Stenting of Symptomatic M1 Stenosis of Middle Cerebral Artery
An Initial Experience of 40 Patients
Objective— To assess the safety and clinical efficacy of stenting for patients with symptomatic M1 stenosis of middle cerebral artery (MCA), and to assess the significance of classification based on location, morphology, and access of intracranial stenosis (LMA classification) in MCA stenting.
Methods— Forty patients with 42 symptomatic M1 stenoses refractory to medical therapy were enrolled in this study. The lesions were situated at M1 trunk (n=13), M1 origin (n=12), and M1 bifurcation (n=17), respectively, which were classified into type N (nonbifurcation lesions, n=13) and type A (prebifurcation, n=11), B (postbifurcation, n=14), C (lesion across the nonstenotic ostium of its branch, n=1), D (across the stenotic ostium of its branch, n=2), F (combinative lesions of prebifurcation and its small branch ostium, n=1) locations, morphologically into type A (n=15), B (n=23) and C (n=4) lesions, and into type I (mild-to-moderate tortuosity and smooth access, n=17), II (severe tortuosity and/or irregular arterial wall, n=18), and III (excessively severe tortuosity, n=7) accesses.
Results— The technical successful rate was 97.6% for total lesions and 100%, 100%, and 85.7% for types I, II, and III accesses, respectively. The total complication rate was 10%. The mortality was 2.5% (1/40 patients), and 0%, 0%, and 25% for types A, B, and C lesions, respectively. During the median 10 months follow-up, there was no recurrence of transient ischemic attack or stroke in 38 available patients. Among 8 stenting vessels of seven patients with six-month follow-up angiography, 7 showed good patency and one showed restenosis.
Conclusion— Stenting appears to be an effective and feasible therapy for symptomatic M1 stenoses, but also appears to have the higher periprocedural complications, which need strict procedural and periprocedural management to reduce the mortality and morbidity. The LMA classification seems to be helpful to work out the individual therapy and predict the results of stenting. A further study is needed to confirm the benefits of stenting of MCA stenosis.
Atherosclerotic stenosis of the major intracranial arteries is an important cause of ischemic stroke among Asians, Hispanics, and blacks, especially in Chinese populations.1–3 The traditional medical treatment for the symptomatic stenosis of intracranial artery is antiplatelet or anticoagulation therapy.2,4 Despite medical therapy, the stroke risk of the patients with intracranial stenosis remains high.4–6 Different types of surgical bypass, which are technically demanding with significant rates of complications, have been attempted with varying prognoses.5,6 Percutaneous transluminal angioplasty is an alternative to surgical bypass; however, the results are less than satisfactory in most series.7–11 In recent years, more compact and flexible stents have been developed and can be implanted in the intracranial major arteries to reduce the incidence of vessel recoil and restenosis associated with percutaneous transluminal angioplasty alone as well as treating dissection.12–16 However, intracranial stenting is not without risk, despite the potential benefits.17,18 Because of difficulty in tracking, only a few cases of selective stenting of middle cerebral artery (MCA) stenosis have been reported,22 since Gomez et al16 first introduced it in 2000. In this series, we report our initial experiences of stenting in 40 patients with symptomatic M1 stenosis refractory to medical therapy. This is, as yet, the most numerous group of MCA stenting. We also introduce an angiographic classification (based on location, morphology and access [LMA classification]) of intracranial stenosis.
Materials and Methods
Forty patients with the symptomatic M1 stenosis were enrolled between January 2002 and August 2003. The study population included patients with recurrent low-flow transient ischemic attacks (TIAs) or nondisabling ischemic stroke in MCA territory, despite antiplatelet or anticoagulation therapy (refractory to medical therapy), and the corresponding M1 stenosis of ≥50% reduction in diameter proved by angiography, which had evidence of atherosclerotic risk factor or dissection. The diameter reduction was calculated in relation to the adjacent distal normal vessel diameter, analogous to the North America Symptomatic Carotid Endarterectomy Trial method.19 Patients with any of the following were excluded: (1) presence of an intracranial tumor or arteriovenous malformation; (2) severe disability because of stroke or dementia; (3) stroke within 6 weeks; (4) total occlusive lesion; (5) Moya-Moya disease or vasculitis; (6) coexistent ipsilateral ICA stenosis; or (7) inability to give informed consent.
