Population model of Human Chorionic Gonadotropin to predict resistance in low risk gestational trophoblastic neoplasia patients
Ludivine Fronton(1), Benoit You(1,2), Marie Pollet-Villard(3), Gilles Freyer(1,2), Michel Tod(1), Olivier Colomban(1), Francois Golfier(3), Pascal Girard(1)
(1)Université Lyon 1, EA3738, CTO, Faculté de Médecine Lyon Sud, F-69600 Oullins, France; (2)Service d'Oncologie Médicale, Centre Hospitalier Lyon Sud, Hospices Civils de Lyon, F-69310, Pierre-Bénite, France; (3)Hospices Civils de Lyon, Hôtel Dieu, Centre de Référence des Maladies Trophoblastiques, F-69288, Lyon, France; Hospices Civils de Lyon, Centre Hospitalier Lyon Sud, Service de gynécologie obstétrique, F-69310 Pierre Bénite, France
Objectives: Conventional methotrexate regimen (MTX) is recommended for treatment of patients with FIGO (International Federation of Gynecology and Obstetrics) low risk Gestational trophoblastic neoplasia (GTN). The treatment efficiency is assessed by Human Chorionic Gonadotropin (hCG) concentrations follow up. In case of MTX resistance or intolerable toxicity, no clear guidelines are available about the method used to analyse hCG kinetic curve [1]. The aim of this study was to evaluate the influence of hCG pharmacokinetic parameters, as prognostic factors able to inform early MTX resistance after the treatment start. Present work presents the population model building of hCG.
Methods: All data derived from a French retrospective study involving 154 patients treated with MTX (IM 1 mg/kg on days 1,3,5,7 q2w) for a low risk GTN between 2000 and 2008. A kinetic population approach using NONMEM VI was performed to model hCG decrease between day 0 (D0) and day 40 (D40) during the first 3 MTX cycles. Since no information on endogenous hCG production was available, its volume of distribution was set to a literature value 3.4L [2]. One, two and three compartments were tested, and extensive testing of model covariates was performed using FOCE INTER method.
Results: The present modelling analysis was performed on a total of 845 hCG concentrations sampled between D0 and D40, representing an average of 5.5 per patient. As data included concentrations below the limit of quantification (BLOQ), the first concentration in a series of BLOQ observations was replaced by LOQ/2 and later observations were censored [3]. hCG decrease was optimally described by a mono-exponential model: hCG(t) = 3900 * e -0.149 * t. Baseline hCG measurement before methotrexate administration showed large inter-individual variability (273.5%) and moderate one for elimination rate constant (53%). On average, apparent individual clearance was estimated to be equal to 0.57 L/day. None of the available covariates was able to explain variability in baseline hCG or K.
Conclusion: The population “PK” model is a useful tool to estimate individual Bayesian posthoc CL. This work is ongoing and those CL will be used in a next step to better predict MTX resistance in hCG patients, database which is under construction.
References:
[1] Foulmann K, Guastalla JP, Caminet N, et al. Gynecol Oncol 102:103-110, 2006.
[2] Norman RJ, Buchholz MM, Somogyi AA, et al. J Endocrinol 164:299-305, 2000.
[3] Beal S. JPP 28 :481-504, 2001.