Abstract
BACKGROUND AND PURPOSE: Periprocedural intracranial hemorrhage is one of common complications after stent placement for symptomatic intracranial atherosclerotic stenosis. This study was conducted to demonstrate predictors and long-term outcomes of periprocedural intracranial hemorrhage after stent placement for symptomatic intracranial atherosclerotic stenosis.
MATERIALS AND METHODS: We retrospectively analyzed patients with symptomatic intracranial atherosclerotic stenosis stent placement in a prospective cohort at a high-volume stroke center. Clinical, radiologic, and periprocedural characteristics and long-term outcomes were reviewed. Periprocedural intracranial hemorrhage was classified as procedure-related hemorrhage (PRH) and non-procedure-related hemorrhage (NPRH). The long-term outcomes were compared between patients with PRH and NPRH, and the predictors of NPRH were explored.
RESULTS: Among 1849 patients, 24 (1.3%) had periprocedural intracranial hemorrhage, including PRH (4) and NPRH (20). The postprocedural 30-day mRS was 0−2 in 9 (37.5%) cases, 3−5 in 5 (20.8%) cases, and 6 in 10 (41.7%) cases. For the 14 survivors, the long-term (median of 78 months) mRS were 0−2 in 10 (76.9%) cases and 3−5 in 3 (23.1%) cases. The proportion of poor long-term outcomes (mRS ≥3) in patients with NPRH was significantly higher than those with PRH (68.4% versus 0%, P = .024). Anterior circulation (P = .002), high preprocedural stenosis rate (P < .001), and cerebral infarction within 30 days (P = .006) were independent predictors of NPRH after stent placement.
CONCLUSIONS: Patients with NPRH had worse outcomes than those with PRH after stent placement for symptomatic ICAS. Anterior circulation, severe preprocedural stenosis, and recent infarction are independent predictors of NPRH.
ABBREVIATIONS:
- CASSISS
- China Angioplasty & Stenting for Symptomatic Intracranial Severe Stenosis Study
- ICAS
- intracranial atherosclerotic stenosis
- NPRH
- non-procedure-related hemorrhage
- PRH
- procedure-related hemorrhage
- SAMMPRIS
- Stenting versus Aggressive Medical Management for Preventing Recurrent Stroke in Intracranial Stenosis
- VISSIT
- Vitesse Intracranial Stent Study for Ischemic Therapy
SUMMARY
PREVIOUS LITERATURE:
With experienced interventionalists and proper patient selection, patients with symptomatic intracranial atherosclerotic stenosis (ICAS) at high risk of stroke recurrence for whom medical treatment failed may still benefit from endovascular treatment. Periprocedural intracranial hemorrhage is a common complications after stent placement for symptomatic ICAS. Compared with ischemic events after intracranial stent placement, studies focusing on hemorrhagic complications were rare. Previous studies have reported 30-day outcomes of periprocedural intracranial hemorrhage, but long-term outcomes remain unclear.
KEY FINDINGS:
This study demonstrated a low rate of periprocedural intracranial hemorrhage (1.3%), and most occurred within 24 hours after stent placement for symptomatic ICAS. Patients with non-procedure-related hemorrhage (NPRH) had worse outcomes than those with procedure-related hemorrhage. Anterior circulation, severe preprocedural stenosis, and recent infarction are independent predictors of NPRH.
KNOWLEDGE ADVANCEMENT:
Based on the findings of this study, for patients with severe anterior circulation stenosis and infarction within 30 days, submaximal or staged angioplasty or stent placement may be considered in prevention of periprocedural hemorrhage.
