Abstract
Background
Reducing intracranial hemorrhage (ICH) can improve patient outcome in acute ischemic stroke (AIS) intervention. We sought to identify ICH risk factors after AIS thrombectomy.
Methods
This is a retrospective review of the Stroke Thrombectomy and Aneurysm Registry (STAR) database. All patients who underwent AIS thrombectomy with available ICH data were included. Multivariable regression models were developed to identify predictors of ICH after thrombectomy. Subgroup analyses were performed stratified by symptom status and European Cooperative Acute Stroke Study (ECASS) grade.
Results
The study cohort comprised 6860 patients. Any ICH and symptomatic ICH (sICH) occurred in 25% and 7% of patients, respectively. Hemorrhagic infarction 1 (HI1) occurred in 36%, HI2 in 24%, parenchymal hemorrhage 1 (PH1) in 22%, and PH2 in 17% of patients classified by ECASS grade. Intraprocedural complications independently predicted any ICH (OR 3.8083, P<0.0001), PH1 (OR 1.9053, P=0.0195), and PH2 (OR 2.7347, P=0.0004). Race also independently predicted any ICH (black: OR 0.5180, P=0.0017; Hispanic: OR 0.4615, P=0.0148), sICH (non-white: OR 0.4349, P=0.0107), PH1 (non-white: OR 3.1668, P<0.0001), and PH2 (non-white: OR 1.8689, P=0.0176), with white as the reference. Primary mechanical thrombectomy technique also independently predicted ICH. ADAPT (A Direct Aspiration First Pass Technique) was a negative predictor of sICH (OR 0.2501, P<0.0001), with stent retriever as the reference.
Conclusions
This study identified ICH risk factors after AIS thrombectomy using real-world data. There was a propensity towards a reduced sICH risk with direct aspiration. Procedural complications and ethnicity were predictors congruent between categories of any ICH, sICH, PH1, and PH2. Further investigation of technique and ethnicity effects on ICH and outcomes after AIS thrombectomy is warranted.
INTRODUCTION
After several independent randomized controlled trials found mechanical thrombectomy (MT) to be superior to medical management alone in acute ischemic stroke (AIS) patients with large vessel occlusion, it has been adopted as the standard of care treatment.1–3 Recently, investigations have focused on improving treatment efficiency by reducing time to revascularization, optimizing patient selection with the use of advanced imaging technologies, and evaluating the design of novel thrombectomy devices.1 Both thrombectomy and thrombolysis carry associated risks of hemorrhage due to mechanical trauma to the vessel wall, blood–brain barrier disruption, and reperfusion-related free radical production.4 Identifying modifiable risk factors for treatment-related complications is an avenue for investigation that may enhance outcomes in AIS patients.
The risk of any post-thrombectomy intracranial hemorrhage (ICH) and symptomatic ICH (sICH) has been estimated to range from 15.2–36.0% and 2.3–13.8%, respectively.2 5–7 When compared with the major clinical trials, there is a higher reported risk of hemorrhage among real-world data. 5 8 This has been attributed to the use of stricter patient selection criteria and greater technical homogeneity in the clinical trials setting.8 The European Cooperative Acute Stroke Study (ECASS) radiographic criteria has been used in AIS thrombectomy clinical trials to estimate ICH severity.9 We hypothesized that the risk factors for ICH after AIS thrombectomy may vary depending on ECASS grade. The aims of this multicenter retrospective study were to identify potentially modifiable ICH risk factors after AIS thrombectomy in the real-world patient setting, and to assess for differences in these risk factors based on the presence or absence of symptoms and the ECASS grade.
METHODS
Study design
We retrospectively reviewed a database of AIS patients who underwent MT at 33 stroke centers participating in the Stroke Thrombectomy and Aneurysm Registry (STAR) from January 2013 to December 2020. The study was approved by the institutional review board (IRB) at each individual institution and follows the guidelines set forth by the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement.10 Patient consent was waived by the IRB. The data were obtained retrospectively and collected according to a standardized protocol. Verification, de-identification, and attestation of data accuracy were performed by investigators at each contributing institution. Individual patient data from each contributing institution were pooled by investigators at the Medical University of South Carolina and these pooled data were analyzed by an independent statistician.
Patient identification and selection
The inclusion criteria for the present study were (1) age ≥18 years, (2) underwent MT for AIS, and (3) available ICH data. Patient selection for MT and post-procedural stroke cares were performed according to participating institutional standards. At most institutions, AIS treatment standards followed the guidelines set forth by the American Heart Association/American Stroke Association.11
Baseline data and variables
Baseline data collected by the consortium included patient, clinical, and AIS treatment parameters. Patient variables comprised age, sex, and race. Clinical data comprised medical history data (history of diabetes mellitus, hypertension, atrial fibrillation, hyperlipidemia, congestive heart failure and/or prior stroke), pre-stroke smoking status (non-smoker, former smoker, and current smoker), pre-stroke modified Rankin Scale (mRS) score, primary vessel occlusion location (anterior or posterior circulation), National Institutes of Health Stroke Scale (NIHSS) score at presentation, and the Alberta Stroke Program Early CT Score (ASPECTS). Clinical assessments and interpretation of radiographic studies at presentation and throughout the hospital stay were performed independently by neuroradiologists, neurologists, and neurointerventionalists according to participating institutional standards.
AIS treatment parameters included administration of intravenous tissue plasminogen activator (IV-tPA), balloon guide catheter use, primary MT technique (A Direct Aspiration First Pass Technique (ADAPT) vs no ADAPT (comprising Solumbra, stent retriever, and other)), time from onset to groin puncture, time from puncture to clot engagement, time from puncture to reperfusion, total procedure time, and post-MT reperfusion grade. The post-MT reperfusion grade was reported using the revised modified Thrombolysis In Cerebral Infarction (mTICI) system. This was assessed by the treating neurointerventionalist by comparing the initial (pre-MT) to the final (post-MT) catheter digital subtraction angiography.12 Successful reperfusion was defined as mTICI grade 2b or greater. AIS treatment-related complications included distal embolization and intraprocedural complications. Intraprocedural complications comprised access site complications, vessel injury, patient agitation or neurological instability during the procedure. Modern MT devices, comprising stent retrievers and aspiration catheters, were used in all procedures. The technique was reported according to the first recanalization attempt.
Outcomes
The primary outcome was any ICH after MT for AIS. This was assessed by neuroradiologists at each participating center and defined as blood at any site in the brain on the post-procedure CT and/or MRI studies. Secondary outcomes included sICH, the ECASS grade, and good neurological outcome. sICH was defined as any ICH with concurrent documentation of a clinical deterioration according to any one of the following criteria: (1) increase in the NIHSS score of 4 or more points, (2) increase in the NIHSS score of 2 or more points in one category, or (3) need for intubation, hemicraniectomy, external ventricular drain placement, or other major medical or surgical intervention.8 9 The ECASS grade was assigned by neuroradiologists at each participating center according to established CT criteria whereby hemorrhagic infarction 1 (HI1) was defined as small petechiae along the margins of the infarct, hemorrhagic infarction 2 (HI2) as confluent petechiae within the infarcted area without mass effect, parenchymal hemorrhage 1 (PH1) as blood in 30% or less of the infarcted area with mild mass effect, and parenchymal hemorrhage 2 (PH2) as blood in more than 30% of the infarcted area with significant mass effect.9 Good neurological outcome was defined as mRS score 0–2 at 90 days.13
Statistical analysis
Statistical analyses were performed using JMP PRO (version 16.2, SAS Institute, Cary, NC) and Microsoft Excel (Redmond, WA). Baseline patient, clinical, and AIS treatment characteristics were compared between groups dichotomized by the primary and secondary outcomes using univariable logistic and multinomial regression analyses, as appropriate. Odds ratios (OR) and 95% confidence intervals (95% CI) were reported. For categorical variables at more than two levels, reference levels were established as the basis for odds ratios. Independent predictors of any ICH and sICH were identified using multivariable logistic regression analyses. For any ICH and good outcome, multivariable model 1 included each of the predictors with P<0.05 from the univariable analysis, and multivariable model 2 included each of the predictors with P<0.05 from model 1. For sICH, multivariable model 1 included each of the predictors with P<0.05 from the univariable analysis, and multivariable model 2 used an ad hoc adaptation of forward stepwise regression for predictor variables with P<0.05. Interactions with technique (ADAPT vs not ADAPT) and subgroup predictors of interest were tested separately in the derived multivariable models. Subgroup predictors included final mTICI score (≤2a vs >2a), time from onset to groin puncture (6 hours vs 6–24 hours vs ≥24 hours), time from puncture to clot engagement (≤10 min vs >10 min), and time from puncture to reperfusion (≤60 min vs >60 min). For the outcomes of any ICH and sICH, multivariable logistic regression interaction models included respective predictors from model 2 and the interaction term ADAPT×subgroup-defining variable. Groups were then stratified by the ECASS grade, using HI1 as the reference variable. Univariable predictors of each ECASS grade were identified using multinomial logistic regression analyses. Several multivariable multinomial regression models were fitted for the outcome of ECASS grade. The goodness of fit was determined by the Akaike information criterion curve (AICc) using model comparison methods, and the model with the best fit was selected for the final analysis. Statistical significance was defined as P<0.05, and all tests were two-tailed. Missing data were not imputed.
