Acute pancreatitis is the inflammation of the pancreas, a serious emergency with no specific treatment. The pancreas, a digestive gland, can become inflamed for many reasons, but mainly as a complication from gallstones or excess alcohol intake. If severe, the pancreas may lose its blood supply, a complication called pancreatic necrosis that can be detected by computed tomography (CT) scanning. Death can occur either early in the disease process in association with uncontrolled inflammatory responses, causing multiple organ-system failure (MOSF), or late when the necrotic tissue becomes infected, which might necessitate major surgery to remove the infection, with the risk of death rising from 10% to over 40%. Antibiotics may prevent later infection and reduce the risk of death, but could also encourage bacterial antibiotic resistance and fungal infections. Controlled trials looking at the value of using prophylactic antibiotics have produced conflicting results.
This review aims to determine the effectiveness and safety of prophylactic antibiotics in CT-proven necrotising acute pancreatitis. A previous version published in 2006 suggested a survival advantage overall, and a decrease in pancreatic infections for some types of antibiotic therapy (beta-lactam antibiotics). Since that review, two further studies have been published: both were double-blinded, randomised, clinical trials (RCTs). These studies have now been included and our conclusions have changed as a result.
In the current review, data were found and analysed from 7 trials involving 404 patients randomly allocated to receive antibiotics or placebo. Although death occurred less after antibiotics (8.4%) than placebo (14.4%), as did infected pancreatic necrosis (19.7% versus 24.4%) and other infections (23.7% versus 36%), the differences were not statistically significant and so genuine benefit cannot be confirmed. There were no major problems with antibiotic resistance, and fungal infections were similar (3.9% versus 5%). The quality of studies was variable and only two were ‘blinded’, whereby investigators and patients were unaware of which treatment patients received. Many different regimens were used, and of the two main types of antibiotics used, a beta-lactam appeared to work better. Only one type of antibiotic (imipenem) was considered on its own, showing a significant decrease in infection of the pancreatic necrosis.
Although we cannot confirm benefit from the use of prophylactic antibiotics in this condition, consistent trends towards a beneficial effect nevertheless remain. Further, better designed studies, ideally with beta-lactam antibiotics, are required.