Bioequivalence and Bioavailability Forum

Hello

I performed NCA analysis for concentration data at steady state:

(AUCtau, Ctau, and fluctuation) PK parameters were missing for some subjects who gave the last sample (at tau) earlier than scheduled.
Upon checking: WinNonlin was not able to calculate Lambda_z for those subjects (using best fit method) since each of them has only one sample after Tmax.
Accordingly, WinNonlin didn't perform extrapolation and was unable to calculate the mentioned steady state parameters.
Can we induce lambda_z calculation using two sampling points only (Tmax and Tlast)? Will this be acceptable by the authorities (specifically: the European authorities)? any other possible solutions?

Also, few subjects have missing lambda_z for another reason (concentrations were still increasing at tau).
Shall we exclude them from statistical analysis? In this case, the sampling time deviation (early collection of the sample) will be the main cause for excluding the subject.
Or shall we replace the actual time with the theoretical time (tau)?

As a note: WinNonlin has calculated AUCTtau and Ctau for subjects with missing samples at tau. Is it acceptable to exclude such subjects from statistical analysis?

Thank you

---

Edit: Category changed; see also this post #1 and #4. [Helmut]