Exposure-Response Analysis of a DPP-IV Inhibitor, PF-00734200 on HbA1c in Type 2 Diabetic Subjects on Stable Metformin Treatment
Srividya Neelakantan (1), Kevin Sweeney (1), Haiqing Dai (1), Steven G Terra (2), Thomas G Tensfeldt (1)
(1) Clinical Pharmacology-Primary Care, Pfizer Inc.; (2) Clinical Sciences–Primary Care, Pfizer Inc.
Objectives: PF-00734200 is a dipeptidyl peptidase-IV (DPP-IV) inhibitor intended for use in the treatment of type 2 diabetes mellitus. The purpose of this analysis was to investigate the relationship between PF-00734200 exposure and change in glycosylated hemoglobin (HbA1c) at Week 12 in type 2 diabetic subjects.
Methods: Two randomized, placebo-controlled, parallel group, multiple-dose studies were conducted for 12 weeks in diabetic subjects on stable treatment with metformin in combination with once daily PF-00734200 doses ranging from 2 to 30 mg. A population PK analysis was performed using a nonlinear mixture model. The predicted PF-00734200 exposures (average steady-state concentrations) were subsequently incorporated into an inhibitory Emax model to characterize the effect of PK on the change from baseline in HbA1c at Week 12. Covariates of interest were identified to help explain the inter-individual variability in the model parameters. The predicted median change from baseline in HbA1c response at Week 12 for the various doses and the 95% confidence intervals (CIs) was derived from the parameters estimated from 1000 replicate bootstrap parameter vectors. The probabilities of having greater reduction in the HbA1c compared to other treatment arms, particularly 20 vs 10 and 30 vs 20 mg doses were also calculated from the bootstrap parameter vectors and the observed median exposures.
Results: Data from a total of 482 subjects who had both baseline and Week 12 HbA1c data were analyzed. The Emax, EC50 and E0 (placebo response) parameters were estimated to be -1.01%, 20.7 ng/mL and 0.133% respectively. Baseline HbA1c and age were identified as influential covariates describing the maximal HbA1c response (Emax). The potential clinically relevant doses of 10, 20 and 30 mg resulted in mean (95% CI) percentage change from baseline in HbA1c of -0.632 (-0.759, -0.501), -0.758(-0.925, -0.585) and -0.801 (-1.02, -0.604) respectively. The probability of observing a greater than 0.1% decrease in the HbA1c at Week 12 for the 20 mg over 10 mg and for 30 mg over 20 mg doses are 0.614 and 0.064 respectively.
Conclusions: The inhibitory Emax model adequately described the exposure-change from baseline HbA1c response. In conclusion, the 20 mg dose is superior to the 10 mg dose in reducing the HbA1c response over 12 weeks whereas the 30 mg and 20 mg doses produced clinically similar HbA1c reductions.