Bioequivalence and Bioavailability Forum

Hi,

I have a very interesting question that I have not come across previously. A study drug in question must pass bioequivalence criteria for two analytes. The ISV for analyte 1 is much higher than for analyte 2. The sample size planned would be about 200 subjects for 80% power to determine bioequivalence for analyte 1. However, for analyte 2, this is overkill, as only about 50 subjects are needed for 80% power. The Sponsor wants to save some money in analytical costs by analyzing samples for all completed subjects for analyte 1, but only 50 of the 200 subjects for analyte 2.

How would this go over for a European submission? My first thoughts are: Doesn't this introduce bias in the study? Which 50 would be selected? To be fair a separate randomization scheme can be generated once the study is completed but before the samples are analyzed, so that a balanced number of subjects can be selected? Or, for instance, could the first 50 be analyzed and the remaining 150 subjects be used as alternate subjects?

Does anyone have any experience in planning a study where the sample size calculation is different for two analytes and so different numbers of subjects are used for each analyte to obtain 80% power?

Many thanks in advance,

David Dubins