Model‐based population pharmacokinetic analysis of vinorelbine in pediatric patients: impact of body composition and pharmacogenetic factors on drug exposure variability

XVIIIes Journées du GPCO

M. Hamimed1, J. Ciccolini1, 13, A. Iliadis1, N. André1, 2, E. Tresch Bruneel3, A. Probst4, P. Chastagner5, A. Pagnier6, E. DeCarli7, N. Entz-Werlé8,
J. Grill9, F. Doz10, D. Frappaz11, A. Bertozzi12, P. Leblond14

  1. Aix-Marseille University, Center for Research on Cancer of Marseille- UMR Inserm 1068- CNRS UMR 7258SMARTc Unit, Marseille, France

  2. APHM- Hôpital La Timone, Department of Pediatric Oncology, Marseille, France

  3. Centre Oscar Lambret, Department of Biostatistics, Lille, France

  4. Oscar Lambret Comprehensive Cancer Center,

    Département de la Recherche Clinique et Innovation, Lille, France

  5. Nancy University Hospital, Service d’hémato-oncologie pédiatrique, Nancy, France

  6. Grenoble University Hospital, Service d’hémato-oncologie pédiatrique, Grenoble, France

  7. Angers University Hospital, Service d’hémato-oncologie pédiatrique, Angers, France

  8. Université de Strasbourg, CHRU Hautepierre- Pédiatrie Onco-Hématologie-

    UMR CNRS 7021, Strasbourg, France

  9. Université Paris Saclay, Gustave Roussy- Département de Cancérologie

    de l’Enfant et de l’Adolescent et UMR 8203 du CNRS, Villejuif, France

10. Université Paris Descartes, Institut Curie- Oncology Center SIREDO, Paris, France
11. IHOPe, Centre Léon Bérard, Lyon, France
12. Toulouse University Hospital, Service d’hémato-oncologie pédiatrique, Toulouse, France
13. APHM- Hôpital La Timone, Pharmacokinetics and Toxicology, Marseille, France
14. Oscar Lambret Comprehensive Cancer Center, Department of Pediatric Oncology, Lille, France

BACKGROUND/OBJECTIVES

Vinorelbine (VNR) is a semi-synthetic vinca alkaloid used primarily in the treatment of advanced breast and non-small cell lung cancers. With a better patient’s quality of life in comparison to the intravenous route, oral VNR represents an attractive treatment option for the management of tumors requiring prolonged treatment duration and also for the emerging metronomic chemothe- rapy approach notably in the pediatric population. There is a crucial need for pharmacokinetic (PK) data of oral VNR in children. The objective of this work was to characterize the population PK of VNR and to identify potential demographic, physiopathological, and genetic factors influencing its exposure variability among children.

DESIGN/METHODS

VNR PK data were obtained from a multicentric phase II study, conducted in children with recurrent or progressive primary Low-Grade Glioma. Patients received 60 mg/m2 of oral VNR on Days (D) 1, 8 and 15 of Cycle (C) 1 and 80 mg/m2 on D1, 8 and 15 of C2 to C12. Population PK analysis was performed using a nonlinear mixed-effects modeling approach. Parameters were estimated by the SAEM algorithm, implemented in Monolix software. A covariate analysis was conducted; covariates were evaluated in a stepwise procedure with forward addition (α = 0.05) followed by backward elimination (α = 0.01).

RESULTS

A three-compartments model, with a delayed double zero-order absorption and a linear elimi- nation, best described VNR pharmacokinetics. Typical population estimates (between-subject variability, betweenoccasion variability) for the apparent central volume of distribution (Vc/F) and elimination rate constant (k10) were 803 L (47%, 41%) and 0.60 h-1 (18%, 16%), respectively. Model diagnosis showed good descriptive and predictive performances. Following covariate ana- lysis, BSA, white blood cells count, and ABCB1rs2032582 SNP on the Vc/F were clinically relevant. Additionally, ABCG2-rs2725264 A/G or G/G carriers and UGT2B7-rs7668258 T/T carriers showed a higher k10 (i.e., 30% and 42% compared to reference alleles, respectively) and included in the final model. Conversely, age and sex had no significant effect on VNR PK.

CONCLUSION

High inter-individual variability of VNR systemic exposure was observed in French children with LGG. Univariate and multivariate simulations showed that the effects of all identified covariates on VNR AUC0-inf and Cmax were outside the established bounds for an irrelevant clinical change in VNR safety and/or efficacy. Thus, a personalized dosing strategy considering those covariates would help reducing PK variability to optimize the efficacy/toxicity balance of VNR.

Keywords: low-grade glioma, pediatric, population pharmacokinetic analysis, vinorelbine, inter-individual variability.