Wednesday, April 16, 2014

Knee Deep in the Hoopla - "A Triumph of Medical Technology" Sans Evidence of Superiority from Published Randomized Double Blind Controlled Trials

With the publication of several new articles in the prestigious New England Journal of Medicine, the buzz about Sovaldi (sofosbuvir - Gilead), a new oral treatment of hepatitis C, has become feverish.  The drug had been hailed as near miraculous, and the only debate has been about its near stratospheric cost (look here). 

This week, of all weeks, however we should be extremely skeptical about expensive drugs promoted as the cures of our most feared ills.  This week a series of articles appeared in the British Medical Journal showing that there was no evidence to support the clinical value of the  antiviral drugs stockpiled for governments as proof against dreaded influenza epidemics (look here to start).  

The Latest Buzz

As we posted earlier, the discussions contrast the ostensibly near miraculous aspects of the drug's performance with its extremely high price.  For example, a few days ago the Los Angeles Times reported,

In a series of clinical trial results, a new generation of antiviral medications was able to clear the liver-ravaging virus from virtually all patients' bloodstreams in as little as eight weeks. Even in patients with the most stubborn infections, the new drugs were capable of suppressing the virus completely at rates well over 90%. The treatments, however, come with a steep price tag.


The new medications are 'a triumph of modern medical technology,' said Dr. Jeffrey Tice, a UC San Francisco physician who was not involved in any of the clinical trials.

Accompanying the research reports in the New England Journal was a commentary by Dr Raymond T Chung and Dr Thomas F Baumert.  Its title proclaimed Sovaldi and similar drugs also to be part of "The Arc of Medical Triumph."(1)  Buried in the commentary's supplementary material was a link to disclosure forms that showed that:

Dr Chung reports grant support from Gilead Sciences [the manufacturer of Sovaldi], Mass Biologics, Vertex and Merck, and personal fees from Abbvie, Enanta and Idenix.

Note that Abbvie and Merck are also at work on antiviral treatments for hepatitis C.


Dr Baumert reports several patents ... and pending patents [apparently that could be conflicts of interest]

It is clear that there have been major scientific advances in the characterization of the hepatitis C virus, and development of multiple new antiviral agents which promise to treat that infection.  Furthermore, clearly there is reason for hope for treatment of hepatitis C that prevents its long-term complications and averts premature death. (Of course, that hope could be a bit stronger in the hearts of those who have financial relationships with the corporations that stand to profit from sales of these drugs.)

But it is not obvious that there is sufficient clear evidence from clinical research, particularly from double-blind randomized controlled trials (RCTs), that shows the newest treatments will provide triumphal benefits to patients that outweigh their possible harms.

One Published Unblinded RCT Compared Sovaldi to Peg-Interferon

In our previous post, written after the immense price of the new drug in the US became apparent, we questioned whether the evidence from controlled trials then available would justify the hype, and the high price.

We looked at what appeared to be the then single best available study. The most prominent randomized controlled trial of Sovaldi for patients who had not had previous therapy for hepatitis C that had been published by March, 2014, appeared in the New England Journal of Medicine in May, 2013.(2)   While it provided some reason for hope, it did not provide clear evidence that Sovaldi was a miracle drug.

The study showed that Sovaldi plus ribavirin produced the same rate of sustained virologic response (SVR), 67%, as did the previously accepted standard treatment, peg-interferon plus ribavirin, again 67%.  (SVR means that the hepatitis C virus has become undetectable in the patient's blood, and is believed, but not proven to represent a cure.)

While Sovaldi had lower rates of unpleasant side effects than did peg-interferon, these rates were not negligible. For example, the rate of nausea after peg-interferon was reported to be 29%, but after Sovaldi it was still 18%.  Given that the trial was unblinded, and that some of these side effects have a subjective element, these figures could have been biased due to patient expectations that the new drug would have fewer side effects.

Furthermore, Sovaldi appeared to produce higher rates of severe adverse effects (3%) than peg-interferon (1%).  However, the article did not mention any deaths in either treatment group.  When I later reviewed the  supplemental data available on the web from the May, 2013 article, I found that buried within this information were the facts that two people died during treatment with Sovaldi, and one after treatment with peg-interferon.  So, the rate of  severe adverse effects or death after Sovaldi treatment was more than double (9/256 ) than that after peg-interferon (4/243).  It is very curious and concerning that the authors chose not to mention deaths of study subjects in the paper that reported their trial results.   

