by Victoria Fisher, M.P.H., and Jennifer Loukissas, M.P.P.
Genetic epidemiologists in the Division of Cancer Epidemiology and Genetics have been collaborating with investigators from the University of Milan in Italy to uncover the critical steps that lead to the development of lung cancer and identify opportunities to interrupt that progression. The population-based study in the northern Lombardy region combined elements that had never before been included together in a study of lung cancer: substantial population size; rigorous study design; detailed epidemiologic, clinical, behavioral, and occupational data; and exceptional tissue and biospecimen collection. The rich integrative platform of the Environment and Genetics in Lung Cancer Etiology (EAGLE) study resulted in extensive data on 2,100 lung cancer patients and over 2,100 healthy population-based controls.
By evaluating these data together, EAGLE investigators are able to draw conclusions not only about host and environmental risk factors, but also on the effect of those factors on the tumors themselves and the risk of local recurrence, distant metastasis, and survival.
DCEG researchers Maria Teresa Landi, M.D., Ph.D., and Neil Caporaso, M.D., Chief of the Genetic Epidemiology Branch, lead the EAGLE study team with their close collaborators from the University of Milan: Drs. Pier Alberto Bertazzi, Dario Consonni, and Angela Pesatori. Other DCEG investigators include Dr. Andrew Bergen (formerly with DCEG), Lynn R. Goldin, Ph.D., Alisa M. Goldstein, Ph.D., Jay H. Lubin, Ph.D. (DCEG retired), Melissa Rotunno, Ph.D., Margaret A. Tucker, M.D., Director of the Human Genetics Program, and Sholom Wacholder, Ph.D. With over 100 collaborators, laboratory staff, surgeons, and support staff involved throughout the various stages of the study, this effort required meticulous organization.
The team selected the Lombardy region for its many unique features that make it well suited to this type of epidemiologic pursuit. A large proportion of the population smokes and lung cancer is common. In addition to tobacco, residents are exposed to other environmental factors known to influence lung cancer risk, particularly occupational exposures from manufacturing and farming, and air pollution. Further, Lombardy is served by a network of modern hospitals, medical schools, and a regional health service, which supported the recruitment effort. With the active involvement of hospital physicians, researchers were able to reduce participation biases.
As new cases were diagnosed and enrolled in the study, plans were in place to collect samples during initial surgical resection. Fresh tissues were pathologically examined and immediately frozen in liquid nitrogen in the operating rooms, while tissue blocks were collected after histological examination.
“We were in the operating rooms collecting multiple samples of the same tumors from each patient, as well as areas adjacent to the tumors, and those far distant,” Dr. Landi said. “In addition, we collected whole blood, serum, and buccal samples, and blood spots on Guthrie cards.”
Thanks to these efforts, EAGLE researchers were able to evaluate tumor, peritumor, and normal tissue for DNA and RNA, and complete integrated analyses of DNA sequencing, copy number aberration, gene expression and genome-wide methylation, features not possible in most studies. The blood specimens were transported to a central laboratory within four hours of collection to avoid the loss of certain analytes and degradation of RNA components. At the lab, they were processed in several ways to allow a variety of uses in future studies.
Lung cancer is the largest single cause of cancer mortality worldwide. A deeper understanding of its etiology, from inception to outcome, is needed to improve treatment and prevention strategies.
Early work in EAGLE focused on exposure-based association studies. Investigators conducted a series of descriptive analyses, examining the role of occupational exposures, dietary components, infections, previous lung diseases, hormonal factors, and circulating immune markers in the etiology of lung cancer. They found increased lung cancer risk in association with several occupations, particularly those involving exposure to asbestos, silica, and nickel-chromium; frequent meat intake; and some previous lung diseases. Investigators believe that many of these lifestyle, host, and environmental factors can influence chronic inflammation, a known pathway for tumorigenesis. In contrast, quercetin-rich foods (fruits and vegetables), history of pneumonia or mood disorders, low alcohol intake, and endogenous hormones during premenopausal years were inversely associated with lung cancer risk. They are currently analyzing the influence of air pollution, using data from detailed exposure maps covering the entire region.
The vast quantities of samples and data from EAGLE allow investigators to study gene-environment interactions, rapidly replicate previous findings, and investigate potentially new associations.
“[Our study] is unique because we have the population, materials, and samples to study any idea or question we may have,” Dr. Landi said. For example, DCEG fellow Guoqin Yu, Ph.D., M.S., and colleagues are currently exploring the role of microbiome in the lung, characterizing the typical bacteria found in histologically normal tissue from lung cancer patients.
