Low Cerebral Blood Volume Index, Thrombectomy, and Prior Stroke Are Independently Associated With Hemorrhagic Transformation in Medium‐Vessel Occlusion Ischemic Stroke
Vivek Yedavalli, Manisha Koneru, Meisam Hoseinyazdi, Elisabeth B. Marsh, R. Llinás, Victor Urrutia, Richard Leigh, Luis Fernando Muñoz González, Risheng Xu, Justin M. Caplan, Judy Huang, Hanzhang Lu, Max Wintermark, Adam A. Dmytriw, Adrien Guenego, Gregory W. Albers, Licia Luna, Jeremy J. Heit, Kambiz Nael, Argye E. Hillis
Abstract
Background Hemorrhagic transformation (HT) is a major complication in patients with acute ischemic stroke (AIS) treated with endovascular thrombectomy. However, HT in patients with AIS due to medium‐vessel occlusions has still not been well researched. In large‐vessel occlusions, collateral status is an important determinant of postprocedural HT, and the cerebral blood volume (CBV) index is a reliable surrogate of collateral status. The aim of our study is to identify an optimal CBV index threshold associated with HT in patients with AIS due to medium‐vessel occlusion and evaluate additional parameters that are independently associated with HT in this group. Methods This retrospective analysis of our prospectively collected database from 2 comprehensive stroke centers consisted of patients presenting with AIS due to medium‐vessel occlusion from 2019 to 2023. The primary outcome was the presence of HT on follow‐up imaging. Optimal CBV index cutoff for HT was derived from a univariate logistic regression analysis. Multivariable logistic regression analysis for HT was derived from the dichotomized CBV index and other covariates. The receiver operator characteristic curve yielded area under the curve. Statistical significance was P <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"> <mml:semantics> <mml:mo>≤</mml:mo> <mml:annotation encoding="application/x-tex">$ \le $</mml:annotation> </mml:semantics> </mml:math> 0.05. Results Of 111 patients (median age, 70 years; 43.2% women) included, 26 (23.4%) patients had HT. The optimal CBV index cutoff was 0.7. From multivariable regression analysis, significant variables included prior stroke (adjusted odds ratio [aOR], 7.18 [95% CI, 1.60–32.16]; P = 0.01), endovascular thrombectomy attempt (aOR, 7.86 [95% CI, 1.78–34.68]; P = 0.01), and CBV index ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"> <mml:semantics> <mml:mo>≤</mml:mo> <mml:annotation encoding="application/x-tex">$ \le $</mml:annotation> </mml:semantics> </mml:math> 0.7; aOR, 4.23 [95% CI, 1.02–17.59]; P = 0.04). The area under the curve was 0.82 (95% CI, 0.69–0.91). Conclusion A CBV index <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"> <mml:semantics> <mml:mo>≤</mml:mo> <mml:annotation encoding="application/x-tex">$ \le $</mml:annotation> </mml:semantics> </mml:math> 0.7 was independently associated with HT in patients with AIS due to medium‐vessel occlusion. Endovascular thrombectomy attempt and prior stroke history were also independently associated with HT in this population.