Infected Peri-Pancreatic Necrosis Causing Gallbladder Necrosis by Direct Extension

Acute acalculous cholecystitis may develop in patients suffering from necrotizing pancreatitis. Conversely, acute pancreatitis may complicate acute gallbladder disease. We present a case that lends support to the existence of another possibility: gallbladder necrosis caused by direct extension of the necrotizing pancreatitic process.


INTRODUCTION
Acute acalculous cholecystitis is well known to be associated with critical diseases and to develop in patients suffering from burns, trauma, receiving parenteral nutrition, following major operations or any debilitating state I. Necrotizing pancreatitis and its infected complications constitute a severe disease which require extensive surgical treatment2. Obviously, acalculous cholecystitis can also develop in acute pancreatitis patients. Conversely, acute pancreatitis is known to occur following any surgical procedure or critical disease3. In fact, the mechanisms which initiate the necrotizing process in both acute pancreatitis and acalculous cholecystitis could be identical4. Consequently, the association between the two entities, acalculous cholecystitis developing in patients suffering from acute pancreatitis, or the latter complicating the former, has been reported sporadically1'5.
The other possibility, that necrosis of the gallbladder may occur due to direct extension of the necrotizing peri-pancreatic process, is not mentioned in the literature. The following case supports our contention that this may indeed occur.

CASE REPORT
A 62 years old male patient was admitted with severe acute alcoholic pancreatitis (APACHE II -17). Contrast enhanced CT scan was performed during the second week of the disease and disclosed patchy pancreatic uptake and a large subhepatic Address correspondence to: M. Schein collection. Percutaneus drainage revealed necrotis material and pus. Concomitantly the patient developed massive upper gastrointestinal haemmorhage; endoscopy revealed chronic duodenal ulcer with an adherent clot. Haemorrhage persisted and laparotomy was undertaken. At operation the bleeding ulcer was underrun and truncal vagotomy/pyloroplasty performed. The pancreas appeared hard and viable but there was an extensive process of infected peripancreatic necrosis occupying both gutters and extending into the duodenohepatic ligament, up the porta hepatis. The gallbladder was adherent to the involved duodenohepatic ligament; it was necrotic at this point and perforated with a free subheptic collection of bile. Because of the obliteration in Calot's triangle partial cholecystectomy was performed6. Brisk bleeding from the edges of the resected gallbladder attested to the patency of the cystic artery. The necrotic process was debrided and drained. Postoperatively the abdominal infection persisted, complicated by leakage from the closure of the pyloroplasty, and multi-organ failure. Subsequently the abdominal cavity was managed with "laparostomy" and 12 planned re-laparotomies were necessary to reverse the septic process. The duodenal fistula closed spontaneously and he was discharged after 7 weeks of hospitalization. Examination of the gallbadder specimen disclosed an oedematous and thickened wall with areas of necrosis and perforation. There were yellow spots of fat necrosis. Histological examination revealed acute, necrotizing cholecystitis. There was evidence of fat necrosis and calcification in the outer layers of the wall.

DISCUSSION
Numerous publications have documented that the necrotizing pancreatic process may involve adjacent visceral structures. Colonic necrosis is the most frequently described example but small-bowel and gastric involvement are less well recognized conditions. Miller et al. have reported a patient in whom pancreatic necrosis eroded the intrapancreatic portion of the common bile duct1. If most juxtapancreatic viscera can be affected by the pancreatitic necrotizing process why not the gallbladder?Wwe believe that this, in fact, happened in our patient.
Various theories have been proposed to explain the associated necrosis of juxtapancreatic organs7. Firstly, it may be involved by direct contact with a highly septic exudate, rich in vasoactive substances and enzymes. Lending support to this hypothesis is the histologic observation (shown also in the present case) that the external layers of the viscus are more affected than the inner layers, as necrosis spreads "from outside to the inside". Secondly, thrombosis of supplying arteries or veins is another possible mechanism. Finally, shock and the occurrence of a lowflw state may account for ischaemia and subsequent gangrene.
Previously we have hypothesized that acute pancreatitis may adversely affect the healing of adjacent intestinal anastomoses3. The fact that the present case developed a leak from a pyloroplasty suture line, bathed in necrotizing pancreatic secretions, lends credibility to that view.
We conclude that the necrotizing pancreatic process may involve, by direct extension, the nearby gallbladder as well as any adjacent organ.