One-Year Outcomes After Minor Stroke or High-Risk Transient Ischemic Attack
Korean Multicenter Stroke Registry Analysis
Background and Purpose—Patients with minor ischemic stroke or transient ischemic attack are at high risk of recurrent stroke and vascular events, which are potentially disabling or fatal. This study aimed to evaluate contemporary subsequent vascular event risk after minor ischemic stroke or transient ischemic attack in Korea.
Methods—Patients with minor ischemic stroke or high-risk transient ischemic attack admitted within 7 days of symptom onset were identified from a Korean multicenter stroke registry database. We estimated 3-month and 1-year event rates of the primary outcome (composite of stroke recurrence, myocardial infarction, or all-cause death), stroke recurrence, a major vascular event (composite of stroke recurrence, myocardial infarction, or vascular death), and all-cause death and explored differences in clinical characteristics and event rates according to antithrombotic strategies at discharge.
Results—Of 9506 patients enrolled in this study, 93.8% underwent angiographic assessment and 72.7% underwent cardiac evaluations; 25.1% had symptomatic stenosis or occlusion of intracranial arteries. At discharge, 95.2% of patients received antithrombotics (antiplatelet polytherapy, 37.1%; anticoagulation, 15.3%) and 86.2% received statins. The 3-month cumulative event rate was 5.9% for the primary outcome, 4.3% for stroke recurrence, 4.6% for a major vascular event, and 2.0% for all-cause death. Corresponding values at 1 year were 9.3%, 6.1%, 6.7%, and 4.1%, respectively. Patients receiving nonaspirin antithrombotic strategies or no antithrombotic agent had higher baseline risk profiles and at least 1.5× higher event rates for clinical event outcomes than those with aspirin monotherapy.
Conclusions—Contemporary secondary stroke prevention strategies based on thorough diagnostic evaluation may contribute to the low subsequent vascular event rates observed in real-world clinical practice in Korea.
Patients who have had a minor ischemic stroke or transient ischemic attack (TIA) are at high risk of recurrent stroke and vascular events. In earlier UK and US studies, during the first 90 days after stroke or TIA, the recurrent stroke rates ranged from 10% to 17%,1,2 and the risk of subsequent vascular events was as high as 25%.3,4 The immediate evaluation and initiation of contemporary vascular prevention strategies substantially reduced the 90-day rate of recurrent stroke by 80% to ≈ 2%.1,5 Consequently, in the recent multinational TIAregistry.org project, which comprised centers capable of urgent evaluation and treatment, the rates of recurrent stroke and of vascular events after minor stroke or TIA were 5.1% and 6.2%, respectively, at 1 year,6 which were substantially lower than rates reported previously.
In parallel with the higher stroke incidence in Asian populations compared with Western populations,7 there has been a concern that Asian patients with stroke might be at higher risk of recurrent stroke than Western patients with stroke. In 2 Japanese observational studies conducted in the mid-2000s, the risk of recurrent stroke after minor stroke or TIA was 10.4% at 90 days and 12.8% to 18.9% at 1 year.8,9 The CHANCE (Clopidogrel in High-Risk Patients With Acute Non-Disabling Cerebrovascular Events) trial, conducted in China, showed stroke recurrence rates in the first 90 days after a minor stroke or TIA of 11.7% in patients treated with aspirin and 8.2% in those receiving aspirin plus clopidogrel; corresponding rates of major vascular events were 11.9% and 8.4%, respectively.10 These rates were higher than observed in the TIAregistry.org project. However, there have been no comparable reports on rates of stroke recurrence and vascular events in Korea.
The aim of the present study was to estimate the rate of subsequent vascular events and recurrent stroke after minor ischemic stroke or TIA in patients in Korea, using data from a nationwide South Korean multicenter registry.11 An additional objective was to explore the association of different antithrombotic strategies with vascular event risk.
