From the AGA Journals

Barrett’s esophagus risk factor profile may predict progression


Key clinical point: Male sex, older age, smoking, greater segment length, and low-grade dysplasia separately predicted progression of Barrett’s esophagus.

Major finding: Pooled odds ratios for risk ranged from 4.3 (low-grade dysplasia) to 1.03 (older age).

Study details: Systematic review and meta-analysis of 20 studies published through May 2016.

Disclosures: The reviewers disclosed no external funding sources or conflicts of interest.

Source: Krishnamoorthi R, et al. Clinical Gastroenterol and Hepatol. 2017 Nov 30.

View on the News

Risk factors other than BE length may be important

Endoscopic surveillance is currently recommended for nondysplastic Barrett's esophagus (BE), but there are conflicting results on the effectiveness of surveillance on esophageal adenocarcinoma outcomes. This meta-analysis by Krishnamoorthi et al. found several risk factors associated with BE progression (i.e., age, male sex, smoking, BE length) among patients with nondysplastic BE or low-grade dysplasia. Current recommendations for BE surveillance intervals are solely based on dysplasia grade without consideration for other high-risk features (i.e., smoking, BE length, age). This meta-analysis demonstrates that some patients with nondysplastic BE are at a higher risk of neoplastic progression, and the AGA recommendation for BE surveillance every 3-5 years may not be suitable for all.

Dr. Mimi C. Tan

In addition, proton pump inhibitor, statin, and nonsteroidal anti-inflammatory drug use were associated with lower risk of BE progression, although inconsistently in studies that adjusted for age, sex, and BE characteristics. Current studies on medication chemoprevention of neoplastic progression in BE are limited by residual confounding inherent in observational studies. I anticipate that the results of the Oxford AspECT clinical trial on chemoprevention with esomeprazole with or without aspirin will conclusively answer this question.
IMG: 2400A107.SIG Tan_Mimi_TEXAS_web

Parasa et al. recently developed a risk prediction model to stratify risk of progression in patients with nondysplastic BE based on BE length, male sex, smoking, and baseline low-grade dysplasia. Patients with one or more of these risk factors are at highest risk of neoplastic progression and may benefit from shorter surveillance intervals or endoscopic eradication therapy.

Mimi C. Tan, MD, MPH, is a postdoctoral fellow in gastroenterology and hepatology, T32 research track at Baylor College of Medicine, Houston, and an investigator at the Center for Innovations in Quality, Effectiveness, and Safety at the Michael E. DeBakey VA Medical Center, Houston. She has no conflicts.



Older age, male sex, smoking, longer segment length, and low-grade dysplasia were significant risk factors for progression of Barrett’s esophagus in a meta-analysis of 20 studies.

“Individuals with these features should undergo more intensive surveillance or endoscopic therapy,” Rajesh Krishnamoorthi, MD, of Mayo Clinic in Rochester, Minn., and his associates wrote in Clinical Gastroenterology and Hepatology. “Smoking is a modifiable risk factor for cancer prevention in patients with BE.”

“Currently, gastrointestinal societies’ guidelines on BE surveillance are solely based on dysplasia grade and do not take into account any of the other risk factors,” the reviewers concluded. Their findings could form the backbone of a risk score that identifies high-risk BE patients with baseline low-grade dysplasia or nondysplastic BE “who would benefit from intensive surveillance or endoscopic therapy.”

Esophageal adenocarcinoma is on the rise and fewer than one in five patients survive 5 years past diagnosis. Endoscopic surveillance for esophageal adenocarcinoma is recommended in Barrett’s esophagus, but only about one in 10 esophageal adenocarcinoma patients has a preceding BE diagnosis. “This ostensible discrepancy has raised concerns about the effectiveness of current screening and surveillance programs,” the reviewers noted. Studies also have yielded conflicting evidence about the value of endoscopic surveillance as currently performed. To help prioritize BE patients for surveillance, the reviewers searched EMBASE, MEDLINE, and Web of Science from inception through May 2016 for cohort studies of risk factors for progression of BE among patients with either no dysplasia or low-grade dysplasia.

The 20 studies covered 1,231 BE progression events among 74,943 patients. In separate pooled estimates, progression of BE correlated significantly with older age (odds ratio, 1.03; 95% CI, 1.01–1.05), male sex (OR, 2.2; 95% CI, 1.8-2.5), current or former smoking (OR, 1.5; 95% CI, 1.09-2.0), and greater BE segment length (OR, 1.3; 95% CI, 1.16-1.36). Results tended to be homogeneous among studies, said the reviewers. Low-grade dysplasia correlated strongly with progression (OR, 4.3; 95% CI, 2.6-7.0), while use of proton pump inhibitors (OR, 0.55; 95% CI, 0.32–0.96) and statins (OR, 0.48; 95% CI, 0.31-0.73) showed the opposite trend. “Alcohol use and obesity did not associate with risk of progression,” the reviewers added.

Thirteen studies in the meta-analysis were from Europe, six were from the United States, and one was from Australia. Ten were multicenter studies, 13 were deemed high-quality, three were deemed medium-quality, and four were deemed low-quality. The reviewers were unable to assess dose-response relationships for relevant factors, such as alcohol, tobacco, and medications, and not all studies accounted for potential confounding.

Only four studies included multivariate analyses to control for the confounding effects of age, sex, and BE characteristics (length and dysplasia). When the reviewers analyzed only these studies, older age and smoking no longer predicted BE progression. Use of proton pump inhibitors remained protective, and use of nonsteroidal anti-inflammatory drugs (NSAIDs) became protective, while statin use lost significance.

The reviewers disclosed no external funding sources or conflicts of interest.

SOURCE: Krishnamoorthi R, et al. Clinical Gastroenterol and Hepatol. 2017 Nov 30. doi: 10.1016/j.cgh.2017.11.044

Next Article: