Initial Stroke Severity in Patients With Atrial Fibrillation According to Antithrombotic Therapy Before Ischemic Stroke

Introduction

Atrial fibrillation (AF) is the main cause of ischemic stroke.¹ Compared with patients with ischemic stroke without AF, those with AF exhibit high stroke severity and poor outcomes. For patients with AF, an oral anticoagulant (OAC) is an effective medication for reducing stroke risk.^2,3 For patients with AF with an elevated CHA₂DS₂-VASc score (≥2 for men or ≥3 for women), OACs are recommended based on the current American Heart Association/American College of Cardiology Foundation/Heart Rhythm Society 2019 guidelines.⁴ The European Society of Cardiology 2016 guidelines recommend OACs in male patients with a CHA₂DS₂-VASc score of 1 and female patients with a score of 2.⁵ In addition to reducing the risk of ischemic stroke, several studies have reported that the use of therapeutic anticoagulants (novel OACs [NOACs] or warfarin) might be associated with less stroke severity and favorable outcomes when a stroke occurs in patients with AF.^6–9 However, in the real world, several patients with AF are still not prescribed OACs due to various reasons.^10–13 In this study, we evaluated the trends of antithrombotic medication use in patients with AF before the occurrence of ischemic stroke in the recent decade. We also assessed the association of preceding antithrombotic medication use with initial stroke severity and outcomes at discharge using data from the nationwide stroke registry.^14,15

Methods

The data that support the findings of this study are available upon reasonable request to corresponding author. The request requires approval from the ethical committee of the Korean Stroke Registry (KSR).

Study Population and Clinical Variables

The KSR is a prospective multicenter cooperative hospital-based stroke registry that was established in 1999.¹⁴ In each participating hospital, data are registered online through a web-based system (www.strokedb.or.kr). The methodology of case ascertainment and adjudication has been extensively described elsewhere.^14–17 The auditing process includes monthly monitoring for data consistency and validity by specialized auditors, generation of queries, and corrections provided by researchers.¹⁸ A research protocol with a statistical analysis strategy was developed and submitted to the KSR committee for approval before analysis.

Between June 2008 and December 2018, a total of 77 579 patients with acute ischemic stroke (AIS; <7 days after onset) from 39 secondary and tertiary hospitals in Korea were enrolled in the KSR. From the registry, we only selected the patients who exhibited AF before stroke admission. When the number of registered patients with stroke was <100 per year in any hospital, we excluded the whole data for that year to facilitate trend validation, and these data were also not used in other analyses.

Preceding antithrombotic medication was defined as the documentation of patients receiving antithrombotic agents within 7 days before hospital admission. For the purpose of the study, preceding antithrombotic medication was categorized into 3 mutually exclusive groups: (1) no antithrombotic; (2) antiplatelet; and (3) OAC (warfarin, NOAC). The patients who received both antiplatelet and OAC before stroke belonged to the OAC group. For each patient, the data on demographic characteristics, risk factors, stroke subtype, initial National Institutes of Health Stroke Scale (NIHSS) score, and modified Rankin Scale (mRS) score before hospitalization (pre-mRS) and at discharge were collected from the KSR. Hypertension, cardiac arrhythmia, diabetes mellitus, hyperlipidemia, congestive heart failure, peripheral arterial disease, and transient ischemic attack (TIA) were ascertained on the basis of their diagnosis and treatment at admission. To determine the presence of anemia, chronic kidney disease (estimated glomerular filtration rate ≤60), and severe renal disease (estimated glomerular filtration rate ≤30), we used the laboratory test results available at the time of admission. We calculated the prestroke CHA₂DS₂-VASc score for each patient based on previously published criteria.¹⁹ Briefly, we assigned 2 points for age ≥75 years and history of stroke/TIA/systemic thromboembolism and 1 point for congestive heart failure, hypertension, diabetes mellitus, vascular disease, and age 65 to 74 years. We also calculated the Anticoagulation and Risk Factors in Atrial Fibrillation (ATRIA) bleeding score, which is used for stratifying the bleeding risk in patients who take warfarin.²⁰ During the calculation of the ATRIA bleeding score, 3 points were assigned for anemia or severe renal disease, 2 points for age ≥75 years, and 1 point for history of hemorrhage or hypertension. To investigate the proportion of patients with AF receiving OAC therapy without a high potential for bleeding complications, we selected patients with a CHA₂DS₂-VASc score ≥2 and an ATRIA score ≤4 as OAC therapy candidates.²¹