Materials and Methods
Study Design and Participants
This descriptive observational study used data from the Comprehensive Registry Collaboration for Stroke in Korea (CRCS-K) registry, which is a prospective, nationwide, multicenter, web-based acute stroke registry of consecutive patients with acute ischemic stroke admitted to academic or regional stroke centers in Korea.11 The study forms part of the multinational Assessment of Real-World Evidence in Stroke/TIA program and complements the SOCRATES (Acute Stroke or Transient Ischemic Attack Treated With Aspirin or Ticagrelor and Patient Outcomes) trial.12
Among a total of 17 416 individuals registered in the CRCS-K database between November 2010 and October 2013, we identified patients aged ≥20 years at stroke symptom onset who had been admitted to participating centers because of an acute ischemic stroke or TIA and whose admission was in the 7 days after stroke onset. The study population was restricted to patients with a minor ischemic stroke (National Institutes of Health Stroke Scale score ≤5) or high-risk TIA (defined as the presence of either acute lesions on diffusion-weighted MRI or ≥50% stenosis or occlusion in intracranial and extracranial cerebral arteries relevant to stroke symptoms). In addition, a SOCRATES-like subpopulation, meeting SOCRATES trial eligibility criteria (Table I in the online-only Data Supplement),12 was identified among the study participants.
Collection of clinical information for the purposes of monitoring and improving the quality and outcomes of stroke care was approved by the local institutional review boards of all participating centers of the CRCS-K registry with a waiver of consent because of study patients’ anonymity and minimal risk to these patients. Use of the registry database and additional review of medical records for the present study were also approved by the local institutional review boards.
We collected data from the CRCS-K registry on patient demographics, medical history, clinical and laboratory characteristics, stroke characteristics, medication prescriptions (before the qualifying stroke or TIA, in the first 48 hours after hospitalization and at discharge), and medication adherence (assessed by interview at 3 months after the qualifying event using the 6-item modified Morisky Medication Adherence Scale; high adherence was defined as motivation and knowledge domain scores ≥2). Information on diagnostic evaluations for etiologic work-up (including brain imaging and cerebral and cervical artery imaging) and cardiac evaluations was also obtained.
Using a predefined protocol from the CRCS-K registry, we prospectively captured vascular events during hospitalization and follow-up at 3 months and 1 year after the qualifying event at patients’ routine clinic visits or by telephone interviews with patients or their caregivers. To assure the accuracy of the outcome capture and to minimize the difference of the outcome capture process according to the interviewers, a set of uniform structured questionnaires (Table II in the online-only Data Supplement) was used, and regular education was provided for the interviewers. The primary outcome was a composite of stroke recurrence, myocardial infarction, or all-cause death. Secondary outcomes were stroke recurrence, occurrence of a major vascular event (a composite of stroke recurrence, myocardial infarction, or vascular death), all-cause death, vascular death, and myocardial infarction. Information on the occurrence of symptomatic intracranial hemorrhage (SICH) was also captured.
During the first 21 days after the qualifying stroke or TIA, recurrent stroke (fatal or nonfatal, ischemic or hemorrhagic) was defined as a new neurological deficit of rapid onset or worsening of an existing focal neurological deficit after a period of stability or improvement that lasted for ≥24 hours and was associated with relevant neuroimaging findings. After 21 days from the qualifying event, recurrent stroke was defined as rapid onset of focal neurological deficits presumptively attributable to vascular origin and lasting for ≥24 hours.
Vascular death was defined as any death during hospitalization because of the qualifying stroke and, after discharge, because of recurrent stroke, myocardial infarction, heart failure, or any sudden death without an identifiable nonvascular cause.
Myocardial infarction was defined as the presence of at least 2 of the following: symptoms of myocardial ischemia; enzyme changes indicative of myocardial infarction; and ECG changes suggesting myocardial ischemia.
SICH was defined as extravascular blood in the brain or cranium confirmed by neuroimaging and associated with neurological deterioration of ≥4 points on the National Institutes of Health Stroke Scale either from baseline or from the lowest score after baseline.
