Identification of non-adherence to adjuvant letrozole using a population pharmacokinetics approach in hormone receptor-positive breast cancer patients

Background Letrozole, an aromatase inhibitor metabolised via CYP2A6 and CYP3A4/5 enzymes, is used as adjuvant therapy for women with hormone receptor (HR)-positive early breast cancer. The objective of this study was to quantify the impact of CYP2A6 genotype on letrozole pharmacokinetics (PK), to identify non-adherent patients using a population approach and explore the possibility of a relationship between non-adherence and early relapse. Methods Breast cancer patients enrolled in the prospective PHACS study (ClinicalTrials.gov NCT01127295) and treated with adjuvant letrozole 2.5 mg/day were included. Trough letrozole concentrations (Css,trough) were measured every 6 months for 3 years by a validated LC-MS/MS method. Concentration-time data were analysed using non-linear mixed effects modelling. Three methods were evaluated for identification of non-adherent subjects using the base PK model. Results 617 patients contributing 2534 plasma concentrations were included and led to a one-compartment PK model with linear absorption and elimination. Model-based methods identified 28% of patients as non-adherent based on high fluctuations of their Css,trough compared to 3% based on patient declarations. The covariate analysis performed in adherent subjects revealed that CYP2A6 intermediate (IM) and slow metabolisers (SM) had 21% (CI95%=12 – 30%) and 46% (CI95%=41 – 51%) lower apparent clearance, respectively, compared to normal and ultrarapid metabolisers (NM+UM). Early relapse (19 patients) was not associated with model-estimated, concentration-based or declared adherence in the total population (p=0.41, p=0.37 and p=0.45, respectively). Conclusions These findings will help future investigations focusing on the exposure-efficacy relationship for letrozole in adjuvant setting.