Epidural anesthesia improves pancreatic perfusion and decreases the severity of acute pancreatitis

Abstract

Aim: To study the safety of epidural anesthesia (EA), its effect on pancreatic perfusion and the outcome of patients with acute pancreatitis (AP).

Methods: From 2005 to August 2010, patients with predicted severe AP [Ranson score ≥ 2, C-reactive protein > 100 or necrosis on computed tomography (CT)] were prospectively randomized to either a group receiving EA or a control group treated by patient controlled intravenous analgesia. Pain management was evaluated in the two groups every eight hours using the visual analog pain scale (VAS). Parameters for clinical severity such as length of hospital stay, use of antibiotics, admission to the intensive care unit, radiological/clinical complications and the need for surgical necrosectomy including biochemical data were recorded. A CT scan using a perfusion protocol was performed on admission and at 72 h to evaluate pancreatic blood flow. A significant variation in blood flow was defined as a 20% difference in pancreatic perfusion between admission and 72 h and was measured in the head, body and tail of the pancreas.

Results: We enrolled 35 patients. Thirteen were randomized to the EA group and 22 to the control group. There were no differences in demographic characteristics between the two groups. The Balthazar radiological severity score on admission was higher in the EA group than in the control group (mean score 4.15 ± 2.54 vs 3.38 ± 1.75, respectively, P = 0.347) and the median Ranson scores were 3.4 and 2.7 respectively (P = NS). The median duration of EA was 5.7 d, and no complications of the epidural procedure were reported. An improvement in perfusion of the pancreas was observed in 13/30 (43%) of measurements in the EA group vs 2/27 (7%) in the control group (P = 0.0025). Necrosectomy was performed in 1/13 patients in the EA group vs 4/22 patients in the control group (P = 0.63). The VAS improved during the first ten days in the EA group compared to the control group (0.2 vs 2.33, P = 0.034 at 10 d). Length of stay and mortality were not statistically different between the 2 groups (26 d vs 30 d, P = 0.65, and 0% for both respectively).

Conclusion: Our study demonstrates that EA increases arterial perfusion of the pancreas and improves the clinical outcome of patients with AP.

Keywords: Computed tomography; Epidural anesthesia; Pancreatic necrosectomy; Pancreatic perfusion; Severe acute pancreatitis.

Figure 1

CONSORT diagram showing randomization and allocation of the study cohort.

Figure 2

Perfusion computed tomography color map and standard computed tomography of the pancreas obtained in a 75-year-old woman admitted for acute severe pancreatitis, randomized to group 1 (epidural anesthesia). A: Axial image obtained at admission shows a positive enhancement integral (PEI) value of 15.9 in the pancreatic body; B: Axial image obtained with control computed tomography (CT), 72 h after admission, shows a 29% improvement of the perfusion in the pancreatic body (PEI of 20.5) when compared to admission values; C: Standard CT axial oblique image at the level of the pancreas obtained on admission, during portal phase.

Figure 3

Perfusion computed tomography color map and standard computed tomography of the pancreas obtained in a 73-year-old man admitted for acute severe pancreatitis, randomized to group 2 (control group). A: Axial image obtained on admission shows a positive enhancement integral (PEI) value of 16.5 in the pancreatic head; B: Axial image obtained with control computed tomography (CT), 72 h after admission, shows a 53% impairment of the perfusion in the pancreatic head (PEI of 7.8) when compared to admission values; C: Standard CT axial oblique image, at the level of the pancreas obtained on admission, during portal phase.

Figure 4

Changes in pancreatic perfusion measurements at 72 h compared to the measurements. On admission in the epidural anesthesia (EA) and control group (significant increase, decrease or no change in perfusion). ^aP = 0.0025 vs control.