A plasma-based protein marker panel for colorectal cancer detection identified by multiplex targeted mass spectrometry

• Published in: Clinical colorectal cancer 2016-06, Vol.15 (2), p.186-194.e13
• Main Authors: Jones, Jeffrey J, Wilcox, Bruce E, Benz, Ryan W, Babbar, Naveen, Boragine, Genna, Burrell, Ted, Christie, Ellen B, Croner, Lisa J, Cun, Phong, Dillon, Roslyn, Kairs, Stefanie N, Kao, Athit, Preston, Ryan, Schreckengaust, Scott R, Skor, Heather, Smith, William F, You, Jia, Hillis, W.Daniel, Agus, David B, Blume, John E
• Format: Article
• Language: English
• Subjects: Adult; Aged; Area Under Curve; Biomarkers, Tumor - blood; Classification; Colorectal cancer; Colorectal Neoplasms - blood; Colorectal Neoplasms - diagnosis; Early Detection of Cancer - methods; Female; Gastroenterology and Hepatology; Hematology, Oncology and Palliative Medicine; Humans; Machine learning; Male; Mass Spectrometry - methods; Mass spectrometry, Multiple reaction monitoring; Middle Aged; ROC Curve; Sensitivity and Specificity
• ISSN: 1533-0028; 1938-0674
• DOI: 10.1016/j.clcc.2016.02.004

Abstract Introduction/Background Colorectal cancer testing programs reduce mortality; however, approximately 40% of the recommended population who should participate does not. Early colon cancer detection in patient populations ineligible for testing, such as the elderly or those with significant co-morbidities, may have clinical benefit. Despite many attempts to identify individual protein markers of this disease there has been little progress. Targeted mass spectrometry, employing multiple-reaction-monitoring (MRM) technology, enables the simultaneous assessment of groups of candidates for improved detection performance. Materials and Methods A multiplex assay was developed for 187 candidate marker proteins, utilizing 337 peptides monitored through 674 simultaneously measured MRM transitions in a 30 minute liquid chromatography-mass spectrometry analysis of immunodepleted blood plasma. To evaluate combined candidate marker performance, this study utilized 274 individual patient blood plasma samples, 137 with biopsy confirmed colorectal cancer and 137 age- and gender-matched controls. Utilizing two well-matched platforms running 5 days a week, all 274 samples were analyzed in 52 days. Results Using half of the data as a discovery set (69 disease and 69 control), Elastic Net feature selection and Random Forest classifier assembly were used in cross-validation to identify a 15-transition classifier. The mean training receiver-operating-characteristic area-under-the-curve was 0.82. After final classifier assembly using the entire discovery set, the 136-sample (68 disease and 68 control) validation set was evaluated. The validation area-under-the-curve was 0.91; At the point of maximum accuracy (84%), the sensitivity was 87% and specificity was 81%. Conclusion These results demonstrate the ability of simultaneous assessment of candidate marker proteins via high-multiplex, targeted-mass spectrometry to identify a subset group of colorectal cancer markers with significant and meaningful performance.