Christophe Passot (1), David Ternant (2,3), Rebecca Sberro-Soussan (4), Dominique Bertrand (5), Sophie Caillard (6), Christelle Barbet (7), Betoul Schvart (8), Cécile Vigneau (9), Camille Domenger (10), Gilles Paintaud (2,3), Jean-Michel Halimi (7), Philippe Gatault (7).
(1)Western Cancerology Institute, Biopathology Department, Angers, France; (2)GICC team, Tours, France; (3)University of Medicine, Pharmacology and Toxicology Department, Tours, France; (4)Nephrology Department, Necker hospital, Paris, France; 5Nephrology Department, Rouen hospital, Rouen, France; (6)Nephrology Department, Strasbourg hospital, Strasbourg, France; (7)Nephrology Department, Tours hospital, Tours, France; (8)Nephrology Department, Reims hospital, Reims, France; (9)Nephrology Department, Rennes hospital, Rennes, France; (10)Nephrology Department, Grenoble hospital, Grenoble, France.
Objectives:
Eculizumab is a monoclonal antibody toward C5 fraction of the complement system. It is approved to treat atypical hemolytic uremic syndrome (aHUS), an orphan disease, at a fixed dose of 1200 mg every 2 weeks. It has been demonstrated that a serum concentration of 99 mg/L is sufficient to suppress complement activity [1]. A pilot study revealed that serum trough concentrations were excessively high in many patients and that administration schedule may be individualized [2]. This prospective multicenter study aimed at (i) quantifying pharmacokinetic variability of eculizumab in aHUS patients and (ii) to design a Bayesian model for individual dosing adjustment of eculizumab.
Methods:
This work is a two-step study. The first part was conducted on 40 patients with aHUS from 6 hospitals. Eculizumab trough concentrations were determined on blood samples drawn in 2016 with a validated ELISA assay [3]. A population PK model was made using Monolix 4.3.3 software (Lixoft, Orsay, France). Body weight, height, sex and age were tested as covariates on PK parameters. Individual time since last dose to reach 100 mg/L steady-state trough concentration (TLD100) was computed for each patient. Patients whose theoretical individual administration interval was at least 21 days were selected for dose tapering. The second part consisted in the development and validation of a Bayesian model for dosing adjustment of eculizumab based on samples drawn in 2017 from patients with TLD100>21 days. Bayesian model used structural, interindividual, residual and covariate model parameters estimated during the first part, as prior information. Individual PK parameters assessed with the Bayesian model were used to predict eculizumab trough concentrations after interval lengthening. Predicted concentrations were then compared with observed concentrations after lenghtening.
Results:
During the first part, a total of 170 eculizumab serum trough concentrations were available in the 40 eligible patients. Eculizumab was administered every 2 weeks in 37 (92.5%) patients out of 40 patients of the first step (92.5%) and every 3 weeks in 3 patients (7.5%). A median of 4 blood samples were available for each patient (range: 2-5). The mean individual eculizumab trough concentration was 476.1 mg/L (range: 124.3-1064.6 mg/L) Eculizumab pharmacokinetics was best described using a one-compartment model with first order elimination rate constant. Population volume of distribution and clearance were 3.1 L and 0.13 L/day, respectively. Body weight increased significantly elimination clearance (bWT_CL=1.16, p=0.0001). TLD100 was >21 days in 37 patients (92.5%). In 13 out of these 37 patients, individual PK parameters were assessed with the Bayesian model. Eculizumab trough concentrations after dose tapering were predicted and compared with observed concentrations. The correlation coefficient of observed vs predicted eculizumab concentrations was 0.92, with a mean of absolute bias of 47.2 mg/L.
Conclusions:
In this ongoing work, we confirmed that eculizumab administration could be individualized through therapeutic drug monitoring. The estimation of PK parameters in a relatively large cohort for this orphan disease allowed to construct a Bayesian model. This model allowed a satisfying prediction of eculizumab trough concentrations after interval lengthening. It is currently in use to determine PK parameters of eculizumab in newly enrolled patients and to determine if administration lengthening can be performed.
References:
[1] Jodele S, Dandoy CE, Danziger-Isakov L, Myers KC, El-Bietar J, Nelson A, Wallace G, Teusink-Cross A, Davies SM. Terminal Complement Blockade after Hematopoietic Stem Cell Transplantation Is Safe without Meningococcal Vaccination. Biol Blood Marrow Transplant. 2016 Jul;22(7):1337-1340.
[2] Gatault P, Brachet G, Ternant D, Degenne D, Récipon G, Barbet C, Gyan E, Gouilleux-Gruart V, Bordes C, Farrell A, Halimi JM, Watier H. Therapeutic drug monitoring of eculizumab: Rationale for an individualized dosing schedule. MAbs. 2015;7(6):1205-11.
[3] Passot C, Desvignes C, Ternant D, Bejan-Angoulvant T, Duveau AC, Gatault P, Paintaud G. Development and validation of an enzyme-linked immunosorbent assay to measure free eculizumab concentration in serum. Bioanalysis. 2017 Aug;9(16):1227-1235.
Reference: PAGE 27 (2018) Abstr 8715 [www.page-meeting.org/?abstract=8715]
Poster: Drug/Disease Modelling - Other Topics