Bioequivalence and Bioavailability Forum

Hi Dan & lizhao,

❝ ❝ If a PD endpoint is an endogenous hormone in human body. The PD metric is the change from baseline after taking the medicine. And it is also the PD metric is a concentration, so I should log-transform it.

Pharmacokinetics may be simply
defined as what the body does
to the drug, as opposed to
pharmacodynamics which may be
defined as what the drug does to
the body.*

With hormones you may stir up a hornets’ nest. Depending on the hormone you may see an increase, decrease, or even a nasty feed-back loop (i.e., steady state is far from what you would predict from a single dose). Can you be more specific? Which hormone? Population? Single/multiple dose?

❝ You determine the baseline for each individual separately and you get a value X. This value is subtracted from each concentration of your PD-profile and then you log-transform the data ln(C1-X, C2-X, C3-X.....) and perform the ANOVA. So I do not understand what you mean by ln(post_treatment-baseline) as bioequivalence metric.

Subtraction (if we have an increase, of course) is only the first step. I would not log-transform the corrected concentrations. Simply calculate the C_max / AUC and then log-transform. Important: For AUC don’t use AUC_t but include the last triangle if zero is the result of two values >LLOQ.

---

• Benet LZ
  Pharmacokinetics: Basic Principles and Its Use as a Tool in Drug Metabolism
  In: Horning MG, Mitchell J (eds) Drug Metabolism and Drug Toxicity
  New York, Raven Press (1984) p199
  ISBN 0-89004-997-1