Modeling and Simulation for Determination of the Therapeutic Window of MK-2295: a TRPV1 Antagonist

Transcription

1 Modeling and Simulation for Determination of the Therapeutic Window of MK-2295: a TRPV1 Antagonist William S. Denney *, Yaming Hang *, Marissa Dockendorf *, Chi-Chung Li *, Samer R. Eid *, Robert Valesky *, Tine Laethem *, Pascale Van Hoydonck *, Inge De Lepeliere *, JNJM De Hoon +, Michael Cruthlow *, Rebecca Blanchard * * Merck & Co, Inc., + Univ. Hospital Gasthuisberg, Belgium

2 Outline The Mechanism of TRPV1 Antagonism One molecule, many models Consolidation of understanding Summary

3 TRPV1 is a Polymodal Nociceptor

4 TRPV1 is a Polymodal Nociceptor

5 One Molecule, Many Models Capsaicin induced dermal vasodilation (CIDV) Core Body Temperature Warmth Sensation Hand Withdrawal Time Hot Water Sipping

6 Capsaicin-Induced Dermal Vasodilation (CIDV) Model Rhesus monkeys Laser Doppler Perfusion Imager Capsaicin rubber O -rings Hershey et al., 2005

7 Capsaicin-Induced Dermal Vasodilation (CIDV) Model Capsaicin VR-1 receptor CGRP CGRP receptor Vasodilation MK-2295 Telcagepant (MK-0974) The mechanisms of drugs/receptors are competitive inhibition. Double right arrows indicate the multiple steps between VR-1 binding capsaicin and CGRP release and also multiple steps between CGRP binding its receptor and vasodilation.

8 Capsaicin-Induced Dermal Vasodilation (CIDV) Model F Emax, capsdcaps = E ED50, caps + + C Dcaps ED50, caps CMK 2295 EC Parameter Parameter Estimate 50, MK 2295 % RSE for Parameter E max, CGRP MK 0974 ω Estimate + C EC MK , MK 0974 % RSE for ω Estimate E 0 (arb) E max,caps (arb) E max,cgrp (fraction) NA NA ED 50, caps EC 50, MK NA NA EC 50, MK NA NA ΔE 0, pilot NA NA ΔE 0, NA NA Proportional residual Error

9 Capsaicin-Induced Dermal Vasodilation (CIDV) Model Flow (arb. units) Actual Flow Individual Predicted Population Predicted MK-2295 Plasma Concentration (nm)

10 Core Body Temperature Model Variable Estimate Description T obs - T 0 A ω t E max C EC Observed temperature ( C) Baseline temperature ( C) Diurnal variation in temperature Diurnal period set to π/24 hr -1 Time of day (hr) Maximum druginduced temperature change ( C) Concentration of MK (nm) Concentration of MK required to cause half-maximal temperature increase. T obs πtdrug (ºC) = T 0 + A sin ( ωt) Plasma Concentration (nm) E C + C EC max Observed Population Predicted Individual Predicted 50

11 Warmth Sensation Model Variable Estimate Description T ws - C E 0 E max EC minimum temperature at which warm sensation is detected on the right hand following an oral dose of MK-2295 ( C) MK-2295 plasma concentration (nm) Minimum temp. at which warm sensation is detected at C = 0 ( C) The temp. at which warm sensation would be detected at C = ( C) Concentration of MK-2295 required to cause halfmaximal increase in T ws. (nm) Tws T WS = E 0 + ( E E ) max C + EC 0 50 C Observed Individual Predicted Population Predicted Concentration (nm)

12 Hand Withdrawal Time Model Data was censored due to the fact that subjects were instructed to remove their hands at 150 sec.

13 Hand Withdrawal Time Model Weibull distribution survival model. Scale parameter was fit as an Emax function. PDF for censored data was the integral of PDF above 150 sec. Intra- and inter-individual data variability only allowed naïve pooled fitting. Fit using SAS 9.1 NLMIXED procedure f ( t; β, η) β 1 β β t t exp, t 0 = η η η 0, t < 0 η = E + E EC Parameter Estimate SE p max Shape (β) < E 0 (sec) < EC 50 (nm) E max (sec) < C + C

14 Hot Water Sipping Model P Subjects given 70 C water to sip and then asked if it was safe to drink freely. Graph shows individual and binned responses along with model predictions ± SE. Logistic regression: exp( a + b Conc) Safe = 1+ exp a + b Conc ( ) ( ) Rel. Std. Parameter Estimate Err (%) a b (nm -1 ) Observed or Probability of Stating that 70C Water is Safe to Drink Freely MK-2295 Plasma Concentration (nm)

15 Consolidation of Understanding PK Target Engagement at Steady State C max Steady State TID Dose (mg) C max (nm) Actual Predicted Actual C trough (nm) Predicted Competitive CIDV (%) (Max effect=2.56, EC50=57.9 nm) Core Body Temperature (%) (Max effect =1.06ºC, EC50=69.9 nm) Warm sensation (%) (Max effect=52.6ºc, EC50=242 nm) Pct. of Max Hand Withdrawal Time (%) (EC50=292 nm) Hot Water Sipping P(Safe to drink at 70 ) *Based on a 2-compartment PK model (not shown) arbitrary units Circadian rhythm accounted for an additional 0.623ºC population mean response Represents on-target effects Represents undesired effects

16 Summary Models were developed that described both on-target and undesired effects. Simulations suggest that it is not possible to decouple the loss of temperature sensitivity from the on-target effects. These models taken as a combined set can help inform decision making through the understanding of the therapeutic window for the compound.

17 Acknowledgments Dick Simpson (currently with United Phosphorous)