Before the procedure, brain CT, MRI/MRA, transcranial Doppler (sonography or ultrasonography) (TCD), ultrasonography, and diagnostic cerebral angiography were performed in all patients, and CT-perfusion was performed in some patients. Aspirin and either ticlopidine or clopidogrel were administered for at least 3 days. Nimodipine was routinely infused intravenously (0.6 mg to 2 mg/hour) 2 hours before the procedure and lasted 24 hours after stenting. LMA classifications were evaluated before the procedure.
First, we classified lesions according to the locations (to pay attention to ostium features of the larger branches in lesion site) shown in Figure 1: type A, prebifurcation lesion; type B, postbifurcation lesion; type C, lesion across the nonstenotic ostium of its branch; type D, lesion across the stenotic ostium of its branch; type E, ostium lesion of branch alone; type F, the combinative lesions of prebifurcation and its small branch ostium; and type N, nonbifurcation lesion. Ostium lesion of M1 was defined as a lesion <3 mm from vessel ostium, and might be considered as type B location. We usually regarded the largest branch of M1 as M1 “trunk” and then categorized it.
Second, we used morphological classification of the target lesion itself, which was based on the coronary artery classification20 and the cerebral angiography of Mori et al,11,15 as follows: type A, <5 mm in length, concentric or moderately eccentric, smooth stenosis; type B, 5 mm to 10 mm in length, extremely eccentric, or angulated (>45°), or irregular stenosis, or total occlusion (<3 months old); type C, >10 mm in length, extremely angulated (>90°) stenosis, or total occlusion (>3 months old), or lesion with a number of neovascultures all around.
Third, we classified the access (between the guide-catheter and the target lesion) as follows: type I, mild-to-moderate tortuosity and smooth access; type II, severe tortuosity and/or irregular arterial wall; and type III, excessively severe tortuosity.
After placement of 6-French femoral access sheath and angiography of “offending” vessel, a 6-French Envoy guide-catheter (Cordis) attached to saline solution with heparin was placed in the distal cervical segment of internal carotid artery (ICA) over a 0.035-inch, 260-cm exchange wire. Quantitative angiography was then performed for sizing up the stent, using the guide-catheter as reference. Then, a bolus dose of heparin (3000 U), followed by 800 U every hour, was administered intravenously during the procedure.
After induction of general anesthesia, intravenous 5 mg to 6 mg of propofol per kilogram weight per hour, conducted by the neuroanesthesia team, a floppy-tipped 0.014-inch microguide wire, such as 182-cm Choice PT (Boston Scientific Scimed) or 180-cm Wizdom wire (Cordis Neurovascular) guided by a road-map image, was then advanced carefully through the stenosis to M3 or M4 segment of MCA. Assembly of a Prowler14 microcatheter (Cordis Neurovascular) and a 0.014-inch, 300-cm Wizdom wire was used in some cases where negotiating siphon or lesion with single microguide wire was difficult. Then, the coronary balloon stent delivery system was advanced over the microguide wire and accurately positioned across the lesion. The balloon was then inflated gradually at 6 to 9 atm, depending on the type of stent and its location. The stent, usually equal to or slightly less than the diameter of adjacent distal normal vessel and slightly more (1 mm or 2 mm) than the length of lesion, was subsequently deployed. Mainly, 2 kinds of stents were used in this series (Table): BiodivYsio stent (Biocompatibles) and S660 stent (AVE Ireland Limited, Medtronic). After technical success was achieved, defined as ≤20% of residual stenosis, the balloon was withdrawn and the microguide wire was left in the original site for 30-minute observation until general anesthesia was discontinued. After ensuring angiographical patency and evaluating the National Institutes of Health Stroke Scale (NIHSS) score, the microguide wire and guiding catheter were withdrawn (Figure 2).
CT scans were routinely performed immediately after stenting to rule out hemorrhage; blood pressure was controlled under 120/80mm Hg by adjusting the flow rates of nimodipine; TCD examination was routinely done to evaluate poststenting hemodynamic status; heparin treatment was continued for 3 hours, and the arterial sheath was removed 3 hours after discontinuing heparin treatment. The majority of patients were discharged within 7 days with a daily dose of 300 mg of aspirin (either 250 mg of ticlopidine or 75 mg of clopidogrel for 30 days) and 500 mg of probucal (an antioxidation agent) orally, twice a day for 6 months. Modifications of other risk factors for atherosclerosis were continued.
