Abstract
INTRODUCTION: Prostate-specific antigen (PSA) can circulate bound to extracellular vesicles (EVs) and its measurement (ev-PSA) can be useful in prostate cancer. Although not designed with that purpose, total PSA assays react with ev-PSA. We evaluated the analytical performance of several total PSA assays in ev-PSA quantification and the impact of ev-PSA on total PSA measurement. MATERIALS AND METHODS: Extracellular vesicles were isolated from 83 serum samples from prostate cancer patients by size exclusion chromatography or ultracentrifugation. PSA was quantified in serum, EVs, International Standard for PSA 17/100 from the World Health Organization (WHO IS 17/100) and exosomes from lymph node carcinoma of the prostate (LNCaP) cell line, using commercial immunoassays (Elecsys, Atellica, Immulite, Liaison and Kryptor). RESULTS: Nanoparticle tracking analysis showed that the WHO IS 17/100 contains significantly less EVs than serum (P < 0.001). The sensitivity to detect ev-PSA followed this order: Elecsys ~ Atellica > Immulite > Liaison > Kryptor. Ev-PSA could be detected in all serum samples with Elecsys and Atellica, but not with Immulite (87.8%), Liaison (58.5%) or Kryptor (48.8%). Bland-Altman analysis showed a proportional bias in ev-PSA quantification between Elecsys and other methods. Addition of ev-PSA to serum samples caused a proportional bias in PSA measurement between Elecsys and Immulite methods, with a relationship (r(2) = 0.99; P < 0.001) between ev-PSA and the difference in total PSA concentration between both methods. CONCLUSIONS: While ev-PSA can be measured using commercial kits, notable differences exist between methods, which could lead to potential discrepancies in serum total PSA results across various assays.