Dr. Haibe-Kains and Dr. Zhaleh Safikhani
Research led by Princess Margaret Cancer Centre Scientist Dr. Benjamin Haibe-Kains and Postdoctoral Fellow Dr. Zhaleh Safikhani revealed that gene isoforms can serve as a useful resource for predicting the effectiveness of anti-cancer therapies. (Photo: UHN)

You may have heard the saying, 'genes make you who you are'. For cancer, this is generally thought to be true: most research has focused on finding out which genes are turned on or off in cancer cells.

Now, recent findings from researchers at Princess Margaret Cancer Centre have shown that the different versions of genes, known as gene isoforms, turned on in tumors may constitute a rich source of information that could be used to tailor therapy to individual patients.

The research, led by Dr. Benjamin Haibe-Kains, Scientist at the Princess Margaret, and Dr. Zhaleh Safikhani, Postdoctoral Fellow, revealed gene isoforms can be used to predict whether anti-cancer drugs will be effective.

Until recently, specific gene isoforms couldn't be easily measured in tumors; only an average of all the levels of isoforms could be measured accurately. Advances in sequencing technologies enabled the quantification of all gene isoforms in a robust and cost-effective way.

"We can now study each of the 200,000 known gene isoforms in cancer cells. This high resolution is a game changer in cancer research," says Dr. Zhaleh Safikhani, first author of the study and postdoctoral fellow in Dr. Haibe-Kains' lab.

Previous investigations into the predictive value of gene isoforms for cancer were limited to a few genes, leaving this wealth of information largely unexplored.

To address this gap in knowledge, the team used publicly available genomic data from more than 1,400 cancer models. To dig deeper beyond simply examining which genes are turned on, they applied advanced computational methods to create profiles of all of the genes and their various isoforms in these models.

They determined which isoforms could be used to predict response to 148 anti-cancer drugs. The researchers found gene isoforms can better predict the response to anti-cancer drugs than the profile of genes that are turned on.

They then selected their four most promising isoform-based tests and found the tests could accurately predict whether cancer cells responded to the drugs lapatinib, erlotinib, AZD6244 and paclitaxel in subsequent experiments.

"Our study is the first large-scale initiative to demonstrate that gene isoforms are a rich resource for developing predictive tools," explains Dr. Haibe-Kains. "This work reveals a new area of focus—one that could be used to develop more robust methods of selecting effective anti-cancer therapies for individual patients."

This work was supported by the Terry Fox Research Institute, Stand Up To Cancer Canada, the Cancer Research Society, the Canadian Cancer Society Research Institute, the Canadian Institutes of Health Research, the Natural Sciences and Engineering Research Council of Canada and The Princess Margaret Cancer Foundation.​

data computer
To analyze the large data sets in this study, the researchers applied machine learning methods, in which computers use mathematical algorithms to parse information. (Photo: iStock)

Share This Story

Share Tweet Email