Population PK/PD evaluation of the effect of dienogest on Hoogland Score in young healthy women
Sathej Gopalakrishnan (1), Barbara Schuett (1), Stefan Klein (2), Stefanie Reif (1)
(1) Clinical Pharmacology; (2) Global Pharmacometrics, Bayer Schering Pharma AG, Berlin, Germany
Objectives: Dienogest (DNG) is a 19-nortestosterone derivative that exhibits selective binding to the progesterone receptor and displays neither agonist nor antagonist activity on androgenic, glucocorticoid or mineralocorticoid receptors. It is used for oral contraception and menopause management, as part of a combination treatment, and, as mono-preparation, for the treatment of endometriosis. The aim of our study was to develop a sequential PK/PD Model to characterize the concentration-response relationship with regard to Hoogland Score, a common tool to assess ovarian function.
Methods: The data was taken from 86 healthy subjects, who received DNG in daily oral doses of 0.5 mg, 1 mg, 2 mg and 3 mg over 72 days, in a single-center, randomized, double-blinded, dose-controlled study [1]. Individual AUC was selected as a predictor for the PD parameter Hoogland Score (HS), a categorical measure of the extent of follicular development. HS data were grouped into three levels: HS 1 or 2: no or minimal ovarian activity, HS 3 or 4: residual ovarian activity, and HS 5 or 6: ovulation. Subsequent to the development of a Population Pharmacokinetic (popPK) model, an EMAX-based proportional odds model was implemented in NONMEM VI to construct a PK/PD description.
Results: The popPK model developed was a linear two compartment model with first order absorption and linear elimination from the central compartment. Moderate interindividual variability (~25%) was estimated for oral clearance (CL/F) and central volume of distribution (V2/F). An EMAX based proportional odds PD model, driven by AUC,of the form Effect = EMAX*AUC^G/(AUC^G+AUC50^G) with G=1, was used. The model converged successfully with acceptable relative standard errors (<35%) for all parameters. The estimates were EMAX = 10.1, AUC50 = 0.3 mcg*h/mL (fixed to observed mean AUC for medium effect - i.e. a HS of 4), ODDS1 (for HS less than or equal (LE) 2) = -6.27 and ODDS2 (for HS LE 4) = 3.53. Different Hill coefficient possibilities were looked at and G=1 adequately described the data. The predicted probabilities were in excess of 0.5 for a HS LE 2 and in excess of 0.9 for a HS LE 4, even for relatively moderate DNG exposures of 0.5 mcg*h/mL, for the 2 mg dose group.
Conclusions: The PK/PD evaluation shows that DNG AUC is a good predictor for PD response in terms of Hoogland Score using a proportional odds model. The model may be further used for the prediction of ovulation inhibition based on DNG exposure.
References:
[1] Klipping C, Duijkers I, Faustmann T, Klein S, Schuett B. Pharmaco-dynamic study of four dosages of Dienogest. Fert&Ster 2010; 94(4) Suppl: 181
