Bioequivalence and Bioavailability Forum

Hi ElMaestro,

❝ You are on the right track. I think this was largely an answer to a question I didn't ask, and not answer to the question I asked.

That’s the “thread-drift” phenomenon common to technical fora. There is always a smart ass giving explanations not asked for. If it happens often, possibly a troll.

❝ I believe I once again perhaps did not explain myself properly enough.

No, you did.

❝ You can rearrange that also in another outright horrible fashion:

❝ (AUC_T/D_T) / (AUC_R/D_R) = F_T/F_R

❝ Ouch. Potency correction is the wonderful tool I can use to prove that 5x250 mg Schützomycin tablets are bioequivalent to 1x250 mg Schützomycin .

Exactly. Same molar dose (in the BE definition) refers to the content stated in the label. Measuring actual content and giving the results already in the protocol is mandatory. The EMA and others decided that

• If Δ ≤5% ⇒ assume that D_T=D_R (don’t adjust).
  
• If Δ >5% ⇒ adjust for measured content (if no closer batches can be found).

Still the question is: Where to draw the line, i.e., what is the upper limit for an adjustment? I would say that a Δ of 400% is way too high (even for Schützomycin).
Of course, the GL uses an elastic clause (in exceptional cases […] content correction could be accepted.)

Think about my [msg]example above[/msg]. Release spec’s ±10%. The test is perfect (assayed content 100%).
You send a guy/doll on a trip to the member states of the EEA to buy batches of the reference. All are lousy (just passed the spec’s). Add assay error and you end up with Δ 12%. Nice justification (you tried really hard – not only three pharmacies in the neighborhood), stated in the protocol. Study passes after correction, but could be accepted.
You can’t be sure.