Ashley D. Bondl, Michael D. Burkittl, David Sawbridgel, Bernard M. Corfe2,3 , Chris S. Probertl
1) Unit of Gastroenterology Research, Dept of Molecular and Cellular Physiology, Institute of Translational Medicine, University of Liverpool, Liverpool;
2) Molecular Gastroenterology Research Group, Department of Oncology;
3) Insigneo Institute for in silico Medicine, University of Sheffield Sheffield, UK
Background & Aims: Colorectal cancer screening programmes that target detection and excision of adenomatous colonic polyps have been shown to reduce colorectal cancer related mortality. Many screening programmes include an initial faecal occult blood test (FOBt) prior to colonoscopy. To refine the selection of patients for colonoscopy other faecal-based diagnostic tools have been proposed, including tumour M2-pyruvate kinase (tM2-PK). To determine whether tM2-PK quantification may have a role in diverse settings we have assessed the assay in a cohort of patients derived from both the England bowel cancer screening programme (BCSP) and symptomatic individuals presenting to secondary care.
Method. Patients undergoing colonoscopy provided faecal samples prior to bowel preparation. Faecal tM2-PK concentrations were measured by ELISA. Sensitivity, specificity, positive predictive value, negative predictive value and ROC analyses were calculated.
Results. Ninety-six patients returned faecal samples: 50 of these with adenomas and 7 with cancer. Median age was 68. Median faecal tM2-PK concentration was 3.8 U/mL for individuals without neoplastic findings at colonoscopy, 7.7 U/mL in those with adenomas and 24.4 U/mL in subjects with colorectal cancer (both, p=0.01). ROC analysis demonstrated an AUROC of 0.66 (sensitivity 72.4%, specificity 48.7%, positive predictive value 67.7%, negative predictive value 36.7%). Amongst BCSP patients with a prior positive FOBt faecal tM2-PK was more abundant (median 6.4 U/mL, p=0.03) and its diagnostic accuracy was greater (AUROC 0.82).
Conclusion. Our findings confirm that faecal tM2-PK ELISA may have utility as an adjunct to FOBt in a screening context, but do not support its use in symptomatic patients.