From July 2002 to December 2009, patients with ES and GOO caused by corrosive injury in a university-affiliated tertiary care center were recruited into this study. These patients had received early upper gastrointestinal endoscopy (GIF-Q240; Olympus Optical Co., Ltd., Tokyo, Japan) within 48 hours of ingestion. Mucosal burns of the esophagus, stomach, and duodenum were graded following a method previously reported by Zargar et al.: grade 0, normal examination; grade I, edema and hyperemia of the mucosa; grade II, subdivided into grade IIa (friability, hemorrhages, erosions, blisters, whitish membranes, exudates and superficial ulcerations), grade IIb (grade IIa plus deep discrete or circumferential ulceration), and grade III, multiple ulcerations and areas of necrosis [10]. If the patients demonstrated symptoms of upper gastrointestinal stricture, including dysphagia or easy satiety with postprandial vomiting, endoscopy was performed at four week after corrosive injury to examine the upper gastrointestinal tract. EBD was performed subsequently to patients with ES and GOO who satisfied the selection criteria by using through-the-scope balloon dilators. If active ulceration was noted within narrowing segment, the dilation procedure would be postponed and reevaluated two weeks later. Other exclusion criteria were (1) patients who had not perform an early endoscopy within 48 hours of ingestion, (2) patients who demonstrated symptoms of stricture but no upper gastrointestinal stricture on endoscopic examination, and (3) patients who decided to receive surgical intervention but not EBD. The patients were then divided into three groups esophageal stricture alone (ES), gastric outlet obstruction alone (GOO) and combination of ES and GOO (ES + GOO).
Endoscopic balloon dilation
With informed consent of each patient, ES and GOO were dilated under direct visualization by using controlled radial expansion (CRE) balloon catheter (Microvasive, Boston Scientific Corporation, Natick, MA, USA) but without guide wire and fluoroscopic guidance. Intramuscular hyoscine butylbromide 20 mg as an antispasmodic agent and intramuscular meperidine hydrochloride 50 mg as an analgesic agent was given approximately 10 minutes before staring the procedure unless contraindicated. Before dilation, the diameter of stricture was estimated by comparing with an open biopsy forceps. Then we selected a balloon catheter according to the diameter of stricture and negotiated the balloon catheter through the working channel of the endoscope across the stricture without fluoroscopic monitoring. The balloon was inflated with water to the recommended pressure for 60 seconds. In each session, the patient received three consecutive dilations with increment of dilation diameter not more than 3 mm following the rule of three [11]. The stricture length of ES and GOO was measured from distal end to proximal end after dilation while withdrawing the endoscope. The patients were kept fasting for four hours after the procedure and proton pump inhibitors were prescribed to suppress gastric acid. Inpatients received two sessions a week and outpatients one session a week. Serial dilations were performed by gradually increasing the balloon diameters up to a maximum of 15 mm until solid or semisolid food could be tolerated. If GOO were encountered after ES was dilated, subsequent EBD for GOO was performed. Under such circumstances, the dilations of ES and GOO were counted together in one session. If symptoms of stricture recurred, additional dilations were performed until symptoms were relieved again. The treatment outcome was considered successful when patients were able to maintain a solid or semisolid diet without having to perform an additional dilation for the next 12 months.
Clinical follow-up
Patients were treated with antacids or proton pump inhibitors for gastric acid suppression after each dilation session. Symptoms such as dysphagia and postprandial fullness sensation were recorded for each patient during the follow-up periods. Repeat dilations were performed for those patients with symptom relapse and proven to be stricture recurrences on clinical follow-up.
Study definitions
The treatment success was reached when patients could ingest solid or semisolid diet for more than 12 months without additional dilation needed. The presence of any untoward event after endoscopic treatment was considered a complication such as gastrointestinal tract perforation or bleeding with clinical signs of hematemesis, coffee-ground vomitus, hematochezia, or melena, or significant pain requiring hospitalization, was defined as major complications. This study was approved by both the Institutional Review Board and Ethics Committee of Chang Gung Memorial Hospital (98-2106B).
Statistical analysis
Continuous variables are given by mean and standard deviation. The continuous variables were analyzed by using the Mann–Whitney U test. Categorical variables were given in total and as percentages. They were analyzed by using the Fisher’s exact test. Two-sided P value of < 0.05 was considered significant. All statistical operations were performed using SPSS WIN version 15.0 (SPSS Inc., Chicago, IL, USA).