An Untargeted Urine Metabolomics Approach for Autologous Blood Transfusion Detection

Purpose Autologous blood transfusion is performance enhancing and prohibited in sport but remains difficult to detect. This study explored the hypothesis that an untargeted urine metabolomics analysis can reveal one or more novel metabolites with high sensitivity and specificity for detection of autologous blood transfusion. Methods In a randomized, double-blinded, placebo-controlled, crossover design, exercise-trained men (n = 12) donated 900 mL blood or were sham phlebotomized. After 4 wk, red blood cells or saline were reinfused. Urine samples were collected before phlebotomy and 2 h and 1, 2, 3, 5, and 10 d after reinfusion and analyzed by ultraperformance liquid chromatography-quadrupole time-of-flight mass spectrometry. Models of unique metabolites reflecting autologous blood transfusion were attained by partial least-squares discriminant analysis. Results The strongest model was obtained 2 h after reinfusion with a misclassification error of 6.3% and 98.8% specificity. However, combining only a few of the strongest metabolites selected by this model provided a sensitivity of 100% at days 1 and 2 and 66% at day 3 with 100% specificity. Metabolite identification revealed the presence of secondary di-2-ethylhexyl phtalate metabolites and putatively identified the presence of (iso)caproic acid glucuronide as the strongest candidate biomarker. Conclusions Untargeted urine metabolomics revealed several plasticizers as the strongest metabolic pattern for detection of autologous blood transfusion for up to 3 d. Importantly, no other metabolites in urine seem of value for antidoping purposes.

Automated summary

This study found that an untargeted urine metabolomics analysis can reveal the metabolic pattern of autologous blood transfusion, providing an effective detection method. The strongest model was obtained 2 hours after reinfusion, and combining only a few of the strongest metabolites selected by this model provided high sensitivity and specificity for detection.