Intracranial atherosclerotic stenosis (ICAS) is one of the major causes of ischemic stroke worldwide, especially in Asian populations. Endovascular treatment has not been regarded as standard protocol for patients with severe symptomatic ICAS due to the high risk of ischemic and hemorrhagic complications.1,2 However, with experienced interventionalists and proper patient selection, the Wingspan Stent System Post Market Surveillance and the Wingspan 1-year Vascular Events and Neurologic Outcomes studies had demonstrated low rates of perioperative complication (2.6%) and long-term recurrent stroke (8.5%) for ICAS stent placement.3,4 Therefore, patients with symptomatic ICAS (70%–99%) or occlusion at high risk of stroke recurrence who failed to respond to medical treatment may still benefit from endovascular treatment.5⇓⇓-8 Compared with ischemic events after intracranial stent placement, studies focusing on hemorrhagic complications were rare.9⇓-11 The periprocedural hemorrhagic stroke rate in the Stenting versus Aggressive Medical Management for Preventing Recurrent Stroke in Intracranial Stenosis (SAMMPRIS) study was 6.3% (13/208),12,13 and in the Vitesse Intracranial Stent Study for Ischemic Therapy (VISSIT) study was 8.6% (5/58).14 However, the long-term outcomes in patients with periprocedural hemorrhagic stroke remains uncertain. The mechanisms of periprocedural intracranial hemorrhage include procedure-related hemorrhage (PRH) and non-procedure-related hemorrhage (NPRH). The PRH occurs due to vessel perforation or rupture during the manipulation of devices, while the NPRH occurs mainly due to postprocedural hyperperfusion and antithrombotic treatment.12,13,15 The differences in the long-term outcomes with variable causes (PRH versus NPRH) also remain unclear. In addition, the predictors of NPRH remain to be investigated.
Considering that endovascular treatment remains one of the treatment choices for patients with symptomatic ICAS who fail to respond to aggressive medical treatment, we conducted this study to evaluate the long-term outcomes of periprocedural intracranial hemorrhage with variable causes, and explore predictors of NPRH after stent placement for symptomatic ICAS.
MATERIALS AND METHODS
Study Design
This was a retrospective analysis of patients with symptomatic ICAS stent placement in a prospective cohort at a high-volume stroke center. The study was performed according to the guidelines from the Helsinki Declaration, and approved by the institutional review board of Beijing Tiantan Hospital (Approval number: KY2022-077–02). All patients or their legally authorized guardians were fully informed during hospitalization that the patient's data might be used for research purposes and written informed consents were obtained. An independent data and safety monitoring board oversaw the conduction, safety, and efficacy of the study. The data that support the findings of this study are available from the corresponding author upon reasonable request.
Study Population, Endovascular Treatment Strategy, Periprocedural Management, and Medical Treatment
Due to the word limit, the details of study population, endovascular treatment strategy, periprocedural management, and medical treatment are presented in the Online Supplemental Data.
Definition of Hemorrhage Causes
Intraprocedural hemorrhage (parenchymal or subarachnoid hemorrhage) adjacent to the target artery was defined as PRH, ascribed to inadvertent vessel perforation or rupture during the manipulation of endovascular devices, including microwire, microcatheter, intermediate catheter, balloon, and stent.
Intraprocedural or postprocedural hemorrhage in the distribution area of the stented artery, excluding iatrogenic injury, with remarkable hyperperfusion in the distribution of the stented artery indicated by CT perfusion imaging, was defined as hyperperfusion hemorrhage. However, some patient’s conditions were critical after periprocedural hemorrhage and CT perfusion was difficult to perform in clinical practice, and possible hyperperfusion hemorrhage was defined based on the elevated flow velocity distal the stented artery indicated by transcranial Doppler (either >30% compared with contralateral artery for anterior circulation with normal contralateral vessels or >50% compared with preprocedural blood flow distal to the stented segment for anterior circulation with abnormal contralateral vessels and posterior circulation) (Fig 1). Postprocedural hemorrhage beyond the distribution area of the stented artery and complicated with significant coagulation dysfunction was regarded as antithrombosis-related hemorrhage. The hyperperfusion hemorrhage, possible hyperperfusion hemorrhage, antithrombosis-related hemorrhage, and hemorrhage with uncertain causes were uniformly classified as NPRH.
