OC-102 A prospective, observational study of surgical and endoscopic rapid evaporative ionisation mass spectrometry (REIMS) for real time analysis of the colonic mucosal lipidome in colorectal cancer

L Muirhead,JM Kinross,R Preece,A Speller,O Golf,R Goldin,A Darzi,Z Takats
DOI: https://doi.org/10.1136/gutjnl-2015-309861.102
IF: 24.5
2015-06-01
Gut
Abstract:REIMS permits real time analysis of tissue specific molecular information contained within an electrocautery smoke plume.1This study validated its diagnostic accuracy in colorectal cancer (CRC) and developed a novel endoscopic methodology. Patients undergoing elective surgery for CRC were recruited at St. Mary’s hospital London. 1. A standard electrosurgery handpiece using monopolar diatheramy on 30 W cutting mode (ValleylabTM) was modified to allow aspiration of the electrosurgical aerosol to a Xevo G2-S iKnife QTof mass spectrometer (Waters Corporation). Ex-vivo analysis was performed and H&E staining and histological examination was carried out for validation. Multivariate analysis was performed using principal component analysis and linear discriminant analysis. Classification of each tissue type was performed using leave-one-patient-out cross-validation and receiver operating curves were generated. 2. Ex-vivocolorectal tissue was sampled using an endoscopic snare (Olympus Medical) modified to allow aspiration of the electrosurgical aerosol. 40 consecutive patients were recruited (22 male, median age 68 y, range 47–90). 23 tumour samples were analysed: 10 rectal adenocarcinomas and 13 colonic adenocarcinoma were analysed with TNM stages T2 (8), T3 (11) T4 (4), N0 (12), N1 (6) N2 (5), M0 (22), M1 (1). Distinction of malignant colorectal tissue was achieved with a classification accuracy of 94.4% and a sensitivity of 92.4%, Specificity 96.8% (ROC AUC 0.98). The diagnostic accuracy for dysplasia was 93.7% (Specificity 95.1%, sensitivity 85.7%, AUC 0.97). Distinction of tissue histology was possible based on membrane lipid ratios. Increases in glycerophospholipids (p = 0.0027), and triacylglycerols (p = 0.0004) were seen in healthy mucosa and increased prostaglandin D2 was found in malignant tissue (p = 0.0002). Endoscopic analysis was technically feasible and could identify colonic wall sub layers with a tissue classification accuracy of 88%. Discrimination of healthy colorectal tissue, adenomatous polyps and cancerous tissue (12 patients) was achieved with a tissue classification accuracy of 87.5%. REIMS is able to diagnose dysplasia and malignant colorectal tissue with a high degree of accuracy in real time. This technology could improve the efficacy and safety of the endoscopic treatment of colon cancer. L. Muirhead: None Declared, J. Kinross Grant/ Research Support from: Imperial College BRC funding, Conflict with: Director of getwell media, R. Preece: None Declared, A. Speller: None Declared, O. Golf: None Declared, R. Goldin: None Declared, A. Darzi: None Declared, Z. Takats: None Declared. Balog J, Sasi-Szabo L, Kinross J, et al. Intraoperative tissue identification using rapid evaporative ionization mass spectrometry. Sci Transl Med. 2013;5(194):194ra93
gastroenterology & hepatology
What problem does this paper attempt to address?