• Thrombectomy

We read with interest the article by Mokin et al ‘Thrombus density predicts successful recanalization with Solitaire stent retriever thrombectomy in acute ischemic stroke’.1 Based on their results, the authors conclude that thrombus density assessed by plain cranial CT scans predicts recanalization after Solitaire stent retriever thrombectomy for acute ischemic stroke. Some methodological shortcomings of this study may have led to erroneous interpretation of the results. The authors investigated patients treated solely by Solitaire stent retriever thrombectomy, which reduced the cohort to 41 patients, and does not represent current practice as stent retrievers are very often combined with preceding intravenous thrombolysis and additional intra-arterial recombinant tissue plasminogen activator administration.2

We are concerned about the assessment of thrombus density on non-contrast CT using 5 mm slice thickness. This technique is very insensitive for detecting proximal vessel occlusion using the hyperdense middle cerebral artery sign as a marker due to limited coverage of the distal internal carotid artery and middle cerebral artery origin. Riedel et al3 reported acceptable results when measuring thrombus length from CT scans with a slice thickness of 2.5 mm or less. We recently conducted a retrospective analysis of 171 stroke patients with proximal occlusion of the anterior circulation who underwent endovascular stroke treatment (37.4% stent retrievers). Assessment of thrombus density was not possible in 76 cases (44%) because cranial CT was acquired with 4.5 mm slice thickness. Overall, we found no difference in thrombus density between patients with successful and failed recanalization. We do not think that a slice thickness of 4.5 mm allows adequate determination of thrombus volume, as also mentioned by Moftakhar et al,4 who used 2.5 mm slice thickness to calculate thrombus volume.

The authors correctly point out the conflicting results regarding the usefulness of thrombus density for predicting recanalization after revascularization treatment. Several studies are biased by using a slice thickness of 4.5 or 5 mm, which is currently widely used in clinical admission multimodal CT protocols.5 We identified only three studies that used non-contrast CT with 2.5 mm slice thickness. These reported similar results, showing that thrombus with lower Hounsfield unit values predicts a lower likelihood of recanalization after intravenous thrombolysis, intra-arterial thrombolysis, a combination of both, or mechanical thrombectomy.4 ,6 ,7 Apart from the technical limitations of assessing thrombus density, the central question remains—if, from a pathophysiological point of view, thrombus composition has an influence on the entrapment mechanism of a stent retriever. Thrombus composition determines the mechanical features of the clot8 and different devices are likely to respond differently to a thrombus of identical composition.9 ,10 Data obtained in studies with accurate assessment of thrombus composition are still limited.

We strongly recommend the use of a thin slice thickness in acute stroke CT protocols to ensure reliable assessment of thrombus density and length. However, the study of Mokin et al shows the potential of thrombus density as a surrogate marker for success of recanalization after endovascular treatment.