A patient presented with progressive weakness in the right limbs for 28 days. The DWI of preprocedural MR) revealed subacute infarction in the right internal watershed area (A). Preprocedural CT perfusion showed extensive hypoperfusion in the right intracranial ICA distribution area, and CBF decreased significantly (B). Preprocedural DSA showed severe stenosis in the terminal segment of the right ICA, with a TICI scale of grade 2a (C). The lesion was treated with a 2.5 × 8 mm balloon-mounted Apollo stent (MicroPort Neuro Tech), with a residual stenosis rate of 10% (D). The CT scan immediately after the procedure showed slightly high attenuation in the right frontal lobe and was considered to be leakage of contrast media (E). The transcranial Doppler indicated elevated flow velocity in the right middle cerebral artery. The patient developed lethargy, slurred speech, and aggravation of left limb weakness 5 hours after the procedure, and a re-examination of head CT revealed cerebral hemorrhage in the right frontal lobe (F). After that, antiplatelet and anticoagulant drugs were stopped, and the CT scan 10 hours after the procedure revealed hemorrhage expansion (G). The CT scan 48 hours after the procedure showed that the hematoma had expanded again. The legal representative refused hematoma removal, and the patient died of respiratory and circulatory failure 7 days after the procedure (H).
Follow-Up and Long-Term Outcomes
All patients were scheduled for outpatient follow-up at 1, 3, 6, and 12 months after endovascular treatment. After that, the patients were followed up via face-to-face or telephone interview once per year. Periprocedural information on clinical and radiologic characteristics and long-term outcomes were collected and reviewed by trained personnel who were blinded to treatment strategies and periprocedural complications. The imaging data were interpreted by 2 neuroradiologists with consensus. The poor outcome was defined as an mRS score ≥3.
Statistical Analysis
Continuous variables were presented as mean ± standard deviation (normal distribution data) or median with interquartile range (skew distribution data), as appropriate. Categoric variables were presented as numbers and percentages. The χ2 test (categoric variables), t test (normal distribution), Mann-Whitney U test (skewed distribution), and logistic regression were used to analyze the differences in characteristics between patients with or without periprocedural NPRH. Variables with P < .05 in univariate analysis were included for multivariate analysis. The statistical analysis was performed by using a commercial statistical software package (SPSS for Windows, version 25.0, IBM).
RESULTS
Patient Baseline Characteristics
From June 2012 to August 2019, a total of 1849 patients (59.3 ± 8.6 years old) with ICAS were recruited, including 1486 (80.4%) men and 363 (19.6%) women. The main qualifying events included stroke in 1297 (70.1%) cases and TIA in 552 (29.9%) cases. The incidence of ischemic stroke within 90 days and 30 days were 51.1% (944/1849) and 19.5% (360/1849), respectively. The stenotic/occlusive lesion sites included intracranial ICA in 282 (15.3%) cases, MCA in 264 (14.3%) cases, intracranial vertebral artery in 691 (37.4%) cases, basilar artery in 598 (32.3%) cases, and vertebrobasilar junction in 14 (0.8%) cases. All baseline characteristics are presented in the Online Supplemental Data.
Periprocedural Intracranial Hemorrhage Features
Of the 1849 patients, 24 (1.3%) experienced intraprocedural or postprocedural intracranial hemorrhage, including 4 (16.7%) with PRH and 20 (83.3%) with NPRH. The primary causes of NPRH included definite hyperperfusion hemorrhage (12 cases), possible hyperperfusion hemorrhage (5 cases), and possible antithrombosis-related hemorrhage (3 cases). Periprocedural intracranial hemorrhage occurred in 14 (58.3%) cases in the anterior circulation and 10 (41.7%) cases in the posterior circulation. NPRH occurred mainly in the anterior circulation (2.4% versus 0.5%, P < .001), while there was no significant difference in the incidence of PRH between the anterior and posterior circulation (0.2% versus 0.2%, P > .999). All 4 cases of PRH were due to microwire perforation, with 1 case (25%) in the anterior circulation and 3 cases (75%) in the posterior circulation. The hemorrhage locations included lobar hematoma (25.0%), basal ganglia (37.5%), cerebellum (12.5%), brainstem (8.3%), intraventricular hemorrhage (4.2%), and subarachnoid hemorrhage (12.5%). Of the 24 cases with periprocedural intracranial hemorrhage, 5 (20.8%) occurred during the procedure, 17 (70.8%) occurred within 24 hours, and 2 (8.4%) occurred beyond 24 hours after the procedure. The characteristics of periprocedural intracranial hemorrhage are presented in Table 1.
