Long-term Risk of Prostate Cancer Mortality Among Men with Baseline Prostate-specific Antigen Below 3 ng/ml: Evidence from the Finnish Randomized Study of Screening for Prostate Cancer

Abstract

BACKGROUND AND OBJECTIVE: Despite the evidence for prostate-specific antigen (PSA) screening reducing prostate cancer (PCa) mortality, the optimal PSA cutoff and the clinical significance of low initial PSA levels in predicting long-term PCa mortality remain subjects of ongoing research. We assessed PCa mortality among men with initial PSA levels below 3 ng/ml during the first screening round of the Finnish Randomized Study of Screening for Prostate Cancer (FinRSPC).METHODS: A retrospective cohort study was conducted, including 20 268 men from the screening arm of the FinRSPC with an initial PSA level of <3 ng/ml, with follow-up spanning up to 20 yr. Hazard ratios (HRs) and their 95% corresponding confidence intervals (CIs) were estimated using a Cox regression analysis.KEY FINDINGS AND LIMITATIONS: During a median follow-up of 17.8 yr, 1840 PCa cases were diagnosed and 128 PCa deaths occurred, with the cumulative PCa mortality of 0.6% and a mortality rate of four per 10 000 person-years. PCa mortality was five-fold higher at PSA levels of 2-2.99 ng/ml (HR 5.0, 95% CI 3.1-8.1) than at <1 ng/ml. Deaths from cases with Gleason score <7 and European Association of Urology low-risk group tumors showed a stronger association with PSA, particularly in the 2-2.99 ng/ml range versus <1 ng/ml. Additionally, PCa mortality in younger men (55-58 yr at entry) exhibited a stronger association with PSA than that in older men (67-71 yr at baseline). Addition of the cumulative number of PSA tests slightly improved the overall prediction of PCa death based on Harrell's C-statistic (base model 0.683 vs 0.717). The relatively small number of deaths, particularly among men with low-risk disease, may potentially limit the statistical precision of the results.CONCLUSIONS AND CLINICAL IMPLICATIONS: Our findings highlight the importance of a nuanced approach to PSA in PCa screening, suggesting utility for combining PSA with other tests at low levels and indicating minimal risk associated with discontinuing screening at ages 67-71 yr when PSA is low.PATIENT SUMMARY: In this study, we analyzed prostate cancer deaths in Finnish men with low initial prostate-specific antigen (PSA) levels. We found that the risk of prostate cancer death increases in relation to PSA, especially in younger men. Screening might safely be stopped at ages 67-71 yr if PSA remains low.

Background and objective: Despite the evidence for prostate-specific antigen (PSA)screening reducing prostate cancer (PCa) mortality, the optimal PSA cutoff and the clinical significance of low initial PSA levels in predicting long-term PCa mortality remain subjects of ongoing research. We assessed PCa mortality among men with initial PSA levels below 3 ng/ml during the first screening round of the Finnish Randomized Study of Screening for Prostate Cancer (FinRSPC).Methods : A retrospective cohort study was conducted, including 20 268 men from thescreening arm of the FinRSPC with an initial PSA level of <3 ng/ml, with follow-up spanning up to 20 yr. Hazard ratios (HRs) and their 95% corresponding confidence intervals (CIs) were estimated using a Cox regression analysis.Key findings and limitations: During a median follow-up of 17.8 yr, 1840 PCa cases were diagnosed and 128 PCa deaths occurred, with the cumulative PCa mortality of 0.6% and a mortality rate of four per 10 000 person-years. PCa mortality was five-fold higher at PSA levels of 2–2.99 ng/ml (HR 5.0, 95% CI 3.1–8.1) than at <1 ng/ml. Deaths from cases with Gleason score <7 and European Association of Urology low-risk group tumors showed a stronger association with PSA, particularly in the 2–2.99 ng/ml range versus <1 ng/ml. Additionally, PCa mortality in younger men (55–58 yr at entry) exhibited a stronger association with PSA than that in older men (67–71 yr at baseline). Addition of the cumulative number of PSA tests slightly improved the overall prediction of PCa death based on Harrell’s C-statistic (base model 0.683 vs 0.717). The relatively small number of deaths, particularly among men with low-risk disease, may potentially limit the statistical precision of the results.Conclusions and clinical implications: Our findings highlight the importance of anuanced approach to PSA in PCa screening, suggesting utility for combining PSA withother tests at low levels and indicating minimal risk associated with discontinuingscreening at ages 67–71 yr when PSA is low.Patient summary: In this study, we analyzed prostate cancer deaths in Finnish menwith low initial prostate-specific antigen (PSA) levels. We found that the risk of prostate cancer death increases in relation to PSA, especially in younger men. Screening might safely be stopped at ages 67–71 yr if PSA remains low.