Cancer cell, illustration
Cancer cell. Computer illustration of cancer cells, showing the blood vessel formation providing the the cells with oxygens and nutrigens. The cells with their nuclei are shown in blue.

Not all gene variants in cancer are created equal. With that in mind, researchers at the Yale School of Public Health (YSPH) recently completed a mathematical analysis of all recurrent single nucleotide variants (SNVs) in 22 major types of cancer, in order to quantify the relative importance of each and which are most likely to cause the spread of the disease.

The research, conducted in the lab of Jeffrey Townsend, Ph.D., the Elihu Professor of Biostatistics and of Ecology and Evolutionary Biology, will allow for the analysis of SNVs present in cancer to determine the relative size of the effect each SNV has on the progression of the disease. The team anticipates their findings—published in the Journal of the National Cancer Institute—will be applicable for developing clinical trials, in helping to inform clinical development of new cancer therapies based on their genetic targets, and ultimately in aiding precision medicine tumor boards tailor care to individual patients.

“For the past 10 years we've been able to calculate from tumor sequencing which mutated genes are winners and losers—which mutated genes help the cancer survive and reproduce, and which do nothing,” Townsend said in a YSPH blog post . “But we haven't been able to compute their cancer effect size—how important one mutation is compared to another. Now we can.”

Tumor sequencing has typically focused on how frequently mutations are seen in cancers and as a statistical measure to show whether a specific gene is overburdened with mutations, but neither measures the relative importance of specific mutations in tumorigenesis and the spread of the disease. Researchers in Townsend’s lab sought to better quantify the cancer effective size of these mutations.

To do this, the investigators categorized the frequency that a mutation is observed in tumors into two contributing factors: the baseline mutation rate, and the degree of selection for the mutation in the cancer lineage. Both mutation and selection contribute to the frequency of variants among cells. Using diverse genome-scale data to calculate the mutation rate to divide out the contribution of mutation from the frequency that mutations were observed in tumors, the researchers were able to calculate the cancer effect size.

Townsend noted that his background in evolutionary biology allowed him to look at this project with fresh eyes. “Whereas in the cancer world the focus has always been on mutation rates, the focus in evolutionary biology has been on the process of natural selection on those mutations,” he said. “The quantification of cancer effect sizes is a great example of how interdisciplinary research is not only helpful, but essential to scientific progress.”

The result of this work is a new ability to effectively identify which mutations contribute significantly to the growth and reproduction of cancer cells and which mutations that may be associated with cancer, don’t effectively drive the growth and spread of the disease. This has significant implications better informing clinical trials and individual patient care.

“By looking at cancer through the lens of evolution, we can harness the wealth of molecular data made available through tumor DNA sequencing to both better understand what is driving cancer and to also expand and refine evolutionary theory,” noted Vincent Cannataro, Ph.D., a postdoctoral associate and first author of the paper.

Also of Interest