Introduction: PSA density (PSAD) is used to distinguish elevated serum levels of PSA due to benign prostatic hyperplasia (BPH) from high serum levels caused by malignant disease. Several studies have demonstrated an association between high PSAD and risk for both biochemical recurrences after treatment with curative intent, as well as the risk for upstaging after prostatectomy after an initial strategy of active monitoring. In this study, our objective was to examine whether a baseline metric of PSAD is associated with prostate cancer-specific mortality after thirty years of follow-up. Methods: In the late nineteen eighties, 2400 men between the ages of 55 and 70 were randomly sampled from a background population of about 27 000 men in the same age group living in a defined part of Stockholm, Sweden. 1779 men accepted the invitation for the study. The sampled cohort was invited to undergo screening for prostate cancer with a PSA test, digital rectal exam, and ultrasound of the prostate. PSAD was calculated for each patient. Outcomes, ie prostate cancer-specific mortality were extracted from the Swedish national cause of death registry. Hazard ratios for cancer-specific deaths were estimated using a Cox proportional hazard regression model. The area under the receiver operating curve (AUC) was determined using conditional logistic regression for men with a baseline PSA of 3.0 or more. Results: PSAD was available for 1748 men. Median PSAD was 0.095 ng/mL/cm3 (IQR 0.073-0.14). The age-adjusted hazard ratio for lethal prostate cancer in the cohort of men with PSA>3.0 ng/mL was 3.5 (95% CI 2.5-4.9) per unit increase in PSAD. 463 men had a PSA of 3.0 ng/mL or more; in this cohort, there were 50 deaths attributed to prostate cancer. All 50 men who died from prostate cancer had a PSAD above the median at baseline. A model including PSAD, palpation, and ultrasound findings could distinguish lethal prostate cancer from non-cases at 30 years of follow-up with an AUC of 0.78. Conclusions: Several risk stratification models or tools have been evaluated to predict high-grade prostate cancer. If the endpoint of high-grade cancer is substituted with cancer-specific mortality a simple measure of PSA density performs similarly. For men with PSA in the gray area, PSAD predicts the lifelong risk for lethal prostate cancer. Limitations to this study include the lack of clinical data during follow-up, e.g not all nonlethal cancers are indolent which should be acknowledged when interpreting the results. SOURCE OF Funding: Grafströms foundation for urological research
