Metallic Nanocluster-Based Sniffing Device for Identification of Malignancy in Gastric Cancer Tissues

Analysis of volatile organic compounds (VOCs) emitted from the human organism can provide useful information about normal physiological processes as well as pathological disorders. The aim of this work is to develop a portable, inexpensive, on-site, fast, small design paper-based device based on the electronic nose (E-nose) concept for the detection of malignancy in gastric cancer (GC) by analyzing volatile organic compounds emitted from human tissue. For this purpose, cancerous and noncancerous tissue samples were provided from 22 patients diagnosed with GC, esophageal cancer (EC), and neuroendocrine carcinoma (NEC). In this system, eight types of fluorescent metallic nanoclusters (NCs) as sensing elements were immobilized on a paper substrate. A microcentrifuge tube was used to create a reaction environment to communicate between a paper device containing NCs and VOCs emitted from tissue samples. After exposing the sensors to headspace of the tissues, the emitted VOCs reacted with NCs and led to changes in the fluorescence intensity of NCs. For pattern recognition and statistical analysis, chemometric methods including principal component analysis, linear discriminant analysis, and partial least square discriminant analysis were applied. The sensor array was able to discriminate between fresh cancerous and normal tissues with 95% accuracy. In summary, we fabricated an optical nose device that can sniff the odor of tissues and then identify the malignant tissues in less than 4 h. So, it has the potential for point-of-need applications in the hospital and can offer results in the same time duration required for a patient to recover from anesthesia.

Automated summary

This study aimed to develop a portable, inexpensive, on-site, fast, small design paper-based device based on the electronic nose (E-nose) concept for the detection of malignancy in gastric cancer (GC) by analyzing volatile organic compounds emitted from human tissue. The sensing elements, eight types of fluorescent metallic nanoclusters, were immobilized on a paper substrate, and reacted with the emitted VOCs from tissue samples, leading to changes in fluorescence intensity. Chemometric methods were applied for pattern recognition and statistical analysis, and achieved 95% accuracy in discriminating between fresh cancerous and normal tissues.