False negative results in testosterone doping in forensic cases: Sensitivity of the urinary detection criteria T/E and T/LH

At the Swedish national forensic toxicology laboratory, a measured testosterone/epitestosterone (T/E) ratio ≥ 12 together with testosterone/luteinizing hormone (T/LH) in urine > 400 nmol/IU is considered as a proof of exogenous testosterone administration. However, according to the rules of the World Anti-Doping Agency (WADA), samples with T/E ratio > 4 are considered suspicious and shall be further analysed by gas chromatography-combustion-isotope ratio mass spectrometry (GC-C-IRMS) to confirm the origin of testosterone and its metabolites. The aim of this study was to investigate the possibility of false negative results and to estimate the frequency of negative results using the current criteria for detection of abuse of testosterone in forensic investigations. Urine and serum samples were collected by the police at suspected infringement of the doping law in Sweden. Fifty-eight male subjects were included in the study. Urinary testosterone was determined by gas chromatography-mass spectrometry (GC-MS), serum testosterone and LH-by immunoassay. The origin of testosterone and its metabolites was confirmed by means of GC-C-IRMS. Twenty-six of the 57 analysed subjects tested positive for exogenous testosterone using the criteria T/E ≥ 12 combined with T/LH > 400 nmol/IU. The IRMS analyses confirmed 47 positives; thus, 21 were considered false negatives. Negative predictive value was 32% (95% confidence interval [CI]: 16%-50%) and sensitivity 55%. No false positive subjects were found. The number of false negative cases using the current criteria for the detection of testosterone abuse and hence the low sensitivity indicates a need to discuss introduction of new strategies in forensic doping investigations.