The primary outcome was the initial stroke severity at admission. The NIHSS score was used as a measure of stroke severity (range: 0–42, with a higher score indicating greater stroke severity).^22,23 Mild initial stroke severity was defined by an NIHSS score of 0 to 5.²⁴ The secondary outcome was functional outcome at discharge according to mRS score (range: from 0 [no symptoms] to 6 [death]).²⁵ Favorable outcome was defined by an mRS score of 0 to 2.²⁶ In addition, we examined the annual incidence trends of AF-related ischemic stroke as well as the preventive antithrombotic medication pattern before hospitalization.

Statistical Analysis

Continuous variables were presented as mean±SD or median and interquartile range for quantitatively skewed variables. Categorical variables were expressed as frequencies and percentages. Baseline characteristics were compared across the 3 therapeutic groups using the χ² test or Fisher exact test for categorical variables and the ANOVA or Kruskal-Wallis test for continuous variables. A Bonferroni correction was applied to determine statistical significance for multiple comparisons. We categorized the data by age: <40, 40 to 49, 50 to 59, 60 to 69, 70 to 79, and >80 years. We also used logistic regression models to investigate the factors that affected mild initial stroke severity. These analyses were adjusted for baseline demographic and clinical variables before the ischemic stroke event, including age, sex, hypertension, diabetes mellitus, congestive heart failure, hyperlipidemia, chronic kidney disease, peripheral arterial disease, previous TIA, anemia, and severe renal disease, to identify the independent contributing factors for the use of OAC therapy. Potential determinants based on the results of univariate analysis (P<0.1) and variables that were already known to be related to OAC therapy were selected for multivariate analysis. Odds ratios (OR) with 95% CI were reported. The Cochran-Mantel-Haenszel test was used to investigate the linear trend in proportion of patients that required preadmission antithrombotic therapy over time.

In addition, trends in AF incidence over time were investigated among a total of 77 579 patients with AIS using the Cochran-Armitage trend test with AF as the dependent variable. Two-sided P values of 0.05 were considered statistically significant. All analyses were performed using SAS version 9.4 (SAS Institute Inc, Cary, NC).

Ethics

The institutional review boards of all the participating institutions approved this study using data from KSR. Written informed consent or assent from relatives was obtained from patients.

Results

Of 77 579 patients with AIS, 6786 (8.7%) had a medical history of AF before stroke admission. The number of female patients was 3460 (51.0%). There was a significant difference in mean age among the three groups classified based on preceding antithrombotic use (P=0.027; Table 1). The mean age of patients was the lowest (72.3±10.5 years) in the OAC group compared with that in the no antithrombotic (72.8±11.0 years) or antiplatelet group (73.1±10.0 years). There was no significant difference in the pre-mRS scores among these groups. There were differences in the proportions of patients with hypertension, diabetes mellitus, hyperlipidemia, peripheral arterial disease, TIA, and anemia among these groups. The median CHA₂DS₂-VASc score was 3 (interquartile range, 2–4) on admission and the mean (±SD) scores in the no antithrombotic, antiplatelet, and OAC groups were 2.8±1.3, 3.0±1.2, and 3.0±1.3, respectively (P<0.001). The median ATRIA score was 3 (interquartile range, 1–5 in the no antithrombotic group and 1–4 in the other groups), and the mean (±SD) scores in the no antithrombotic, antiplatelet, and OAC groups were 3.0±2.4, 3.1±2.4, and 3.1±2.4, respectively (P=0.241). Intravenous thrombolysis was administered more frequently to patients in the no antithrombotic or antiplatelet group than to those in the OAC group. However, intra-arterial thrombolysis was performed more frequently in the OAC group than in the other 2 groups.

In total, 4009 patients had a prestroke CHA₂DS₂-VASc score ≥2 and ATRIA score ≤4 (OAC therapy candidates). Of these patients, 1082 (27.0%) were receiving OACs before stroke, 1649 (41.1%) were receiving antiplatelets, and 1278 (31.9%) were not receiving any antithrombotic medication (Table 2). The Figure shows that the proportion of OAC therapy candidates using OAC has been increasing since 2014 when NOAC was covered by health insurance. In contrast, the proportion of patients using antiplatelets decreased.

