VOC pattern recognition of lung cancer: a comparative evaluation of different dog- and eNose-based strategies using different sampling materials

• Published in: Acta oncologica Vol. 58; no. 9; pp. 1216 - 1224
• Main Authors: Biehl, Wiebke, Hattesohl, Akira, Jörres, Rudolf A., Duell, Thomas, Althöhn, Ulrike, Koczulla, Andreas Rembert, Schmetzer, Helga
• Format: Journal Article
• Language: English
• Published: England Taylor & Francis 02.09.2019
• Subjects: Adult; Aged; Aged, 80 and over; Animals; Breath Tests - instrumentation; Breath Tests - methods; Conditioning, Psychological; Dogs; Electronic Nose; Exhalation; Female; Glass; Humans; Lung Neoplasms - diagnosis; Male; Masks; Middle Aged; Pets; Reproducibility of Results; Sensitivity and Specificity; Smell; Species Specificity; Teaching; Volatile Organic Compounds - analysis
• ISSN: 0284-186X; 1651-226X
• DOI: 10.1080/0284186X.2019.1634284

Background: It has been reported that canine scent tests offer the possibility to screen for cancer. Assuming that breath samples can be collected with carrier materials, we tested the practicability of different carrier materials to be presented to dogs and validated and compared results with an electronic nose (eNose). Moreover, we hypothesized that cancer detection ability of dogs differs according to their working experience. Methods: In a methodological approach, two dog teams participated, one using experienced working dogs and the other ordinary household dogs to find the most qualified dogs and training method. To find best carrier material for breath sampling we compared charcoal containing glass tubes and fleece masks. In a second validating part, experienced working dogs were trained with improved training strategies. For breath sampling, two different, previously successfully tested fleece-based carrier materials were used: one was used with the dog team and both materials were compared with eNose. Results: In the methodological approach, it turned out that the charcoal-based sampling strategy qualified not sufficiently for VOC-detection. Moreover, we could determine that using experienced working dogs provided several advantages. Overall results of dogs in the validating part regarding specificity were 83%, regarding sensitivity 56%, but with great variability among dogs. Using eNose for breath analysis collected with both fleece carrier materials, specificity was 97% and sensitivity 89-100%. Conclusion: Our data confirmed that the diagnostic accuracy of dogs depended on the type of dog training and on the carrier materials. A comparison of breath samples analysis with an eNose achieved better results for both, sensitivity and specificity than for dogs. The use of fleece masks or fleeces in glass tubes as a sampling material can be recommended as successful VOC carriers, encouraging their use for clinical screenings.