Characterization of Subarachnoid Hyperdensities After Thrombectomy for Acute Stroke Using Dual-Energy CT
Arturo Renú, Carlos Laredo, Alejandro Rodríguez-Vázquez, Daniel Santana, Mariano Werner, Laura Llull, Antonio López‐Rueda, Xabier Urra, Salvatore Rudilosso, Vı́ctor Obach, Sergi Amaro, Ángel Chamorro
Abstract
<h3>Background and Objectives</h3> The presence of postinterventional subarachnoid hyperdensities (SA-HD) is a relatively common finding after mechanical thrombectomy (MT). We aimed to assess the incidence, characteristics, clinical relevance, and predictors of SA-HD after MT as categorized through the use of postinterventional dual-energy CT (DE-CT). <h3>Methods</h3> A single-center consecutive series of patients with acute stroke treated with MT was retrospectively reviewed. Posttreatment SA-HD were defined as incident extraaxial hyperdensities in a follow-up DE-CT performed within a median of 8 hours after MT. SA-HD were further classified according to their content (isolated contrast extravasation vs blood extravasation) and extension (diffuse [hyperdensities in more than one extraparenchymal compartment] vs nondiffuse). Adjusted logistic regression models assessed the association of SA-HD with pretreatment and procedural variables and with poor clinical outcome (shift towards worse categories in the ordinal Rankin Scale at 90 days). <h3>Results</h3> SA-HD were observed in 120 (28%) of the 424 included patients (isolated contrast extravasation n = 22, blood extravasation n = 98). In this group, SA-HD were diffuse in 72 (60%) patients (isolated contrast extravasation n = 7, blood extravasation n = 65) and nondiffuse in 48 (40%) patients (isolated contrast extravasation n = 15, blood extravasation n = 33). Diffuse SA-HD were significantly associated with worse clinical outcome in adjusted models (common odds ratio [cOR] 2.3, 95% confidence interval [CI] 1.36–4.00, <i>p</i> = 0.002), unlike the specific SA-HD content alone. In contrast with the absence of SA-HD, only the diffuse pattern with blood extravasation was significantly associated with worse clinical outcome (cOR 2.4, 95% CI 1.36–4.15, <i>p</i> = 0.002). Diffuse SA-HD patterns were predicted by M2 occlusions, more thrombectomy passes, and concurrent parenchymal hematomas. <h3>Discussion</h3> In our cohort of patients imaged within a median of 8 hours after MT, postinterventional SA-HD showed a diffuse pattern in 17% of thrombectomies and were associated with more arduous procedures. Diffuse SA-HD but not local collections of blood or contrast extravasations were associated with an increased risk of poor outcome and death. These findings reinforce the need for improvement in reperfusion strategies. <h3>Classification of Evidence</h3> This study provides Class IV evidence that in individuals with proximal carotid artery territory occlusions treated with MT, diffuse postinterventional SA-HD on imaging 8 hours postprocedure are associated with worse clinical outcomes at 90 days.