Clarification on Phase III H&N trial
Investor Question to CEO
Dear Dr. Thompson,
Just a quick question:
Suppose that
1. The existing Head & Neck phase III returns p=.06 for mets and p=.07 for local regional.
2. A somewhat larger stage 2 trial is run in mets only, returning p=.049 for that trial.
Would the FDA require yet another trial? (IE. would the first trial count towards the required two randomized trials if it didn't return p=.05 or better?)
____________________
He responded as follows:
------------
The original study design was to enrol a first stage (80 patients), and then perform a predictive probability of success for a second stage which was very likely to be necessary. The probability of success for this analysis was 0.7 or greater which loosely translates into a p of 0.15 to 0.20. If we had equalled or
surpassed the 0.7 then we would have proceeded to the second stage. The only way we would not have proceeded, on success, to the second stage ( for the US) was to have a probability of success of .98 or .99, which gives a very low p value.
We doubled up on patient numbers last year to account for the two groups (loco regional and mets only) and will do an analysis of p value for each group. If the p value is 0.15 to 0.2 or better for each or either group, we will proceed to a second stage as before. The size of the second stage will be about the same as before, about 150 to 200 per patient group. If the p value is extra low ( unlikely with such small patient numbers ) as before we will potentially file with the single stage data.
Please post this if you choose,
Thanks,
Brad
Sent from AOL Mobile Mail