RESULTS
Predictors of the primary outcome
From the STAR registry, 6860 patients were eligible and categorized by the primary outcome of ICH (n=1728) versus no ICH (n=5132) after MT for AIS. Table 1 outlines the univariable and multivariable predictors of the primary outcome. Prior history of diabetes (P=0.0273), former smoker (P=0.0008; with non-smoker as the reference category), anterior circulation occlusion (P<0.0001), higher admission NIHSS (P<0.0001), lower ASPECTS (P<0.0001), use of IV-tPA (P=0.0148), use of balloon angioplasty (P=0.0143), use of ADAPT (P<0.0001; with stent retriever as the reference category), longer time from onset to groin puncture (P<0.0001), shorter time from puncture to clot engagement (P<0.0001), longer time from puncture to reperfusion (P=0.0164), distal embolization (P<0.0001), and intraprocedural complications (P<0.0001) were positive univariable predictors of the primary outcome. Race (black: P=0.0144; Hispanic: P<0.0001; non-white: P<0.0001; white as reference), use of other MT technique (P<0.0001; stent retriever as reference), and higher final mTICI score (P<0.0001) were negative univariable predictors of the primary outcome.
Table 1.
Predictors of intracranial hemorrhage after mechanical thrombectomy for acute ischemic stroke
| Intracranial hemorrhage (n=1728) | No intracranial hemorrhage (n=5132) | Univariable analysis | Multivariable analysis model 1* | Multivariable analysis model 2† | ||||
|---|---|---|---|---|---|---|---|---|
| OR (95% CI) | P value | OR (95% CI) | P value | OR (95% CI) | P value | |||
| Age, years (SD) | 69.7 (14.0) | 69.3 (14.9) | 1.0018 (0.9980 to 1.0055) | 0.3551 | – | – | – | – |
| Female, n (%) | 874/1728 (50.6) | 2510/5126 (50.5) | 1.0666 (0.9564 to 1.1895) | 0.2460 | – | – | – | – |
| Race, n (%) | <0.0001 | 0.0012 | 0.0156 | |||||
| White | 1074/1511 (71.1) | 2496/3910 (63.8) | Ref | Ref | Ref | Ref | Ref | Ref |
| Black | 190/1511 (12.6) | 518/3910 (13.2) | 0.8528 (0.7113 to 1.0215) | 0.0144 | 0.5180 (0.3433 to 0.7815) | 0.0017 | 0.7147 (0.5047 to 1.0121) | 0.0584 |
| Hispanic | 45/1511 (3.0) | 275/3910 (7.0) | 0.3803 (0.2753 to 0.5254) | <0.0001 | 0.3409 (0.1531 to 0.7591) | 0.0084 | 0.4615 (0.2479 to 0.8594) | 0.0148 |
| Non-white | 202/1511 (13.4) | 621/3910 (15.9) | 0.7559 (0.6315 to 0.8997) | <0.0001 | 1.0051 (0.5720 to 1.7662) | 0.9859 | 1.1370 (0.7600 to 1.7010) | 0.5323 |
| Diabetes mellitus, n (%) | 494/1690 (29.2) | 1330/5027 (26.5) | 1.149 (1.106 to 1.297 | 0.0273 | 1.4018 (0.9782 to 2.0088) | 0.0658 | – | – |
| Hypertension, n (%) | 1236/1691 (73.1) | 3695/5036 (73.4) | 0.9859 (0.8706 to 1.1164) | 0.8226 | – | – | – | – |
| Atrial fibrillation, n (%) | 598/1687 (35.4) | 1833/5026 (36.5) | 0.9566 (0.8525 to 1.0733) | 0.4490 | – | – | – | – |
| Hyperlipidemia, n (%) | 679/1689 (40.2) | 2141/5031 (42.6) | 0.9075 (0.8112 to 1.0152) | 0.0893 | – | – | – | – |
| Congestive heart failure, n (%) | 114/895 (12.7) | 299/2329 (12.8) | 0.9910 (0.7866 to 1.2485) | 0.9389 | – | – | – | – |
| Prior stroke, n (%) | 268/1569 (17.1) | 680/4294 (15.8) | 1.0948 (0.9377 to 1.2782) | 0.2539 | – | – | – | – |
| Smoking status, n (%) | 0.0028 | 0.3178 | – | – | ||||
| Non-smoker | 599/946 (63.3) | 1628/2359 (69.0) | Ref | Ref | Ref | Ref | – | – |
| Former smoker | 186/946 (19.7) | 359/2359 (22.1) | 1.4081 (1.1525 to 1.7204) | 0.0008 | 0.9801 (0.6498 to 1.4781) | 0.9235 | – | – |
| Current smoker | 161/946 (17.0) | 372/2359 (15.8) | 1.1763 (0.9561 to 1.4472) | 0.1247 | 0.7101 (0.4484 to 1.1246) | 0.1445 | – | – |
| Anterior circulation, n (%) | 1603/1716 (93.4) | 4479/5076 (88.2) | 1.89052 (1.5350 to 2.3300) | <0.0001 | 1.4542 (0.2985 to 7.0839) | 0.6430 | – | – |
| Pre-morbid mRS, mean (SD) | 0.6 (1.1) | 0.7 (1.1) | 0.976286 (0.9231 to 1.0326) | 0.4014 | – | – | – | – |
| Admission NIHSS, mean (SD) | 16.3 (6.9) | 15.2 (7.1) | 1.0235 (1.0156 to 1.0315) | <0.0001 | 3.7699 (1.6122 to 8.9171) | 0.0022 | 1.0385 (1.0197 to 1.0580) | <0.0001 |
| ASPECTS, mean (SD) | 7.8 (1.8) | 8.3 (1.7) | 0.8501 (0.8211 to 0.8802) | <0.0001 | 0.7629 (0.3345 to 1.7546) | 0.5222 | – | – |
| IV-tPA, n (%) | 769/1712 (44.9) | 2454/5079 (48.3) | 0.8723 (0.7815 to 0.9737) | 0.0148 | 1.9390 (1.3556 to 2.7735) | 0.0003 | 1.7000 (1.2594 to 2.2948) | 0.0005 |
| Angioplasty, n (%) | 121/1041 (11.6) | 196/2213 (8.9) | 1.3535 (1.0651 to 1.7120) | 0.0143 | 2.8566 (1.5556 to 5.2456) | 0.0007 | 1.6232 (1.0472 to 2.5162) | 0.0314 |
| ADAPT, n (%) | 710/1525 (46.6) | 1632/4691 (34.8) | 1.7782 (1.5213 to 2.0770) | <0.0001 | 1.4668 (0.9074 to 2.3711) | 0.1180 | – | – |
| Time from onset to groin puncture, mean min (SD) | 434.4 (449.2) | 380.2 (387.5) | 1.0003 (1.0002 to 1.0004) | <0.0001 | 33.3044 (7.1428 to 169.1229) | <0.0001 | 1.0007 (1.0003 to 1.0011) | 0.0002 |
| Time from puncture to occlusion site, min (SD) | 11.8 (16.2) | 19.9 (27.9) | 0.9800 (0.9727 to 0.9874) | <0.0001 | 0.0762 (0.0121 to 0.4302) | 0.0046 | 0.9753 (0.9655 to 0.9839) | <0.0001 |
| Time from puncture to ≥mTICI 2b, min (SD) | 54.1 (45.7) | 51.2 (40.2) | 1.0016 (1.0003 to 1.0029) | 0.0164 | 2.5857 (0.4776 to 13.8528) | 0.2684 | – | – |
| Procedure time, min (SD) | 51.2 (47.1) | 48.7 (42.5) | 1.0012 (0.9992 to 1.0033) | 0.2377 | – | – | – | – |
| Final mTICI score, n (%) | <0.0001 | 0.0356 | 0.0002 | |||||
| 0 | 102/1682 (6.1) | 312/4484 (6.4) | Ref | Ref | Ref | Ref | Ref | Ref |
| 1 | 40/1682 (2.4) | 108/4884 (2.2) | 1.1330 (0.7398 to 1.7348) | 0.8602 | 1.5577 (0.1843 to 13.1666) | 0.6840 | 0.8085 (0.2367 to 2.7618) | 0.7345 |
| 2a | 172/1682 (10.2) | 295/4884 (6.0) | 1.7835 (1.3317 to 2.3885) | <0.0001 | 2.1638 (0.7384 to 6.3410) | 0.1594 | 1.3626 (0.6230 to 2.9804) | 0.4384 |
| 2b | 636/1682 (37.8) | 1546/4884 (31.7) | 1.2584 (0.9880 to 1.6027) | 0.1863 | 1.4976 (0.6600 to 3.3981) | 0.3340 | 1.1064 (0.6037 to 2.0275) | 0.7436 |
| 2c | 156/1682 (9.3) | 382/4884 (7.8) | 1.2492 (0.9338 to 1.6710) | 0.4330 | 1.5878 (0.6513 to 3.8709) | 0.3092 | 1.0724 (0.5422 to 2.1210) | 0.8408 |
| 3 | 576/1682 (34.2) | 2241/4884 (45.9) | 0.7862 (0.6175 to 1.0010) | <0.0001 | 0.8551 (0.3768 to 1.9406) | 0.7082 | 0.5501 (0.3059 to 0.9895) | 0.0460 |
| Distal emboli, n (%) | 309/1316 (23.5) | 585/3704 (15.8) | 1.6360 (1.4010 to 1.9104) | <0.0001 | 1.8155 (1.1657 to 2.8276) | 0.0083 | 1.6455 (1.1265 to 2.4038) | 0.0103 |
| Intraprocedural complication, n (%) | 194/1481 (13.1) | 220/4125 (5.3) | 2.6756 (2.1839 to 3.2780) | <0.0001 | 4.8303 (2.6102 to 8.9387) | <0.0001 | 3.8083 (2.4030 to 6.0356) | <0.0001 |
Interactions were explored for the outcome of intracranial hemorrhage between ADAPT and subgroup-defining variable (ADAPT × variable) with no ADAPT as reference adjusting for other covariates in model 2. Subgroup defining variables tested were: final mTICI score (≤2a vs ≥2b; P=0.4444), time from onset to groin puncture (≤6 hours vs 6–24 hours vs ≥24 hours; P=0.1413), time from puncture to occlusion site (≤10 min vs >10 min; P=0.1696), and time from puncture to ≥mTICI 2b (≤1 hour vs >1 hour; P=0.3049).
