though, so click through for the full thing if you’re interested.

    Claim: Real-world response rates are likely to be lower than those in clinical trials.

    This is a pretty uncontroversial claim, because effectiveness in the real world is almost always lower than in clinical trials. The question is whether this is particularly true of sofosbuvir — effectiveness would be even lower than we’d expect — because of the study design. While CEBP suggests that the study populations were biased towards getting better results, it doesn’t seem like that’s necessarily more problematic than, say, the trials reporting telaprevir. Clinical trials tend to pick the least complicated patients to isolate the effect of the drug. That’s nothing new here.

    It’s also possible that real-world effectiveness will be closer to efficacy than in other drugs because of the ease of adherence. Regardless, it’s not clear that this is a fatal flaw in the evidence. CTAF’s analysis uses clinical-trial data for all options and seemed to justify sofosbuvir even compared to ideal trials of the other options. Apples-to-apples, given the evidence we have, sofosbuvir seems to improve outcomes.

    Claim: The studies under-report adverse effects of sofosbuvir.

    The CEBP report says that because the evidence wasn’t based on long-term studies, it’s likely that there are long-term side effects of the drug that aren’t clear yet. That’s true, of course, but the question becomes one of a trade-off in drug-approval policies more generally: is it better to conduct longer pre-approval studies to capture more long-term effects, or is it better to get the treatment on the market sooner and monitor it later for long-term effects? In this case, the FDA appears to have erred on the side of bringing the drugs to market sooner. It warrants long-term management, but it’s a problem inherent to many new drugs.

    Claim: Many patients are likely to relapse after treatment with sofosbuvir.

    The CEBP report rightly points out that relapse rates for sofosbuvir range from 5-28% in various studies, dispelling the notion that sofosbuvir is a universal “cure,” even for patients who achieve SVR. That’s true, but PR has a relapse rate of about 28% , while telaprevir has a relapse rate of about 9%. Sofosbuvir’s right in that range, suggesting that while it’s no magic bullet, it’s not any different from previous standards of treatment.

    Claim: The studies were noncomparative in nature, and at a high risk of bias.

    This is perhaps the most serious and compelling accusation of the CEBP report. The studies of sofosbuvir never actually compared sofosbuvir to previous standards head-to-head. That’s because the FDA made a decision to allow drug companies to study new HCV drugs without standard-of-care comparators. It’s a complex decision summarized well here. While it meant drugs like sofosbuvir reached the market much faster, it also meant there was no head-to-head comparison of the drugs in randomized, controlled trials. That’s concerning from an evidence standpoint, because SVR rates from one study can’t necessarily be compared to rates from another study, which would have had a different population. Even worse, no currently-ongoing study compares sofosbuvir to previous standards of care. This is a big hole in the evidence for sofosbuvir, which leads to the final claim.

    Claim: “There is not yet clear evidence that this drug should be used routinely to treat patients.”

    This one’s probably true, if you focus on the routine portion of it. The research base simply isn’t strong enough to recommend that every patient with hepatitis C take this expensive drug. But we’ve got enough positive evidence to suggest that it should be tried in patients with advanced disease. That’s where we’re headed in the next post.


    Comments closed