A serum proteomic pattern for the detection of colorectal adenocarcinoma using surface enhanced laser desorption and ionization mass spectrometry.

PURPOSE

New serum biomarkers are needed to improve the early detection of colorectal adenocarcinoma. We performed surface enhanced laser desorption and ionization time-of-flight mass spectrometry (SELDI-TOF-MS) to screen for differentially expressed proteins in serum and build a proteomic diagnostic pattern for the detection of colorectal adenocarcinoma to improve the prognosis of patients with this disease.

EXPERIMENTAL DESIGN

In an attempt to improve current approaches to the serologic diagnosis of colorectal cancer, we analyzed serum samples from subjects with or without colorectal cancer using SELDI-MS. Using a case-control study design, SELDI-MS profile of serum samples from 74 colorectal adenocarcinoma patients were compared with 48 age-and sex-matched healthy subjects using a ProteinChip reader, PBSII-C. Proteomic MS spectra were generated using IMAC3 chips, and protein peaks clustering and classification analyses were performed to build a proteomic pattern that could differentiate patients with colorectal adenocarcinoma from healthy subjects utilizing Biomarker Wizard and Biomarker Patterns software packages, respectively. The constructed pattern was then used to test an independent set of masked serum samples from 60 colorectal cancer patients and 39 healthy subjects.

RESULTS

Among the differentially expressed protein peaks identified by SELDI-MS profiling that had the ability to distinguish between patients and healthy subjects, we determined a minimum set of two protein peaks for system training and for developing a decision classification pattern. Masked analysis of an independent set of serum samples showed the diagnostic pattern could differentiate patients with different stages of colorectal cancer from healthy subjects with a sensitivity of 95.00 percent and specificity of 94.87 percent.

CONCLUSION

SELDI-TOF-MS profiling of serum proteins combined with bioinformatics tools can be applied to accurately differentiate patients with colorectal cancer from healthy subjects. The high sensitivity and specificity achieved by the constructed clustering analysis algorithm show great potential for the early detection of colorectal cancer.