II-04 Maurice Ahsman

Modelling and simulation of oral GnRH Antagonist TAK-385 and testosterone-lowering response in Prostate Cancer Patients to Optimize Trial Design and Dose Selection

Maurice Ahsman (1), Hélène M. Faessel (2), Nelleke Snelder (1), David MacLean (2) and Peter Vis (1)

(1) LAP&P Consultants BV, Archimedesweg 31, 2333 CM Leiden, the Netherlands, (2) Millennium Pharmaceuticals, Inc., Cambridge, MA, USA, a wholly owned subsidiary of Takeda Pharmaceutical Company Limited

Objectives: To develop a PK/PD model of testosterone (T) suppression with the non-peptide gonadotropin-releasing hormone (GnRH) receptor antagonist TAK-385, and use it in prospective confirmatory trial simulations to select a dose resulting in >90% of pts with T ≤ 50 ng/dL (medical castration) for 48 wks.

Methods: PK and T data were obtained from data from 3 phase I/II clinical trials. In total, 104 healthy males and 174 prostate cancer patients contributed 2465 PK and 3445 T observations, after treatment with various TAK-385 maintenance doses (40-160 mg daily) for up to max 48 weeks. A PK/PD model was developed using non-linear mixed-effects modelling in NONMEM V7.2.0 [1]. Simulations were done to construct the exposure-effect curve, and predict the fraction of subjects with sustained T ≤ 50 ng/dL over a 48-week treatment period (80-120 mg OD, after a loading dose on day 1).

Results: A three-compartment model with first-order delayed absorption and first-order elimination and an exponential error model adequately described TAK-385 PK. T levels were described using a semi-mechanistic PK/PD model, which combined indirect response-based modelling of production and degradation processes (GnRH, testosterone) with competitive and reversible inhibition of endogenous GnRH binding by TAK-385, and down-regulation mechanisms of GnRH receptors. Age was included as a covariate for the endogenous agonist (GnRH) concentration at baseline to account for different baseline T between healthy men and prostate cancer patients. Simulations showed that the proportion of patients with sustained medical castration reached a maximum at doses of 100 mg OD and above, with minimal added benefit beyond 120 mg OD. Higher doses were associated with a more robust T lowering response vs lower dose regimens, taking the 95% CI of expected responders into account. Large variability in PK and PD responses required a sufficiently high maintenance dose to ensure that >90% pts achieve and maintain T ≤ 50 ng/dL.

Conclusion: This analysis provided an integrated understanding of the relationships between TAK-385 dose, exposure and efficacy to inform trial design and decision-making in oncology drug development. A clinical trial of 610 patients receiving TAK-385 120 mg OD, which can produce a larger and consistent treatment effect, has a prospective power of >90%, even when allowing for a 15% dropout. 

References:
[1] Beal SL, Sheiner LB, Boeckmann AJ & Bauer RJ (Eds.) NONMEM Users Guides. 1989-2011. Icon Development Solutions, Ellicott City, Maryland, USA

Reference: PAGE 25 (2016) Abstr 5847 [www.page-meeting.org/?abstract=5847]

Poster: Drug/Disease modeling - Oncology