| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Stroke. 2007;38:1274.)
© 2007 American Heart Association, Inc.
Original Contributions |
From the Department of Neurology (A.C.F., W.S.S.), University of California, San Francisco, San Francisco, Calif.; the Departments of Radiology (G.R.D.) and Neurology (D.S.L.), University of California, Los Angeles, Los Angeles, Calif.; and the Department of Radiology (R.F.B.), Riverside Methodist Hospital, Columbus, Ohio.
Correspondence to Alexander C. Flint, MD, PhD, University of California, San Francisco, Neurovascular Service and Neurocritical Care, 505 Parnassus Ave, San Francisco CA 94143-0114. E-mail alex.flint{at}ucsfmedctr.org
| Abstract |
|---|
|
|
|---|
Methods All patients with acute stroke from intracranial ICA occlusion were identified in the MERCI and Multi MERCI Part I trials. We determined the success rate of ICA recanalization with endovascular thrombectomy and then assessed clinical outcomes according to whether vessel recanalization was successful.
Results Eighty patients with acute stroke from intracranial ICA occlusion were identified. Of these 80 patients, 53% had successful ICA recanalization with the Merci Retriever alone and 63% had ICA recanalization with use of the Merci Retriever plus adjunctive endovascular treatment. Baseline patient characteristics and procedural complications did not differ between the recanalized and nonrecanalized groups. Good clinical outcome, defined by a modified Rankin Scale of 0 to 2 at 90 days, occurred in 39% of patients with ICA recanalization (n=19 of 49) and in 3% of patients without ICA recanalization (n=1 of 30) (P<0.001; one patient was lost to follow up for 90-day modified Rankin Scale). Ninety-day mortality was 30% (n=15 of 50) in the recanalized group and 73% (n=22 of 30) in the nonrecanalized group (P<0.001). Symptomatic hemorrhage was not significantly different between the recanalized (6% [n=3 of 50]) and nonrecanalized (16.7% [n=5 of 30]) groups (P=0.14). Hemorrhage rates were also not found to be influenced by use of intravenous thrombolysis before mechanical thrombectomy. Multivariable logistic regression identified ICA recanalization (OR=28.4, 95% CI=2.6 to >99.9) and lack of history of hypertension (OR=0.15, 95% CI=0.04 to 0.57) as significant predictors of a good 90-day outcome. Failure to recanalize the ICA (OR=0.16, 95% CI=0.05 to 0.51) and age (per decade, OR=1.07, 95% CI=1.03 to 1.13) were significant predictors of mortality at 90 days.
Conclusions Mechanical thrombectomy of acute intracranial ICA occlusion using the Merci Retriever device, alone or in combination with adjunctive endovascular therapy, has a high rate of successful vessel recanalization. Subjects with successful ICA recanalization by this method have improved poststroke clinical outcome and survival compared with subjects in which the ICA is not successfully recanalized.
Key Words: acute stroke endovascular treatment internal carotid artery thrombectomy
| Introduction |
|---|
|
|
|---|
Because of the poor prognosis and poor response of ICA occlusion to intravenous (IV) thrombolysis, various procedural approaches to acute recanalization have been explored, including emergency carotid endarterectomy for acute extracranial occlusions,4 balloon angioplasty,5 intraarterial (IA) thrombolysis and combined IV/IA thrombolysis,2,610 IA thrombolytic plus balloon angioplasty,11 an endovascular microsnare device,12 and mechanical clot disruption plus thrombolytics.13,14 The recent clearance in the United States of an endovascular thrombectomy device, the Merci Retrieval System (Concentric Medical, Inc., Mountain View, Calif.), raises the possibility that mechanical thrombectomy might have a role in the treatment of acute ICA occlusions.
We therefore pooled data from the MERCI trial and Part I of the Multi MERCI trial to explore the clinical outcomes in patients with acute stroke from intracranial ICA occlusion. We describe the safety and efficacy of mechanical thrombectomy in 80 patients with intracranial ICA occlusion and report 3-month clinical outcomes in these patients according to recanalization status.
