Endovascular treatment of acute ischemic stroke: device type and complications
Due to the COVID-19 crisis the PhD defence of Marie Louise Bernsen will take place (partly) online.
The PhD defence can be followed by a live stream.
Marie Louise Bernsen is a PhD student in the research group Clinical Neurophysiology (CNPH). Her supervisors are prof.dr. J. Hofmeijer from the Faculty of Science and Technology (S&T) and prof.dr. W.H. van Zwam (University Medical Center Maastricht).
This thesis addresses technical developments to further improve the benefits of endovascular treatment (EVT) in acute ischemic stroke. In part I the two most used thrombectomy methods are compared. In part II prediction and diagnosis of two much encountered complications are studied.
Part I: Comparison of stent retriever and aspiration thrombectomy
In Chapter 2, aspiration and stent retriever thrombectomy were compared using data from the MR CLEAN Registry.12 Analysis of 1175 patients included in the first part of this registry showed that, in routine clinical practice, similar technical and clinical results are achieved when endovascular treatment is performed by direct aspiration and stent retriever as first approach. In patients treated with direct aspiration as first treatment technique, procedure times were shorter (median 57 minutes versus 70 minutes) and, as a consequence, times from symptom onset to reperfusion were shorter. However, we observed no difference in clinical outcome.
Occlusions that qualify for endovascular treatment in anterior circulation stroke are located in the proximal part of the middle cerebral artery (M1 and M2 segments), proximal part of the anterior cerebral artery (A1 and A2 segments), or in the intracranial internal carotid artery. Most patients eligible for endovascular treatment have an occlusion in the proximal part of middle cerebral artery (M1 segment), followed by patients with an occlusion in the more distal part of this artery (M2 segment), and the intracranial internal carotid artery. Occlusions of the anterior cerebral artery are relatively rare.
In Chapter 3, registry data were used to compare aspiration and stent retriever thrombectomy in relation to occlusion segment.17 In total, 2282 patients were included in this analysis. Again, faster procedures were seen in the aspiration group. In addition, first line aspiration treatment was associated with higher reperfusion rates in the total group, and in all subgroups based on occlusion segments. However, again, there was no difference in functional outcome between the stent retriever and aspiration treatment groups, neither in the total group, nor in subgroups based on occlusion segment. Lack of effect on functional recovery despite faster reperfusion may be related to the heterogeneity of patient groups, where functional recovery depends on factors other than reperfusion, as well. It may also be associated with insufficient sensitivity of the modified Rankin Scale to detect small differences in functional recovery between patient groups.
In Chapter 4, aspiration and stent retriever thrombectomy were compared in posterior circulation stroke. Two hundred and five patients registered in the MR CLEAN Registry were studied. Most patients (79%) had an occlusion in the basilar artery. Reperfusion rates were higher and reperfusion was achieved faster with aspiration first. As opposed to comparisons of stent retriever and aspiration thrombectomy in the anterior circulation, in posterior circulation infarcts the higher reperfusion rates and faster procedures resulted in better clinical outcome after aspiration thrombectomy. Probably, the vulnerability of the brain stem area, combined with the fragile anatomy of the basilar artery and its many small side branches, adds to the technical benefits of aspiration over stent retriever. Together, these lead to better functional outcome in posterior circulation infarcts treated with aspiration first.
Part II: Detection of complications
In Chapter 5, 1445 patients of the MR CLEAN Registry were studied in order to identify early predictors of malignant infarction after EVT. Malignant infarction had an incidence of 6% in our cohort. This percentage is lower than expected in patients with large vessel occlusive stroke, which is probably related to (early) revascularization. Independent predictors of the development of malignant infarction were younger age, higher baseline NIHSS score, ICA occlusion, lower ASPECTS score on baseline CT, low collateral status, longer time from onset to groin puncture, and lack of reperfusion. Associations with low collateral status, longer time from onset to groin puncture, and lack of reperfusion confirm the importance of timely reperfusion to prevent malignant infarction. A prediction model combining the parameters showed good discrimination between patients with and without malignant infarction.
In Chapter 6, further improvement of the prediction model by incorporating the ratio of intracranial cerebrospinal fluid volume (CSF) to intracranial volume (ICV) was investigated. This ratio should relate to the risk of malignant infarction, because more CSF volume serves as a buffer to prevent brain herniation due to swelling. As expected, the effect of age on malignant edema disappeared after adding the CSF/ICV ratio to the prediction model. In general, older patients have a higher CSF/ICV ratio, because of brain atrophy. As age and the CSF/ICV ratio are closely related to each other and to malignant edema, one predictor could be removed from the prediction model. We propose to remove the age variable, because the CSF/ICV added more to model discrimination than age.
Postprocedural imaging in treated patients can result in diagnostic difficulties due to hyperattenuation on CT, which may either represent contrast agent from the endovascular procedure or intracranial hemorrhage. New spectral CT techniques may offer a solution by allowing interrogation between blood and iodine contrast agents. In Chapter 7, results of initial experience with spectral detector CT after endovascular treatment are presented. Sixty-three patients with spectral CT within one week after EVT were studied.
As compared with single energy CT, the diagnosis of postprocedural hemorrhage changed in eight patients (13%). In four patients, the diagnosis of intracranial hemorrhage was rejected after spectral data assessment and in four patients additional hemorrhage was detected. In three patients the grading of ICH was changed.
Based on these results we conclude that spectral detector CT could be a valuable addition in hospitals who perform EVT.