Liquid vs dried blood matrices: Application to longitudinal monitoring of androstenedione, testosterone, and IGF-1 by LC-MS-based techniques

Dried blood spots have recently been approved by the World Anti-Doping Agency as an alternative biological matrix for testing of doping substances. However, their use is limited to the detection of non-threshold compounds without a Minimum Reporting Level due to the numerous issues related to quantitative analyses and the limitation on testing capabilities of a haemolysed matrix.

In this study androstenedione, testosterone and IGF-1 were longitudinally monitored in four different blood matrices to evaluate the potential of liquid capillary blood as an alternative matrix for quantitative determination in doping control analysis.

The analytical protocols developed to pretreat 20 μL of the blood matrices selected were based: i) for testosterone and androstenedione, on supported liquid extraction for liquid blood matrices, and on ultrasonication in the presence of methanol for dried blood matrices; ii) for IGF-1, proteins precipitation followed by evaporation of the supernatant was used to pretreat both liquid and dried blood matrices. The detection for all the target analytes was performed using liquid chromatography coupled to mass spectrometry. The analytical workflows, once optimized, were fully validated according to the requirements of World Anti-Doping Agency and ISO 17025 standard and used for the analysis of venous (serum) and capillary (liquid plasma and dried whole blood collected using either volumetric or non-volumetric devices) blood samples collected from 7 healthy subjects.

The validation results showed satisfactory performance as related to specificity, sensitivity, matrix effects, linearity, accuracy, and precision in all the blood matrices evaluated despite the limited volume of sample used. The analysis of the different blood matrices collected from the subjects showed non-significant differences between the levels of testosterone and androstenedione measured in dried (fixed volume collected) and liquid matrices. An acceptable underestimation (lower than 15 %) was observed in capillary plasma compared to venous serum. The testosterone/androstenedione ratio was similar in all the blood matrices considered (bias lower than 5 %), indicating this parameter was not affected by either the blood matrix or collection device selected. For IGF-1, the levels measured in liquid blood matrices differed significantly (bias higher than 20 %) from those measured in dried whole blood matrices, suggesting haemolyzed blood might represent a challenge for the determination of macromolecules, mainly due to the complexity of the whole blood matrix in comparison to plasma/serum.

The outcomes of our study suggest that liquid capillary blood might open new avenues to blood microsampling in doping control field. It represents an efficient alternative to overcome the issues related to venous blood and dried blood spot sampling. Furthermore, it also allows greater frequency of blood sampling, with minor discomfort and without needing a phlebotomist, for analyses that can only be performed in blood samples, with an increased probability to detect and report Adverse Analytical Finding.