Dr Linscott on Urinary Cell-free Tumor DNA to Predict MRD in High-risk NMIBC

Joshua Linscott, MD, discusses using urinary cell-free tumor DNA to predict minimal residual disease in high-risk non-muscle invasive bladder cancer.

Joshua Linscott, MD, PhD, urologic onologist, Moffitt Cancer Center, discusses findings from a study investigating the use of urinary cell-free tumor DNA (utDNA) to predict minimal residual disease (MRD) prior to repeat transurethral resection of bladder tumors (rTURBT) in patients with high-risk non-muscle invasive bladder cancer (NMIBC).

The study investigated the efficacy of using tumor fraction and copy number burden (CNB) from utDNA as biomarkers for detecting MRD. Linscott explains that genomic alterations detected in utDNA are concordant with those from tissue samples in urothelial carcinoma. This study evaluated whether utDNA alterations and CNB could serve as reliable indicators of MRD before standard-of-care rTURBT.

At the 2024 ASCO Annual Meeting, Linscott and colleagues presented data from 53 patients with high-risk NMIBC, comparing utDNA from urine samples collected before rTURBT with tissue samples from initial TURBT (iTURBT). Low-pass whole-genome sequencing was used to detect utDNA and estimate CNB. Tumor fraction was measured using personalized ultra-deep sequencing MRD probes designed from somatic variants in iTURBT tissue.

Key findings revealed that MRD was detected in 68% of patients through pathologic evaluation. The CNB score from urine samples collected prior to rTURBT showed an area under the curve (AUC) of 0.85, with a CNB threshold of 6.14, yielding a positive likelihood ratio of 11.7 and a negative likelihood ratio of 0.37 for detecting MRD. For personalized urinary tumor fraction (n = 34), the AUC was approximately 0.89. The median urinary tumor fraction was approximately 3.5% for patients with MRD and was undetectable in disease-free patients (P < 0.001). Combining CNB and tumor fraction in a step-wise model improved the performance of the assay, achieving approximately 85.7% sensitivity, 100.0% specificity, and 90% overall accuracy.

Linscott highlights the significance of these findings, noting that utDNA is a robust tool for detecting MRD in patients with high-risk NMIBC. He emphasizes that the combination of structural changes detected by CNB and specific point mutations measured by tumor fraction resulted in high levels of MRD detection accuracy.

The study observed that MRD can be detected with high accuracy using utDNA prior to rTURBT, providing a promising noninvasive biomarker. If validated, this approach could identify patients needing maximal resection before intravesical therapy, ultimately improving patient outcomes.