Predictors of unfavourable repeat biopsy results in men participating in a prospective active surveillance program.

Abstract

Active surveillance (AS) protocols for low-risk prostate cancer (PCa) generally include repeat prostate biopsies at predefined follow-up intervals. To study the outcome of routinely obtained 1-yr repeat biopsies and factors predicting reclassification to higher risk, to contribute to risk stratification for men on AS. We analysed men with low-risk PCa (clinical stage ≤ T2, prostate-specific antigen (PSA) ≤ 10 ng/ml, PSA density <0.2 ng/ml per millilitre, one or two positive biopsy cores, and Gleason score ≤ 6) who had been included in a prospective AS protocol. PSA was measured 3-monthly and the first volume-dependent repeat biopsy was scheduled 1 yr after diagnosis, independent of PSA doubling time (PSA-DT). Reclassification to higher risk disease on repeat biopsy was defined as Gleason score ≥ 7 or ≥ 3 positive cores. We analysed whether baseline patient characteristics and PSA-DT were associated with reclassification to more aggressive PCa on repeat biopsy. A first repeat biopsy was taken in 757 patients after median follow-up of 1.03 yr. The results of repeat biopsies were favourable (no or low-risk PCa) in 594 patients (78.5%) and led to reclassification of risk in 163 (21.5%). Analysis showed that reclassification to higher risk was significantly influenced by the number of initial positive cores (two vs one) (odds ratio [OR]: 1.8; p=0.002) and higher PSA density (OR: 2.1; p=0.003). The outcome was not significantly influenced by age, clinical stage, total number of biopsy cores, or PSA. Adding PSA-DT at time of repeat biopsy to the model showed PSA-DT <3 yr to be significantly associated with reclassification to higher risk (OR: 1.7; p=0.015). Data on tumour involvement per biopsy core were not available. Clinical features at baseline and during follow-up in our AS cohort are significantly associated with short-term reclassification to higher risk on repeat biopsy. These characteristics can potentially be used for risk stratification of men with PCa who are apparently at favourable risk.