| Methods |
|---|
|
|
|---|
Patients and Techniques
MERCI was a single-arm multicenter trial. The study tested whether a mechanical thrombectomy device could safely restore vascular patency at a rate exceeding a prespecified rate of spontaneous recanalization in patients presenting within 8 hours of onset of an acute ischemic stroke.15 Multi MERCI is a single-arm, multicenter trial that included both IV tissue plasminogen activator (t-PA)-ineligible and IV t-PA-treated patients. Patient eligibility in the IV t-PA ineligible arm of Multi MERCI was the same as for MERCI.15,16 Eligibility in the IV t-PA-treated arm was the same as in the IV t-PA-ineligible arm except that patients who had received t-PA within 3 hours of onset under U.S. Food and Drug Administration-labeled indications could be enrolled if t-PA failed to open the intracranial large vessel as determined by angiography. Details of patient enrollment in MERCI and Multi MERCI, including inclusion and exclusion criteria, have been previously described.15,16
Thrombectomy Procedure
Details of the devices and procedures used in MERCI15,17 and Multi MERCI16 have been previously published. Up to six device passes were allowed. Successful recanalization was defined as achieving at Thrombolysis In Myocardial Infarction18 II or III flow in all treatable vessels, as in PROACT-II19 and IMS.9 Successful recanalization for intracranial ICA lesions required the ICA, M1, and M2 branches to have flow restored to at least Thrombolysis In Myocardial Infarction II. Intraarterial thrombolytics were allowed in cases of treatment failure with the device after six passes or to treat distal embolus not accessible to the device after successful proximal thrombectomy. Angioplasty and stenting of any lesion were not allowed.
Clinical Variables and Measurement of Outcome
Measurement of baseline clinical variables and clinical outcomes was similar in MERCI and Multi MERCI, as previously described.15,16 Computed tomography brain imaging was performed at 24 hours or for any decline in neurologic status. Symptomatic intracranial hemorrhage was defined as a 4 or more point worsening of the National Institutes of Health Stroke Scale (NIHSS) score within 24 hours with any blood products identified on head computed tomography scan. All computed tomography scans were reviewed in a core laboratory. Hemorrhages were adjudicated by the Data Safety Monitoring Board (DSMB).
Primary outcomes were recanalization and device-related complications. Device-related complications were defined as vascular perforation, arterial dissection, or embolization of an uninvolved territory. Clinically significant procedure complications were defined as a procedure complication with worsening of NIHSS by
4, any subarachnoid hemorrhage, or any groin complication requiring surgery or blood transfusion. Secondary outcomes included the modified Rankin Scale (mRS) and NIHSS at 30 and 90 days. Good neurologic outcome was prespecified as mRS
2.
Statistical Analysis
Categorical data in contingency tables were analyzed by the Fisher exact test. Continuous data were assessed for normality by the Kolmogorov-Smirnov test; normally distributed continuous data were analyzed by t test, and continuous data not passing this normality test were compared using the Mann-Whitney test. The percentage of patients with successful recanalization with mechanical thrombectomy was compared with the benchmark rate of 18% spontaneous middle cerebral artery (MCA) recanalization from the control arm of PROACT II20 using the binomial test. Univariable analyses were separately performed modeling good outcome (mRS=0 to 2) at 90 days or mortality at 90 days. All variables with P<0.2 in univariable analysis were entered into separate multivariable logistic regression models of good outcome (mRS=0 to 2) at 90 days and mortality at 90 days. Models were built using forward/backward stepwise logistic regression with variables entered into the model at the 0.05 significance level and removed at the 0.10 significance level. Statistical analyses were performed using STATA and SAS.
| Results |
|---|
|
|
|---|
Recanalization of the intracranial ICA was achieved in 53% with the Merci Retriever alone (n=42 of 80). The 53% rate of ICA recanalization after use of the Merci Retriever alone is significantly higher than the 18% benchmark rate15,20 (P<0.001, two-tailed binomial test). Sixty-three percent (n=50 of 80) had vessel recanalization after both Merci Retriever use and adjunctive treatment with IA thrombolytics. The remaining 37% (n=30 of 80) did not have recanalization of the occluded ICA by the end of the procedure. Of the 38 ICA occlusions that did not open with the Merci Retriever alone, 16 received adjunctive IA thrombolysis, and of these, eight went on to recanalize with adjunctive treatment. Adjunctive IA thrombolytics were used in a total of 23 patients: 16 of 38 patients with vessel occlusions that failed to recanalize with the Merci Retriever and seven of 42 patients in which the Merci Retriever produced at least Thrombolysis In Myocardial Infarction grade II flow at the occlusion site and IA thrombolysis was used to treat residual clot burden at the occlusion site or more distally. The adjunctive thrombolytic used was t-PA in 21 of 23 (mean dose=13.2±8.9 mg IA), urokinase in one of 23 (250 000 IU IA), and tenecteplase in one of 23 (5 mg IA). Among the patients with ICA occlusion who were enrolled in Multi MERCI after failed IV thrombolysis, eight of 11 (73%) had successful recanalization by the end of the procedure (Figure 1).
