Low-Dose vs Standard-Dose Alteplase in Acute Lacunar Ischemic Stroke

Objective To determine any differential efficacy and safety of low- vs standard-dose IV alteplase for lacunar vs nonlacunar acute ischemic stroke (AIS), we performed post hoc analyzes from the Enhanced Control of Hypertension and Thrombolysis Stroke Study (ENCHANTED) alteplase dose arm. Methods In a cohort of 3,297 ENCHANTED participants, we identified those with lacunar or nonlacunar AIS with different levels of confidence (definite/according to prespecified definitions based on clinical and adjudicated imaging findings. Logistic regression models were used to determine associations of lacunar AIS with 90-day outcomes (primary, modified Rankin Scale [mRS] scores 2–6; secondary, other mRS scores, intracerebral hemorrhage [ICH], and early neurologic deterioration or death) and treatment effects of low- vs standard-dose alteplase across lacunar and nonlacunar AIS with adjustment for baseline covariables. Results Of 2,588 participants with available imaging and clinical data, we classified cases as definite/probable lacunar (n = 490) or nonlacunar AIS (n = 2,098) for primary analyses. Regardless of alteplase dose received, lacunar AIS participants had favorable functional (mRS 2–6, adjusted odds ratio [95% confidence interval] 0.60 [0.47–0.77]) and other clinical or safety outcomes compared to participants with nonlacunar AIS. Low-dose alteplase (versus standard) had no differential effect on functional outcomes (mRS 2–6, 1.04 [0.87–1.24]) but reduced the risk of symptomatic ICH in all included participants. There were no differential treatment effects of low- vs standard-dose alteplase on all outcomes across lacunar and nonlacunar AIS (all pinteraction ≥0.07). Conclusions We found no evidence from the ENCHANTED trial that low-dose alteplase had any advantages over standard dose for definite/probable lacunar AIS. Classification of Evidence This study provides Class II evidence that for patients with lacunar AIS, low-dose alteplase had no additional benefit or safety over standard-dose alteplase. Clinical Trial Registration Clinicaltrials.gov identifier NCT01422616.


Conclusions
We found no evidence from the ENCHANTED trial that low-dose alteplase had any advantages over standard dose for definite/ probable lacunar AIS.

Classification of Evidence
This study provides Class II evidence that for patients with lacunar AIS, low-dose alteplase had no additional benefit or safety over standard-dose alteplase.
In routine clinical practice, patients with lacunar acute ischemic stroke (AIS) are eligible to receive IV thrombolysis, given comparable favorable outcomes to other common AIS pathologic subtypes. [1][2][3] These results were confirmed in a recent subgroup analysis of the Efficacy and Safety of Magnetic Resonance Imaging-Based Thrombolysis in Wake-Up Stroke (WAKE-UP) trial, where the safety and efficacy of standard-dose IV alteplase were comparable between lacunar and nonlacunar subtypes defined on baseline MRI. 4 Similar consistency of effect of IV alteplase between lacunar and nonlacunar AIS, defined by the Oxfordshire Community Stroke Project (OCSP) syndromic classification, was found in the third International Stroke Trial (IST-3). 5 Despite this evidence, some clinical concern persists over whether the modest risk of thrombolysis-related intracerebral hemorrhage (ICH) could offset the modest benefits of IV thrombolysis for lacunar AIS, where the natural course is generally more benign compared to other AIS subtypes 6 from there being no or small thrombotic lytic target on the presumption of a single penetrating artery occlusion. 7,8 In the alteplase-dose arm of the Enhanced Control of Hypertension and Thrombolysis Stroke Study (ENCHAN-TED), 9 a lower dose (0.6 mg/kg) of IV alteplase was shown to have a lower risk of ICH compared to standard dose (0.9 mg/kg) in thrombolysis-eligible patients with AIS. Whether it is the same for lacunar AIS is unclear. Herein, we report further analyses of the efficacy and safety of low-vs standarddose IV alteplase in the ENCHANTED participants with lacunar (versus nonlacunar) AIS who were identified by the combination of clinical and adjudicated imaging findings.

