Gram-Negative Intestinal Colonization, Acute Pancreatitis, and Pancreatic Infection


Synopsis: Gram-negative intestinal colonization, E. coli excepted, entails a significantly increased risk of pancreatic infections and mortality in patients with severe acute pancreatitis.

Source: Luiten EJT, et al. Prognostic importance of gram-negative intestinal colonization preceding pancreatic infection in severe acute pancreatitis. Intensive Care Med 1998; 24:438-445.

The development of gram-negative pancreatic infection significantly increases the risk of mortality in patients with severe acute pancreatitis. The gastrointestinal tract seems to be the major source of organisms that eventually infect the necrotic pancreas, and several experimental studies have suggested possible routes of translocation. In this light, severe acute pancreatitis could be the ideal indication for selective decontamination of the digestive tract (SDD).1

The role of SDD in severe acute pancreatitis was evaluated by Luiten and colleagues in a randomized, controlled, multicenter trial published in 1995.2 Luiten et al now report the analysis of the results of systematic semi-quantitative cultures from the oropharynx, rectum, and pancreatic tissues taken during the aforementioned study. Surveillance cultures from the oropharynx and rectum were taken on admission and repeated twice weekly. Fine needle aspiration guided by ultrasound or computed tomography with subsequent culture was performed if there was clinical suspicion of infected pancreatic necrosis. Surveillance cultures of the peri-pancreatic devitalized tissues were obtained at each repeat laparotomy, as well as from drainage. Follow-up was continued either until death or the risk of development of a pancreatic infection was considered negligible.

Of the 102 patients studied, 12 were excluded from the analysis because the intestinal surveillance cultures taken were insufficient. Fifteen (17%) of the 90 patients who were analyzed developed a gram-negative pancreatic infection during the course of the disease. Of the 15, three of 49 were receiving adjuvant SDD, and 12 of 41 were in the untreated control group. In each patient, gram-negative pancreatic infection (n = 15) was preceded by intestinal colonization with the same microorganisms. The majority of these infections (73%) occurred within a week of the first isolation from the digestive tract.

Of 48 patients in whom either intestinal colonization did not occur or in whom E. coli was the only microorganism cultured during the course of the disease, none developed a gram-negative pancreatic infection. However, of the 42 patients in whom gram-negative intestinal colonization occurred (with or without additional E. coli), 36% developed pancreatic infection. Gram-negative intestinal colonization was thus associated with a 3.7-fold (P = 0.004) increased mortality risk.


Despite improvements in the supportive care of patients with acute pancreatitis, the morbidity and mortality of infectious complications remain high. Patients with acute necrotizing pancreatitis are at the greatest risk of developing an infection with enteric gram-negative bacteria translocated from the bowel lumen.

The role of prophylactic antibiotics in reducing infectious morbidity and mortality has been debated for decades. The incidence of pancreatic sepsis was reduced in patients with necrotizing pancreatitis demonstrated by computed tomography who received imipenem prophylactically (12.2% vs 30.3%, P < 0.01) but the overall mortality rate was not affected.3 More recently, selective decontamination of the digestive tract significantly reduced both morbidity and mortality (22% vs 35%, P < 0.05) in patients with severe acute necrotizing pancreatitis.2 However, all patients in that study received both intravenous cefotaxime and the oral SDD regimen, and it is still unclear whether the intravenous agent alone may suffice. Moreover, concern remains regarding the possibility that prophylactic antibiotic regimens may eventually lead to problematic infections by organisms resistant to multiple antibiotics.

This study by Luiten and colleagues shows that the appearance of intestinal colonization with gram-negative microorganisms, other than E. coli, is an early, ominous, predictive sign indicating that a gram-negative pancreatic infection can probably be expected within a week of the first intestinal occurrence. It adds important information to this controversial issue by suggesting a new way to identify patients with an increased risk of pancreatic infections, in whom increased efforts to free the digestive tract of nosocomial gram-negative microorganism seem warranted. (Dr. Baigorri is Servei Medicina Intensiva, Consorci Hospitalari del Parc Tauli, Barcelona, Spain.)


    1. Nathens AB, Rotstein OD. Selective decontamination of the digestive tract in acute severe pancreatitis-An indication whose time has come. Clin Infect Dis 1997; 25:817-818.

    2. Luiten EJT, et al. Controlled clinical trial of selective decontamination for the treatment of severe acute pancreatitis. Ann Surg 1995;222:57-65.

    3. Pederzoli P, et al. A randomized, multicenter clinical trial of antibiotic prophylaxis of septic complications in acute necrotizing pancreatitis with imipenem. Surg Gynecol Obstet 1993;176:480-483.