Ironically, data from Johnson & Johnson's (JNJ) recently halted study of its late stage treatment for prostate cancer, Zytiga, very strongly suggest that a competing treatment, Dendreon's (DNDN) Provenge, should be prescribed first.
As might be expected, since it helps a competing company's product, Johnson & Johnson in its podium presentation at the conference failed to mention the relevant data, and instead touted the positive aspects of the Zytiga study: good PSA control (Prostate Specific Antigen levels are a chemical marker for prostate cancer components in a man's blood and strong effects on disease progression, with a trend for an overall survival benefit. J & J also made no mention of the data I am writing about here in its breathless press release spinning the study results.
The company first minimized the important fact that the while the study did show strong benefit for progression free survival, the study in fact had dual primary endpoints, namely progression free survival AND overall survival, and that the overall survival endpoint was NOT yet statistically significant (and may never be). This last led famed biostatistics expert Dr. Susan Halabi, who was assigned by ASCO to analyze the study results and independently present them, to refer to the study in her presentation as a "failed trial". That's the rules in medical studies - co-primary endpoints means BOTH must be statistically significant for trial success. (She even suggested the monitoring committee was wrong to end the study early before a clear survival hit). Now crossover in the trial might prevent there from ever being a clear picture of the possible true survival benefit of Zytiga. Will Dr. Richard Pazdur, the biostatistical purity nut who heads the FDA's oncology branch and his henchmen who are assigned to statistically tear apart studies supporting drug approvals, be OK with approving a treatment whose study failed by the clear rules? Time will tell. But it is not a lock and certainly will likely involve an expert review panel pretty much like that which preceded the FDA's initial refusal to approve Provenge in 2007 in spite of very strong AND statistically significant overall survival data, because of biostatistical impurity (overall survival, while unable to be faked or fudged, was not a pre-specified endpoint).
But I digress - if Zytiga is approved pre-chemo, the big deal here is in the data Johnson & Johnson pretty much chose to bury, and perhaps hope that its hype machine could get doctors to look past or ignore: Although there was, as above, a non-statistically significant trend toward overall survival benefit in the study, the Kaplan-Meier curves clearly showed that there was NO, repeat NO, survival benefit until 18 months after the beginning of treatment. The average course of treatment was 13.8 months and it was only after 18 months that the Kaplan-Meier survival curves began to separate. Further, fully one-fifth of the men in the treatment arm of the study died before any survival benefit showed up.
So how does this help Dendreon and Provenge? Here's the deal - the Provenge trials not only clearly showed a strong, statistically significant, overall survival benefit - and one that improved over time - but in each of its three Phase III studies Provenge's survival benefit clearly showed up at 6 months after treatment and persisted to study's end.
The NCCN already recommends that Provenge be sequenced first, based upon this early separation of the Kaplan-Meier curves and the scientific logic that an immune boosting therapy should be given as soon as possible - when the immune system is the strongest and has not been compromised by other therapies. The Journal of Clinical Oncology has also similarly opined, based upon the same reasoning. And further support for the "Provenge First" thesis comes from recent data Dendreon presented at AUA and ASCO showing immune cell activity in early stage disease in men treated with Provenge, and most tellingly in an analysis of PSA quartiles in men treated in the Provenge trials showing that men who received Provenge when their PSA's were below 22 showed a whopping 13 month overall survival benefit.
And now this coup de grace: the putative competitor drug's own study shows that while Provenge starts its survival benefit after 6 months and the benefit grows, the competitor's drug provides NO statistically significant survival benefit and the trend toward a survival benefit does not even begin until past a point where 20% of the treated men are already dead.
Sequencing studies are already being conducted, but the present controlling conventional wisdom is that since Zytiga must be given with steroids that weaken the immune system and the administration of Provenge requires a "steroid wash out" waiting period if given after Zytiga, Provenge should go first.
While Provenge is administered 3 times over a one month period and then the therapy is done and its benefits can begin to kick in - and the earlier the better - Zytiga first means waiting over a year to begin Provenge benefits, and perhaps never getting those benefits at all, with 20% of men dying before getting ANY survival benefit from Zytiga, and the rest getting no survival benefit for 18 months with many men then progressing to levels of pain that will put them out of Provenge's "minimally symptomatic" label indication.
Based upon the science and upon Johnson & Johnson's own study, Provenge First is an absolute no brainer.
As you read this the Johnson & Johnson spin machine and its cooperative media friends may have had Dendreon shares dropping-but the reported demise of Dendreon and Provenge based upon the Zytiga results is totally off base.
In the pre-chemo advanced prostate cancer space the eventual victory of Zytiga is far from clear. Provenge First, if accepted by docs as it should be, would mean Provenge wins.
Disclosure: I am long DNDN.