Primary Research Question and Evidence Level
Is there are any differential efficacy and safety of low-vs standard-dose IV alteplase between participants with lacunar and nonlacunar AIS in the alteplase dose arm of the ENCHANTED trial? This study provides Class II evidence that for patients with lacunar AIS, low-dose alteplase has no additional benefit or safety over standard-dose alteplase.

Design and Participants
ENCHANTED was an international, multicenter, 2 × 2 quasifactorial, prospective, randomized, open-label, blindedendpoint trial that assessed the effectiveness of low-dose (0.6 mg/kg; 15% as bolus, 85% as infusion during 1 hour) vs standard-dose (0.9 mg/kg; 10% as bolus, 90% as infusion during 1 hour) IV alteplase, and more intensive vs guidelinerecommended control of blood pressure (BP) in adult participants with AIS. The study design, participant characteristics, and main results of the alteplase-dose arm have been reported 9-11 for 3,310 patients with AIS recruited from 111 centers in 13 countries. Key demographic and clinical characteristics were recorded at the time of enrollment, with clinical severity defined according to the NIH Stroke Scale (NIHSS) at baseline, 24 hours, and at day 7 (or on discharge from hospital if earlier). A final clinical diagnosis of AIS subtypes based upon the opinion of site investigations, generally according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST) classification system, 12 was made at day 7, postrandomization (or on discharge from hospital, if earlier). Glossary AIS = acute ischemic stroke; BP = blood pressure; CI = confidence interval; CSVD = cerebral small vessel disease; CTA = CT angiography; DICOM = Digital Imaging and Communications in Medicine; ECASS = European-Australian Cooperative Acute Stroke Study; ENCHANTED = Enhanced Control of Hypertension and Thrombolysis Stroke Study; END = early neurologic deterioration; ICH = intracerebral hemorrhage; IST-3 = the third International Stroke Trial; LVO = large vessel occlusion; MRA = magnetic resonance angiography; mRS = modified Rankin Scale; NIHSS = NIH Stroke Scale; NINDS = National Institutes of Neurological Diseases and Stroke; OCSP = Oxfordshire Community Stroke Project; OR = odds ratio; sICH = symptomatic intracerebral hemorrhage; SITS-MOST = the Safe Implementation of Thrombolysis in Stroke-Monitoring Study; TICI = Thrombolysis in Cerebral Infarction; TOAST = Trial of Org 10172 in Acute Stroke Treatment classification; WAKE-UP = Efficacy and Safety of MRI-based Thrombolysis in Wake-Up Stroke trial.

Standard Protocol Approvals, Registrations, and Participant Consents
The study protocol was approved by the appropriate ethics committee at each participating center and written informed consent was obtained from participants or an appropriate legal surrogate according to the Declaration of Helsinki. The ENCHANTED trial was registered at ClinicalTrials.gov (Unique identifier: NCT01422616).

