Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Modelling and Simulation of the Telephone Sexual Activity Daily Diary Data of patients with female sexual arousal disorder treated with sildenafil (Viagra®) L. Claret(1), E.H. Cox(1), L. McFadyen(2), A. Pidgen(2), P.J. Johnson(2), S. Haughie(2), M. Boolell(2), R. Bruno(1) (1)Pharsight Corporation, Drug Development Consulting Services, Mountain View, CA. (2)Pfizer, PGRD, Sandwich, UK Objectives Introduction Sildenafil (Viagra) is an orally active, selective inhibitor of cGMP specific phosphodiesterase type 5 (PDE5), approved for the treatment of male erectile dysfunction. To develop models to: Although some clinical studies in FSAD have shown statistically significant sildenafil efficacy over placebo others have not. • Characterize the probability of sexual events and their satisfaction scores over time based on Telephone Sexual Activity Daily Diary (TSADD) data obtained in clinical studies of sildenafil in patients with FSAD without concomitant Hypoactive Sexual Desire Disorder (HSDD). Three Phase 2b/3 studies had collected data in a consistent manner thus offering a good opportunity for a combined analysis to look at temporal aspects of clinical response under placebo and active treatment. • Simulate the expected dose-response in various patient populations to assess the impact of patient and disease characteristics on outcome. Experimental evidence suggests that the nitric oxide-cGMP pathway may be important in producing clitoral engorgement, pelvic vasocongestion and vaginal lubrication thus enhancing the female sexual arousal response. Data Sexual Activity Daily Diary • Captured every day … by menopausal status 0.5 0.4 00 55 10 10 00 Time Timeweek week 55 Scores model θ2 q6 , ,q9 h j ,S N 0, 1.0 0.8 0.6 0.4 1.0 0.2 0.4 P(Sorg>=2) 0.6 0.8 1.0 0.0 0 20 40 Dose (mg) 60 80 100 0 20 rand 40 80 100 80 100 P(Ssex>=2) 1.0 P(Ssex>=3) 60 rand Dose (mg) 0.8 E(t) = 374 days where is the Gamma function 0.6 P(Sorg>=2) 0.4 1 1 P(Sorg>=3) 0.8 1% 4% 7% 0.6 0.016 0.012 0.008 0.4 -1.410 0.316 0.125 P(Ssex>=2) and 55% 45% 19% 9% 5% 14% 7% 5% 4% 10% 17% 66% 13% 7% 9% 29% 9% 36% 14% 85% 0.2 ,q 5 CV 0.237 0.237 0.236 0.239 0.258 0.317 0.333 0.355 0.378 0.038 0.057 0.001 0.005 0.059 0.088 0.028 0.136 0.001 0.039 0.059 0.0 θ1 q1 , SE -0.428 -0.521 -1.240 -2.760 -5.140 -2.270 -4.540 -6.880 -9.660 0.380 0.338 0.001 -0.035 0.821 1.010 0.096 1.460 0.003 0.272 0.070 P(Sorg>=3) ln() ln() Sexual satisfaction scores were modeled from predicted orgasm satisfaction 1 exp x Estimate Intercept Intercept Intercept Intercept Intercept Intercept Intercept Intercept Intercept Intercept Time drgtaken x dose x tlevel Age base.osat Drgtaken time exponent base.ssat Menopausal 0.2 57485 Sildenafil effect depends on drug intake, testosterone level, and menopausal status Covariate 0.0 Weibull q1 q2 q3 q4 q5 q6 q7 q8 q9 q10 q11 q12 q13 q14 q15 q16 q17 q20 ln[hj,org)] ln[SD hj,sex)] 1.0 0.2 Posterior predictive check of P(Sorg N), and P(Ssex N) as a function of dose. We compared the 90%CI and median across 100 simulated replicates of predicted proportions to those observed 0.8 1% 8% 0.0 10 10 Time Timeweek week 10 10 Cv 0.016 0.008 1 Probability 0.8 0.0 0.0 0.1 0.1 Areas of tiles are proportional to the number of events P(Ssex>=3) Ste -1.310 0.102 0.6 54 3 2 1 0.5 0.5 0.3 0.3 0.2 0.2 0.3 0.3 Probability Probability 0.4 0.4 Placebo Placebo Sildenafil Sildenafil 0.2 Value ln() E t 0.4 Probability 0.2 0.0 0.2 0.2 0.0 0.0 0.1 0.1 0.2 0.2 0.5 0.5 55 Time Timeweek week 59253 0 Orgasm satisfaction 0.5 0.5 0.4 0.4 0.3 0.3 0.3 0.3 Probability Probability 0.1 0.1 0.0 