The small molecules (‘metabolites’) that appear in a person’s urine provide very useful information about that person’s current health. MTI has analyzed the urine metabolic profiles from many individuals to find patterns that distinguish various disease states from controls. For many diseases, these patterns correspond to relatively small numbers of metabolites (5-15), which leads to an effective easy-to-administer, automated diagnostic system.

MTI’s research and development program has successfully identified the existence of two distinct metabolomic fingerprints in urine samples of patients with either adenomatous polyps, the precursor to colorectal cancer, or CRC itself. These distinct signatures were discovered using 1,200 samples collected in a clinical trial attached to the Edmonton colon cancer screening program (SCOPE) in conjunction with Alberta Health Services (formerly, Capital Health). This clinical trial compared the sensitivity and specificity of the urine metabolomics tests (PolypDx™ and ColoDx™) and fecal-based tests (fecal-guaiac, and fecal-immune) relative to the gold standard, full colonoscopy by expert gastroenterologists. The major challenge going forward is to turn this initial discovery into an effective diagnostic test, which can be used to screen the general population for adenomatous polyps and/or CRC.

Technology Platform

Many researchers have successfully identified metabolomic-based signatures that may potentially be used as diagnostics but the challenge is around cost. Often these metabolomic-based signatures are comprised of a large number (60-90) of metabolites that makes it cost-prohibitive to incorporate into a large population-based screening tool. The innovation that MTI developed is the ability to reduce this signature (using a combination of machine learning and statistical analysis) to small number of metabolites yet maintain high accuracy to be useful as a diagnostic screening tool.

Metabolomics is a new science and research teams are actively engaged in developing new methods to develop highly sensitive low metabolite number panels. The advantage that MTI had early on is its clinical connection and its access, via a clinical trial, to a large clinical sample repository that is well-characterized with both colonoscopies conducted as well as fecal guaiac and fecal immune-based tests. Most research groups working in metabolomics have limited access to large clinical samples for statistical analysis and machine learning to be feasible.

MTI also has also formed several important collaborations to further enhance on their current technology. MTI is currently working with AICML (Alberta Innovates Centre for Machine Learning) to aid in the development of the diagnostic algorithm. AICML is providing MTI with an expertise in machine learning techniques that are designed to handle large and complex data sets. MTI has also forged a new and exciting relationship with BGI-Shenzhen. BGI is a world leader in genetic sequencing with a strong background in the ‘omics’ sciences. BGI and MTI have teamed up to complete a joint project in which the metabolomic profile of colon cancer and adenomatous polyps in the Chinese population will be compared to MTI’s existing metabolomic profiles. BGI also has a large expertise in Mass spectrometry and will be an important part of MTI’s scientific evolution.