Periprocedural characteristics and long-term outcome of patients with periprocedural intracranial hemorrhage after stent placement for symptomatic intracranial atherosclerotic stenosis
Alexander et al16 have recently proposed a new set of stent placement criteria for symptomatic ICAS based on currently available evidence. We performed a subgroup analysis according to their proposed criteria (baseline characteristics are presented in the Online Supplemental Data) and compared the incidence of periprocedural intracranial hemorrhage under 2 different criteria. The results indicated that the incidence of periprocedural intracranial hemorrhage (mainly NPRH) under the criteria of Alexander et al16 was slightly higher than our criteria, but there was no statistically significant difference (P > .05) (Online Supplemental Data).16
Long-Term Outcomes of Patients with Periprocedural Intracranial Hemorrhage
The mRS at discharge was 0−2 in 8 (33.3%) cases, 3−5 in 15 (62.5%) cases, and 6 in 1 (4.2%) case. The 30-day mRS after the procedure was 0−2 in 9 (37.5%) cases, 3−5 in 5 (20.8%) cases, and 6 in 10 (41.7%) cases. Apart from the 10 patients who died within 30 days after the procedure, 13 patients were followed up with a median period of 78 months, and 1 patient was lost to follow-up. The long-term mRS was 0–2 in 10 (76.9%) cases and 3–5 in 3 (23.1%) cases. Therefore, the exact proportions are 76.9% and 23.1% respectively (Table 1).
Outcome Comparison between Patients with PRH and NPRH
Patients with NPRH tended to have a worse 30-day outcome (mRS ≥3) than those with PRH without a statistically significant difference (70.0% versus 25.0%, P = .130). For the survivors beyond 30 days after the procedure, patients with NPRH also tended to have a worse long-term outcomes than those with PRH, though a statistically significant difference was not drawn (33.3% versus 0.0%, P = .497). Overall, the proportion of poor long-term outcomes (mRS ≥3) in patients with NPRH was significantly higher than that in patients with PRH (68.4% versus 0%, P = .024). The distribution of mRS in patients with PRH or NPRH is demonstrated in Fig 2.
The distribution of 30-day (A) and long-term (B) mRS in patients with PRH or NPRH.
Factors Associated with NPRH
NPRH occurred in 20 cases, including 13 cases in the anterior circulation and 7 cases in the posterior circulation (2.4% versus 0.5%, P < .001). The median preprocedural stenosis rate was significantly higher in patients with NPRH than that without periprocedural intracranial hemorrhage (93% versus 80%, P < .001). The postprocedural modification of stenosis rate was also significantly higher in patients with NPRH than that without periprocedural intracranial hemorrhage (80% versus 70%, P < .001). The rate of NPRH was higher in patients with cerebral infarction within 30 days (2.8% versus 0.7%, P = .001) or 90 days (1.6% versus 0.6%, P = .032) before the procedure (Online Supplemental Data). The median baseline NIHSS (1 versus 0, P = .006) and mRS (1 versus 0, P = .041) was higher in patients with NPRH than those without hemorrhage. The rate of NPRH was higher in patients with a preprocedural TICI scale of 0–2a than those with 2b–3 (2.6% versus 0.6%, P = .001) (Online Supplemental Data). Multivariate logistic regression analysis indicated that anterior circulation (OR [95% CI]: 4.395 [1.711–11.285], P = .002), high preprocedural stenosis rate (OR [95% CI]: 1.133 [1.070–1.200], P < .001), and cerebral infarction within 30 days before the procedure (OR [95% CI]: 3.588 [1.446–8.903], P = .006) were independent predictors of NPRH after stent placement for symptomatic ICAS (Table 2).
