Bioequivalence and Bioavailability Forum

Hi Nasty,

» […] information about the records where C_max is the first point (as well as records where AUC_0-t/AUC_0-∞<0.8 and records where pre-dose sample > 5% C_max). Could the regulators ban the results of the BE study with such data? Did you ever faced with this in your practice? What was the percentage of subjects with such data (for example for 0.8 it is common to suppose more than 20%)? If there is only one or two subjects with C_max in the first point should we perform additional analysis excluding them in order to show that this doesn't affect the results if it was not stated in the protocol?

OK, let’s have a look at the guideline:

Sampling times

• The sampling schedule should include frequent sampling around predicted t_max to provide a reliable estimate of peak exposure. In particular, the sampling schedule should be planned to avoid C_max being the first point of a concentration time curve.
  
• The sampling schedule should also cover the plasma concentration time curve long enough to provide a reliable estimate of the extent of exposure which is achieved if AUC_(0-t) covers at least 80% of AUC_(0-∞).

Reasons for exclusion

• Subjects with non-zero baseline concentrations > 5% of C_max. Such data should be excluded from bioequivalence calculation (see carry-over effects below).
  The above can, for immediate release formulations, be the result of […] an insufficient wash-out period, […] and should as far as possible be avoided […] by designing the study with a sufficient wash-out period. The samples from subjects excluded from the statistical analysis should still be assayed and the results listed.
  If there are any subjects for whom the pre-dose concentration is greater than 5 percent of the C_max value for the subject in that period, the statistical analysis should be performed with the data from that subject for that period excluded. In a 2-period trial this will result in the subject being removed from the analysis. The trial will no longer be considered acceptable if these exclusions result in fewer than 12 subjects being evaluable.
  
• AUC_(0-t) should cover at least 80% of AUC_(0-∞). Subjects should not be excluded from the statistical analysis if AUC_(0-t) covers less than 80% of AUC_(0-∞), but if the percentage is less than 80% in more than 20% of the observations then the validity of the study may need to be discussed.

My personal summary:

• First-point C_max “should be avoided”. Though not elaborated in the section about exclusion, I would state it in the protocol. See also what Ohlbe wrote above. I would gather that excluding >20% of subjects will lead to troubles (similar to high residual AUC).
  Note that the FDA is more generous: If you have data from a pilot study showing that the sampling schedule in the pivotal was adequate, you don’t have to worry (just bad luck, data acceptable).
  
• Subjects with carry-over can be excluded (hopefully ≥12 left). No discussion needed if procedure stated in the protocol. Do it better next time.
  
• High residual AUC. Clearly stated in the GL, though I would be happy to discuss “the validity of the study”. For IR products sampling far beyond 2×t_max is scientifically questionable.

If nothing is stated in the protocol, assessors love a sensitivity analysis. However, if the full data set fails and the one after exclusion passes, don’t hope for too much. Then it depends on the “whole body of evidence”. If this is one study of an entire set, chances for approval are still there. If not, very unlikely.