Online Molecular Diagnostics & Personalized Medicine
Skip Navigation LinksHOME : DIAGNOSTIC TESTS : ColonPRS  


> Background Information Perform Analysis Access Results
ColonPRS  predicts recurrence and overall survival in patients with Stage 2 or 3 colon cancer.

The genes used in this classifier were selected by performing Cox regression analysis on gene expression and clinical data from a large series of stage 1-4 colon cancer patients.

Independently validated results

A predictive algorithm was trained using information from the 163 genes found to be related to patient outcome. This algorithm was then applied to genomic profile data from independent validation cohort of stage 2 and 3 colon cancer.  The highly significant results reveal that ColonPRS is a powerful and accurate predictor of outcome.

Multivariate analysis indicates that the classifier is able to identify patients at high or low risk of relapse, independent to age, gender, stage and grade.

The output of the 163-gene algorithm is a unique prognostic index (range: -2.0 to +2.0) that correlates with increased risk of recurrence.

A gene expression heatmap of the 163-gene signature in data from 151 colon cancer patients. Red = high gene expression, Green = low gene expression.

Patients/rows are ordered by increasing prognostic Index, as determined by each patient's expression of the 163 gene set. Recurrence events (within 5 years) are indicated to the right of the heatmap and can be seen to increase in frequency proportional to the prognostic index. the prognostic index.

Related publications:

Van Laar, R.K., An online gene expression assay for determining adjuvant therapy eligibility in patients with stage 2 or 3 colon cancer.

British Journal of Cancer (2010) [Link]

Details and survival risk-group statistics

Kaplan Meier analysis of the predicted risk groups for the 163-gene classifier applied to the training and validation series are shown below. Each series is analyzed by clinical stage and then by the combination of clinical stage and gene expression risk group.

  Training Series: Disease Free Survival Training Series: Disease Specific Survival Independent Validation Series: Disease Free Survival
Risk stratification by clinical staging alone:
Stage 2:
69 Stage 3: 75
Log rank test: P=0.11
Stage 2: 66 Stage 3: 69
Log rank test: P=0.086
Stage 2: 33 Stage 3: 27
Log rank test: P=0.0026
Risk stratification by clinical staging & ChipDX analysis:
Stage 2: Low risk: 32, High Risk: 37
Stage 3: Low risk 35, High Risk: 40

Log rank test: P=0.0005

Stage 2: Low risk: 31, High Risk: 35
Stage 3: Low risk 31, High Risk: 38

Log rank test: P=0.0009

Stage 2: Low risk: 19, High Risk: 14
Stage 3: Low risk 16, High Risk: 11
Log rank test: P=0.0008

Multivariate analysis of risk group predictions revealed that individuals with a high-risk colon tumor (according to the 163-gene prognostic index) is at ~3-4 times greater risk of disease recurrence and disease-specific death during the 5 year period following initial treatment.