| |
Clumsy FDA Antiviral Advisory Committee (Maraviroc hearing)
for both HIV & HCV in the future
|
| |
| |
a little bit of history on these panels
from Jules Levin: if you look at the data presented at the hearing it is a little difficult to understand so I think the FDA panelists mischaracterized the data plus the quality of FDA Antiviral Advisory Committee panelists is not of the high quality it used to be, the FDA is not able to place on the panel experienced HIV researchers anymore. First, at week 48 in the study the data reported 15% discontinued due to lack of viral efficacy on MVC compared to 6% on EFV. The panelists interpreted this to mean MVC was not very potent. this is nonsense. 20% of the study patients were from the USA, 80% from outside the USA, and 25% from South Africa. There was a large percent of patients who withdrew due to adverse events early in the study, within the first 4-8 weeks on EFV, BEFORE they could experience viral failure. Therefore, the higher viral failure rate on MVC. Clinicians were permitted to make their own qualitative decisions about whether the patients was experiencing a lack of viral efficacy but most of the patients failed to achieve <50 or rebounded. Most of the early withdrawals on EFV due to adverse events was however in the Northern Hemisphere, USA & Western Europe. Because another issue the panelists got stuck on was that among viral failure there was a higher rate of patients with drug resistance to the background regimen (62% vs 36%). Nonsense again, sure the rate on MVC was higher, that's because patients remained longer on MVC thus allowing them to develop failure & resistance but a lot of patients on EFV discontinued early due to AEs s could not develop viral failure. If you look at the history of MVC studies including the early monotherapy studies MVC was similar in potency as other ARTs, there is not that much difference, so it should have been obvious to the panelists there were other reasons for the data. BUT no one asked the right questions. Jim Neaton who is a statistician, and a noted statistician, is not a clinician, but he stated on the panel that the study was not like an ITT analysis that it was more like an on-study analysis so you could not interpret the data, particularly the the data that reported discontinuations due to AE & lack of viral efficacy were in different directions, nonsense again.
Discontinuations due to related AEs was 12% for EFV and 4% for MVS. So, the AE profile favored MVC and the lack of viral efficacy discontinuations favored EFV. He found this a result that made the study uninterpretable, unusual, reflected that the study was not ITT, and that therefore MVC should NOT get approval for naives, so he voted no and I think additional panelists followed his lead. Before they took the actual vote they went around the table for comments and the comments were so critical about the lack of viral efficacy & the resistance, that I thought for sure the drug woud not get approval. BUT the FDA at the table Jeff Murray & Debra Birnkrant stepped in to try to correct some misunderstandings. The panelists thought they should vote no because they did not think it appeared the drug should be on the Guidelines recommended list for firstline. Murray & Birnkrant stated that the panel was not voting on whether they thought this drug was a great choice or preferable choice for firstline, they were only voting on whether the drug ought to be available as an option for patients. Despite the consistent critical comments during the comment go-around, surprisingly the committee did in fact vote to approve. And I think this was in large part due to Murray & Birnkrant stepping in to correct misundertstandings and set the record straight. It was obvious that the FDA wants to give naive approval for MVC. Only one or 2 panelists referred to MVC's preferable safety and tolerability profile & only 1 panelist made mention of the lipid profile. In fact MVC's lipid profile is probably the best of any ART and the drug has a good tolerability profile. In the early days there were concerns about malignancies, hepatotoxicity and postural hypotension but if you click the link below to see the FDA Briefing Document the MVC data does not report anything bad about these adverse events, MVC did not have more of these AEs than EFV, in fact EFV in some cases had more of these AEs than MVC. This gets to the infections issues. Because MVC is a CCR5 antagonist there was concern early in its development that it could encourage certain types of infections like West Nile Virus, so Pfizer was required to follow patients for 5 years. In the briefing document they report not finding anything bad yet, however I have some lingering concerns about this. Another point raised by only 1 or 2 panelists was the cost of the tropism assay, which has been a barrier to using the drug. In addition it could take time to get results back from the lab in a timely fashion. For the past year Pfizer has been conducting quite a lot of research to see if you can use genotyping or the 454 sensitive assay to predict tropism as reliably compared to the current tropism assays. Several genotyping studies have been conducted by Pfizer and by academics in Vancouver and in Western Europe who have developed assays. The studies have been reported at recent conferences - IAS, Resistance Workshop, and ICAAC and the results report that genotyping does predict tropism. Perhaps this needs more work to establish but clearly the data is encouraging and provides a premise to move ahead with this. In fact in Europe they use genotyping.
