[Identification of colorectal cancer using proteomic patterns in serum]

At present there is no serological parameter with high sensitivity and specificity to diagnose colorectal cancer. This study aimed at establishing a serum protein fingerprinting technique coupled with a pattern-matching algorithm to distinguish colorectal cancer from benign colorectal diseases and healthy people.

Preliminary group and test group were both random selected serum samples, which each comprises 73 cases of colorectal cancer, 31 healthy people, and 16 cases of benign colorectal diseases. The sera of the test group was combined with the surface of the IMAC3 proteinchip. Then reading data of surface enhanced laser desorption/ionization- time of flight-mass spectrometry (SELDI-TOF-MS) was analyzed by Biomarker Wizard software and Biomarker pattern software to get a classification rule of tree, which is standard configuration that can distinguish the sera of colorectal cancer patients from the sera of benign colorectal disease patients and healthy people, and the standard was approved test validity by double-blindly in the test group.

At the key M/Z values of 4 467 Da, 8 131 Da, 8 939 Da, 9 192 Da, 9 134 Da, 8 221 Da, 5 928 Da, 8 324 Da, 11 732 Da, protein contents of the three classes in preliminary group are obviously different by the software analysis. And the classification rule of tree was discovered, and corresponding correct ratio was 98.33% (118/120), corresponding sensitivity was 97.26% (71/73) and corresponding specificity was 100% (47/47); after double-blind examining the test group with that rule, the corresponding correct ratio was 96.77% (116/120),the corresponding sensitivity was 95.89% (70/73) and the corresponding specificity was 97.87%(46/47).

Via comparative proteomics research in the sera of colorectal cancer patients and benign colorectal disease patients and healthy people, the group of protein was obviously different in pathological process in discovering colorectal cancer.