Embolization of a mesenteric arteriovenous fistula following pancreatic allograft: The steal effect

The first pancreatic transplant was performed in 1966 by Kelly and Lillehei at the University of Minnesota (1, 2). Nearly 30 years later, Ozaki et al., published the first occurrence of an arteriovenous fistula (AVF) in a transplanted allograft bundle (3). From its early days as a radical and experimental procedure in the treatment of insulin-dependent diabetes mellitus, this operation has progressed to become routine in many major medical centers. However, the incidence of technical complications remains relatively high, ranging from 10 to 40% (1). The list of potential complications includes allograft pancreatitis, vascular thrombosis, hemorrhage, pseudoaneurysm formation, anastomotic leaks, intra-abdominal infections, and, although rare, AVF. Lowell et al., in 1993, reported this last complication in 3 of 90 consecutive pancreatic recipients (4). These same authors promoted the theory that AVF formation was directly related to procurement technique: a nonspecific "blind ligation" of mesenteric vessels along the inferior pancreatic border. However, this approach continues to be used commonly. We have identified the occurrence of an allograft superior mesenteric artery-superior mesenteric vein (SMA-SMV) AVF in a pancreas-after-kidney (PAK) transplant recipient.