From Pages to Practice
Published November 30, 2022
The pancreas is a notoriously complex endocrine organ and when inflamed can cause a wide range of clinical symptoms, ranging from simple epigastric pain, nausea, and vomiting, all the way to circulatory collapse and acute respiratory distress syndrome (ARDS). As the old saying goes, "eat when you can, sleep when you can, and don’t mess with the pancreas." More than three centuries have passed since the Dutch anatomist Nicholas Tulp reported the first case of pancreatitis in 1652, and the therapeutic approach remained largely the same until recently.
Conventional wisdom for the management of pancreatitis has relied heavily on the two-pronged approach of prolonged bowel rest by restricting oral intake and aggressive intravenous hydration, in the hope of "flushing out" the system. However, the management of pancreatic inflammation, or pancreatitis, has come under new scrutiny recently since early feeding with a low-fat diet was shown to reduce length of hospitalization, as compared with a prolonged NPO course. Now, investigators have turned their attention to a fluid-resuscitation strategy.
In the recent open-label WATERFALL trial published in NEJM, the ERICA consortium of investigators compared the safety and efficacy of aggressive versus moderate fluid resuscitation in patients with pancreatitis. Patients diagnosed with acute pancreatitis were randomized to receive either aggressive fluids (20 mL/kg lactated Ringer’s bolus over 2 hours followed by a 3 mL/kg/hr infusion) or moderate fluids (10 mL/kg bolus only if hypovolemic, followed by a 1.5 mL/kg/hr infusion). Laboratory studies and physical exam assessments were made at 12, 24, 48 and 72 hours after the initial assessment at 3 hours after initiation of fluids. Fluid resuscitation could be stopped after patients were able to tolerate oral feeding for 8 hours. In both groups, the bedside clinician could adjust or stop hydration if the patient appeared to be developing volume overload.
During the first 48 hours, patients in the aggressive-fluids group received a median of 7.8 liters of lactated Ringer’s solution, versus 5.5 liters in the moderate-fluids group. The trial was stopped because 20% of patients in the aggressive-fluids group developed signs of volume overload (the primary safety outcome), versus 6% in the moderate-fluids group. No significant difference was reported between the two groups for the primary efficacy outcome of progression from early pancreatitis to severe pancreatitis. Patients in the aggressive-fluids group were hospitalized for a median of 1 day longer than those in the moderate-fluids group. The investigators thus concluded that an aggressive-fluid strategy increased the risk of harm to patients, as compared to a moderate-fluid strategy, without a significant difference in the risk of developing severe pancreatitis.
As our understanding of the inflammatory response in pancreatitis evolves, so too does our awareness of the potential risks of each intervention. These data will help refine the treatment approach to pancreatitis to minimize complications while still adequately managing the disease process. One notable challenge of the trial was the open-label design. Because fluids were continuously administered at the bedside, the investigators felt there was no way to safely blind clinicians to fluid strategy. An area for future study involves understanding the best approach for hydration in patients excluded from the trial, including patients with severe pancreatitis, congestive heart failure, uncontrolled electrolyte derangements, chronic pancreatitis, chronic renal failure, and decompensated cirrhosis.
The following NEJM Journal Watch summary provides more details of the study.
Rahul B. Ganatra, MD, MPH, reviewing Ramírez-Maldonado E et al. Ann Surg 2021 Aug
Based on moderate-quality evidence from several previous clinical trials, the American Gastroenterological Association recommends that oral feeding be started within 24 hours of admission for acute pancreatitis to lower complications (Gastroenterology 2018; 154:1096. opens in new tab). In this latest randomized trial of 131 patients hospitalized with mild or moderate acute pancreatitis, researchers at four hospitals in Spain compared immediate refeeding and conventional refeeding (i.e., feeding delayed until abdominal pain and inflammatory markers improved).
Mean length of stay (LOS) was significantly shorter in the immediate-refeeding group than in the conventional-refeeding group (3.4 vs. 8.8 days). Although pain initially was rated higher in the immediate-refeeding group (on the day oral feeding started), more immediate-refeeding patients tolerated a diet (99% vs. 79%), and fewer required opioids (0% vs. 8.3%) or experienced complications (4.2% vs. 18.3%) or progression to severe disease (0% vs. 10%) compared with patients in the conventional-refeeding group.
Comment: Reductions in LOS with early feeding that have been reported in other studies were substantially less than the 5-day reduction noted in this study. This difference could be due to the stringent criteria used for feeding in this study's conventional group, which might have delayed feeding unnecessarily and biased the results toward a bigger difference in LOS. Information on the type and quantity of intravenous fluid used for resuscitation was not provided, which could also affect these outcomes. Nevertheless, this study illustrates the harms associated with delayed refeeding and supports current professional society recommendations.