Bioequivalence and Bioavailability Forum

Hi Pash,

❝ In BE study of modified release drug product for EMEA submission …

Nitpicking: EMEA changed its name to EMA back in 2009.

❝ Is there any regulatory concern if we have not included the drop out subject(s) due to ambulatory missing samples in the statastical analysis.

I agree with ElMaestro. Furthermore keep in mind that

Guidelines
are guidelines
are guidelines.   Henrike Potthast (BfArM, member of the CHMP’s PKWP)

Though the GL speaks about excluding subjects only, you should not take that literally. My procedure:

• If no reliable estimate of λ_z possible → keep C_max and AUC_0–t (exclude AUC_0–∞ of the subject from the comparison).
  
• If no clearly defined C_max (e.g., ~two increasing concentrations before and two to three decreasing concentrations afterwards) → exclude the subject from the comparison.

Since in most cases the variability follows the order C_max > AUC_0–∞ > AUC_0–t and likely the study is powered for the PK-metric with the highest variability the impact on power is low (exception: reference-scaling for C_max together with highly variable AUC).
An SOP is fine but I suggest to state the rules for exclusion already in the protocol. Referring in the report to an SOP (which is not available to the assessor) may rise questions and delay the approval.