Is endoscopic ultrasound-guided angioembolization of feeder vessel as good as targeting submucosal variceal complex in the management of gastric varices: A pragmatic comparative analysis

Aims The management of gastric varices (GV) using endoscopic ultrasound-guided angio-embolization using coil and glue (EUS-CG) is a relatively recent modality showing higher obliteration and lower reintervention rates compared to the conventional endoscopic approach. In cases where it can be delineated, targeting feeder/perforator vessel by EUS seems a logical alternative. However, no study exists comparing this strategy with the conventional strategy of targeting the submucosal variceal complex (SVC). Hence, this study was designed to compare these two strategies of EUS-CG for GV management.

Methods This study included patients of GV requiring obliteration with discernible feeder vessel localized on EUS. Baseline clinical, demographic and laboratory parameters were documented. Baseline imaging and EUS were performed to assess the size of the varices, feeder vessel, and classified as per the Arakawa classification. Patients were divided into 2 arms wherein either feeder or SVC was targeted and obliterated using EUS-CG. Procedural details such as amount of glue, number of coils, obliteration rates, number of sessions required were documented. Follow-up data collected included adverse events, obliteration at 4 weeks, bleeding after index procedure, and need for re-intervention [1] [2].

Results Out of 117 patients (male 84, 71.9%; mean age 46.4±13.3 years) included in the study, feeder vessel (EUS-F) was targeted in 24 cases, and SVC in 93 cases. At baseline, the type of presentation and etiology of liver disease was similar between the two arms. GOV2 (70; 59/8%), and type 1 Arakawa (86; 73.5%) were the most common type of GV. The size of GV were similar in 2 arms (17.6±7.6 vs 18.9±7.6 mm, p=0.46) but size of feeder was larger in EUS-F arm (10.4±3.05 vs 8.4±2.7 mm, p=0.005). The technical success, immediate complete obliteration rates and at 72 hours and the number of sessions required were similar in the 2 arms. The rates of complete obliteration at 4 weeks (23, 95.8% vs 81. 87.1%; p=0.225) were higher and reintervention (1; 4.2% vs 11; 11.8%; p=0.455) and bleeding events (2; 8.3% vs 12; 12.9%; p=0.732) were numerically lower in EUS-F arm, though statistically non-significant. The overall requirement of coils (1.8±1.04 vs 2.75±1.75; p=0.002) and glue (1.77±0.7 vs 2.8±1.74 ml; p<0.0001) were significantly lower in EUS-F arm. On multivariate analysis, only GV size (aOR-1.24; p<0.0001) was a significant predictor of need for reintervention, not the technique used for GV management, even after adjusting for age, sex, etiology of disease and CTP class.

Conclusions For EUS-guided angioembolization of GV, the strategy of targeting the feeder vessel, when feasible, is comparable to the conventional strategy of targeting the large SVC, in terms of similar obliteration, reintervention and subsequent bleeding rates, with significantly lower requirements of coil and glue. This can translate to a more cost-effective strategy for the subset of cases with discernible feeder vessel.

Publication History

Article published online:
27 March 2025

• References

• 1 Dhar J, Samanta J.. Endoscopic ultrasound-guided vascular interventions: An expanding paradigm. World J Gastrointest Endosc 2023; 15 (04): 216-239
• 2 Samanta J, Nabi Z, Facciorusso A. et al. EUS-guided coil and glue injection versus endoscopic glue injection for gastric varices: International multicentre propensity-matched analysis. Liver Int 2023; 43 (08): 1783-1792