Introduction Colorectal carcinoma is the second most common cause of cancer death in the western world. Alarm features such as weight loss, change in bowel habit, iron deficiency anaemia, rectal bleeding or a palpable abdominal mass have been proposed as the optimal method of identifying patients in primary care who need early assessment and investigations. However, the accuracy of these clinical features in identifying high-risk patients is quite variable. Tumour M2PK is the dimeric isoform of pyruvate kinase, which is upregulated in proliferating tissues. Several studies show good sensitivity and specificity of faecal M2PK for colorectal cancer. There is still a paucity of data for M2PK in identifying high-risk patients in secondary care.

Methods To investigate the clinical utility of faecal M2PK in detection of colorectal cancer, colonic polyps or any other significant colonic pathology in routine clinical practice and assess its suitability as a tool for identifying patients referred to secondary care for early lower gastro-intestinal investigations. Patients undergoing colonoscopy or barium enema (routine/urgent/cancer pathway) were recruited from both general gastroenterology and surgical outpatient clinics. Patients returned a stool sample in the specimen container provided. A quantitative analysis by sandwich ELISA using two different monoclonal antibodies to specifically recognise the dimeric form of M2PK was used. A positive test result was defined as >4 kU/l. Data were collected prospectively including demographics, clinical presentation and indication for colonoscopy as well as findings at lower GI investigations. The data were analysed for sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV).

Results 48 patients (60.4% females) were recruited from March to October 2007. The mean age was 59.6±15.1 years, with no significant difference in M2PK levels between males and females. Elevated M2PK levels were found in 43.8% of patients. Abnormal pathology was found in 58.3% of patients. The diagnoses included polyps (n=9), sigmoid colon cancer (n=1), diverticular disease (n=16), collagenous colitis (n=1), and miscellaneous (n=2). Raised stool M2PK was found in 55.6% of patients with polyps. The sensitivity of M2PK for detecting significant colorectal pathology (polyps/cancer) was 50%. Specificity was 57.9% at 4 kU/l, and increased to 84.2% with a modified cut-off of 8 kU/l (sensitivity unchanged). PPV was 23.8% and NPV 81.5% (8 kU/l: 45.5% and 86.5%, respectively).

Conclusion In this highly selected cohort of patients referred to secondary care, a modified cut-off of M2PK may act as a useful adjunct in identifying high-risk patients.