Adherence to the CDK 4/6 Inhibitor Palbociclib and Omission of Dose Management Supported by Pharmacometric Modelling as Part of the OpTAT Study

Abstract

The cyclin-dependent kinase 4/6 inhibitor (CDK4/6i) palbociclib is administered orally and cyclically, causing medication adherence challenges. We evaluated components of adherence to palbociclib, its relationship with pharmacokinetics (PK), and drug-induced neutropenia. Patients with metastatic breast cancer (MBC) receiving palbociclib, delivered in electronic monitors (EM), were randomized 1:1 to an intervention and a control group. The intervention was a 12-month interprofessional medication adherence program (IMAP) along with monthly motivational interviews by a pharmacist. Implementation adherence was compared between groups using generalized estimating equation models, in which covariates were included. Model-based palbociclib PK and neutrophil profiles were simulated under real-life implementation scenarios: (1) optimal, (2) 2 doses omitted and caught up at cycle end. At 6 months, implementation was slightly higher and more stable in the intervention (n = 19) than in the control (n = 19) group, 99.2% and 97.3% (Δ1.95%, 95% CI 1.1−2.9%), respectively. The impact of the intervention was larger in patients diagnosed with MBC for >2 years (Δ3.6%, 95% CI 2.1−5.4%), patients who received >4 cycles before inclusion (Δ3.1%, 95% CI 1.7−4.8%) and patients >65 (Δ2.3%, 95% CI 0.8−3.6%). Simulations showed that 25% of patients had neutropenia grade ≥3 during the next cycle in scenario 1 versus 30% in scenario 2. Education and monitoring of patient CDK4/6i cycle management and adherence along with therapeutic drug monitoring can help clinicians improve prescription and decrease toxicity.

Keywords: CDK4/6 inhibitor; electronic adherence monitoring; interprofessionality; medication adherence; motivational interviewing; neutropenia; oral anticancer therapy; palbociclib; pharmacists.

Figure A1

NB: The end of each ON (21 days) and OFF phase (7 days) is represented in red lines and numbers. Empirical implementation is represented with the light blue curves, GEE modelisation is represented with the thick blue curve in each ON and OFF phase.

Figure 1

Flow of patient enrollment and follow up from inclusion to data analysis. ^a study discontinuation before randomization; NB: EM = electronic monitor.

Figure 2

GEE models estimating patient implementation (the vertical red dotted line indicates the randomization at day 21, date on which the intervention starts—i.e., after the baseline period). (a) Patients’ palbociclib implementation in each randomized group, (b) dichotomization in each group according to time since MBC diagnosis, (c) dichotomization in each group according to the number of palbociclib cycle received before inclusion, and (d) dichotomization in each group according to age. NB: In (a), the two discontinuations are represented by dots in the green curves depicting the number of participants over time.

Figure 3

Discontinuation and censoring times algorithm.

Figure 4

Population median prediction of palbociclib plasma concentrations profiles–solid lines in (a,c)—and absolute neutrophil count–solid lines in (b,d)—and their 95% prediction intervals–dotted lines. Orange and red dotted lines represent, respectively the threshold for neutropenia grade ≥3 and ≥4. NB: the dots in (a) represent the observed plasma palbociclib concentrations in a real patient included in the OpTAT study in both A and B1 scenarios. In scenario A, the concentrations are lower than the median observed in the simulated patients. This observation is consistent over this patient’s cycles and could be attributed to several extrinsic factors potentially increasing palbociclib elimination for this patient (e.g., co-administration of proton-pump inhibitors in fasting conditions [35], increase in creatinine clearance or decrease in alkaline phosphatase [36]).