|
Baseline patient characteristics for the overall patient population and for the successful recanalization and failed recanalization groups are shown in Table 1. No significant differences in baseline characteristics were detected between the recanalized and nonrecanalized groups.
|
Good clinical outcome, as determined by a mRS of 0 to 2 at 90 days, occurred more often in patients who had successful ICA recanalization (Table 2; Figure 2). Among patients with postprocedure ICA recanalization, 39% (n=19 of 49) had a mRS of 0 to 2 at 90 days, whereas 3% (n=1 of 30) of patients with failed recanalization had an mRS of 0 to 2 at 90 days (P<0.001). Among survivors at 90 days, an mRS of 0 to 2 was seen in 55.9% (n=19 of 34) of the recanalized group and 12.5% (n=1 of 8) of the nonrecanalized group (P=0.047). Overall, 25% of the ICA occlusion patients in MERCI and Multi MERCI Part I (n=20 of 79) had a good neurologic outcome at 90 days poststroke (mRS=0 to 2). One patient in the recanalized group was lost to follow up for assessment of the mRS.
|
|
There was also a difference in 90-day mortality observed between the recanalized and nonrecanalized groups. At 90 days poststroke, 30% (n=15 of 50) of the recanalized patients were dead, whereas 73% (n=22 of 30) of the nonrecanalized patients were dead (P<0.001). The overall mortality was 46% (n=37 of 80) at 90 days poststroke.
In contrast, there was no significant difference in the rate of symptomatic intracranial hemorrhage between the recanalized and nonrecanalized groups. Symptomatic hemorrhage was observed in 6% (n=3 of 50) of the patients with ICA recanalization and in 16.7% (n=5 of 30) of the patients without recanalization (P=0.14). The symptomatic hemorrhage rate was 10% (n=8 of 80) for the overall population, 9.1% (n=1 of 11) for the subgroup of patients who received IV thrombolysis before endovascular treatment, and 10.1% (n=7 of 69) among patients who did not receive IV thrombolysis (P>0.99). Symptomatic hemorrhage occurred in one of eight patients who failed IV t-PA but went on to have successful ICA recanalization after thrombectomy, whereas none patients who failed IV t-PA and failed to recanalize with thrombectomy had symptomatic hemorrhage (P>0.99). In the six patients given IV t-PA but no adjunctive IA thrombolysis, there were no symptomatic hemorrhages. Similarly, none of the four patients who received IA t-PA after IV t-PA had a symptomatic hemorrhage. Of the 23 patients who received adjunctive IA thrombolysis, one patient had a symptomatic hemorrhage; this patient received IA tenecteplase after receiving IV t-PA before attempted thrombectomy. Of the 51 patients who received no thrombolytics, either IV or IA, seven (13.7%) had symptomatic hemorrhage. Of the eight patients with symptomatic hemorrhage in the total patient population, seven of eight (87.5%) died by 90 days poststroke compared with 30 of 72 (41.7%) patients without symptomatic hemorrhage who died by 90 days (P=0.02).
Of the eight symptomatic hemorrhages in the total patient population (n=80), three were associated with a vessel perforation, and all three were in the nonrecanalized group. Therefore, 3.8% of all treated patients (n=3 of 80) had symptomatic hemorrhage from a procedural complication; all three of these patients died. Two other clinically significant procedural complications occurred. One patient had an intracranial vessel dissection associated with the balloon guide catheter without symptomatic hemorrhage. This patient died between the 30- and 90-day visits. Another patient had a complication of femoral artery access that required surgical repair. One procedural complication occurred that was adjudicated to be not clinically significant in which an intracranial dissection occurred without symptomatic hemorrhage. This patient was alive with a good neurologic outcome (mRS=1) at 90 days. Of the five clinically significant adverse events in the total patient population, two occurred among the 50 recanalized patients and three occurred among the 30 patients who did not have ICA recanalization (P=0.36). When the five patients with significant adverse events are excluded, the relationships between recanalization and good outcome and recanalization and mortality are not affected.
