Abstract:

Objective To investigate the diagnostic efficacy of total prostate-specific antigen （t-PSA），the percentage of free prostate-specific antigen（f-PSA）to t-PSA （f-PSA/t-PSA%），prostate health index （PHI） and prostate imaging reporting and data system （PI-RADS） score in the diagnosis of prostate cancer. Methods Totally，277 patients with prostate diseases diagnosed by pathological examination from July 2022 to February 2024 in the First Hospital of Jilin University were classified into benign disease group （143 cases） and prostate cancer group （134 cases）. They were classified into t-PSA < 20 ng·mL-1 group （147 cases，including 100 cases of benign disease and 47 cases of prostate cancer） and t-PSA ≥ 20 ng·mL-1 group （130 cases，including 43 cases of benign disease and 87 cases of prostate cancer） according to t-PSA level. The serum t-PSA，f-PSA and isoform[-2] prostate-specific antigen（p2PSA）of all the patients were determined，and the f-PSA/t-PSA% was calculated. The PI-RADS score was evaluated. Logistic regression analysis was used to assess the factors for prostate cancer. The diagnostic efficacy of each indicator was evaluated by receiver operating characteristic （ROC） curve. Results The t-PSA and PHI in prostate cancer group were higher than those in benign disease group （P<0.001），and f-PSA/t-PSA% was lower than that in benign disease group （P<0.001）. The proportion of PI-RADS score ≥4 was higher than that of benign disease group （P<0.001）. The areas under curves（AUC）of t-PSA，f-PSA/t-PSA%，PHI and PI-RADS for the diagnosis of prostate cancer were 0.752，0.633，0.913 and 0.881，respectively. PHI and PI-RADS score were independent risk factors for prostate cancer [odds ratios （OR） were 1.035 and 2.520，95% confidence intervals （CI） were 1.021-1.049 and 1.672-3.800，respectively，P<0.001]. In t-PSA <20 ng·mL-1 group，compared with benign disease patients，PHI and PI-RADS score≥4 proportion in prostate cancer patients were increased （P<0.001），f-PSA/t-PSA% was decreased （P<0.05），and there was no statistical significance in t-PSA level （P>0.05）. In t-PSA≥20 ng·mL-1 group，compared with benign disease patients，t-PSA，PHI and PI-RADS score≥4 proportion in prostate patients were increased （P<0.001），while f-PSA/t-PSA% had no statistical significance （P>0.05）. The AUC of t-PSA，f-PSA/t-PSA%，PHI and PI-RADS in the diagnosis of prostate cancer with t-PSA < 20 ng·mL-1 were 0.561，0.611，0.836 and 0.834，respectively. The AUC of the combined determination was the highest （0.900）. The AUC for the diagnosis of t-PSA≥20 ng·mL-1 prostate cancer were 0.822，0.520，0.963 and 0.875，respectively. The AUC of t-PSA+f-PSA/t-PSA%+PHI and t-PSA+f-PSA/t-PSA%+PHI+PI-RADS score were 0.970 and 0.972，respectively. Conclusions For patients with t-PSA< 20 ng·mL-1，the combined determination of t-PSA，f-PSA/t-PSA%，PHI and PI-RADS score can improve the determination rate of prostate cancer. For patients with t-PSA≥20 ng·mL-1，PHI is a good indicator for diagnosing prostate cancer，and the combined determination of t-PSA，f-PSA/t-PSA% and PHI can improve diagnostic efficacy.

Key words: Total prostate-specific antigen, Free prostate-specific antigen, Prostate health index, Prostate imaging reporting and data system score, Prostate cancer