When to Pull the Trigger: Persistent Upper Gi Bleeding in the Setting of Splenic Artery Pseudoaneurysm and Gastrostomy Stent Ultimately Requiring Coil Embolization

Abstract

A 49-year old female with history of chronic pancreatitis, 14 cm peri-pancreatic pseudocyst with cystogastrostomy stent placed 1 year earlier, and splenic vein thrombosis presented with massive hematemesis and pre-syncope requiring aggressive volume resuscitation, high dose IV pantoprazole and octreotide drip.CASE PRESENTATION: Initial CT angiography (CTA) of the abdomen revealed a 4 mm splenic artery pseudoaneurysm without evidence of rupture or extravasation, and remonstrated a previously identified splenic vein thrombosis with reduction in size of existing pancreatic pseudocyst. Upper endoscopy showed a large blood clot in the gastric body, suspected to be caused by irritation due to the cystogastrostomy stent. Her hematemesis promptly resolved but she developed frequent melena. The next day, she became severely hypotensive and repeat CTA of the abdomen showed the splenic artery pseudoaneurysm abutting the pseudocyst had enlarged from 6x4 mm to 8x6 mm. There was no evidence of active bleeding but there was a large blood clot in the stomach. She continued to be hemodynamically unstable, and developed hematochezia 4 days later. Repeat CTA showed no active extravasation of contrast and stable splenic artery pseudoaneurysm. Repeat upper endoscopy did not show active bleeding and the clot in her stomach had cleared. She developed recurrent hematemesis which prompted intubation for airway protection. A repeat CTA of the abdomen showed gastric blood, and lack of opacification of the splenic pseudoaneurysm with no active extravasation of contrast later confirmed by splenic artery angiogram. A subtle defect was noted within the expected location of the pseudoaneurysm with spasm within the mid/distal splenic artery. Coil embolization of the proximal splenic artery across the region of the pseudoaneurysm was performed. Post embolization angiogram showed no antegrade flow within the splenic artery. She continued to have melena, cystogastrostomy stent was removed, and upper endoscopy revealed only old blood. At 1 month follow-up, her melena and hematochezia had resolved, CT Abdomen did not show the splenic artery pseudoaneurysm. DISCUSSION: Our case outlines one of the less frequent causes of gastrointestinal bleeding and highlights the necessity of a multi-disciplinary team in critical care. Careful management of our persistently hemodynamically unstable patient and close coordination with the gastroenterology and interventional radiology teams led to the patients survival.CONCLUSIONS: Splenic artery pseudoaneurysms, while rare, are typically caused by pancreatitis, trauma, or peptic ulcer disease and almost always necessitate ICU stay when bleeding occurs. The multitude of precipitating factors in this case with focus on the cystogastrostomy stent outlines the necessity of considering splenic artery pseudoaneurysm as a potential cause of GI bleeding while addressing stent-related complications.