To determine which baseline patient variables and trial-related variables impacted outcome and survival, we performed regression analysis on the outcomes of 90-day mRS 0 to 2 and 90-day mortality. In multivariable logistic regression analysis, lack of history of hypertension and ICA recanalization significantly predicted good outcome (mRS=0 to 2) at 90 days, and NIHSS had a borderline significant effect on good outcome (Table 3). Failure of ICA recanalization and age group by decade significantly predicted 90-day mortality (Table 3). Thus, successful ICA recanalization remained a positive predictor of 90-day good outcome (mRS=0 to 2) and a negative predictor of 90-day mortality when other variables were held constant in multivariable regression. When the same models were built without the five patients with significant adverse events, lack of history of hypertension, ICA recanalization, and NIHSS still predicted good outcome (mRS=0 to 2) at 90 days, and recanalization failure and age group by decade still predicted mortality at 90 days (data not shown).
|
| Discussion |
|---|
|
|
|---|
The rate of successful ICA recanalization with the Merci Retriever was high (Figure 1), but there are no angiographic control data to establish the rate of spontaneous ICA recanalization in acute stroke. The 18% angiographically confirmed spontaneous recanalization of proximal MCA occlusions in the control arm of PROACT-II was used as a benchmark rate15 and may overestimate spontaneous ICA recanalization.3 In one recent series of patients with intracranial ICA occlusion studied with transcranial Doppler, 17% of patients not treated with thrombolytics showed complete or partial recanalization in the first 24 hours.21
The rate of ICA recanalization with the Merci device compares favorably with case series of thrombolysis for ICA occlusion. ICA recanalization in one study of intravenous t-PA occurred in 26%.3 The rate of recanalization of tandem MCA occlusions was 59% (complete in 35%, partial in 24%).3 Larger series of isolated proximal MCA occlusions support this discrepancy between middle cerebral artery and ICA recanalization rates.22 In another series of patients comparing IV t-PA treatment in patients with ICA or MCA occlusions, there was also a much lower rate of ICA recanalization (31%) compared with MCA recanalization (88%) at 3 days after t-PA.1 In two studies that used angiography after IV t-PA administration, only approximately 8% of patients with intracranial ICA occlusion showed early recanalization with t-PA.23,24 One study of patients with intracranial ICA occlusion followed by transcranial Doppler found that 12 of 18 patients treated with IV t-PA showed recanalization,21 a much higher rate of recanalization than reported elsewhere. Among patients who received IA urokinase for acute ICA occlusions, recanalization was seen in 42%.8 In a series reporting treatment of ICA occlusions with intravenous t-PA, IA t-PA, or IA urokinase, the rate of recanalization was low (12.5%).2 In another small series using IA thrombolysis, 63% of ICA occlusions had partial recanalization and 46% had complete recanalization.7 Although few patients with intracranial ICA occlusion were reported in the IMS "bridging" trial (IV+IA t-PA), successful ICA recanalization was seen in nine of 20 patients.9 In a series of 18 patients treated with either IA or combined IV/IA thrombolytics, complete recanalization was reported in 62% of patients receiving IA therapy alone and in 80% of patients receiving combined IV/IA thrombolysis.6 In the same report, a summary of previous small ICA series reported a combined IV/IA rate of recanalization of 69% (nine of 13 patients) and a 10% to 38% rate of recanalization with IA treatment alone.6 The combination of mechanical clot disruption with IA thrombolysis was found to produce recanalization in nine of 11 patients in one series.14
The literature on the natural history of ICA occlusion is limited, but it is clear that untreated ICA stroke carries a poor prognosis, except when the presenting stroke symptoms are mild and the degree of collateral flow is robust. Among 40 patients with ICA occlusion, the majority (22 of 40) had a severe stroke, and nearly all of those with a severe stroke were significantly disabled (nine of 22 [41%]) or dead (12 of 22 [55%]) at 2-year follow up.25 In a series of 179 patients with clinically diagnosed carotid occlusion, mortality appeared to vary based on alteration of consciousness at presentation (41% of patients with abnormal level of consciousness died compared with 2% of those with normal-level consciousness).26
Clinical outcomes in patients with ICA occlusions treated with thrombolytics are also poor in most series. In one small series, treatment of angiographically confirmed intracranial ICA occlusion with either IV or IA thrombolysis resulted in an outcome of severe neurologic disability or death in 80%.2 A series of patients with transcranial Doppler-documented intracranial ICA occlusion treated with IV t-PA reported an outcome of severe disability or death in 40%.3 In a small series of patients treated with IA urokinase, patients with distal ICA occlusions had worse outcomes compared with other sites of occlusion.27 Among 18 patients treated with IA and IV thrombolytics, the 90-day mortality rate was 50%, and symptomatic intracranial hemorrhage was seen in 22%.6 In another study, 17% had a good outcome at 90 days (mRS=
2) and 42% died by 90 days.7 However, better outcomes were reported in a small series of ICA patients treated with combined IA urokinase and mechanical clot disruption with good outcomes in four of 11 patients and mortality in only one of 11.14
The MERCI and Multi MERCI trials have not included a nontreatment arm. Thus, these data cannot directly demonstrate that mechanical thrombectomy of acute intracranial ICA occlusions is superior to nontreatment. However, our data do show that the Merci device, together with adjunctive therapy, allows for a high rate of ICA recanalization and that ICA recanalization is associated with better clinical outcome and survival. The baseline patient characteristics did not differ between those who were successfully recanalized and those who were not, and therefore it is unlikely that confounding biased our results. Removal of the five patients with clinically significant adverse events from our analyses did not alter the results, so procedure-related complications in the nonrecanalized group cannot explain the apparent effect of recanalization on good outcome or mortality. In multivariable logistic regression, ICA recanalization was a positive predictor of 90-day good outcome and negative predictor of 90-day mortality. Interestingly, the time from stroke onset to procedure was not an independent predictor of either outcome measure (but all patients were treated in the <8-hour time window). Because the rate of ICA recanalization with the Merci Retriever far exceeds the expected rate of spontaneous early recanalization, mechanical thrombectomy for acute ICA occlusion <8 hours from stroke onset in an appropriately selected patient population is likely to be clinically beneficial.
