Background: The aim of this study was to determine the utility of dried blood spots (DBS) compared with conventional plasma collection methods for characterization of efavirenz pharmacokinetics, in the setting of a large-scale, global clinical trial (ENCORE1).
Methods: Six hundred thirty patients were recruited from 38 sites and had single matched whole blood DBS and plasma samples (mid-dose interval) taken at weeks 4 and 12 of treatment. In addition, a subgroup of patients underwent intensive DBS and plasma sampling (0-24 hours) to provide full-profile data for pharmacokinetic parameters. Efavirenz concentrations were determined by validated high-performance liquid chromatography-mass spectrometry methods. A DBS-predicted plasma concentration was derived and linear regression and Bland-Altman plots were used to compare DBS-predicted plasma concentrations with that of measured plasma concentrations.
Results: Efavirenz DBS and plasma concentrations were significantly correlated (R = 0.904, P < 0.001; n = 1094), and DBS concentrations were, on average, 53% ± 9.5% lower than plasma. In the main study, the DBS-predicted plasma values significantly underestimated the true measured concentration of efavirenz in plasma; the mean difference (95% confidence interval) between efavirenz DBS-predicted concentrations and measured plasma concentrations was -0.451 mg/L (-0.504 to -0.398) at week 4 (n = 561). However, in the intensive study, the mean difference was only 0.086 mg/L (-0.006 to 0.178) at 12 hours after dose (n = 46) and was not statistically significant.
Conclusions: Our data show a high correlation between measurements of efavirenz concentrations in plasma and in DBS. However, DBS concentrations significantly underestimated the true measured plasma concentrations in the sparse samples taken in this large multinational ENCORE1 trial.