Rupture of portal vein pseudoaneurysm caused by impinging infected walled off pancreatic necrosis (WOPN): a rare complication

Discussion

Focal aneurysmal dilatation of portal vein is a rare entity. It can be categorised as congenital, acquired (portal hypertension, trauma, tumour invasion) and pseudoaneurysm (secondary to necrotising pancreatitis).1–3 There is one report in the literature of portal vein pseudoaneurysm caused by impinging pseudo cyst.4 5 Present case is the first in the literature reporting rupture of portal vein pseudoaneurysm caused by impinging WOPN.

Arterial pseudoaneurysms are more common than venous aneurysm in patients with DPS in WOPN and are much more common when lumen apposing metal stents (LAMS) were used (8.7% in LAMS vs 0.7% in DPS).6 Diagnosis of pseudoaneurysm of portal vein can be challenging as subtle irregularity in the wall of portal vein was the only sign visible on CT scan initially (figures 1 and 2). However, after EUS-guided drainage of WOPN, subsequent CT scan revealed outpouching of weakened portal vein (figure 3), which probably was due to sudden decrease in the intracystic pressure by DPS stents. Patient subsequently presented with warning bleed in the form of melaena and 24 hours later with massive haematemesis. Though it is difficult to pinpoint the cause of rupture but it is likely due to residual necrosum or decubitus of transgastric pigtail stents.

Management of PVA depends on aetiology and presence or absence of complications. Congenital PVA are likely to have uncomplicated course compared with acquired PVA. Uncomplicated PVA has been treated conservatively previously.2 However, surgery is warranted for PVA presenting with bothersome symptoms, increase in size over time, mass effect, rupture or for non-thrombotic PVA >3 cm.7 8 On the other hand pseudoaneurysms of portal vein would merit intervention due to high risk of rupture.

There are two cases of portal vein erosion in acute pancreatitis described in literature, one case was managed with exploratory laparotomy and initial packing of abdominal cavity for haemostasis and subsequently percutaneous stent graft placement. Other case was managed conservatively.9 However, portal vein pseudoaneurysm occurring in the setting of WOPN managed with DPS stent, cannot be left alone. It is important to highlight that our patient recovered for 3 weeks and then presented with rupture.

Surgery to control bleeding from portal vein in necrotic cavity has its own challenges due to presence of friable, inflamed and oedematous tissues which are likely to give in to the compression of vascular clamps and may further damage the vein. We realised that in such a scenario best way to control haemorrhage was to apply digital pressure on the rent itself. Digital compression of portal vein above and below the rent was ineffectual as bleeding from splenic vein side was continuous and obscured the narrow operative field. One has to suture in a synchronised way as assistant lifts digital pressure from the rent and surgeon takes sutures and immediately assistant puts back digital pressure to occlude the rent thus limiting blood loss. We found that half circle, 20 mm size, needle with 4–0 polypropylene was helpful as this size needle was easy to visualise and handle in presence of bleeding. One has to be careful while taking suture and tying knots as suture can cut through inflamed wall of portal vein. At the same time, care has to be taken not to accidently take posterior wall. It will be prudent to remove any residual necrosum in the cavity. This case also brings to attention the fact that presence of portal vein pseudoaneurysm in WOPN with DPS stents in vicinity should be subjected to definitive management (endovascular/surgical) depending on availability of expertise even if patient is asymptomatic.

To summarise, haemorrhage following endoscopic interventions in acute necrotising pancreatitis occurs most commonly from rupture of arterial pseudoaneurysm, however, in the present case we found that haemorrhage occurred from rupture of portal vein pseudoaneurysm following endoscopic transgastric pigtail catheter drainage of WOPN. The pseudoaneurysm was picked up on CT scan and patient underwent emergent laparotomy and venorrhaphy to control exsanguinating haemorrhage.