144
THE ROLE OF NMR URINE METABOLOMICS AS A NEW METHOD FOR SCREENING COLORECTAL CANCER - A PRELIMINARY ANALYSIS
H Wang2, C Slupsky1, D Schiller2, M Sawyer3, C Wong1, R Fedorak1
1Department of Medicine; 2Department of Surgery; 3Department of Oncology, University of Alberta, Edmonton, Alberta
Aims: Colorectal cancer is a major public health concern since it is among the leading causes of death in North America, but it can be curable if identified early, even better if found at the adenomatous polyp stage. Current screening methods have limitations or potential risks associated with them. With the emerging field of metabolomics and the recent discovery of a urine metabolomic signature for colorectal cancer (in a pilot study by Sawyer and Slupsky et al.), we hypothesize that urine metabolomics could represent a more sensitive, non-invasive and accessible method of identifying asymptomatic subjects with colorectal cancers, and perhaps even precursor adenomatous polyps.
Methods: This is the preliminary analysis from a 1000 subject prospective, multi-centered trial to assess whether a simple urine test can play a role in the screening of colorectal cancer. Urine samples were collected from 52 subjects going through an established colorectal screening program (29 had no polyps, 10 adenomatous polyps, 12 hyperplastic polyps, and 1 malignant polyp). Using nuclear magnetic resonance (NMR) spectroscopy, the 1H NMR spectrum of each urine sample was analyzed to define a unique metabolomic signature, and it was compared to the recently established metabolomic signature of colorectal cancer.
Results: Using the metabolomic signature, the cancer subject was correctly predicted as having cancer. Similarly, of the 10 adenomatous polyp subjects, and 12 hyperplastic polyp subjects, 9/10 and 8/12, respectively, were correctly predicted by the metabolomic signature. Of those patients with no polyps, only 13/29 were identified as "normal". Comparison of those with adenomatous and hyperplastic polyps revealed no clear distinction using multivariate statistical analysis.
Conclusions: These results are encouraging; as it appears that urine metabolomics has a high sensitivity for diagnosis of colorectal cancer and identifying colonic polyps. Additional patient analyses will likely increase specificity and decrease false-positives. This preliminary analysis generated interesting and encouraging results which are being studied in a larger population. Urine metabolomics has the potential to provide society with an accurate, non-invasive, and inexpensive screening tool for colorectal cancer.