Abstract
BACKGROUND: The α2,3-sialyl N-glycosylated free prostate-specific antigen ratio (S2,3PSA%) was approved in Japan as a prostate cancer (PCa) diagnostic test. We evaluated the analytical characterization and real-world diagnostic performance of S2,3PSA%. METHODS: The precision testing, dilution linearity, measurement sensitivity, preanalytical stability, and interferences of S2,3PSA, α2,6-sialyl N-glycosylated free PSA (S2,6PSA), and S2,3PSA% were performed. The diagnostic accuracy detecting PCa of S2,3PSA% was prospectively evaluated in 253 men (Cohort 1, vs. PI-RADS) and in 145 men (Cohort 2, vs. PI-RADS, prostate health index, phi) who scheduled MRI-targeted biopsy by area under the receiver operating characteristics curve (AUC). RESULTS: The precision of the S2,3PSA, S2,6PSA, and S2,3PSA% were all < 3.4% coefficient of variation. The dilution linearity of S2,3PSA had a correlation coefficient of 0.9949-0.9987. The detection limit of S2,3PSA and S2,6PSA was 0.044 and 0.029 ng/mL, respectively. Serum S2,3PSA and S2,6PSA concentrations were stable at 6°C for 24 h and -20°C for 90 days, while S2,3PSA% was unchanged at 6°C and -20°C for 90 days or under five freeze-thaw cycles. Serum S2,3PSA and S2,3PSA% were not affected by any interferences and drugs. In Cohort 1, AUC of S2,3PSA% (0.776, 95% CI 0.719-0.832) detecting PCa was comparable to that of PI-RADS (0.746, 0.685-0.807, p = 0.7996). In Cohort 2, AUC of S2,3PSA% detecting PCa (0.837, 0.774-0.901) was comparable to PI-RADS (0.779, 0.703-0.854, p = 0.3037), and phi (0.867, 0.809-0.926, p = 0.3000). CONCLUSIONS: Analytical characteristics of S2,3PSA% perform well and the diagnostic performance of S2,3PSA% was comparable to phi and MRI.