NEW YORK (Precision Oncology News) – As cigarette smoking — a leading risk factor for lung cancer — is on the decline, new studies are exploring genetic and other environmental risk factors that affect whether or not someone might develop the disease.
According to the US Centers for Disease Control and Prevention, about 14 percent of US adults were smokers in 2017, down from 20.9 percent in 2005. At the same time, lung cancer among people who have never smoked is becoming proportionately higher, and never smokers now account for between 10 percent and 25 percent of lung cancer patients.
Many previous studies of lung cancer, including by The Cancer Genome Atlas project, have relied on lung cancer samples that largely came from people who had been smokers. Eighty-one percent of patients whose tumors were analyzed in the TGCA lung adenocarcinoma project were former or current smokers. Still, these studies hinted that the mutational profiles of lung tumors from smokers and non-smokers differ.
Some new studies aim to examine the mutational landscape of lung cancer among non-smokers to not only tease out what genetic alterations arise in their disease, but also to pinpoint environmental or other factors that contribute to disease risk. The findings may then be applied to identify people who are at increased risk of developing lung cancer or suggest treatment approaches for those who have already developed the disease.
"We know a few important established risk factors like radon or asbestos or passive smoking or some previous lung diseases, but we don't know what causes the large majority of these cases," Maria Teresa Landi, a senior investigator at the National Cancer Institute, said in an interview. "And there is no effective treatment for them. So, we really wanted to study them for this reason."
She and her colleagues launched Sherlock-lung, a genomic epidemiologic study of lung cancer in people who have never smoked, to try to uncover the etiology of the disease among nonsmokers. They plan to collect 2,000 to 2,500 lung cancer samples from people living in a range of geographic areas and from people of a range of ethnicities. That way, she said, the study will encompass a number of exposures and genetic backgrounds and be applicable to a wide population.
So far, Landi said they've conducted whole-genome sequencing of paired tumor and blood samples from 235 lung cancer cases and are currently sequencing another 600 cases.
While that number is small, they already can pick up differences in the mutational signature patterns seen in their nonsmokers, as compared to what has been detected in smokers.
Similarly, Fred Hutchinson Cancer Research Center's Alice Berger has teamed with researchers from the Women's Health Initiative, a long-term national study that started in the 1990s, to also study gene mutations in lung cancers from women who never smoked. (Men and women, she noted, also appear to harbor differences in their mutational profiles.)
Berger and her colleagues plan to sequence lung tumors from about 130 women, about 110 of which will be from never smokers and the rest from women who did smoke. So far, they've collected DNA from 99 lung tumors and sequenced about a third of them and have plans to finish sequencing by later this year. Berger noted they are in particular ensuring they can detect frequent translocations, as there are various gene fusions that appear to be more common among non-smokers.
Both studies also aim to delve into environmental factors that could influence lung cancer risk, such as asbestos, air pollution, or radon exposure. Smoking, Berger noted, leaves a particular mutational mark on the tumor genome, and so too could these other types of exposures.
She and her colleagues also plan to examine secondhand smoke exposure, as do NCI's Landi and her colleagues.
"We have information about if they live with people who do [smoke], [or] work in workplaces where they're exposed to smoke," Berger said. "If you work in a casino or something like that, you might have a high rate of secondhand smoke."
Through Sherlock-lung, Landi and her colleagues want to explore the links between those known environmental exposures among individuals in their study — such as radon or indoor air pollution — or endogenous mutational processes — such as DNA repair or DNA replication — to the mutational signatures they see in patients' tumors.
Once they do that, they plan to see whether those mutational signatures present in people with lung cancer who were exposed to a high level of a known environmental risk factor can be teased out among individuals with lung cancer whose environmental exposures are low or even unknown. They plan on then validating these signatures through in silico simulations.
Researchers at Johns Hopkins University and their colleagues, meanwhile, have already begun to dig into how environmental exposures influence lung cancer risk among never smokers. At the annual American Association for Cancer Research meeting in April, Hopkins' Batel Blechter presented data on lung cancer risk among non-smoking women in China. She and her colleagues were particularly interested in the influence of coal as a cooking fuel — a known lung cancer risk factor — in conjunction with genetic factors on the risk of developing lung cancer.
Using data from the Female Lung Cancer Consortium in Asia (FLCCA) study, the Hopkins team examined the women's risk of developing lung cancer based on their genetic profiles by building a polygenic risk score based on 10 SNPs from a genome-wide association study conducted among women in Asia. Women with high risk scores were, they found, more likely to develop lung cancer.
At the same time, they examined the rate of lung cancer among women who had or who never had used coal in the home to find that those who had used coal in their homes had an increased risk of lung cancer.
"We often find that gene and environment both have an effect," Hopkins' Nilanjan Chatterjee, a biostatistician and Blechter's advisor, said in an interview. "If you look at the effect of the gene, it is similar whether you have that [environmental] parameter or not."
But in this case, the Hopkins-led team found the environmental factor — here, coal use — was a stronger risk factor among women with low genetic risk for lung cancer.
While he cautioned he did not have a strong theory for why this occurred, he noted that some researchers have speculated that, in some cases, the influence of genetic factors could already make risk so high that it cannot go much higher even with the addition of some environmental factor.
"That is interesting, but we do not know what it means biologically," Chatterjee said of their finding.
Through studies like these, researchers hope to combine both genetic and environmental risk factors to gauge people's risk of developing lung cancer as well as uncover means of preventing or treating disease.
Chatterjee and his colleagues argued in a 2016 JAMA Oncology paper that such a model could help guide cancer screening efforts and the understanding of individual risk.
In that paper — which used breast cancer as an example — he and his colleagues built a model that combined the influence of about 100 genetic risk variants and epidemiological factors like BMI, smoking status, whether someone was taking hormone replacement therapy, and more to determine breast cancer risk. The addition of these environmental factors to the polygenic risk score model, Chatterjee noted, improved its performance predicting risk in the general population.
This, he added, can then help guide screening efforts. He noted that there has been back-and-forth in the US about when to, in the case of breast cancer, begin mammography screening. Current guidelines rely heavily on age, but he said that these sorts of risk models could identify people who, though they are young, may be at higher risk of disease and who might benefit from earlier screening. At the same time, it could be used to identify people who are older, but are at lower risk of disease, and for whom screening could possibly start later in life.
"You can see that if you use this type of risk model, you will probably recommend screening it in a different way for many people," Chatterjee said, noting that a similar approach could work in lung cancer with CT scans.
By identifying which genes are mutated within those who end up developing cancer, researchers might then be able to stratify patients into different groups. NCI's Landi said that the additional information about environmental exposures alongside other information on methylation status, histological features, radiological imaging, and more could be combined with deep learning to identify lung cancer subtypes among never smokers.
"We hope to be able to identify subgroups of lung cancer cases that are associated with specific exposures or endogenous processes," Landi said of her Sherlock-lung study. "And then we could identify potential strategies to prevent them or treat them."