Abstract
Background
Atrial fibrillation (AF) increases risk of ischemic stroke, and oral anticoagulation (OAC) increases risk of intracerebral hemorrhage (ICH). This study aimed to compare OAC‐treated AF patients with an ischemic stroke/transient ischemic attack (TIA) or spontaneous ICH as their first lifetime cerebrovascular event, especially focusing on patients with therapeutic international normalized ratio (INR).
Hypothesis
We assumed that in AF patients suffering ischemic stroke/TIA or ICH, patient characteristics could be different in patients with therapeutic INR than in patients with warfarin.
Methods
FibStroke is a multicenter, retrospective registry collating details of AF patients with ischemic stroke/TIA or intracranial hemorrhage in 2003–2012. This substudy included AF patients on OAC with first lifetime ischemic stroke/TIA or spontaneous ICH.
Results
A total of 1457 patients with 1290 ischemic strokes/TIAs and 167 ICHs were identified. Of these, 553 (42.9%) strokes/TIAs and 96 (57.5%) ICHs occurred in patients with INR within therapeutic range. During OAC with therapeutic INR, congestive heart failure (odds ratio [OR]: 2.33, 95% confidence interval [CI]: 1.18–4.58) and hypercholesterolemia (OR: 2.52, 95% CI: 1.51–4.19) were more common in patients with ischemic stroke/TIA, whereas a history of bleeding (OR: 0.30, 95% CI: 0.11–0.82) was less common when compared with patients with ICH. In the whole cohort, renal impairment (OR: 1.86, 95% CI: 1.23–2.80) and mechanical valve prosthesis (OR: 4.41, 95% CI: 1.32–14.7) were overrepresented in patients with stroke/TIA, whereas aspirin use (OR: 0.52, 95% CI: 0.30–0.91) and high INR (OR: 0.40, 95% CI: 0.33–0.48) were overrepresented in patients with ICH.
Conclusions
In anticoagulated AF patients with therapeutic INR and first lifetime cerebrovascular event, congestive heart failure and hypercholesterolemia were associated with ischemic stroke/TIA and history of bleeding with ICH.
Keywords: Anticoagulation, Atrial Fibrillation, Intracerebral Hemorrhage, Stroke
1. INTRODUCTION
Atrial fibrillation (AF) is associated with an increased risk of thromboembolic complications, such as stroke and transient ischemic attack (TIA). Oral anticoagulation (OAC) therapy prevents two‐thirds of AF‐related ischemic strokes,1 but it increases the risk of bleeding complications, with intracerebral hemorrhage (ICH) being the most feared due to its high morbidity and mortality. In recent randomized clinical trials on patients with nonvalvular AF, the risk of ischemic stroke in patients receiving warfarin therapy remained at 1.3% to 2.4% per year with a lower risk of ICH (0.4%–0.8% per year).2, 3, 4, 5 In clinical practice, the risk for ischemic stroke in OAC‐naïve AF patients can be evaluated with the CHA2DS2‐VASc score and the bleeding risk estimated with the HAS‐BLED score.6, 7 Nevertheless, little is known about the differences in the characteristics of patients with thrombotic vs bleeding events. Especially, data on patients with renal impairment—a well‐known risk factor for both thrombotic and bleeding events—are scarce. In addition, despite the fact that effectiveness is dependent on ensuring that warfarin therapy remains within the therapeutic range, the majority of bleeding events occur while patients' international normalized ratio (INR) values are in the therapeutic range. Therefore, it would be useful to identify clinical predictors of these events.
We compared the clinical characteristics, risk factors, and therapeutic control in AF patients with their first ischemic stroke/TIA or ICH. In addition, we explored whether patients suffering from ischemic stroke and ICH differ with respect to renal impairment.
2. METHODS
The FibStroke study (http://www.ClinicalTrials.gov identifier NCT02146040) is part of a wider ongoing protocol assessing thrombotic and bleeding complications of AF in Finland.8 The FibStroke registry included information on all consecutive patients with AF who suffered a stroke, TIA, or ICH during the study period. The primary population was collected from the discharge registries of 2 university hospitals and 2 central hospitals, which treat all acute strokes for a residential population of 1.2 million inhabitants. The initial screening was conducted by identifying all patients from the hospital discharge records from 2003 to 2012 (2006–2012 in 1 central hospital) with the following criteria: the patient had been diagnosed with (1) an ischemic stroke, TIA, or ICH and (2) AF or atrial flutter. The temporal relationship between the stroke and a diagnosis of AF was not restricted and also included patients with AF diagnosed after the stroke. A total of 5676 patients with 6715 strokes, TIAs, or ICHs met these criteria. After the initial screening, all patient files were reviewed individually and data were collected in a structured electronic case‐report form.
