Genomic comparison of urothelial mucinous adenocarcinoma and colorectal mucinous adenocarcinoma biomarkers and therapeutic implications.

Abstract

650 Background: Urothelial mucinous adenocarcinoma (UC-MUC-AC), arising from the urothelium, is a rare condition (<2%), typically presenting at an advanced stage with an unfavorable prognosis. The lack of a standard treatment protocol adds complexity, requiring the extension of treatments from UC or organ-specific MUC-AC such as colorectal. Notably, UC-AC bears a histopathological resemblance to intestinal tumors, constituting the predominant subtype. Alternatively, treatment decisions based on gene expression similarities in lieu of histological similarities are an evolving concept in rare cancers. This is an unexplored area in UC-MUC-AC treatment, prompting our hypothesis that an all-encompassing genomic profile could reveal analogous biomarkers shared with CRC-MUC-AC. Methods: Three techniques: DNA sequencing, RNA sequencing (WTS), and immunohistochemistry (IHC) sequencing (protein) with varying measures for gene expression were analyzed at Caris Life Sciences in Phoenix, AZ. Histopathology was used to categorize five cancer site groups: (1) UC arising in the bladder, ureter, urethra, or renal pelvis with any transitional cell carcinoma (TCC)-non-MUC-nonAC [N156 samples], (2) UC nonTCC-nonMUC-AC [N83], (3) UC-MUC-AC [N16], (4) CRC-MUC-AC [N3100] and (5) CRC nonMUC-AC [N38205]. Our primary aim was to determine which group was closest to Group 3, our control. Our secondary aim was to determine if Group 2 was genetically closer to Group 1 or 5. Only genes with positive expression were included in the cluster analysis. ANOVA analysis was conducted to further explore the relationships between the groups. A log transformation was performed to achieve normality. To further investigate pairwise comparisons, a post hoc Dunnett test was employed. Results: In our first analysis, Group 3 UC-MUC-AC gene expression ratio was predominately associated with UC Groups 1 & 2 (24 of 33), and farther to CRC Groups 4 & 5 (9/33) (all p-values >0.05). In the second analysis, the ANOVA overall model comparing gene expression between the histological groups 1, 2 & 5 was statistically significant (p=0.0018). The Dunnett test revealed a farther difference between Group 2, UC nonTCC-nonMUC adenocarcinoma, and 5, CRC non-MUC-AC (p-value =0.001), but closest between Group 2 and 1, UC-TCC-nonMUC-nonAC (p =0.100). Conclusions: Our findings indicate that next-generation sequencing technology has the potential to aid in treatment decisions based on the site of origin, such as UC-AC, rather than relying on traditional histopathological descriptive approaches for extrapolating evidence from treatment protocols used for other cancer origins like CRC. Future research on artificial intelligence could aid physicians with more informed treatment decisions by providing additional insight into how closely tumors match cancer types' genomic and transcriptomic signatures.