Validation of a Novel, Sensitive, and Specific Urine-Based Test for Recurrence Surveillance of Patients With Non-Muscle-Invasive Bladder Cancer in a Comprehensive Multicenter Study

Abstract

Bladder cancer (BC), the most frequent malignancy of the urinary system, is ranked the sixth most prevalent cancer worldwide. Of all newly diagnosed patients with BC, 70-75% will present disease confined to the mucosa or submucosa, the non-muscle-invasive BC (NMIBC) subtype. Of those, approximately 70% will recur after transurethral resection (TUR). Due to high rate of recurrence, patients are submitted to an intensive follow-up program maintained throughout many years, or even throughout life, resulting in an expensive follow-up, with cystoscopy being the most cost-effective procedure for NMIBC screening. Currently, the gold standard procedure for detection and follow-up of NMIBC is based on the association of cystoscopy and urine cytology. As cystoscopy is a very invasive approach, over the years, many different noninvasive assays (both based in serum and urine samples) have been developed in order to search genetic and protein alterations related to the development, progression, and recurrence of BC. TERT promoter mutations and FGFR3 hotspot mutations are the most frequent somatic alterations in BC and constitute the most reliable biomarkers for BC. Based on these, we developed an ultra-sensitive, urine-based assay called Uromonitor^®, capable of detecting trace amounts of TERT promoter (c.1-124C > T and c.1-146C > T) and FGFR3 (p.R248C and p.S249C) hotspot mutations, in tumor cells exfoliated to urine samples. Cells present in urine were concentrated by the filtration of urine through filters where tumor cells are trapped and stored until analysis, presenting long-term stability. Detection of the alterations was achieved through a custom-made, robust, and highly sensitive multiplex competitive allele-specific discrimination PCR allowing clear interpretation of results. In this study, we validate a test for NMIBC recurrence detection, using for technical validation a total of 331 urine samples and 41 formalin-fixed paraffin-embedded tissues of the primary tumor and recurrence lesions from a large cluster of urology centers. In the clinical validation, we used 185 samples to assess sensitivity/specificity in the detection of NMIBC recurrence vs. cystoscopy/cytology and in a smaller cohort its potential as a primary diagnostic tool for NMIBC. Our results show this test to be highly sensitive (73.5%) and specific (93.2%) in detecting recurrence of BC in patients under surveillance of NMIBC.

Keywords: FGFR3 mutation; TERT promoter mutation; Uromonitor®; non-muscle invasive bladder cancer; urinary test.

Figure 1

Urine testing workflow. In patients under surveillance for non-muscle-invasive bladder cancer (NMIBC), a minimum of 10 ml of urine is collected before cystoscopy. This 10 ml of urine is then filtered through a 0.8-µm filter and stored at 4°C. DNA extraction and Uromonitor^® test are then performed. If a positive result is obtained, confirmatory cystoscopy and transurethral resection of eventual recurrences are recommended. If a negative result is obtained, it is recommended that the test should be repeated on next follow-up appointment.

Figure 2

Technical principles of the test. (A) Real-time qualitative method optimized for TERTp c.1-124C > T detection. Two competitive fluorescent probes targeting normal (WT- c.1-124C) and mutated (Mut- c.1-124C > T) alleles incorporating Locked Nucleic Acid bases are used to detect the mutations. (B) Real-time qualitative method optimized for TERTp c.1-146C > T detection. Two competitive fluorescent probes targeting normal (WT- c.1-146C) and mutated (Mut c.1-146C > T) alleles incorporating Locked Nucleic Acid bases are used to detect the mutations. (C) Real-time qualitative method optimized for FGFR3 c.742C > G detection. A mutation allele-specific primer, a phosphorylated wild-type allele blocker that completely suppresses the amplification of the wild-type allele, a locus reverse primer, and a fluorescent probe for real-time detection of the generated amplicon are used. (D) Real-time qualitative method optimized for FGFR3 c.746C > T detection. A mutation allele-specific primer, a phosphorylated wild-type allele blocker that completely suppresses the amplification of the wild-type allele, a locus reverse primer, and a fluorescent probe for real-time detection of the generated amplicon are used.

Figure 3

(A) Serial dilution detection of TERTp c.1-124C > T. TERTp c.1-124C > T mutated DNA with 50% (wild-type) WT/mutation ratio was diluted in a twofold dilution (eight dilutions) in WT DNA. Detection limit was fixed at the presence of 6.25% of TERTp c.1-124C > T alteration in the total DNA in a reaction with 25 ng of total DNA. Below this limit, mutation detection is not guaranteed. (B) Serial dilution detection of TERTp c.-146C > T. TERTp c.1-146C > T mutated DNA with 50% WT/mutation ratio was diluted in a twofold dilution (eight dilutions) in WT DNA. Detection limit was fixed at the presence of 6.25% of TERTp c.1-146C > T alteration in the total DNA in a reaction with 25 ng of total DNA. Below this limit, mutation detection is not guaranteed.

Figure 4

Performance of different screening methods in non-muscle-invasive bladder cancer (NMIBC) follow-up recurrence detection and in NMIBC diagnosis.

Figure 5

Mutation distribution across follow-up cohort (A) and initial-diagnosis cohort (B).

Figure 6

Cohort’s tumor stage distribution and Uromonitor performance in recurrence detection across tumor stages.

Figure 7

Cohort’s tumor grade and uromonitor performance in recurrence detection across tumor grades.