Candidate biomarker verification: Critical examination of a serum protein pattern for human colorectal cancer
Article first published online: 25 APR 2012
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
PROTEOMICS - Clinical Applications
Volume 6, Issue 3-4, pages 182–189, April 2012
How to Cite
Albrethsen, J., Rikke Bøgebo, Møller, C. H., Olsen, J. A., Raskov, H. H. and Gammeltoft, S. (2012), Candidate biomarker verification: Critical examination of a serum protein pattern for human colorectal cancer. Prot. Clin. Appl., 6: 182–189. doi: 10.1002/prca.201100095
- Issue published online: 25 APR 2012
- Article first published online: 25 APR 2012
- Accepted manuscript online: 4 JAN 2012 10:02AM EST
- Manuscript Accepted: 16 NOV 2011
- Manuscript Revised: 11 NOV 2011
- Manuscript Received: 26 OCT 2011
- Colorectal cancer;
- Translational research
We critically examine a candidate serum protein pattern for human colorectal cancer (CRC) with respect to reproducibility, sample handling, and disease specificity.
Serum samples from CRC patients, patients with benign colon tumors and healthy individuals, were obtained at two collection sites and analyzed by SELDI-TOF MS on 8 days, over a period of 5 weeks. The spectra were subjected to multivariate analysis. Tissues from normal colon and CRC were analyzed by SELDI-TOF MS. Selected mass peaks were identified.
Using an elaborate experimental design we developed a multivariate classifier that correctly classified CRC and control serum measured on an independent day. The classifier did not discriminate between samples from CRC patients and patients with benign colon tumors, and, secondly, did not correctly classify serum from an independent collection site. All discriminatory mass peaks were identified as high abundant plasma proteins. Tissue profiling provided support of increased proteolytic activity in CRC tissue.
Conclusion and clinical relevance
Critical verification did not justify advancing the identified CRC serum protein pattern into clinical validation without improvement. We believe that proteomics biomarker research could benefit if the presented, or a similar, verification scheme was more commonly employed in explorative biomarker studies.