There were 31 men and 9 women ranging in age from 16 to 67 years (≤45, n=23; > 45 years, n=17). The following risk factors of stroke were found in all patients except one (case 16, spontaneous dissection): hyperlipidemia (n=26), smoking (n=25), hypertension (n=15), homocystinemia (n=10), diabetes mellitus (n=5), and other arterial disease (n=7). As such, atherosclerotic stenosis was diagnosed in 39 patients, and spontaneous dissection alone in 1. The patients all experienced recurrent low-flow TIAs (n=29), reversible stroke (n=10), or minor stoke (n=1). The neurological function deficit before the procedure was found in one patient (NIHSS: 3 score, case 16). The time from last attack to stenting was within 24 hours (n=1), 1 to 7 days (n=2), 8 to 30 days (n=19), and >30 days (n=7) for patients with TIAs alone, and >6 weeks for patients with stroke (n=11).
There were a total of 42 M1 stenoses ranging in diameter reduction from 50% to 99%, as shown (Table). Thirteen stenoses (31.0%) were situated at M1 trunk, 17 at M1 bifurcation (40.5%), and 12 at M1 origin (28.6%). According to the location classification, type A, B, C, D, F, and N locations were detected in 11, 14, 1, 2, 1, and 13 lesions, respectively. According to the morphological features of lesion itself, type A, B, and C lesions were 15, 23, and 4, respectively. According to the access classification, type I, II, and III accesses were 17, 18, and 7, respectively.
The placement of microguide wire in the destination of M3 or M4, across lesion, was done successfully in all cases. Single microguide wire technique was applied in 22 lesions. Assembly was done in 20 lesions. The readjustment of micro-guidewire was needed in all type III and some type II accesses.
The technical success rate was 97.6% as a whole (41/42) (100% [17/17], 100% [18/18], and 85.7% [6/7] for type I, II, and III accesses, respectively). A 50% to 75% enhancement in lumen diameter was obtained in 18 lesions, and >75% enhancement was obtained in 23 lesions. The poststenting residual stenosis ranged from 0% to 20% (0 to 10%, n=34; 11%–20%, n=7). Technical failure occurred in one patient (case 29).
The total complication rate was 10% (4/40 patients) during the periprocedural period. Acute occlusion (2.5%) occurred in case 4 within 6 minutes, immediately after stenting. Fortunately, complete patency was obtained by intrathrombus thrombolysis, without any of neurological sequelae.
Subarachnoid hemorrhage (SAH) occurred in 3 patients (7.5%, 3/40) with brain parenchyma hemorrhage (case 16) or without (cases 6 and 17). Two of the subjects did not report headache (asymptomatic hemorrhage), and SAH was detected by immediate CT. Patients 16 and 17 were discharged in 14 and 30 days, respectively, with no increase of NIHSS score compared with the scores before stenting. The remaining case (case 6) was diagnosed as massive SAH by CT, after she reported a severe headache (symptomatic hemorrhage), 2 hours after stenting. Unfortunately, she died 36 hours later despite medical efforts. No occlusion of major branches and deep-perforating arteries was noticed clinically, or on angiogram.
Clinical follow-up was available in 38 patients. During median 10 months of period, ranging from 2 to 21 months, no patient experienced recurrent TIA or stroke again. Angiographical follow-up was obtained in 8 vessels of 7 patients, in which restenosis was found in only one patient (case 20).
Endovascular Therapy of Intracranial Stenosis
The symptomatic intracranial artery stenosis remains an important disease because of its relatively higher incidence among Hispanics, blacks, and Asians, especially in Chinese populations, where there is a lack of effective therapies and poorer prognosis.1–6 Angioplasty has been proposed as a therapeutic approach for symptomatic patients with intracranial artery stenosis since 1980.7 Although the results are encouraging, angioplasty without stenting may produce complex results of dissection, stroke, and restenosis in the presence of improvement of lumen diameter.8–11,21
Artery stenting is a well-established technique for the atherosclerotic lesions of coronary and peripheral arteries, reducing the risks of dissection, acute occlusion, and restenosis related to angioplasty alone. Because of the inability of early stents to track well into the intracranial arteries, stenting of intracranial artery stenosis has not been reported until recently, when the new generation coronary stents with more flexible and compact capabilities were developed.12–18 In 2000, Mori et al15 introduced stenting of the distal ICA, intracranial vertebral artery, and basilar artery stenoses. Because of difficulty in tracking, only a few cases of selective stenting of the symptomatic MCA stenosis have been reported,22 since Gomez et al16 first introduced it in 2000.
