Continuous regional arterial infusion (CRAI) of protease inhibitor nafamostat mesilate (NM) is used in the context of predicted severe acute pancreatitis (SAP) to prevent the development of pancreatic necrosis. Although this therapy is well known in Japan, its efficacy and safety remain unclear.

This investigator-initiated and -driven, multicenter, open-label, randomized, controlled trial (UMIN000020868) enrolled 39 patients with predicted SAP and low enhancement of the pancreatic parenchyma on computed tomography (CT). Twenty patients were assigned to the CRAI group, while 19 served as controls and were administered NM at the same dose intravenously (IV group). The primary endpoint was the development of pancreatic necrosis as determined by CT on Day 14, judged by blinded central review.

There was no difference between the CRAI and IV groups regarding the percentages of participants who developed pancreatic necrosis (more than 1/3 of the pancreas: 25.0%, range 8.7–49.1% vs. 15.8%, range 3.4–39.6%, respectively, P = 0.694; more than 2/3 of the pancreas: 20%, range 5.7–43.7% vs. 5.3%, range 0.1–26.0%, respectively, P = 0.341). The early analgesic effect was evaluated based on 24-h cumulative fentanyl consumption and additional administration by intravenous patient-controlled analgesia. The results showed that the CRAI group used significantly less analgesic. There were two adverse events related to CRAI, namely bleeding and splenic infarction.

CRAI with NM did not inhibit the development of pancreatic necrosis although early analgesic effect of CRAI was superior to that of IV. Less-invasive IV therapy can be considered a viable alternative to CRAI therapy.