In-Hospital Complication Rates After Stent Treatment of 388 Symptomatic Intracranial Stenoses
Results From the INTRASTENT Multicentric Registry
Background and Purpose— Stenting is increasingly used as an adjunct to medical therapy in symptomatic intracranial stenoses. High periprocedural adverse event rates are one of the limitations of endovascular treatment. Data from the INTRASTENT multicentric registry should demonstrate in-hospital complications at the current stage of clinical development of the stent procedure.
Methods— Participating centers entered the records of all their consecutive intracranial stent procedures into the database. To determine the clinical outcome in the acute phase, we distinguished transient ischemic attack/nondisabling stroke (modified Rankin Scale <2), disabling stroke, death, and intracranial hemorrhage as clinical complications and analyzed whether they were associated with patient- or stenosis-related risk factors.
Results— Data from 372 patients with 388 stenoses proved 4.8% disabling strokes and 2.2% deaths. Transient or minor events were detected in 5.4% of the cases. Hemorrhagic events (3.5%) occurred more frequently after treatment of middle cerebral artery stenoses (P=0.004) and were associated with significantly higher morbidity and mortality rates. Ischemic strokes by compromise of perforating branches were detected mainly in the posterior circulation. However, the overall rate of severe adverse events was not dependent from location, degree, and morphology of the stenosis or from patient’s age, gender, vascular risk factors, or type of qualifying event.
Conclusion— The complication rates within the registry are within the limits of previously published data. Severe adverse events were equally distributed between potential risk groups with similar rates but different types of main complications in the anterior and posterior circulation.
Intracranial atherosclerosis is the most common cause of cerebral ischemia in the Asian population being detected in 33% of all strokes and in 51% of patients presenting with transient ischemic attack.1,2 The proportion of strokes caused by intracranial stenoses is less pronounced in other ethnicities. Nevertheless, intracranial large artery disease proved to be responsible for 8% of strokes in a cohort study performed in Northern Manhattan.3 In Germany, a large prospective analysis of 4157 patients found intracranial stenosis as the presumed cause of cerebral ischemia in 6.5% of all patients.4 Therefore, the impact of symptomatic intracranial atherosclerosis is relevant not only in Asian, but also European patients with stroke.
Whereas asymptomatic stenoses seem to bear a relatively low risk,5,6 symptomatic intracranial stenoses cannot be regarded as benign lesions. The Warfarin and Aspirin for Symptomatic Intracranial Arterial Disease (WASID) trial showed an overall 1-year risk of stroke of 11% despite medical therapy.7 For a subgroup of patients with >70% luminal narrowing, the risk of stroke in the territory of the affected vessel was 19% within the first year.8 These high rates of stroke recurrence induced the search for alternative therapies.
Intracranial stent placement can be an additional treatment option, but up to now, there is no high-class scientific evidence for the therapeutic effectiveness. INTRASTENT, a European multicentric registry for stent treatment of intracranial stenoses, was supposed to serve as a platform for quality control of participating centers and provided the opportunity to analyze procedure-related complication rates and possible associated risk factors. Our objective here was to report on the technical success and clinical complication rates of stent treatment for symptomatic atherosclerotic stenoses derived from the registry reflecting a “real-world” clinical setting.
Materials and Methods
The INTRASTENT project was approved by the ethics committee of the University of Frankfurt, the local data protection authorities of Hessen, Germany, as well as by the local ethics committees of the participating centers. Eighteen hospitals composed the INTRASTENT study group. The responsible local investigator of each center signed an agreement to include cases consecutively. Any intended treatment of a symptomatic intracranial lesion with a degree of stenosis of >50% qualified for inclusion in the registry. Prospective inclusion from the time of accession was mandatory. Inclusion of retrospective cases was optional but had to be performed consecutively and completely if done. Clinical follow-up was scheduled at 3 and 6 months. Clinical follow-up data exceeding the 6-month interval could be entered on an optional basis. The decision for performing control angiography was left to the discretion of each center.
Data set entry into the Internet-based registry was done online by the participating centers and was checked for plausibility by a local monitor at the University of Frankfurt.
For the current analysis, we considered all elective stent treatments of symptomatic atherosclerotic lesions with a degree of stenosis of at least 50% measured on digital subtraction angiography images. A degree of stenosis >70% was regarded as “high grade” and 50% to 70% as “midgrade.” We omitted treatments in the setting of an acute vessel occlusion or progressive strokes as well as patients with intracranial dissections and inflammatory stenoses.
Baseline clinical data included patient age, gender, and vascular risk factors. The following vascular risk factors were evaluated: history of arterial hypertension, diabetes mellitus, lipid disorders and coronary artery disease, and current smoking habits. The modified Rankin Scale of each patient was recorded before the intervention and on the day of discharge. Information was to be given about the type of symptoms that led to presentation (transient ischemic attack or stroke).
