Craig S Brown, Chi Wang, PhD, Jay L Goldstein, MD, John G Linn, MD, Ervin W Denham, MD, Stephen P Haggerty, MD, Mark S Talamonti, MD, John A Howington, MD, Michael B Ujiki, MD. NorthShore University Health Systems, Pritzker School of Medicine at the University of Chicago.
Introduction: Barrett’s esophagus is the most predictive risk factor for development of esophageal adenocarcinoma, a malignancy with the fastest increasing incidence rate in the US. Currently, there is a lack of consensus as to the proper management of patients who have a positive biopsy followed by a negative biopsy during a subsequent endoscopy. The aim of this study was to investigate the differences in exposures, demographics, and comorbidities between patients who regress and those who do not regress.
Methods and Procedures: We retrospectively collected and analyzed data from a cohort of Barrett’s Esophagus patients participating in this single-center study comprised of all patients diagnosed with Barrett’s Esophagus at NorthShore University Health Systems hospitals and clinics over a 10 year period. Patients were followed in order to identify those progressing from Barrett’s Esophagus to high-grade dysplasia (HGD), esophageal adenocarcinoma (EAC), or those regressing to no Barrett’s. Mean follow up period was 5.4 years (9088 patient-years). We collected information from the patient’s electronic medical records regarding demographic data, endoscopic findings, histological findings, exposures, and history of antireflux surgery. Chi–squared and student’s t-tests were performed where appropriate.
Results: This study included 2563 total Barrett’s esophagus patients, 473 of which had their Barrett’s esophagus regress (had a negative endoscopy confirmed by biopsy). The group that regressed was 53.1% male, while the group that did not regress was 64.8% male (p<0.0001). Mean age was 65.1 (±13.6) and 61.8(±13.2) years for non-regressing and regressing patients, respectively (p<0.0001). No difference was seen in BMI between groups, with non-regressing patients having a mean BMI of 28.04±10.47 compared to regressing patients at 28.13±12.70 (p=0.88). No difference was seen between groups with respect to PPI use (91.1% using for non-regressing patients vs 93.5% using for regressed patients, p=0.12). Interestingly, groups differed in their rates of vitamin D use, with regressing groups taking vitamin D more commonly (48.1%) than those who did not regress (37.9%) (p=0.0001). As expected, groups differed with respect to their segment length at diagnosis. The group that regressed had an average segment length of 3.28 cm±3.07 cm, in contrast to the group that did not regress which had an average length of 1.46 cm±1.66 cm (p<0.0001). Interestingly, not a single patient in the regression group progressed to HGD/EAC, while 47 patients of the non-regressing patients in the cohort progressed to HGD/EAC, a result that was found to be independent of long segment vs. short segment on multivariate analysis (p=0.029).
Conclusions: Currently, several studies have shown risk factors that can predict progression from non-dysplastic Barrett’s esophagus to HGD/EAC, but few if any studies are available investigating predictors for regression. Our study reports several factors that can be used to predict patients who will regress from Barrett’s esophagus and those who likely will not, tools that will be useful in tailoring therapeutic and surveillance strategies.