Prediction of progression in Barrett’s esophagus: does inflammation hold the key?

The clinical significance of Barrett’s esophagus (BE) lies in it being the only known precursor to esophageal adenocarcinoma (EAC), which occurs via the development of dysplasia. As endoscopic therapy of dysplasia can prevent EAC, surveillance is recommended to detect dysplasia. Endoscopic surveillance is recommended every 3–5 years in those without dysplasia (NDBE). Most gastrointestinal society guidelines do not personalize surveillance intervals on the basis of patient or BE characteristics [ 1 ]. Unfortunately, endoscopic surveillance as currently performed is only modestly effective in reducing EAC mortality [ 2 ]. Hence, risk prediction algorithms to tailor surveillance and management on the basis of progression risk is highly appealing. Several investigators have focused on genomic or epigenetic biomarkers to predict BE progression risk. However, most have not progressed beyond Phase 1 or Phase 2 studies given the challenge with prospective validation [ 3 ]. In this issue of Endoscopy , Peleg et al. take a novel approach to this elusive quest, by using the neutrophil to lymphocyte ratio (NLR) in peripheral blood samples to predict progression in a retrospective analysis of a prospectively followed BE cohort [ 4 ]. The NLR is the ratio of the absolute neutrophil count (ANC) and absolute lymphocyte count (ALC) in peripheral blood. In systemic inflammatory states, elevation of the ANC and depression of the ALC leads to elevation of the NLR. Several studies have associated an elevated NLR with poor outcomes (overall and cancer-free survival, poor response to chemotherapy) in a variety of cancers including esophageal carcinoma [ 5 ] [ 6 ]. While the exact mechanism of this phenomenon is unknown, inhibition of antitumor lymphocytes and natural killer cells by neutrophilic peritumoral inflammation is postulated [ 7 ]. Cytokines and vascular endothelial growth factor produced by neutrophils may also promote tumor growth by facilitating genomic alterations. “Confounding of the association between metaplasia and neoplasia and elevated neutrophil to lymphocyte ratio (NLR) by other variables such as age, sex, obesity (which causes a systemic inflammatory state), and aspirin or statin use (noted to be higher in the high NLR group) also needs to be carefully assessed in subsequent studies.” Peleg et al. [ 4 ] hypothesized that given the known association between esophageal inflammation and neoplasia, an elevated NLR may be associated with an increased BE progression risk. They defined progression appropriately as high grade dysplasia (HGD) or EAC at least 1 year after index endoscopy in patients with NDBE or low grade dysplasia (LGD). The NLR was calculated from blood samples taken within 3 months of the index endoscopy. Surveillance was standardized and most patients remained on twice-daily proton pump inhibitors (PPI). Dysplasia was confirmed by an expert gastrointestinal pathologist. Tissue inflammation was scored (as mild, moderate, severe) in a subset of patients. Of 324 patients in the cohort (74 % NDBE), progression was seen in 13 (annual incidence of progression 1.0 %) over a median follow-up of 3.7 years. However, only three of these progressors had baseline NDBE. Using an optimal NLR cutoff of 2.4, the area under the receiver operating characteristic curve for prediction of progression with NLR was 0.88 (sensitivity 88 %, specificity 77 %). This estimate was not validated either internally or externally in an independent cohort. The predictive value of NLR in a Cox model adjusting for covariates (LGD, BE length, presence of a visible lesion) showed NLR to be an independent predictor of progression (hazard ratio 3.2, 95 % confidence interval 1.8–5.8). However, important variables such as age and sex were not included in this model. The authors are to be congratulated for taking a fresh perspective to this vexing issue, which could have a substantial impact on the management of BE patients, if validated. Obvious advantages of such a “biomarker” are the ease of obtaining a sample and measuring the “variables”: both the ANC and ALC are routinely reported in complete blood counts from blood samples, making this less expensive and more accessible than esophageal biopsies. Integration into a clinical BE prediction risk score such as the Progression in BE (PIB) score [ 8 ] could improve its predictive ability. The NLR cutoff used in this study was similar to that used in other studies and was derived from a robust bootstrap validation analysis. The progression rate in the cohort also appears to be in line with other cohort studies. However, several issues need to carefully considered as we think of next steps. The biological plausibility of this concept needs to be explored further: can a single NLR value from 3 months before index endoscopy truly predict BE progression several years later? While NLR predicts poorer outcomes in malignancy, its association with predicting progression in BE is biologically more tenuous, particularly as most patients in the cohort were on twice-daily PPIs, with likely reasonable control of reflux-induced inflammation. The authors did attempt to grade tissue inflammation, but this was done in less than 50 % of participants and methods were not clearly described. While NLR correlated with the degree of tissue inflammation, the authors did not analyze this as a predictor of progression. Tissue-level NLR needs to be further explored in a rigorous manner with validated and standardized techniques. Indeed, in a previous study, tissue-level NLR (assessed by immunohistochemistry for neutrophil- and lymphocyte-specific markers) in tumor nests, was shown to also predict clinical outcomes in patients with esophageal squamous cell carcinoma [ 5 ]. Additionally, in the NDBE group, there were only three progressors and hence multivariable analysis could not be conducted in this subset. Therefore, these results need to be validated in an independent cohort of patients and the additive value of NLR to a clinical model (such as the PIB score) needs to be assessed, particularly in a sample of patients with NDBE, given recommendations for ablation in those with confirmed LGD. Confounding of this association by other variables such as age, sex, obesity (which causes a systemic inflammatory state), and aspirin or statin use (noted to be higher in the high NLR group) also needs to be carefully assessed in subsequent studies. Development of such a score may allow true personalization of management recommendations for patients with BE: those with a low-risk score being surveyed less frequently, while those at a higher risk undergoing careful intensive surveillance and endoscopic treatment of dysplasia if detected. In conclusion, the authors of this study have drawn attention to a hitherto mostly unstudied, easily analyzable and intriguing “circulating biomarker” of potential value in predicting progression in patients with BE. Elucidation of a potential mechanism, assessment of tissue correlation, and robust external validation are important next steps to explore the value of NLR in influencing patient care. Publication History Publication Date: 27 July 2021 (online) © 2021. Thieme. All rights reserved. Georg Thieme Verlag KG Rüdigerstraße 14, 70469 Stuttgart, Germany Refers to: Neutrophil to lymphocyte ratio and risk of neoplastic progression in patients with Barrett’s esophagus Endoscopy 2021; 53(08): 774-781 DOI: 10.1055/a-1292-8747