Statistical analyses were performed using SAS version 9.4 (SAS Institute Inc, Cary, NC), STATA version 11 (StataCorp, College Station, TX), and R (version 3.1.1, R Foundation for Statistical Computing, Vienna, Austria [64-bit]) software. Based on primary outcome rates in previous studies in minor ischemic stroke or TIA populations (3.7%,4 8.4%,7 11.7%,2 and 11.9%10), a primary outcome event rate of 10% within 3 months after the qualifying stroke or TIA was assumed for this study, and 10 000 participants would provide a 95% confidence interval within an error margin (ie, precision) of 0.6%.
Baseline characteristics were presented as frequency (percentage) for categorical variables and as mean±SD or median with interquartile range for continuous variables. Cumulative 3-month and 1-year event rates of primary and secondary outcomes, together with 95% confidence intervals, were calculated using the Kaplan–Meier (product-limit) method. In addition, we compared the recurrent stroke rate between patients admitted within 24 hours of stroke onset and those after 24 hours using the log-rank test.
Differences in baseline characteristics and event rates according to antithrombotic strategies were evaluated using Pearson χ2 test or Fisher exact test for categorical variables and ANOVA or the Kruskal–Wallis test for continuous variables. Variables with missing values for >5% of total observations were excluded from the analysis. Statistical significance was declared when a 2-tailed P value was <0.05.
Of 17 416 patients registered in the CRCS-K database between November 2010 and October 2013, 16 016 (92%) were hospitalized within 7 days of symptom onset because of an acute ischemic stroke or TIA. The final study population (n=9506) consisted of 9016 patients with a minor ischemic stroke (63% of all patients with ischemic stroke in the database) and 490 patients with a high-risk TIA (30% of all patients with a TIA in the database). Median follow-up was 369 days (interquartile range, 361–391 days). The SOCRATES-like subpopulation (ie, including age >40 years, not having a previous history of hemorrhagic stroke, and not being at high risk of cardioembolism) comprised 7121 patients with a minor ischemic stroke and 367 patients with a high-risk TIA. Eligibility criteria and the number of individuals included or excluded according to each criterion are summarized in Figure 1.
Key demographic characteristics at baseline are listed in Table 1. Table III in the online-only Data Supplement provides a more extensive list of demographics and information on diagnostic evaluations and treatment status. The mean age at presentation was 66 years, and 61% of the study participants were men. Hospitalization occurred in the first 24 hours after stroke symptom onset in 62.2% of patients; large artery atherosclerosis was the most common stroke subtype. In total, 25.1% of patients had symptomatic stenosis or occlusion in the relevant intracranial arteries.
Before the qualifying stroke or TIA, 28.4% of patients were receiving at least 1 antithrombotic agent, with aspirin being the most common (20.5%), and 14.4% were receiving a statin. Baseline characteristics of the SOCRATES-like subpopulation were similar to those of the overall study population, except for the distribution of stroke subtype because patients at high risk of cardioembolism were excluded from this subpopulation (Figure 1; Tables I and IV in the online-only Data Supplement).
Diagnostic Evaluations and Treatments
Diffusion-weighted MRI, angiographic evaluations, and cardiac evaluations were performed in 94.1%, 93.8%, and 72.7% of the study population, respectively (Table III in the online-only Data Supplement). In total, 5.4% of the patients received reperfusion therapies (intravenous thrombolysis: 4.7%; intra-arterial therapy: 1.2% [combined therapy: 0.5%]). Emergency carotid revascularization during index hospitalization was undertaken in 2.0% of individuals. Antithrombotic agents were administrated in the first 48 hours after admission in 96.1% of patients; aspirin was the most frequently used agent (87.4%), and more than one third of patients received antiplatelet polytherapy.
At discharge, 95.2% of the study population was prescribed antithrombotic agents; aspirin was the most common (76.5%), followed by clopidogrel (35.5%). Antiplatelet monotherapy and polytherapy were prescribed in 42.8% and 37.1% of patients, respectively, at discharge, and 15.3% received anticoagulant therapy. Clopidogrel plus aspirin was the most commonly prescribed antiplatelet polytherapy (28.2%), and aspirin was the most common drug prescribed as part of antiplatelet polytherapy (36.3%); 0.9% of patients received ≥3 antiplatelet agents.
