Development of a dried blood spot sampling method towards therapeutic monitoring of radotinib in the treatment of chronic myeloid leukaemia

Abstract

What is known and objective Dried blood spot (DBS) sampling is a minimally invasive method of blood sampling that enables monitoring of drug concentrations to be more convenient. This study aimed at developing a DBS sampling method for an accurate and precise prediction of radotinib plasma concentrations (C-p) in patients with chronic myeloid leukaemia (CML). Methods Dried blood spot and venous blood samples were simultaneously collected from fifty CML patients who had been receiving radotinib for at least a week. Radotinib concentrations were measured using a high-performance liquid chromatographic method with tandem mass spectrometric detection. Unmeasured C-p was predicted directly based on a Deming regression between DBS concentrations (C-DBS) and C-p. Unmeasured C-p was also predicted from C-DBS corrected by each patient's haematocrit (Hct). Both prediction methods were evaluated for their accuracy and precision using Deming regression and Bland-Altman analysis. Results and discussion The Deming regression equation between C-DBS and C-p was obtained as follows: C-p = 1.34 center dot C-DBS + 4.26 (r(2) = .97). C-p was directly predictable using C-p,C-pred1 = 1.34 center dot C-DBS + 4.26. With Hct correction, C-p was alternatively predictable using C-p,C-pred2 = C-DBS/ (1-Hct + Hct(2)). The slopes of Deming regression line between predicted and measured C-p were 0.99 and 1.02 for the direct and Hct-corrected method, respectively. The mean biases (accuracy) were -0.44% and 1.6% with the 95% limits of agreement (precision) of -22.4% to 21.5% and -20.5% to 23.7%, respectively. More than 93% of predicted and measured C-p pairs had their differences within 20% of the mean of each pair in both methods. What is new and conclusions Radotinib C-DBS are highly correlated with radotinib C-p. Radotinib C-p can be accurately and precisely predicted from C-DBS using direct or Hct-corrected prediction methods. Both appear to be appropriate for the therapeutic monitoring of radotinib in patients with CML.