Dr. Maha Hussain (left) is calling for an open access program for Dendreon’s Provenge prostate cancer vaccine (Sipuleucel-T) pending completion of an ongoing, 500 man clinical trial, She made this suggestion March 29 during the FDA Advisory Committee hearing after voting “No” on the vaccine’s efficacy. She followed up with a letter to the FDA reiterating her “No” position and again calling for an open access program.

Dr. Hussain is the second oncologist among 4 members of the panel who voted “No” to the efficacy of the drug who has followed up with a letter to the FDA.

A letter by Dr. Howard Scher to the FDA appeared earlier this month and was first printed on April 17, in the The Cancer Letter (see Scher to FDA about Provenge hearing). Dr. Hussain’s letter to the FDA also first appeared in The Cancer Letter, on April 27. Dr. Hussain writes:

It is with concern and professional obligation that I write to you as a member of the FDA’s Advisory Committee that recently reviewed Sipuleucel-T on March 29, 2007. My concerns relate to medical, scientific and procedural aspects of the meeting and the precedence that will be set for future reviews.

By way of introduction, I am an academic medical oncologist with expertise in GU oncology, extensive clinical trials experience and have been the PI of several NCI sponsored multi-center trials including randomized phase II and III trials. Currently, I am the PI of a Prostate Cancer Clinical Trials grant funded by the Department of Defense that focuses on phase I and II trials in prostate cancer. My experience also includes co-chairing the prostate cancer subcommittee
of SWOG overseeing development of national trials for advanced prostate cancer for the past 13 years. I have served as an ad hoc FDA consultant for several years and currently serve as a member of the Oncology Drugs Advisory Committee. I was a member of and chaired the ODAC special session on prostate cancer endpoints, March 3, 2005, and have been actively involved in the development of endpoints for this disease, a summary of which was recently presented at the 2007 Prostate Cancer ASCO meeting.

I was one of the four members who voted “No” to whether the submitted data on Sipuleucel-T established “efficacy” or “demonstrated substantial evidence of benefit” in the intended population at the recent advisory committee meeting.

From the medical and scientific aspects the recommendations for approval that may be inferred
from the vote are based on data that can only be characterized at best as “suggestive” of possible benefit. As the discussant for Q5 regarding the persuasiveness of the efficacy evidence my comments are public record but to summarize my conclusion was that the data presented is not conclusive. The context here is not “is the treatment promising” or “does it open the door for more immunotherapy research,” the context here is “is the treatment effective and are the results solid” such that this therapy should be offered as “The Standard of Care” by physicians to thousands of patients with the confidence that their recommendations truly serves the best interest of the patients. First of all the lead trial (study 1) was a small trial by any standard with 127 patients in total of whom only 82 were treated with Sipuleucel-T. The study was not powered for survival nor was survival an end point. A post hoc analysis indicated a significant survival difference but there were no significant differences between the
Sipuleucel-T and placebo group with regard to any of the disease manifestations including PSA, time to disease progression (primary endpoint) or pain. This coupled with a clear imbalance in the arms with the control arm having more patients with bone and soft tissue disease thus potentially bulkier disease, more patients with higher Gleason scores, more % with prior chemotherapy and questions regarding the nature of the agent administered as the control (please see comments below), the small sample size, the fact that survival was not powered for and is a post hoc analysis could lead to a plausible conclusion that the observed survival difference may be related to other factors or chance alone and not to the treatment effect. Please contrast this data with the two phase III trials (TAX-327 with 997 patients, SWOG -9916 with 770 patients) that led to the approval of docetaxel. Both of these trials had very consistent results across them and conclusively demonstrated a survival advantage with notable effects on other disease manifestations.

The sponsor presented a second “supportive trial” which was also a small prematurely terminated trial which showed about a 3 month difference in survival which was not statistically significant. The trial results were especially problematic since both arms had a poorer survival (15.7 and 19.0 months) than expected for asymptomatic patients and worse than the survival observed in study 1. This occurred despite similar eligibility criteria to study 1. Furthermore, even the Sipuleucel-T treated patients had a median survival of (19 months) which is comparable to the “asymptomatic” subgroup of men treated on the mitoxantrone arm of the Tax327 trial (19.8 months, Berhold et al, ASCO Prostate Cancer Symposium 2007). Please note that mitoxantrone is not considered the standard first line therapy in general or for symptomatic patients.

This clearly raises concern regarding the true efficacy of the agent and reproducibility and reliability of the data hence the application in the intended population at large. Furthermore, considering that the “placebo” treated patients had an unexpected poor survival of 15.7 months, which is worse than the median survival of patients on mitoxantrone arm of the TAX-327 of 16.4 months (NEJM 04) which also included symptomatic patients, raising questions regarding a negative effect from the placebo thus leading to an apparent survival benefit. Issues regarding CVAs, particularly in the intended population, are also of concern without mature toxicity data and in the context of inconclusive efficacy data.

As you know, a definitive trial is in progress and is within 100 patients of achieving target accrual. This trial will lead to definitive answers as to the true efficacy and safety of this agent. These questions will never be answered if the decision regarding this agent is not deferred at this time until all patients are accrued and data are mature, for obvious reasons.

From the scientific and procedural aspects, in general, it would seem that at the end of the day what should determine a positive verdict in any therapeutic trial is the strength of the evidence as critically reviewed by an Advisory Committee with the proper expertise in the context at hand (ODAC in the case of a therapeutic cancer trial), with clear guidance on the questions posed to the committee within the framework of the regulatory guidelines and requirements of the FDA for approval. This needs to be coupled with an atmosphere that is conducive to an objective discussion and vote.

Another concern, based on this case, is the appearance of discordance in the burden of proof
required for regulatory approval between CBER and CDER. In the meeting regarding endpoints in 2005, ODAC reaffirmed the importance of powering trials for endpoints that measure true clinical benefit. But fundamentally here this particular agent did not even meet criteria for its primary endpoint.

In conclusion, as physicians we owe it to our patients to maintain the highest scientific standards and rigor. We owe them our objectivity and the assurance that when we make recommendations for treatment that we are basing our decisions on strong conclusive data. We need your help to ensure maintaining this high standard.