Venous anatomy of the left ventricular summit: Therapeutic implications for ethanol infusion

Background Venous ethanol ablation (VEA) is effective for treatment of left ventricular (LV) summit (LVS) arrhythmias. The LVS venous anatomy is poorly understood and has inconsistent nomenclature. Objective The purpose of this study was to delineate the LVS venous anatomy by selective venography and 3-dimensional (3D) mapping during VEA and by venous-phase coronary computed tomographic angiography (vCTA). Methods We analyzed (1) LVS venograms and 3D maps of 53 patients undergoing VEA; and (2) 3D reconstructions of 52 vCTAs, tracing LVS veins. Results Angiography identified the following LVS veins: (1) LV annular branch of the great cardiac vein (GCV) (19/53); (2) septal (rightward) branches of the anterior ventricular vein (AIV) (53/53); and (3) diagonal branches of the AIV (51/53). Collateral connections between LVS veins and outflow, conus, and retroaortic veins were common. VEA was delivered to target arrhythmias in 38 of 53 septal, 6 of 53 annular, and 2 of 53 diagonal veins. vCTA identified LVS veins (range 1–5) in a similar distribution. GCV–AIV transition could either form an angle close to the left main artery bifurcation (n = 16; 88° ± 13°) or cut diagonally (n = 36; 133°±12°) (P ≤.001). Twenty-one patients had LV annular vein. In 28 patients only septal LVS veins were visualized in vCTA, in 2 patients only diagonal veins and in 22 patients both septal and diagonal veins were seen. In 39 patients the LVS veins reached the outflow tracts and their vicinity. Conclusion We provide a systematic atlas and nomenclature of LVS veins related to arrhythmogenic substrates. vCTA can be useful for noninvasive evaluation of LVS veins before ethanol ablation. Venous ethanol ablation (VEA) is effective for treatment of left ventricular (LV) summit (LVS) arrhythmias. The LVS venous anatomy is poorly understood and has inconsistent nomenclature. The purpose of this study was to delineate the LVS venous anatomy by selective venography and 3-dimensional (3D) mapping during VEA and by venous-phase coronary computed tomographic angiography (vCTA). We analyzed (1) LVS venograms and 3D maps of 53 patients undergoing VEA; and (2) 3D reconstructions of 52 vCTAs, tracing LVS veins. Angiography identified the following LVS veins: (1) LV annular branch of the great cardiac vein (GCV) (19/53); (2) septal (rightward) branches of the anterior ventricular vein (AIV) (53/53); and (3) diagonal branches of the AIV (51/53). Collateral connections between LVS veins and outflow, conus, and retroaortic veins were common. VEA was delivered to target arrhythmias in 38 of 53 septal, 6 of 53 annular, and 2 of 53 diagonal veins. vCTA identified LVS veins (range 1–5) in a similar distribution. GCV–AIV transition could either form an angle close to the left main artery bifurcation (n = 16; 88° ± 13°) or cut diagonally (n = 36; 133°±12°) (P ≤.001). Twenty-one patients had LV annular vein. In 28 patients only septal LVS veins were visualized in vCTA, in 2 patients only diagonal veins and in 22 patients both septal and diagonal veins were seen. In 39 patients the LVS veins reached the outflow tracts and their vicinity. We provide a systematic atlas and nomenclature of LVS veins related to arrhythmogenic substrates. vCTA can be useful for noninvasive evaluation of LVS veins before ethanol ablation.