The total enrolment was 1513, but the population for the primary analysis was 1239. 274 patients randomized six months before the hold were excluded by a prospectively agreed to protocol adjustment to account for the confounding influence of the hold.
"Sensitivity analyses revealed that there was no OS benefit in pts randomized 6 m before the hold (HR 1.09, CI 0.75-1.56, p=0.663). "
The HR of 1.09 indicates that for this group of 274, BLP 25 underperformed placebo.
The data was broken down into 2 subgroups, the concurrent chemo/radio group and the sequential group. Both metaanalysis and a ph 3 trial have demonstrated concurrent therapy is superior to sequential therapy. See "Sequential vs Concurrent Chemoradiation for Stage III Non-small Cell Lung Cancer."
In the sequential subgroup (n=433) BLP 25 again underperformed placebo. "median OS with sequential CRT was 19.4 m (L-BLP25) vs. 24.6 m (PBO; HR 1.12, 95% CI 0.87-1.44, p=0.38;"
In the hopeful concurrent subgroup the following result was seen. " pts with concurrent CRT (n=806) had median OS of 30.8 m (L-BLP25) vs. 20.6 m (PBO; HR 0.78, 95% CI 0.64-0.95, p=0.016)"
Here is the part that is troubling. As expected, the BLP concurrent did better than the BLP sequential, 30.8 months vs 19.4 months. Unexpectedly, the placebo concurrent arm did worse than the placebo sequential arm, 20.6 months vs 24.6 months. This is an unexpected and counterintuitive result indicating that one or both of these statistics is not a representative sampling.
"Sensitivity analyses showed a smaller treatment effect due to the clinical hold, suggesting that longer uninterrupted treatment with L-BLP25 is required."
The above is a bit confusing. The solution to at least partly account for the hold allowed Merck to exclude the results of 274 patients for whom survival was tilted in favor of placebo, HR=1.09.
When I first read the abstract, I focused on the positive implications of the subgroup result. Now I'm not so sure.
If the trial failed on the basis of this data - HR 0.88, 95% CI 0.75-1.03, p=0.123 which is interpreted as "no difference', then wouldn't HR and p values be more relevant when looking at seq BLP25 arm, concurrent Plac or seq Plac?? The HR and, more importantly, p values say exactly the same thing for these 3 groups.... "no difference" among these arms.
I don't think that the potential for efficacy is up for debate. This, after all, is an investment board and ONTY needs this to be approved for share price to go up. Sorry to be callous. No disrespect to those who are sick, but the numbers aren't looking like they are there for approval at this juncture and it's going to take more time. I've got a bunch of shares not worth too much and it does no one any good to grasp at straws. Earliest possible approval of this drug would likely be late 2014, but more realistically, 2016-2017 timeframe as FDA will want more data. This is just the reality of the situation.