A rare inherited mutation may sharply raise lung cancer risk, including among people who have never smoked. A study published in Science found that EGFR T790M was associated with 25 times the risk of lung cancer compared with the risk among people without the mutation.

The difference was even larger within the group that had never smoked. Carriers who never smoked had about 62 times the odds of lung cancer compared with people who never smoked and did not carry the variant.

Investigators from Dana-Farber Cancer Institute and 23andMe Research Institute led the study, analyzing information from more than 3.3 million people. The gene variant showed no increased risk across 17 other cancers included in the research.

The findings could challenge a screening system centered largely on tobacco exposure. “Today, lung cancer screening is driven almost entirely by smoking history,” study author Jaclyn LoPiccolo, an attending physician and lung cancer researcher at Dana-Farber Cancer Institute, said in a press release.

LoPiccolo said the findings raise the possibility that inherited genetic risk could eventually help determine who receives screening. If further research confirms a benefit, genetic testing could identify people with EGFR T790M and personalized CT screening could seek tumors when they are at their most curable stage.

Researchers also uncovered a geographic and genealogical pattern among carriers in the United States. Most were traced to shared ancestry connected to British and Irish settlers who arrived in Southern Appalachia about 200 to 225 years ago.

Despite its potent association with lung cancer, the mutation remains rare across the country. Researchers estimated that it affects roughly one in every 15,000 to 16,000 people nationwide.

The mutation appears more frequently in certain parts of Southern Appalachia. In those areas, researchers estimate that as many as one in 2,000 people could carry it.

LoPiccolo said the vast majority of carriers appeared to inherit the mutation from the same ancestral line. Researchers traced that line to British and Irish settlers in the United States and found that the mutation became more concentrated following a founder event and genetic bottleneck in Southern Appalachia about 200 years ago.

The findings illustrate how migration and genealogy can influence disease risk across generations, according to LoPiccolo. The regional concentration also means that the study’s risk estimates may not apply equally to all populations.

Alexander Gusev, a quantitative geneticist at Dana-Farber, called the strength of the effect one of the study’s remarkable findings. “To my knowledge, it's one of the strongest, if not the strongest, cancer risk-increasing mutations that has ever been found,” he said in the release.

Gusev said smoking and the mutation each raise lung cancer risk, while having both produces the sum of those risks. His message for carriers and others remained direct: Smoking should still be avoided.

The researchers suggested that people consider speaking with a genetic counselor if they have a strong family history of lung cancer, multiple lung nodules or tumors, or family roots in parts of the southeastern United States. A counselor could discuss whether genetic testing or lung cancer screening makes sense for them.

Important uncertainties remain because the mutation is exceptionally rare. Even with millions of participants, the researchers identified relatively few carriers, leaving the precise size of the increased lung cancer risk uncertain.

The analysis also relied heavily on people participating in 23andMe research, and those participants may not represent the broader population. The mutation’s uneven geographic distribution creates another limitation when applying the findings to people throughout the country.

Although the study establishes a strong association between EGFR T790M and lung cancer risk, it does not show that genetic testing improves mortality or other health outcomes. The National Institutes of Health and the American Cancer Society provided part of the study’s funding.