IMAJ | volume 17
Journal 9, September 2015
pages: 567-570
Summary
Background:
In recent years several reports have been published describing dogs’ ability to detect, by scent, patients with cancer. This ability is based on the sniffing of volatile organic elements that are secreted by malignant cells, react to them.
Objectives:
To evaluate the ability of trained dogs to detect (i) breast cancer cell cultures (MCF7) compared to the control pseudo-normal keratinocyte cell line (HaCaT), and then (ii) melanoma (BG) and (iii) type 2 epithelial lung carcinoma (A549) malignant cell cultures to which they were not previously exposed in the course of their training.
Methods:
Cell cultures were prepared in a standard manner. Two Belgian Shepherd dogs were trained and then tested in a single-blind test (for dogs and trainers) on their ability to detect the "target specimen," a MCF7 breast cancer cell culture. Following this, the ability of the dogs to detect cancer cell cultures that they were not previously exposed to (i.e., A549, BG) was tested. In each test round, four specimens placed in identical blocks were arranged in a line with one meter between them: one target specimen (MCF7, A549, BG), two control specimens (HaCaT), and a sample containing cell culture medium only.
Results:
The two dogs picked out all the target specimens of MCF7 breast cancer cell cultures that they were trained to detect (10/10) as well as all the target specimens that they were not previously exposed to [A549 (5/5) and BG (5/5)], but did not pick out the control specimens or the cell culture medium. Thus, the sensitivity, specificity, and positive and negative predictive values for both dogs were 100%.
Conclusions:
The results of this study support the assumption that cancer cells have a unique odor pattern, and that this odor pattern is common to different types of cancer.