Further study is needed to determine the effect of ICA recanalization on outcome in subsets of patients with particular ICA occlusion patterns (ICA versus ICA/MCA, ICA/MCA/anterior cerebral artery, or "T occlusion") and according to the degree of collateral blood flow before recanalization. Given that adjunctive use of thrombolytics with attempted thrombectomy appears to be safe in this population, selective adjunct use of thrombolytics should be considered, for example, to recanalize the carotid terminus to improve collateral flow should thrombectomy fail to recanalize the ICA. In light of the high recanalization rate of mechanical thrombectomy for ICA occlusion, a comparison of mechanical thrombectomy to intravenous thrombolysis in the <3-hour time window is warranted.
| Acknowledgments |
|---|
A.C.F. has nothing to disclose; G.R.D. has stock ownership in Concentric Medical; R.F.B. is a consultant for Concentric Medical; D.S.L. is a consultant for Concentric Medical; W.S.S. was the Principal Investigator, MERCI and Multi MERCI trials and has stock ownership in Concentric Medical.
Received September 19, 2006; revision received November 8, 2006; accepted November 15, 2006.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
O. Ozdemir, M. Bussiere, A. Leung, I. Gulka, D. Lee, R. Chan, J.D. Spence, and D. Pelz Intra-Arterial Thrombolysis of Occluded Middle Cerebral Artery by Use of Collateral Pathways in Patients with Tandem Cervical Carotid Artery/Middle Cerebral Artery Occlusion AJNR Am. J. Neuroradiol., September 1, 2008; 29(8): 1596 - 1600. [Abstract] [Full Text] [PDF] |
||||
![]() |
E.S. Rosenthal, L.H. Schwamm, L. Roccatagliata, S.B. Coutts, A.M. Demchuk, P.W. Schaefer, R.G. Gonzalez, M.D. Hill, E.F. Halpern, and M.H. Lev Role of Recanalization in Acute Stroke Outcome: Rationale for a CT Angiogram-Based "Benefit of Recanalization" Model AJNR Am. J. Neuroradiol., September 1, 2008; 29(8): 1471 - 1475. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Gruber Interventional Management of Stroke Stroke, June 1, 2008; 39(6): 1663 - 1664. [Full Text] [PDF] |
||||
![]() |
T.N. Nguyen, S. Lanthier, and D. Roy Iatrogenic Arterial Perforation during Acute Stroke Interventions AJNR Am. J. Neuroradiol., May 1, 2008; 29(5): 974 - 975. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. J. Quinn and K. R. Lees Advances in Emerging Therapies 2007 Stroke, February 1, 2008; 39(2): 255 - 257. [Full Text] [PDF] |
||||
![]() |
D. M. Pelz, E. I. Levy, and L. N. Hopkins Advances in Interventional Neuroradiology 2007 Stroke, February 1, 2008; 39(2): 268 - 272. [Full Text] [PDF] |
||||
![]() |
H. B. van der Worp and J. van Gijn Acute Ischemic Stroke N. Engl. J. Med., August 9, 2007; 357(6): 572 - 579. [Full Text] [PDF] |
||||
![]() |
H. P. Mattle Intravenous or Intra-Arterial Thrombolysis?: It's Time to Find the Right Approach for the Right Patient Stroke, July 1, 2007; 38(7): 2038 - 2040. [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Stroke Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2007 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |