Eva Helmer 1, Nadja Haas 1, Olga Teplytska 1, Fenja Klima 2,3, Jonas Huhn 4, Georg Hempel 5, Charlotte Kloft 2,3, Oliver Scherf-Clavel 4, Ulrich Jaehde 1
1 Department of Clinical Pharmacy, Institute of Pharmacy, University of Bonn (Bonn, Germany), 2 Freie Universität Berlin, Institute of Pharmacy, Department of Clinical Pharmacy and Biochemistry (Berlin, Germany), 3 Graduate Research Training Program PharMetrX (Berlin/ Potsdam, Germany), 4 Department of Pharmacy, Faculty of Chemistry and Pharmacy, Ludwig Maximilian University of Munich (Munich, Germany), 5 Institute of Pharmaceutical and Medicinal Chemistry – Clinical Pharmacy, University of Muenster (Muenster, Germany)
INTRODUCTION
Oral anticancer drugs (OADs) play a vital role in modern cancer therapy. However, they require a high degree of personal responsibility from patients with regard to taking their medication. Since up to 80% of OADs exhibit an exposure-efficacy and/or exposure-toxicity relationship, adherence is of crucial importance [1].
Based on the ON-TARGET study, the influence of non-adherence on the estimation of pharmacokinetic (PK) parameters of cabozantinib was investigated.
METHODS
The ON-TARGET study investigated renal cell carcinoma patients receiving axitinib or cabozantinib in a prospective, multicenter setting [1]. Patients documented their daily medication intake in diaries covering 28 days. Blood samples were taken at the study centers and the prescribed therapy was documented. Overall and daily adherence were calculated based on the diaries and the therapy documented at the study center with regard of documented therapy breaks. For patients exhibiting non-adherent behavior within the 15 days prior to blood sampling, individual PK parameters were estimated using a previously published PK model with post-hoc Bayesian estimation [2]. PK parameters were derived either (1) assuming idealized 100% adherence or (2) incorporating diary-corrected observed adherence. In addition, average daily adherence till the day of blood sampling (scaled 0-1) was evaluated as a continuous, time-dependent covariate on apparent clearance (CL/F) within the population PK model. The covariate effect was considered statistically significant if inclusion resulted in a reduction of the objective function value (OFV) of ≥ 3.84, corresponding to p < 0.05.
RESULTS
Among the 40 enrolled patients 20 patients received cabozantinib and submitted at least one diary. Of these, 16 were male and four were female with a mean age of 70.65 years (± 13.04). In total, 232 cabozantinib diaries were submitted between March 24, 2021, and December 29, 2024, with a median of 6 diaries per patient (range: 2–43). Overall, 4,560 diary entries were included in the adherence analysis. The daily adherence per patient ranged from 76.5% to 100%, with a median of 100%. Thirteen patients demonstrated 100% adherence, four patients had adherence between 90.0% and 99.9%, and one patient showed adherence below 90%. Comparison of individual PK estimates in non-adherent patients showed that adherence-informed predictions yielded plasma concentrations that were closer to the observed values. Incorporating non-adherence reduced the relative mean prediction error (MPE) by 3,03% (absolute MPE reduction of 21.14 ng/mL), resulting in concentrations that were, on average, 4.12% (23.55 ng/mL) closer to the measured values. Compared with the assumption of ideal adherence, diary-corrected modeling resulted in lower estimates of apparent clearance (CL/F) and apparent volume of distribution (V/F), with mean decreases of 0.0403 L/h or 1,52% and 5.19 L or 2.95%, respectively. Covariate analysis further demonstrated that incorporating non-adherence as a continuous, time-dependent covariate on clearance resulted in a reduction of the OFV, indicating improved model fit.
CONCLUSIONS
In conclusion, adherence among patients with renal cell carcinoma receiving cabozantinib in the ON-TARGET study was generally high. The collection of 232 diaries demonstrated the successful implementation of this indirect adherence measure. Accounting for non-adherence improved the accuracy of PK predictions and may help explain discrepancies between predicted and observed plasma concentrations in the context of therapeutic drug monitoring.
References:
1. Mc Laughlin AM et al. Cancers (Basel) 2021; 13(24):6281
2. Nguyen L et al. J Clin Pharmacol 2019; 59(11):1551-1561
Reference: PAGE 34 (2026) Abstr 11956 [www.page-meeting.org/?abstract=11956]
Poster: Drug/Disease Modelling - Oncology