Clinical outcomes in metastatic bladder cancer associated with mutations in Trithorax-group protein genes.

835 Background: The Trithorax-group (TrxG) comprises a large family of multi-protein complexes that play a key role in epigenetic regulation, cell division and inflammatory response; mutations therefore have a potential role as biomarkers or treatment targets. Two crucial complexes in TrxG are COMPASS (histone methylation) and SWI/SNF (chromatin remodeling). Certain genes in these complexes have been implicated as tumor suppressors and oncogenes, respectively. In urothelial cancer, mutations have been associated with more aggressive tumor biology, an immune-dysregulated microenvironment, and worse clinical outcomes. However, research into the overall impact of mutations in these complexes, particularly in metastatic setting, is scarce. We hypothesized mutations would predict worse outcomes following treatment, specifically investigating metastatic urothelial carcinoma (mUC). Methods: We retrospectively reviewed data from 63 patients with mUC treated at Brown Health Cancer Institute with next generation sequencing available from tumor or ctDNA. Demographic, pathologic, treatment and response data were collected. The presence of somatic alterations in the COMPASS complex ( KDM6A, KMT2B/C/D, SETD1/2, EP300 ) and SWI/SNF complex ( SMARCA2/4, SMARCB1, SMARCC1/2, ARID1A/B, ARID2 ) genes were noted. Log rank tests were used to analyze differences in overall survival (OS) and progression free survival (PFS) stratified by mutation status and first line treatments received. All outcomes were censored as of 12/31/22. Results: Of the 63 patients, the majority were male (41, 65%), white (57, 90%), had a smoking history (41, 65%). Median age was 74 (± 9.5). 28 (44%) received first-line chemotherapy and 31 (49%) received immune checkpoint inhibitor (ICI) monotherapy. 3 (5%) died prior to treatment and 1 (2%) entered a clinical trial. 41 (65%) had a mutation in any investigated TrxG gene; most common were KMT2D (17, 27%) and KDM6A (14, 22%). Patients with mutations were more likely to be older (75 vs 70, p = 0.04); there was no statistically significant difference in sex, smoking history, race, or sites of metastasis. The presence of a mutation was not associated with OS in the full cohort (median 14 months vs 13, p = 0.63, HR = 0.86 [0.45 – 1.6]). In the subset of patients who received first line chemotherapy, mutation presence (17, 61%) was associated with improved OS (21 months vs 10, p = 0.015, HR = 0.26 [0.08 – 0.77]) and PFS (7.9 vs 3.8, p = 0.051, HR = 0.36 [0.13 – 1.0]). Among patients who received first line ICI, mutation presence (19, 61%) had a trend for reduced PFS, (4.3 months vs 17, p = 0.06, HR = 2.2 [0.96 – 5.0]). Conclusions: Mutations in COMPASS and SWI/SNF genes were common in patients with mUC and associated with improved outcomes with 1st line chemotherapy and inferior responses to ICIs. TrxG mutations hold potential as predictive biomarkers, and merits further exploration in larger datasets.