Division of Urology, Sunnybrook Health Sciences Center, University of Toronto, Toronto, ON, Canada
Amanda Elizabeth Hird , Refik Saskin , Lisa Del Giudice , Girish S. Kulkarni , Robert Nam
Background: Baseline prostate specific antigen (PSA) is a strong predictor of clinically significant prostate cancer. In a large cohort of patients undergoing opportunistic testing, our objective was to assess the association between first serum PSA in combination with other PSA predictors and risk of prostate cancer. Methods: This was a retrospective, population-based study in Ontario, Canada between 2010-2019. Men 40-75 years of age who underwent incident PSA testing with at least two PSA tests during the study period were included. Among men who underwent PSA testing, the PSA levels of patients who were diagnosed with prostate cancer were compared to patients who were not. Univariable and multivariable logistic regression analysis was used to compare any prostate cancer and clinically significant prostate cancer diagnosis (International Society of Urological Pathology [ISUP] grade group 2-5) between groups. Results: A total of 508,238 patients were included in our cohort (12,444 cases and 495,794 controls). The median follow-up time was 8.2 years (IQR: 7.0-9.1). Patients who were diagnosed with prostate cancer were older (median age 62 [IQR: 56-67] versus 56 years [IQR:50-63], standardized difference [SD]:0.61) and had a higher first PSA than patients who were not (median 4.79 [IQR:3.29-6.96] versus 0.96 ng/mL [IQR:0.58-1.69], SD:1.93). Fewer than 0.1% (248/261,463) of patients with a first PSA less than 1.0 ng/mL were diagnosed with prostate cancer during the study period. Our final multivariable model revealed that a first PSA above 2.0 ng/mL (adjusted odds ratio [OR] 6.64, 95%CI: 6.13-7.20, p < 0.001), a final PSA between 4.0 to 9.9 ng/mL and 10.0 to 20.0 ng/mL (adjusted OR 22.09, 95%CI: 20.58-23.71, p < 0.001 and adjusted OR 47.46, 95%CI: 43.28-52.05, p < 0.001, respectively) and change from first to final PSA per 365 days of 20.0 to 99.9% (adjusted OR 3.40, 95%CI: 3.21-3.60, p < 0.001) and more than 100% (adjusted OR 6.91, 95%CI: 6.29-7.59, p < 0.001) were strongly associated with the diagnosis of prostate cancer. The model performed well when predicting any prostate cancer, clinically significant prostate cancer, and when stratified by age. Conclusions: This study suggests that first PSA in combination with final PSA and rate of change can be used to predict prostate cancer diagnosis. Future studies will be conducted to determine the association with metastatic and lethal prostate cancer.
Disclaimer
This material on this page is ©2024 American Society of Clinical Oncology, all rights reserved. Licensing available upon request. For more information, please contact licensing@asco.org
Abstract Disclosures
2022 ASCO Genitourinary Cancers Symposium
First Author: Alicia K. Morgans
2023 ASCO Annual Meeting
First Author: Daniel Eidelberg Spratt
2023 ASCO Annual Meeting
First Author: Alexandre Akoulitchev
2024 ASCO Genitourinary Cancers Symposium
First Author: Alexandru George Floares