Researchers from the Universidad Politécnica de Madrid's Facultad de Informática, the Universidad del País Vasco, the Gaiker Technology Centre (Biscay) and the Hospital de Cruces in Barakaldo, have developed a tentative model for the diagnosis of colon cancer based on a biomarker panel. According to study results, it classifies carcinogenic and non-carcinogenic samples with 94.45% accuracy.
Facultad de Informática researchers Pedro Larrañaga and Rubén Armañanzas, of the Computational Intelligence Group, have participated in this research. The results of this research were published in the BMC Cancer journal.
This is a mathematical model that uses a set of biomarker expression patterns to objectively diagnose colon cancer. To do this, DNA chip or microarray technology is used in combination with bioinformatics analyses of data supplied by the DNA chips.
Two-channel glass unhybridized DNA microarray slide (left). Image with the reading of the gene expression for some of the thousands of genes scanned simultaneously (right). The crosses through some measures indicate hybridization or scanning process errors.
A DNA chip is a collection of DNA fragments attached to a solid surface. DNA chips are used to analyse differential gene expression, simultaneously sifting through thousands of genes.
Using the above technology, the researchers managed to identify a biomarker panel for early colon cancer diagnosis, consisting of seven genes capable of correctly discriminating carcinogenic and non-carcinogenic genes.
Colon cancer is the third most common type of cancer and the second highest cause of death from cancer worldwide. However, the survival rate varies from country to country. Whereas the US survival rate is 62%, only 43% of sufferers survive in Europe.
The reasons behind such inconsistent behaviour are not at all clear. Some researchers suggest that preventive screening programmes (designed for the early diagnosis of important diseases) and care quality can play a key role in the reduction of colon cancer mortality.
Based on these data, the medical scientific community is attaching major importance to screening programmes to reduce the incidence and mortality of this type of cancer through the detection of pre-malign polyps and the diagnosis of cancer in the early stages of the disease.
The research analysed the gene expression patterns in human colon cancer from 31 tumour samples that matched different stages of the disease and another 33 tumour-free samples.
The gene expression profiles derived from microarrays were analysed and interpreted using machine learning and data mining techniques. The researchers used qPCR, a variation on polymerase chain reaction (PCR) used to amplify and simultaneously quantify the product of the amplification of deoxyribonucleic acid (DNA ) in absolute terms, to confirm the results.
The next project goal is to verify the data collected to date on a population of 200 people (50% healthy and 50% suffering from colorectal cancer). If they were confirmed, the test would be scaled up for use on 7000 individuals, after which the technology could be transferred to industry with the aim of assuring that society ultimately benefits from the project results.
Recent advances in genomics and proteomics have contributed to the molecular-level understanding of colon cancer through the evaluation of gene expression and protein profiles in carcinogenic and non-carcinogenic surrounding tissue and body fluids.
In this context, genomic techniques like DNA microarrays are highly effective because they are reliable and fast at analysing genes, offer high data availability and ultimately increase the possibilities of discovering potential biomarkers.
Fast and effective cancer analysis at the early stages of the disease will unquestionably lead to early diagnosis and the development of effective treatments for this and other types of cancer in the future.
Identification of a biomarker panel for colorectal cancer diagnosis, Amaia García-Bilbao,Rubén Armañanzas, Ziortza Ispizua, Begoña Calvo, Ana Alonso-Varona, Iñaki Inza, Pedro Larrañaga, Guillermo López-Vivanco, Blanca Suárez-Merino and Mónica Betanzos. BMC Cancer 2012, 12:43. DOI:10.1186/1471-2407-12-43
Facultad de Informática de la Universidad Politécnica de Madrid