Canine Olfactory Differentiation of Cancer: A Review of the Literature

TL;DR

This review examines the ability of dogs to detect various cancers through olfactory methods, highlighting variable results, methodological concerns, and the potential for VOC analysis inspired by canine detection.

Contribution

The paper synthesizes existing research on canine cancer detection, critically assesses reproducibility issues, and discusses implications for future VOC-based diagnostic methods.

Findings

Canines detect multiple cancer types with variable sensitivity and specificity.

Methodological inconsistencies affect reproducibility of results.

Dogs' detection inspires VOC analysis for cancer diagnosis.

Abstract

Numerous studies have attempted to demonstrate the olfactory ability of canines to detect several common cancer types from human bodily fluids, breath and tissue. Canines have been reported to detect bladder cancer (sensitivity of 0.63-0.73 and specificity of 0.64-0.92) and prostate cancer (sensitivity of 0.91-0.99 and specificity of 0.91-0.97) from urine; breast cancer (sensitivity of 0.88 and specificity of 0.98) and lung cancer (sensitivity 0.56-0.99 and specificity of 8.30-0.99) on breath and colorectal cancer from stools (sensitivity of 0.91-0.97 and specificity of 0.97-0.99). The quoted figures of sensitivity and specificity across differing studies demonstrate that in many cases results are variable from study to study; this raises questions about the reproducibility of methodology and study design which we have identified herein. Furthermore in some studies the controls used have resulted in differentiation of samples which are of limited use for clinical diagnosis. These studies provide some evidence that cancer gives rise to different volatile organic compounds (VOCs) compared to healthy samples. Whilst canine detection may be unsuitable for clinical implementation they can, at least, provide inspiration for more traditional laboratory investigations.