Transrectal ultrasound for staging prostate carcinoma prior to radiation therapy

Abstract

Despite its subjectivity and inaccuracy, digital rectal examination (DRE) has a long history of well-documented prognostic significance in patients with prostate carcinoma. To the authors' knowledge, very few studies have evaluated the relative prognostic merits of transrectal ultrasound (TRUS) versus DRE. This question is addressed in this study. The outcome for 558 men with T1-T3, N0, M0 adenocarcinoma of the prostate who underwent both DRE and TRUS and received external beam radiation without androgen ablation was evaluated relative to the prognostic information from DRE, TRUS, or both. The outcome endpoints were no evidence of disease (NED) (no relapse or rising prostate specific antigen level) and freedom from metastases. Prognostic factors were evaluated with univariate and multivariate techniques. The median follow-up was 55 months. Both purely DRE-based and purely TRUS-based T categories correlated significantly with NED status. For DRE T categories, 6-year NED rates for T1/T2 and T3 disease were 64% and 36%, respectively (P < 0.001). For TRUS T categories, the rates for T1/T2 and T3 were 63% and 39%, respectively (P < 0.001). There were significant differences in patient composition between DRE and TRUS T categories. Only 40% of patients were in the same DRE and TRUS category, but the majority of the reclassification based on TRUS was within rather than between major T categories (T1/T2 vs. T3). Changes between the prognostically significant T1/T2 versus T3 categories occurred in < or =25%. This accounted for the similarity in NED outcome for DRE and TRUS T categories. However, TRUS categories did not discriminate significantly for metastatic recurrence between T1/T2 and T3 categories, whereas DRE categories did. Upstaging or downstaging by TRUS relative to DRE did not alter the DRE prognostic groupings substantially. There was no clinically meaningful superiority of TRUS over DRE in the definition of prognostically useful T categories. Moreover, the addition of TRUS to DRE did not enhance the prognostic value of DRE findings in any meaningful way. Despite its subjectivity and inaccuracy, DRE provides prognostic information at least equivalent to TRUS and is preferable because of its low cost.