Discrepancy between local and central pathological review of radical prostatectomy specimens.

摘要

PURPOSE: Pathological assessment of radical prostatectomy specimens has not been uniform among pathologists. We investigated interobserver variability of radical prostatectomy specimen reviews between local and central pathologists. MATERIALS AND METHODS: We collated data from 50 institutions on 2,015 patients with cT1c-3 prostate cancer who underwent radical prostatectomy between 1997 and 2005. All radical prostatectomy specimens were retrospectively reevaluated by a central uropathologist. Gleason score, extracapsular extension, seminal vesicle invasion, lymph node involvement, positive surgical margin, year of diagnosis and pathology volume were recorded. RESULTS: The exact concordance rate of Gleason score between local and central review was 54.8%, and under grading and over grading rates at local review were 25.9% and 19.2%, respectively. Spearman's rank correlation coefficient was 0.61 for local and central radical prostatectomy Gleason score. The exact concordance rate of Gleason score 8-10 at local review was significantly lower than that of Gleason score 5-6, 3 + 4 and 4 + 3 at local review (p = 0.011, <0.001 and 0.006). Exact concordance rates between local and central review for extracapsular extension, seminal vesicle invasion, lymph node involvement and positive surgical margin were 82.5%, 97.6%, 99.6% and 87.5%, respectively. High volume institutions and recently diagnosed cohorts showed significantly higher exact concordance rates between local and central review for radical prostatectomy Gleason score and other pathological features (all p <0.001). CONCLUSIONS: High volume institutions and recent series show higher concordance between local and central review of radical prostatectomy pathology. However, concordance for high grade Gleason score, extracapsular extension and surgical margin status remains poor. Radical prostatectomy specimens should be reevaluated in a multi-institutional study for more accurate pathological data.