Exploration of biomarkers for TRUCE-01: Development of a urine-based efficacy prediction model.

Abstract

4622 Background: Preliminary results from the TRUCE-01 phase II trial (NCT04730219) indicate encouraging antitumor activity and a manageable safety profile for the combination of the immune checkpoint inhibitor tislelizumab and nab-paclitaxel in patients with T2-4a N0/X M0 muscle-invasive bladder cancer (MIBC) of predominant urothelial carcinoma histology. To identify patient subgroups that may derive particular benefit, we conducted a comprehensive biomarker analysis. Methods: We performed whole-exome sequencing (WES) on baseline tumor tissue from 44 patients. Concomitantly, urine samples were collected during treatment. Baseline urine from 44 patients was subjected to 808-panel circulating tumor DNA (ctDNA) detection, and low-depth whole-genome sequencing (WGS) was performed across three time points (pre-treatment, during treatment, and post-treatment) for 47 patients. Results: The classic biomarker PD-L1 expression failed to predict complete response (CR) Urine-derived biomarkers showed high concordance with tissue-based mutation profiles. Specifically, urinary mutations in RB1, ERCC2, and AR were associated with CR, while mutations in ELF3, KDM6A, and TP53 were linked to non-clinical response (NCR). Baseline ctDNA levels (p=0.016) and baseline (p=0.011)/post-treatment (p=0.0011) WGS scores were significantly higher in the NCR versus CR group, potentially reflecting higher tumor burden. Interestingly, only two patients who achieved CR later experienced disease progression; both showed a rising trend in WGS scores during treatment, suggesting that CR does not ensure long-term favorable prognosis and that dynamic urine monitoring may be crucial during treatment. Finally, we developed a predictive model integrating urinary WGS scores, ctDNA scores, and efficacy-associated mutation genes. This integrated model achieved an area under the curve (AUC) of 0.930, significantly outperforming models based on WGS score alone (AUC=0.695) or ctDNA score alone (AUC=0.749). Conclusions: Urine-based biomarker assays show promise for predicting response to chemo-immunotherapy in MIBC. Further studies with expanded sample sizes and multi-center validation are warranted to confirm these findings and assess their clinical utility in broader patient cohorts. Clinical trial information: NCT04730219 .