Abstract
Paclitaxel is an anticancer agent efficacious in various tumors. There is large interindividual variability in drug plasma concentrations resulting in a wide variability in observed toxicity in patients. Studies have shown the time the concentration of paclitaxel exceeds 0.05 µM is a predictive parameter of toxicity, making dose individualization potentially useful in reducing the adverse effects. To determine paclitaxel drug concentration, a venous blood sample collected 24 h following the end of infusion is required, often inconvenient for patients. Alternatively, using a microsampling device for self-sampling would facilitate paclitaxel monitoring regardless of the patient's location. We investigated the feasibility of collecting venous and capillary samples (using a Mitra(®) device) from cancer patients to determine the paclitaxel concentrations. The relationship between the venous plasma and whole blood and venous and capillary blood (on Mitra(®)) paclitaxel concentrations, defined by a Passing-Bablok regression, were 0.8433 and 0.8569, respectively. Demonstrating a clinically acceptable relationship between plasma and whole blood paclitaxel concentration would reduce the need to establish new target concentrations in whole blood. However, in this study, comparison of venous and capillary blood using Mitra(®) for sampling displayed wide confidence intervals suggesting the results from the plasma and whole blood on this device may not be interchangeable.