Profile of gastric varices among Sri Lankan cirrhotics

Abstract

BACKGROUND AND AIMS: Gastric varices (GV) can result in life threatening bleeding with a higher mortality than esophageal varices. There have been no studies on the characteristics of GV among Sri Lankan cirrhotics. Aim of this study was to perform a descriptive analysis of GV among a cohort of Sri Lankan cirrhotic population. METHODS: We analyzed medical records of all upper gastrointestinal endoscopies performed on cirrhotics, at the University Endoscopy Unit, Colombo North Teaching Hospital, Ragama, Sri Lanka from 2006 to 2011. Characteristics of GV, demographics, indications and fi ndings at endoscopy were analyzed and they were compared among patients with Oesophageal varices (EV). RESULTS: Out of 641 cirrhotics screened, 628 had a complete data set for analysis. GV was detected in 70 (11%) patients; male:female 8.7:1.3; mean age 55 (SD = ± 10.7) years. From these 48/70 had EV (Gastro Oesophageal Varices GOV1 – 18/48, GOV2 – 30/48) in addition to GV. Only 22/70 had Isolated GV (IGV1–10, IGV2–12). Among patients with GV 38 (54%) had portal hypertensive gastropathy and 3 (4%) had gastric antral vascular ectasia. Nineteen (27%) of GV were detected on presentations with UGIB (6 with IGV, 13 with GOV), whereas 51 (73%) were detected on routine screening. EV was detected in 288 (46%) of cirrhotics (Isolated EV 240, GOV 48). Seventy seven (32%) of EV were detected on presentations with UGIB, whereas 163 (68%) were detected on routine screening. There was no statistically significant difference on presentation with UGIB between isolated EV (77/240) vs. IGV (6/22) patients (p = 0.64; χ2 = 0.2). CONCLUSION: The profi le of GV among our cirrhotics is comparable to previous reports from other centres. Findings suggest that in cirrhotic patients presenting with UGIB, a careful search for the presence of GV is as important as identifying EV, even among patients who have EV.