For risk factor control, antihypertensive drugs, antidiabetic agents, and statins were prescribed at discharge in 45.8%, 25.3%, and 86.2% of the study population, respectively, corresponding to 64.2%, 71.3%, and 97.5% of patients with hypertension, diabetes mellitus, and dyslipidemia, respectively. Information on medication adherence at 3 months was available for 88.9% of the study population, >80% of whom reported high adherence (domain scores ≥2).
A total of 895 primary outcome events occurred, with cumulative event rates of 5.9% at 3 months and 9.3% at 1 year (Table V in the online-only Data Supplement; Figure 2). Cumulative event rates for stroke recurrence, major vascular events, and all-cause death at 3 months were 4.3%, 4.6%, and 2.0%, respectively, and at 1 year, they were 6.1%, 6.7%, and 4.1%, respectively. SICH occurred in 0.4% of patients during hospitalization, and 0.3% of patients experienced intracranial bleeding after discharge. Sensitivity analyses showed 3-month and 1-year stroke recurrence rates of 4.6% and 6.4%, respectively, in patients admitted in the first 24 hours after stroke onset and of 3.6% and 5.6%, respectively, in those hospitalized >24 hours after onset (P=0.15).
In the SOCRATES-like subpopulation, primary outcome event rates were 5.3% at 3 months and 8.4% at 1 year, and secondary outcome event rates were generally similar to those in the overall study population (Table VI in the online-only Data Supplement).
Compared with patients prescribed aspirin monotherapy, those receiving either other antithrombotic therapies or no antithrombotic agents were more likely to be older; have comorbid coronary heart disease, atrial fibrillation, hypertension, diabetes mellitus, or dyslipidemia; and have symptomatic stenosis or occlusion of intracranial and extracranial cerebral arteries (Table 2; Table VII in the online-only Data Supplement). SICH rate during hospitalization was 0.2% in patients who received antiplatelet polytherapy or antiplatelet plus anticoagulant therapy at discharge and 4.1% in individuals who received no antithrombotic drug prescriptions (Table 2). Cumulative event rates at 1 year for the primary outcome, stroke recurrence, major vascular events, and all-cause death were lower in the group receiving aspirin monotherapy (6.2%, 4.2%, 4.5%, and 2.4%, respectively) than the group receiving antiplatelet polytherapy (9.6%, 6.9%, 7.4%, and 3.3%, respectively) and combination of antiplatelet and anticoagulant (13.6%, 8.5%, 9.7%, and 7.0%, respectively). One-year event rates for the clinical outcomes were the highest in the group with no antithrombotic therapy (20.2% for primary outcome, 10.7% for stroke recurrence, 14.1% for major vascular event, and 16.6% for all-cause death; Figure 3; Figure in the online-only Data Supplement).
This large observational study, conducted in a Korean multicenter acute stroke registry, had several key findings. First, it provided current data on the event rates of stroke recurrence, myocardial infarction, vascular death, and all-cause death after a minor ischemic stroke or TIA in a representative Korean population, showing rates that were lower than expected. Second, diagnostic investigations for stroke pathogeneses were extensive: most patients underwent diffusion-weighted MRI and angiographic evaluation, and >70% of the patients underwent cardiac evaluation. Third, more than one third of patients were prescribed at least 2 antiplatelet agents at discharge.
The composite primary outcome event rate at 3 months was 5.9%, which was slightly lower than that found in the multinational SOCRATES trial (6.7%–7.5%).12 Three-month rates for major vascular events were also lower in our study than in recent clinical trials (4.6% in our study versus 6.4%–7.2% in the SOCRATES trial12 and 8.4%–11.9% in the CHANCE trial10) and were substantially lower than those reported in older observational studies from the United States and United Kingdom (25%3 and 18%,4 respectively).
