Pancreaticoduodenostomy for treatment of giant duodenal ulcer.

Definitive acid-reducing peptic ulcer surgery currently plays a less prominent role than medical treatment in the management of peptic ulceration. The utilisation of proton pump inhibitors and Helicobacter eradication are proven effective therapies. Giant duodenal ulcers, first described by Brdiczka in 1931, 3 pose a unique challenge to this shift in therapy. There are various described surgical options to deal with these ulcers depending on the operative findings and we describe a simple new technique that we recently used successfully. A 48-year-old man in full-time employment was noted to use alcohol and cigarettes excessively. He initially presented with a very strong dyspeptic history on 28 January 1999. He had severe epigastric pain radiating to the back and melaena. His haemoglobin concentration on 7 December 2000 was 3.9 g/dl. Endoscopic examination on 8 December showed a normal oesophagus and stomach and an oedematous fold of the mucosa in the duodenal bulb, with a superficial ulcer on it that was not bleeding. He had been on proton pump inhibitors and antimicrobial therapy since 18 May 2000. The current admission on 28 March 2002 was similar, with pain, vomiting and melaena and epigastric tenderness. At endoscopy the stomach was found to be full of fresh blood and he suddenly collapsed. He was resuscitated and rushed to the operating theatre. At laparotomy a giant duodenal ulcer was found with destruction of the bulb and proximal second part of the duodenum. The gastroduodenal artery and its branches were bleeding profusely, and were suture-ligated. The base of the ulcer was formed by the superior and anterior surface of the head of the pancreas. Duodenal destruction spared the papilla-bearing medial wall of the second part of the duodenum. The omentum, transverse colon, liver edge and small bowel concealed a large anterior perforation. Bilateral truncal vagotomy and antrectomy were performed (Fig. 1). Gastro-intestinal continuity was completed with a pancre-aticoduodenostomy (PD) with invagination of the medial surface of the pancreatic head into the freshly debrided distal second part of the duodenal end. A transduodenal T-tube bil-iary-enteric drain was left in situ (Fig. 2) and a retrocolic gas-trojejunostomy was done. A Penrose drain was left in Morrison's pouch. A postoperative T-tube cholangiogram on day 10 showed a patent biliary system and an intact PD (Fig. 3). This was clamped and removed after a further 24 hours without any complications. A barium meal examination showed an intact gastrojejunos-tomy and PD (Fig. …