Model includes statistically significant variables from univariate analyses.
Model includes statistically significant variables from model 1.
ADAPT, A Direct Aspiration First Pass Technique; ASPECTS, Alberta Stroke Program Early CT Score; IV-tPA, intravenous tissue plasminogen activator; mRS, modified Rankin Scale; mTICI, modified Thrombolysis In Cerebral Infarction; NIHSS, National Institutes of Health Stroke Scale.
In both multivariable models, higher admission NIHSS (model 1: OR 3.7699, P=0.0022; model 2: OR 1.0385, P<0.0001), use of IV-tPA (model 1: OR 1.9390, P=0.0003; model 2: OR 1.7000, P=0.0005), use of balloon angioplasty (model 1: OR 2.8566, P=0.0007; model 2: OR 1.6232, P=0.0314), longer time from onset to groin puncture (model 1: OR 33.3044, P<0.0001; model 2: OR 1.0007, P=0.0002), shorter time from puncture to clot engagement (model 1: OR 0.0762, P=0.0046; model 2: OR 0.9753, P<0.0001), distal embolization (model 1: OR 1.8155, P=0.0083; model 2: OR 1.6455, P=0.0103), and intraprocedural complications (model 1: OR 4.8303, P<0.0001; model 2: OR 3.8083, P<0.0001) were positive independent predictors of the primary outcome. In both multivariable models, Hispanic race (model 1: OR 0.3409, P=0.0084; model 2: OR 0.4615, P=0.0148; white as reference) was a negative independent predictor of the primary outcome. In model 1 only, black race (model 1: OR 0.5180, P=0.0017; white as reference) was a negative independent predictor of the primary outcome. In model 2 only, higher final mTICI score (model 2: P=0.0460) was a negative independent predictor of the primary outcome. Primary MT technique did not independently predict the primary outcome in either model.
Subgroup analyses for the primary outcome
Exploratory subgroup analyses did not show a difference in the odds of the primary outcome for patients treated with ADAPT MT technique among the subgroups of final mTICI score (≤2a vs >2a; Pinteraction=0.4444), time from onset to groin puncture (6 hours vs 6–24 hours vs ≥24 hours; Pinteraction=0.1413), time from puncture to clot engagement (≤10 min vs >10 min; Pinteraction=0.1696), and time from puncture to reperfusion (≤60 min vs >60 min; Pinteraction=0.3049).
Predictors of symptomatic ICH
Data on symptom status was available for 1696 of 1728 patients who experienced ICH and they were dichotomized into sICH (n=474) and no sICH (n=1222). Table 2 outlines the univariable and multivariable predictors of sICH. Prior history of congestive heart failure (P=0.0105), Hispanic race (P<0.0001; with white as the reference category), lower ASPECTS (P=0.0131), lower admission NIHSS (P=0.0230), use of IV-tPA (P=0.0425), shorter time from onset to groin puncture (P=0.0270), longer time from puncture to reperfusion (P<0.0001), longer total procedure time (P=0.0003), and intraprocedural complications (P<0.0001) were positive univariable predictors of sICH. ADAPT technique (P<0.0001) was a negative univariable predictor of sICH.
Table 2.
Predictors of symptomatic intracranial hemorrhage after mechanical thrombectomy for acute ischemic stroke in patients
| Symptomatic hemorrhage (n=474) | Asymptomatic hemorrhage (n=1222) | Univariable analysis | Multivariable analysis model 1* | Multivariable analysis model 2† | ||||
|---|---|---|---|---|---|---|---|---|
| OR (95% CI) | P value | OR (95% CI) | P value | OR (95% CI) | P value | |||
| Age, years (SD) | 70.4 (13.3) | 69.4 (14.2) | 1.0146 (0.8207 to 1.2543) | 0.1761 | – | – | – | – |
| Female, n (%) | 238/474 (50.2) | 618/1222 (50.6) | 0.9856 (0.7972 to 1.2185) | 0.8936 | – | – | – | – |
| Race, n (%) | 0.0007 | 0.0115 | 0.0698 | |||||
| White | 273/400 (68.3) | 770/1080 (71.3) | Ref | Ref | Ref | Ref | Ref | Ref |
| Black | 51/400 (12.8) | 139/1080 (12.9) | 1.0349 (0.7298 to 1.4675) | 0.8475 | 0.8173 (0.4052 to 1.6486) | 0.5732 | 0.9048 (0.5057 to 1.5658) | 0.7259 |
| Hispanic | 25/400 (6.3) | 20/1080 (1.9) | 3.5256 (1.9272 to 6.4498) | <0.0001 | 0.7820 (0.1942 to 3.1495) | 0.7294 | 1.3458 (0.4121 to 3.7702) | 0.5990 |
| Non-white | 51/400 (12.8) | 151/1080 (14.0) | 0.9526 (0.6739 to 1.3466) | 0.7835 | 0.2344 (0.0949 to 0.5787) | 0.0017 | 0.4349 (0.2191 to 0.8283) | 0.0107 |
| Diabetes mellitus, n (%) | 145/467 (31.1) | 346/1209 (28.6) | 1.1232 (0.8903 to 1.4169) | 0.3271 | – | – | – | – |
| Hypertension, n (%) | 356/468 (76.1) | 871/1209 (72.0) | 1.2335 (0.9638 to 1.5786) | 0.0955 | – | – | – | – |
| Atrial fibrillation, n (%) | 172/467 (36.8) | 422/1207 (35.0) | 1.0846 (0.8685 to 1.3544) | 0.4738 | – | – | – | – |
| Hyperlipidemia, n (%) | 175/466 (37.6) | 499/1209 (41.3) | 0.8557 (0.6869 to 1.0659) | 0.1643 | – | – | – | – |
| Congestive heart failure, n (%) | 40/224 (17.9) | 74/662 (11.2) | 1.7274 (1.1364 to 2.6257) | 0.0105 | 1.9404 (0.9668 to 3.8946) | 0.0622 | – | – |
| Prior stroke, n (%) | 72/427 (16.9) | 194/1128 (17.2) | 0.9764 (0.7257 to 1.3139) | 0.8749 | – | – | – | – |
| Smoking status, n (%) | 0.927 | |||||||
| Non-smoker | 143/223 (64.1) | 450/714 (63.0) | Ref | Ref | – | – | – | – |
| Former smoker | 44/223 (19.7) | 141/714 (19.8) | 0.9820 (0.6667 to 1.4464) | 0.9267 | – | – | – | – |
| Current smoker | 36/223 (16.1) | 123/714 (17.2) | 0.9210 (0.6074 to 1.3967) | 0.6986 | – | – | – | – |
| Anterior circulation, n (%) | 438/473 (92.6) | 1134/1211 (93.6) | 0.8497 (0.5614 to 1.2862) | 0.4413 | – | – | – | – |
| Pre-morbid mRS, mean (SD) | 0.7 (1.1) | 0.6 (1.1) | 1.4861 (0.8812 to 2.5062) | 0.1374 | – | – | – | – |
| Admission NIHSS, mean (SD) | 15.7 (7.2) | 16.6 (6.8) | 0.9822 (0.9671 to 0.9976) | 0.0230 | 1.0029 (0.9659 to 1.0412) | 0.8806 | – | – |
| ASPECTS, mean (SD) | 7.6 (2.0) | 7.8 (1.8) | 0.9192 (0.8601 to 0.9825) | 0.0131 | 0.9221 (0.8081 to 1.0522) | 0.2296 | – | – |
| IV-tPA, n (%) | 228/469 (48.6) | 524/1215 (43.1) | 1.2475 (1.0075 to 1.5448) | 0.0425 | 1.4328 (0.8373 to 2.4519) | 0.1895 | – | – |
| Angioplasty, n (%) | 33/275 (12.0) | 86/757 (11.4) | 1.0640 (0.6939 to 1.6313) | 0.7762 | – | – | – | – |
| ADAPT, n (%) | 124/397 (31.2) | 580/1097 (52.9) | 0.3398 (0.2503 to 0.4613) | <0.0001 | 0.2283 (0.1306 to 0.2991) | <0.0001 | 0.2501 (0.1568 to 0.3968) | <0.0001 |
| Time from onset to groin puncture, mean min (SD) | 392.9 (401.7) | 450.6 (466.2) | 0.9997 (0.9993 to 0.9999) | 0.0270 | 0.2345 (0.0098 to 5.6006) | 0.3704 | – | – |
| Time from puncture to occlusion site, min (SD) | 14.2 (25.0) | 11.2 (13.1) | 1.0093 (0.9966 to 1.0222) | 0.1508 | – | – | – | – |
| Time from puncture to ≥mTICI 2b, min (SD) | 62.5 (48.6) | 51.1 (44.4) | 1.0050 (1.0028 to 1.0073) | <0.0001 | 1.0003 (0.9885 to 1.0122) | 0.9624 | – | – |
| Procedure time, min (SD) | 64.7 (50.9) | 47.3 (45.5) | 1.0070 (1.0032 to 1.0108) | 0.0003 | 1.0099 (1.0002 to 1.0197) | 0.0488 | 1.0073 (1.0016 to 1.0129) | 0.0119 |
| Final mTICI score, n (%) | 0.0766 | |||||||
| 0 | 29/407 (7.3) | 71/1104 (6.4) | Ref | Ref | – | – | – | – |
| 1 | 16/407 (3.9) | 24/1104 (2.2) | 1.6322 (0.7588 to 3.5110) | 0.2100 | – | – | – | – |
| 2a | 57/407 (14.0) | 111/1104 (10.1) | 1.2572 (0.7347 to 2.1513) | 0.4036 | – | – | – | – |
| 2b | 169/407 (41.5) | 460/1104 (41.7) | 0.8995 (0.5641 to 1.4342) | 0.6563 | – | – | – | – |
| 2c | 45/407 (11.1) | 107/1104 (9.7) | 1.0296 (0.5912 to 1.7933) | 0.9178 | – | – | – | – |
| 3 | 136/407 (33.4) | 426/1104 (38.6) | 0.7816 (0.4870 to 1.2545) | 0.3074 | – | – | – | – |
| Distal emboli, n (%) | 70/324 (21.6) | 238/976 (24.4) | 0.8546 (0.6317 to 1.1561) | 0.3081 | – | – | – | – |
| Intraprocedural complication, n (%) | 74/382 (19.4) | 116/1067 (10.9) | 1.9700 (1.4322 to 2.7089) | <0.0001 | 1.7587 (0.8608 to 3.5937) | 0.1215 | – | – |
Interactions were explored for the outcome of symptomatic intracranial hemorrhage between ADAPT and subgroup-defining variable (ADAPT × variable) with no ADAPT as reference adjusting for other covariates in model 2. Subgroup defining variables tested were: final mTICI score (≤2a vs ≥2b; P=0.2934), time from onset to groin puncture (≤6 hours vs 6–24 hours vs ≥24 hours; P=0.3893), time from puncture to occlusion site (≤10 min vs >10 min; P=0.2065), and time from puncture to ≥mTICI 2b (≤1 hour vs >1 hour; P=0.7705).
