Early Detection of Lung Tumors by Sniffer Dogs - Evaluation of Sensitivity and Specificity

Overview

Some groups reported that sniffer dogs can be applied to detect lung cancer in the exhaled breath of patients. Therefore, breath samples (BS) of patients are collected. Five sniffer dogs are trained to distinguish between the BS of patients with lung cancer and healthy individuals (controls). In a prospective, randomized blinded study the dog's ability to differentiate between BS of i) patients with lung cancer, ii) patients with inflammatory airway disease, but no evidence of cancer and iii) healthy individuals is tested.

All breath samples (BS) are collected in patients on the basis of radiologic findings before any intervention (bronchoscopy, biopsy or surgery) has been performed. The allocation to study groups is made following diagnostic work up and surgery. Three study groups are defined using the following inclusion and exclusion criteria: * lung cancer: male \& female, age 18-80, competent, confirmed lung cancer in diagnostic work up and surgery, no history or present other tumor disease * healthy individual: male \& female, age 18-80, competent, no history or present other tumor disease including lung cancer, no pathological lung function tests * chronic obstructive lung disease: male \& female, age 18-80, competent, no history or present other tumor disease including lung cancer, pathological lung function tests For each patient, i) history, ii) present medication, iii) lung function tests are documented. For the lung cancer patients, the tumor stage following surgery is documented. Five sniffer dogs are trained using BS of lung cancer patients and healthy individuals. After completion of the training the ability of the dogs to differentiate between the groups is tested: * 5 BS are presented in 1 experiment: * Test I: lung cancer vs healthy individual * Test II: lung cancer vs chronic obstructive pulmonary disease * Test III: lung cancer vs healthy individuals \& bronchitis * for every experiment 1 BS of a lung cancer patient is used * the lung cancer sample is placed randomly in one of the five test tubes * the other test tubes are used for BS of healthy individuals or chronic obstructive pulmonary disease patients or both * neither the dog handler nor the dog know the loading of the test tubes * experiments are repeated 5 to 10 times For analysis, the sensitivity and specificity of the dog's reaction is determined. The patient's medication, smoking habits, age and gender are analysed to rule out confounders or bias.