Number of subject and population PK/PD modeling

From: "Gordi, Toufigh" Toufigh.Gordi@cvt.com Subject: [NMusers] Number of subject and population PK/PD modeling Date: Thu, January 13, 2005 8:29 pm Dear all, I am involved in planning a concentration-response study in dogs, with a maximum of 6 dogs included. Each dog will receive a minimum of 2 different (single) doses. I am pushing for a proper PK/PD evaluation of the compound, including a population approach. I believe everybody would agree that 6 is a small number. However, this is quite normal to have few larger animals in pre-clinical studies. One of the problems I face is the word "population" approach, which to many people means that one must have a large number of subjects in order to be able to apply the methodology. I don't think that's the case but have problems putting it in simple words why such approach can be taken even with a small "population" of 6 animals. Are there any publications that discuss this issue? In general, are there any situations where a "normal" PK modeling approach (e.g. using ADAPT or WinNonlin) is superior to a mixed-effect modeling approach? I am not concerned with population models being more complicated or take longer time. I am more interested to know whether the former produces better and more reliable estimates than the latter. Best regards, Toufigh Gordi

From: "Serge Guzy" GUZY@xoma.com Subject: RE: [NMusers] Number of subject and population PK/PD modeling Date: Fri, January 14, 2005 12:40 pm My experience with both Winnonlin and the population approach is that the mixed effect approach always gave me better estimates of at least the population means. Using Winnonlin and averaging the PK estimates never gave me superior average values of the main PK parameters. On the other hand, sometimes I saw problems with the estimates of the covariance components of the population variance covariance matrix when dealing with small number of patients and in a rich data environment. Correlation sometimes would be drifted to 1 with the log-likelihood being flat across a big range of correlation values. In that case, rich data analysis using Winnonlin would give me better estimates of the true correlation between the PK parameters. My experience was also that population variances were estimated equally in a rich data environment and better of course in a semi rich data environment (some patients did not have enough information to be analyzed with Winnonlin). My conclusion was that a mixed effect approach is always as good as and most of the time better than a Winnonlin approach expect for very rich data environment where we are interested to estimate the population covariance. Serge Guzy President POP_PHARM

From: "Mats Karlsson" mats.karlsson@farmbio.uu.se Subject: RE: [NMusers] Number of subject and population PK/PD modeling Date: Mon, January 17, 2005 4:33 am Dear Toufigh, We investigated the properties of non-linear mixed effects modeling and standard-two-stage for some subject-sparse (n=8), data-rich, situations and found that MEM was better, at least when using FOCE (AAPS PharmSci. 2000;2(3):E32). Best regards, Mats

From: daren.j.austin@gsk.com Subject: Re: [NMusers] Number of subject and population PK/PD modeling Date: Thu, January 27, 2005 9:40 am Gordi, I've used pop-pk for toxicokinetic studies for both dogs and rats. This approach has been used predominantly where data was censored (BQL) in some animals at late time points. The population method correctly uses all of the data without imputation and will, in general provide a more accurate (higher) estimate of AUC(0-inf). These analyses were not done prospectively. Do bear in mind that the variability is likely to be low, even in six subjects, and certainly lower than you will be used to in humans. Kind regards, Daren Dr. Daren J. Austin Director, CPK/Modelling & Simulation Clinical Pharmacology Discovery Medicine GlaxoSmithKline R&D Work: 7-711 2073 or +44 (0) 20 8966 2073 Mobile: 07712 670097 ________________________ voli i cervi volanti, guidi le bici _______________________________________________________