Imaging Analysis
Uncompressed digital images of all baseline and follow-up digital CT, MRI, and angiographic images were uploaded into the study brain imaging database in Digital Imaging and Communications in Medicine (DICOM) format identified only by the participant's unique study identification number. Images were analyzed centrally for any ICH by a trial adjudication panel, blind to clinical data, treatment, date, and sequence of scan. Assessors graded any identified symptomatic ICH (sICH) using a range of standard definitions from the Safe Implementation of Thrombolysis in Stroke-Monitoring Study (SITS-MOST), National Institute of Neurologic Disease and Stroke (NINDS), the European-Australian Cooperative Acute Stroke Study II (ECASS), ECASS III, and IST-3 (additional Methods I, doi.org/10.5061/dryad.t1g1jwt0s).
The ENCHANTED Imaging Analysis Project was established in August 2016, with the aim of defining the presence, extent, and severity of, and swelling from, acute ischemic changes (including arterial territory, border zone, small subcortical and brainstem/cerebellar infarcts), coexisting old vascular lesions and their subtypes, white matter lesions, and brain volume loss on all collected images by an imaging analysis team of trained individuals, blind to all clinical data, using an electronic scoring system modified from IST-3. 13 All observed infarct lesions on baseline (prerandomization) CT or MRI were coded according to the IST-3 criteria for infarct site and size. Separately and subsequent to primary scan reads, a neuroradiologist (Z.Z.) and neurosurgeon (X.C.) sought the ischemic lesion on 24-hour follow-up images while viewing the baseline images for those with no infarct lesion identified at baseline. They also assessed large vessel occlusion (LVO) on baseline CT angiography (CTA) or magnetic resonance angiography (MRA) according to a modified Thrombolysis in Cerebral Infarction (TICI) score for an abnormal artery in IST-3. 14 All the imaging data were cross-checked (Z.Z.) and a final rating made before unmasking the clinical data and randomization code for analyses.
Definitions of Lacunar and Nonlacunar AIS Different levels of confidence (definite/probable/possible) were used around the definitions of lacunar and nonlacunar AIS based on adjudicated imaging findings, clinical severity, and clinical diagnosis (additional Methods II, doi.org/10. 5061/dryad.t1g1jwt0s). In brief, definite lacunar AIS was defined when all 4 criteria were met: (1) the presence of acute infarct lesion (maximum diameter ≤20 mm) in the territory of penetrating arteries, with a rounded, ovoid, or tubular shape on axial CT or diffusion-weighted imaging/apparent diffusion coefficient map (typical examples are shown in figure 1) 15 ; (2) no LVO adjudicated centrally (on CTA/MRA) or reported by site investigators (on CTA/MRA/digital subtraction angiography); (3) the final diagnosis was reported as "small vessel or perforating vessel 'lacunar' disease" according to the TOAST criteria that involved any of the standard clinical lacunar syndromes with the lack of large vessel atheroma or cerebral cortical dysfunction; and (4) infarct side on images is consistent with that reported by site investigators. Definite nonlacunar AIS was defined as having acute infarct lesion with maximum diameter >20 mm or LVO on angiography. Participants were classified as nonlacunar if they had lacunar and nonlacunar infarcts.
Given that the clinical diagnosis of lacunar syndrome plus baseline NIHSS score <7 had a high specificity to predict imaging-confirmed lacunar stroke in IST-3, 16 probable lacunar and nonlacunar AIS were discriminated mainly by baseline NIHSS scores and final diagnosis in the situation that there was no acute infarct lesion identified on images or the images were not collected from the sites. For those with conflicting clinical and adjudicated imaging information that compromised the confidence of discrimination, we classified as possible lacunar or nonlacunar AIS according to the clinical diagnosis and LVO status.

Outcomes
The primary outcome of these analyses was the composite endpoint of disability or death (modified Rankin Scale [mRS] scores 2-6) at 90 days postrandomization. Secondary efficacy outcomes included major disability or death (mRS 3-6), death (mRS 6), and ordinal shift of the full range of mRS scores at 90 days. Secondary safety outcomes were sICH defined according to several criteria from other studies, fatal ICH within 7 days, ICH identified by central adjudicators, and any ICH adjudicated centrally or reported by site investigators. Other clinical outcomes included early neurologic deterioration (END) (≥4-point increase in NIHSS scores) or death within 24 hours or 7 days.

Statistical Analysis
Continuous or categorical variables at baseline were presented as mean (SD), median (interquartile range), or number (percentage). Baseline differences between participants with lacunar and nonlacunar AIS were evaluated using analysis of variance, χ 2 test, or Wilcoxon signed-rank test, as appropriate. Associations of lacunar AIS with 90-day function, safety, and other secondary outcomes were estimated in logistic regression models with adjustment for randomized treatment and key prognostic covariates (age, sex, ethnicity, baseline NIHSS score, time from symptom onset to randomization, premorbid function [mRS score 0 or 1], prior use of antithrombotic agents, history of diabetes or cardiovascular disease, and assigned to intensive blood pressure-lowering group). The treatment effect of low-vs standard-dose alteplase was determined in logistic regression models and the heterogeneity of alteplase dose effect across participants with lacunar and nonlacunar AIS was estimated by adding an interaction term to statistical models. Proportional odds regression models were used to analyze ordinal mRS scores. The primary analyses pertain to participants with definite/probable lacunar and nonlacunar AIS after excluding those with a possible diagnostic classification. Sensitivity analyses of the treatment effects of low-vs standard-dose alteplase were performed in participants with definite lacunar/nonlacunar AIS and in all participants with possible lacunar/nonlacunar AIS. We also performed an exploratory analysis of the treatment effects in a subset of lacunar AIS identified at baseline (infarct size ≤15 mm and no adjudicated LVO). Data were reported as odds ratios (ORs) and 95% confidence intervals (CIs) and a 2-sided p < 0.05 was considered statistically significant. All analyses were performed using SAS version 7.1 and Stata version 12.0.