Multivariate logistic regression of NPRH prediction
DISCUSSION
With the largest number of cases focusing on periprocedural intracranial hemorrhage after stent placement for symptomatic ICAS, this study demonstrated a periprocedural intracranial hemorrhage rate of 1.3% (24/1849), which was lower than that in the SAMMPRIS (6.3%, 13/208) and VISSIT studies (8.6%, 5/58), and comparable with that in the China Angioplasty & Stenting for Symptomatic Intracranial Severe Stenosis (CASSISS) study (2.3%, 4/176) and a previous multicenter registry study (0.3%, 1/300).2,12⇓-14,17 The subgroup analysis indicated that the incidence of periprocedural intracranial hemorrhage (mainly NPRH) under the criteria of Alexander et al16 was slightly higher than our criteria without statistically significant difference. In this study, the proportion of periprocedural intracranial hemorrhage within 24 hours, 48 hours, 72 hours, and 96 hours was 91.6%, 91.6%, 95.8%, and 100%, respectively. Therefore, intensive care unit monitoring within 24 hours after the procedure is critical. Also, a postoperative observation period of 3 to 4 days in the hospital may be reasonable for those at high risk of perioperative intracranial hemorrhage.
Previous studies have reported 30-day outcomes of periprocedural intracranial hemorrhage after intracranial stent placement, but long-term outcomes remain unclear. In this study, the poor 30-day outcome (mRS ≥3) occurred in 62.5% (15/24) of the patients with periprocedural hemorrhage, which was comparable to that in the SAMMPRIS study (61.5%, 8/13) and the CASSISS study (50.0%, 2/4).2,13 Also, this study demonstrated a poor long-term outcome rate of 56.5% (13/23) with a median follow-up period of 78 months.
In this study, periprocedural hemorrhage cause was classified as PRH (16.7%) and NPRH (83.3%). The SAMMPRIS and VISSIT studies had not systematically demonstrated the specific causes of periprocedural hemorrhage. The SAMMPRIS study reported a PRH proportion of 30.8% (4/13), and the CASSISS study reported a PRH proportion of 50.0% (2/4), which were relatively higher than that in the present study.2,13 In this study, the primary cause of NPRH was postprocedural hyperperfusion. The patients with NPRH had worse long-term outcomes compared with those with PRH.
The risk of PRH can be effectively reduced with the improvement of endovascular manipulation skills, experiences, and devices.2,4 However, it is difficult to effectively predict the NPRH so far. Given the high rates of disability and mortality, it is imperative to strictly screen and identify the patients at high risk of NPRH before endovascular treatment. Our study indicated that anterior circulation, severe preprocedural stenosis, and recent infarction were independent predictors of NPRH after stent placement for symptomatic ICAS.
Compared with those in the posterior circulation, arteries in the anterior circulation have larger vessel diameters, higher blood flow, and more significant increase in cerebral blood flow after stent placement.18 A few pathophysiologic mechanisms contributing to the development of cerebral hyperperfusion hemorrhage have been proposed, and the most accepted mechanism is the impairment of cerebral autoregulation.19 Under normal conditions, cerebral autoregulation constricts the brain vessels in response to a sudden increase in blood flow to maintain normal cerebral perfusion within an acceptable range of mean arterial pressure.20,21 However, this autoregulation is often impaired when arteriosclerosis or atherosclerotic stenosis occurs.22⇓⇓-25 Our results suggested that anterior circulation was an independent risk factor of NPRH after intracranial stent placement. The cerebral blood flow could abnormally increase after stent placement for ICAS due to the impairment of cerebral autoregulation, especially in the anterior circulation.23,24,26,27 As a result, postprocedural hyperperfusion hemorrhage is more likely to occur in the anterior circulation.
