Evidence-Based Treatment of Acute Pancreatitis: Antibiotic Prophylaxis in Necrotizing Pancreatitis

To the Editor:

With interest we read the article by Heinrich et al.¹ The authors have to be congratulated for their attempt to review the literature in acute pancreatitis in a rigorous “evidence-based” manner. We agree with the authors that there is a clear need for further clinical trials on acute pancreatitis. For this reason, in 2004, the Dutch Acute Pancreatitis Study Group embarked on a randomized double-blind placebo-controlled multicenter trial on probiotic prophylaxis in patients with predicted severe acute pancreatitis.² Recently, our group initiated a second randomized controlled multicenter trial in 20 hospitals in which patients with infected necrotizing pancreatitis are randomized to single necrosectomy with postoperative lavage versus a minimally invasive “step-up approach,” exactly the trial design suggested by Heinrich et al.³ Concerning the author's recommendations on antibiotic prophylaxis, it should be noted that these may need tempering.

Based on the same literature as reviewed by Heinrich et al, an international group of renowned pancreatologists in 2004 recommended against the routine use of prophylactic systemic antibacterial or antifungal agents in patients with necrotizing pancreatitis. The main reason for this was the inconclusive evidence and divided expert opinion.⁴ The 2005 U.K. guidelines for the management of acute pancreatitis stated: “The evidence to enable a recommendation about antibiotic prophylaxis against infection of pancreatic necrosis is conflicting and difficult to interpret. Some trials show benefit, others do not. At present there is no consensus on this issue… (recommendation grade C)”.⁵ Furthermore, we recently analyzed the methodological quality of the randomized controlled trials on systemic antibiotic prophylaxis in acute pancreatitis based on a previously published scoring system.⁶ We demonstrated that the better the trial, the less impact of antibiotic prophylaxis on mortality was observed.⁷ Therefore, it is surprising that Heinrich et al conclude that there is “level A” evidence that imipenem or meropenem prophylaxis decreases the risk of infected necrosis and mortality. In their analysis, the authors excluded the mortality data of the Isenmann trial and added a particular subgroup analysis. In our opinion, any argument to exclude all mortality data from methodologically the best trial yet can hardly be considered “evidence based.” Furthermore, the authors state that they excluded all non-English studies but did include the Schwarz et al trial,⁸ published only in German, which favors antibiotic prophylaxis. In contrast, the authors did not include 2 randomized controlled Czech trials of Spicak et al, both of which failed to show an effect of antibiotic prophylaxis.^9,10

In addition, we would like to note that Dellinger et al recently presented the results of a randomized, double-blind, placebo-controlled, multicenter trial with meropenem prophylaxis in 100 patients with severe acute pancreatitis, including 57 patients with >30% pancreatic necrosis. This trial failed to demonstrate a beneficial effect of meropenem prophylaxis on infection of pancreatic necrosis and mortality.¹¹ Maybe that with the publication of this well-designed study the discussion on antibiotic prophylaxis in necrotizing pancreatitis can finally be closed. Nevertheless, we would like to encourage (national) multicenter groups to embark on well-designed, randomized controlled trials studying promising prophylactic or surgical strategies to unequivocally shift treatment paradigms in acute pancreatitis.

Marc G. H. Besselink, MD
Hjalmar C. van Santvoort, MD
Erik Buskens MD, PhD
Hein G. Gooszen, MD, PhD
Dutch Acute Pancreatitis Study Group
University Medical Center Utrecht
Utrecht, The Netherlands
m.besselink@umcutrecht.nl