Endoscopy of the Vertically Stapled Stomach

S SULLIVAN, R HOLLIDAY. Endocopy of the vertically stapled stomach. Can J Gastroenterol 1993;7(7):568-570. Over a five-year period, 145 morbidly obese patients were treated by vertical banded gastroplasty. ln a follow-up ranging from six months to five and one-half years, 29 developed upper gastrointestinal symptoms severe enough to require consultation and endoscopy. Vomiting of solid food was the most common symptom. Eight patients had stomal stenosis ( which was treated by endoscopic and bougie dilation in five, while three required gastrogastrostomy to relieve obstruction). In four patients the only endoscopic abnonnality was marked laxity of the lower esophageal sphincter which was confinned manometrically. Only one patient had Marlex mesh erosion.


PATIENTS AND METHODS
Between January 1985 and December 1989, 145 morbidly obese patients (20 males, 125 females) underwent VBG at the authors' institution. All operations were performed by one surgeon using the method described by Mason (1 ). A 50 to 60 mL vertical pouch was created using stapling instruments. The first 80 patients ha<l Marlex mesh placed about the lesser curve gastric stoma, which was ca librated with a 32 French Maloney bougie. The last 65 patients had the stoma reinforced with Gortex (WL Gore and Associates Ltd, United Kingdom). T he Gortex and Marlex mesh were premeasured (7 .5 cm by 1.5 cm) and secured with an overlap of approximately l cm to give a circumference of 5.5 cm. The rest of the technique was done in routine fashion. The patients were reviewed regularly and those who developed troublesome upper gasLrointestmal symptoms were referred for endoscopy.

RESULTS
Until June 1990, 29 patients developed gastrointestinal symptoms severe enough to require endoscopy. Twentyone (72%) presented with postprandial nausea and vomiting of solids. Other symptoms included abdominal pain (six, 20%), dysphagia (four, 14%), heartburn (four, 14%) and regurgitation without heartburn (five, 17%). These patients underwent 46 endoscopic procedures. The endoscopic findings are listed in Tahle 1. The most common significant finding was that of stomal stenosb (found 111 eight). These patiems underwent 19 dilations with an olympus GIF type Q 20 endoscope ( 11 mm) or scmirigid Savary-Gilliard bougies over a guidewire; the maximum <l1ameter bougie used was 17 mm, with most patients managed with dilation up to 14 mm. The symptoms of three pauents with stomal stenosis could nor be managed by repeated dilation, and they were treated with gastrogasrrostomy.
Only one patient had erosion of the Marlex mesh into the scoma. This patient had been taking ibuprofen. The ulceration healed after stopping ibuprofen and use of cimetidine, but patency of the stoma could nor be maintained and she required gastrogasrrostomy.
Six patients with postprandial vomiting had complete incompetence of the lower esophageal sphincter at endoscopy. Inflation of the esophagus with air 5 cm above the sphincter allowed one to look down to the stoma directly; there was no evidence of a functional sphincter. Four of these patients who had otherwise normal endoscopies without stomal srenosis underwent esophageal manometry with a pneumohydraulically perfused catheter system. All had weak lower esophageal sphincters with pressures of 5, 5, 8 and 10 mmHg. One patient with dysphagia had a tightly closed sphincter at endoscopy. She had esophageal manomerry and radionuclide esophageal transit studies, highly suggestive of early achalasia. Sixty-eight per cent (34 of 50) had stomal stenosis and 24% ( 12 of 50) had developed Marlex erosion into the stoma. Many of these patients had to be managed by gastrogastrostomy or conversion to a VBG. This unacceptably high rate of complications in HBG has been recognized by others (4).
Accordingly, in 1985 the current authors changed their method of surgical management of morbid obesity to VBG. Compared with HBO, we are very pleased with the weight loss and the low rate of complications with VBG (as are others (2,(5)(6)(7)(8)(9)(10)(11)(12)). In the presented pauent population, only one patient developed Marlex mesh erosion, but this patient was also taking ibuprofen, a medication with known ulcerogenic potential. This low rate of mesh or ring erosion m VBG 1s wellrecognized, (5,(7)(8)(9)13,14). However, during the current study one patient who had been operated on at another institution and developed stomal ulceration and mesh erosion while being treated with 40 mg omeprazole daily for esophagitis was seen. This dose of omeprazole virtually abolishes gastric acid secretion and had healed his esophagi tis. He was treated by gastrogastroscomy but others have described the endoscopic removal of eroded mesh (15). Clearly, acid is not needed for mesh erosion.

CONCLUSIONS
The symptoms and endoscopic findings of the current patients are very simi lar to those recently reported by Wayman (17). However, an unexpected endoscopic finding which has not been mentioned by others was marked incompetence of the lower CAN J GA~TROENTEROL VOL 7 NO 7 SEPTEMBER/Oc'TOBER 1993  esophageal sphincter 111 four patients with vomiting and regurgitation who had otherwise normal endoscopies. Weakness of the lower esophageal sphincter was confi rmed manomcrrically. In spite of their massive obesity none of these patients had troublesome symptoms of gastroesophagcal reflux prior to VBG, raising the possibility that overeating and distension of the pouch and d istal esophagus by food and fluid might lead to incompetence of the lower esophageal sphincter. This possibility would be worth investigating prospectively by a preoperative esophageal manometry and follow-up manometry in patients who developed symptom\ of regurgitation postoperatively.