• Download figure
• Download PowerPoint

The severity of the neurological deficits, mRS score at discharge, and mortality are presented in Table 3. The NIHSS score at admission (median, interquartile range) varied significantly on the basis of preceding medication use: 8 (3–15), 7 (2–15), and 6 (2–14) in the no antithrombotic, antiplatelet, and OAC groups, respectively (P<0.001). We also observed a significant difference in favorable and nonfavorable mRS scores at discharge (P<0.001). The severity of neurological deficits (NIHSS score at admission and discharge and nonfavorable mRS score at discharge) was higher in patients in the no antithrombotic group. However, there was no significant difference between patients in the antiplatelet and OAC groups. There was no significant difference in mortality rates among the 3 groups (P=0.062). Because the use of thrombolytics might influence the discharge outcome, we performed subgroup analysis only for patients without thrombolysis. Favorable outcome at discharge was not different among the subgroups.

Table 4 shows the independent predictors of mild initial stroke severity (NIHSS score ≤5) in patients with AIS with AF. Compared with the no antithrombotic group, the antiplatelet (OR, 1.16 [95% CI, 1.03–1.31]) and OAC (OR, 1.25 [95% CI, 1.10–1.43]) groups exhibited a high chance of mild initial stroke severity. A history of hyperlipidemia (OR, 1.24 [95% CI, 1.11–1.38]) and TIA (OR, 1.45 [95% CI, 1.02–2.04]) were also associated with a high chance of mild initial stroke severity. Old age (OR, 0.81 [95% CI, 0.76–0.85]), female sex (OR, 0.67 [95% CI, 0.60–0.77]), high pre-mRS scores (OR, 0.78 [95% CI, 0.74–0.81]) were associated with a low chance of mild initial stroke severity. The baseline characteristics of patients with NIHSS scores ≤5 and >5 are presented in Table I in the Data Supplement.

Moreover, the proportion of patients with AIS with AF (previous AF before event and newly diagnosed AF at the time of admission) has been significantly increasing every year (Table II in the Data Supplement; P<0.001).

Discussion

In this large study with a 10-year study period using a nationwide stroke registry, we found that 27.0% of patients with AIS with indications for OAC use (known history of AF, CHA₂DS₂-VASc score ≥2, and ATRIA score ≤4) were taking OACs before stroke onset despite indications according to numerous international guidelines.^4,5 We assessed their baseline disability and bleeding risk using the pre-mRS and ATRIA scores, respectively. The scores were not significantly different among the groups according to their preceding medications. However, the reasons for not prescribing anticoagulants were not obtained from the registry. In the AF registries, they reported higher rate of OAC use than ours. For example, the GARFIELD-AF (Global Anticoagulant Registry in the Field-Atrial Fibrillation) and ORBIT-AF II (Outcomes Registry for Better Informed Treatment of Atrial Fibrillation) registries reported that 69% and 87% of patients with AF with CHA₂DS₂-VASc score ≥2 were taking OACs, respectively.²⁷ Despite geographic variability and the differences in the registration period (GARFIELD-AF: 2010–2016, ORBIT-AF II: 2013–2016), the proportion of patients using OACs in our study was lower than those in these registries. In the GARFIELD-AF study, OAC use was noted in 58% of South Korean patients with new-onset AF and CHA₂DS₂-VASc ≥2 within 6 weeks. This was higher than the 10-year average rate of OAC use in our study. The proportion of preceding OAC use in our study was lower than the proportion of OAC use for primary prevention in the AF registry; this may be explained by a lower occurrence of stroke in patients with AF using OACs than in those not receiving any antithrombotic medications. In addition, the use of OACs might be exaggerated in AF registry studies than in real-world practice. Obtaining informed consent could have resulted in a higher percentage of participating patients receiving OAC treatment than the proportion seen among all patients.²⁸ Previous studies that investigated the preceding OAC in patients with previous AF who experienced ischemic stroke reported the proportion of OAC use ranged from 15.7% to 29.8%.^8,29–31 After the emergence of NOACs, the use of OACs has increased from 22.2% in 2013 to 40.6% in 2018. Some other studies have reported an increase in NOAC use in patients with AF and OAC therapy candidates.^28,32 This could be attributed to stable pharmacokinetic profiles and nonrequirement for blood monitoring as well as the efficacy and safety of NOACs.³³

