A new cancer study reveals that lab cultures of cancerous cells that are an important part of cancer research and developing cancer drugs are not faithful representations of the cancerous tumors that the drugs are meant to treat.
Not only do these cultures not accurately replicate tumors as they appear in the body, but the study found significant differences between cultures grown in different laboratories, even when their source was the same tumor sample. These findings could explain the gaps between the success of many drugs in the laboratory and their failures in clinical trials, and vice versa. The researchers suggest a solution for reducing these gaps and achieving a genetic profile in lab samples that are more representative of the actual tumors.
The study was published Wednesday on the website of the prestigious journal Nature, and will appear as the journal’s cover story next week. The researchers say this discovery in of enormous importance and will allow cancer researchers around the world to “speak the same language” and focus cancer research in more promising directions.
The study was conducted by a team from the Broad Institute, a biomedical and genomic research center that is jointly run by Harvard University and the Massachusetts Institute of Technology. Its cancer research project focuses on the genetic basis of the disease and on ways to harness genetics to cure it. The study was done in the lab of Prof. Todd Golub, director of the institute’s cancer program, and was led by Dr. Uri Ben-David, an Israeli researcher doing postdoctoral work at the institute.
- Israeli scientists grow 'tiny brains' to study how its folds are created
- Doctor may have failed to give cancer tests to thousands of women
- Israel drops controversial study on links between pollution and cancer in Haifa area
The research focused on cell lines — cancer cells derived from patient tumors that continue to grow and divide under laboratory conditions. These are used to understand the mechanisms of the disease and as platforms for testing new drugs.
“Cell lines are a very basic system, the bread and butter of drug development,” Ben-David told Haaretz. “When you want to do experiments on human cancer cells you need to use a reliable and stable model that holds over time and on a wide scale.”
Over the years, scientists thought cell lines were a faithful representation of the tumors from which they were produced, even as they continued to grow and divide in laboratory cultures. “Most of the scientific community is working on several hundred cell lines that have been produced since the 1970s, and the assumption is that they are stable over time.”
Now it turns out that this is not the case. In the article, the researchers demonstrate that while the cells grow and divide in the lab, they are continuing to change genetically and develop different mutations, which changes their protein expressions and their response to drugs.
The researchers began to suspect that the cell lines were developing differently when they performed a genetic sequencing of 106 cell lines from two different stockpiles of cell lines used for testing cancer drugs. While in theory the lines should have been identical in both stockpiles, the team found numerous differences between them; a fifth of all the mutations found were found in only one of the stockpiles and not in the other.
The researchers then collected dozens of versions of the most popular cell lines — those of breast and lung cancer — from dozens of labs all over the world. They then assessed the properties of the samples’ DNA sequences, and assessed the gene expression patterns of each one. They then exposed the cells to hundreds of different anti-cancer compounds, and precisely analyzed the genetic differences and the differences in how each of the cell lines responded to the compounds.
“We discovered that 75 percent of the genetic mutations that were found in each of the cell lines did not appear in all the versions,” says Ben-David. “It turns out that these cells have lives of their own, and they distance themselves from the original tumor as time passes. We show in our study that they are much less stable than previously thought, and this instability is very significant in terms of their performance and use in research.”
The researchers believe the cells develop and change differently once they are taken out of the body and moved to a lab, and that the environment of a particular lab also makes a difference. “The cells are subject to differing conditions and pressures. In the patient’s body, tumor cells are influenced, inter alia, by the interaction with the immune system and other systems. The more cell lines continue to grow, the more they are distinguishing themselves from the original tumor and are less representative,” Ben-David explains.
The cells are not just responding to being out of the body, but also to differences between the cultures on which they are grown in different labs. The medium in which the cells are cultured, the density of the cell samples and the oxygen concentration in the lab could all have an effect. “Even if these difference seem minor at first, when they take place over a long time they are translated into significant differences that influence the cultures and dynamics between different subsets of cancer cells.”
In a press statement, Golub stresses that the findings don’t mean that cell lines are not a good model for cancer research, but that their limitations must be understood and adjusted for. The solution proposed by the researchers is to always compare the characteristics of the cell line being used to information about its genetic characteristics at an earlier stage in a comprehensive database. The research team even opened a free internet portal, called Cell STRAINER, to help researchers do this.
“Cell lines are still a very good model for cancer research and we don’t need to stop using them,” said Ben-David. “We are arguing that we must become deeply familiar with our models so we can understand their limitations.” He added that the method they developed will help researchers avoid reporting false results, explain the differences in findings between different research teams and conserve resources.