Model includes statistically significant variables from univariate analyses.
Model includes statistically significant variables from model 1.
ADAPT, A Direct Aspiration First Pass Technique; ASPECTS, Alberta Stroke Program Early CT Score; IV-tPA, intravenous tissue plasminogen activator; mRS, modified Rankin Scale; mTICI, modified Thrombolysis In Cerebral Infarction; NIHSS, National Institutes of Health Stroke Scale.
In both multivariable models, non-white race (model 1: OR 0.2344, P=0.0017; model 2: OR 0.4349, P=0.0107; white as reference) and ADAPT technique (model 1: OR 0.2283, P<0.0001; model 2: OR 0.2501, P<0.0001) were negative independent predictors of sICH. Longer total procedure time was a positive independent predictor of sICH in both models (model 1: OR 1.0099, P=0.0488; model 2: OR 1.0073, P=0.0119).
Subgroup analyses for sICH
Exploratory subgroup analyses did not show a difference in the odds of the primary outcome for patients treated with the ADAPT MT technique among the subgroups of final mTICI score (≤2a vs >2a; Pinteraction=0.6702), time from onset to groin puncture (6 hours vs 6–24 hours vs ≥24 hours; Pinteraction=0.3893), time from puncture to clot engagement (≤10 min vs >10 min; Pinteraction=0.2065), and time from puncture to reperfusion (≤60 min vs >60 min; Pinteraction=0.7705).
ECASS grades
Data on ECASS grade was available for 1312 of 1728 ICH patients and were categorized as HI1 (n=480, 36.6%), HI2 (n=320, 24.4%), PH1 (n=290, 22.1%) or PH2 (n=222, 16.9%) (table 3). Table 4 outlines the univariable predictors of the ECASS categories, with HI1 as the reference category. Univariable positive predictors of HI2 were distal embolization (P=0.0485) and longer time from puncture to reperfusion (P=0.0137). The univariable negative predictor of HI2 was former smoking status (P=0.0055; non-smoker as reference). Univariable positive predictors of PH1 were admission NIHSS (P=0.0009), longer procedure time (P=0.0036), intraprocedural complications (P=0.0195), and non-white race (P<0.0001; white as reference). Univariable negative predictors of PH1 were prior history of diabetes (P=0.0097) and former smoking status (P=0.0022; non-smoker as reference). Univariable positive predictors of PH2 were longer procedure time (P=0.0004), intraprocedural complications (P=0.0195), and other race (P<0.0001; white as reference).
Table 3.
Distribution of hemorrhage categories by ECASS
| ECASS categories | Frequency (%) |
|---|---|
| HI1 | 480/1312 (36.6) |
| HI2 | 320/1312 (24.4) |
| PH1 | 290/1312 (22.1) |
| PH2 | 222/1312 (16.9) |
ECASS, European Cooperative Acute Stroke Study; HI, hemorrhagic infarction; PH, parenchymal hemorrhage.
Table 4.
Univariate predictors of hemorrhage categories using multinomial logistic regression models
| ECASS | Variable | Relative OR (95% CI) | P value |
|---|---|---|---|
| HI1 | |||
| Ref | Ref | Ref | |
| HI2 | |||
| Distal embolization | 1.4460 (1.0024 to 2.0859) | 0.0485 | |
| Diabetes mellitus | 0.8484 (0.6200 to 1.1609) | 0.3041 | |
| Hypertension | 1.0015 (0.7282 to 1.3773) | 0.9927 | |
| Atrial fibrillation | 1.0936 (0.8108 to 1.4751) | ||
| Hyperlipidemia | 1.2419 (0.9276 to 1.6627) | ||
| Congestive heart failure | 0.5307 (0.2656 to 1.0605) | ||
| Prior stroke | 1.2530 (0.8481 to 1.8512) | ||
| Smoking status | |||
| Non-smoker | Ref | Ref | |
| Current smoker | 0.7512 (0.4429 to 1.2738) | 0.2882 | |
| Former smoker | 0.4776 (0.2835 to 0.8045) | 0.0055 | |
| Admission NIHSS | 1.0103 (0.9887 to 1.0324) | 0.3513 | |
| Anterior circulation | 1.0405 (0.5653 to 1.9152) | 0.8984 | |
| Time from onset to groin puncture | 0.9998 (0.9994 to 1.0002) | 0.2862 | |
| Angioplasty | 0.9105 (0.4920 to 1.6847) | 0.7651 | |
| ADAPT | 1.0053 (0.7394 to 1.3668) | 0.9730 | |
| Time from puncture to occlusion site | 1.0181 (0.9926 to 1.04540) | 0.1685 | |
| Time from puncture to ≥mTICI 2b | 1.0039 (1.0008 to 1.0070) | 0.0137 | |
| Procedure time | 1.0032 (0.9965 to 1.0099) | 0.3435 | |
| Final mTICI score | |||
| 0 | Ref | Ref | |
| 1 | 1.7789 (0.6447 to 4.9084) | 0.2659 | |
| 2a | 1.6565 (0.7973 to 3.4418) | 0.1760 | |
| 2b | 0.7951 (0.4196 to 1.5066) | 0.4818 | |
| 2c | 0.7861 (0.3685 to 1.6772) | 0.5335 | |
| 3 | 0.8440 (0.4447 to 1.3205) | 0.6038 | |
| Complications | 1.6109 (0.9225 to 2.8129) | 0.0936 | |
| Race | |||
| White | Ref | Ref | |
| Black | 1.0520 (0.6795 to 1.6506) | 0.8253 | |
| Hispanic | 0.1447 (0.0185 to 1.1300) | 0.0652 | |
| Non-white | 1.2656 (0.7654 to 2.0927) | 0.3585 | |
| PH1 | |||
| Distal embolization | 0.8361 (0.5576 to 1.2539) | 0.3866 | |
| Diabetes mellitus | 0.6407 (0.4573 to 0.8977) | 0.0097 | |
| Hypertension | 0.8904 (0.6445 to 1.2300) | 0.4811 | |
| Atrial fibrillation | 1.0991 (0.8087 to 1.4939) | 0.5459 | |
| Hyperlipidemia | 0.8286 (0.6088 to 1.1279) | 0.2320 | |
| Congestive heart failure | 0.8658 (0.4789 to 1.5652) | 0.6333 | |
| Prior stroke | 1.0489 (0.6920 to 1.5900) | 0.8219 | |
| Smoking status | |||
| Non-smoker | Ref | Ref | |
| Current smoker | 0.6559 (0.3861 to 1.1143) | 0.1187 | |
| Former smoker | 0.4498 (0.2694 to 0.7510) | 0.0022 | |
| Admission NIHSS | 1.0387 (1.0158 to 1.0623) | 0.0009 | |
| Anterior circulation | 0.9986 (0.5364 to 1.8588) | 0.9964 | |
| Time from onset to groin puncture | 0.9996 (0.9992 to 1.0000) | 0.0678 | |
| Angioplasty | 1.1226 (0.6162 to 2.0451) | 0.7055 | |
| ADAPT | 0.9470 (0.6956 to 1.2894) | 0.7296 | |
| Time from puncture to occlusion site | 1.0144 (0.9872 to 1.0426) | 0.2948 | |
| Time from puncture to ≥mTICI 2b | 0.9983 (0.9946 to 1.0018) | 0.3485 | |
| Procedure time | 1.0088 (1.0030 to 1.0149) | 0.0036 | |
| Final mTICI score | |||
| 0 | Ref | Ref | |
| 1 | 0.8000 (0.2100 to 3.0473) | 0.7436 | |
| 2a | 1.2632 (0.5456 to 2.9244) | 0.5854 | |
