In cystectomy specimens with bladder cancer whole organ embedding increases the detection rate of histopathological parameters, but not of those with prognostic significance.

Histological tumor subtyping, staging, and grading are of utmost importance to stratify patients with bladder cancer for treatment and should be as precise as possible. In the presented study, we investigated the prognostic impact of standard clinicopathological parameters in cystectomy patients and compared embedding of the entire bladder with standard partial embedding via a virtual superimposed approach. The study included 121 cystectomy specimens, which were completely embedded. Clinical and histopathological data of patients were obtained (median follow-up 21.5 months; range 1-67 months). For 88 patients two-dimensional tumor maps (macrophotographs and histology-based maps) were prepared, and embedding of the entire bladder was compared with a virtual standard partial embedding, created by a virtual overlay and data extraction of the tumor maps. Kaplan-Meier plots, Cox regression estimators, Chi-square, and McNemar tests were used. In a multivariate Cox regression model for overall survival, only venous invasion (p = 0.008, HR = 3.35, 95 % CI 1.375-8.161) and organ-confined (pTis-pT2) versus non-organ-confined diseases (pT3-pT4; p = 0.021, HR 2.669, 95 % CI 1.157-6.159) were found significant. Advanced versus standard embedding revealed significant improvement in the detection of carcinoma in situ (50 versus 61, p = 0.003) and lymphatic invasion (18 versus 24, p = 0.041), but no significant advantage in the detection of tumor stage, tumor multifocality, or venous invasion (all p > 0.05). TNM classification, including lymphatic and venous invasion, is of utmost importance to stratify patients with advanced invasive bladder cancer. Histopathological details are detected more reliably by whole organ embedding, but this approach showed no significant benefit in terms of outcome-related parameters (max. tumor stage, venous invasion) in our cohort.