All treated lesions were characterized by location, degree of luminal narrowing, and morphology.
In the anterior circulation, the intracranial carotid artery was divided into petrous, cavernous, and intradural segments. Of the middle cerebral artery (MCA), only M1 stenoses (common trunk to bifurcation) were included. For the posterior circulation, we considered the extraspinal extradural segment (V3), the intradural segment (V4) of the vertebral artery, and the basilar artery (BA). The degree of stenosis was categorized into midgrade (50% to 70%) and high-grade (>70%) lesions. Regarding the morphology, we classified short (<5 mm), moderate (5 to 10 mm), and long (>10 mm) stenoses. The operator had to judge whether the plaque configuration generated a concentric or eccentric lesion.
Antiplatelet pretreatment was performed according to the local protocol of each center based on a double antiplatelet regimen with aspirin and clopidogrel. The choice of stent was left to the discretion of the operator.
A procedure was counted as a technical failure if stent placement was not possible and/or measurement after the intervention revealed a residual stenosis of >50%.
All clinical adverse events associated with the procedure had to be reported. For analysis of the severity of complications, we chose 3 categories. The first category contained transient ischemic attacks and nondisabling strokes with a discharge modified Rankin Scale of <2. Disabling stroke was defined as any stroke leading to a deterioration of the modified Rankin Scale with a score ≥2 at discharge. Imaging was done in case of a clinical event. The choice of imaging method was not prespecified and dependent on the technical resources of each center. Cerebral events were divided in hemorrhagic and nonhemorrhagic. Discrimination between the 2 groups was made by postinterventional imaging. If CT or MRI revealed parenchymal and/or subarachnoid blood, the incident was counted as hemorrhagic irrespective of concomitant or pre-existing brain ischemia.
Statistical analyses of potential risk factors for procedure-related complications were made with the assistance of BiAS.9.02. We used χ2 test with Yates correction or Fisher exact test if the number of expected events was <5 or n<20. Statistical tests were performed for all events as well as patient death and disability and hemorrhagic events.
The predefined inclusion criteria were met by 372 patients with 388 stenoses treated in 374 separate procedures from September 1998 until February 2009 (median year of treatment 2007). Of the 388 data sets, 239 were entered retrospectively and 149 prospectively. The median duration of hospitalization was 4 days (range, 0 to 50 days). Table 1 shows the clinical baseline data. The location and morphological characteristics of the treated lesions are summarized in Tables 2 and 3⇓.
Intracranial stenting was successful in 90.2% of all attempts. Eleven procedures (2.8%) were changed into a balloon angioplasty (percutaneous transluminal angioplasty) alone. Technical failure was not associated with an increased procedure-related event rate (Table 2).
Patient death or disability occurred 26 times (7%) with a patient mortality rate of 2.2% and 4.8% disabling strokes. The rate of nondisabling events was 5.4% (Table 1).
Gender, vascular risk factors, and type of presenting symptoms did not influence the periprocedural event rate (Table 1). Testing for anatomic locations, we did not find significantly increased complication rates for any prespecified site of stenosis. Treating posterior circulation or intradural lesions was not associated with higher complication rates. Also, none of the morphological characteristics of the stenoses showed increased event rates (Table 2).
Of all recorded clinical incidents, 13 (3.5%) were hemorrhagic and 12 of those resulted in patient disability or death. The rate of death and disability was significantly increased in brain hemorrhage compared with all other events (P=0.003). Four of the hemorrhagic events were caused by guidewire perforation and 4 were classified as reperfusion hemorrhages. One patient bled after application of intra-arterial thrombolytics for a periprocedural vessel occlusion, another patient bled into a pre-existent infarction, and there was one vessel rupture. In 2 cases, the etiology of the hemorrhage was unknown. Bleeding events occurred significantly more often in MCA stenoses than in any other location (P=0.004; Table 2).
Of the remaining 33 cerebral events, 27 were ischemic. Of those, 4 were thromboembolic, 9 were due to occlusion of perforating arteries, 4 were caused by stent thromboses, and 10 were not further classified.
Remarkably, all but one perforator ischemic events occurred after treatment of posterior circulation stenoses and all stent thromboses were located in the anterior circulation.
One severe contrast reaction resulted in a disabling stroke.
Five transient cerebral events were not clearly attributable to either ischemia or toxic effects of the contrast agent.
The data of the INTRASTENT registry provided a detailed insight into technical success, complication rates, and patient outcomes after intracranial stent treatment in routine practice throughout 18 participating centers and a high number of 372 patients. Therefore, the data represent a “real-world” setting with a relatively broad application of this method.
