Bioequivalence and Bioavailability Forum

Dear Detlew!

❝ ❝ ... ‘partial Cmax’ (hey – a new term!)

❝ Excerpt from my next report: "We have evaluated the partial Cmax according to Schütz ..."

Well I think it’s a stupid term. Maybe roman indices would work better (C_max,I and C_max,II).

❝ ❝ It might well be that a couple of subjects show a t_max,0-4 at 4 h despite there was a decrease from an earlier peak or there is no clearly defined peak in one of the two sections at all.

❝ In the extremal case this would mean that the Cmax values are the same in both phases! See the red curve of your example picture.

That’s why I said:

❝ ❝ […] the maximum concentration (C≤4 and C>4) ignoring the actual shape …

In my example (red line) the first one is at 4 h and the second one at 4.5 h. But it’s counterintuitive to call a concentration in the decreasing part of a profile C_max.

❝ Your cut-off point T=4 h is driven by clinical reasoning I think, like the arguments in the linked presentation?

Exactly. All references above show data in fed state.

❝ I worry that the sentence in the Q&A calls for a cut point that is solely driven by the separation of the two phases in the concentration time courses. This would lead in the example courses you have given to T somewhere between 2-3h.

MR MPH is a nice model because variability is very low (both intrinsic and analytical). Formulations contain IR and SR parts (50/50 or 30/70). Due to variability in gastric emptying the second peak is more or less pronounced. Sometimes the first peak is missing (gradual increase – just a ‘shoulder’), the second peak is ± shifted, etc. I did a lot of PK modeling as well (both classical and PopPK), but finding a data-driven cut-off is simply impossible (+ against the Q&A). Even eyeballing fails in many cases (does the blue curve have three peaks or just two – the peak at 1.5 h being only noise?)…

❝ But eventually I'm too anxiety psychotic regarding taking guidances literally (me sitting in front of a snake).

Yeah. Though I’m evaluating both C_max-values since 1999 – only as a descriptive metric. The initial ramp (rapid onset) was shown to be important for efficacy, so AUC_0-4 makes sense. Some authors discussed a potential development of acute tolerance (tachyphylaxis) – but this was never proven in a clinical study. These authors advocated the second peak to be higher than the first one, which also mimics IR MPH given 4 hours apart. But receptors don’t give a damn on single concentrations. AUC_4-t is therefore suitable as well. BTW, AFAIK the hybrid authorisations were granted in clinical studies: MR vs. 2×IR (BE of AUC) vs. placebo (superiority of MR and IR). All of them were shown to be safe and efficacious despite differences in profiles. Of course they are not interchangeable and a generic application is not that easy.

Zolpidem is even more tricky. You get an idea looking at the mean curves in Kam’s slides. I once saw some spaghetti plots – there’s never a ‘first peak’. Requiring C_max,1.5 is very strange.