A new approach to personalized cancer medicine: listening to the internal messages inside the tumor cells

In this study, Big Data was used - a database that includes information on treatments for ten types of cancer, a system developed by Prof. Sol Afroni at Bar Ilan University

A successful fight against cancer depends to a large extent on choosing the right treatment, but the doctors' toolbox currently does not have a sufficient selection of reliable methods for adapting optimal treatment to each patient. Scientists from the Weizmann Institute of Science and the Broad Institute in Cambridge, Massachusetts, have developed a new method for tailoring treatment to patients based on messages transmitted within their tumor cells. This method, the findings of which were recently published in the scientific journal Nature Communications, may help not only in choosing a treatment but also in identifying molecular targets for the development of future drugs.

The most common way today to adjust treatment for cancer patients is by looking for certain mutations in the tumor cells. However, identifying the mutations in the tumor does not guarantee that the appropriate treatment will actually work in a particular patient, and in any case, many treatments do not target mutations at all. In addition, despite many attempts to predict the effectiveness of cancer treatment based on the level of expression of certain genes in the tumor, and not based on its mutations, there is very little success in choosing personalized treatments using this method. In the new study, the research groups of Dr. Ravid Straussman from the Department of Molecular Biology of the Cell at the Weizmann Institute of Science and Dr. Gadi Getz from the Broad Institute of MIT and Harvard came together with the goal of developing a more effective method for adjusting treatment for cancer patients. The new method they developed is not based on searching for mutations or on the level of expression of individual genes, but on listening to messages inside the tumor cells: biochemical signal pathways that affect crucial issues - from changes in the cell's metabolism to the decision whether to divide or die. Many genes are expressed in each such pathway, so sophisticated approaches are required to reveal their activity.

The researchers were able to predict the sensitivity of different tumors to more than 30 anti-cancer drugs currently in use

In recent years, huge databases containing a lot of information about cancer cells and their sensitivity to various treatments have been created in the world. The scientists, led by post-doctoral researcher Dr. Rotem Ben-Hamo, turned to these databases and analyzed data on signaling pathways in approximately 460 types of cancerous tissue from ten types of cancer. Using advanced bioinformatics software called PathOlogist - developed by Prof. Sol Efroni from Bar-Ilan University - the scientists attached to each pathway in each tissue a "score" that reflects the activity level of that pathway. The scores were based not only on the level of expression of the genes in the pathway but also on the structure of the pathway, the interrelationships between the various genes active in it and whether a particular gene blocks or enhances the message of the entire pathway. The scientists then compared the scores they received with databases containing information on the sensitivity of different cancer cells to about 500 drugs.

The researchers discovered that the level of activity of some of the pathways allowed them to predict the sensitivity of different tumors to more than 30 anti-cancer drugs currently in use. In other words, they created a "profile" for cancerous tissues that may direct doctors to optimal drugs to eliminate the cancer. For example, when certain lung cancer cells contained a high level of activity in the pathway leading to programmed cell death called apoptosis, there was a high chance that drugs of the type "microtubule inhibitors" would kill these cells.

The scientists later showed that they could use the findings not only to predict which drug would be effective for which patient, but also to change the response of cancer cells to the drug. They took a sample of lung cancer tissue which, according to their analysis, lacked the activity of the pathway leading to apoptosis. In a laboratory experiment, cells of this tissue were indeed resistant to the "microtubule inhibitor" drug, but when a substance that increases the apoptosis pathway was added to the culture, the drug effectively killed the cancer cells.

The scientists also compared the level of activity of signal pathways in the cells with another type of data: which genes play key roles in different tumors, so that silencing them would lead to the elimination of the tumor. They found that here, too, the activity level of the pathways allowed them to identify "sensitive" genes in different crops. For example, they discovered that breast tumors with a certain activity in the pathway containing the BRCA gene - activity that indicated a mutation in this gene - are highly dependent on a gene called MAD2L1. According to this finding, silencing the gene will lead to the elimination of the tumor cells in patients with the BRCA mutation. These findings can serve as a starting point in the search for existing or new drugs for these patients.

Dr. Adi Jacob Berger, Dr. Nancy Gebert and Dr. Yara Tsong from the Department of Molecular Cell Biology of the Weizmann Institute of Science participated in the study; Dr. Mandy Miller from Dr. Getz's lab at the Broad Institute; Dr. Guy Pines from Kaplan Medical Center; Dr. Roni Oren from the Department of Veterinary Resources of the Weizmann Institute of Science; Dr. Eli Pikarski and Dr. Tzachi Neuman from the Hebrew University of Jerusalem; and Dr. Cyril Bence from Massachusetts General Hospital (MGH).

for the scientific article