| 2b | 1.2291 (0.6061 to 2.4924) | 0.5674 | |
| 2c | 0.9787 (0.4259 to 2.2492) | 0.9596 | |
| 3 | 1.2695 (0.6247 to 2.5796) | 0.5095 | |
| Complications | 1.9053 (1.1090 to 3.2734) | 0.0195 | |
| Race | |||
| White | Ref | Ref | |
| Black | 1.0123 (0.6194 to 1.6545) | 0.9609 | |
| Hispanic | 0.1841 (0.0235 to 1.4387) | 0.1066 | |
| Non-white | 3.1668 (2.0317 to 4.9361) | <0.0001 | |
| PH2 | |||
| Distal embolization | 1.0237 (0.6574 to 1.5942) | 0.9174 | |
| Diabetes mellitus | 1.0106 (0.7169 to 1.4246) | 0.9519 | |
| Hypertension | 1.1232 (0.7813 to 1.6148) | 0.5303 | |
| Atrial fibrillation | 0.9754 (0.6963 to 1.3665) | 0.8850 | |
| Hyperlipidemia | 0.9211 (0.6604 to 1.2847) | 0.6281 | |
| Congestive heart failure | 0.9738 (0.5134 to 1.8469) | 0.9352 | |
| Prior stroke | 0.6286 (0.3770 to 1.0481) | 0.0750 | |
| Smoking status | |||
| Non-smoker | Ref | Ref | |
| Current smoker | 0.8030 (0.4395 to 1.4670) | 0.4753 | |
| Former smoker | 0.8126 (0.4788 to 1.3790) | 0.4417 | |
| Admission NIHSS | 0.9983 (0.9744 to 1.0227) | 0.8876 | |
| Anterior circulation | 1.0071 (0.5111 to 1.9841) | 0.9838 | |
| Time from onset to groin puncture | 0.9996 (0.9991 to 1.000) | 0.0734 | |
| Angioplasty | 1.6062 (0.9023 to 2.8591) | 0.1072 | |
| ADAPT | 0.8241 (0.5741 to 1.1830) | 0.2942 | |
| Time from puncture to occlusion site | 1.0097 (0.9766 to 1.0411) | 0.5462 | |
| Time from puncture to ≥mTICI 2b | 1.0052 (1.0018 to 1.0085) | 0.0022 | |
| Procedure time | 1.0113 (1.0053 to 1.0179) | 0.0004 | |
| Final mTICI score | |||
| 0 | Ref | Ref | |
| 1 | 1.6545 (0.5003 to 5.4718) | 0.4092 | |
| 2a | 1.5550 (0.6533 to 3.7014) | 0.3183 | |
| 2b | 1.1488 (0.5425 to 2.4326) | 0.7170 | |
| 2c | 0.9052 (0.3717 to 2.2048) | 0.8264 | |
| 3 | 0.9907 (0.4641 to 2.1150) | 0.9808 | |
| Complications | 2.7347 (1.5588 to 4.7974) | 0.0004 | |
| Race | |||
| White | Ref | Ref | |
| Black | 0.8677 (0.5013 to 1.5019) | 0.6122 | |
| Hispanic | 0.6626 (0.1819 to 2.4133) | 0.5325 | |
| Non-white | 1.8689 (1.1149 to 3.1329) | 0.0176 |
ADAPT, A Direct Aspiration First Pass Technique; HI, hemorrhagic infarction; mTICI, modified Thrombolysis In Cerebral Infarction; NIHSS, National Institutes of Health Stroke Scale; PH, parenchymal hemorrhage.
In the multivariable model, higher admission NIHSS (OR 1.0278, P=0.0383) and longer time from puncture to reperfusion (OR 1.0065, P=0.0054) were positive independent predictors of category) were positive independent predictors of PH1. Prior HI2. Higher admission NIHSS (OR 1.0400, P=0.0034), intrap-history of diabetes was a negative independent predictor of procedural complications (OR 2.0122, P=0.0255), and non-PH1 (OR 0.5879, P=0.0092). Higher admission NIHSS (OR white race (OR 3.0318, P<0.0001; with white as the reference 1.0314, P=0.0430), intraprocedural complications (OR 2.3290, P=0.0105), and non-white race (OR 1.9144, P=0.0234; with white as the reference category) were positive independent predictors of PH2 (table 5).
Table 5.
Multivariable multinomial logistic regression model for hemorrhage categories
| ECASS | Variables | Relative OR (95% CI) | P value |
|---|---|---|---|
| HI1 | |||
| Ref | Ref | Ref | |
| HI2 | |||
| Complications | 1.4690 (0.7750 to 2.7844) | 0.2384 | |
| Diabetes mellitus | 0.8530 (0.5832 to 1.2469) | 0.4111 | |
| Race (non-white) | 1.0166 (0.5854 to 1.7653) | 0.9535 | |
| Race (Hispanic) | 0.1602 (0.020 to 1.2726) | 0.0832 | |
| Race (black) | 1.2384 (0.7614 to 2.0145) | 0.3888 | |
| Admission NIHSS | 1.0278 (1.0015 to 1.0548) | 0.0383 | |
| Time from puncture to ≥mTICI 2b | 1.0065 (1.0019 to 1.0112) | 0.0054 | |
| PH1 | |||
| Complications | 2.0122 (1.0894 to 3.7167) | 0.0255 | |
| Diabetes mellitus | 0.5879 (0.3943 to 0.8767) | 0.0092 | |
| Race (non-white) | 3.0318 (1.8814 to 4.8855) | <0.0001 | |
| Race (Hispanic) | 0.1741 (0.0218 to 1.3877) | 0.0987 | |
| Race (black) | 0.9437 (0.5529 to 1.6107) | 0.8318 | |
| Admission NIHSS | 1.0400 (1.0130 to 1.0676) | 0.0034 | |
| Time from puncture to ≥mTICI 2b | 0.9988 (0.9938 to 1.0038) | 0.6374 | |
| PH2 | |||
| Complications | 2.3290 (1.2188 to 4.4506) | 0.0105 | |
| Diabetes mellitus | 1.2272 (0.8036 to 1.8741) | 0.3432 | |
| Race (other) | 1.9144 (1.0917 to 3.3570) | 0.0234 | |
| Race (Hispanic) | 0.7211 (0.1909 to 2.7233) | 0.6295 | |
| Race (black) | 0.9918 (0.5421 to 1.8147) | 0.9787 | |
| Admission NIHSS | 1.0314 (1.0010 to 1.0628) | 0.0430 | |
| Time from puncture to ≥mTICI 2b | 1.0041 (0.9987 to 1.0095) | 0.1374 |
AICc = 2448.27.
AICc, Akaike information criterion curve; ECASS, European Cooperative Acute Stroke Study; HI, hemorrhagic infarction; mTICI, modified Thrombolysis In Cerebral Infarction; NIHSS, National Institutes of Health Stroke Scale; PH, parenchymal hemorrhage.
Neurological outcome
Data on neurological outcome at 90 days was available for 2938 of 6860 MT patients and they were dichotomized by good (mRS 0–2; n=1167, 39.7%) and poor (mRS 3–6, n=1771, 60.3%) outcome. Univariable positive predictors of good outcome were current smoking status (P<0.0001; non-smoker as reference), higher ASPECTS (P<0.0001), and higher final mTICI score (P<0.0001). Univariable negative predictors of good outcome were older age (P<0.0001), non-white race (P=0.0004; white as reference), prior history of diabetes (P<0.0001), prior history of hypertension (P<0.0001), prior history of atrial fibrillation (P<0.0001), prior history of stroke (P<0.0001), higher pre-morbid mRS (P<0.0001), higher admission NIHSS (P<0.0001), use of IV-tPA (P<0.0001), ADAPT technique (P=0.0130), longer time from onset to groin puncture (P<0.0001), longer time from puncture to reperfusion (P<0.0001), longer procedure time (P<0.0001), distal embolization (P=0.0485), and intraprocedural complications (P<0.0001) (table 6).
Table 6.
