Genomic Determinants of Response to Trimodality Therapy for Muscle-Invasive Bladder Cancer.

Trimodality therapy (TMT), which consists of transurethral resection of bladder tumor (TURBT) followed by chemoradiation, is an alternative to radical cystectomy for muscle-invasive bladder cancer (MIBC), but molecular determinants of response are poorly understood. We characterized the genomic and transcriptomic landscape associated with response to TMT through whole exome sequencing (WES) and expression profiling. MIBC patients treated with TMT at a single institution were identified. DNA and RNA were extracted from available FFPE pre-treatment TURBT samples, and WES and transcriptome-wide expression profiling were performed. WES data were analyzed by standard mutation analysis pipelines with a common variant filter to account for lack of matched germline samples. Somatic mutations, insertion/deletions, and copy number variations were identified. Molecular subtypes and DNA damage response (DDR) pathway scores were derived from expression profiles. Responders (R) were defined as patients with complete response at post-treatment cystoscopy and no recurrence. Nonresponders (NR) were defined as patients with incomplete response at post-treatment cystoscopy, those who underwent salvage cystectomy, or those with regional/distant disease progression. Pre-treatment TURBT FFPE samples from 77 selected TMT patients were successfully sequenced and passed quality control metrics, 68 of which had matched transcriptomic profiling. The median age of the cohort was 72.7 years; 23% were female. All patients had MIBC (87% T2, 13% T3/T4) and the majority were treated with concurrent cisplatin-based chemotherapy. Median follow-up was 42 months in alive patients. Thirty-nine patients were classified as NR, while 38 were R. Mean somatic tumor mutational burden among R and NR was 13.5 (range, 4.2-31.5) and 11.3 (range, 3.4-26.8) mutations/megabase, respectively (p=0.08). Exploratory analyses of DDR pathways revealed more patients classified as R had mutations in ERCC2 (n=7) as compared to NR (n=1), p=0.023. Six of these 8 patients with ERCC2 mutations received cisplatin-based chemotherapy; of the 2 that did not, 1 was NR. Samples with ERCC2 mutations had higher hallmark DNA repair scores (median 0.110 versus -0.310, p=0.02), suggesting dysregulation of DDR. ERCC2 mutant tumors were also enriched for GSC luminal subtype compared to ERCC2 wildtype (83% vs. 37% luminal; p=0.03). Patients with ERCC2 mutations had improved disease-specific survival (HR 0.17 (95%CI: 0.064-0.41), p=0.04). Consistent with these results, in vitro studies showed ERCC2 mutant bladder cancer cell lines are significantly more sensitive to cisplatin and radiation than isogenic ERCC2 wildtype cells. Our data demonstrate that alterations in ERCC2, a crucial DDR gene, may be associated with improved response to TMT. These findings require validation in larger cohorts. [ABSTRACT FROM AUTHOR]