Technical success according to the predefined criteria was achieved in 90.2% and therefore was slightly lower than in other reported series with technical success rates ranging from 91.7% to 99%.9,10 This may in part be due to the fact that some operators preferred to change their strategy into a percutaneous transluminal angioplasty rather than forcing the stent procedure. Percutaneous transluminal angioplasty alone was performed in 2.8% of all attempted treatments.
Clinical complication rates in this registry were 4.8% for disabling strokes and patient mortality was 2.2%. In 5.4% of all treatments, we observed clinical symptoms that were transient or did not lead to relevant functional impairment.
The rate of hemorrhagic cerebrovascular events was 3.5% and cerebral bleeding was associated with a significantly increased rate of death and disability. The high morbidity and mortality in this type of complication may in part be explained by the standard double antiplatelet premedication that probably aggravates hemorrhages and therefore the accompanying clinical symptoms. On the other hand, sufficient platelet inhibition is essential to prevent in-stent thrombosis as a source of ischemic stroke and can therefore not be omitted. The frequency of hemorrhages we observed in the registry was in good agreement with a series from Japan reporting a bleeding rate of 3% after endovascular treatment of intracranial stenoses.11
We were not able to define a subgroup with an increased risk for procedure-related morbidity and mortality taking into account patient age, gender, vascular risk factors, and presenting symptoms. Therefore, we cannot propose any of these characteristics as a selection criterion for patients considered for intracranial stenting. Within the US Multicenter Wingspan Registry, stroke as a presenting symptom was associated with a higher rate of cerebrovascular complications. This finding could not be reproduced in the European INTRASTENT registry.12
There was no specific location of stenoses associated with a significantly increased risk of procedure-related morbidity; nevertheless, there was a trend of increased complication rates treating MCA and basilar artery stenoses. An extradural location is often regarded as a “lower-risk lesion” by interventionalists, but the data of the registry do not confirm this assumption. Although overall complication rates, morbidity, and mortality did not differ, we saw some differences in the types of complications occurring at different sites. MCA stenoses were associated with a significantly higher rate of hemorrhagic events. Reaching a stenosis in the MCA can be challenging due to the curved anatomy of the carotid siphon, which could lead to unintentional movement of guidewires carrying the risk of vessel perforation. In addition, all but one ischemic stroke classified as caused by perforator occlusion occurred in the posterior circulation. The mean diameter of perforating branches in the MCA is known to be 472 μm and therefore is substantially larger than perforating branches in the basilar trunk and vertebral artery measuring 391 μm and 243 μm, respectively.13–15 The size of side branches is a known predictor of antegrade flow after stenting of coronary arteries.16 Therefore, the smaller average size of the perforating arteries in the posterior circulation might contribute to the higher arterial branch vulnerability in this anatomical location. Data collection within this registry does not allow to draw definitive conclusions regarding the etiology of cerebrovascular events because a relevant proportion was not precisely classified, but the results might serve to generate new hypotheses for further studies on different types of cerebral events in intracranial stenting procedures.
The length of the stenosis, morphology, and degree of stenosis did not have an impact on all or severe procedure-related complications. With a morbidity and mortality of 7%, a benefit might exist for patients with high-grade stenoses and a high risk of recurrence on medical therapy alone. In patients with a lesser degree of luminal narrowing, the complication rates did not differ significantly from high-grade stenoses. Assuming a 1-year risk of stroke of 7% to 8% in the subgroup with midgrade stenoses on medical therapy alone, caution should be paid offering these patients intracranial stent treatment, because procedure-related death and disability may outbalance the 1-year stroke risk. Our results imply that treatment should be restricted to high-grade lesions with a higher expected risk of recurrent stroke.
Data collection in the registry was not purely prospective and the angiographic results and images of intracranial complications were not validated by a core laboratory. Due to the lack of a prespecified treatment protocol regarding premedication, the procedure itself, and postprocedural management of the patients, some important factors influencing patient outcome might have been missed.
This article is about in-hospital complication rates and does not include the 30-day follow-up data. Therefore, adverse events occurring during this period of time were not considered and 30-day morbidity/mortality is probably higher than the in-hospital complication rates reported here.
Patient morbidity and mortality within the INTRASTENT registry was in good agreement with other reported case series and meta-analyses. Because complication rates treating midgrade and high-grade stenoses did not show significant differences, interventional therapy should be confined to high-grade lesions being at high risk for recurrent stroke with medical therapy alone. We were not able to define any risk factors for all-cause procedure-related death and disability taking into account patient age, gender, vascular risk factors, location of stenosis, length of stenosis, and configuration. Complications seem to be rather inherent to the procedure itself than patient-dependent. Causes of death and disability were very variable and some anatomic locations seem to be more prone to develop certain types of events. MCA stenoses were significantly more often associated with hemorrhagic stroke than any other location and all but one perforator stroke occurred in the posterior circulation. For future studies, we suggest more detailed and prospective analyses of different types of complications to optimize patient management and to better adapt the material with the aim to substantially reduce the risk of the intracranial stent procedure.
