- Analysis of tumor samples from 1,150 Latin American patients suggests correlation between ancestry and specific mutations in lung cancer.
- Findings may shed light on the evolution of the disease and spur further research for disease prevention and early detection.
Even as science has removed all doubt about the link between environmental factors like tobacco smoke and lung cancer, the role of genetics in people’s risk of the disease has been much harder to pin down.
A study by Dana-Farber investigators provides new evidence that, in people with lung cancer, genetic ancestry can influence the molecular signature of their cancer — the mix of mutated and normal genes in their tumor cells. By analyzing lung cancer tissue from more than 1,100 patients in Mexico and Colombia, they found that those with a higher degree of Native American ancestry were especially likely to have mutations in the cancer gene EGFR but not in the cancer-related genes KRAS and STK11.
The challenge now is to discover which inherited genomic features make it likely that people who develop lung cancer will have a specific pattern of mutated and non-mutated genes in their tumor cells. The information will help researchers trace the pathways by which the disease develops, and potentially derail it with new therapies.
“Lung cancer kills more people in Latin America than any other malignancy,” says the study’s senior author Matthew Meyerson, MD, PhD, of Dana-Farber and the Broad Institute of MIT and Harvard. “The most common form of the disease on the continent is lung adenocarcinoma, a type of non-small cell lung cancer. We know that the percentage of patients whose tumors harbor EGFR mutations varies widely from country to country — from roughly 14% in Argentina, to 25-34% in Colombia, Brazil, and Mexico, to 51% in Peru.
“It hasn’t been clear, however, whether this variation stems mostly from differences in people’s ancestry or from environmental factors in particular regions of these countries. We were interested in which of these has the larger influence.”
The link between ancestry and EGFR mutation
The study stems from a landmark 2004 paper in which Dana-Farber scientists showed that lung adenocarcinomas with EGFR mutations were susceptible to drugs that target these abnormalities. One of the surprises of that research was a less-heralded discovery that lung cancers in East Asian patients were far more likely to have EGFR mutations than those in people of European or African descent, regardless of whether they were smokers.
The finding raised the question of why lung cancer is molecularly different in different ethnic groups. Latin America, with its mix of people with Native American and European roots, represented an ideal region to look for answers. In Argentina, for example, most people are of European descent, while in Peru the population is largely of Native American ancestry. Colombia, Mexico, and Brazil have a more even representation of the two ancestries.
For the new study, researchers led by Oscar Arrieta, MD, in Mexico City and by Andres Cardona, MD, in Bogotá, Colombia, provided their colleagues in Boston with lung tumor samples from 1,153 patients, almost half of whom were non-smokers. The team at Dana-Farber and the Broad Institute analyzed them for abnormalities in all 547 genes known to be cancer “drivers” — that is, genes whose imperfections help spur cancer growth. They found that in both the Mexican and Colombian patients’ tumors, EGFR mutations were more common than KRAS mutations.
The researchers also conducted an “off-target” analysis of the tumors, scanning regions of the genome outside those 547 genes. Their quarry: sites on the genome where a single letter of DNA varies from one patient to the next. Such pinpoint changes, known as single nucleotide polymorphisms, or SNPs, can help indicate the mixture of African, European, or Native American ancestry within an individual.
Ideally, one would look for SNPs in an individual’s normal tissue to determine ethnic ancestry. In the absence of such tissue, Jian Carrot-Zhang, PhD – the study’s first author and a postdoctoral fellow in Meyerson’s lab who now leads her own lab at Memorial Sloan Kettering Cancer Center – and Dana-Farber’s Alexander Gusev, PhD, developed a technique for inferring ancestry from tumor tissue alone.
Combining the results of their on- and off-target analyses, the investigators found that EGFR mutations occurred more frequently, and KRAS and STK11 mutations less frequently, in lung tumors from patients with higher levels of Native American ancestry compared to European or African ancestry. This essentially mirrors the pattern of mutations in lung cancers from patients with East Asian backgrounds.
Examining inherited versus environmental risk
Researchers next examined whether this pattern is due to an inherited region of the genome associated with Native American ancestry or to environmental or socioeconomic influences on people of that ancestry. A unique aspect of this population is the abundance of admixed individuals with different “local” ancestry across their genome, a phenomenon that could be used to disentangle the effects of genetics from environment.
Carrot-Zhang and Gusev developed a “local ancestry risk score,” a measure of the likelihood that lung cancer in people of specific ancestry, within a specific region of the genome, will carry EGFR mutations. They found that, for people with Native American ancestry, the score was higher than the worldwide risk score for such mutations. The implication, Meyerson says, is that “we’re likely to be seeing an inherited risk factor rather than an environmental one.”
Researchers hope to conduct follow-up studies using a larger number of tissue samples. “If we can identify the inherited factors that predispose people to develop EGFR-mutant or KRAS-mutant lung adenocarcinomas, we’ll have a better understanding of the evolution of the disease,” Meyerson remarks. “That, in turn, can suggest strategies for preventing the disease or detecting it early in Latin American countries and beyond, particularly for non-smokers.”
Check out other posts from Insight: