Wednesday, May 28, 2014

Sovaldi - a "Revolution" in Clinical Care, or in Marketing and Public Relations?

The continuing public discussion of the sky high price Gilead has set for Sovaldi (sofosbuvir,) its new antiviral drug for hepatitis C, continues to avoid considering the lack of good evidence that the drug is as safe and effective as its proponents claim.

We first posted about the Sovaldi debate on March 27, 2014, and its focus on price rather than the quality of the evidence underlying loud claims about the miraculous qualities of the drug.  We suggested that there is no good evidence supporting claims that most hepatitis C patients have very bad outcomes if untreated, treatment prevents most bad outcomes, Sovaldi cures nearly all patients, and Sovaldi has very few side effects.

In a subsequent post we wrote that initially, "no one but your humble blogger seemed to be publicly skeptical about published assertions that the drug was some sort of modern miracle, and a triumph of medical science."

On May 7, we noted an assessment by the German Institute for Quality and Efficiency in Health Care (IQWiG) of information submitted by "industry" (presumably Gilead) to the German government.  This assessment found multiple problems with the evidence, including limited generalizability, problems with randomization, lack of information about important outcomes, and lack of ability to quantify benefit.   Also  the US based Institute for Clinical and Economic Review noted that there was little evidence that compared to previous treatments, Sovaldi is better, and no evidence that Sovaldi produces cures in the long term.

Nonetheless, the notions that the evidence supporting Sovaldi (and perhaps other similar drugs) is weak, and that the drugs therefore should not yet be considered miracle cures have not seemingly affected the public discussion. 

Examples of the Latest Discussion

Washington Post/ Kaiser Health News

On May 12, 2014, in an article on the dilemma the drug's US price of $1000/ pill presents to Medicare, Richard Knox wrote this about a patient with the infection:

Previous drug treatments didn't clear the virus from Bianco's system. But it's almost certain that potent new drugs for hep-C could cure him.

In other words, the article asserted that Sovaldi and similar drugs cure nearly everyone with hepatitis C, even those not cured by previous treatment.


On May 20, 2014, in an article about how US health insurers are balking at the price of Sovaldi, was this statement by the main trade organization for US for-profit health care insurers, America's Health Insurance Plans (AHIP),

Sovaldi has shown tremendous results, and it's the kind of medical innovation we need to sustain. 

In the article's text was the assertion,

The new drug has demonstrated an ability to cure well over 90 percent of patients in just 12 weeks or less with few side effects.

Prior to the Sovaldi approval, hepatitis C treatments took 24 or 48 weeks, cured about 75 percent of patients and involved many more pills as well as injectable interferon that causes flu-like symptoms and other side effects that led many people to avoid or discontinue treatment.

In other words, the article asserted that Sovaldi can cure over 90% of patients compared to the cure rate of 75% provided by previously available treatments, and implied Sovaldi has fewer side effects.


On May 22, 2014, in an article on the high costs of new drugs, a quote from Dr Douglas Dieterich, "a liver disease specialist at Mount Sinai Hospital," who "has consulted for pharmaceutical companies, including Gilead," appeared,

I don’t think there’s any question that treating patients with hepatitis C will lower overall health-care costs in the coming 20 years,...

If we could get rid of the liver disease in these patients with hepatitis C, prevent them from dying of liver cancer, cirrhosis and liver failure, then there’s no question the cost will be less.

The implication was that the new drugs can get rid of the disease, that is, cure it, and in doing so prevent early mortality, cirrhosis, and liver failure.


The discussion of Sovaldi in Canada seems similar. On May 25, 2014, an article on the high cost of hepatitis C drugs in Canada from CTV, "Canada's largest private broadcaster," (look here) quoted a Canadian physician,

 Dr. Curtis Cooper, director of the viral hepatitis program at the Ottawa Hospital, said the drugs Sovaldi and Galexos offer a revolution for patients with the hepatitis C virus (HCV).

'It only requires 12 weeks of treatment and (they) are producing cure rates of 90 to even 100 per cent,' he told CTV News.

And later, addressing the treatment of a particular patient,

We're talking about curative therapies, which could potentially save her from liver failure, save her from liver cancer

Again, the assertions were that the drug cures 90%, maybe 100% if patients, and that cure will prevent liver failure and cancer.

Is There Any Good Evidence?

Again, while there is much discussion, and some outrage over the $1000 per pill price of Sovaldi in the US (in Canada, a bargain at Canadian $650 per pill), all the discussion seems to assume that the pill is really a "revolution" (as per CTV), that provides
- cure rates of 90 - 100%, much better than previously available treatments
- lower adverse effect rates than than those caused by previously available treatments
-  prevention of complications of hepatitis C, including cirrhosis, liver failure, liver cancer, and early death.
I have found just one recent media article that throws a bit of evidence-based cold water on these claims, and refers to a new systematic review that should be generating a lot of interest, and provoking much more skepticism about the drug, but so far is not.

