Clinicogenomic characterization of patients with rapid tumor progression versus sustained response to frontline systemic therapy for metastatic urothelial carcinoma (mUC).
Albert Jang,Sulin Wu,Hamsa Latha Sampath Kumar,Ravi Kumar Kyasaram,Chen-Han Wilfred Wu,Laura Bukavina,Adam C. Calaway,Jorge A. Garcia,Pedro C. Barata,Prateek Mendiratta,Jason R Brown
DOI: https://doi.org/10.1200/jco.2024.42.4_suppl.685
IF: 45.3
2024-01-31
Journal of Clinical Oncology
Abstract:685 Background: mUC often carries a poor prognosis, and a subset of patients (pts) progresses within weeks of starting frontline therapy. While clinicopathologic predictors such as the Bellmunt score have been proposed, predicting this high-risk subset is challenging. Next generation sequencing (NGS) may identify predictive biomarkers for pts with rapid tumor progression, but the predictive value of variant allele frequency (VAF) of genomic alterations is uncertain. Methods: An IRB-approved, HIPAA-compliant single institution retrospective database of pts diagnosed with mUC from 1/2013-3/2023 with somatic NGS and were treated with standard of care therapy was generated. Baseline demographics, clinicopathologic features, and NGS results including pathogenic/likely pathogenic mutations, fusions, copy number alterations, and VAF were extracted from the electronic health record. Progression-free survival (PFS) was calculated from the start of frontline (1L) treatment until radiographic/pathologic progression or death from cancer, whichever came first. PFS 180 days was considered sustained response (SR). Statistical analysis included chi-square and student paired t-test. Results: 260 pts with mUC were identified, and 86 pts had available NGS data. Clinical data cutoff was 9/2023. NGS results used in this analysis were based on the primary tumor (64.0%), distant metastasis (27.9%), or blood (5.8%). Comparing the RP cohort (n=29) vs SR cohort (n=40), there were no significant differences for median age at mUC onset (68 v 68 yrs), proportion of female pts (34 v 28%), Black ancestry (14 v 5%), or smoking history (66 v 68%). Notable variant histologies included squamous cell (28 v 10%, p=0.051) and small cell (0 v 10%, p=NA). 1L treatment involved cisplatin (38 v 43%), carboplatin (17 v 35%), oxaliplatin (0 v 2.5%), or immune checkpoint inhibitor monotherapy (45 v 20%). At mUC, RP cohort pts had more liver metastases (31 v 10%, p=0.027). Median TMB was 10.5 (n=17) vs 6.6 mut/Mb (n=27). No significant difference between RP and SR was observed in the incidence of alterations in TP53 (52 v 58%), TERT promoter (50 v 71%), PIK3CA (41 v 28%), ARID1A (27 v 23%), and KMT2D (14 v 26%). Pts in the RP cohort were significantly more likely to have lower VAF for TP53 ( p=0.006) and KMT2D ( p=0.028). Amongst 1L platinum-treated pts, low TP53 VAF remained significantly associated with the RP cohort ( p=0.010) and KMT2D VAF nearly significant ( p=0.077). Conclusions: Pts with mUC who rapidly progressed on frontline therapy had a significantly higher liver involvement at mUC diagnosis, a trend toward squamous histology, and a significantly lower VAF for TP53 and KMT2D alterations. VAF is a potential predictive biomarker that should be further investigated. We plan to validate these findings in larger cohorts.
oncology