Digital Rectal Examination in Stockholm3 Biomarker-based Prostate Cancer Screening

Abstract

BackgroundPathological digital rectal examination (DRE) is suggestive of prostate cancer but has low sensitivity and specificity. DRE is incorporated in many clinical risk calculators, but there is less evidence on how DRE performs in the setting of blood biomarkers and polygenic risk prediction models other than prostate-specific antigen (PSA) associated with prostate cancer. The Stockholm3 test combines a blood test and clinical variables including DRE. ObjectiveTo assess the predictive performance of DRE for finding clinically significant prostate cancer in systematic biopsy and evaluate its added value to the multivariable diagnostic test Stockholm3. Design, setting, and participantsThis population-based study in the screening by invitation setting included 5543 men aged 50–69 yr with PSA ≥3 ng/ml who were referred for systematic prostate biopsy between 2012 and 2015. The STHLM3 study is registered with ISRCTN.com as ISRCTN84445406. Outcome measurements and statistical analysisPredictive performance was assessed via estimates of sensitivity and specificity and in logistic regression. Clinically significant cancer was defined as International Society of Urological Pathology grade group ≥2 (GG ≥2) cancer on systematic biopsy. Results and limitationsWe found that 11% of men with PSA ≥3 ng/ml had a suspicious DRE. A suspicious DRE was associated with a 3.16-fold higher risk (95% confidence interval [CI] 2.83–3.52) of GG ≥2 cancer and greater length of cancer on biopsy. The risk of nonsignificant cancer was similar regardless of the DRE finding. The risk of GG ≥2 cancer was 46.2% (95% CI 42.2–50.3%) for men with a suspicious DRE versus 14.6% (95% CI 13.7–15.7%) for men with a negative DRE. The elevated risk of GG ≥2 cancer persisted after adjusting for the other Stockholm3 test parameters (odds ratio 2.88, 95% CI 2.32–3.57). For detection of GG ≥2 cancer among men with PSA ≥3 ng/ml, DRE had sensitivity of 27.8% (95% CI 25.1–30.7%) and specificity of 92.8% (95% CI 92.1–93.6%). ConclusionsIn this screening-by-invitation setting we found that for men with PSA ≥3 ng/ml, a suspicious DRE indicates more than threefold higher risk of harboring significant prostate cancer. DRE as a variable adds significant precision to the Stockholm3 prediction model. Men with a suspicious DRE should be referred for further diagnostic workup, including biopsy. Patient summaryWe investigated the ability of digital rectal examination to predict if a patient has clinically significant prostate cancer. We found that digital rectal examination provides valuable information and can help doctors in making an informed decision on whether to recommend prostate biopsy.