Predictors of good outcome after mechanical thrombectomy for acute ischemic stroke
| Good outcome (n=1167) | Poor outcome (n=1771) | Univariable analysis | Multivariable analysis model 1* | Multivariable analysis model 2† | ||||
|---|---|---|---|---|---|---|---|---|
| OR (95% CI) | P value | oR (95% CI) | P value | oR (95% CI) | P value | |||
| Age, years (SD) | 65.6 (0.4) | 73.2 (0.3) | 0.0387 (0.0240 to 0.0618) | <0.0001 | 0.3619 (0.0118 to 0.1079) | <0.0001 | 0.0300 (0.0114 to 0.0775) | <0.0001 |
| Female, n (%) | 555/1167 (47.6) | 877/1771 (49.5) | 0.9244 (0.7973 to 1.0718) | 0.2977 | – | – | – | – |
| Race, n (%) | 0.0040 | 0.0085 | 0.0055 | |||||
| White | 553/851 (65.0) | 736/1257 (58.6) | Ref | Ref | Ref | Ref | Ref | Ref |
| Black | 132/851 (15.5) | 206 (1257 (16.4) | 0.8528 (0.6677 to 1.0893) | 0.2024 | 0.6499 (0.4311 to 0.9795) | 0.0395 | 0.6284 (0.4269 to 0.9249) | 0.0185 |
| Hispanic | 31/851 (3.6) | 41/1257 (3.3) | 1.0063 (0.6231 to 1.6252) | 0.9795 | 0.7048 (0.3152 to 1.5758) | 0.3941 | 0.5685 (0.2770 to 1.1671) | 0.1239 |
| Non-white | 135/851 (15.9) | 274 (21.8) | 0.6557 (0.5191 to 0.8284) | 0.0004 | 1.9114 (1.1246 to 3.2486) | 0.0167 | 1.5346 (0.9742 to 2.4175) | 0.0647 |
| Diabetes mellitus, n (%) | 261/1164 (22.4) | 563/1754 (32.1) | 0.6114 (0.5157 to 0.7250) | <0.0001 | 0.9432 (0.6653 to 1.3372) | 0.7425 | – | – |
| Hypertension, n (%) | 783/1164 (67.3) | 1393/1761 (79.1) | 0.5429 (0.4590 to 0.6422) | <0.0001 | 0.9358 (0.6461 to 1.3555) | 0.7258 | – | – |
| Atrial fibrillation, n (%) | 376/1159 (32.4) | 732/1755 (41.7) | 0.6711 (0.5746 to 0.7839) | <0.0001 | 1.2250 (0.8747 to 17157) | 0.2368 | – | – |
| Hyperlipidemia, n (%) | 540/1163 (46.4) | 802/1755 (45.7) | 1.0300 (0.8877 to 1.1951) | 0.6971 | – | – | – | – |
| Congestive heart failure, n (%) | 119/1027 (11.6) | 207/1517 (13.7) | 0.8294 (0.6519 to 1.0553) | 0.1257 | – | – | – | – |
| Prior stroke, n (%) | 127/999 (12.7) | 286/1501 (19.1) | 0.6187 (0.4934 to 0.7759) | <0.0001 | 0.8261 (0.5406 to 1.2623) | 0.3762 | – | – |
| Smoking status, n (%) | <0.0001 | 0.4394 | – | – | ||||
| Non-smoker | 680/1045 (65.1) | 1075/1531 (70.2) | Ref | Ref | Ref | Ref | – | – |
| Former smoker | 166/1045 (15.9) | 268/1531 (17.5) | 0.9792 (0.7889 to 1.2154) | 0.8488 | 1.0031 (0.6830 to 1.4732) | 0.9874 | – | – |
| Current smoker | 199/1045 (19.0) | 188/1531 (12.3) | 1.6734 (1.3412 to 2.0878) | <0.0001 | 1.3224 (0.8489 to 2.060) | 0.2166 | – | – |
| Anterior circulation, n (%) | 1043/1160 (89.9) | 1557/1753 (88.8) | 1.1222 (0.8809 to 1.4295) | 0.3488 | – | – | – | – |
| Pre-morbid mRS, mean (SD) | 0.3 (0.7) | 1.0 (1.3) | 0.0348 (0.0208 to 0.0569) | <0.0001 | 0.0454 (0.0172 to 0.1133) | <0.0001 | 0.0607 (0.0259 to 0.1356) | <0.0001 |
| Admission NIHSS, mean (SD) | 12.7 (6.7) | 17.2 (6.8) | 0.0169 (0.0101 to 0.0278) | <0.0001 | 0.0181 (0.0072 to 0.0443) | <0.0001 | 0.0137 (0.0053 to 0.0345) | <0.0001 |
| ASPECTS, mean (SD) | 8.5 (1.5) | 7.7 (1.9) | 15.1704 (9.1126 to 25.5575) | <0.0001 | 11.5580 (4.8282 to 28.3468) | <0.0001 | 11.7959 (5.1502 to 27.6201) | <0.0001 |
| IV-tPA, n (%) | 582/1166 (49.9) | 695/1764 (39.4) | 0.6524 (0.5619 to 0.7574) | <0.0001 | 1.0447 (0.7489 to 1.4575) | 0.7968 | – | – |
| Angioplasty, n (%) | 105/883 (11.9) | 147/1306 (11.3) | 1.0641 (0.8152 to 1.3889) | 0.6481 | – | – | – | – |
| ADAPT, n (%) | 422/1081 (39.0) | 718/1638 (43.8) | 0.8205 (0.7017 to 0.9594) | 0.0130 | 0.9693 (0.6669 to 1.4088) | 0.8700 | – | – |
| Time from onset to groin puncture, mean min (SD) | 363.9 (386.6) | 466.7 (523.6) | 0.03641 (0.0103 to 0.1187) | <0.0001 | 0.0362 (0.0040 to 0.2498) | 0.0005 | 0.0350 (0.0049 to 0.1996) | <0.0001 |
| Time from puncture to occlusion site, min (SD) | 15.7 (23.0) | 18.8 (27.4) | 0.3009 (0.0749 to 1.0961) | 0.0691 | – | – | – | – |
| Time from puncture to ≥mTICI 2b, min (SD) | 36.8 (31.1) | 47.7 (36.6) | 0.0420 (0.0183 to 0.0938) | <0.0001 | 0.3136 (0.0090 to 11.4607) | 0.5253 | – | – |
| Procedure time, min (SD) | 39.1 (37.4) | 56.8 (49.2) | 0.0021 (0.0004 to 0.0107) | <0.0001 | 0.0351 (0.0014 to 0.7661) | 0.0329 | 0.0036 (0.0003 to 0.0415) | <0.0001 |
| Final mTICI score, n (%) | <0.0001 | 0.0003 | <0.0001 | |||||
| 0 | 39/1148 (3.4) | 147/1736 (8.5) | Ref | Ref | Ref | Ref | Ref | Ref |
| 1 | 6/1148 (0.5) | 41/1736 (2.4) | 0.5516 (0.2184 to 1.3933) | 0.2082 | 0.8801 (0.0660 to 11.7353) | 0.9230 | 0.7434 (0.1264 to 4.3725) | 0.7429 |
| 2a | 30/1148 (2.6) | 140/1736 (8.1) | 0.8077 (0.4758 to 1.3712) | 0.4290 | 1.0963 (0.2818 to 4.2648) | 0.8945 | 0.8045 (0.2661 to 2.4322) | 0.7000 |
| 2b | 343/1148 (29.9) | 587/1736 (33.8) | 2.2025 (1.5102 to 3.2120) | <0.0001 | 3.0485 (0.9704 to 9.5769) | 0.0563 | 3.0038 (1.2166 to 7.4161) | 0.0171 |
| 2c | 128/1148 (11.2) | 166/1736 (9.6) | 2.9064 (1.9065 to 4.4307) | <0.0001 | 3.5746 (1.0664 to 11.9816) | 0.0390 | 3.4819 (1.3011 to 9.3174) | 0.0130 |
| 3 | 602/1148 (52.4) | 655/1736 (37.7) | 3.4642 (2.3929 to 5.0151) | <0.0001 | 4.9856 (1.5979 to 15.5552) | 0.0057 | 4.5460 (1.8365 to 11.2526) | 0.0011 |
| Distal emboli, n (%) | 183/1021 (17.9) | 321/1522 (21.1) | 0.8170 (0.6677 to 0.9998) | 0.0485 | 1.4048 (0.8977 to 2.1984) | 0.1368 | – | – |
| Intraprocedural complication, n (%) | 46/1028 (4.5) | 149/1564 (9.5) | 0.4449 (0.3165 to 0.6253) | <0.0001 | 0.3073 (0.1434 to 0.6585) | 0.0013 | 0.4160 (0.2182 to 0.7933) | 0.0077 |
Model includes statistically significant variables from univariate analyses.
Model includes statistically significant variables from model 1.
ADAPT, A Direct Aspiration First Pass Technique; ASPECTS, Alberta Stroke Program Early CT Score; IV-tPA, intravenous tissue plasminogen activator; mRS, modified Rankin Scale; mTICI, modified Thrombolysis In Cerebral Infarction; NIHSS, National Institutes of Health Stroke Scale.
In both multivariable models, higher ASPECTS (model 1: OR 15.1704, P<0.0001; model 2: OR 11.7959, P<0.0001) was a positive independent predictor of good outcome. Non-white race (OR 1.9114, P=0.0167; white as reference) was a positive independent predictor of good outcome in model 1 only. In both multivariable models, older age (model 1: OR 0.3619, P<0.0001; model 2: OR 0.0300, P<0.0001), black race (model 1: OR 0.6499, P=0.0395; model 2: OR 0.6284, P=0.0185), higher pre-morbid mRS (model 1: OR 0.0454, P<0.0001; model 2: OR 0.0607, P<0.0001), higher admission NIHSS (model 1: OR 0.0181, P<0.0001; model 2: OR 0.0137, P<0.0001), longer time from onset to groin puncture (model 1: OR 0.0361, P=0.0005; model 2: OR 0.0350, P<0.0001), longer total procedure time (model 1: OR 0.0351, P=0.0329; model 2: OR 0.0036, P<0.0001), and intraprocedural complications (model 1: OR 0.3073, P=0.0013; model 2: OR 0.4160, P=0.0077) were negative independent predictors of good outcome. Primary MT technique did not independently predict good outcome in either model.
