With the use of nanotechnology-based sensors, researchers validate the potentials of analyzing the amounts of volatile organic compounds in exhaled breath of patients diagnosed with disease. The findings suggest that each disease results to a unique volatile organic compound composition.
Since the ancient past, physicians have learned to evaluate their patients based on the smell of their breath, stools, urine, and other bodily fluids. Often, these excreted volatile organic compounds (VOC) are linked to diseases. Exhaled breath is the most accessible and useful VOC for detecting physiological changes in health. To be beneficial in a clinical setting, breath analysis tools should be low-cost, low-energy, miniaturized, easily repeated, and non-invasive to the patient. These criteria are met by nanotechnology-based chemical sensor matrices, but they can only detect a narrow spectrum of disease. Thus, there is still a need to extend this approach so that breath analysis can also classify disease based on the same etiology or manner of causation. From there, the specific disease can be identified through process of elimination from a ranking in order of probability and severity based on the etiology.
A study published by the ACS Nano aims to assess the ability of nanotechnology-based chemical sensor matrices, or nano-arrays, to diagnose and classify diseases by categories (cancerous, inflammatory and neurological). The nano-array consists of a sensing layer, which is a recognition element for the VOCs based on their chemical makeup. The study consisted of 1404 subjects who were either healthy controls or diagnosed with one of the following diseases: chronic kidney failure, idiopathic Parkinson’s disease, atypical Parkinsonism, multiple sclerosis, Crohn’s disease, ulcerative colitis, irritable bowel syndrome, pulmonary arterial hypertension, pre-eclampsia in pregnant women, head and neck cancer, lung cancer, colorectal cancer, bladder cancer, kidney cancer, prostate cancer, gastric cancer, and ovarian cancer.
Each disease had its own unique combination of sensing response in the nanoarrays based on their VOC composition. A series of discriminant factor analyses (DFA) were created based this combination, which allows different diseases to be distinguished. The confounding factors that correlated with the response were mainly age and smoking status, and a few are correlated with gender, which were all taken into account and adjusted. In order to validate the accuracy, 120 random pairs of diseases were blindly classified according to their DFA. In some cases, the accuracy was low, such as in the comparison with gastric vs. bladder cancer, but100% accuracy was also found in 13 different comparisons, with an overall accuracy of 86%.
Altogether, the study reports that there was a strong resemblance between diseases with common pathophysiologies. For instance, the response profiles were highly similar amongst cancerous diseases, diseases with inflammatory activity as well as Parkinsonian-related diseases. These results further demonstrate that similar pathophysiological processes share similar breath patterns, which drives the potential of nano-arrays to categorize diseases.
To gain understanding of the results with the nano-arrays, a second set of breaths were analyzed using gas chromatography linked to mass spectrometry (GC-MS), which is a lab-based technique that can identify different VOCs. The study reports that 13 VOCs differed significantly among diseases. For example, nonanal was significantly lower in Chron’s disease, irritable bowel syndrome, and pre-eclampsia compared to other diseases. Although VOC cannot be used to identify a specific disease, there is, once again, a remarkably distinct combination for each disease after analyzing the amount of 13 VOCs associated with it.
Overall, both the nano-arrays and the GC-MC methods support the idea that diseases can be identified using breath analysis. Although the nano-array would have been a quick and easy diagnostic tool, it gave low accuracy in some trials from this study, so it will be ineffective in clinical use. However, many adjustments can be made to improve these sensors, and more validation studies are required to pave the way towards obtaining this latest type of useful tool that allows for personalized screening, diagnosis and successful intervention of diseases.
Written By: Kim Gotera, BMSc