Background: Breast cancer is a leading cause of cancer death worldwide. Several studies have demonstrated that dogs can sniff and detect cancer in the breath or urine sample of a patient. This study aims to assess whether the urine sample can be used for breast cancer screening by its fingerprints of volatile organic compounds using a single trained sniffer dog. This is a preliminary study for developing the “electronic nose” for cancer screening. Methods: A nine-year-old female Labrador Retriever was trained to identify cancer from urine samples of breast cancer patients. Urine samples from patients histologically diagnosed with primary breast cancer, those with non-breast malignant diseases, and healthy volunteers were obtained, and a double-blind test was performed. Total of 40 patients with breast cancer, 142 patients with non-breast malignant diseases, and 18 healthy volunteers were enrolled, and their urine samples were collected. Results: In 40 times out of 40 runs of a double-blind test, the trained dog could correctly identify urine samples of breast cancer patients. Sensitivity and specificity of this breast cancer detection method using dog sniffing were both 100%. Conclusions: The trained dog in this study could accurately detect breast cancer from urine samples of breast cancer patients. These results indicate the feasibility of a method to detect breast cancer from urine samples using dog sniffing in the diagnosis of breast cancer. Although the methodological standardization is still an issue to be discussed, the current result warrants further study for developing a new breast cancer screening method based on volatile organic compounds in urine samples.

(1) Background: Previous reports have indicated that cancers of the stomach, lung, and pancreas can be detected by dog sniffing, but results have been varied. Here, a highly trained dog was used to determine whether urine from patients with cervical premalignant lesions and malignant tumors have a cancer-specific scent. (2) Methods: A total of 195 urine samples were collected from patients with cervical cancer, cervical intraepithelial neoplasia grade 3 (CIN3), benign uterine diseases, and healthy volunteers. Each test was performed using one urine sample from a cancer patient and four samples from different controls. Each of the five urine samples was placed in a separate box. When the cancer sniffing dog stopped and sat in front of the box with a sample from a cancer patient, the test was considered as positive. (3) Results: 83 patients with cervical cancer (34 cases of cervical cancer and 49 cases of cervical intraepithelial neoplasia grade 3), 49 patients with uterine benign diseases, and 63 healthy volunteers were enrolled, and their urine samples were collected. In 83 times out of 83 runs in a double-blind test, the trained dog could correctly identify urine samples of cervical cancer patients. (4) Conclusion: A trained dog could accurately distinguish the urine of all patients with cervical cancer or CIN3, regardless of the degree of cancer progression.

Cyclo-oxygenase (COX)-2 is not usually detectable in normal tissues but is induced in inflammation and carcinogenesis. The level of COX-2 is elevated in cancer tissues of the colon, bladder, and skin. In the esophagus, squamous cell carcinoma and adenocarcinoma are known to express COX-2. The purpose of this study was to clarify the association of COX-2 expression with clinicopathological factors of squamous cell carcinoma. The immunohistochemical expression of COX-2 was examined in 48 surgical specimens of esophageal squamous cell carcinoma. Although COX-2 over-expression was more frequently observed in tumors invading the submucosa (T1b, 76.4%), muscularis propria (T2, 57.1%), adventitia, or adjacent organs (T3∼4, 83.3%), even 33.3% of mucosal cancers, such as T1a, showed COX-2 over-expression. COX-2 over-expression was present in 82.3% of lymph node-negative patients but in only 54.8% of lymph node positive patients. There was no difference in COX-2 over-expression between the earlier stages (0 and I, 60%) and more advanced stages (II∼IV, 69.6%). COX-2 over-expression did not correlate with survival during 3 years of follow-up. These findings suggest that COX-2 is associated with the phenotype of the esophageal squamous cell carcinoma cells, including superficial cancer cells, and may be related to tumor growth in esophageal squamous cell carcinoma.