Midlife baseline prostate-specific antigen, velocity, and doubling time association with lethal prostate cancer and mortality

Background: Midlife baseline prostate-specific antigen (MB PSA), defined as a single PSA value measured between 40-59 years of age, has been proposed as a tool that can limit potential harms of PSA screening. This study aimed to examine the ability of MB PSA versus PSA doubling time (PSADT) and PSA velocity (PSAV) in assessing the likelihood of developing of lethal prostate cancer (PCa) in a diverse and contemporary North American population.

Methods: Men 40-59 years old, who received their first PSA between the years 1995 and 2019, were included. For MB PSA values, the first PSA test result was included. For PSADT, the first two PSA test results were included. For PSAV, the first three PSA test results within 30 months were included. Selection criteria resulted in a total of 77,594 patients with at least two PSA test results and 11,634 patients with at least three PSA test results. Multivariable Fine-Gray regression was used to examine the impact of the value of the PSA testing methods on the development of lethal PCa (defined as death from PCa or development of metastatic disease either at diagnosis or during follow-up). Time-dependent receiver operating characteristic/area under the curve (AUC) at 5, 10, and 15 years were plotted.

Results: In the main cohort, patients were most frequently in the 50-54 age category (32.8%), had a Charlson comorbidity index of 0 (70.5%), and were White (63.2%). Of these, 9.3% had the midlife baseline PSA in the top 10th percentile, and 0.4% had a PSADT 0-6 months. Lethal PCa was diagnosed in 593 (0.8%) patients. The median (interquartile range) time to lethal PCa was 8.6 (3.2-14.9) years. In the main cohort, MB PSA and PSADT showed significant associations with the occurrence of lethal PCa, with a hazard ratio (HR) of 6.10 (95% confidence interval [CI], 4.85-7.68) and HR of 2.20 (95% CI, 1.07-4.54) for patients in the top 10th percentile MB PSA group and in the PSADT between 0 to <6 months group, respectively. In patients with three PSA results available, MB PSA and PSAV showed significant associations with the occurrence of lethal PCa, with a HR of 3.95 (95% CI, 2.29-6.79) and 3.57 (95% CI, 2.17-5.86) for patients in the top 10th percentile MB PSA group and in the in the PSAV >0.4 ng/mL/year group, respectively. PSADT and PSAV did not exhibit higher AUCs than MB PSA in assessing the likelihood of lethal PCa. Specifically, they were 0.818 and 0.708 at 10 and 15 years, respectively, for the PSADT; 0.862 and 0.756 at 10 and 15 years, respectively, for the PSAV; and 0.868 and 0.762 at 10 and 15 years, respectively, for the MB PSA (all p > .05).

Conclusions: The study findings are that PSAV or PSADT were not superior to midlife baseline in assessing the likelihood of developing lethal PCa. This suggests that these variables may not have practical use in enhancing PSA screening strategies in a clinical setting.