The Single Skeptical Article

On May 22, 2014, a MedPage Today article noted a new systematic review of sofosbuvir (Sovaldi) that had a very different message from the articles above.  In summary,

The evidence base for one of the star hepatitis C drugs is poor and the guidelines for its use are flawed, according to a report obtained by the National Association of Medicaid Directors.

According to the report, studies of sofosbuvir (Sovaldi) are generally of poor quality, mostly directed by the drug's maker, and don't answer key questions, including whether the drug is better and safer than the current standard of care.

The only available guidelines for its use -- guidelines created by the American Association for the Study of Liver Diseases and the Infectious Diseases Society of America -- are 'methodologically flawed,' according to the report, which was prepared by the Center for Evidence-Based Policy at Oregon Health and Science University in Portland (OHSU).

In addition, their authors and sponsors had 'multiple and significant conflicts of interest,' the report argued.

The Report from the Center for Evidence-Based Policy

The report is now public, and directly addresses all four of the major claims made about Sovaldi that we discussed in our first post n the topic.

- Most hepatitis C patients have very bad outcomes if untreated

The report cited the best data about outcome prevalence,

approximately 15% to 25% of people infected with HCV will clear the virus during the acute stage without treatment.  Seventy-five to 85% of infected individuals will develop a chronic HCV infection, and 60% to 70% of patients with chronic infection will develop chronic liver disease.  Over 20 to 30 years, 5% to 20% of infected patients will develop cirrhosis and 1% to 5% will die of cirrhosis or liver cancer.

Although a majority of chronically infected patients will develop some liver disease, only a minority will develop cirrhosis or liver cancer.

- Treatment prevents most bad outcomes

The report stated,

Because of the slow progression of the disease, clinical trials have not evaluated these patient-important conditions [cirrhosis, hepatocellular carcinoma, decompensated liver disease, liver transplant, or death] as trial outcomes.  Instead a surrogate endpoint of sustained virologic response (SVR) has been used to measure success of treatment.  The SVR is defined as undetectable HCV-ribonucleic acid (RNA) levels.  The standard measure of treatment success has been SVR at 24 weeks post treatment (SVR24).

Several long-term studies of patients with chronic HCV infection have shown an association between achieving SVR24 and patient-important clinical outcomes.

So there is no direct evidence that treatment prevents any of the bad outcomes.  There is only indirect evidence that treatment may reduce the rate of some bad outcomes.  For example,

A 2014 observational study of a VA population found that ... the 5180 (4%) of patients who were able to achieve an undetectable viral load with interferon-based treatment had a 45% reduction in the risk of death ... and 27% reduction in the composite clinical endpoint ... of newly diagnosed cirrhosis, HCC [hepatocellular carcinoma], or liver related hospitalization.

However, at best, based on an observational analysis that could have been biased, treatment only may have prevented a minority of bad outcomes.

- Sovaldi cures nearly all patients

As we noted earlier, there have been only two randomized controlled trials published that compared sofosbuvir with anything else.  One compared it to placebo, and one was the trial we discussed earlier that compared it sofosbuvir with ribavirin to pegylated interferon with ribavirin.  (Look here.)  

All other studies were designed to refine drug dose, drug combination or duration of treatment.


All studies were rated as having a high risk of bias.  No study was judged to have good applicability....  The overall summary judgement for each of the published studies yielded a rating of poor.

Furthermore, the review found even more problems with the only study that compared sofosbuvir to active treatment (peg-interferon) than we did.  In particular, that study did not compare sofosbuvir with ribavirin to the current standard regimen of PEG plus weight-based ribavirin.  Instead, it compared it to a regimen that included low dose ribavirin.  By comparing sofosbuvir plus a higher dose or ribavirin, an active drug, to a peg-interferon plus a lower dose of ribavirin, the study design seemed designed to artificially enhance the efficacy of the sofosbuvir containing regimen.   Furthermore, the study did not assess the measure of sustained viral response at 24 weeks currently accepted as the best surrogate variable.  Instead it used SVR at 12 weeks, which may be correlated with SVR24 but is often higher. Thus this choice of endpoint seemed designed to increase the apparent efficacy of the regimens it evaluated. 

So sofosbuvir has so far been compared to another active anti-viral regimen against hepatitis C in only one study.  That study was poorly designed and implemented, and its problems seemed likely to enhance the apparent efficacy of sofosbuvir.  Nonetheless, that one study did not show that sofosbuvir was more efficiacious, or safer than peg-interferon.

