Introduction:

While the carcinogenesis and histology of upper tract urothelial carcinoma (UTUC) and urothelial carcinoma of the bladder (UCB) are similar, they emerge from differing molecular landscapes that may impact outcomes. Of all urothelial carcinomas, only 5-10% of originate in the upper tract and much of the treatment paradigm for UTUC is based on empiric evidence and extrapolation from trials in UCB. In advanced and metastatic urothelial cancer, many trials enroll both UCB and UTUC patients. Whether or not clinical outcomes for non-metastatic UCB versus UTUC are the same at comparative pathologic stages is not well defined. This study aims to compare overall survival (OS) by stage for patients with high grade (HG) urothelial cancer who underwent extirpative surgery for UTUC versus UCB to determine if site of origin impacts clinical outcomes in non-metastatic urothelial cancer.

Methods:

Retrospective medical record review of radical nephroureterectomy (RNU), ureterectomy, radical cystectomy, and partial cystectomy performed for urothelial cancer at a single cancer center from Nov 2004 to September 2022 identified 2,213 patients. Patients with metastatic disease at diagnosis, procedure for benign disease, low grade final pathology, receipt of NAC, and those lost to follow up were excluded. NAC was excluded as this improves OS and UTUC patients are less likely to receive NAC due to staging challenges.

Propensity score matching (PSM) was performed matching by age, biological sex, smoking status, ECOG score, and pathologic stage. Pearson chi squared analyses, and Kruskall-Wallis tests were used to compare clinicopathologic characteristics between UTUC and UBC. Kaplan-Meier methods were used to estimate survival. Cox-regression was performed controlling for patient demographic variables, ECOG performance status, and receipt of adjuvant therapy.

Results:

Following PSM, 580 patients were well matched with 290 in each group. Clinicopathologic variables are shown in table 1. Median OS was worse for UCB patients with muscle invasive (MI) pathology compared to UTUC patients with MI (29.0 months [95%CI 18.8-37.3] vs 56.3 months [31.9-NR], p=0.002, figure 1). UCB patients with positive lymph nodes had lower OS (16.8 months [13.5-29.6]), but median survival for UTUC patients was not reached, and the difference was not significant (p=0.059). OS was similar between groups with non-muscle invasive (NMI) pathology or negative lymph nodes. Cox-regression, controlling for demographic and clinicopathologic variables, including receipt of adjuvant chemotherapy, LVI and margin status showed UCB was independently associated with reduced OS compared to UTUC for both MI (OR=2.03 [95%CI 1.42-2.92], p<0.001) and NMI (OR=2.21 [95%CI 1.15-4.24], p=0.017). UTUC patients had a higher likelihood of any cancer recurrence, driven by increased urothelial recurrence (p<0.001, table 1).

Conclusion:

For propensity score matched patients who underwent extirpative surgery for urothelial cancer without prior NAC, UCB was independently associated with worse overall survival when compared to UTUC. Patients with UTUC are more likely to experience urothelial disease recurrence, as expected, but this does not lead to worse OS and likely reflects localized bladder recurrences. These findings help inform future clinical trials and suggest that non-metastatic UTUC and UCB should not be grouped in urothelial carcinoma trial design, as UCB patients at the same stages of invasion have worse survival outcomes.

Funding: n/a