Pancreatic infection is the leading cause of death from acute pancreatitis. Patients with severe necrotizing pancreatitis are most at risk. Early computed tomography and percutaneous fine-needle aspiration microbiology of areas of pancreatic necrosis enable early diagnosis. Pancreatic infection should be treated surgically, although sterile necrosis may be managed conservatively. The role of antimicrobial drugs is uncertain.

The routes of spread of pathogens into the pancreas in acute pancreatitis were investigated. Four experiments were performed: (1) cats with and without acute pancreatitis were given 107 Escherichia coli (E colt) intravenously, (2) in cats with acute pancreatitis 108 E coli was placed in the colon. In half of them the colon was then enclosed in an impermeable bag to prevent transmural spread. (3) E coli (104) was placed in the pancreatic duct in cats with and without acute pancreatitis. (4) In cats with acute pancreatitis 105 E coli was placed in the gall bladder. In half of them the common bile duct was ligated to prevent biliary-pancreatic reflux. After 24 hours, intravenous E coli infected the pancreas in six ofnine cats with acute pancreatitis and three of 10 controls. After 72 hours E coli spread to the pancreas from the colon in six ofnine cats with acute pancreatitis. This was prevented by enclosing the colon in an impermeable bag (p=0.02). In five of six cats with acute pancreatitis and five of six controls E coli placed in the pancreatic duct colonised the pancreas within 24 hours. Pancreatic colonisation from the gall bladder occurred in five ofsix cats with a patent common bile duct and in three of six with an obstructed common bile duct. In conclusion, in cats E coli can spread to the pancreas by the blood stream, transmurally from the colon, and by reflux into the pancreatic duct. (Gut 1994; 35: 1306-1310 and only a brief account will be given here. Fasted adult mongrel cats were used in all experiments. No antibiotics were given at any stage. Cats were anaesthetised with xylazine hydrochloride (1 mg/kg body weight, intramuscular) and sodium pentobarbital (25 mg/kg body weight, intraperitoneal). All operations were performed using an aseptic technique and infusions were made through in line bacterial filters (02 ,um pore size).A femoral vein was cannulated with a heparinised plastic catheter (external diameter 0965 mm). The abdomen was entered and the pancreatic duct was cannulated in the tail of the gland with a plastic catheter (external diameter 061 mm) primed with standard perfusate. Acute pancreatitis was induced by perfusing 0.5 ml of 7.5 mM glycodeoxycholic acid into the pancreatic duct over one hour. Then 0.5 ml of pooled pancreatic juice (activated by prior incubation with enterokinase) was perfused over a second hour. For two hours while the pancreatic duct was being perfused 16,16 dimethyl prostaglandin E2 (2 ,g/kg body weight/hour) was infused intravenously to cause the development of acute pancreatitis. A similar procedure was performed in control animals except the pancreatic duct was perfused with standard perfusate for two hours and 09% isotonic saline was infused intravenously instead of prostaglandin E2. Cats were then allowed to recover. Reoperation was performed on anaesthetised animals to remove tissues for microbiology. In previous experiments we showed that pathogens may spread to the pancreas from the colon, gall bladder, and kidney.3 The only possible route of the s...

Chronic pancreatitis is characterized by persistent and severe pain, which can be relieved by decompression of the main pancreatic duct (MPD). Both ductal and interstitial pressures have been shown to be increased in chronic pancreatitis in patients. A study was carried out of pancreatic interstitial pressure and pancreatic blood flow in normal cats and those in which chronic obstructive pancreatitis had been induced 5 weeks earlier to determine the effect of decompression of the MPD. In the normal pancreas, median(interquartile range (i.q.r.)) basal interstitial pressure was 0.05(1.2) mmHg and median(i.q.r.) basal pancreatic blood flow 58.3(24.3) ml per min per 100 g. Secretory stimulation did not change the interstitial pressure significantly, but was associated with a 40 per cent increase in median(i.q.r.) blood flow to 81.8(45.8) ml per min per 100 g. In contrast, in chronic obstructive pancreatitis, the median(i.q.r.) basal interstitial pressure was 2.0(1.5) mmHg, which was significantly higher than in the normal gland, and median(i.q.r.) pancreatic blood flow was 38.3(9.8) ml per min per 100 g, significantly lower than in the normal pancreas. Furthermore, secretory stimulation was associated with a significant increase in median(i.q.r.) interstitial pressure to 3.3(1.6) mmHg and a simultaneous decrease in median(i.q.r.) blood flow to 31.5(13.7) ml per min per 100 g. After decompression of the MPD in cats with chronic obstructive pancreatitis, the median(i.q.r.) basal interstitial pressure was 2.0(1.4) mmHg and on secretory stimulation 1.8(1.5) mmHg. Decompression thus prevented the increase in interstitial pressure seen in the animals with obstruction. In contrast, ductal decompression improved the median(i.q.r.) basal pancreatic blood flow to 45.9(38.4) ml per min per 100 g and, furthermore, this increased significantly on secretory stimulation to a median(i.q.r.) of 81.4(47.8) ml per min per 100 g. Decompression thus restored the normal pattern of secretory hyperaemia. Within the confines of this model, these observations demonstrate that chronic obstructive pancreatitis exhibits a compartment syndrome that is relieved by duct drainage.

Based on these findings, unrecognized disruption of small artery collaterals during colorectal resection might be implicated in anastomotic leak and in particular might explain the higher leak rate in low anterior resection.