NONMEM and drug-drug interaction

From: "BRUNO, Rene" <Rene.BRUNO@rp-rorer.fr> Subject: NONMEM and drug-drug interaction Date: 17 Dec 1997 11:05:54 -0500 Hi everybody, I have one experience of a negative result and a few comments: We have prospectively integrated pop PK/PD approach in the development of docetaxel (TaxotereR), an anticancer agent and came up with a data base of 577 patients from 22 Phase II studies in the registration dossier (based on Phase II studies only). We identified covariates predictive of PK variability and demonstrated that PK variability was a significant predictor of several safety endpoints. Docetaxel is a CYP3A substrate. In our data base no patients were coadministered known inhibitors of CYP3A and we did not look at drug-drug interactions in the main analysis, mainly because the dosage regimen of co-administered drugs (and especially the actual time of intake of the co-administered drug relative to the time of docetaxel administration) was not well documented. One side-effect of docetaxel treatment is fluid retention. During the course of clinical studies several premedication regimens were tested to prevent this problem. A five-day dexamethasone premedication (8 mg po b.i.d. starting the day before docetaxel infusion) significantly delayed the onset and reduced the severity of docetaxel induced fluid retention. However, the concern was raised by a regulatory agency that the effect of dexamethasone could be due to an induction of docetaxel CL (and the resultant decrease of drug exposure) via CYP3A induction rather than by a direct effect of dexamethasone. This question was of importance since it conditioned the approvability of the premedication regimen. A retrospective look at the population PK data base indicated that 82 patients received dexamethasone with various schedules including 25 treated with the recommended 5 day schedule. We tested the effect of dexamethasone premedication on docetaxel CL using the population PK model, and found none. In the absence of these data, a specific clinical trial would have been required to assess docetaxel PK in patients receiving the recommended dexamethasone premedication. I think, one major limitation is the documentation of dosage regimen of coadministered drugs. If the analysis is planned in advance however, the dosage regimen of the drug of interest can be documented and may be subjected to some control without jeopardizing the clinical study objectives. With respect to Janet's point regarding false negatives. I haven't look at back to my files regarding the dexamethasone analysis but I think that when you test an effect when there is none (say change in CL in patients having received a given drug), the parameter estimate related to this effect is typically going to be very small (say less than 1% of 0.5% change in clearance) with a large confidence interval. Am I wrong ? Rene Bruno Rhone-Poulenc Rorer Reference : Bruno R. Integration of Population PK/PD in the Clinical Development and the Registration Dossier of Docetaxel. In: L. Aarons et al. editors. The Population Approach: Measuring and managing variability in response, concentration and dose. European Commission, Luxembourg, 253-262, 1997.