Modelling plasma data across 1 or 2 doses?

Phoenix Certara Forums Modelling and Simulation
1

Dear colleagues,

In a clinical study 2 equal doses of compound X (formulated as mini drink, administered before or after a meal; inclusion and exclusion criteria were applied) were given to 10 healthy females with a 6 hour interval.

Thereafter we observed classic plasma oral dosing curves (30 plasma concentrations per 2 doses/peaks). In most cases -not always- concentrations returned to baseline before the second dosis was administered. In addition, Cmax was different across those 2 doses, sometimes > 20%. Finally, the variety across those 2 peaks in terms of Tmax was small compared to that of Cmax.

My question is: should I estimate PK parameters (chose the model..) based on plasma data from 1 or 2 doses? (Based on 1 dosis yields relative small residual errors, but doesn’t capture variability that’s there etc.)

Thanks in advance for your suggestions / thoughts.

Pieter

2

I think particularly as you say sometimes the concentration has not always returned to baseline by the time the second dose is administered; that you should fit the data as a single profile of two doses.

A better questions is once you have fit your data; what dod you want to do with it? e.g. is it to better design your next study, predict future dosing regimens?

Simon.