W.K. and T.N.-H. received modest speakers honoraria from Micrus Endovascular. J.B. received modest honoraria as a consultant of Micrus Endovascular. B.E. received modest speakers honoraria from Boston Scientific. There are no other conflicts of interest.
Collaborators of the INTRASTENT Study Group: Germany—University of Frankfurt: W. Kurre, S. Kamek, J. Berkefeld, M. Sitzer, T. Neumann-Haefelin, and M. Lorenz; University of Göttingen: M. Knauth, S. Pilgram-Pastor, R. Schramm, and J.H. Buhk; A.K. Altona Hamburg: B. Eckert, and A. Leppien; Asklepios Klinik Barmbek Hamburg: R. Brüning and T. Fitting; University of Hamburg: J. Fiehler and O. Wittkugel; University of Düsseldorf: B. Turowski; University of Dresden: R. von Kummer, D. Mucha, and V. Pütz; University of Kiel: O. Jansen and M. Tietke; Wedau Kliniken Duisburg: F. Brassel, D. Meila, and S. Schotes; University of Mainz: W. Müller-Forell; University Hospital Rechts der Isar Munich: T. Liebig and F. Dorn; University of Marburg: I. Kureck; and Helios Klinikum Erfurt: J. Klisch and V. Sychra; Klinikum Herford: M. Sitzer. Austria—Landesnervenklinik Wagner-Jauregg Linz: J. Trenkler, M. Sonnberger, and H.P. Haring; and University of Graz: G.E. Klein, T. Mikolits, K. Niederkorn, and S. Horner. Great Britain—University of Oxford: W. Kueker. Greece—Papanikolaou General Hospital Thessaloniki: V. Katsaridis. Czechoslovakia—Central Military Hospital Prague: J. Maskova and H. Parobkova.
- Received September 15, 2009.
- Revision received October 27, 2009.
- Accepted November 20, 2009.
Wong KS, Huang YN, Gao S, Lam WM, Chan YL, Kay R. Intracranial stenosis in Chinese patients with acute stroke. Neurology. 1998; 50: 812–813.
Huang YN, Gao S, Li SW, Li JF, Wong KS, Kay R. Vascular lesions in Chinese patients with transient ischemic attacks. Neurology. 1997; 48: 534–525.
Sacco RL, Kargmann DE, Gu Q, Zamanillo MC. Race–ethnicity and determinants of intercranial atherosclerotic cerebral infarction: the Northern Manhattan Stroke Study. Stroke. 1995; 26: 14–20.
Nahab F, Cotsonis G, Lynn M, Feldmann E, Chaturvedi S, Hemphill JC, Zweifler R, Johnston K, Bonovic D, Kasner S, Chimowitz MC; for the WASID Study Group. Prevalence and prognosis of coexistent asymptomatic intracranial stenosis. Stroke. 2008; 39: 1039–1041.
Kasner SE, Chimowitz MI, Lynn MJ, Howlett-Smith H, Stern BJ, Hertzberg VS, Frankel MR, Levine SR, Chaturvedi S, Benesch C, Sila CA, Jovin TG, Romano JG, Cloft HJ; for the Warfarin Aspirin Symptomatic Intracranial Disease (WASID) Investigators. Circulation. 2006; 113: 555–563.
Jiang WJ, Xu XT, Jin M, Du B, Dong KH, Dai JB. Apollo stent for symptomatic atherosclerotic intracranial stenosis: study results. AJNR Am J Neuroradiol. 2007; 28: 830–834.
Suh DC, Kim JK, Choi JW, Choi BS, Pyun HW, Choi YJ, Kim MH, Yang HR, Ha HI, Kim SJ, Lee DH, Choi CG, Hahm KD, Kim JS. Intracranial stenting of severe symptomatic intracranial stenosis: results of 100 consecutive patients. AJNR Am J Neuroradiol. 2008; 29: 781–785.
Nahab F, Lynn MJ, Kasner SE, Alexander MJ, Klucznik R, Zaidat OO, Chaloupka J, Lutsep H, Barnwell S, Mawad M, Lane B, Chimowitz MB; for the NIH Multicenter Wingspan Intracranial Stent Registry Study Group. Risk factors associated with major cerebrovascular complications after intracranial stenting. Neurology. 2009; 72: 1974–1975.