Critical assessment of prebiopsy parameters for predicting prostate cancer metastasis and mortality

Introduction: Value of characteristics assessed prior to diagnosis predicting aggressive prostate cancer, metastases and mortality in men participating in a screening study were identified. Materials and methods: This study included 19950 men, aged 55 to 74 years at first screening, in the European Randomized Study of Screening for Prostate Cancer. Age, Charlson comorbidity, prostate cancer family history, vasectomy status, International Prostate Symptom Score (IPSS) score, digital rectal examination (DRE) status, transrectal ultrasound (TRUS) findings, prostate volume and prostate-specific antigen (PSA) level were assessed. Men were followed for median 11.1 years after first screening visit. Multivariate estimates of the probability of aggressive prostate cancer [stage ≥ T2c, or N1, M1, PSA > 20 ng/mL, or Gleason score ≥ 8], developing distant metastases and dying from prostate cancer stratified for predictors measured before prostate biopsies. Harrell's concordance index (c-index) was used for predictive accuracy. Results: Among 19950 men, 2420 men (12.1%) were diagnosed with prostate cancer, of which 623 men (3.1%) had aggressive prostate cancer, 157 men (0.8%) developed metastases and 104 men (0.5%) died due to a prostate cancer related cause of death. In multivariate analysis, PSA, DRE, TRUS findings and prostate volume had a significant association with detection of aggressive prostate cancer, metastases and prostate cancer mortality. Family history was significantly associated with aggressive prostate cancer. Accuracy for predicting aggressive prostate cancer c-index = 0.90, distant metastases c-index = 0.87, and prostate cancer specific mortality c-index = 0.87. Conclusions: In a large population of men who were screened for prostate cancer, detection of aggressive prostate cancer, metastases and prostate cancer mortality was predicted based on predictors available before biopsy. These results support the value of a multivariate risk assessment and stratification tools.

• View at Pubmed