Multidimensional Analysis Provides More Accurate Prognostic Biomarkers in Colorectal Cancer


Key Points

  • Using a multidimensional approach that integrated data from microRNA, gene, and protein analysis on colorectal cancer tumors, researchers discovered that the biomarkers had a predictive accuracy of nearly 90%.
  • Combining information from this multidimensional approach provides better clinical management of colorectal cancer and can help determine the cause of poor outcome.

Using a multidimensional platform that combined nanofluidic technology for quantitative polymerase chain reaction (PCR) analysis and quantitative fluorescent immunohistochemistry for protein analysis, researchers at the Danbury Hospital Biomedical Research Institute were able to accurately predict tumor behavior in patients with colorectal cancer. Their findings have the potential to determine more effective treatments based on the unique molecular characteristics of patients’ tumors. The study by Mariani et al is published in PLOS ONE.

Study Methodology

The researchers analyzed a cohort of 267 patients with colorectal cancer to find possible predictors of outcome. Tumor samples were collected when the patients had their first surgery and before receiving any treatment. The patients then underwent treatment and were followed for outcomes.

Initially, the researchers conducted separate analysis of microRNA, genes, and proteins in the tumor samples. Several markers of outcomes were discovered within each group. The researchers then divided the patients according to those who survived less than 3 years (aggressive disease) and those who survived longer than 3 years (nonaggressive disease).


All factors identified as predictive of outcomes, including microRNA, genes, and proteins were tested to determine whether they were a predictor of aggressiveness. The researchers’ analysis showed that none of the factors identified had a high predictive accuracy, meaning that clinical decisions could not be made based on the information.

They then combined the information from all three of the dimensions analyzed and found biomarkers that had an accuracy approaching 90%.

“[Colorectal cancer] aggressiveness is a complex trait that cannot be predicted with suitable accuracy by the use of an individual, single dimensional factor (microRNA, gene, or protein). In contrast, a multidimensional integrated approach which utilizes data from microRNA, gene, and protein analysis can generate accurate predictors of biological behavior, foster better clinical management of [colorectal cancer], and shine a spotlight on molecules and molecular pathways which are associated with and potentially the cause of poor outcome,” concluded the researchers. 

Cristiano Ferlini, MD, of Danbury Hospital Research Institute, is the corresponding author for the PLOS ONE article.

Funding for this study was supported by the Ruth C. Donovan Cancer Research Program and Mr. and Mrs. Ruggles. The study authors reported no conflicts of interest.

The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.