Data Availability
Additional methods (I and II) and data (supplementary tables 1-3) are available from Dryad (doi.org/10.5061/dryad. t1g1jwt0s). Individual de-identified participant data used in this analysis will be shared by request from any qualified investigator via the Research Office of The George Institute for Global Health.

Baseline Characteristics
Among 3,297 AIS participants in the ENCHANTED alteplase dose arm, 2,588 (78.5%) were classifiable (definite lacunar, n = 195; probable lacunar, n = 295; definite nonlacunar, n = 1,697; and probable nonlacunar, n = 401 AIS) for inclusion in the primary analysis (figure 2). Compared to the 709 excluded participants, they were more likely to be older, have higher baseline NIHSS scores, be Asian, have a history of cardiovascular disease, and have a final diagnosis of LVO, but they also had shorter time interval from symptom onset to randomization (supplementary table 1, doi.org/10.5061/dryad. t1g1jwt0s). Table 1 shows that all the baseline clinical characteristics were significantly different between definite/ probable lacunar and nonlacunar AIS except for sex, history of diabetes, and prior use of statin/other lipid-lowering agents. Participants with lacunar (versus nonlacunar) AIS were younger and had milder neurologic impairment, higher baseline BP, and a lower proportion with conventional cardiovascular risk factors except smoking. In keeping with the lacunar pattern of stroke, few participants had multiple lesions in both anterior and posterior circulation, but they were more likely to have a lesion only in the posterior circulation. They were also less likely to have brain atrophy or a hyperdense vessel sign on CT or hyperintense arteries on MRI.

Lacunar AIS and Outcomes
Compared to participants with definite/probable nonlacunar AIS, those with definite/probable lacunar AIS had better 90day functional outcomes, whether defined by the outcome of mRS scores 2-6 (unadjusted OR 0.26, 95% CI 0.21-0.33), mRS scores 3-6 (0.20, 0.15-0.26), ordinal shift in the full range of scores (0.27, 0.23-0.33), or death alone (0.04, 0.01-0.12) (table 2). They were also less likely to have ICH and END or death after IV thrombolysis. The findings persisted with adjustment of baseline covariables and randomized alteplase dose.
Specifically, in the definite subgroup of lacunar AIS, there were no significant differences on the primary efficacy outcome (mRS 2-6) (33.7% vs 32.9%, adjusted OR 0.96, 95% CI 0.49-1.87) or major disability or death (mRS 3-6) (20.2% vs 15.3%, adjusted OR 1.31, 95% CI 0.54-3.19) between lowdose and standard-dose alteplase groups (figure 5). There was one case of sICH (0.9%) meeting NINDS and IST-3 criteria in participants with definite lacunar AIS treated by low-dose alteplase, but no case of sICH was observed after use of standard-dose alteplase. In participants with definite lacunar

