Impact of variant allele frequency (VAF) of <i>TP53</i> alterations and Signatera circulating tumor DNA (ctDNA) monitoring for patients (pts) with advanced urothelial carcinoma (aUC) treated with enfortumab vedotin (EV).

Abstract

830 Background: EV is now a standard of care for pts with aUC, both as monotherapy and with pembrolizumab (EVP). Factors predictive of primary refractory disease (PRD) remain poorly understood. Prior investigation correlated low VAF pathogenic TP53 alterations on next-generation sequencing (NGS) with rapid progression on frontline (1L) platinum chemotherapy. Thus, we hypothesized that low TP53 alteration VAF would associate with inferior outcomes on EV. Methods: We generated an IRB-approved, HIPAA-compliant single institution retrospective database of pts with aUC who received EV monotherapy or in combination, starting 1/2019. Data cutoff was 9/2024. Baseline demographics, clinicopathologic features, and somatic NGS data were extracted from the electronic health record. PRD was defined as progressive disease (PD) on the first restaging scan after starting EV or earlier clinical progression resulting in death from disease. Signatera ctDNA was collected every 1-3 months. Statistical analysis included Fisher’s exact test. Results: We identified 82 pts with aUC who received EV, of whom 35 had available NGS data on tumor tissue. 22 pts (63%) had a TP53 alteration with reported VAF, with a median of 48.7% (range 12.1%-81.9%). Amongst the 35 pt cohort, at EV start, median age was 69 yrs, 65% male, 83% white, 40% never smokers, 23% with variant predominant histology, 17% upper tract disease, and metastatic sites included lymph node (77%), lung (26%), liver (20%), and bone (9%). 21 pts received EV as 1L therapy, and 14 pts received EV in the refractory setting (2L+). 21 pts received EVP, 1 pt received EVP with carboplatin, and 13 pts received EV monotherapy. Median follow-up time from EV start was 6.5 months, and 54% of pts were alive at data cutoff. For 1L EV, 29% (6/21) had PRD. Observed response rate (ORR) was 81% (13/16) at first restaging scan. For 2L+ EV, PRD rate was 43% (6/14), and ORR at first restaging scan was 55% (6/11). 8 of 22 pts (36%) with TP53 alteration had PRD. Pts with TP53 VAF below the median had 64% with PRD versus 9% with TP53 VAF above the median ( p =0.024). Among 22 pts with TP53 alterations, 4 pts had serial Signatera testing, of which 1 exhibited PRD. The pt with PRD had low TP53 VAF and persistently rising ctDNA values. 1 pt with low TP53 VAF had a decline in ctDNA that correlated with partial response on initial restaging scan before ctDNA rise corresponded with PD on the following scan, 1 pt with low TP53 VAF had persistent undetectable values, and 1 pt with high TP53 VAF had an initial detectable value become undetectable. Conclusions: Low TP53 alteration VAF was significantly associated with PRD to EV treatment, highlighting its potential as a predictive biomarker. ctDNA monitoring may be useful to monitor treatment response on EV. These findings are exploratory and must be validated in larger cohorts.