Original contributionVariant morphology in upper urinary tract urothelial carcinoma: a 14-year case series of biopsy and resection specimens

•Upper urinary tract urothelial carcinoma exhibits a spectrum of variant morphology.•Variant morphology frequency in biopsies and resections has not been reported.•The frequency of variant morphology is around 12% in the general population.•Previously reported incidences are most likely inflated because of a consultation bias.

SummaryUpper urinary tract urothelial carcinoma exhibiting variant morphology, especially in higher-grade tumors, is a recognized phenomenon but has not been comparatively studied in biopsy versus resection material. We studied the morphologic patterns and clinicopathological features, and provide a comparison between biopsy and resection specimens. Consultation cases were evaluated separately to investigate for possible consultation bias. A total of 383 in-house cases from 352 patients including 314 resection specimens and 69 biopsies from 2001 to 2014 were reviewed from a single institution. Histologic type, tumor grade, invasion, pathologic stage, nodal status, metastasis, and the presence and type of variant morphology for each case were evaluated. Variant morphology was identified in 5 biopsy specimens (7.2%) and 42 resection specimens (13.4%). The most common variant morphologic pattern was squamous differentiation (16 cases, 4.5%) followed by an inverted growth pattern (8 cases, 2.2%). The presence of variant morphology in resection specimens had a significant association with higher tumor grade, higher pT stage, and nonpapillary configuration. Of 69 patients with biopsies, 31 had a subsequent resection. In comparison, 181 consultation cases from 168 patients showed variant morphology in 6 biopsies (7.1%) and 27 resections (28.1%). In conclusion, the frequency of recognizing variant morphology in biopsies is about one-half of that in resections. The inclusion of consultation cases can inflate the incidence of variant morphology. As a result, the frequency of variant morphology in our in-house cases is lower than the percentage reported in the literature, most likely secondary to a consultation bias.