Society: AGA

Introduction: Shifts towards high-fat, low-fiber diets, may contribute to the rising incidence of esophageal adenocarcinoma (EAC). In mice high-fat diet promotes EAC, possibly through effects on the gut microbiome and the systemic bile acid pool. Reflux bile acids are thought to promote EAC, but the role of systemic bile acids is unknown. The aim of this study was to assess associations between systemic bile acids and stages of progression in Barrett’s esophagus (BE).
Methods: Subjects with and without BE were enrolled from three sites. Targeted bile acid profiling on serum collected on the day of endoscopy was performed by LC-MS. A subset (73%) of subjects completed a 12-month food frequency questionnaire. MANOVA was performed to assess global bile acid differences across groups. Principal components analyses were performed to identify drivers of global bile acid composition. Individual bile acid comparisons across groups were performed only for Kruskal-Wallis ANOVA p<0.05. Logistic regression analyses were performed to assess associations between logarithmically-transformed bile acid levels and neoplastic stages of BE.
Results: A total of 141 subjects were enrolled with bile acids profiled (49 non-BE; 92 BE: 44 no dysplasia (ND), 9 indefinite (IND), 17 low-grade dysplasia (LGD), 16 high-grade dysplasia (HGD), 6 EAC). Lower Healthy Eating Index score (HEI), older age, and higher BMI were associated with increased levels of several individual bile acids, with an inverse association with PPI use. Global bile acid pools were distinct comparing non-BE and BE subjects (p=0.002) and between non-BE and stages of BE neoplasia (p=0.004). Principal components analyses demonstrated that associations between bile acids and HGD/EAC were driven by the unconjugated primary bile acids cholic acid (CA) and chenodeoxycholic acid (CDCA) as well as forms of ursodeoxycholic acid (UDCA). There were clear shifts in individual bile acids with progression to HGD/EAC. (Fig 1) Increasing CA was associated with HGD/EAC in univariate analyses (Fig 2) and remained significant after adjusting for established EAC risk factors (vs. non-BE; aOR 2.03, 95%CI 1.11-3.71). Similar trends were seen after adjustment for HEI, fat, and fiber intake. CA and CDCA were highly correlated; the two combined were also associated with HGD/EAC (aOR 1.81, 95%CI 1.04-3.13). There were non-significant trends towards decreased levels of unconjugated and conjugated forms of UDCA associated with HGD/EAC.
Conclusions: Alterations in serum bile acids are independently associated with advanced neoplasia in BE and may contribute to neoplastic progression. Future studies should explore associated gut microbiome changes such as bacterial deconjugation via bile salt hydrolase, pro-neoplastic effects of CA and CDCA and protective effects of UDCA, and whether these bile acids represent viable therapeutic targets.

<b>Figure 1. </b>Heatmap of relative risk ratios from polytomous logistic regressions for associations of BE stages compared to controls for serum bile acid levels.

Figure 1. Heatmap of relative risk ratios from polytomous logistic regressions for associations of BE stages compared to controls for serum bile acid levels.

<b>Figure 2.</b> Cholic acid was significantly increased across stages of progression to EAC (p for trend=0.002).

Figure 2. Cholic acid was significantly increased across stages of progression to EAC (p for trend=0.002).

Introduction A confirmed diagnosis of low-grade dysplasia (LGD) on two endoscopies is currently considered a predictor for neoplastic progression in patients with Barrett’s esophagus (BE), and, therefore, ablation therapy is offered to these patients. However, recent studies show that aberrant expression of the p53 tumor suppressor protein might be a stronger risk factor for neoplastic progression than dysplasia status. The aim of this study was to evaluate the value of p53 expression compared to confirmed LGD in predicting neoplastic progression in BE patients.

Method Data was analyzed from a large prospective cohort study of 1031 BE patients (73% males, median age 61 years) with index diagnosis of non-dysplastic BE (NDBE), indefinite for dysplasia (IND), and LGD. Confirmed LGD was defined as a LGD diagnosis on 2 separate occasions, each LGD diagnosis was confirmed by two pathologists. P53 immunohistochemistry staining was performed on biopsies of 1545 endoscopies from 655 (64%) patients; both overexpression and loss of expression were considered aberrant. Neoplastic progression was defined as high-grade dysplasia (HGD) and/or esophageal adenocarcinoma (EAC). Cox regression analysis was used to determine neoplastic progression risk.

Results During a median follow-up of 6.2 (IQR 3.2-11.4) years, 73/1031 (7%) patients developed HGD/EAC. Neoplastic progression was found in 23/756 (3%) NDBE patients, 27/176 (15%) patients with a single diagnosis of LGD, and 23/99 (23%) patients with confirmed LGD. P53 expression was aberrant in 28/411 (7%) NDBE patients, 50/149 (34%) single LGD patients, and 51/95 (54%) confirmed LGD patients. Aberrant p53 expression was strongly associated with an increased risk of neoplastic progression after adjusting for age, gender, segment length, oesophagitis, and grade of dysplasia (HR 13.5, 95% CI 7.1-25.8). In NDBE patients, the absolute neoplastic progression risk increased from 2% with normal p53 expression to 50% with aberrant p53 expression. In patients with a single diagnosis of LGD, the progression risk increased from 8% to 37% when p53 expression was aberrant. Similar risks where found in patients with confirmed LGD, with a neoplastic progression risk of 6% with normal p53 expression and 47% with aberrant p53 expression.

Conclusion Aberrant p53 expression is a strong predictor for neoplastic progression in BE patients. The predictive value of aberrant p53 expression is independent of the grade of dysplasia. Aberrant P53 expression may be a better parameter than (confirmed) LGD to identify patients who need close surveillance or could benefit from ablation therapy.

Pauline Zellenrath

Appearances