Accurate pathological staging of urothelial neoplasms requires better cystoscopic sampling

PURPOSE: The frequency with which muscularis propria was sampled by urologists and the sources of interpretive discrepancies among pathologists were studied in a community practice setting. MATERIALS AND METHODS: A total of 217 consecutive cases of urothelial neoplasm were independently reviewed by 3 pathologists. The presence or absence of muscularis propria as well as interpretive discrepancies were recorded. RESULTS: Despite clinical emphasis on accurate pathological staging portions of muscularis propria were absent from samples of histologically documented urothelial neoplasms in up to 51% of cases. Failure to obtain muscularis propria varied widely among urologists but was most often associated with cases of low grade papillary neoplasms, in which invasion is less likely. Muscularis propria was usually present in cases of noninvasive carcinoma in situ but this may have represented inadvertent sampling of structures in close proximity. The incidence of interpretive discrepancies among pathologists who are required to assess the status of muscularis propria was significant (24%). Almost all problems were related to artifacts and most may have been avoided if careful attention had been given to specimen sampling and processing. CONCLUSIONS: The well documented tendency toward cystoscopic under staging has not necessarily resulted in a high incidence of muscularis propria in bladder cases of urothelial neoplasms. Even when muscle may have been sampled, artifacts that were often due to thermocoagulation hampered accurate pathological staging.