Furthermore, based on the 2017 USPSTF update, there is still a Grade D recommendation for men 70 years and older undergoing PSA-based screening for prostate cancer. The objective of this study was to determine if a PSA level in a man between the ages of 60-70 predicted future lethal prostate cancer, potentially allowing low-risk men to be exempt from future PSA screening.
The authors conducted a nested case-control study within the Physicians’ Health Study of primarily white, US male physicians initiated in 1982. Men 60 (57.5‒62.5), 65 (62.5‒67.5), and 70 (67.5‒72.5) years of age who gave blood before enrollment in this initiative were eligible. Baseline PSA levels were available for 110 lethal prostate cancer cases that were matched to 330 age-matched controls or non-lethal prostate cancer cases. Lethal prostate cancer was defined as a composite outcome defined as metastatic (to bones or distant organs) or fatal prostate cancer. Conditional logistic regression was used to estimate odds ratios (ORs) for the association between PSA and risk of lethal disease. For the control group, median PSA was 1.10 ng/mL for men aged 60, 1.51 ng/mL for men aged 65, and 1.52 ng/mL for men aged 70. Additionally, the 90th percentile of PSA levels among controls was 3.97 ng/mL for men aged 60, 5.38 ng/mL for men aged 65, and 5.17 ng/mL for men aged 70. Among lethal cases, the median time from blood draw to lethal event was 15.6 years. Risk of lethal prostate cancer was strongly associated with baseline PSA levels: for men aged 60, the OR comparing PSA in the >90th percentile vs. ≤ median was 14.2 (95%CI 5.2-39.1), OR 27.9 (95%CI 5.5-142.3) for men aged 65, and OR 25.5 (95%CI 3.4-120.5) for men aged 70. Interestingly, 87% of lethal cases across all age groups were in men with baseline PSA above the median, and 70% had PSA above the 75th percentile. The strength of this study is the baseline PSA draw many years prior to lethal disease, however a limitation is the low number of lethal events per age bracket, resulting in wide confidence intervals on multivariable modeling.
In summary, the authors conclude that a pre-diagnostic PSA level at age 60‒70 strongly predicts future risk of lethal prostate cancer in a cohort of men subject to opportunistic screening. This may support risk-stratified screening, with consideration of exempting men in this age bracket with PSA below the median from further screening. However, men in the upper PSA percentiles with long life expectancy are at significant risk for lethal prostate cancer and should be monitored accordingly.
Presented By: Mark A. Preston, MD, MPH, Brigham and Women’s Hospital, Boston, MA, USA
Co-Authors: Mary Kathryn Downer, Travis Gerke, Sigrid V Carlsson, Howard D Sesso, Adam S Kibel, Quoc-Dien Trinh, Hans Lilja, Andrew J Vickers, Kathryn M Wilson, Lorelei Mucci
Written By: Zachary Klaassen, MD, Urologic Oncology Fellow, University of Toronto, Princess Margaret Cancer Centre Twitter: @zklaassen_md at the 72nd Canadian Urological Association Annual Meeting - June 24 - 27, 2017 - Toronto, Ontario, Canada