Looks like there is a protocol published for a Cochrane Review - http://onlinelibrary.wiley.com/doi/10.1002/14651858.CD010587/pdf . Interestingly they are planning to include DAA vs placebo trials. Not sure that's a good idea given we know IFN based regimens are somewhat effective. I can't see a timeline for completion of this review but may not be looking in the right place?

Best wishes,

Tom

On 16 April 2014 16:06, Poses, Roy <[log in to unmask]> wrote:

Thanks for all the comments, only some of which are below.

I have only seen extreme optimism about the drugs in the media and the journal articles I have managed to read.  I have seen NO skepticism at all about the drugs or the evidence supporting them.  The only concerns are about the price.  But again, if the drugs offer miraculous cures without any serious downside, who would really contest the price ;-) ?

I am neither an infectious disease nor a liver specialist, so am not fully acquainted with all the details, especially about particular virus types. So I am not the ideal person to take the lead on this particular issue.\

Yet  I am concerned that this horse is already out of the barn, and there already is a consensus that the drugs are near miraculous, never mind the evidence.

The Cochrane and EBH groups may include the only people who might bring some evidence-based skepticism to address this issue, so my hope was to alert you all, and hope some of you will have a closer look before the horse has run too far. 

On Wed, Apr 16, 2014 at 8:48 AM, Tom Yates <[log in to unmask]> wrote:

In the Afdahl study, with the 90+% SVR rates, they were not looking at Sofosbuvir alone, as they did in the head to head trial, but at various Ledipasvir–Sofosbuvir based regimens in patients that had previously failed IFN/RBV. I agree patient selection is also likely to be important in explaining these results but I don't know enough about HCV to know whether this might explain the very high SVR rates described in the latter study. I am yet to speak to an infectious disease specialist who isn't excited by these results.

Great discussion,

Tom 

On 16 April 2014 13:29, Poses, Roy <[log in to unmask]> wrote:

But am I the only one who finds it very strange that in the only comparison of sofosbuvir with another drug (peg-interferon) the SVR for sofosbuvir (and the comparator) was about 67% 12 weeks after therapy, but in all the other studies of sofosbuvir the SVR was much higher?

I note that in every study I have looked at, the inclusion and exclusion criteria were extremely extensive, suggesting that the studies were done on the healthiest of the healthy patients (still with HCV).  My concern is that the very high SVR's in studies in which sofosbuvir was not compared to other drugs (or placebo) reflect patient prognoses rather than treatment, but that these very high SVR's are being used to promote the ridiculously high price of the drug. 

On Wed, Apr 16, 2014 at 8:01 AM, Ansari, Mohammed <[log in to unmask]> wrote:

I think in this particular case I find the FDA flexibility reasonable. Its true that we may not have a comparator that is standard of care…but we do have historical knowledge, and expectations, of change in outcomes in response to standard of care. If we have no major concerns about the study population and study validity, then we know the standard of care counterfactual in terms of virologic response from valid RCTs – our response benchmark. That outcomes are soft and not hard is altogether another point.

 

From: Evidence based health (EBH) [mailto:[log in to unmask]] On Behalf Of Tom Yates
Sent: Wednesday, April 16, 2014 3:34 AM
To: [log in to unmask]
Subject: Re: Questions about new, heavily hyped treatments for hepatitis C

 

I agree and it is regrettable that it was not done right the first time. I now worry, however, that we are approaching the point where sufficient non RCT data is accumulating and public perception is such that it would be very difficult to do the proper head to head clinical trial. Is anyone in the HCV word voicing concerns about this? I hear only optimism. 

 

Best wishes,

Tom

 

On 15 April 2014 19:59, Poses, Roy <[log in to unmask]> wrote:

It seems to me that a controlled trial only tells you about the outcomes of the treatment compared to the comparator.  So if the comparator is, say, just a different dose or duration of the same drug, all that trial can say is whether there is a difference between the doses or durations. 

To prove that the drug is better than the standard of care, one must compare it to the standard of care, no?

But from what I can tell, in the case of Sovaldi/ sofosbuvir, the only comparison to the standard of care at best only proved "non-inferiority," but did not prove overall superiority (of benefits vs harms). 

 

On Tue, Apr 15, 2014 at 7:09 AM, Tom Yates <[log in to unmask]> wrote:

Hi Roy,

 

I think these are really important issues. I am told there is another head to head study vs IFN called BOSON but can't find the paper.

 

A recent editorial in the NEJM contained the following paragraph (http://www.nejm.org/doi/full/10.1056/NEJMp1400986)...