Stroke recurrence rates in our study were lower than those previously reported (including in the CHANCE and SOCRATES trials10,12)1,3,10,12 although they were similar to those found in the TIAregistry.org project (Table VIII in the online-only Data Supplement).6 It is noteworthy that the risk of recurrent stroke in our study was about half or less of that in other recent studies in Asian populations,10,13 potentially because of proper control for vascular risk factors and application of high-intensity and stroke subtype-specific secondary prevention strategies to our patients.
Diagnostic evaluation for determining stroke subtype was performed extensively in our study. Angiographic evaluation was undertaken in 95% of patients, which is higher than the 80% to 88% in the TIAregistry.org project,6 and the 62% to 88% in the Ontario Stroke Registry14 (a statewide stroke registry capturing data on all consecutive patients with acute stroke or TIA at 11 regional stroke centers). The proportion of patients who underwent cardiac evaluations was also higher in our study (Table 1) than in previous studies. The use of a validated magnetic resonance–based diagnostic algorithm to assist with the TOAST classification (Trial of ORG 10172 in Acute Stroke Treatment)15 might have led to more accurate diagnosis of stroke pathogeneses and proper determination of antithrombotic strategies.
Regarding the antithrombotic strategies, the decision to prescribe antiplatelet polytherapy or add antiplatelet agents to anticoagulants in our study might have been related to the consideration of a high number of prevalent vascular risk factors. Although combination of antiplatelet agents is not recommended in current guidelines, 37% of the patients in our study were prescribed antiplatelet polytherapy, mainly clopidogrel plus aspirin, at discharge, which is much higher than the 27% in a community-based observational study in Scotland16 and the 13% in the TIAregistry.org project.6 Results from CHANCE and several clinical trials suggest that the addition of clopidogrel to aspirin might be more effective for secondary stroke prevention than aspirin monotherapy, particularly in the large artery atherosclerosis stroke subtype10,17,18; ≈30% of patients in the ticagrelor arm of the SOCRATES trial were assumed to receive short-term dual antiplatelet therapy because they were taking aspirin at the time of randomization.12 Statin was prescribed in 96% of patients in our study, which is much higher than the 67% in the TIAregistry.org project and the 42% in the CHANCE trial.6,10 Compared with the patients receiving anticoagulant monotherapy, those receiving antiplatelet agents in addition to anticoagulant therapy had more coronary artery diseases and intra- and extracranial artery diseases. In a meta-analysis, high-intensity and moderate-intensity oral anticoagulant therapies appeared to be effective in reducing the myocardial infarction and stroke rate but increased the risk of bleeding.19 In our study, the SICH rate in the antiplatelet plus anticoagulant group (0.2%) was not high. The incidence was highest in those with no antithrombotic agent (Table 2), potentially because patients at high risk of hemorrhage did not receive antithrombotics at discharge. Because our study is an observational cohort study, however, we could not conclude that more intensive antithrombotic strategies might not be effective for the prevention of clinical event even though the event rates of clinical outcomes were higher in patients receiving nonaspirin antithrombotic strategies and no antithrombotic agent than those with aspirin monotherapy.
