ENDOSCOPIC ULTRASOUND-GUIDED COIL AND GLUE IS COMPARABLE TO RADIOLOGICAL ANGIOEMBOLISATION FOR THE MANAGEMENT OF VISCERAL ARTERY PSEUDOANEURYSMS: A LARGE MULTICENTER EXPERIENCE

Introduction: Visceral artery pseudoaneurysms (PsA) can develop in patients with acute (AP) and chronic pancreatitis (CP) and intervention radiology (IR)-guided intervention is the standard of care. EUS-guided angioembolisation (EUS-A) using coil and cyanoacrylate (CYA) glue injection is a relatively newer modality with limited experience. Comparison between these two modalities is lacking and hence this study was planned.

Methods: This study was conducted at two academic tertiary-care centres in India between September 2018-September 2023 and included all consecutive AP/CP patients with visceral artery PsA, who underwent either EUS-guided or IR-guided angioembolisation (IR-A). EUS-A was carried out using coil and CYA injection, while IR-A was performed using varying combinations of coil, glue, thrombin or gelfoam depending on the anatomy and operator preference. The number of coils used depended on the size of the PsA. Baseline characteristics, amount of coil and other agents needed, technical success, obliteration rates, and adverse events were documented.

Results: A total of 121 patients (mean age 39.43±12.1 years; 102; 84.3% males) with 123 PsA were included of which 40 underwent EUS-A and 83 undergoing IR-A. Most common underlying diagnosis was CP (97; 80.2%), and majority presented with gastrointestinal bleed (85; 69.1%). Splenic artery was involved in 86 (69.9%) followed by gastroduodenal artery in 25 (20.3%).
On comparison, EUS-A had significantly larger PsA (maximum diameter 27.8±17.0 mm vs 15.12±14.4mm; p<0.001) with similar baseline disease, presentation, hemoglobin levels, and transfusion requirements compared to IR-A. The number of coils used were similar in the two arms while the amount of glue/thrombin used was significantly higher in the EUS-A (2.25±0.8 ml vs. 1.89±0.6 ml; p=0.026). With similar technical success, the number of sessions required for obliteration were comparable in the two arms (1.15±0.4 vs. 1.10±0.3). Complete obliteration achieved at 72 hrs. (EUS-A – 90.0% vs. IR-A – 93.8%) and at 1 month (EUS-A – 97.5% vs. IR-A – 96.3%) were similar. On multivariate analysis, the size of the PsA was the key determinant for complete obliteration at 1 month (aOR – 1.10; CI 1.05-1.16; p<0.0001) even after adjusting for mode of intervention (EUS-A vs IR-A), diagnosis or the artery involved. While adverse events were similar between two arms, the length of hospital stay was longer in the IR-A arm (7.77±4.2 vs. 3.88±2.4 days; p<0.001) (Table 1).

Conclusion: EUS-A using coil with CYA is a feasible and safe modality with performance comparable to IR-A for the management of visceral artery PsA. The size of the PsA rather than the modality of intervention is the key factor determining complete obliteration.