Incidence and survival outcomes in patients with upper urinary tract urothelial carcinoma diagnosed with variant histology and treated with nephroureterectomy

Objective: To evaluate the incidence and survival outcomes of histological variants of upper tract urothelial carcinoma (UTUC) treated with radical nephroureterectomy (RNU). Materials and Methods: We retrospectively analysed data from 1610 patients treated with RNU for clinically non-metastatic UTUC between 1990 and 2016 in several centres participating in the UTUC Collaboration. Histological variants were classified as micropapillary, squamous, sarcomatoid and other, including other rare variants (<10 cases for each). Multivariable competing risk analyses were conducted to assess the effect of variant histology on overall recurrence and cancer-specific mortality (CSM). Results: Overall, 1460 patients (91%) had pure urothelial carcinoma (PUC), whereas 150 (9%) were diagnosed with a variant histology, including 89 (5.0%), 41 (2.0%), 10 (1.0%) and 10 (1.0%) cases of micropapillary, squamous, sarcomatoid and other tumours, respectively. Variant histology was associated with the presence of adverse pathological features compared with PUC, including non-organ-confined disease (59% vs 38%; P < 0.001), lymph node invasion (28% vs 24%; P = 0.02), high-grade disease (88% vs 71%; P < 0.001), tumour necrosis (28% vs 16%; P = 0.001) and positive surgical margins (15% vs 8%; P = 0.01). In competing risk analysis, micropapillary variant was the only factor associated with worse recurrence (sub-hazard ratio [SHR] 2.27, 95% confidence interval [CI] 1.25–4.79; P = 0.02) whereas sarcomatoid variant was associated with worse CSM (SHR 16.8, 95% CI 6.86–41.17; P < 0.001). Conclusion: We found that one out of 10 patients with UTUC treated with RNU had variant histology. Only micropapillary and sarcomatoid variants were associated with poorer oncological outcomes after adjusting for available confounding factors.