506P - Development of a sensitive urinary cfDNA profiling assay for detection and therapy efficacy monitoring in bladder cancer

Background

Urothelial bladder cancer (UBC) is one of the most common lethal cancer worldwide. Current standard of care relies on cystoscopy, which is invasive and potentially inaccessible in low-resource settings. The fast advancement of next-generation sequencing (NGS)-based assays offers a great opportunity for analyzing tumor-derived signals from various bodily fluids. Here we report the development and validation of a NGS-based urinary tumor DNA (utDNA) profiling assay for noninvasive detection and surveillance of UBC.

Methods

The OncoCompass Target^TM panel covers genomic variants commonly found in UBC, which span genomic regions of around 410kb. Noise suppression was achieved by unique molecular identifiers (UMI)-facilitated deep sequencing and in-house bioinformatics toolsets was implemented to detect SNV (single nucleotide variant), InDel (insertion and deletion), SV (structural variation) and CNV (copy number variation). Limit of blank (LoB) was determined by profiling urine samples from 63 age-matched healthy donors. Limit of detection (LoD) was established by a series of titrations with somatic variant allele frequencies (sVAF) ranging from 0.125% to 1%. The precision study was carried out by testing 20 replicates per condition across different operators and instruments.

Results

The analytical performance was assessed using contrived samples and human urine samples. At 20ng of urine cell-free DNA input, The LoD for SNV, short Indels, and SVs were 0.31, 0.36%, and 0.30% respectively. In addition, the variant-level error rate is 5.5X10^-6 across the entire panel in negative samples. Excellent reproducibility was observed with complete concordance for all samples performed in replicates.

Conclusions

The urine-based OncoCompass Target^TM deep-sequencing assay has demonstrated robust sensitivity for low-frequency SNVs, indels, and fusions, offering an opportunity for non-invasive diagnostics and treatment response monitoring in patients with UBC.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

Burning Rock Dx.

Disclosure

J. Su, X. Yang, L. Zhang, S. Zhang, L. Huang, X. Mao, B. Li: Financial Interests, Personal, Full or part-time Employment: Burning Rock Dx.