Finally, the trial, like just about all previous trials of treatment of hepatitis C, did not follow patients long-term, and hence could not show whether treatment actually resulted in better clinical outcomes, for example, longer survival, decreased rates of severe liver disease, etc.  We thus concluded that there really was no clear evidence that Sovaldi was really a miracle drug, which could cure more people, in fact, nearly everyone, compared to standard therapy, yet with fewer side effects and more safety.

Any Evidence from Other Controlled Trials?

We looked through the newly published articles, the articles they cited, and the official US label for sofosbuvir to see if there is any other evidence from controlled trials to support the triumphant claims about sofosbuvir.  To simplify this casual, not systematic review, we only looked at articles about patients who had not been previously treated for hepatitis C.

The drug label mentioned a controlled trial that compared sofosbuvir to placebo.  The POSITRON study (study 107) was a trail that compared Sovaldi and ribavirin to placebo in patients who are "interferon intolerant, ineligible or unwilling."  The overall SVR in the treated group was 78%.  The label did not mention adverse effects rates in this trial, and as far as I can tell, it was not published.  Again, while the SVR was good, it was NOT over 90%.  As far as I could tell, the results of this study have not otherwise been published. 

The new studies in the New England Journal, and the studies cited by one of them were NOT double-blind randomized controlled trials of sofosbuvir versus some other treatment, or versus placebo for that matter.

For example, the study by Afdhal et al in the New England Journal considered sofosbuvir in combination with another new drug from Gilead, ledipasvir.  However, this was an open label (unblinded) study that compared patients given different durations of treatment with both sofosbuvir and ledipasvir with or without ribavirin.  It did NOT directly compare sofosbuvir (or ledipasvir) to any other treatment, or to placebo.(3) Another study by Sulkowski et al in the NEJM focused on previously treated patients.

Cited by Afdhal et al were studies by Osinusi et al (an unblinded study that compared different doses of ribavirin given to patients also who all received sofosbuvir)(4); Lawitz et al (an unblinded study that compared different durations of treatment for both sofosbuvir and ledipavir with or without ribavirin)(5); and Gane et al (a study that compared various drugs again all added to sofosbuvir)(6).  Again, NO study compared sofosbuvir to any other treatment, or to placebo.

Thus so far I have been unable to find any additional studies that compared Sovaldi (sofosbuvir) to any other potential treatments for hepatitis C.  Without a comparison to another treatment, it is not possible to tell whether the high rates of patients whose blood tests show no detectable hepatitis C virus in the short-term after treatment were due to the excellence of the treatment, or due to selection of patients with particularly good prognoses.   Recall that there is evidence that some patients with hepatitis C spontaneously clear their blood of hepatitis C, and become spontaneously cured (see our previous discussion, and particularly the study by Seeff et al[7]). 

There is evidence that the studies above may have been designed to only include patients with the best prognoses.  They generally were designed with very complicated, restrictive and subjective inclusion and exclusion criteria that could have eliminated all but the healthiest patients.  For example, according to the study protocol found in the supplementary material for the study by Afdhal et al, patients with "clinically-significant illness (other than HCV) or any other major medical disorder that may interfere with subject treatment, assessment or compliance with the protocol" would have been excluded, as would have "subjects currently under evaluation for a potentially clinically-significant illness (other than HCV)...."

So, while many articles about sofosbuvir (Sovaldi) have appeared recently, in my humble opinion they do not provide strong clear evidence that this drug is extremely effective and remarkably safe, i.e., that the drug is a "triumph" that will cure nearly everyone without risk or harm to them, and therefore warrants a princely price.


While the buzz about Sovaldi intensifies, there still seems to be little clear evidence to justify extravagant claims made for it, or the extravagant price demanded for it.  Yet there seems to be a bandwagon building to somehow pay whatever it takes to provide the "triumph" of medical science to every patient who needs it.  At the current price, that should make the CEO of Gilead even richer (increases in the stock price after the $84,000/ course drug price was announced already turned him into a billionaire on paper, look here.)  As noted above, it is not clear what benefits such a costly policy would bring to patients, and at what risks.

