Proteomics in human disease: cancer, heart and infectious diseases

Serum protein fingerprinting for diagnosis and prognosis evaluation of colorectal cancer

AIM: To identify differentially expressed proteins for diagnosis and prognosis evaluation of colorectal cancer by serum protein fingerprint in colorectal cancer patients.

METHODS: Serum protein fingerprinting was performed by surface-enhanced laser desorption ionization time-of-flight mass spectrometry (SELDI-TOF-MS) in 45 colorectal cancer patients, 14 colorectal cancer having a good prognosis (no postoperative recurrence and metastasis), 13 colorectal cancer patients having a poor prognosis (having recurrence or metastasis), 24 patients with benign gastrointestinal disease, and 155 healthy controls. The Biomarker Wizard software was used to identify differential proteins. Two respective artificial neural network (ANN) models were developed for diagnosis and prognosis evaluation of colorectal cancer.

RESULTS: Seven proteins that displayed significant differential expression were identified (all P < 0.01), and their molecular weight was 4 955 Da, 5 325 Da, 5 890 Da, 6 615 Da, 7 739 Da, 8 109 Da, and 8 575 Da, respectively. Using these seven protein markers, we developed an artificial neural network model for diagnosis of colorectal cancer. Furthermore, five proteins that had a molecular weight of 4 955 Da, 5 325 Da, 5 890 Da, 6 615 Da, and 7 739 Da were used to develop an artificial neural network model for evaluation of the prognosis of colorectal cancer. The sensitivity, specificity, negative predictive value, positive predictive value, and accuracy of the diagnostic model were 82.22%, 80.45%, 94.74%, 51.39% and 80.80%, respectively. The coincidence rate of the prognostic model for evaluation of recurrence and metastasis was 62.96%.

CONCLUSION: SELDI-TOF-MS serum protein fingerprinting allows identification of differentially expressed proteins in colorectal cancer to develop models for diagnosis and prognosis evaluation of the disease.