Esophageal Adenocarcinoma: Disease Detectives Study Etiology, Biomarkers, and Risk
, by Jennifer K. Loukissas, M.P.P.
For the majority of cancers, the likely precursors rarely go on to become malignant, or occur in tissues in the body that are less amenable to treatment, making it hard to develop cost-effective prevention strategies that balance risk reduction with unnecessary medical intervention. Such is the case for an increasingly common condition, known as Barrett’s esophagus (BE), which may evolve into esophageal adenocarcinoma (EA).
EA is the malignant endpoint of the continuum of disease which may initiate with severe gastroesophageal reflux disease (GERD), a condition thought to cause erosion of the lining of the esophagus. In the case of BE, the body sometimes mistakenly re-lines the esophagus with tissue similar to the lining of the intestine. This metaplastic tissue change is named for an Australian-born surgeon named Norman Rupert Barrett, who published his observations in 1950, though others noticed the pattern before him.
Interest in BE arose out of the investigation of alarmingly high annual increases in the rate of EA in the United States. Diagnosis with BE increases a patient’s risk of EA up to 10–30 times that of the general population. Although still a relatively rare cancer, EA incidence has gone up more than 600 percent in the United States since 1973, from 0.36/100,000 to 2.68/100,000 per year in 2014. In 1997, adenocarcinoma surpassed squamous cell carcinoma as the predominant type of esophageal cancer in the United States. The two cancers are histologically distinct, with differing etiologies.
Michael Cook, Ph.D., of the Metabolic Epidemiology Branch (MEB), leads investigations of EA and BE with an aim to clarify etiologic factors and improve risk prediction. According to Dr. Cook, there are two main clinical questions in this research field: Can we develop a cost-effective means to identify and risk triage patients with BE so they can be clinically managed with precision? And, is there a biomarker that could serve as a low cost, highly sensitive test to diagnose EA?
Dr. Cook uses both classical and molecular epidemiology in cross-sectional, case-control, and prospective studies to describe pathogenesis and identify biomarkers.
This work is rife with complex epidemiological and methodological challenges. Some of the known and potential risk factors are linked, making it difficult to tease apart their various contributions and elucidate putative biological mechanisms. For instance, GERD and obesity—particularly abdominal obesity—have been identified as risk factors for both BE and EA, and, although obesity increases risk for GERD, it also likely confers risk for BE and EA via systemic inflammation and metabolic syndrome. Second, for reasons yet unknown, EA has one of the highest male-to-female sex ratios of any malignancy: between five and eight men develop EA for every woman diagnosed. Third, the rate of progression from BE to EA is fairly low: about one case for every 400 person-years of follow-up, which makes prospective studies of the carcinogenic process difficult. Lastly, according to an analysis of data from the NCI SEER program and the National Program of Cancer Registries by Dr. Cook and colleagues, there is marked racial and geographic variability within the United States.
Seeking to tackle these problems, Dr. Cook uses both classical and molecular epidemiology in cross-sectional, case-control, and prospective studies to describe pathogenesis and identify biomarkers. If there were an effective method to discern patients at low- or high-risk of progression to EA, interventions such as ablation/endoscopic surgery aimed at removing the affected tissue and surveillance (e.g., periodic endoscopy) could be targeted to individuals with the highest risks, likely resulting in a reduction in the number of cases presenting with large or late-stage malignancies, which are rapidly fatal.
To compile a sufficient number of cases for study, Dr. Cook collaborates with the International Barrett’s and Esophageal Adenocarcinoma Consortium (BEACON), composed of over three dozen investigators with completed or ongoing epidemiologic studies of BE and/or EA. Using this resource, Dr. Cook has provided evidence that obesity—which traditionally had been hypothesized to have a direct mechanical effect on risk of EA by interfering with the gastroesophageal sphincter—likely exerts indirect, non-mechanical effects. Obesity is known to wreak havoc on various metabolic pathways—including the production and metabolism of sex steroid hormones and inflammatory cytokines—and related health conditions, like metabolic syndrome and diabetes. To build on this evidence base, Dr. Cook and colleagues are engaging the full range of next-generation technologies to compare metabolic patterns to tease apart the various contributing factors.
The search for better biomarkers
Dr. Cook has led and designed an NCI Cohort Consortium project utilizing data from seven prospective studies for a total of 300 EA cases and an equal number of controls. In collaboration with the laboratory of Dr. Ligia Pinto at NCI Frederick, the research team has tested 70 markers of inflammation and immune response—including several adipocytokines—using prediagnostic serum.
Findings from this study identify nine markers associated with future risk of EA—in particular, soluble TNF receptor II. This association did not change with increased time between blood draw and diagnosis, increasing the likelihood that EA itself need not be present to observe the association. Importantly, level of excess body fat only had slight attenuating effects on these relationships. This suggests that the relationships between these circulating biomarkers and cancer risk are not mediated by obesity and highlights the importance of a heightened systemic inflammation in the etiology of EA. Work is ongoing to determine whether a panel of these circulating immune-inflammation markers can assist with triaging of patients at high risk for EA.
Identification of such a high-risk population is currently difficult due to the fact that less than 10 percent of EA cases have a preceding diagnosis of BE. Dr. Cook has used cutting-edge platforms to search for a tissue microRNA signature that distinguishes BE with high sensitivity and high specificity from normal subjects without BE. For example, he is now using this BE tissue microRNA signature in collaboration with Dr. Rebecca Fitzgerald at the University of Cambridge, England. Dr. Fitzgerald has designed and implemented the Cytosponge—a low-cost esophageal sampling device in which patients swallow a capsule containing a sponge. Once in contact with gastric acid in the stomach, the sponge expands and is then pulled up through the esophagus, collecting cells on its way. Together, the researchers have been able to detect microRNAs from their BE tissue microRNA signature in these Cytosponge samples, which provided extremely high sensitivity and specificity for BE. This evidence is a critical step toward the utilization of the Cytosponge to diagnose BE.
“With smart design of human epidemiologic studies, we can quickly discover and implement molecular signatures of precancers that further the central tenets of Precision Medicine.”
Work is ongoing to determine whether circulating immune-inflammation markers or microRNAs can assist with risk-triaging of patients at high risk for EA. Since microRNAs can be actively and passively exported by cells into the blood, Dr. Cook is also assessing whether tissue microRNA signatures are detectable in blood, possibly paving the way for a cheap, minimally-invasive EA diagnostic test.
“That is a big leap of faith, and a much more high-risk project, but we are starting with a small pilot,” said Dr. Cook. Early results using HTG EdgeSeq appear promising: two tissue microRNAs provided significant distinction between cancer cases and controls when measured in blood.
Next steps: Building the next study population
Dr. Cook is a big advocate of prospective studies. “We cannot reconstruct the natural history of a disease with data from stage-specific, cross-sectional studies; and rare cancers are no exception,” he said. To address this gap, he is assessing the feasibility of a multi-institutional, retrospective case-control study of BE to discover tissue biomarkers associated with progression to EA.
Dr. Cook imagines a future that maximizes the potential of precancerous states. “With smart design of human epidemiologic studies, we can quickly discover and implement molecular signatures of precancers that further the central tenets of Precision Medicine.”