No Association Between Thrombus Perviousness and Cardioembolic Stroke Etiology in Basilar Artery Occlusion Stroke

Hemodynamic Characteristics of Intracranial Atherosclerotic Stenosis: A Pilot Study of Contrast Enhancement Time-Density Curves Based on Regions of Interest

OBJECTIVE

The present study aimed to analyze the hemodynamic characteristics of occluded vessels responsible for acute ischemic stroke and to diagnose the occlusion types.

METHODS

Multimodal computed tomography (CT) was used to accurately identify the range of occlusion of large intracranial vessels. Regions of interest (ROI 1-3 ) were manually delineated at sites 2 mm away from the proximal, middle, and distal portions of each occlusion, generating 3 contrast enhancement time-density curves. The peak CT attenuation values, or Hounsfield units (H 1-3 ), and time-to-peak values (T 1-3 ) were extracted from each curve. H 0 and T 0 of the time-density curve, based on ROI 0 of the automatically recognized input artery, were used as the baseline values with which the odds ratios of each parameter, H 1-3/0 and T 1-3/0 , were obtained. The present study aimed to establish prediction models for intracranial atherosclerotic stenosis (ICAS) based on each ROI's time-density curve.

RESULTS

Among the 33 acutely occluded intracranial vessels, 10 were found to have ICAS, whereas 23 did not, based on the diagnostic criteria. Significant differences were observed in patient sex, neutrophil-to-lymphocyte ratio upon admission, Alberta Stroke Program Early CT Score 24-48 hours after reperfusion therapy, and H 1/0 , H 3/0 , and T 3/0 between the ICAS and non-ICAS groups ( P < 0.05). The prediction model (model 3) based on the ROI 3 time-density curve showed the best performance for the diagnosis of ICAS (area under the curve, 0.944; 95% confidence interval, 0.854-1.000). The prediction models based on ROI 1 (model 1) and ROI 2 (model 2) showed moderate diagnostic performance (area under the curve, 0.817 vs 0.822, respectively). The best visualization for proximal occlusions was in the first phase (arterial phase) of multiphase CT angiography, and in the second phase (early venous phase) for distal occlusions.

CONCLUSIONS

The contrast enhancement time-density curves of the ROIs at all evaluated portions of the acute ischemic stroke occlusions provided a visual display of the blood flow characteristics of the responsible vessels. The time-density curve of the ROI placed 2 mm from the distal occlusion was a combined effect of residual blood flow and collateral establishment, thus providing good performance for the diagnosis of ICAS.

CT Marker in Emergency Imaging of Acute Basilar Artery Occlusion: Thrombosis vs. Embolism

Purpose: Acute basilar artery occlusion, a neurovascular emergency leading to high rates of morbidity and mortality, is usually diagnosed by CT imaging. The outcome is partly dependent on etiology, with a worse outcome in occlusions with underlying basilar artery stenosis. As this occlusion type requires a more complex angiographic therapy, this study aimed to develop new CT markers in emergency admission imaging to rapidly identify underlying stenosis. Methods: A total of 213 consecutive patients (female n = 91, age in years (mean/SD/range): 72/13/28−97), who received endovascular treatment at a single comprehensive stroke center for acute basilar artery occlusion, were included in this study. After applying strict inclusion criteria for imaging analyses, novel CT imaging markers, such as ‘absolute density loss’ (ADL) and relative thrombus attenuation (CTA-index), that measure perviousness, were assessed for n = 109 patients by use of CT-angiography and correlated to different occlusion patterns (thrombotic vs. embolic). Inter-observer agreement was assessed using an intraclass correlation coefficient for independent measures of a radiologist and a neuroradiologist. Associations between the imaging markers and clinical and interventional parameters were tested. Results: CT markers differ between the subgroups of basilar artery occlusions with and without underlying stenosis (for ADL: 169 vs. 227 HU (p = 0.03), for CTA-index: 0.55 vs. 0.70 (p < 0.001)), indicating a higher perviousness in the case of stenosis. A good inter-rater agreement was observed for ADL and CTA-index measures (ICC 0.92/0.88). For the case of embolic occlusions, a more pervious thrombus correlates to shorter time intervals, longer procedure times, and worse reperfusion success (p-values < 0.05, respectively). Conclusions: ADL and CTA-index are easy to assess in the emergency setting of acute basilar artery occlusion with the use of routinely acquired CT-angiography. They show a high potential to differentiate thrombotic from embolic occlusions, with an impact on therapeutic decisions and angiographic procedures. Measurements can be quickly performed with good reliability, facilitating implementation in clinical practice.