I think that most physicians and investors were aware that the analysis of Dendreon's (NASDAQ:DNDN) Provenge in the Phase III IMPACT Trials understated its effect on survival because of the use of frozen Provenge (Frovenge) in the control group. In a new scientific paper, the authors concluded that after adjusting for this factor, the median overall survival for Provenge is probably about 7.8 months rather than the 4.1 months that is shown in the prescribing information. Overall survival of 4.1 months is quite good in a metastatic setting, but 7.8 months is very good.
Because this analysis is retrospective, I think that there is no chance that this 7.8 months will be included in the product labeling. However, I think that it could have a very positive impact on physician prescribing and attitudes. This should be an important incentive to make sure that their patients are exposed to Provenge at some point during the treatment cycle.
There is a temptation to compare this 7.8 months median survival to the 3.8 months improvement in median overall survival seen with Zytiga and 4.8 months seen with MDV 3100 in post chemotherapy prostate cancer patients. However, the patient population is different as Provenge was studied in the pre-chemotherapy setting. We await results of Johnson & Johnson's (NYSE:JNJ) Zytiga and Medivation's (NASDAQ:MDVN) MDV-3100 in their phase III trials in pre-chemotherapy patients. However, I would be surprised if they reach 7.8 months for median overall survival.
To some extent, this is old news. However, I think that the analysis will be of significant benefit in the marketing of Provenge. I continue to recommend purchase of Dendreon. My basic hypothesis is that most metastatic prostate cancer patients are likely to be given Provenge, Zytiga and MDV 3100 at some point in the treatment of their disease. Because of its slow onset and safety profile, Provenge is more likely to be used if the disease is not rapidly progressing while Zytiga would be given if the disease is rapidly processing. MDV 3100 is likely to be used in combination with Provenge or Zytiga in either of these settings.
Background on Use of Frovenge in Pivotal Provenge Phase III IMPACT Trials
The initial IMPACT Trials that were the basis of approval for Provenge enrolled 512 men of whom roughly two thirds were given Provenge and one third were given placebo. The original design of the trials proscribed the primary endpoint as progression free survival or in other words the time it takes for the tumor to begin to grow. This proved to be a problem for Provenge as analysis showed that in many patients the tumor resumed growth even though overall survival was increased. My hypothesis is that Provenge may have little impact on shrinking the large tumor masses that are measured to determine if the tumor is growing, but does have a positive effect on metatasteses that are more likely to cause death.
The initial design of the IMPACT trials allowed control patients to receive cryopreserved Provenge, often referred to as Frovenge, once their tumors began to grow. However, after the trials were in process, Dendreon concluded that Provenge could increase overall survival even if the tumors resumed growth and with the approval of the FDA changed the endpoint of the IMPACT trials to overall survival.
Unfortunately, a good portion of the control group was given Frovenge. Altogether, 109/171 (64%) of the control group received Frovenge. In the final analysis of the trial, Provenge was compared to the control group which was comprised of 165 Frovenge patients and 84 true placebo patients. By this analysis, Provenge had a median survival advantage of 4.1 months. To a significant extent Provenge was being compared to Provenge (Frovenge) in the final data analysis as Frovenge improved length of survival in the control group. Previous papers have concluded that median survival for Frovenge patients was 20.0 months as compared to 9.8 months in the control group, which is a 10.2 month improvement in median survival.
New Paper Does Additional Analysis of the Effect of Frovenge in the IMPACT Trials Data Anaysis
On January 31, 2012, Dendreon announced that a paper called "An Analysis to Quantify the Overall Survival Benefit of Sipuleucel-T Accounting for the Crossover in the Control Arm of the IMPACT Study," will be presented at the American Society of Clinical Oncology 2012 Genitourinary Cancers Symposium, taking place February 2-4, 2012 in San Francisco, California. This paper used a rank-preserving structural failure time model, to quantify how treatment with Frovenge might have impacted the overall survival of the Phase III IMPACT trial by adjusting for the positive treatment effect of Frovenge in the control arm. I must be honest and say that I don't understand how this was done.
This paper pointed out that the FDA prescribing information states that median survival is 4.1 months, but does not take into account the Frovenge effect. Based on their model and assuming that Provenge and Frovenge produce the same therapeutic effect, the authors concluded that the real median overall survival advantage of Provenge is a much greater 7.8 months relative to control..
Disclosure: I am long DNDN.