Tamoxifen, a selective estrogen receptor modulator, is currently the standard-of-care adjuvant treatment of breast cancer. Tamoxifen is a prodrug and particularly exerts its effect through its most active metabolite endoxifen. Cytochrome P450 (CYP) enzymes, in particular CYP2D6, convert tamoxifen to endoxifen. Polymorphisms in the CYP2D6 gene can hamper CYP2D6 activity and subsequently lead to decreased concentrations of endoxifen. Madlensky et al. found a direct association between endoxifen concentrations and breast cancer recurrence in a retrospective cohort. Patients with endoxifen concentrations below 16 nmol/L had a 30% higher risk of breast cancer recurrence than patients with endoxifen concentrations above this threshold. Madlensky et al. also found that CYP2D6 intermediate- and poor metabolizer phenotypes were associated with endoxifen levels below the 16 nmol/L threshold. The association between CYP2D6 phenotypes and endoxifen levels has since been confirmed by several other studies. In several retrospective studies, approximately 20-24% of tamoxifen patients do not reach the 16 nmol/L endoxifen threshold at steady state. Therapeutic drug monitoring (TDM) could be used to increase the probability of reaching this threshold to 89% after 6 months. With TDM, the dose is corrected after reaching steady state and patients are often only adequately treated after 3 to 6 months. To counter this problem and predict the correct tamoxifen dose at baseline, model-informed precision dosing (MIPD) could be used. In prior research at the Erasmus MC a population-pharmacokinetic (POP-PK) model has been developed. POP-PK-modeling is a mathematical modeling technique that describes the pharmacokinetics of a drug for each individual based on patient characteristics. A POPPK model can describe and predict the absorption, distribution, metabolism and elimination of a drug in the body and predict blood concentration-time profiles prior to actual administration of the drug. In previous, not yet published research we have developed a POP-PK model to describe tamoxifen and endoxifen pharmacokinetics. In this model we have evaluated the activity of different single nucleotide polymorphisms (SNP's) on a continuous scale. In addition the concomitant administration of CYP3A4 and CYP2D6 inhibitors influenced endoxifen formation. Whereas age significantly influenced tamoxifen clearance, BMI and height affected the endoxifen formation rate and tamoxifen clearance respectively.
After careful retrospective validation the validity of our model can be tested by prospectively predicting the best dose for each patient. Using Monte-Carlo simulations we estimated that when using the standard dose of 20 mg tamoxifen, 23% of all patients will not reach endoxifen steady-state concentration \>16 nM. Using model-informed precision dosing, the proportion of patients that reach steady-state endoxifen concentrations above 16 nmol/L will be 91%. Out of these final 9%, 66% of all patients will not reach 16 nM using the highest registered dose of 40 mg. If the POP-PK model could adequately identify this patient group, that will not reach the 16 nM threshold with the highest prescribed dose of 40 mg, they could in the future be treated differently from the start of adjuvant therapy. An example of this are aromatase inhibitors.
The primary aim of this study is to increase the proportion of patients that reach an endoxifen level of 16 nM after reaching steady state endoxifen plasma concentrations using MIPD. In this study we will be prospectively validating a POP-PK model and evaluate the feasibility of MIPD for routine clinical use.
