Colorectal Cancer

Cologic® Clinical Studies

Our path to discovery of GTA-446 and the clinical story of developing our Cologic® test:

Initial metabolomics discovery  

In our pilot discovery study, we used non-targeted metabolomics technology based on high resolution mass spectrometry to discover a reduction in the circulating levels of a new family of metabolites called gastric tract acids (GTAs), and specifically the isoform GTA-446. The study included serum samples from 40 pre-treatment colorectal cancer patients and 50 disease-free controls from the Genomics Collaborative Global Repository. Using computational methods and tandem mass spectrometry, we determined that GTA-446 was a 28-carbon long-chain polyunsaturated fatty acid. Statistical analysis resulted in a GTA-446 receiver-operator characteristic curve area (ROC-AUC) for colorectal cancer of 0.94.

Repeat verification of metabolomics discovery  

To confirm the discovery described above, we performed another independent non-targeted metabolomics discovery experiment on a separate set of serum samples from 26 pre-treatment colorectal cancer patients and 25 matched controls (in collaboration with Seracare Lifesciences Inc). We confirmed the discovery of the GTA family of metabolites, and again observed the reduction of GTA-446 in colorectal cancer patients. The GTA-446 ROC-AUC was 0.93 for colorectal cancer versus controls.

Tandem MS validation study  

We developed a first-generation tandem mass spectrometry method to semi-quantitatively measure GTA-446 levels, and used it to analyze serum from 70 additional pretreatment colorectal cancer patients and 70 matched controls (Seracare Lifesciences). The ROC-AUC for colorectal cancer was 0.88.

Validation 1 in Japanese patients   

In this study, we investigated the performance of GTA-446 in a non-Caucasian population, and asked whether GTA-446 reduction was the result of tumor load by comparing the levels before and after surgery. We tested 46 colorectal cancer patients prior to surgery, 34 of the 46 following surgery, and 35 controls. GTA-446 levels were not significantly different between pre and post-surgery, suggesting the tumor was not causing the reduction. The ROC-AUC for colorectal cancer was 0.93.

Validation 2 in Japanese patients   

We performed this clinical study to confirm the lack of surgery effect observed in the Validation 1 study, and to further validate GTA-446 performance in another Japanese population. The study included serum collected prior to and following surgery from 40 colorectal cancer patients and 40 controls. Similar to the Validation 1 study described above, we observed no difference in GTA-446 levels before or after surgery, confirming that low GTA-446 levels were not due to the presence of the tumor. The ROC-AUC for colorectal cancer was 0.99.

Validation with Bioserve  

The largest risk factor for colorectal cancer is age, with only about 6% occurring under age 50 (Canadian Cancer Statistics, 2012). Therefore, people under age 50, all factors considered, represent a low-risk population. The goal of this study was twofold: first we determined the distribution of GTA-446 in the general population and showed that there was inverse association with age. We used this distribution to determine a GTA-446 cut-off based on the lowest 10^th percentile of subjects aged 40-49, meaning 10% of 40-49 year-olds would be considered to have low levels. Second, we measured GTA-446 in 150 Caucasian colorectal cancer patients following chemo-radiation therapy and compared it to the cut-off. The ROC-AUC for colorectal cancer was 0.98.

Prospective validation of GTA-446 levels in a population undergoing colonoscopy  

This was a prospective clinical trial to investigate the levels of GTA-446 among subjects undergoing colonoscopy. Over a two-year period, we enrolled nearly 5000 subjects undergoing colonoscopy at Regina General and Pasqua hospitals in Regina, Saskatchewan, Canada. GTA-446 levels were considered low if they were below the 10^th percentile of subjects aged 40-49, based on a reference sample population collected from the Saskatchewan Disease Control Lab (SDCL). The ROC-AUC for colorectal cancer versus low-risk subjects was 0.89.

Longitudinal studies

Validation with Biomira:  

In this study, we longitudinally investigated GTA-446 levels in 12 stage IV colorectal cancer patients as they underwent a novel vaccination therapy. We collected an average of 8 samples per patient for up to 61 weeks, of which all were below the GTA-446 cut-off described above (100% sensitivity). GTA-446 levels appeared stable among all patients for several months, and were not affected by any combination of treatment.

Longitudinal investigation in disease-free healthy subjects:  

The purpose of this study was to investigate the intra-individual stability of GTA-446 in healthy subjects. We collected monthly longitudinal blood samples from seven healthy volunteers for more than a year. GTA-446 levels for all subjects and all timepoints were in the normal range. Although there was inter-subject variability (each of the seven subjects showed different levels), the intra-subject levels (samples from the same person) over several months were very stable.

 

The Cologic® test is intended for use in risk assessment and monitoring; it is not a standalone diagnostic test, and is not a screening test for colon cancer.