Discussion
In these post hoc analyzes of the ENCHANTED trial, we did not identify any benefit, nor any harm, from the use of low-dose alteplase vs standard-dose alteplase to treat patients with lacunar AIS compared to those with other subtypes of AIS. As well as having a range of significantly different characteristics, the 90-day outcomes were better for those with lacunar than nonlacunar AIS, which provided some internal consistency for the classifications used in our study. However, given the low event rate of sICH, with fewer than 5 events in the primary analysis for definite or probable lacunar AIS, we are limited in the conclusions that can be drawn as to whether a lower dose of IV alteplase should be preferred because of the good prognosis for lacunar AIS.
Our results on thrombolysis outcomes for lacunar AIS are consistent with prior observational studies. 2,[17][18][19][20][21] However, the net benefit of thrombolysis for lacunar AIS is still debated, mainly because the evidence is drawn from subgroup analyzes of trials, such as WAKE-UP 4 and IST-3, 5 where there is low statistical power. In addition, accurate identification of lacunar AIS is challenging, especially in the absence of an acute lesion on the initial CT, and even MRI (in nearly one third of patients with nondisabling stroke). 22 The pragmatic approach of applying a lacunar syndrome classification system in studies has moderate diagnostic sensitivity and specificity, 15 which may potentially mix patients with nonlacunar AIS with the target population of lacunar AIS, and nondifferentially bias results towards IV thrombolysis.
We were unable to confirm in ENCHANTED participants any benefit of low-dose over standard-dose alteplase in lacunar AIS. The fact that there were few cases of sICH in the lowdose alteplase group, and no sICH in the standard-dose group, highlights the potential for chance and imprecise estimates of treatment effects when there are few events. Even with current imaging techniques and clinical criteria, it is difficult to discriminate lacunar AIS due to occlusion of a deep penetrating arteriole presumed caused by progressive lipohyalinosis from thrombosis related to atherosclerosis or embolus. Platelet activation triggered by disintegration of the endothelium from intrinsic cerebral small vessel disease (CSVD) may also be relevant in this type of AIS. 8 It is possible, therefore, that IV thrombolysis may have a differential effect dependent on the cause of lacunar stroke, being more effective when there is underlying thromboembolism. In lacunar AIS, we noted a significant imbalance in the frequency of background white matter lesions between the low-dose and standard-dose alteplase groups (41.8% vs 30.7%), which could partly account for more ICH in the former (supplementary table 3, doi.org/10.5061/dryad.t1g1jwt0s). 23 Again, however, due to the few sICH events in patients with lacunar AIS, we cannot confirm whether the increase in sICH by lowdose alteplase was confounded by CSVD. Some strengths of our study include the large, prospective, multicenter cohort of patients with AIS who had systematic, complete, and high-quality data collected prospectively, where we were able to adjust for multiple covariables in statistical models. Furthermore, the imaging assessment was completed blind to clinical features and other data, using a rigorously defined approach developed for the IST-3 study. However, we acknowledge limitations that include insufficient statistical power and inevitable selection bias from the data being derived from a clinical trial where a large number of participants were from Asia and had mild to moderate stroke. Moreover, given the pragmatic nature of ENCHANTED, few participants had a baseline brain MRI, and the identification of lacunar AIS required analysis of follow-up images with comparison to those obtained at baseline. Whereas this approach may have altered the imaging appearances of acute ischemic lesions after use of IV thrombolysis 24 and  a Refers to the effect of IV thrombolysis in definite/probable lacunar stroke vs nonlacunar stroke after pooling the 2 groups of randomized alteplase dose as one cohort. b Adjusted for key prognostic covariates (age, sex, ethnicity, baseline NIHSS score, time from stroke onset to randomization, premorbid function [mRS score 0 or 1], prior use of antithrombotic agents [aspirin, other antiplatelet agent, or warfarin], history of diabetes or cardiovascular disease [stroke, atrial fibrillation, coronary artery disease, valvular or other heart disease], assigned to intensive blood pressure-lowering group, and randomization to low-dose alteplase group) for functional outcomes. Adjusted for minimization and key prognostic covariates (age, baseline NIHSS score, time from stroke onset to randomization, assigned to intensive blood pressure-lowering group, and randomization to low-dose alteplase group) for safety outcomes and neurologic deterioration within 24 hours and 7 days. c Significant. limited the identification of all true lacunar AIS, our results are comparable with previous work showing that nearly one-third of patients with nondisabling AIS lack an infarct lesion on acute MRI (median 4 days poststroke). 22 In the ENCHANTED alteplase arm, 27.1% (789/2,916) of participants had no infarct lesion on either the baseline or 24-hour follow-up images. Thus, we had to use a combination of clinical and adjudicated imaging data to classify as many cases as possible into lacunar and nonlacunar AIS, which likely closely represents that used in routine practice. Relatively small samples in lacunar AIS compromised the power of a reliable assessment of any interaction, especially for sICH. Moreover, regarding the outcomes of major disability or death, a p interaction of 0.07 might have been due to chance rather than true differential treatment effects of low-vs standard-dose alteplase across definite/probable lacunar and nonlacunar AIS. Future research in systematic reviews and clinical registries may be required to confirm or refute these findings.
We found no clear evidence that low-dose IV alteplase was any better or safer than standard-dose alteplase in the