Department of Neurology, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, USA.
Department of Neurosurgery, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, USA.
J Neuroimaging. 2024 May-Jun;34(3):348-355. doi: 10.1111/jon.13197. Epub 2024 Mar 30.
Thresholds for abnormal transcranial Doppler cerebrovascular reactivity (CVR) studies are poorly understood, especially for patients with cerebrovascular disease. Using a real-world cohort with cerebral arterial stenosis, we sought to describe a clinically significant threshold for carbon dioxide reactivity (CO2R) and vasomotor range (VMR).
CVR studies were performed during conditions of breathing room air normally, breathing 8% carbon dioxide air mixture, and hyperventilation. The mean and standard deviation (SD) of CO2R and VMR were calculated for the unaffected side in patients with unilateral stenosis; a deviation of 2 SDs below the mean was chosen as the threshold for abnormal. Receiver operating characteristic (ROC) curves for both sides for patients with unilateral and bilateral stenosis were evaluated for sensitivity (Sn) and specificity (Sp).
A total of 133 consecutive CVR studies were performed on 62 patients with stenosis with mean±SD age 55±16 years. Comorbidities included hypertension (60%), diabetes (15%), stroke (40%), and smoking (35%). In patients with unilateral stenosis, mean±SD CO2R for the unaffected side was 1.86±0.53%, defining abnormal CO2R as <0.80%. Mean±SD CO2R for the affected side was 1.27±0.90%. The CO2R threshold predicted abnormal acetazolamide single-photon emission computed tomography (SPECT) (Sn = .73, Sp = .79), CT/MRI perfusion abnormality (Sn = .42, Sp = .77), infarction on MRI (Sn = .45, Sp = .76), and pressure-dependent exam (Sn = .50, Sp = .76). For the unaffected side, mean±SD VMR was 39.5±15.8%, defining abnormal VMR as <7.9%. For the affected side, mean±SD VMR was 26.5±17.8%. The VMR threshold predicted abnormal acetazolamide SPECT (Sn = .46, Sp = .94), infarction on MRI (Sn = .27, Sp = .94), and pressure-dependent exam (Sn = .31, Sp = .90).
In patients with multiple vascular risk factors, a reasonable threshold for clinically significant abnormal CO2R is <0.80% and VMR is <7.9%. Noninvasive CVR may aid in diagnosing and risk stratifying patients with stenosis.
目前，人们对异常经颅多普勒脑血管反应性（CVR）研究的阈值了解甚少，尤其是对于脑血管疾病患者。本研究使用伴有脑动脉狭窄的真实队列，旨在描述二氧化碳反应性（CO2R）和血管舒缩范围（VMR）的临床显著阈值。
在正常呼吸空气、呼吸 8%二氧化碳混合空气和过度通气条件下进行 CVR 研究。在单侧狭窄患者中，计算未受影响侧的 CO2R 和 VMR 的平均值和标准差（SD）；选择平均值减去 2 个 SD 作为异常的阈值。评估单侧和双侧狭窄患者两侧的接收者操作特征（ROC）曲线的敏感性（Sn）和特异性（Sp）。
共对 62 例狭窄患者的 133 例连续 CVR 研究进行了分析，患者平均年龄为 55±16 岁。合并症包括高血压（60%）、糖尿病（15%）、卒中和吸烟（35%）。在单侧狭窄患者中，未受影响侧的 CO2R 平均值和标准差为 1.86±0.53%，将异常 CO2R 定义为<0.80%。受影响侧的 CO2R 平均值和标准差为 1.27±0.90%。CO2R 阈值预测异常乙酰唑胺单光子发射计算机断层扫描（SPECT）（Sn=0.73，Sp=0.79）、CT/MRI 灌注异常（Sn=0.42，Sp=0.77）、MRI 梗死（Sn=0.45，Sp=0.76）和压力依赖性检查（Sn=0.50，Sp=0.76）。对于未受影响的一侧，VMR 的平均值和标准差为 39.5±15.8%，将异常 VMR 定义为<7.9%。对于受影响的一侧，VMR 的平均值和标准差为 26.5±17.8%。VMR 阈值预测异常乙酰唑胺 SPECT（Sn=0.46，Sp=0.94）、MRI 梗死（Sn=0.27，Sp=0.94）和压力依赖性检查（Sn=0.31，Sp=0.90）。
在存在多种血管危险因素的患者中，临床显著异常 CO2R 的合理阈值为<0.80%，VMR 的合理阈值为<7.9%。非侵入性 CVR 可能有助于诊断和对狭窄患者进行风险分层。