The present study of stenting for 42 M1 stenoses in 40 patients has the highest patient number treated. The technical successful rate of stenting was 97.6% for lesions and 97.5% for patients; the case fatality rate was 2.5% and a higher perioperative complications rate (10%) was observed; good short-term results were shown in 38 patients available for follow-up who remained free from TIA or stroke during the median 10-month period. It is noteworthy that young patients are slightly more than those >45 years of age in this series. A further study of long-term results of MCA stenting, which may relate to distribution difference of cerebral artery stenosis among different age groups in Chinese populations, is necessary as the present study is relatively short-term.
Mori’s classification11,15 is helpful in predicting the likelihood of technical and clinical success. However, the technical success of stenting appears more closely related to the circumstances of access; the risk of complications as well as that of restenosis is related to the morphological features of the target lesion itself, while the occlusion risk of larger branches is related to the location of lesion. Based on this observation and concept, the authors put forward the LMA classification as described above, which served as a basis to work out the individual therapy planning.
Lateral projection of the cerebral arteries is used for evaluation of access. In this series, the only technical failure was caused by the inability of the stent to negotiate the extremely tortuous siphon of ICA (case 29, type III access). The successful rate of stenting of type III access (85.7%) would be lower than type II and I (100%). Access classification redounds to determining whether single wire or assembly technique should be used to place microguide wire to the appropriate site quickly as well as safely; for example, single wire for type I access. To avoid stenting failure and artery perforation for type III and some type II access, one should pay attention to the phenomena of withdrawing and “sprinting” of microguide wire at the time of stent delivery and balloon withdrawal. In our series, single wire technique was applied in 52.4% of lesions (12 type I and 10 type II accesses), and in 47.6% of assembly 1 in all type III (n=7), 8 type II, and 5 type I accesses with dissection lesion, type C lesion, or lesion at bifurcation.
Morphology of lesion itself is another important parameter. Assembly technique is necessary for dissection and for some type C lesions to make sure that microguide wire enters the true lumen. Type A and B lesions may be indications of intracranial stenting, while type C should be treated with meticulous caution. In this series, stenting was attempted in only 4 patients with type C lesion, and one of them died of massive SAH. At presentation of absent neurostent for intracranial stenosis, it seems that 1 coronary stent 10 mm in length was preferable and selectively placed in the narrowest part of type C lesion, rather than using one longer coronary stent or two shorter stents in tandem, to obtain technical success and reduce complications as though with suboptimal result, which was used in 3 of 4 cases with type C lesions.
As is well known, lesion location is an important morphological characteristic of coronary atherosclerotic lesion before coronary angioplasty,20 though this point is not discussed in detail by Mori et al.11,15 In fact, atherosclerotic stenoses often involve origin of the major artery. In this series, the incidence of lesion at bifurcation or ostium is 69.0% (29/42). Stenting may result in occlusion of branches due to the atheromatous components squashed into the origin of branches, especially in patients with type D, C, and F locations, resulting subsequently in infarction when the pial collaterals are in short supply. Thus, preprocedural evaluation of collaterals is necessary. The deep-perforating arteries of M1 trunk stenosis should be studied carefully because of dread consequence of their occlusion. In order to prevent occlusion, undersized stent, such as 0.9:1.0 ratio of stent and normal vessel in diameter, and lower inflation pressure, such as 6 to 8 atm, should be adopted. Location classification may help to decide how to place the stent across the major branch. For example, we place the stent from upper branch to left M1 trunk for type D location when patient’s symptoms are episodes of Broca’s aphasia and hemiparesis, and territory of the lower branch of MCA can be supplied by posterior cerebral artery via pial collaterals.
The limited publications18 and our results suggest that the most severe complication of stenting of M1 stenosis seems to be intracranial hemorrhage, which occurred in 3 patients (7.5%). Stenting of M1 stenosis may be a high-risk condition of hyperperfusion syndrome, especially for severe stenosis without good pial collaterals. The cause of hemorrhage may also relate to double stents in tandem, to prior damage to the thinner wall of MCA by the distal end of first stent while delivering the second one, or to artery perforation distal to lesion by the microguide wire. Two of 3 patients with double stents in tandem experienced this complication. From our experience we deduce the following: (1) strict control of blood pressure below 120/80 mm Hg immediately after stenting can prevent hyperperfusion; (2) CT brain scan done immediately after stenting to detect early hemorrhage may prevent continuing anticoagulation and antiplatelet therapy used as a routine; (3) gentle manipulation of microguide wire and deliberative use of double stents in tandem, as well as avoidance of any additional vascular dilator agents that may increase the risk of hemorrhage are essential; (4) it is vital to control blood pressure below 110/70 mm Hg and to neutralize heparin with protamine sulfate, and the emergent salvage should be started immediately, once intracranial hemorrhage is detected. Two of 3 patients with complication of SAH were, thus, saved successfully.
