Aurelia H.M. de Vries Schultink (1), Thomas P.C. Dorlo (1), Jos H. Beijnen (1,2), Alwin D.R. Huitema (1,3)
(1) Antoni van Leeuwenhoek – The Netherlands Cancer Institute, Amsterdam, the Netherlands, (2) Utrecht Institute for Pharmaceutical Sciences, Utrecht University, Utrecht, the Netherlands, (3) University Medical Centre Utrecht, Utrecht University, Utrecht, the Netherlands
Objectives:
Five years of adjuvant treatment with tamoxifen lowers estrogen receptor (ER)-positive breast cancer recurrence and mortality rates. Despite tamoxifen’s effectiveness, resistance to treatment occurs. Variability in response has been attributed to variability in pharmacokinetics (PK), more specifically to variability in endoxifen concentrations, an active metabolite of tamoxifen. Both endoxifen concentrations and the CYP2D6 genotype, responsible for bio-activation of tamoxifen, have been proposed as predictive markers for recurrence. Therapeutic Drug Monitoring (TDM) of endoxifen seems the best way forward to tailor tamoxifen treatment, since variability in concentration of endoxifen can only partially be attributed to CYP2D6 genotypes. An endoxifen concentration >5.97 ng/mL has been associated with 26% lower risk of breast cancer recurrence, in a retrospective analysis (1). This target can be applied to tailor tamoxifen treatment, recommending an increase in tamoxifen dose from 20 mg to 40 mg daily if endoxifen concentrations are below this threshold. In order to evaluate the feasibility of a prospective validation of this threshold, a PK model that represents clinical practice is needed. Models describing the PK of tamoxifen and its metabolites are sparse and based on data from trials only. However, TDM is implemented in clinics, and might have different proportions of patients reaching the target. Therefore, the aim of this analysis is to evaluate target attainment based on endoxifen TDM applied in a clinical cohort and compare to model-based predictions from a PK model based on trial data.
Methods:
Breast cancer patients treated with tamoxifen in the adjuvant setting and for whom TDM of endoxifen was applied in the Netherlands Cancer Institute, were included. Samples were drawn >3 months after start of treatment and after dose adaptation, ensuring steady state. Since PK data in this population was sparse, target attainment before and after dose adaption was compared to predicted proportions based on a previously reported 4-compartment PK model (2). We assumed an underlying CYP2D6 phenotype distribution with 52.5% extensive/ultra-rapid, 45% intermediate and 2.5% poor metabolizers, as previously described (3), other covariates were imputed as the median. Subsequently, 10,000 patients were simulated using a dose adaptation simulation script (R version 3.3.1) including inter-individual variability and residual error. Dose was adapted based on a sample at day 90 and reevaluated on day 180.
Results:
In total, 976 samples of 713 patients were available, of which 658 patients had a first sample taken during treatment of 20 mg tamoxifen daily. In this group, 203 patients (33%) did not reach the target concentration. Of these 203 patients, 120 patients received a dose increment to 30 or 40 mg depending on the measured endoxifen concentration, of which 82 patients (68.3%) reached the target at the second sample. The simulation analysis demonstrated that 32.6% of the simulated patients did not reach the TDM target of 5.97 ng/mL. This was in accordance with the proportions in clinical practice (33%). However, after dose adaption to 40 mg of the initial patients below target, 85.1% simulated patients reached the target, compared to only 68.3% of patients who received a similar dose increment in the clinical setting. In total, after applying TDM to the clinical cohort, 73% of all patients attain the target at the second sample, compared to 88.4% in the simulated cohort. Since only a part of the patients in the clinical cohort received a dose increment, a sensitivity analysis was conducted. This analysis showed that if all patients with below-target endoxifen concentrations would have received a dose increment, 83.1% compared to 73% of all patients in the clinical cohort would reach therapeutic levels of endoxifen.
Conclusions:
This analysis demonstrated that the effect of TDM on endoxifen target attainment is over predicted by considering only trial data. Around 15% of the subjects are falsely assumed to reach the target. This is partly caused by the assumption that all below-target patients receive a dose increment, though side effects hamper this in the clinical setting. Additionally, patients included in a trial are assumed to be more adherent to therapy than patients in clinical practice. These findings should be considered when evaluating feasibility of prospective validation of TDM for endoxifen.
References:
[1] Madlensky L. et al. Clin Pharmacol Ther (2011) 89(5),718–25.
[2] Dahmane EBA et al. 2013;1–184. Available from: http://archive-ouverte.unige.ch/unige:33429
[3]Ter Heine R. et al. Br J Clin Pharmacol (2014) 78(3),572–86.
Reference: PAGE 27 (2018) Abstr 8648 [www.page-meeting.org/?abstract=8648]
Poster: Drug/Disease Modelling - Oncology