DISCUSSION
It is evident that endovascular therapy is associated with improved neurological recovery after AIS.3 7 However, hemorrhagic transformation after AIS, particularly sICH, is associated with worse clinical outcomes.14 Major clinical trials have not reported a significant difference in any ICH or sICH rates between AIS patients treated with MT versus without MT.3 7 However, the risk factors for ICH and sICH may be different among MT-treated patients, and reducing the incidence of hemorrhagic transformation after MT may improve their neurological outcomes. In this study, we sought to identify modifiable risk factors for ICH after MT for AIS among patients treated at 33 stroke centers between 2013 and 2020. We found the rates of any ICH and sICH, which were 25% and 7%, respectively, to be comparable to those reported in the major clinical trials.3 7 This contrasts with several retrospective studies that reported higher rates of any ICH and sICH.5 6 8 This may be due to improvements in thrombectomy technique and periprocedural management over time, including device familiarity, standardization of stroke center certification guidelines, increased efficiency, and higher rates of successful reperfusion.15
In the ASTER and COMPASS trials, a difference in the rate of ICH or sICH between the use of aspiration versus stent retriever was not found.1 2 While we did not find technique-associated differences in the risk of any ICH, technique was an independent predictor of sICH in the STAR registry. Compared with stent retriever use, ADAPT was associated with a lower risk of sICH, suggesting that ADAPT-associated ICHs are predominantly asymptomatic. We did not find any interactions between use of ADAPT and TICI grade, time from onset to puncture, or time from puncture to reperfusion. This suggested that the observed association between thrombectomy technique and risk of sICH was independent of the degree of reperfusion or the efficiency of the procedure. We also found balloon angioplasty to be independently associated with increased risk of any ICH, but not sICH. Our observation of a technique-associated difference in risk of sICH did not translate into a difference in 90-day neurological outcomes between ADAPT and stent retriever. Although the effects of ICH after MT on clinical outcomes were not within the scope of the present study, a recent meta-analysis by Tang et al found asymptomatic ICH to be associated with worse 3-month clinical outcome among patients who underwent MT (adjusted OR 1.89, P=0.007).16 This suggests that efforts to reduce any ICH, asymptomatic or otherwise, are important in the neurological recovery of AIS patients after MT.
The primary molecular driver for hemorrhagic transformation after AIS is thought to be reperfusion injury.17 Reperfusion activates inflammatory cascades and reactive oxygen species, both of which can damage capillary endothelial cells and lead to increased vascular permeability.18 A downstream consequence is the activation of matrix metalloproteinases (MMPs), particularly MMP-9, which incite destruction of basal lamina collagen, thereby promoting vasogenic edema formation and injury to the surrounding parenchyma.19 The combination of poor baseline collaterals and successful recanalization after AIS has been associated with an increased risk of ICH, lending support to the hypothesis for reperfusion injury in patients experiencing ICH after MT.20 Direct mechanical trauma from the device itself may cause vascular injury, and periprocedural thrombus fragmentation can lead to distal migration, both of which may promote ICH after MT.21 22 The ECASS group first developed their grading system for hemorrhagic transformation in an attempt to characterize clinically significant ICH after AIS reperfusion.23 24 Initial emphasis was placed on distinguishing hemorrhagic infarction (HI), from parenchymal hematoma (PH). In the original ECASS trials, only PH2 was found to be associated with neurological deterioration and 3-month mortality after AIS. 9 Although these trials were primarily designed to address the safety and efficacy of thrombolysis for AIS, PH occurrence has also been associated with reduced favorable outcome and increased mortality after MT.14
We divided patients by ECASS grade to assess for differences in ICH predictors for each ECASS subtype. In contrast to our any ICH and sICH analyses, we did not find thrombectomy technique to be an independent predictor of ECASS grade, with HI1 as the reference category. MT-related complications were modifiable independent predictors of both PH1 and PH2, which is consistent with studies reporting vessel wall trauma to be a potential mechanism underlying clinically significant hemorrhage after recanalization for AIS.21 22 This highlights the importance of the technical aspect in stroke thrombectomy. Shorter time from puncture to clot engagement predicted any ICH, whereas longer time from puncture to reperfusion predicted HI1 but not sICH, PH1 or PH2. Longer total procedure time predicted sICH, PH1, PH2, and poor neurological outcome. It is conceivable that faster intracranial navigation methods using large bore catheters may increase the risk of vessel wall trauma. Taken together, these findings suggest that reducing time to reperfusion and total procedure time is an important goal for MT. Further investigation is needed to identify the mechanism for the observed relationship between shorter time to puncture to clot engagement and any ICH. We also found race to be an independent predictor of both PH1 and PH2. Relationships between race and hemorrhagic complications have been identified among patients receiving thrombolysis for AIS, and proposed mechanisms include differences between genetic risk factors, susceptibility to reperfusion therapies, underlying atherosclerotic disease burden, and mechanical sensitivities of the vessel wall.25 In our ECASS analysis, non-white race was associated with increased risk of both PH1 and PH2 and concurrently with reduced risk of sICH. The differential effects of race on risk of ICH after MT and the impact of ethnicity on outcomes following post-MT ICH warrant further investigation.
Higher NIHSS score at admission was associated with increased risk of PH1 and PH2 in our ECASS analysis, which is consistent with the established connection between greater stroke burden and risk of hemorrhagic transformation.26 Concordant with recent clinical trials investigating outcomes after thrombectomy, we did not find differences in the risk of ICH with or without use of adjunctive tPA.27 Interestingly, we found prior history of diabetes to be a negative independent predictor of PH1. Diabetes is an established risk factor for hemorrhagic transformation after thrombolysis for AIS.28 Hyperglycemia dysregulates blood–brain barrier hemostasis by altering endothelial cell function.29 However, the degree of stress hyperglycemia response, when defined as admission serum glucose relative to the baseline glycated hemoglobin (HbA1c), may be a stronger predictor of hemorrhagic transformation than pre-morbid blood glucose levels.30 Sulfonylurea medications may also have protective effects by dampening the inflammatory response after AIS.31 Future studies should seek to understand the mechanism by which pre-existing diabetes mellitus modifies the risk and severity of hemorrhagic transformation after MT for AIS.
There are several limitations to the present study, which affect the validity of our results. Importantly, the study protocol was not constructed to investigate specifically the association between AIS MT and risk of ICH. Therefore, our retrospective analysis is subject to confirmation bias as variables were chosen based on data availability and hypothesis generation. Specifically, data on blood pressure parameters, laboratory values including blood glucose levels, use of monitored anesthesia care versus general anesthesia, timing of post-MT neuroimaging studies, location of balloon angioplasty, and presence versus absence of collaterals were not available, and may have been relevant risk factors for post-MT ICH. Our results are also contingent on the accuracy and reliability of reported data, which were primarily based on medical chart abstraction. Data were self-reported and the principal investigators at each participating center were directly responsible for its fidelity and validity. It was not externally adjudicated by an independent data monitoring board. Therefore, this study is limited by potential heterogeneity in patient selection and data reporting between centers. Additionally, because data were obtained retrospectively, the total number of eligible patients and reasons for exclusion were not readily available. The uncorrected multiple comparisons performed within the study could have elevated the false discovery rate. Outcomes data were not available to test our hypothesis that post-MT hemorrhagic transformation would be associated with worse neurological recovery in the STAR patient cohort. Specifically, we were not able to compare differences in neurological outcomes between patients with any ICH, sICH or each of the ECASS grades. Therefore, we were unable to determine which of the identified modifiable risk factors in the present study is most critical to improving patient outcomes. Future studies should seek to address this limitation by investigating associations between post-MT hemorrhagic transformation risk factors and neurological outcomes in real-world datasets of patients stratified by ICH symptom status and by ECASS grade.
CONCLUSIONS
In this large multicenter retrospective cohort of real-world patient data, this study has identified potentially modifiable factors that may reduce ICH risk after AIS thrombectomy. Contrary to the findings of the ASTER and COMPASS trials, the STAR database indicated a reduced symptomatic hemorrhage risk with direct aspiration, when compared with stent retriever. When stratified by ECASS grade, MT-related complications and race were the risk factors that were congruent predictors between each respective category of overall ICH, symptomatic hemorrhage, PH1, and PH2. The findings of the present study suggest that additional investigation of technique and ethnicity effects on hemorrhagic transformation and patient outcomes after AIS thrombectomy is warranted.
Supplementary Material
WHAT IS ALREADY KNOWN ON THIS TOPIC
Mechanical thrombectomy has been adopted as the standard of care treatment for patients with acute ischemic stroke. Hemorrhage after acute ischemic stroke thrombectomy is associated with worse neurological outcomes.
WHAT THIS STUDY ADDS
This study identified modifiable and unmodifiable risk factors for hemorrhage after acute ischemic stroke thrombectomy using real-world data from a multicenter, multi-national database.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Thrombectomy technique appeared to affect risk of hemorrhage after acute ischemic stroke thrombectomy. There was a propensity towards an increased overall hemorrhage but reduced symptomatic hemorrhage risk with direct aspiration when compared with stent retriever.
Funding
The STAR registry currently receives funding from Penumbra, Stryker and Medtronic. MRL: Grants from the NIH (R01NS105692, R01NS088072, U24NS100654, UL1TR002319, R25NS079200) and the American Heart Association (18CDA34110295). Unrestricted educational grants from Medtronic, Stryker and Philips Volcano. Consultant for Medtronic. Minor equity/ownership interest in Proprio, Cerebrotech, Synchron. Adviser to Metis Innovative. JAG: Grants from the Georgia Research Alliance. Consultant for Cognition Medical. AMS: Research support from Penumbra, Stryker, Medtronic, and Siemens. Consultant for Penumbra, Stryker, Terumo, and Arsenal. MM: Consultant for Medtronic, and Cerenovus. Stock ownership in Serenity Medical, Synchron, and Endostream. RMS: Grants from the NREF, Joe Niekro Foundation, Brain Aneurysm Foundation, Bee Foundation, the NIH (R01NS111119-01A1, UL1TR002736, KL2TR002737), the National Center for Advancing Translational Sciences, the National Institute on Minority Health and Health Disparities, and Medtronic. Consultant for Penumbra, Abbott, Medtronic, InNeuroCo and Cerenovus.
