Stability of immune microenvironment changes associated with progression risk in Barrett’s esophagus.

Abstract

450 Background: Barrett’s esophagus (BE) is the principal risk factor for esophageal adenocarcinoma (EAC). Since only a small number of patients with BE progress to high-grade dysplasia (HGD) or EAC, biomarkers of risk are in high demand. We previously reported alterations in the relative abundance of lamina propria lymphocyte and macrophage subsets in individuals with BE at the time of diagnosis. In this study, we extended our findings to include serial surveillance samples obtained from these patients to assess the stability of these alterations over time. Methods: Formalin-fixed, paraffin-embedded tissue samples from patients with non-dysplastic BE undergoing endoscopic biopsy were examined. Initial diagnostic samples (n=58) and serial surveillance samples (n=61) from 58 unique patients were examined, among whom 25 were subsequently diagnosed with HGD or EAC (“progressors”), while 33 patients did not progress during at least 5 years of surveillance (“non-progressors”). All patients had an initial and at least one follow-up surveillance sample. Immunohistochemical staining was performed with antibodies specific for CD3 and CD8 (lymphocytes), CD68/CD86 co-positive (“M1- like”) macrophages, and CD68/CD206 co-positive (“M2-like”) macrophages. Regions of interest (ROI) selected from the lamina propria in the region of BE were subjected to digital image analysis, and subgroup comparisons of immune cell densities were carried out using the Wilcoxon rank sum test for the outcome of progression. Results: A total of 1920 ROIs were examined for non-progressors (1505 at initial diagnosis and 415 in follow-up biopsies) and 1333 ROIs were examined for progressors (738 initial/595 follow-up). Relative to non-progressors, progressor biopsies contained a markedly greater density of CD3+ lymphocytes (4900 vs. 2634, p=0.0015) and a significantly lower density of CD8+ lymphocytes (929 vs. 1359, p=0.04). This difference persisted at follow-up biopsy, suggesting that it is an enduring feature of the BE microenvironment (CD3+:1528 vs. 1248, p<0.0001; CD8+: 339 vs. 422, p=0.005). Concomitantly, relative to non-progressors, M1-like macrophages were significantly elevated in progressors, (91.4 vs. 33.8, p<0.001), whereas M2-like macrophages were decreased (33.3 vs 81.9, p<0.001). Conclusions: The immune microenvironment of BE demonstrates substantial differences in lamina propria lymphocyte and macrophage populations in patients who are destined to progress to HGD or EAC, relative to those who are not. These findings, which we show here to persist in serial biopsies over time, warrant further investigation to deepen our understanding of this altered immune microenvironment, both as a biomarker of progression risk and a potential target for immunomodulatory prevention strategies.