Portal de Periódicos da CAPES

Gastric Mucosal Injuries in Children from Inflatable Low-Profile Gastrostomy Tubes

1997; Lippincott Williams & Wilkins; Volume: 24; Issue: 1 Linguagem: Inglês

10.1097/00005176-199701000-00017

1536-4801

Sadaf Kazi, Thirumazhisai S. Gunasekaran, James Berman, Hymie Kavin, Jerome R. Kraut,

Intestinal Malrotation and Obstruction Disorders

Feeding gastrostomy tubes (G-tubes) are being inserted with increased frequency in children in order to supplement nutrition in a variety of chronic conditions such as mental motor retardation, failure to thrive, cystic fibrosis, Crohn's disease, and metabolic diseases that require avoidance of fasting (1-3). A variety of complications have been reported with both the Foley-type G-tubes and the low-profile G-tubes (Button, Bard Interventional Products, C.R. Bard Inc., Tewksbury, MA) (2,4). These complications include infection, obstruction, perforation, aspiration, and fistula formation. A new generation of inflatable/deflatable low-profile G-tubes (ILPGT, Mic-Key, Ballard Medical Products, Draper, Utah) has been introduced. Their advantage is that balloon deflation makes the devices easier to insert and remove. We report a spectrum of gastric mucosal injuries in five children after ILPGT insertion. PATIENTS AND METHODS Five patients with ILPGT were evaluated for a variety of clinical presentations: recurrent upper gastrointestinal bleeding, vomiting, chronic cough, or retching and gagging with infusion of feeds through the G-tubes. All patients were seen by the Division of Pediatric Gastroenterology and Nutrition, Lutheran General Children's Hospital, between January 1994 and February 1995. The patient characteristics are given in Table 1. All patients previously had a fundoplication. The ILPGT had been placed between 1 and 4 months before the diagnostic endoscopy. The ILPGT were of variable sizes (18-24 French, 1.2-2.5 cm). None of the patients were receiving aspirin or nonsteroidal antiinflammatory drugs (NSAIDs). Upper gastrointestinal endoscopy was done with a Pentax EG2901 video endoscope under general anesthesia. ENDOSCOPIC FINDINGS AND OUTCOME Case 1 At endoscopy, a single 6-mm diameter gastric ulcer was noted on the lesser curvature of the distal body directly opposite the tip of the G-tube. The tip of the G-tube abutted the ulcer during the procedure (Fig. 1). Therefore, the ILPGT was changed to a Malecot G-tube and, at a later date, to a button-like G-tube. Followup endoscopy two months later showed complete healing of the ulcer despite the fact that the patient had not been treated with H2 blockers or antacids. Case 2 Endoscopy showed multiple gastric erosions in a circumscribed area on the lesser curvature of the proximal antrum directly opposite the tip of the G-tube (Fig. 2). The patient was treated with ranitidine and cisapride, but this failed, as he continued to have persistent vomiting. Because of failure of the fundoplication and prolonged gastric emptying, a repeat fundoplication with proximal positioning of the G-tube and a pyloroplasty were performed. The same size ILPGT was used. Ranitidine and cisapride were stopped. Followup endoscopy 3 months later showed normal gastric mucosa. Case 3 At endoscopy the findings were almost identical to those in Case 2 but were more proximally located in the distal body: multiple erosions over a localized area of the gastric wall directly opposite the tip of the G-tube. This patient was treated with ranitidine and sucralfate without removal of the ILPGT. He remained asymptomatic for 6 months before returning with two episodes of coffee-ground emesis and irritability. Repeat endoscopy showed mild improvement of the localized erosive gastritis. Case 4 At endoscopy, a circumscribed, nodular, erythematous area was noted on the posterior wall of the distal body opposite the tip of the G-tube (Fig. 3). Biopsies showed fundic mucosa with nonspecific chronic inflammation. The ILPGT was not removed. In addition, there was erosive distal esophagitis despite a previous fundoplication. Presently, the patient is receiving omeprazole and has had no further vomiting or hematemesis. Case 5 At endoscopy, the gastric mucosa of the lesser curve of the distal body of the stomach was adherent to the tip of the G-tube (Fig. 4). The surrounding mucosa showed multiple erosions in a circumscribed area. Because the patient's symptoms were thought to be due to a tight fundoplication, a lower esophageal myotomy and a partial (Thal) fundoplication were undertaken without removal of the ILPGT. The patient's cough improved, and there has been no recurrence of gastrointestinal symptoms despite the fact that he was not treated with H2 blockers or antacids. DISCUSSION Historically, G-tubes have long shafts and extend several inches exterior to the stoma. Low-profile G-tubes have been introduced in order to provide a less cumbersome and more cosmetically acceptable tube for feeding. There are currently two types of low-profile G-tubes: the button-like and the ILPGT: there are important differences between them (Table 2). The following mechanisms, singly or in combinations, could have been the cause of the gastric lesions described in our five patients. When not filled with air, fluid, or food, the lumen of the stomach is obliterated and the gastric walls are in apposition. With a G-tube in place on the greater curvature of the stomach or its anterior wall, the opposite wall is apposed onto the tip of the G-tube, which is “pointed” in the ILPGT. During episodes of increased intragastric pressure-for example, during retching, gagging, coughing or seizures (5)-there may be frictional forces and forceful movement of the stomach wall against the G-tube, resulting in injury and the development of ulcers and erosions. Hence, children with frequent retching and gagging, complicating fundoplication (6,7), and those with cough and seizures may be at increased risk for the development of ILPGT-associated gastric injuries. The size of the stomach may be another important factor determining ILPGT-induced gastric injury. The child in Case 1, the smallest child of the group, sustained the deepest injury, a chronic gastric ulcer. The larger children developed erosions and chronic inflammation. The position of the ILPGT may also be important. All of the lesions described in our patients occurred in the distal portion of the body or proximal antrum of the stomach, opposite the ILPGT placement site, where the lumen is narrow relative to the gastric body. This is supported by the observation that in Case 2, mucosal healing occurred after revision of the fundoplication, at which time the same ILPGT was positioned more proximally into the more capacious portion of the body of the stomach. Trauma during suction may be another factor that may explain the above observations. The endoscopic findings in Case 5 were striking. The gastric mucosa was firmly adherent to the tip of the ILPGT, which suggested suction-induced injury. The button-like G-tubes have two internal openings, which provide venting during aspiration, thereby decreasing the risk of suction injury. To prevent this kind of injury to the gastric mucosa, the manufacturers of ILPGTs advise that fluid should be infused back in order to release the adherent mucosa. We believe that the design of the ILPGT is an important factor that predisposes to gastric mucosal injury. Care should be taken during placement of ILPGT, particularly in young children and in children with chronic cough, seizures, retching, or gagging. In these children, the smallest length of ILPGT or a button-like G-tube should be used. We have shared our observations with the manufacturers of the ILPGT, and we understand that they are redesigning it. Acknowledgment: The authors thank Eric Hassal, M.D., and Juan Engel, M.D., for reviewing the manuscript, Michele Rekett and Sheetal Shah for assistance in preparing the manuscript, and Pentax Instrument Corp. for support with printing of color pictures.FIG. 1: . Single gastric ulcer, 6 mm (A), with the tip of the G-tube overlying the ulcer (B).FIG. 2: . Multiple gastric erosions in a circumscribed area.FIG. 3: . Multiple nodules with redness in a circumscribed area.FIG. 4: . Gastric mucosa adherent to the tip of the G-tube (A) and erosions around it (B).