Differential MMR protein stability and heterogeneity in bladder versus non-bladder urothelial carcinomas

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Abstract Background Mismatch repair deficiency (dMMR) is a predictive biomarker for immunotherapy response in various cancers, but its stability and heterogeneity across different anatomic sites in urothelial carcinoma (UC) remain poorly defined. This study aimed to evaluate dMMR status in bladder (BUC) versus non-bladder UC (upper tract [UTUC] and prostatic [PUCP]), and to assess whether it remains stable during disease progression. Methods We analyzed 249 tissue specimens from 108 UC patients (72 BUC, 35 UTUC, 1 PUCP). For a subset of patients, multiple samples across spatiotemporal dimensions (primary, locally invasive, metastatic, recurrent) were collected. MMR protein expression (MLH1, MSH2, MSH6, PMS2) was assessed by immunohistochemistry (IHC). dMMR concordance across samples was evaluated. Results dMMR was absent in all BUC cases (0/72). In all BUC patients with multiple samples, dMMR status was 100% concordant across all spatiotemporally distinct specimens. In contrast, dMMR was identified in 2/35 UTUC (5.7%) and 1/1 PUCP cases. In the few dMMR‑positive non‑bladder cases, examples of divergent MMR status were observed between primary and metastatic lesions. Conclusions In conclusion, our study suggests a potential dichotomy in the behavior of dMMR between bladder and non‑bladder urothelial carcinomas. In this cohort, dMMR was exceptionally rare in BUC and remained stable across all paired specimens examined, while the few dMMR‑positive non‑bladder cases showed examples of divergent MMR status between primary and metastatic lesions. These findings suggest that primary tumor assessment may be sufficient for BUC, while for non‑bladder UC, they raise the question of whether a more flexible approach to biopsy sampling may be warranted. However, given the limited number of dMMR events, these observations should be considered hypothesis‑generating and require validation in larger, independent cohorts.