To Study the Incidence of Barrettʼs Esophagus in Patients with Liver Cirrhosis

Abstract

Purpose: GERD is increasing in the general population. In cirrhotic patients, incidence may even be higher because of the use of alcohol, presence of ascites and use of drugs like beta-blockers, which reduce the LES pressure. de la Pena et al showed that the sclerosant used for sclerotherapy provokes GERD. COX-2 gene is overexpressed in patients with cirrhosis and hypothetically may worsen GERD (Grover et al). In patients with esophageal varices, esophageal transit time is delayed and GERD is common in up to 64% of the patients (Fass R et al). Sclerotherapy can decrease the mean amplitude of esophageal contractions and normal peristalsis may be occasionally replaced by non-propagating simultaneous contractions that may cause chest pain and/or dysphagia in the absence of stricture. Cirrhotics have screening EGD's to evaluate for esophageal and gastric varices. The investigators have noted an increased incidence of high-grade dysplasia in cirrhotics. We hypothesize that there may be an increased risk of Barrett's esophagitis in patients with cirrhosis. Methods: Inclusion criteria: All patients with biopsy proven cirrhosis who underwent EGD's at UMass between 1/1/98 to 6/30/03 and control patients who underwent EGD in the month of April 2004 for any other indication. Exclusion criteria: Inability to find the patients in the Meditech system. Also, patients with known Barretts who presented for a surveillance EGD were excluded. Data collection: 249 cirrhotics and 153 controls met the criteria. The control group was divided into two groups, namely subjects with symptoms of GERD (pain, heartburn) and those with all other indications for the EGD. Results: In the cirrhotic group(ages 17 to 82 years, 80 females,168 males), 19/249 had endoscopic appearance of Barretts esophagus, with 14/19 having had a biopsy. 8/14(3.2%) patients had biopsy findings consistent with Barretts esophagus. In the controls with GERD symptoms (ages 29–80 years, 26 females, 10 males), 2/37 had endoscopic Barretts, with 1 with positive biopsy (2.7). In the third group (ages 22–83 years, 51 males and 54 females), 3/105 had endoscopic appearance but only 1/3 (0.95%) had biopsy proven Barretts. Conclusions: This data supports the hypothesis that patients with cirrhosis may be at increased risk of developing Barretts esophagus. A larger, prospective study is needed. If this relationship can be proven, it might become a standard of care to regularly screen cirrhotic patients for Barretts esophagus.