Accuracy of Transurethral Resection of the Bladder in Detecting Variant Histology of Bladder Cancer Compared with Radical Cystectomy

Background: Correct identification of variant histologies (VHs) of bladder cancer (BCa) at transurethral resection of the bladder (TURB) could drive the subsequent treatment. Objective: To evaluate the concordance in detecting VHs between TURB and radical cystectomy (RC) specimens in BCa patients. Design, setting, and participants: We retrospectively analyzed 1881 BCa patients who underwent TURB and subsequent RC at seven tertiary care centers between 1980 and 2018. VHs were classified as sarcomatoid, lymphoepithelioma-like, neuroendocrine, squamous, micropapillary, glandular, adenocarcinoma, nested, and other variants. Outcome measurements and statistical analysis: Concordance between TURB and RC was defined as the ability to achieve histological subtypes at TURB confirmed at RC specimen, and was expressed according to Cohen's kappa coefficient. Results and limitations: Of the patients, 14.6% and 21% were diagnosed with VH at TURB and RC specimens, respectively. The most common VHs at TURB were squamous, neuroendocrine, and micropapillary carcinoma (5.2%, 1.5%, and 1.5%, respectively). At RC, the most frequent VHs were squamous, micropapillary, and sarcomatoid carcinoma (7.2%, 3.0%, and 2.7%, respectively). The overall concordance in detecting VH was defined as slight concordance (coefficient: 0.18). Moderate concordance was found for neuroendocrine, adenocarcinoma, and squamous carcinoma (coefficient: 0.49, 0.47, and 0.41, respectively). Micropapillary, glandular, and other variants showed slight concordance (coefficient: 0.05, 0.17, and 0.12, respectively), while nested and sarcomatoid carcinoma showed fair concordance (coefficient: 0.32 and 0.26, respectively). Results may be limited by the absence of centralized pathological analysis. Conclusions: A non-negligible percentage of patients were diagnosed with VH at both TURB and RC. TURB showed relatively low accuracy, ranging from poor to moderate, in detecting VHs. Our study underlines the need of additional diagnostic tools in order to identify VHs properly at precystectomy time and to improve patient survival outcomes. Patient summary: In this report, we underlined the low accuracy of transurethral resection of the bladder in detecting variant histologies and the need for additional diagnostic tools.