EAGLE is also investigating how risk differs by histology and the genetic variations associated with specific histologic subtypes. Lung cancer is classified into two main groups: small cell and non-small cell. The latter includes adenocarcinoma, squamous cell carcinoma, and large cell carcinoma, along with rarer subtypes. The different histologies have diverse molecular characteristics, which may reflect differences in carcinogenesis, etiology, and response to treatment.
For example, in a genome-wide methylation analysis, Dr. Landi together with Jianxin Shi, Ph.D., and colleagues, found that methylation patterns, and the inherited variations that influence them, strongly differentiate risk for various subtypes of non-small cell carcinoma.
Another genetic analysis provided the first evidence of inherited susceptibility to a distinct lung cancer subtype. In a study of 33,000 participants from EAGLE and other cohort studies, a single nucleotide polymorphism in the TERT region was associated with risk of adenocarcinoma but not squamous cell carcinoma or small cell lung cancer. Similarly, a pathway analysis of inflammation-related genes showed that the RAD52 region was distinctly associated with squamous cell carcinoma risk. This finding was validated in three other study populations.
Epidemiologic and other scientific investigations have clearly implicated tobacco smoking as the primary cause of lung cancer, accounting for ~80% of the worldwide burden in males and at least 50% in females. However, risk varies substantially among smokers; smoking intensity, duration, and cumulative exposure (pack-years) can affect risk.
In a recent report, Dr. Caporaso, Fangyi Gu, M.Med., Sc.D., and colleagues found that participants who are highly addicted to nicotine–those who smoke their first cigarette within five minutes after waking up in the morning–are at higher risk of developing lung cancer than those who wait for an hour or more to smoke. This simple measure of nicotine dependency improved lung cancer risk prediction beyond standard smoking measures, such as cigarettes per day and pack-years. These findings underscore the need for even light smokers to quit, because even light smokers who are, or who become dependent smokers, were observed to be at substantial risk for developing lung cancer.
“The idea is that you target smokers at highest risk, and use behavioral, biomarker, and genomic tools to refine a risk model to select screening candidates.”
Ongoing and future work on disease progression within the EAGLE population is helping to address the clinical challenges of managing lung cancer patients. Recently, Dr. Landi, Mitchell H. Gail, M.D., Ph.D., and collaborators published results of an analysis of lung cancer prognosis, describing patterns of indicators for recurrence, metastases, and survival that may one day inform the treatment of patients. Indicators related to lung cancer recurrence, metastases and survival can inform treatment of patientsRead more about these findings in the NCI News Note.
In addition, EAGLE is contributing data to several consortia, including the International Lung Cancer Consortium (ILCCO) and the Transdisciplinary Research in Cancer of the Lung (TRICL) consortium. In a GWAS analysis with imputed data, Dr. Landi and TRICL collaborators identified a BRCA2 variant associated with lung cancer risk. For a smoker carrying this variant (which occurs in about 2% of the population), the risk of developing lung cancer was approximately doubled, which may have implications for identifying high-risk subjects for lung cancer screening. Additionally, future study of the effects of PARP inhibition in smokers with lung cancer carrying this BRCA2 variant may be warranted. Together with collaborators, EAGLE investigators are trying to identify additional highly predictive biomarkers. Ultimately they hope to design a risk prediction model utilizing the new markers.
“The idea is that you target smokers at highest risk, and use behavioral, biomarker, and genomic tools to refine a risk model to select screening candidates,” Dr. Caporaso said.
Is one lung tumor biopsy enough to define the cancer? The answer isn’t clear to scientists or clinicians. Two recent studies published conflicting evidence, but both were from small studies. EAGLE investigators hope to add more evidence to the discussion by analyzing samples collected from multiple regions of the same tumor.
“If we find the same mutations in every region, then one biopsy may be enough,” Dr. Landi said. “But, if there is heterogeneity, that is, different mutations, then one biopsy might not be enough.”
EAGLE investigators are also studying clonal evolution in lung cancer using their collection of tissue samples. “We think it’s important to understand the origin of the cancer,” Dr. Landi said. “If we know the most important mutations that occur in the initial cancerous cells, then potentially we can provide more targeted treatment.”
“EAGLE provides a superb framework for diverse studies and collaborative efforts to understand the etiology of lung cancer,” said Stephen J. Chanock, M.D., DCEG Director. “Our hope for the next wave of results—which will provide clues from next-generation sequencing, methylome, microbiome, metabolomics, and other approaches—is to provide additional insights to the critical questions for screening, prevention, and clinical management of this malignancy.”
Read other articles in the Summer 2015 issue of Linkage newsletter.