In summary, the FDA panels these days do not have the quality review they used to have. The FDA feels precluded from empaneling leading researchers due to that all the leading researchers these days are affiliated with drug companies as consultants, advisors and doing CMEs, so the FDA feels there are conflicts of interest potentially. As a result we get these panels that are unable to properly evaluate and ask the right questions, this does not serve patients very well. When we start reviewing the new oral HCV drugs I'm not convinced the FDA panels will be adequately composed to give good review. Barbars McGovern and Doris Strader were on this panel and I think that's because the FDA is preparing for the HCV oral drugs and certainly I know Barbara McGovern and she is a good person for the panel, and I have passing knowledge of Strader, who is an African-American and a gestroenterologist and has published, so I think she will be good to. But unless the FDA adds more experienced and knowledgeable people from HCV the panel will not be very good. The other researchers and docs on this panel do not have adequate knowledge and experience in HCV. Despite this the oral HCV drugs I expect will pass muster and get approval unless some bad data emerges which I don't think will happen, still you want good people on the panel who can have good discussions about the HCV drugs and care and treatment as this helps public education, design of future studies, understanding adverse events and toxicities and patient management. A very important subject that will come up at the HCV drug hearings will be HCV drug resistance and without doubt you WILL in fact need a good panel to properly discuss and evaluate HCV drug resistance, which remains today a very controversial issue in HCV. By 2011 when the 2 HCV protease inhibitors will be ready for review we may have better information to improve our understanding of this very thorny issue, but today this remains very controversial. SO, to the FDA please hear this! However, lets make no mistakes about these panels. In March of 1996 the Antiviral Advisory Committee was undecided about approving the first protease inhibitors indinavir & ritonavir. I think I recall the vote on indinavir was very surprisingly close, I recall it was 8-6 to approve. But the FDA Commissioner David Kessler again was smart as he sat at the table so he could make sure these drugs got approved; as Birnkrant & Murray, Kessler as well as the Anriviral Section head David Feigal, also understood they were making no mistakes in placing too much trust in the panel to make such very important decisions. And the panel was unable to reach a decision at all about ritonavir. Kessler had to keep the panel & the Abbott people overnight in the hotel to meet privately through the evening and reconvene the public panel hearing in the morning, when they finally did in fact approve ritonavir. In addition HIV viral load was similar, I was at the hearing, I have not missed a panel hearing i HIV or HCV except for one, tipranavir. The panel did not want to approve HIV viral load, this hearing was shortly after the first 2 protease inhibitors were approved. The panel felt the data on viral load was not convincing. I went to the microphone and said the FDA just approved 3 protease inhibitors based on their affect on viral load and now you won't approve viral load, how can patients evaluate their response to these drugs without using viral load!!! The panel then changed their opinion and voted to approve viral load. The rest is history, because HIV viral load was firmly established as a reliable surrogate marker, which is crucial for care and treatment and studies today. We may face a similar concern regarding viral load in HCV as some Federal officials at the NIH and a few others I believe have expressed doubts that clearing HCV has long-term clinical benefits. Now I have not heard them say this for several years but they did say this several years ago.
Jules Levin
The documents were posted on the FDA's website here
|
|
| |
| |
|
|
|
