Abstract 1145: Therapeutic drug monitoring of Sunitinib in real-world patients

Abstract Therapeutic Drug Monitoring (TDM) with model-based adaptive dosing is expected to quickly ensure that exposure levels of anticancer agents are in the adequate range of plasma concentrations. Here, we have compared such TDM + in silico approach with empirical changes in dosing performed upon clinical signs in patients with metastatic renal cell carcinoma treated with Sunitinib. A total of 31 patients were monitored in our institute. All patients were treated with a standard starting dose of 50 mg QD following the 4/2 schedule. Sunitinib and n-desethyl sunitinib trough levels were assayed at steady state by a fully validated LC-MS/MS analysis. A PK-pop model with Bayesian estimate implemented on Monolix helped to identify individual PK parameters from a single time-point. Two different target exposures were considered (i.e., trough levels comprised between 50 and 100 ng/ml or 0-24h AUC comprised between 1250 and 2150 ng/ml.h). The PK model enabled to simulate both trough levels and 0-24h AUC from a single sample, regardless of the sampling time. Overall, dosing was modified empirically in about 54% of the patients after treatment started, mostly (46%) for reducing the dosing after that toxicities showed. Clinical benefit (i.e., stable disease + partial response + complete reponse) was achieved in 54% of patients but treatment discontinuation was observed in 58% of the patients eventually, mostly because of severe side-effects. A relationship between baseline exposure levels and early onset treatment-related toxicities was found. Because of the various changes in dosing, no such relationship was found between baseline exposure levels and efficacy evaluated at 3 months. TDM showed that only 45% of the patients were in the right window for trough levels (i.e., 50-100 ng/ml) and only 26% when considering AUC as the target exposure (i.e., 1200-2150 ng/ml.h). Consequently, the PK/PD model would have suggested to rapidly customize dosing in a much larger part of the patients (i.e., 84% with respect to the target AUCs). Importantly, Sunitinib dosing was empirically reduced only in 41% patients who displayed early-onset severe toxicities, whereas modelling would have immediately proposed to cut the dosing in more than 80% of those patients, thus suggesting that the safety of Sunitinib could have been improved when using our model. Although performed on a limited number of patients, this real-world study supports the hypothesis that TDM associated with PK modelling could help the prescriber to identify the best Sunitinib dosing in a quicker way than empirical change in dosing. Citation Format: Laurent Ferrer, Jonathan Chauvin, Dr Jean-Laurent Deville, Joseph Ciccolini. Therapeutic drug monitoring of Sunitinib in real-world patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1145.