Optimization of nonmuscle invasive bladder cancer recurrence detection using a urine based FGFR3 mutation assay

Abstract

Purpose: FGFR3 mutations occur in 70% of nonmuscle invasive bladder tumors. Although urine based FGFR3 mutation analysis can detect recurrence, its sensitivity may be limited if samples have few or no tumor cells. We determined whether test sensitivity depends on tumor size and the time point of urine collection, and how to increase sensitivity.

Materials and methods: A total of 440 urine samples from 18 patients with a suspicious bladder lesion at cystoscopy were collected during 6 days before surgery. Eight patients (300 samples) had an FGFR3 mutant tumor, including 4 each with a tumor greater than 3 and less than 1.5 cm. Polymerase chain reaction based FGFR3 analysis was done on all tumors and urine samples.

Results: FGFR3 mutations were detected in 257 of the 300 urine samples (86%) from patients with an FGFR3 mutant tumor. Assay sensitivity was 100% for tumors greater than 3 cm and 75% for tumors less than 1.5 cm. It increased to 100% in patients with a less than 1.5 cm tumor when samples were pooled during 24 hours. Sensitivity was not influenced by the time of urine collection. All urine samples from patients with an FGFR3 wild-type tumor were negative for FGFR3 mutation.

Conclusions: The sensitivity of tumor detection increased with tumor size. FGFR3 assay sensitivity depends on the number of shed tumor cells and improves by increasing urine volume. These findings suggest that there is an upper limit to the sensitivity of the FGFR3 assay when 1 urine sample is analyzed. This may also apply to other DNA or RNA based assays.