In this study, the median preprocedural stenosis rate was significantly higher in patients with NPRH than those without periprocedural hemorrhage (93% versus 80%, P < .001), which was indicated to be an independent risk factor of NPRH after stent placement. It should be emphasized that patients with occlusive lesions have a significantly increased risk of NPRH after stent placement compared with those with stenotic lesions (5.0% versus 0.9%, P = .001). In addition, there was no significant difference in postprocedural residual stenosis rates between the patients with or without NPRH in this study (P = .827). Therefore, the patients with higher preprocedural stenosis rate tended to have more improvement in stenotic diameter, and more significant increase in postprocedural cerebral blood flow, which may contribute to the higher risk of hyperperfusion hemorrhage after stent placement.23,24
Our results suggested that the rate of NPRH was higher in patients with cerebral infarction within 30 days before the procedure with a statistically significant difference (2.8% versus 0.7%, P = .001), which was also indicated to be an independent risk factor of NPRH after stent placement. Recent cerebral infarction is often a result of hemodynamic compromise and decompensation, which is also an indication of cerebral autoregulation impairment.27 On the other hand, patients with recent cerebral infarction are often complicated with damaged and not yet repaired vascular bed and blood-brain barrier, so they may be faced with a higher risk of postprocedural hemorrhage (Fig 1).28,29 Thus, patients with ICAS with new cerebral infarction within 30 days may benefit from stent placement (reducing the risk of ischemic stroke recurrence) but with an increased risk of periprocedural intracranial hemorrhage.
It has to be noted that hypertension history and periprocedural blood pressure levels are supposed to be closely associated with periprocedural hemorrhage.19,30 In this study, however, there was no significant difference in the proportion of hypertension history between patients with NPRH and those without hemorrhage (90.0% versus 79.7%, P = .388). Although periprocedural systolic blood pressure was recommended to be kept between 100 and 120 mm Hg during and for 3 days after the procedure, there was a lack of records of exact periprocedural blood pressure values, so whether a more strict management of periprocedural blood pressure is exempt from intracranial hemorrhage is uncertain.
In this study, most patients received dual antiplatelet therapy (aspirin and clopidogrel) during the perioperative period. There was no significant difference in the antiplatelet therapy and inhibition rate of arachidonic acid and adenosine diphosphate between patients with NPRH and those without hemorrhage. However, the resistance testing for antiplatelet drugs was examined in only 56.9% of the patients. Therefore, with the development of antiplatelet drugs (such as ticagrelor, cilostazol, and tirofiban) and protocols, the effect of antiplatelet therapy and inhibition rate for platelets on periprocedural intracranial hemorrhage need to be further explored in future prospective studies.
Poor collateral circulation and compensation had been indicated to influence periprocedural hyperperfusion hemorrhage.19,31 In this study, most treated arteries were located in the posterior circulation (70.5%), so we did not analyze the relationship between the condition of collateral circulation and periprocedural hemorrhage. It is difficult to accurately and objectively assess the condition of posterior circulation collaterals due to the limitation of the existing collateral circulation scoring system.32,33
Based on the findings of this study, for patients with severe anterior circulation stenosis and infarction within 30 days, submaximal or staged angioplasty or stent placement may be considered in prevention of postprocedural hemorrhage, but need to be confirmed in future studies.34⇓⇓-37
Potential limitations of this study should be mentioned. First, the patients were enrolled from a single center, so that potential selection bias may be inevitable. Second, in some patients, the definition of hyperperfusion hemorrhage was based on the elevated flow velocity in the stented artery indicated by transcranial Doppler, instead of hyperperfusion in the distribution of stented artery indicated by CT perfusion imaging. Third, the cohort spans a long time and some stent placement indications (such as TIA and ischemic stroke more than 30 days) may not be suitable for future clinical practice and research with the increasing understanding of symptomatic ICAS.
CONCLUSIONS
The patients with NPRH had worse outcomes than those with PRH after stent placement for symptomatic ICAS. Anterior circulation, severe preprocedural stenosis, and recent infarction are independent predictors of NPRH.
Footnotes
Kaijiang Kang and Peipei Gong contributed equally to this work.
This work was supported by the National Natural Science Foundation of China (Grant number: 81471390 and 82171894 to N.M.) and Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences (Grant number: 2019-I2M-5-029 to X.Z.).
Prof. Ning Ma, Zhongrong Miao and Xingquan Zhao are co-corresponding authors.
Disclosure forms provided by the authors are available with the full text and PDF of this article at www.ajnr.org.
References
- Received March 20, 2024.
- Accepted after revision June 2, 2024.
- © 2024 by American Journal of Neuroradiology
Jump to section
Related Articles
Cited By...
- No citing articles found.