A-032 Accuracy of Immunoassay Testosterone Assays in Low Testosterone Patients

M Deebajah,S Patel,M Keating,O Palmer,S E Wheeler

DOI: https://doi.org/10.1093/clinchem/hvae106.032

IF: 12.114

2024-10-27

Clinical Chemistry

Abstract:BackgroundThe measurement of testosterone in patients expected to have low testosterone levels can present a challenge to the clinical laboratory for turn-around-time and cost. At our institution send-out for measurement by mass spectrometry (MS) is the current standard in patients for whom we expect low testosterone levels. We aimed to investigate the utility of three immunoassay methodologies as a first line screen for testosterone levels by comparing results to mass spectrometry with an emphasis on specimens with low testosterone levels.Methods48 remnant plasma samples each from healthy men and women were aliquoted and stored at -80°C within 24 hours of draw (n=96 total specimens). Samples are remnants from out-patients being drawn for lipid profiles or other tests indicative of an annual wellness visit. Immunoassays were performed at the University of Pittsburgh Medical Center within the CAP accredited clinical laboratories. This study was approved by the University of Pittsburgh Institutional Review Board. The assays assessed were Access® Testosterone, Beckman Coulter (B-TST); Elecsys® Testosterone II, Roche Diagnostics (R-TST); Atellica® IM Testosterone II, Siemens Healthineers (S-TST). MS was sent to ARUP for testosterone analysis. The CDC Hormone Standardization program (HoSt) provides information of the accuracy of testosterone assays for participants. Both the S-TST and ARUP MS assays participate in this program.ResultsResults by MS ranged from 0.07 to 27.31 nmol/L. Correlation between testosterone by MS and immunoassays was strong with R2 >0.95 for all methods (B-TST: R2= 0.9570, y= 0.8588x + 0.9431; R-TST: R2= 0.9797, y= 1.085x + 0.3119; S-TST: R2= 0.9556, y= 1.124x + 0.0917). Assessment of specimens with low concentrations (< 3 nmol/L by MS, n=41) found that correlations were similarly poor between immunoassay methods and highest for B-TST (B-TST: R2= 0.5883, y= 1.259x + 0.3866; R-TST: R2= 0.4460, y= 0.9681x + 0.2760; S-TST: R2= 0.5449, y= 0.6352x + 0.3202). The root mean squared error (RMSE) was also calculated for the low concentration specimens and was significantly lower for S-TST than B-TST or R-TST (B-TST: 0.6708 nmol/L; R-TST: 0.6872 nmol/L; S-TST: 0.3698 nmol/L).ConclusionsIn patients where we expect low testosterone levels, MS continues to be the standard of care and immunoassays do not compare robustly at levels below 3 nmol/L. However, in cases where clinical decision making requires faster turn-around-time screening by immunoassay may be feasible. Clinical laboratories should be aware of the performance of their testosterone immunoassay methodologies at low testosterone concentrations as performance may vary between manufacturers. Manufacturer participation in the CDC HoSt program may provide additional performance information for clinical laboratories. Additional work is needed to assess performance across a larger cohort and additional immunoassay manufacturers.

medical laboratory technology