Metaproteomics based subtypes in muscle invasive bladder carcinoma and association with survival.

812 Background: Muscle-invasive bladder carcinoma (MIBC) is associated with high recurrence and mortality rates, and current neoadjuvant chemotherapy (NACT) regimens offer limited efficacy and considerable toxicity. These limitations underscore the need for reliable biomarkers to improve treatment management and patient outcomes. Metaproteomics profiling allows for the simultaneous characterisation of human and microbial proteins within tumors, offering a new perspective on patient-microbiota interactions in cancer. Our group has identified three microbiota-related metaproteomics clusters (Microbiota1-3) associated with response to NACT in MIBC. Our objective is to evaluate the prognostic relevance in terms of disease-free survival (DFS) of these microbiota-related metaproteomics clusters. Methods: Fifty-eight FFPE transurethral resection (TURBT) samples from patients diagnosed with MIBC were analyzed using data-independent acquisition (DIA) mass spectrometry. Protein identification was performed against a custom database including the human proteome and bacterial proteomes from bacteria genera identified by 16S rRNA gene sequencing in the same samples. After quality control, bacterial and human proteins were analyzed using Perseus software and probabilistic graphical models (PGMs) to identify functional relationships. Patients were classified into three metaproteomic clusters (Microbiota1-3) according to their tumor microbiota protein profiles by hierarchical clustering. Disease-free survival (DFS) was estimated by the Kaplan-Meier method and compared among clusters using the log-rank (Mantel-Cox) test. Results: Kaplan-Meier analysis revealed significant differences in DFS between the three defined metaproteomics subtypes (Microbiota1-3) (p=0.01). Patients classified as Microbiota3 had the most favorable DFS and a lower proportion of relapse events, with median DFS not reached during the follow-up period. This group was previously associated with a higher proportion of responders to NACT, consistent with the favorable DFS observed. In contrast, Microbiota2 was associated with earlier relapse and shorter DFS (median DFS: 23 months), in line with its higher proportion of non-responders to NACT. Conclusions: To our knowledge, this is the first metaproteomics study focused in biomarker discovery using FFPE samples from bladder carcinoma patients. Distinct MIBC metaproteomics profiles are associated with differential DFS following NACT. The identification of these three microbiota-related proteomics clusters highlights the potential prognostic relevance of the tumor-associated microbiota. These findings support metaproteomics as a promising approach for refining patient stratification and guiding future biomarker-driven strategies in bladder cancer.