Among patients with hypertension, diabetes mellitus, and dyslipidemia, 64%, 71%, and 98% received antihypertensive agents, antidiabetic agents, and statin, respectively, at discharge. Although we were unable to collect information about whether the target levels for blood pressure, hemoglobin A1c, or low-density lipoprotein cholesterol were established and achieved or not, self-reported adherence to medication was high (Table III in the online-only Data Supplement). The treatment rates in our study were much higher than those (35% for hypertension, 13% for diabetes mellitus, and 42% for dyslipidemia) during the follow-up period in the CHANCE trial.10
This study has several limitations. First, retrospective analyses lead to several inevitable limitations in retrospective observational studies. However, most of the data, especially event outcomes, were gathered prospectively according to predetermined protocols, and data quality was assured by central monitoring, research coordinator training, and regular audits.11 Second, generalizability of the study results might be limited because the CRCS-K registry consists of university hospitals or regional stroke centers although universal public health insurance provided by the national healthcare system in Korea allows easy access to participating centers and the representativeness of the CRCS-K registry is supported by an almost identical distribution of age and sex in the registry as in the broader Korean stroke population.20 Nevertheless, we could not exclude the possibility that the event rates in this study were underestimated compared with those in the general stroke population in Korea because being treated in referral stroke centers might increase the likelihood of adherence to urgent evaluations and rigorous secondary prevention strategies. Third, loss to follow-up, an unavoidable weak point of observational studies, was 5.3% at 1 year in our study which is lower than the 8.4% in the TIAregistry.org project although higher than in clinical trials.10,12 Fourth, it would have been useful to have information on the laterality of the lesion in patients with recurrent stroke because the stenosis degree or plaque vulnerability might increase the risk of ipsilateral stroke occurrence. However, we did not gather the information about whether the recurrent strokes were ipsilateral or not. Fifth, the lower stroke recurrence rates in our study compared with those found in clinical trials10,12 could indicate underestimation of outcome events in our study, which enrolled patients hospitalized in the 7 days after stroke. However, sensitivity analyses excluding patients who were hospitalized >24 hours after stroke onset showed similar recurrence rates to those in the overall study population. The lower recurrence rates could also be attributed to events simply having been missed during the outcome capture process. Despite our efforts of standardizing the process, the possibility of underestimation of the event rates because of simple missing is likely to have been higher in this study than clinical trials. Finally, information on oral anticoagulation doses, rates of complications (eg, gastrointestinal bleeding), and potential changes in treatment after discharge was not available.
The rates of outcome events after a minor ischemic stroke or high-risk TIA seem to be lower in Korea than in other populations from Western countries and contemporary China or Japan. Extensive diagnostic work-ups for accurate classification of stroke subtypes, thorough application of secondary stroke prevention strategies corresponding to current guidelines, and more intensive antithrombotic strategies may contribute to the observed differences in outcome event rates after the qualifying minor stroke or TIA. We still have opportunities to reduce clinical event outcomes after minor ischemic stroke or TIA by achieving optimal antithrombotic strategies and patient educations.
Sources of Funding
This study was supported by AstraZeneca and was partly funded by a grant from the Korea Healthcare Technology R&D Project, Ministry for Health and Welfare, Republic of Korea (grant number: HI10C2020). Editorial support was provided by Dr Anja Becher of Oxford PharmaGenesis, Oxford, United Kingdom, and was funded by AstraZeneca.
Drs Johansson and Lee are employees of AstraZeneca. The other authors report no conflicts.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.117.018045/-/DC1.
- Received May 14, 2017.
- Revision received August 16, 2017.
- Accepted September 1, 2017.
- © 2017 American Heart Association, Inc.
- Rothwell PM,
- Giles MF,
- Chandratheva A,
- Marquardt L,
- Geraghty O,
- Redgrave JN,
- et al
- Coull AJ,
- Lovett JK,
- Rothwell PM
- Selvarajah JR,
- Smith CJ,
- Hulme S,
- Georgiou RF,
- Vail A,
- Tyrrell PJ
- Hackam DG,
- Spence JD
- Ueshima H,
- Sekikawa A,
- Miura K,
- Turin TC,
- Takashima N,
- Kita Y,
- et al
- Kiyohara T,
- Kamouchi M,
- Kumai Y,
- Ninomiya T,
- Hata J,
- Yoshimura S,
- et al
- Ng VT,
- Bayoumi AM,
- Fang J,
- Burton KR,
- Stamplecoski M,
- Edwards JD,
- et al
- Kennedy J,
- Hill MD,
- Ryckborst KJ,
- Eliasziw M,
- Demchuk AM,
- Buchan AM
- Wong KS,
- Chen C,
- Fu J,
- Chang HM,
- Suwanwela NC,
- Huang YN,
- et al