0.0 exp x 0.2 0.2 Probability Probability 0.1 0.1 10 10 P S j ,sex m | z j , S j ,org g exp θ2 Esex h j , sex Eorg q13 age+q14 base.Sorg+q15 drgtaken+ q12 q 20 meno dose drgtaken tlevel q11 1-e-q16 time 00 expon. P( T * t ) exp t , Esex q10 S 2j ,org q17 base.Ssex q11 1 e q16 time 10 10 Parameter 5 0.3 Probability Obj 4 Placebo Placebo Sildenafil Sildenafil Time to event model Model 3 Satisfying Satisfying orgasms orgasms with with drug drug intake intake Testosterone Testosterone level level << 0.9 0.9 pg/ml pg/ml Time Time week week t 1 P( T * t ) exp h( u )du , with for Weibull distribution h t t 0 2 Satisfying Satisfyingorgasms orgasms with withdrug drugintake intake Testosterone Testosteronelevel level << 0.9 0.9pg/ml pg/ml 0.1 55 Time (week) 10 10 0.0 00 with g x 55 Parameter estimates • Time between events and scores are independent, time to event distribution was estimated by a Weibull distribution model in NONMEM: 10 0.2 0.5 0.4 0.3 0.2 0.1 Probability 1: not satisfied … 5: extremely satisfied Satisfying Satisfyingorgasms orgasms with withdrug drugintake intake Testosterone Testosteronelevel level >= >= 0.9 0.9pg/ml pg/ml Time Timeweek week 0.0 Final Model P S j ,org m | z j g exp θ1 Eorg h j ,org Overall Sexual Satisfaction 55 5 Tim e week proptest_v aria_1127_estim_DAAT.ssc Time Timeweek week 0.0 0.0 00 0 Placebo Sildenaf il 0.0 Covariates: Treatment duration (time) sexual satisfaction baseline score (base.Ssex) 10 10 Satisfying Satisfying orgasms orgasms with with drug drug intake intake Testosterone Testosterone level level >= >= 0.9 0.9 pg/ml pg/ml • Orgasm, and sexual satisfaction scores were modeled simultaneously in NONMEM: Orgasm 00 0.4 0.4 0.4 55 Time Time week week 0: no orgasm 1: not satisfied … 5: extremely satisfied 0.3 0.1 0.0 00 Model Covariates: Treatment duration (time) Orgasm satisfaction baseline score (base.Sorg) Age Drug intake (drgtaken) Sildenafil Dose Testosterone level (tlevel) Menopausal status (meno) 0.2 Probability 0.3 0.0 0.1 0.2 Probability 0.4 0.5 Satisfying Satisfying orgasms orgasms with with drug drug intake intake 0.5 Satisfying Satisfying orgasms orgasms Dose to be taken 30 mins to 4 hours prior to sexual event Sexual event 10 10 Time Timeweek week All studies have 2-6 weeks treatment free and 12 wks treatment phase. Time since previous event Probability Probability 0.4 0.4 0.5 0.5 Satisfying Satisfyingorgasms orgasms with withdrug drugintake intake Study 1127 Proportions of satisfying orgasms (Sorg≥3) versus time (90% confidence interval) 55 10 Sexual satisfaction 1 Satisfying Satisfyingorgasms orgasms 00 Satisfying orgasms with drug intake Testosterone level >= 0.9 pg/ml in pre-menopausal Correlation of sexual satisfaction and orgasm satisfaction • Time of drug intake before sexual event – Less than 30’ minutes – 30’ to 4 hours – Greater then 4 hours Study 1123 - 98 women – Premenopausal with minimum physiological level of estradiol and free testosterone – Dose: Placebo, 5, 10, 25, 50, 100 mg with 83, 41, 43, 42, 42, 43 patients respectively 5 Tim e week • Orgasm – Presence and absence – Satisfaction scale (1-5) (Sorg) – Satisfying events defined as above Study 1082 - 71 women – Postmenopausal on HRT with minimum physiological level of estradiol ( 40 pg/ml) and free testosterone ( 0.9 pg/ml) – Dose: Placebo, 5, 10, 25, 50, 100 mg with 21, 11, 9, 10, 10, 10 patients respectively The structure of the model was determined by the nature of the clinical endpoint that is derived from the TSADD 0 0.6 Study 1127 - 248 women – Premenopausal (n = 43) and postmenopausal (n= 205) women on HRT – Minimum physiological level of estradiol ( 40 pg/ml) except for patients on HRT – Stratified according to free testosterone level ( 0.9 pg/ml (n = 121) or <0.9 pg/ml (n = 127)) – Dose: placebo (n = 124), 50 mg or adjusted doses (25 (6%) or 100 (75%) mg) (n = 124) 1.0 • Sexual events – Presence or absence – Satisfaction scale (1-5) (Ssex) – Event rated moderately to extremely satisfying (scale 3 to 5) were rated as satisfying Satisfying orgasms with drug intake Testosterone level >= 0.9 pg/ml in post-menopausal 0.4 Randomized, double blind, placebo controlled, multicenter studies to evaluate efficacy, safety and toleration of oral sildenafil administered for 12 weeks to women with FSAD 0 20 40 60 80 100 0 20 40 Dose (mg) rand 60 rand Dose (mg) simul_scores3_post_rdtk3_meno.ssc Model Simulations Absolute difference from placebo (%) 0.3 (-0.2, 0.9) 0.7 (0.1, 1.3) 1.7 (0.8, 2.7) 50 38.1 (34.4, 42.0) 3.5 (1.9, 5.2) 100 41.6 (37.0, 46.3) 6.9 (3.7, 10.0) 0 5 10 1 15 80 5 60 40 6 10 10 8 P(Sorg>=3), absolute difference from placebo (%)) 25 p(Sorg3) (%) 34.7 (31.3, 38.3) 35.0 (31.6, 38.7) 35.4 (31.9, 39.0) 36.4 (32.9, 40.1) Sildenafil Dose (mg) 100 mg sildenafil placebo 0 20 40 60 80 0 Sildenafil Dose (mg) P(Sorg>=3)(%) 100 20 80 0 60 0.1+ thru 0.9 pg/ml (Q1) 0.9+ thru 1.2 pg/ml (Q2) 1.2+ thru 1.6 pg/ml (Q3) 1.6+ thru 18.4 pg/ml (Q4) 4 40 100 mg Sildenafil 2 15 10 5 20 There is a substantial decline in orgasm satisfaction score with age. However, the impact of age on the treatment effect is minimal. 0 –500 replicates –90% prediction intervals are represented –A population of 5,000 subjects is simulated for each replicate • Relevant patient covariates are sampled from the 1123/1127/1082 study population • Response rate is evaluated after 12 weeks of study duration (unless noted otherwise) 0 Treatment responses for 100 mg Sildenafil may vary according to testosterone level from 4.0% (1st quartile) to 12% (4th quartile). P(Sorg>=3), absolute difference from placebo (%)) • Simulations are integrated across model uncertainty: difference from placebo (%)) –As such the performance of sildenafil in the expected patient population can be evaluated and is not affected by variability due to a limited trial size. –Subsequent trial simulations can then be performed to evaluate how this population performance can be translated into the performance of a trial (of limited size). 0 • The probability of achieving a Orgasm Satisfaction Score, P(Sorg), 3 is used throughout the simulations. • The response rate is evaluated for the expected patient population. The expected absolute treatment effect for 100 mg Sildenafil may be 7% p(Org. Sat. >=3,p(Org. absolute from placebo (%)) >=3, absolute Sat. difference Simulation of orgasm satisfaction score 100 20 Sildenafil Dose (mg) 30 40 50 60 70 20 30 Age (years) median (90% prediction interval) Conclusions • A model for the time-course of diary data observed in 3 Phase 2/3 studies of sildenafil was developed • A Weibull distribution best described the probability density function of the time between sexual events • Satisfaction scores were simultaneously modelled with overall sexual satisfaction conditional on orgasm satisfaction • Simulations were performed to evaluate the expected clinical response in the FSAD patient population • p(Sorg3) ranges from 34.7% for placebo to 41.6% for 100 mg sildenafil. Thus, the absolute treatment effect (difference from placebo) for sildenafil may be up to 6.9% for 100 mg sildenafil. • treatment effect of sildenafil is increased in post-menopausal women with high testosterone level. Pfizer Global Research & Development Sandwich CT13 9NJ Kent 40 50 60 70