Accuracy of genome-wide mutational analysis for tumor-informed ctDNA-guided MRD monitoring in bladder cancer.

Abstract

4589 Background: Neoadjuvant chemotherapy (NAC) followed by radical cystectomy (RC) is the standard management of localized muscle-invasive bladder cancer (MIBC). However, about 45% of patients develop metastases within 2 years after RC. Biomarkers for treatment efficacy evaluation and early detection of minimal residual disease (MRD) are needed for earlier treatment initiation and monitoring of treatment response. Tumor-informed detection of mutations in cell-free DNA (cfDNA) has shown promising results to monitor MRD. However, the low tumor fraction after surgery and limited input material obtained from a typical plasma sample limits the probability of detecting low metastatic burden scenarios. Here we implemented and applied a whole-genome sequencing (WGS) approach to monitor circulating tumor DNA (ctDNA) for sensitive ctDNA detection. Methods: 110 MIBC patients undergoing NAC and RC were enrolled. cfDNA was extracted from ~1mL plasma (n=978) and procured from longitudinal plasma sampling during NAC and pre-RC and post-RC. WGS of tumor/germline pairs (30x/20x) and plasma cfDNA (>20x) was performed, facilitating detection of genome-wide genomic alterations and quantification of ctDNA using the C2inform method. Results: For each patient we developed a tumor-informed WGS model by integrating genome-wide mutation and copy number variation data coupled with advanced signal processing and AI-based error suppression. Patient-specific somatic variant patterns were used for detection and measuring ctDNA levels in low-input blood samples by WGS. Post-RC ctDNA analysis identified patients with recurrence with 78% sensitivity and 95% specificity and with a median lead time over radiographic imaging of 118 days (full follow-up included). When restricting follow-up time to 12 months following the latest ctDNA test, a sensitivity of 86% and specificity of 94% was achieved. ctDNA status was associated with recurrence-free (p<0.0001) and overall survival (p<0.0001). Furthermore, ctDNA clearance during NAC was also associated with recurrence-free survival (p=0.0051). The APOBEC associated signatures SBS2 and SBS13 were identified as the primary contributors to the mutational landscape of the primary tumors and TP53, KMT2D, PIK3CA, RB1 and KDM6A were the main affected driver genes. Half of the tumors displayed whole genome doubling. Moreover, our analysis of plasma samples after treatment showed the presence of chemotherapy-induced mutational signatures not present in the primary tumor, along with increases in copy numbers on chromosomes 19q and 20, and a focal amplification of the FGFR3 gene on chromosome 4. Conclusions: Our results highlight the clinical potential of personalized genome-wide mutation integration as an ultra-sensitive, non-invasive method for MRD detection and treatment response monitoring.