A novel methodology for in vivo endoscopic phenotyping of colorectal cancer based on real time analysis of the mucosal lipidome: A prospective observational study of the iknife

Jl Alexander1, J Balog1, A Scott1, L Gildea1, A Speller1, L Muirhead1, O Golf1, K Veselkov1, J Kinross1, S Rasheed2, J Teare1, P Tekkis2, A Darzi1, Z Takats1. 1Imperial College London, 2Royal Marsden Hospital

Introduction

Rapid Evaporative Ionization Mass Spectrometry (REIMS) provides the operating surgeon with tumour phenotypic data in near real time based on analysis of the mucosal lipidome.(1) This study determined the accuracy of REIMS for detecting established histological biomarkers of poor prognostic outcome in colorectal cancer.

Methods

This was a prospective, observational multi-centred cohort study. Patients undergoing elective surgical resection for colorectal cancer were recruited at St. Mary’s Hospital London and The Royal Marsden Hospital, UK. Ex-vivo analysis was performed on resected tumour tissue using a standard electrosurgery handpiece using monopolar diathermy on cutting mode (ValleylabTM) with aspiration of the electrosurgical aerosol to a Xevo G2-S iKnife QTof mass spectrometer (Waters Corporation). Histological examination was carried out for validation purposes. Multivariate analysis was performed using principal component analysis and linear discriminant analysis. Classification of each individual tissue type was performed using leave-one-patient-out cross-validation and receiver operating curves (ROC) were generated.

Results

28 consecutive patients were recruited (12 male, median age 71, range 35-89). REIMS was able to distinguish the histopathological tissue type (Adenoma, adenocarcinoma, mucinous, GIST or signet ring) with 80.4% overall accuracy. It was able to diagnose well or moderately differentiated tumours with a sensitivity of 89.5% and a specificity of 65.9%, ROC AUC = 0.90. The AUC for Lymph node micrometastases, Extra Mural Vascular Invasion (EMVI), and tumour budding was 0.80, 0.86 and 0.84 respectively. Distinction of tissue histology was possible based on differences in membrane lipid ratios. Increases in glycerophospholipids (p=0.0027), and triacylglycerols (p=0.0004) were seen in healthy mucosa and an increased prostaglandin D2 was found in malignant tissue (p= 0.0002).

Conclusion

REIMS can accurately establish histological features of poor prognostic outcome in colorectal cancer based on chemical analysis of the mucosal lipidome. This technology could assist in the real time assessment of colorectal tumours during endoscopic and surgical resections. We have developed a modified endoscopic snare to allow deployment of this technology in vivo, and a clinical study is now underway.

 

1.         Balog J, Sasi-Szabo L, Kinross J, Lewis MR, Muirhead LJ, Veselkov K, et al. Intraoperative tissue identification using rapid evaporative ionization mass spectrometry. Science translational medicine. 2013;5(194):194ra93.

« Return to SAGES 2016 abstract archive