The acute intrastent thrombus formation can be prevented by effective antiplatelet therapy before stenting as well as by performing anticoagulation during the procedure. In this series, all patients receiving aspirin and either ticlopidine or clopidogrel before stenting were free from it, except for 1 who did not receive antiplatelet agents before stenting (case 4). Fortunately, he obtained complete patency by intrathrombus thrombolysis, without any of neurological sequelae.
Restenosis is a constant issue of percutaneous transluminal angioplasty and stenting.15,17 In the present study, restenosis occurred in 1 of 7 patients who remained free from TIA, and who were available at a 6-month angiographic follow-up.
Preliminary results suggest that stenting appears to be an effective and feasible therapy for patients with symptomatic M1 stenoses but who have higher periprocedural complications that need strict procedural and periprocedural management to reduce mortality and morbidity. The LMA classification seems to be helpful in working out the individual therapy and in predicting the results of stenting. What is called for is the development of special neurostents for intracranial stenosis, with excellent flexibility, more compact capabilities, lower profile, and nominative pressure of 5 to 6 atm. Also needed is further study of a larger patient population with long-term clinical and angiographic follow-up, and a subsequently randomized comparison study.
- Received February 3, 2004.
- Accepted February 22, 2004.
Wong KS, Li H, Lam WWM, Chan YL, Kay R. Progression of middle cerebral artery occlusive disease and its relationship with further vascular events after stroke. Stroke. 2002; 33: 532–536.
Arenillas JF, Molina CA, Montaner J, Abilleira S, Gonzalez-Sanchez MA, Alvarez-Sabin J. Progression and clinical recurrence of symptomatic middle cerebral artery stenosis: a long-term follow-up transcranial doppler ultrasound study. Stroke. 2001; 32: 2898–2904.
Warfarin-Asprin Symptomatic Intracranial Disease (WASID) Study Group. Prognosis of patients with symptomatic vertebral or basilar artery stenosis: the Warfarin-Asprin Symptomatic Intracranial Disease (WASID) Study Group. Stroke. 1998; 29: 1389–1392.
McKenzie JD, Wallace RC, Dean BL, Flom RA, Khayata MH. Preliminary results of intracranial angioplasty for vascular stenosis caused by atherosclerosis and vasculitis. AJNR Am J Neuroradiol. 1996; 17: 263–268.
Takis C, Kwan ES, Pessin MS, Jacobs DH, Caplan LR. Intracranial angioplasty: experience and complications. AJNR Am J Neuroradiol. 1997; 18: 1661–1668.
Mori T, Fukuoka M, Kazita K, Mori K. Follow-up study after intracranial percutaneous transluminal cerebral balloon angioplasty. AJNR Am J Neuroradiol. 1998; 19: 1525–1533.
Al-Mubarak N, Gomez CR, Vitek JJ, Roubin GS. Stenting of symptomatic stenosis of the intracranial internal carotid artery. AJNR Am J Neuroradiol. 1998; 19: 1949–1951.
Dorros G, Cohn JM, Palmer LE. Stent deployment resolves a petrous carotid artery angioplasty dissection. AJNR Am J Neuroradiol. 1998; 19: 392–394.
Mori T, Kazita K, Chokyu K, Mima T, Mori K. Short-term arteriographic and clinical outcome after cerebral angioplasty and stenting for intracranial vertebrobasilar and carotid atherosclerotic occlusive disease. AJNR Am J Neuroradiol. 2000; 21: 249–254.
Gomez CR, Misra VK, Campbell MS, Soto RD. Elective stenting of symptomatic middle cerebral artery stenosis. AJNR Am J Neuroradiol. 2000; 21: 971–973.
Ryan TJ, Bauman WB, Kennedy JW. Guidelines for percutaneous transluminal coronary angioplasty. A report of the American Heart Association/American College of Cardiology Task Force on Assessment of Diagnostic and Therapeutic Cardiovascular Procedures (Committee on Percutaneous Transluminal Angioplasty). Circulation. 1993; 88: 2987–3007.
Sam Shin Y, Yong Kim S, Young Bang O, Jeon P, Han Yoon S, Hong Cho K, Gi Cho K. Early experiences of elective stenting for symptomatic stenosis of the M1 segment of the middle cerebral artery: reports of three cases and review of the literature. J Clin Neurosci. 2003; 10: 53–59.