Marylise Sterlé1, Bernard Royer2, Sylvia Ilie3, Isabelle Desmoulins3, Didier Mayeur3, Manon Reda3, Audrey Hennequin3, Jean-David Fumet4, Leïla Bengrine-Lefevre3, Courèche Kaderbhai, Antonin Schmitt1
1Pharmacy Department, Centre Georges-François Leclerc; INSERM U1231, Université Bourgogne Europe ;Groupe de Pharmacologie Clinique Oncologique GPCO-Unicancer, 2Pharmacology and Toxicology Laboratory, CHRU Besançon; Groupe de Pharmacologie Clinique Oncologique GPCO-Unicancer, 3Oncology Department, Centre Georges-François Leclerc, 4Oncology Department, Centre Georges-François, INSERM U1231, Université Bourgogne Europe
Objectives Abemaciclib, an oral cyclin-dependent kinase 4/6 inhibitor, is approved for the treatment of HR+/HER2- advanced breast cancer and as adjuvant therapy for HR+/HER2- early breast cancer at high risk of recurrence. Despite its clinical benefits, its use is limited by adverse effects, especially hematologic disorders. The mechanisms underlying these toxicities remain poorly understood, and identifying contributing factors is key for optimizing treatment. The aim of this study was to describe the pharmacokinetic/pharmacodynamic (PK/PD) relationships between abemaciclib plasma concentrations and the main adverse effects in a real-world population, to propose a target concentration for therapeutic drug monitoring (TDM). Methods A retrospective analysis was conducted at the Dijon Clinical Cancer Center, including patients who began abemaciclib treatment between April 2021 and December 2023. As TDM of abemaciclib was routinely performed, assay results were available. The data required were collected from the patients’ medical records. To develop the PK/PD models, a non-linear mixed-effects regression approach was applied using Monolix® software (Lixoft SAS, a Simulations Plus company). Individual PK parameters were estimated by bayesian approach on one observation using a previously published population PK model [1] and were applied as constants (regressors in Monolix®) in the PK/PD models to estimate abemaciclib plasma concentrations. Covariate selection was performed using a Forward/Backward approach. Final model selection was based on comparisons of objective function values, relative standard errors (RSE, in %), convergence of Monolix® to typical population values, and graphical diagnostics. The final models were used to calculate the incidence of neutropenia and thrombocytopenia based on plasma trough concentrations (Cres) for 1,000 virtual subjects to define target exposure levels. Results A total of 405 neutrophil and 435 platelet concentrations were available for 65 patients. The two PK/PD models that best described the relationship between abemaciclib plasma concentrations and the evolution of neutrophil and platelet concentrations during the first 6 months of treatment were based on Friberg’s model [2], which consists of five compartments (one for stem cells, three transit compartments, and one for the circulating compartment), a feedback mechanism (reflecting the effect of hematopoietic factors), and a linear drug effect that leads to a decrease in stem cell proliferation. For both models, all parameters were estimated with RSEs less than 21% for neutrophils and 18% for platelets, and residual errors are 24 % and 18% respectively. The only significant covariate, improved the platelet model by reducing inter-individual variability on the slope of sensitivity to abemaciclib-induced thrombopenia by 12% compared to the base model was “age.” Simulations showed that 20% of simulated patients with Cres around 200 µg/L developed = 3 grade neutropenia, while much higher Cres were required to induce = 3 grade thrombopenia in 20% of simulated patients. Conclusions Two PK/PD models describing the evolution of neutrophils and platelets as a function of time in patients treated with abemaciclib, based on real-life data, were developed. A target concentration linked to the onset of grade =3 neutropenia was identified, which could guide dosage adjustment during TDM. References [1]Chigutsa E and al., « Development and Application of a Mechanistic Population Modeling Approach to Describe Abemaciclib Pharmacokinetics. », CPT Pharmacomet. Syst. Pharmacol., doi: 10.1002/psp4.12544. [2]L. E. Friberg and al. « Model of Chemotherapy-Induced Myelosuppression with Parameter Consistency Across Drugs », J. Clin. Oncol., sept. 2016, doi: 10.1200/JCO.2002.02.140.
Reference: PAGE 33 (2025) Abstr 11687 [www.page-meeting.org/?abstract=11687]
Poster: Clinical Applications