Collaborators
Marielle Ernst, MD - Universitätsmedizin Göttingen, Göttingen, Germany. Stavropoula Tjoumakaris, MD, M. Reid Gooch, MD, Nabeel Herial, MD - Thomas Jefferson University, Philadelphia, PA, USA. Kyle Fargen, MD - Wake Forest Baptist Health, Winston-Salemn, NC, USA. Dileep R. Yavagal, MD, Eric C. Peterson, MD - University of Miami Health System, Miami, FL, USA. Alex Brehm, MD - Universitätsspital Basel, Basel, Switzerland. Edgar Samaniego, MD, MS - University of Iowa, Iowa City, IA, USA. Nitin Goyal, MD, Daniel Alan Hoit, MD, MPH, Violiza Inoa-Acosta, MD - University of Tennessee Health Science Center/Semmes Murphey Foundation, Memphis, TN, USA. Michael Cawley, MD, Gustavo Pradilla, MD, Brian Howard, MD - Emory University, Atlanta, GA, USA. João Reis, MD, Jaime Pamplona, MD, Rui Carvalho, MD - Centro Hospitalar Universitário de Lisboa Central, Lisbon, Portugal. Brian Hoh, MD, MBA, Nohra Chalouhi, MD- University of Florida, Gainesville, FL, USA. Ryan Hebert, MD - Yale University, New Haven, CT, USA. Louis J. Kim, MD, Melanie Walker, MD, University of Washington, Seattle, WA, USA. Russell Cerejo, MD - Alleghany Hospital, Pittsburgh, PA, USA. Giulia Frauenfelder, MD, Francesco Diana, MD - Aou S. Giovanni di Dio e Ruggi d’Aragona, Salerno, Italy. Fernanda Rodriguez-Erazú, MD - Médica Uruguaya, Montevideo, Uruguay. Waldo Guerrero, MD - University of South Florida, Tampa, FL, USA. Mehmet Akdol, MD - Washington Regional Medical, Fayetteville, AR, USA.
Footnotes
Competing interests PK, JRM, MM and MRL are members of the editorial board of JNIS.
Correction notice Since this article first published, Peter Wludyka has been added as an author.
Patient consent for publication Not applicable.
Ethics approval Not applicable.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.
Data availability statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
REFERENCES
- 1.Lapergue B, Blanc R, Gory B, et al. Effect of endovascular contact aspiration vs stent retriever on revascularization in patients with acute ischemic stroke and large vessel occlusion: the ASTER randomized clinical trial. JAMA 2017;318:443–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Turk AS, Siddiqui A, Fifi JT, et al. Aspiration thrombectomy versus stent retriever thrombectomy as first-line approach for large vessel occlusion (COMPASS): a multicentre, randomised, open label, blinded outcome, non-inferiority trial. Lancet 2019;393:998–1008. [DOI] [PubMed] [Google Scholar]
- 3.Berkhemer OA, Fransen PSS, Beumer D, et al. A randomized trial of intraarterial treatment for acute ischemic stroke. N Engl J Med 2015;372:11–20. [DOI] [PubMed] [Google Scholar]
- 4.O’Carroll CB, Aguilar MI. Management of postthrombolysis hemorrhagic and orolingual angioedema complications. Neurohospitalist 2015;5:133–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Venditti L, Chassin O, Ancelet C, et al. Pre-procedural predictive factors of symptomatic intracranial hemorrhage after thrombectomy in stroke. J Neurol 2021;268:1867–75. [DOI] [PubMed] [Google Scholar]
- 6.Neuberger U, Kickingereder P, Schönenberger S, et al. Risk factors of intracranial hemorrhage after mechanical thrombectomy of anterior circulation ischemic stroke. Neuroradiology 2019;61:461–9. [DOI] [PubMed] [Google Scholar]
- 7.Goyal M, Menon BK, van Zwam WH, et al. Endovascular thrombectomy after large-vessel ischaemic stroke: a meta-analysis of individual patient data from five randomised trials. Lancet 2016;387:1723–31. [DOI] [PubMed] [Google Scholar]
- 8.Zhang X, Xie Y, Wang H, et al. Symptomatic intracranial hemorrhage after mechanical thrombectomy in Chinese ischemic stroke patients: the Asian score. Stroke 2020;51:2690–6. [DOI] [PubMed] [Google Scholar]
- 9.Hacke W, Kaste M, Fieschi C, et al. Randomised double-blind placebo-controlled trial of thrombolytic therapy with intravenous alteplase in acute ischaemic stroke (ECASS II). Second European-Australasian acute stroke study Investigators. Lancet 1998;352:1245–51. [DOI] [PubMed] [Google Scholar]
- 10.von Elm E, Altman DG, Egger M, et al. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Int J Surg 2014;12:1495–9. [DOI] [PubMed] [Google Scholar]
- 11.Powers WJ, Rabinstein AA, Ackerson T, et al. Guidelines for the early management of patients with acute ischemic stroke: 2019 update to the 2018 guidelines for the early management of acute ischemic stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2019;50:e344–418. [DOI] [PubMed] [Google Scholar]
- 12.Zaidat OO, Yoo AJ, Khatri P, et al. Recommendations on angiographic revascularization grading standards for acute ischemic stroke: a consensus statement. Stroke 2013;44:2650–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Uyttenboogaart M, Stewart RE, Vroomen PCAJ, et al. Optimizing cutoff scores for the Barthel index and the modified Rankin scale for defining outcome in acute stroke trials. Stroke 2005;36:1984–7. [DOI] [PubMed] [Google Scholar]
- 14.Boisseau W, Fahed R, Lapergue B, et al. Predictors of parenchymal hematoma after mechanical thrombectomy: a multicenter study. Stroke 2019;50:2364–70. [DOI] [PubMed] [Google Scholar]
- 15.Powers WJ, Derdeyn CP, Biller J, et al. 2015 American Heart Association/American Stroke Association focused update of the 2013 guidelines for the early management of patients with acute ischemic stroke regarding endovascular treatment: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2015;46:3020–35. [DOI] [PubMed] [Google Scholar]
- 16.Tang G, Cao Z, Luo Y, et al. Prognosis associated with asymptomatic intracranial hemorrhage after acute ischemic stroke: a systematic review and meta-analysis. J Neurol 2022;269:3470–81. [DOI] [PubMed] [Google Scholar]
- 17.del Zoppo GJ, von Kummer R, Hamann GF. Ischaemic damage of brain microvessels: inherent risks for thrombolytic treatment in stroke. J Neurol Neurosurg Psychiatry 1998;65:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Arba F, Rinaldi C, Caimano D, et al. Blood-brain barrier disruption and hemorrhagic transformation in acute ischemic stroke: systematic review and meta-analysis. Front Neurol 2020;11:594613. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Zhao B-Q, Wang S, Kim H-Y, et al. Role of matrix metalloproteinases in delayed cortical responses after stroke. Nat Med 2006;12:441–5. [DOI] [PubMed] [Google Scholar]
- 20.Lee T-J, Roh HG, Kim JH, et al. Collateral and permeability imaging derived from dynamic contrast material-enhanced MR angiography in prediction of PH 2 hemorrhagic transformation after acute ischemic stroke: a pilot study. Neuroradiology 2021;63:1471–9. [DOI] [PubMed] [Google Scholar]
- 21.Renú A, Laredo C, Lopez-Rueda A, et al. Vessel wall enhancement and blood-cerebrospinal fluid barrier disruption after mechanical thrombectomy in acute ischemic stroke. Stroke 2017;48:651–7. [DOI] [PubMed] [Google Scholar]
- 22.Kaesmacher J, Boeckh-Behrens T, Simon S, et al. Risk of thrombus fragmentation during endovascular stroke treatment. AJNR Am J Neuroradiol 2017;38:991–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.del Zoppo GJ, Poeck K, Pessin MS, et al. Recombinant tissue plasminogen activator in acute thrombotic and embolic stroke. Ann Neurol 1992;32:78–86. [DOI] [PubMed] [Google Scholar]
- 24.Hacke W, Kaste M, Fieschi C, et al. Intravenous thrombolysis with recombinant tissue plasminogen activator for acute hemispheric stroke. The European Cooperative Acute Stroke Study (ECASS). JAMA 1995;274:1017–25. [PubMed] [Google Scholar]
- 25.Mehta RH, Cox M, Smith EE, et al. Race/ethnic differences in the risk of hemorrhagic complications among patients with ischemic stroke receiving thrombolytic therapy. Stroke 2014;45:2263–9. [DOI] [PubMed] [Google Scholar]
- 26.Puetz V, Dzialowski I, Hill MD, et al. Intracranial thrombus extent predicts clinical outcome, final infarct size and hemorrhagic transformation in ischemic stroke: the clot burden score. Int J Stroke 2008;3:230–6. [DOI] [PubMed] [Google Scholar]
- 27.Renú A, Millán M, San Román L, et al. Effect of intra-arterial alteplase vs placebo following successful thrombectomy on functional outcomes in patients with large vessel occlusion acute ischemic stroke: the CHOICE randomized clinical trial. JAMA 2022;327:826–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Lees KR, Walters MR. Acute stroke and diabetes. Cerebrovasc Dis 2005;20 Suppl 1:9–14. [DOI] [PubMed] [Google Scholar]
- 29.Broderick JP, Hagen T, Brott T, et al. Hyperglycemia and hemorrhagic transformation of cerebral infarcts. Stroke 1995;26:484–7. [DOI] [PubMed] [Google Scholar]
- 30.Yuan C, Chen S, Ruan Y, et al. The stress hyperglycemia ratio is associated with hemorrhagic transformation in patients with acute ischemic stroke. Clin Interv Aging 2021;16:431–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Kunte H, Busch MA, Trostdorf K, et al. Hemorrhagic transformation of ischemic stroke in diabetics on sulfonylureas. Ann Neurol 2012;72:799–806. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.