Effect of Infection on Outcome in Acute Pancreatitis

Abstract & Commentary

Synopsis: Demonstration of pancreatic necrosis by computed tomography identifies a subgroup of patients with the most severe form of acute pancreatitis. Among patients with necrotizing pancreatitis, this study found that the presence of infection in the pancreatic bed increased the overall mortality rate several-fold.

Source: Buchler MW, et al. Ann Surg 2000;232(5): 619-626.

Infection of the necrotic tissue, likely secondary to bacterial translocation from the colon, is a complication associated with a marked increase in mortality, according to most series, and may represent an indication for surgical debridement of the pancreas and retroperitoneal tissues. In a prospective cohort, Buchler and colleagues describe their treatment approach for acute pancreatitis, which relies heavily upon surgical therapy when infected pancreatic necrosis is diagnosed.

Buchler et al describe a series of 204 patients with acute pancreatitis admitted to their hospital in Berne, Switzerland, between January 1994 and June 1999. Of these patients, 86 were found to have acute necrotizing pancreatitis (ANP), based upon CT scan findings. By various indices (Apache II, Ranson’s criteria), these 86 patients had more severe systemic disease than the 188 patients with pancreatitis without necrosis. They had a longer hospital stay (44 days vs 13 days) and a higher mortality (10% vs 0%). Twenty-nine (34%) were diagnosed with infected pancreatic necrosis, based upon fine needle aspiration in 28 and post-mortem examination in one patient (despite 3 sterile, fine-needle aspirates in this patient prior to death).

Additional details of Buchler et al’s therapeutic strategy are presented in the manuscript, but this summary addresses where surgery fits into their treatment armamentarium. The treatment protocol involved surgical decision-making, based primarily upon the results of percutaneous (CT-guided) fine-needle aspiration and culture of the necrotic pancreatic tissue. Growth of organisms mandated surgery, which included resection of necrotic pancreas, followed by post-operative continuous lavage.

In patients with infected necrosis, the case-fatality rate was 24%, compared to 3.5% for patients with presumed sterile necrosis. One of the patients presumed to have sterile necrosis died and was found to have infected necrosis at autopsy. Therefore, 27 of 28 patients (96%) with infected necrosis were diagnosed on the basis of fine-needle aspiration and culture. One of 16 patients with culture-negative fine-needle aspirates was found to have infected necrosis at autopsy. The remaining 15 patients with culture-negative aspirates lived and did not appear to have pancreatic infection based upon clinical recovery. However, we cannot be certain that infection was absent, and simply responded to conservative (intravenous antibiotic) management. Similarly, the remaining 41 patients considered to have sterile pancreatic necrosis did not undergo fine-needle aspiration, and therefore the true status of the pancreas was unknown in these individuals. One of these patients died from progressive organ failure and was found to have sterile necrosis at autopsy.

Comment by Grant E. O’Keefe, MD

Acute pancreatitis is most often a mild-to-moderate and self-limited disease that requires supportive therapy and subsequent treatment of the underlying cause (cholecystectomy for biliary lithiasis, etc). However, severe inflammation may lead to necrosis of the pancreas and adjacent retroperitoneal tissues, local infection, and the systemic complications of multiple organ dysfunction (MOD). The combination of clinical and radiological (computed tomographic) criteria identifies the majority of patients with severe acute pancreatitis. The demonstration of pancreatic necrosis by CT identifies a subgroup of patients with the severest form (ANP); patients who present a great therapeutic challenge and represent the majority of fatalities. Numerous therapeutic interventions have been tried and various treatment guidelines proposed, but there are few compelling data upon which to base recommendations.

Surgery has been considered an important component in the treatment of ANP, although the indications and timing of operative therapy are controversial. Early surgical intervention, including debridement of necrotic tissue has been considered by some to be an important component of treatment. However, surgery is not without complications, and the benefits are not clear or widely accepted.

The findings of this study support the concept of a detrimental effect of infected pancreatic necrosis upon outcome, and the potential use of fine-needle aspiration in directing surgical management. However, there are a number of major limitations that physicians and surgeons caring for these patients must consider. We have no knowledge of the infection status of the 41 patients who did not undergo fine-needle aspiration. Therefore, we must consider that at least some of these patients had infections, which responded to antibiotic therapy and resolved without surgery. We also know that one patient died from the ravages of severe, ongoing systemic inflammation and organ failure despite having "only" sterile necrosis.

It is also critical to note that 12 of the 28 patients underwent more than one fine-needle aspiration. Therefore, only 16 (57%) of the initial aspirates were positive, suggesting that the actual sensitivity may be much lower than the reported 96%. Patients with an ultimately positive FNA and with positive intraoperative cultures had a 24% case-fatality rate. As we do not have a comparison group, with known infected necrosis, it is not certain whether surgery effects the course of ANP once infection is established. Conversely, we cannot be certain that waiting for the diagnosis of infected necrosis before considering operative therapy can be universally applied. Would earlier surgery avert the progression to infection and the systemic complications?

This study raises many of the important issues faced when caring for these complicated, critically ill patients, in whom surgery must be considered in the context of the other supportive therapies. While not definitively answering the questions of: 1) who requires surgery; 2) when should surgery be done; and 3) what procedure should be used, this study provides a rational framework for the consideration of these questions in the management of patients with acute necrotizing pancreatitis.