Introduction

The objective of this study was to investigate whether trained dogs can scent tumour related volatile compounds in dogs’ urine and accurately detect canine urinary transitional cell carcinoma (TCC).

Methods

Urine samples from healthy dogs without urinary tract abnormalities (control), dogs with non-malignant urinary tract disease (control), and dogs with urinary TCC (positive) were collected prospectively. All dogs underwent bladder ultrasound and urinalysis. Urine samples were frozen in glass jars until analysis. TCC was diagnosed upon cytology and/or histopathology. Dogs were naïve of chemotherapy, NSAIDs were permitted.

Dogs previously trained to detect human bladder and prostate cancer were presented with randomly allocated sample lines of 2 control samples, 1 positive sample, and 1 biological sample irrelevant to this study (filler), or lines of only control and filler samples. Dog handlers were blinded to the samples’ identity until the dog’s behavior had been recorded, after which the data collection software revealed the correct response.

The dogs were rewarded when correct.

Results

Three trained dogs were tested under double-blind conditions and challenged with 60 controls, 30 fillers, and 30 TCC samples. Overall, no fillers were incorrectly indicated, 8.3% of controls were incorrectly indicated, and 80% of TCC were correctly indicated. The difference in the proportions of control and target samples indicated was highly significant for all three dogs (overall p value <0.001).

Conclusion

This study suggests that dogs can be trained to detect canine TCC in urine and could lead to the development of an electronic nose as an additional diagnostic test for TCC.

Funding Information

No conflict of interest or funding