Introduction:

Upper tract urothelial carcinoma (UTUC) is a rare and aggressive malignancy that is difficult to diagnose and stage. A urine test capable of detecting UTUC that provides diagnostic and prognostic information might aid diagnosis, improve risk stratification, and aid in treatment decisions. Urinary comprehensive genomic profiling (uCGP) via next-generation sequencing with the UroAmp assay (Convergent Genomics) was developed to identify mutations, diagnose disease, assess molecular grade, and predict recurrence risk. Here, we validate a uCGP approach to identify lesions in UTUC patients, assess risk of invasion, and identify genomic features used for therapy selection, such as FGFR3 mutation status, microsatellite instability (MSI), and tumor mutational burden (TMB). We describe the genetic alterations and test performance characteristics of UroAmp disease and grade prediction in a cohort diagnosed with UTUC. 

Methods:

uCGP was performed on 20 urine specimens from individuals with either a history of UTUC (n=8) or collected at the time a pathology confirmed UTUC tumor was present (n=12). The grade distribution was 23% LG and 77% HG; stage distribution was 20% Ta, 35% T1, 15% T2-T4, 5% CIS, and 25% unknown. Urine DNA was sequenced and profiled across 60 genes to identify six classes of tumor mutations: single-nucleotide variants (SNV), gene-level copy-number variants (CNV), insertion-deletions (INDEL), copy-neutral loss of heterozygosity (LOH), MSI, and whole-genome aneuploidy. UroAmp provides a measure of genomic disease burden (GDB), defined as a percentile ranking of mutational burden compared to previously profiled urine tumor DNA. TMB high was defined as ≥10 mutations per mega base.   

Results:

UroAmp provided disease classifications on all specimens and performed with a sensitivity and specificity of 100 (95% CI: 88.3-100). HG status was predicted with a sensitivity of 60% (95% CI: 30 – 85) and PPV of 100% (95% CI: 67-100). Disease positive patients had a higher number of mutations (mean 14 vs. 2, p<0.001) and GDB (mean 66.6 vs. 6.2, p<0.001) compared to the disease negative group. FGFR3 positive (n = 4), MSI unstable (n = 2), and TMB high status (n = 3) were only observed in disease positive patients. One individual was positive for all three therapeutic selectors. 

Conclusion:

uCGP provides clinicians with diagnostic and prognostic insights into their patient’s disease. This is of particular importance in UTUC where lesions are often small and difficult to detect, stage, and thus treat. uCGP can be used to appropriately risk stratify patients and can identify patients who may benefit from first line systemic therapy, versus immediate surgery, based on FGFR3 and MSI/TMB status. 

Funding: Convergent Genomics