BACKGROUND: Screening for colorectal cancer (CRC) requires non-invasive methods of high diagnostic sensitivity and specificity. We evaluated the measurement of genetic and protein biomarkers of CRC in stool samples with the aim of testing their clinical utility in a CRC screening program. PATIENTS AND METHODS: Individuals aged 53-75 years who were at risk of CRC and immunochemical fecal occult blood test (iFOBT) positive were invited to submit stool samples for molecular testing prior to colonoscopy. KRAS codon 12 Gly→Asp, Gly, Val, and codon 13 Gly→Cys gene mutations were tested using an in-house real-time ARMS PCR method. M2PK levels in stool samples were measured utilizing a commercial ELISA kit. RESULTS: At colonoscopy, 7.6% of patients were found to have CRC, 50% had adenomas, 10.6% had hyperplastic polyps, 20.2% had diverticulosis and hemorrhoids, and 11.6% had normal mucosa. The best sensitivity for CRC (50%) was found in those cases where M2PK and KRAS abnormalities coexisted. M2PK showed a detection rate of 40.3% for adenomas but the combination of M2PK and KRAS abnormalities was found in only 5.7% of adenomas (P <0.01). iFOBT was false positive in 31.8% of cases in which colonoscopy excluded neoplastic lesions, while the coexistence of molecular and enzymatic abnormalities was more specific with false positive rates between 8.3% and 9.0% (P <0.05). CONCLUSION: Our molecular screening approach demonstrates that detection of cancer-associated biomarkers measured in iFOBT-positive stool samples could help separate true from false positives in a FOBT-based screening process. M2PK showed particular promise for the detection of CRC and adenomas.
BACKGROUND: Screening for colorectal cancer (CRC) requires non-invasive methods of high diagnostic sensitivity and specificity. We evaluated the measurement of genetic and protein biomarkers of CRC in stool samples with the aim of testing their clinical utility in a CRC screening program. PATIENTS AND METHODS: Individuals aged 53-75 years who were at risk of CRC and immunochemical fecal occult blood test (iFOBT) positive were invited to submit stool samples for molecular testing prior to colonoscopy. KRAS codon 12 Gly→Asp, Gly, Val, and codon 13 Gly→Cys gene mutations were tested using an in-house real-time ARMS PCR method. M2PK levels in stool samples were measured utilizing a commercial ELISA kit. RESULTS: At colonoscopy, 7.6% of patients were found to have CRC, 50% had adenomas, 10.6% had hyperplastic polyps, 20.2% had diverticulosis and hemorrhoids, and 11.6% had normal mucosa. The best sensitivity for CRC (50%) was found in those cases where M2PK and KRAS abnormalities coexisted. M2PK showed a detection rate of 40.3% for adenomas but the combination of M2PK and KRAS abnormalities was found in only 5.7% of adenomas (P <0.01). iFOBT was false positive in 31.8% of cases in which colonoscopy excluded neoplastic lesions, while the coexistence of molecular and enzymatic abnormalities was more specific with false positive rates between 8.3% and 9.0% (P <0.05). CONCLUSION: Our molecular screening approach demonstrates that detection of cancer-associated biomarkers measured in iFOBT-positive stool samples could help separate true from false positives in a FOBT-based screening process. M2PK showed particular promise for the detection of CRC and adenomas.