- Sovaldi has very few side effects.

The review had this summary,

The FDA compiled reports of adverse events from four trials....  There were no treatment-related deaths reported [note that we found there were deaths in the one trial that compared sofosbuvir to PEG, look here].

Approximately 78% of patients receiving placebo, 88% of patients on SOF + RBV treatment and 95% of patients receiving PEG + SOF +RBV reported a side effect from treatment.  The most common side effects were fatigue, anemia, nausea, rash, headache, insomnia and pain....

Thus it seems likely that Sovaldi has a lot of side effects, and whether it has fewer than standard treatment is unclear.  Furthermore,

studies on sofosbuvir were small, included populations that were healthier than the general hepatitis C population, were of short duration and had limited follow-up.  In many of the studies, the manufacturer was responsible for recording and reporting adverse events.  In general, reporting of adverse events is often incomplete and discrepancies between clinical trial reports and publications are common....  All of these factors would lead to a bias in under-reporting the true nature of adverse events.

Thus there is no good evidence that sofosbuvir has few side effects, or is dramatically safer than older treatments.


While there continues to be concern, if not outrage, that the latest treatment for hepatitis C is priced at $1000 per pill, most of those expressing concern seem to assume that the pill is a wonder drug, promising nearly everyone a cure without major side effects.  However, as we first noted in March, 2014, there is no strong evidence to that effect.  In fact, now three skeptical looks at the evidence by people with more resources and perhaps more expertise than we possess have shown similar conclusions.

It is a tribute to the power of the anechoic effect that there has been almost no recognition by physicians and other health professionals, journalists, and most amazingly, insurance companies who stand to lose billions paying for Sovaldi, that there is little good evidence that Sovaldi works, much less is superior to previous treatments.  Instead, even America's Health Insurance Plans thought the drug had "tremendous results," not very different from the assertion by the drug's manufacturer that the drug provides "a finite cure," (as reported by Reuters).  One might think that the insurance companies have enough money to invest in some real evidence-based medicine experts who could provide a skeptical assessment of the pricy new drugs and devices for which these companies may pay.  Is it that the commercial insurers are so now so dominated by generic managers who know nothing about health care, medicine, or biomedical science, much less evidence-based medicine that they are unable to resist the marketing and public relations hype?

The Sovaldi case is a signal example of how our health care system is awash in marketing hype and public relations buzz that has swamped rational skeptical thinking about logic and evidence.  That marketing and PR is ever enriching managers while it will send the rest of us, health care professionals included, to the poor house.  And all the money we spend will not buy us the promised miracles and triumphs.

As we have said until blue in the face, true health care reform would bring some skeptical thinking and regard for evidence and logic into the health policy discussion.  

ADDENDUM (25 July, 2014) - Web link added for CEBP report. 


Sally said...

The drug is also being portrayed as a miracle in Australia, with the only issue being the price. The lack of critical scrutiny of the evidence appals me.

Thank you for plugging on with this and other issues!

Anonymous said...

So solvaldi is effective in ridding the body of the HCV virus, but you see no value in that? LOL

Clearly if SVR12 is achieved the virus is gone since even a single replicating cell would grow to detectable levels in 12 weeks. Ridding the body of the virus before liver damage occurs would prevent further damage, even if does not reverse the effects that the damage has already caused.

Roy M. Poses MD said...

Anonymous of 29 May, 2014,

- You misread my post. I did not say there is "no value" in SVR12. I did say that is not the same as SVR24, tends to be higher than SVR24, and has not been shown to predict any clinical benefit
- Clearly, you assume that the test is perfect, and the virus cannot hide somewhere other than the blood. I do not believe either is proven.

Anonymous said...

Very interesting posts, Roy. Thank you!

The frustrating lack of convincing evidence may be due in part to
the FDA's fast tracking sofosbuvir as a Breakthrough Therapy
Designation drug aimed to treat serious or life-threatening
disorders (such as cancers:}.

Maggie Mahar said...


Thanks very much. I had been wondering if Solvadi was truly
a miracle drug.

When an expensive new drug comes to market people tend to believe what the manufacturer says about how effective it is. After all, if it wasn't effective, how could they charge so much?

This seems to me another example of the Nazi's "Big Lie" theory: if the lie is big enough (or the price is high enough) people will believe it.

I also tend to agree with the reader who suspects that the "fast-tracking" of the drug may
have meant that the FDA's appraisal was less than thorough.

I have never liked "fast-tracking" unless we are faced with an infectious epidemic--and all patients are aware that the new drugs are just that--very new. (Here, I'm thinking of AIDS. That was a crisis that demanded getting drugs to patients as quickly as possible. But they knew they were guinea pigs.)
Otherwise, fast-tracking just isn't necessary.