In our study, the initial stroke severity varied according to the preceding antithrombotic use. Patients without preceding antithrombotic medication use were more likely to have severe initial neurological impairment. Sufficient warfarin therapy plays an important role in stroke severity and infarct volume.^8,34 The relationship between NOAC use and small infarction volume and decreased risk of greater proximal artery occlusion has been reported.²⁹ Therefore, use of anticoagulants, including warfarin or NOAC, may be associated with a small infarct volume and mild symptoms.

Contrary to our expectation, there was no significant difference in initial stroke severity between the antiplatelet and OAC groups. Antiplatelet therapy in patients with AF decreases the occurrence of ischemic stroke; however, the decrease is less than that observed with OAC therapy.³⁵ A hypercoagulable state has been noted in AF with increased circulating platelet activators and plasma thrombotic complexes.³⁶ Based on this evidence, antiplatelet therapy could be considered to have an antithrombotic effect in patients with AF by partially inhibiting platelet aggregation. Therefore, antiplatelet therapy also has the potential to reduce ischemic stroke. In previous studies on AIS patients with AF, the stroke severity varied according to preceding medication use; stroke severity was lower in patients receiving the standard dose of NOACs or therapeutic international normalized ratio (INR) range of warfarin than in those receiving antiplatelets and no antithrombotics.^7,9,30,37 However, data on adherence to OAC use, INR for warfarin, or dosage of NOACs were not fully available in our study. It is possible that patients with subtherapeutic warfarin use or those receiving suboptimal doses of NOACs could exist in our OAC group, and this might have resulted in the lack of difference in initial stroke severity between the antiplatelet and OAC groups and overestimating the effect of antiplatelet therapy. Our findings should be interpreted with caution owing to the lack of information on the proportion of subtherapeutic INR in warfarin users and the exact doses of NOAC.

The proportion of thrombolysis was low in patients with AF taking OAC, and this might have been influenced by the exclusion criteria involving intravenous thrombolysis, high INR, or use of NOAC. Contrarily, intra-arterial thrombolysis was more frequent in the OAC group than in the no antithrombotic or antiplatelet group.

We should acknowledge the various limitations of this study. First, this was a retrospective observational analysis. Unlike randomized studies, the selection of patients and undocumented confounding factors could affect the validity of our findings. However, it was impossible to randomize patients with AIS with AF according to antithrombotic use before ischemic stroke. Second, we could not obtain data on the previous history of stroke (ischemic or hemorrhagic), except TIA, falling, and hemorrhagic complications, which could be a reason for not using antithrombotics. We assessed the patients’ functional status at baseline using pre-mRS scores, which were not significantly different among the groups according to their preceding medications. Third, we could not get initial INR levels of patients taking anticoagulants and the doses of NOACs in our study. Lastly, there were missing values of NIHSS scores (2.1%) and mRS scores at discharge (2.3%). We regarded those as missing at random without any response mechanism, and we used complete case analysis.

The strengths of our study included its large sample size of consecutive patients and the prospective, comprehensive collection of qualified hospital administrative data from 39 hospitals. This enabled analysis of all comorbidities presented at the time of admission. In addition, the KSR quality committee monitored the data and sent queries regularly to hospitals, which improved the data quality and reduced variation or missing data.¹⁸

Conclusions

During the study period, there was an increase in prestroke OAC use, whereas antiplatelet use decreased in patients with AIS with AF. Although preceding OAC use increased after the introduction of NOACs in clinical practice, several patients with AIS with known AF did not take OACs. The preventive use of OACs in patients with AIS with AF was associated with a significant higher likelihood of a mild initial neurological deficit and favorable outcome at discharge.

Acknowledgments

This study used the Korean Stroke Registry database supported by the Korean Stroke Society (KSR-2019-01). We would like to thank Editage (www.editage.co.kr) for English language editing.

Supplemental Materials

Tables I and II

Footnotes