Big Data Shows Cancer Cell Lines Reflect Patient Tumor Mutations

    New research shows that patient-derived cancer cell lines harbor most of the same genetic changes found in patients’ tumors and could be used to learn how tumors are likely to respond to new drugs. The findings (“A Landscape of Pharmacogenomic Interactions in Cancer”), published in Cell, should help increase the success rate for developing new, more personalized cancer treatments.

    Led by scientists from the Wellcome Trust Sanger Institute, the European Bioinformatics Institute (EMBL-EBI), and the Netherlands Cancer Institute, the international study discovered a strong link between many mutations in patient cancer samples and the sensitivity to particular drugs. This could advance personalized cancer medicine by leading to results that help doctors predict the best available drugs or the most suitable clinical trials for each individual patient.

    This was the first systematic, large-scale study to combine molecular data from patients, laboratory cancer cell lines, and drug sensitivity. The researchers looked at genetic mutations known to cause cancer in more than 11,000 patient samples of 29 different tumor types, built a catalog of the genetic changes that cause cancer in patients, and mapped these alterations onto 1000 cancer cell lines. Next, they tested the cell lines for sensitivity to 265 different cancer drugs to understand which of these changes effect sensitivity.

    The team made two significant discoveries. First is that the majority of molecular abnormalities found in patient’s cancers are also found in cancer cells in the laboratory. This means that cell lines are indeed useful models to identify which drugs would work best for different patients. Second, many of the molecular abnormalities detected in the thousands of patient cancer samples can, both individually and in combination, have a strong effect on whether a particular drug affects a cancer cell’s survival.

    The results suggest cancer cell lines could be better exploited to learn which drugs offer the most effective treatment to which patients.

    According to Francesco Iorio, Ph.D., postdoctoral researcher at EMBL-EBI and the Sanger Institute, “If a cell line has the same genetic features as a patient’s tumor, and that cell line responded to a specific drug, we can focus new research on this finding. This could ultimately help assign cancer patients into more precise groups based on how likely they are to respond to therapy. This resource can really help cancer research. Most importantly, it can be used to create tools for doctors to select a clinical trial which is most promising for their cancer patient. That is still a way off, but we are heading in the right direction.”

    “In this study we compared the genetic landscape of patient tumors with that of cancer cells grown in the lab,” noted Matthew Garnett of the Wellcome Trust Sanger Institute. “We found that cell lines do carry the same genetic alterations that drive cancer in patients. This means that drug sensitivity testing in cell lines can be used to figure out how a tumor is likely to respond to a drug.”

    Previous studies have sequenced the DNA of cancers from patients to identify the molecular abnormalities that drive the biology of cancer cells. Researchers have also shown that large collections of cancer cell lines grown in the laboratory can be used for measuring sensitivity to many hundreds of drugs. However, this is the first study to systematically combine these two sets of information.

“We need better ways to figure out which groups of patients are more likely to respond to a new drug before we run complex and expensive clinical trials,” said Ultan McDermott, Ph.D., of the Sanger Institute. “Our research shows that cancer cell lines do capture the molecular alterations found in tumors, and so can be predictive of how a tumor will respond to a drug. This means the cell lines could tell us much more about how a tumor is likely to respond to a new drug before we try to test it in patients. We hope this information will ultimately help in the design of clinical trials that target those patients with the greatest likelihood of benefiting from treatment.”