We have discussed how our current policy of letting corporations sponsor and run the clinical research meant to assess their own products has lead to widespread manipulation of research to make their products look better, and sometimes suppression of research that cannot be manipulated into making their products look good.  This should make health care professionals, policy makers, and the public extremely skeptical of every over-hyped new "innovation."  So far, there seems to be no public skepticism about this latest, and seemingly most expensive pharmaceutical "triumph."

The US health care system, and to some extent the health care systems in most developed countries are experiencing ever higher prices.  Much of these prices' effects seem to drive up remuneration of health care executives and managers, but whether they buy better care or outcomes for most people is not so clear.  The ever rising prices seem to be buoyed by endless enthusiasm for new tests, treatments, programs, and unbridled faith in the benefits of all new technology.  It is not clear how much of that is due to evidence, how much is based on ideology and unquestioning faith in technological progress, and how much is driven by marketing and public relations, which not always may be entirely honest and free from deception.

Evidence-based medicine rigorously applied suggests that individual health care and health policy decisions should be driven by the best available evidence, mostly from clinical research, about the benefits and harms of tests, treatments, programs, and so on, in the context of what outcomes matter to patients.  The skepticism EBM should engender could lead to health care that is more about patients and their outcomes, and less about ideology, hype, and hucksterism. If only such skepticism were easier to find.

1.  Chung RT, Baumert TF.  Curing chronic hepatitis C - the arc of medical triumph.  N Engl J Med 2014:  Link here.
2.   Lawitz E, Mangia A, Wyles D et al.  Sofosbuvir for previously untreated chronic hepatitis C infection.  N Engl J Med 2013; 368: 1878-1887.  Link here.
3.  Afdhal N et al.  Ledipasvir and sofosbuvir for untreated HCV genotype 1 infection.  N Engl J Med 2014: Link here.
4.  Orinusi A Meissner EG, Lee YJ et al. Sofosbuvir and ribavirin for hepatitis C genotype 1 in patients with unfavorable treatment characteristics: a randomized clinical trial.  JAMA 2013; 310: 804.  Link here.
5.  Lawitz E, Poordad FF, Hyland RH et al.  Sofosbuvir and ledipasvir fixed-dose combination with and without ribavirin in treatment-naive and previously treated patients with genotype 1 hepatitis C virus infection (LONESTAR): an open-label, randomised phase 2 trial.  Lancet 2013; 383: 515.  Link here.
6.  Gane EJ, Stedman CA, Hyland RH et al.  Efficacy of nucleotide polymerase inhibitor sofosbuvir plus the NS5A inhibitor ledipasvir or the NS5B non-nucleoside inhibitor GS-9669 against HCV genotype 1 infection.  Gastroenterology 2013.  Link here.
7.  Seeff LB, Hollilnger B, Alter HJ et al.  Long-term mortality and morbidity of transfusion- associated non-A, non-B, and type C hepatitis: a National Heart Lung and Blood Institute Collaborative Study.  Hepatology 2001; 33: 455-463.  Link here.

ADDENDUM (21 April, 2014) - Also see the post on "breakthrough cures?" by Dr Howard Brody on the Hooked: Ethics, Medicine and Pharma blog


Anonymous said...

Thanks so much, you are asking the right questions!

Anonymous said...

POSITRON results were published in NEJM 4/23/2013 (dated 5/16/2013). The study was published simultaneously with the FUSION study.
POSITRON/FUSION studied patients with genotype 2 and 3 Hep C, where response rates using Sovaldi seem to be about the same in other studies (all conducted by Gilead) as found in POSITRON/FUSION.
The best results come in genotype 1 patients at > 90% SVR12. Roughly 75% of U.S. Hep C patients have genotype 1 disease, so for many patients the > 90% SVR12 rate may apply.

Roy M. Poses MD said...

Anonymous of 18 June, 2014,

Yes, and your point is?

How valid is POSITRON? Does it have major methodologic biases? What about outcomes other than SVR12? What about adverse events? Who were the patients and how generalizable are results.

Note that I am now hardly the only one raising concerns that in general the evidence from clinical research supporting Sovaldi is not clear and not convincing. See: