Detection of BRAF Mutation in Urine DNA as a Molecular Diagnostic for Canine Urothelial and Prostatic Carcinoma

Abstract

Urothelial carcinoma (UC) of the lower urinary tract and prostatic carcinoma (PC) are aggressive genitourinary cancers in dogs, characterized by invasion to surrounding tissues and high metastatic potential. Current diagnosis of canine UC and PC requires histopathological examination of a biopsy. Such specimens require specialized medical equipment and are invasive procedures, limiting the availability of diagnosis by histopathology for many canine patients. Access to a non-invasive means to confirm diagnosis is currently an unmet need. Recently, the canine BRAF V595E mutation was detected in ~80% of canine UCs and PCs. In this study, we developed a droplet digital PCR (ddPCR) assay for detection of the canine BRAF V595E mutation in canine urogenital tumors. The assay was evaluated in DNA samples prepared from biopsy specimens of UC (n = 48) and PC (n = 27), as well and non-neoplastic bladder epithelium (n = 38). In addition the assay was assessed for use with DNA isolated from free catch urine samples derived from canine patients with UC (n = 23), PC (n = 3), as well as from dogs with cystitis and healthy controls (n = 37). In all cases the sensitivity to detect the mutant allele was compared with conventional Sanger sequencing. ddPCR had superior sensitivity for detection of the V595E mutation: 75% of UC, 85% of PC, and 0% of control samples were mutation positive, respectively, and the V595E mutation was detected at a level as low as just 1 in 10,000 alleles (~0.01%). Furthermore, the ddPCR assay identified the mutation in free catch urine samples from 83% of canine UC and PC patients, demonstrating its utility as a non-invasive means of diagnosis. We have shown that ddPCR is a sensitive molecular technique with the potential to facilitate accurate and non-invasive means of canine UC and PC diagnosis.

Fig 1. Performance of the cBRAF V595E assay in a serial dilution of plasmid containing V595E sequence in wild type dog gDNA.

(A) Representative two-dimensional scatter plots of the ddPCR assay in serially-diluted DNA samples. Areas in red rectangles indicate the positive reactions for cBRAF V595E assay. The numbers shown in red rectangles indicate the number of droplets positive for the V595E mutation assay. (B) Correlation of V595E fraction calculated by the dilutions and measured by ddPCR. The V595E mutation was detected up to 0.005% of V595E fraction. Vertical error bars represents the 95% confidence intervals calculated by the Poisson distribution.

Fig 2. Results of cBRAF V595E ddPCR analysis of (A) tissue and (B) urine samples of canine UC and PC patients (x-axis: each sample, y-axis: V595E fraction %).

Black and gray bars indicate UC and PC samples, respectively. Asterisks indicate samples in which the V595E mutation was not detected in Sanger sequencing analysis.

Fig 3. Comparison of the cBRAF V595E fraction between matched tissue and urine samples of six UC/PC cases.

The association between tissue and urine was not evident.

Fig 4. Performance of the cBRAF V595E ddPCR assay at 1–200 ng DNA input.

The V595E fraction showed little variance between a wide range of DNA input. Vertical error bars represents the 95% confidence intervals calculated by the Poisson distribution.