Comparison of endoscopic ultrasound-guided primary and secondary prophylaxis for gastric variceal bleeding.

OBJECTIVES: Endoscopic ultrasound (EUS)-guided injection of cyanoacrylate (CYA) for primary prophylaxis (PP) of gastric varices (GV) is controversial. This study evaluates the safety and efficacy of this intervention. METHODS: Patients treated for PP of GV bleeding by EUS injection of CYA with or without coils were identified. Endoscopic techniques, outcomes, and adverse events (AEs) were reviewed and compared with a group treated for secondary prophylaxis (SP). Patients were followed until: (i) loss to follow-up; (ii) GV bleeding; (iii) interventional radiology or surgery decompression; (iv) liver transplant; or (v) death or comfort care. RESULTS: One hundred and nineteen patients (61 men; mean 59 ± 12 years) underwent EUS for PP (n = 24) or SP (n = 95). The PP group was treated with CYA alone (n = 18) or with coils (n = 4). Eight (33%) mild (n = 6) or moderate (n = 2) AEs and no index GV bleeding occurred during a mean of 6.1 ± 5.9 months follow-up. Repeat EUS in 22 (92%) PP patients showed 7 (32%) residual GVs, which were retreated with CYA alone (n = 6) or with coils (n = 1). Two (29%) mild (n = 1) or moderate (n = 1) AEs occurred after repeat EUS and 1/22 (5%) index GV bleed occurred during a mean 23 ± 25 months follow-up. Compared to the SP group, the PP group had lower Model for End-stage Liver Disease (MELD) score (P = 0.03), fewer GV stigmata (P < 0.001), required less CYA (P = 0.019) during index EUS, and had a longer time between index and surveillance EUS (P = 0.014). The incidence of AEs and GV bleeding between the two groups were similar. CONCLUSION: Posttreatment GV bleeding and AEs are similar following EUS-guided primary and secondary GV prophylaxis.