OBJECTIVE: The aim of this study was to determine whether the volatile organic compounds (VOCs) pattern in colorectal cancer (CRC) patients is modified by curative surgery for a potential application in the oncologic follow-up. BACKGROUND: CRC has been proved to induce metabolic derangements detectable by high through-output techniques in exhaled breath showing a specific pattern of VOCs. METHODS: Forty-eight CRC patients and 55 healthy controls (HC) entered the study. Thirty-two patients (M/F: 1.4; mean age 63 years) attended the oncologic follow-up (mean 24 months) and were found disease-free. Breath samples were collected under similar environmental conditions into a Tedlar bags and processed offline by thermal-desorption gas chromatography-mass spectrometry (TD-GC-MS). VOCs were selected by U test to build a Probabilistic Neural Network (PNN) model to set-up a training phase, which was cross-validated using the leave-one out method. RESULTS: A total of 11 VOCs were finally selected for their excellent discriminant performance in identifying disease-free patients in follow-up from CRC patients before surgery, (sensitivity 100%, specificity 97.92%, accuracy 98.75%, and AUC: 1). The same VOCs pattern discriminated follow-up patients from HC, with a sensitivity of 100%, specificity of 90.91%, accuracy of 94.25%, and AUC 0.959. CONCLUSIONS: Exhaled VOCs pattern from CRC patients is modified by cancer removal confirming the tight relationship between tumor metabolism and exhaled VOCs. PNN analysis provides a high discriminatory tool to identify patients disease-free after curative surgery suggesting potential implications in CRC screening and secondary prevention.
Effects of Curative Colorectal Cancer Surgery on Exhaled Volatile Organic Compounds and Potential Implications in Clinical Follow-up
MEMEO, RICCARDO;
2015-01-01
Abstract
OBJECTIVE: The aim of this study was to determine whether the volatile organic compounds (VOCs) pattern in colorectal cancer (CRC) patients is modified by curative surgery for a potential application in the oncologic follow-up. BACKGROUND: CRC has been proved to induce metabolic derangements detectable by high through-output techniques in exhaled breath showing a specific pattern of VOCs. METHODS: Forty-eight CRC patients and 55 healthy controls (HC) entered the study. Thirty-two patients (M/F: 1.4; mean age 63 years) attended the oncologic follow-up (mean 24 months) and were found disease-free. Breath samples were collected under similar environmental conditions into a Tedlar bags and processed offline by thermal-desorption gas chromatography-mass spectrometry (TD-GC-MS). VOCs were selected by U test to build a Probabilistic Neural Network (PNN) model to set-up a training phase, which was cross-validated using the leave-one out method. RESULTS: A total of 11 VOCs were finally selected for their excellent discriminant performance in identifying disease-free patients in follow-up from CRC patients before surgery, (sensitivity 100%, specificity 97.92%, accuracy 98.75%, and AUC: 1). The same VOCs pattern discriminated follow-up patients from HC, with a sensitivity of 100%, specificity of 90.91%, accuracy of 94.25%, and AUC 0.959. CONCLUSIONS: Exhaled VOCs pattern from CRC patients is modified by cancer removal confirming the tight relationship between tumor metabolism and exhaled VOCs. PNN analysis provides a high discriminatory tool to identify patients disease-free after curative surgery suggesting potential implications in CRC screening and secondary prevention.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
