Bevacizumab treatment is subject to large interpatient variability in efficacy, which may partly be explained by differences in complex bevacizumab pharmacokinetic characteristics that influence bevacizumab exposure. Exposure-response relationships have been identified for other monoclonal antibodies. We aimed to identify possible exposure-survival relationships in bevacizumab-treated patients with metastatic colorectal cancer (mCRC).

Patients with mCRC who started first-line bevacizumab-based chemotherapy between July 2012 and July 2014, and from whom serial blood samples and survival were prospectively collected, were included. Follow-up was carried out until July 2018. Total bevacizumab trough concentrations were measured from cycle 2 to cycle 30 of treatment. The receiver operating characteristic (ROC) curve analysis and Cox analysis were used to identify the relationship between concentrations and overall survival (OS). In addition, OS was compared between different trough concentration groups.

One hundred fifty-seven blood samples from 46 patients were evaluable for analyses. ROC analysis showed a clear separation in survival based on trough levels (area under the curve = 0.739, p = .009). Cox regression also showed a strong positive correlation between trough levels and survival (p = .0004). Three distinct groups of exposure were identified: low (median trough concentration [C ] ≤41.9 mg/L); medium (C 43-87.2 mg/L) with median OS of 12.8 and 36 months, respectively (p = .0003); and high (C ≥7.9 mg/L), where the majority of patients were still alive 60 months after the initiation of treatment.

This study shows that survival was proportional to the magnitude of exposure in patients with mCRC. Further clinical research should focus on clarifying these exposure-outcome relationships in order to optimize dosing.

Bevacizumab-based chemotherapy is standard first-line treatment in metastatic colorectal cancer. Moreover, bevacizumab presents complicated pharmacokinetics, and in many cases, clinical outcomes can be highly variable, with some patients responding remarkably well and others not. This study's results show that patients who experienced longer overall survival also had significantly higher exposure to bevacizumab. Therefore, bevacizumab trough concentrations could be used both as a predictive biomarker and as a tool for treatment monitoring and optimization. Finally, the development of validated, rapid, and sensitive assays for bevacizumab concentration measurements in combination with these results may lead to a therapeutic drug monitoring-guided approach in bevacizumab treatment with better clinical outcomes.