The effect of benign lower urinary tract symptoms on subsequent prostate cancer testing and diagnosis.

BACKGROUND Lower urinary tract symptoms (LUTS) are common and have been associated with the subsequent diagnosis of prostate cancer (PCa) in population cohorts. OBJECTIVE To determine whether the association between LUTS and PCa is due to the intensity of PCa testing after LUTS diagnosis. DESIGN, SETTING, AND PARTICIPANTS We prospectively followed a representative, population-based cohort of 1922 men, aged 40-79 yr, from 1990 until 2010 with interviews, questionnaires, and abstracting of medical records for prostate outcomes. Men were excluded if they had a previous prostate biopsy or PCa diagnosis. Self-reported LUTS was defined as an American Urological Association symptom index score >7 (n=621). Men treated for LUTS (n=168) were identified from review of medical records and/or self report. Median follow-up was 11.8 yr (interquartile range: 10.7-12.3). OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS Associations between self-reported LUTS, or treatment for LUTS, and risk of subsequent prostate biopsy and PCa were estimated using Cox proportional hazard models. RESULTS AND LIMITATIONS Fifty-five percent of eligible men enrolled in the study. Men treated for LUTS were more likely to undergo a prostate biopsy (hazard ratio [HR]: 2.4; 95% confidence interval [CI], 1.7-3.3). Men younger than 65 yr who were treated for LUTS were more likely to be diagnosed with PCa (HR: 2.3, 95% CI, 1.5-3.5), while men aged >65 yr were not (HR: 0.89, 95% CI, 0.35-1.9). Men with self-reported LUTS were not more likely to be biopsied or diagnosed with PCa. Neither definition of LUTS was associated with subsequent intermediate- to high-risk cancer. The study is limited by lack of histologic or prostate-specific antigen level data for the cohort. CONCLUSIONS These results indicate that a possible cause of the association between LUTS and PCa is increased diagnostic intensity among men whose LUTS come to the attention of physicians. Increased symptoms themselves were not associated with intensity of testing or diagnosis.