Oral Glucose Tolerance Test as a Breath Challenge

Oral Glucose Tolerance Test as a Breath Challenge Diagnostic Tool in Lung Cancer (1) * Alkoby Layah (3) Abud Hawa Manal (1) Bar Yair (4) Cancilla John(1) Dekel Shlomi (1) Feinberg Tali (1) Gaimor Naomi (3) Haick Hossam (2) Ilouze Maya (1) Onn Amir (4) Torrecilla Jose (1, 2)* Peled Nir Thoracic Cancer Research and Detection Center, Sheba Medical Center; Thoracic Cancer Unit, Davidoff Cancer Centerl; Technion- Israel Institute of Technology; Complutence University of Madrid Spain Introduction Ø Exhaled breath obtained from affected individuals is unique and thereby can be used to diagnose Lung Cancer Ø The differences in exhaled breath samples are expressed by Volatile Organic Compounds (VOCs) Ø VOCs are organic metabolites that are released by the cancer cells and/or by the surrounding environment Methods Results Research Goals: Pre-Glucose Administration Analysis Ø To evaluate the role of glucose metabolism on the volatile signature Ø To pinpoint the molecules that associate between Warburg and the exhaled breath ACCURACY Ø VOCs reflect the unique metabolic and biochemical activity in malignant cells TOTAL (%) 91. 67 HEALTHY (%) 80 LC (%) 100 Our previous studies showed that: • Cancer cells show specific VOCs signature. 18 active, naïve lung cancer patients and 22 matching high Ø risk control patients were recruited to the study PTR Analysis: successfully identified 14 masses (m/e 38, 44, 50, 52, 6, 72, 107, 109, 124, 125, 126, 136, 148, 164) that enable us to accurately distinguish between high risk for LC and active, naïve LC Results of Δ Glucose ( Post- Pre) Exhaled breath samples were collected from patients before Ø and after an Oral Glucose Challenge Test Feature Selection m/e 43, 44, 131 & 148 Differences in Glucose Uptake Breath Samples were analyzed by PT-MRS for unique mass Ø differences Glucose Uptake • Cancer cells show specific VOCs signature per histology and genetic profile. 100% 50% LC 0% 41 43 44 60 61 108 119 131 132 134 142 145 148 154 167 controls -50% -100% m/e Glucose Uptake Analysis: A feature selection with the combination of m/e 43, 44, 131 & 148 allowed the design of a multi-layer perception model with a K-fold cross-validation accuracy of 90% Conclusion: This study shows that breath analysis can discriminate between the high-risk and LC group. It also demonstrates that glucose metabolism leaves a unique VOC pattern in the LC group. Furthermore, it uses feature selectiona to identify lung cancer by applying these masses in a multi-layer perception model. These findings may assist in the