Carbon dioxide (CO2) vs iodinated contrast digital subtraction angiography during balloon-occluded retrograde transvenous obliteration (BRTO) using foam sclerosant for gastric varices.

PURPOSE: To compare the visualization of the target gastric varices (GV) on balloon-occluded retrograde transvenous venography (BRTV) using iodinated contrast material vs carbon dioxide (CO(2)) in preparation for subsequent balloon-occluded retrograde transvenous obliteration (BRTO) using foam sclerotherapy.

MATERIALS AND METHODS: In 16 consecutive patients with nonruptured GV, BRTV was performed first using iodinated contrast material and then with CO(2). BRTV was repeated whenever there were changes in the catheter or patient position or when coil embolization of collaterals was needed. Each visualization grade of GV (grade 1 = GV only; grade 2 = GV > collaterals; 3 = GV < collaterals; grades 4-5 = collaterals only) was determined by two observers in consensus. During foam BRTO, the GV visualization grade was recorded again and confirmed by C-arm computed tomography (CT).

RESULTS: In 38 pairs of BRTV, GV grades were significantly (P < .0001) lower (ie, favoring BRTO) on CO(2) BRTV (mean ± standard deviation, 1.8 ± 0.8) than on iodine BRTV (3.4 ± 0.8). GV grades on foam BRTO (1.4 ± 0.7) were similar to the grades obtained on the most recent CO(2) BRTV (1.3 ± 0.5) but were significantly smaller (P < .0001) than on iodinated BRTV (3.1 ± 0.9). GV were opacified by foam on initial C-arm CT in 14 patients (87.5%), and complete thrombosis of GV was obtained without any complication in all 16 patients (100%). CO(2) reached the GV even when iodinated contrast material could not (grade 4) in seven of our 16 patients (43.8%), leading to successful BRTO.

CONCLUSIONS: CO(2) BRTV visualized GV better than did iodine BRTV and changed the management of more than 40% of patients by enabling successful foam BRTO in patients in which conventional liquid BRTO could not be performed.