 

'...additional steps were required to get from these interferon-tethered regimens to trials of investigational DAAs in combination. Traditional study designs had, until recently, used a standard-of-care comparator for new agents, which in the case of HCV meant using an interferon-based therapy in the control group — an unappealing prospect for many patients. Buoyed by the success of combination HIV therapies and the knowledge that resistance variants in HCV were not predicted to persist as they do in HIV, the Food and Drug Administration agreed to permit phase 2 trials to use new combinations of HCV DAAs without requiring a standard-of-care comparator. This decision proved catalytic. Because of the short durations of, and rapid enrollment in, these studies, the pace of ensuing clinical drug development has been breathtaking.'

 

I find this sentiment worrying but I'm told there are real difficulties in recruiting patients to trials including IFN based regimens now. The side effects of IFN are, I understand, miserable and most patients are aware of the existence of alternatives. There is a good chance that patients may enroll in such studies hoping to be randomised to DAA and then not take their treatment if randomised to IFN, meaning trials provide good evidence on effectiveness but poor data on relative efficacy. 

 

What trials do you think we should be calling for? The Afdahl trial you reference is interesting, recruiting people who had previously failed IFN based therapy. Do we still have equipoise? Are virological outcomes ok? I would have thought that any licensing should be conditional on ongoing follow up (presumably for many years) to get clinical outcomes.

 

It would be nice to have some non industry trials. There is a notable absence of trials in 'hard to reach' groups - e.g. injecting drug users, people with alcohol dependence, etc. There may be an opening there for rigorous independently conducted studies with highly relevant outcomes. What should such studies look like?

 

All the best,

Tom

 

On 13 April 2014 20:00, Poses, Roy <[log in to unmask]> wrote:

I wondered if anyone on the list had taken a close look on the (at least published) clinical literature on sofosbuvir (Solavdi - Gilead), which is now currently being heavily promoted as practically a miracle cure for hepatitis C, and is prices at $1000/ pill, $84,000 per 12 week course of therapy in the US?  (See: http://hcrenewal.blogspot.com/2014/03/too-good-to-be-true-sovaldi-kerfuffle.html)

We reviewed the May, 2013 article on the drug that appeared in the NEJM(1) for our journal club.  Briefly, this was a non-inferiority trial of sofosbuvir plus ribavirin vs peg-interferon vs ribavirin.  The sustained viral response rates for both were numerically the same at 12 weeks after therapy finished (67%).  The patient population had multiple inclusion and exclusion criteria, buried in an appendix, that suggested a lack of generalizability.  The sofosbuvir treated patients had lower rates of  bothersome adverse effects than the peg-interferon treated patients, but these rates were hardly negligible.  Also, by perusing the appendix, one could find that the severe adverse effect plus death rate was higher in the sofosbuvir group than in the peg-interferon group, but the difference was not statistically tested.  Of course, the study did not follow patients long-term (for a very long-term disease) and did not assess any clinical outcomes.  Most particularly, it did not assess whether treatment prevented any of  the know complications of hepatitis C infection. 

In short, the article did not suggest that the drug is a miracle cure, or even that it cures most patients or has decisively less and less important side effects than peg-interferon.

However, as we discussed here, although there are lots of concerns about the prices in the US, almost no one questioned that the drug is extremely good.

Recently there has been even more publicity about the drug, apparently based in part on new articles appearing in the New England Journal of Medicine.  I took a quick look at the one article on previously untreated patients.(2)  This concerned a combination of sofosbuvir and ledipasvir (also made by Gilead).  Patients were randomized into 4 groups.  However, all four groups received the combination of sofosbuvir and ledipasvir for 12 weeks.  They differed only in whether they got an additional 12 weeks of the same combination, and in whether they did or did not also get ribavirin.  Thus, I cannot see how the trial could have directly assessed the outcomes of treatment of sofosbuvir plus ledipsvir versus any alternative treatment (or even placebo).  It seemed to be virtually a case series of patients treated with that drug.  Because the appendix of the study also suggested that there were multiple inclusion and exclusion criteria, the apparently good results for sustained virologic response may simply have reflected a patient group with a very good prognosis.  I have not yet had a chance to look at the adverse effect data (but it would be hard to interpret without a true control group.)   Of course, there was no long-term follow-up and no assessment of clinical outcomes.

So I am even more concerned that the hype about sofosbuvir and now other new antiviral treatments of hepatitis C is hardly based on clear evidence, and whether we are already heading for our next Tamiflu kind of case.

Has anyone else had a chance to look over these or related trials and is willing to render an opinion? 

References.

1.  Lawitz E et al.  NEJM 2013.  http://www.nejm.org/doi/full/10.1056/NEJMoa1214853

2.  Afdahl NA et al.  NEJM 2014.  http://www.nejm.org/doi/full/10.1056/NEJMoa1402454

--