Application of a Multiscale Physiologically-Based Bone and Calcium Systems Model to Guide the Development of GnRH receptor modulators for the Management of Endometriosis
Matthew Riggs (1), Meg Bennetts (2), Steven W. Martin (2), Piet H. Van Der Graaf (2)
(1)Metrum Research Group, LLC, Tariffville, CT, USA ; (2)Pfizer, Pharmacometrics, Global Clinical Pharmacology, Sandwich, UK
Objectives: To provide model-based decision support toward the selection of doses, endpoints and study durations for gonadotropin releasing hormone (GnRH) antagonist clinical programs intended for the management of endometriosis (EM).
Methods: A previously reported, multiscale physiologically-based calcium homeostasis and bone remodeling model has recently been adopted to describe the effects of estrogen loss during menopause transition . Longitudinal effects of varying estradiol (E2) reductions caused by GnRH suppression on biomarkers of bone turnover (BM) and bone mineral density (BMD) changes were simulated from this model. The relative percent decreases in E2 affected by varying degrees of GnRH inhibition were used to fit a differential equation linking bone marker changes with BMD effects to publicly available elagolix, leuprolide and triptorelin data using Berkeley Madonna. A larger literature-based database (1988-2006) that included clinical study-level summary data from various GnRH agonist treatments was used to provide an external evaluation of the 6-month BMD predictions from the multiscale model. Additionally, a logistic regression model describing the relationship of estrogen and total endometrial symptom severity score (ESSS) was fit to patient-level data from three clinical studies using WinBUGS 1.4
Results: Suppression of E2 through GnRH-mediated effects (either agonism or antagonism) occurs within the first month of treatment. Bone markers showed an increase at 6 months in bone specific alkaline phosphatase (BSAP) and serum c-telopeptide (CTX) of 39% and 88%, respectively. BM and BMD model predictions were consistent with these observed data. The model results indicated that bone marker changes are delayed compared to the E2 decrease. In addition, within the range of acceptable 6-month BMD, there was minimal early differentiation of BM across doses. An extension of the model adequately predicted the broader literature database, indicating that E2 was a reliable predictor of 6-month BMD. Logistic regression indicated that P(ESSS=0) increased to 26% and 29% at E2 values of 40 and 20 pg/mL, respectively.
Conclusions: E2 was shown to be a reliable early predictor of 6-month BMD change, whereas bone markers, as affected through GnRH inhibition, were projected to change too slowly to provide reliable early dose differentiation. Doses within a GnRH antagonist development program that target E2 in the range of >20 to 40 pg/mL are expected to provide efficacious EM pain response while minimizing BMD effects.
 M. Riggs, W. R. Gillespie, M. R. Gastonguay, and M. C. Peterson. Extension of a multiscale model of calcium homeostasis and bone remodeling to include the progressive effects of estrogen loss during menopause transition. In Presented at National Institute of General Medical Sciences Quantitative Systems Pharmacology Workshop II, Bethesda, MD, September 2010.