Colorectal cancer laboratory screening using a multiplex immunoassay based on automated protein-array system for tumor markers

Background: Serum tumor markers (STM) are chemical substances generated by the reactions of human body to certain tumors, or expressed and synthesized by the gene in tumor cell, such as proteins (glycoproteins), enzymes (isoenzymes), or peptide hormones, which may reveal the presence of cancer. Increased serum levels of STM are significantly associated with certain tumor or carcinoma. Thus, the elevation of STM levels in human serum can be useful for early diagnosis of cancer or recurrences, and monitoring curative effect of chemotherapy. Simultaneous measurement of STM can increase the single diagnostic accuracy. Various approaches have been proposed to perform multianalyte immunoassays, including label-free immunoassays and labeled probe methods, which are difficult to develop because of the lack of a detectable protein signal or a signal too weak to quantify the trace amount of analytes. Multilabel and spatially resolved assays can offer amplified detection signals for multianalyte immunoassays. Unfortunately, they need several labels, such as radioisotopes, fluorescence dyes, enzymes, metal ions, or quantum dots, which limits their application. The electrochemical immunoassay has the advantages of a small analyte volume required, low detection limit, simple instrumentation, and minimal manipulation, thanks to a miniaturized assay system. The aim of this study was to determine whether a multiplex immunoassay based on an automated protein-array chip system technology can be useful for early detection of colorectal cancer (CRC). Material and Methods: We measured a panel of STM (carcinoembryonic antigen [CEA], cancer antigen [CA] 19-9 and 72-4, cytokeratin fragment [CYFRA] 21-1, and osteopontin) in a selected homogeneous population of 102 consecutive patients (median age 66 years, range 42-75 years) with confirmed Dukes’ B, G1-2, colorectal adenocarcinoma (cases), and in a group of 99 age- and sex-matched patients suffering from benign colorectal diseases (controls). Written informed consent were obtained from all the participant. CEA was measured by a time-resolved immunofluorimetric assay, while CA 19-9, CA 72-4 and CYFRA 21-1 by immunoenzymometric assays. Osteopontin was measured using an ELISA based on rabbit polyclonal antibodies against peptides 288-304 and 211-228 of human osteopontin. The cutoff values (at 95% specificity) were the followings: 4.9 ng/mL (CEA), 32.6 U/mL (CA 19-9), 8.5 U/mL (CA72-4), 2.7 ng/mL (CYFRA 21-1), and 811.9 pmol/mL (osteopontin). Results: The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were: 0.44, 0.91, 0.93, 0.61, 0.67 (OR=0.08, 95% CI 0.03-0.17, p<0.0001) for CEA; 0.19, 0.27, 0.24, 0.21, 0.23 (OR=0.62, 95% CI 0.30-1.17, p=0.14) for CA 19-9; 0.31, 0.90, 0.76, 0.56, 0.60 (OR=0.05, 95% CI 0.02-0.11, p<0.0001 for CA 72-4; 0.35, 0.37, 0.36, 0.36, 0.39 (OR=0.91, 95% CI 0.51-1.62, p=0.76) for CYFRA 21-1; 0.45, 0.31, 0.40, 0.36, 0.38 (OR=1.80, 95% CI 1.01-3.20, p=0.044) for osteopontin. The combination of five markers achieved 94.1% sensitivity, and 100% specificity. Conclusions: When a simultaneous study of a panel of STM is required and the sample volume is limited, multiplex immunoassay can a reliable tool for studying patients undergoing laboratory screening test for CRC.