Predictive factors of medullary infarction after endovascular internal trapping using coils for vertebral artery...

OBJECTIVE The first choice of treatment in cases of vertebral artery dissecting aneurysms (VADAs) is endovascular internal trapping (EIT) of the dissecting segment using coils. However, this procedure carries the risk of medullary infarction, and the risk factors for this complication are not well understood. This study investigated the risk factors causing medullary infarction. METHODS One hundred patients who underwent EIT for VADAs were included in this study. Ninety-three patients presented with subarachnoid hemorrhage. In cases involving the posterior inferior cerebellar artery (PICA), partial internal trapping targeting the ruptured site was performed to preserve the PICA. The VADAs were classified into the distal VA stump group, proximal VA stump group, and entire VA stump group, according to the location of VA segments without adequate flow-out vessels (such as the PICA [VA stump]) at risk of delayed thrombosis. The occurrence of medullary infarction was examined in each group using diffusion-weighted MRI and/or clinical symptoms. Various measurements were performed on digital subtraction angiography, and the risk factors for medullary infarction were analyzed. RESULTS Medullary infarction occurred in 30 patients, affecting the posterolateral medulla in 27 patients and the anteromedial medulla in 3 patients. Medullary infarction occurred in 3 of 47 patients (6%) in the distal VA stump group, 10 of 19 patients (53%) in the proximal VA stump group, and 17 of 34 patients (50%) in the entire VA stump group. The length of trapping was significantly longer in the infarction group than in the noninfarction group but did not differ among the 3 groups. Total length (length of trapping plus VA stump) was a risk factor for medullary infarction in the proximal VA stumps. CONCLUSIONS The primary risk factor for medullary infarction after EIT is not the length of trapping; rather, it is the anatomical location of the VADAs. The risk of medullary infarction is low in cases with distal VA stumps, but the symptoms are severe. Preservation of the origin of the anterior spinal artery can reduce the risk of medullary infarction. The risk of medullary infarction is high in cases with proximal VA stumps, but the symptoms are mild. A shorter length of trapping, although less likely to lead to complications, cannot prevent medullary infarction because the total length depends on the anatomical location of the PICA and not on the surgical technique. Reconstructive therapy should be indicated for patients with ruptured VADAs at high risk of severe ischemic complications (e.g., patients with hypoplasia of the contralateral VA or cases involving the PICA or anterior spinal artery, which are inappropriate for partial internal trapping) or for patients with unruptured VADAs.