This prespecified substudy included only patients fulfilling the following criteria: (1) ischemic stroke/TIA or spontaneous ICH, (2) permanent or paroxysmal AF diagnosed prior to or at the onset of the cerebrovascular event, and (3) OAC therapy on admission. Patients with previous ischemic stroke/TIA or ICH, as well as patients with ICH due to subdural, subarachnoid, or traumatic hemorrhage, were excluded. The study flowchart is shown in Figure 1. The CHA2DS2−VASc and HAS‐BLED scores were calculated as previously described,6, 9 with the exception that we calculated the modified HAS‐BLED without using labile INR; furthermore, hypertension (HTN) was categorized as history of HTN/treatment for HTN and hypercholesterolemia was categorized as treatment for hypercholesterolemia. In the FibStroke registry, history of bleeding was defined as any previous major bleeding, but from this substudy patients with previous ICH were excluded. At the time of admission, INR values of 2.0 to 3.0 in patients with nonvalvular AF and 2.5 to 3.5 in patients with a mechanical prosthetic valve were considered to be within the therapeutic range.10, 11 Impaired renal function was defined as estimated glomerular filtration rate (eGFR) ≤60 mL/min/1.72 m2, calculated using the Modification of Diet in Renal Disease (MDRD) Study equation.
Figure 1.
The study flowchart. Abbreviations: AF, atrial fibrillation; ICH, intracerebral hemorrhage; OAC, oral anticoagulation; SAH, subarachnoid hemorrhage; SDH, subdural hematoma; TIA, transient ischemic attack
2.1. Statistical analysis
The t test and Mann–Whitney U test were used in the analyses of continuous variables, and χ2 and Fisher exact tests were used to compare differences between dichotomized variables. Multivariate logistic regression was applied to evaluate patient characteristics in patients with ischemic stroke/TIA and ICH. Variables correlating (P < 0.1) with the dependent variable by univariate analyses were entered into the model as covariates. The same analyses were conducted in a subgroup of patients with admission INR within the target range. Statistical analyses were performed using the SPSS software program, version 24.0 (IBM Corp., Armonk, NY). Continuous data are presented as mean ± SD or median (interquartile range) and categorical variables as absolute numbers and percentages. A P value <0.05 was considered significant, and all the tests were 2‐sided.
3. RESULTS
The final study population consisted of 1290 patients with their first lifetime ischemic stroke/TIA and 167 patients with their first lifetime ICH (Table 1). Of the ICHs, 44% (n = 73) were lobar and 53% (n = 89) were deep. The mean age (78.0 ± 11.3 vs 78.4 ± 12.4 years) and sex distribution (male 49.7%, female 54.8%) in the groups were similar. Patients with ischemic stroke/TIA had a higher CHA2DS2‐VASc risk score than did patients with ICH (4 (3‐5) vs 3 (3‐5)). In contrast, there were no differences between the groups in the modified HAS‐BLED score (2 (1‐2) vs 2 (2‐3)). Ischemic stroke/TIA patients had lower eGFR (67 (53‐82) vs 76 (56‐91) mL/min/1.73 m2; P = 0.001) and lower INR values (2.0 (1.6‐2.4) vs 2.7 (2.3‐3.1); P = 0.001) than ICH patients (Figure 2). The proportion of patients with INR <2.0 was higher among the stroke/TIA patients (47.8% vs 9.7%; P = 0.001), and INR >3.0 was more frequent in the ICH patients (8.6% vs 32.1%; P = 0.001). Stroke/TIA patients had more comorbidities such as congestive heart failure (CHF; P = 0.001) and hypercholesterolemia (P = 0.004) compared with ICH patients, as well as a trend toward higher prevalence of mechanical heart valve prosthesis (P = 0.053). On the contrary, ICH patients more frequently had a history of bleeding (6.6% vs 3.3%; P = 0.036). There were no differences between the stroke/TIA and ICH patients with respect to medical therapy, although there was a trend toward more frequent use of nonsteroidal anti‐inflammatory drugs among the stroke/TIA patients (P = 0.064).
Table 1.
Baseline characteristics of study patients
| Whole Cohort, N = 1457 | P Value | INR in Therapeutic Range, n = 649 | P Value | |||
|---|---|---|---|---|---|---|
| Variable | Ischemic Stroke/TIA, n = 1290 | ICH, n = 167 | Ischemic Stroke/TIA, n = 553 | ICH, n = 96 | ||
| Age, y | 78.9 (72.5‐84.0) | 78.0 (71.3‐82.7) | 0.339 | 77.6 (71.5‐83.2) | 78.1 (71.9‐82.9) | 0.888 |
| 65–74 | 314 (24.3) | 42 (25.1) | 0.815 | 144 (26.6) | 25 (26.0) | 0.914 |
| ≥75 | 857 (66.4) | 109 (65.3) | 0.775 | 355 (65.5) | 64 (66.7) | 0.824 |
| Female sex | 708 (54.8) | 83 (49.7) | 0.210 | 290 (53.5) | 53 (55.2) | 0.758 |
| Treatment for | ||||||
| HTN | 887 (68.8) | 111 (66.5) | 0.548 | 364 (67.2) | 60 (62.5) | 0.373 |
| DM | 321 (24.9) | 34 (20.4) | 0.202 | 139 (25.6) | 21 (21.9) | 0.432 |
| Hypercholesterolemia | 501 (39.0) | 46 (27.5) | 0.004 | 231 (42.7) | 24 (25.0) | 0.001 |
| Previous MI | 188 (14.6) | 20 (12.0) | 0.369 | 74 (13.7) | 10 (10.4) | 0.387 |
| CAD | 468 (36.3) | 57 (34.1) | 0.582 | 191 (35.2) | 32 (33.3) | 0.718 |
| Other vascular disease | 111 (8.6) | 8 (4.8) | 0.094 | 51 (9.4) | 4 (4.2) | 0.092 |
| CHF | 350 (27.1) | 25 (15.1) | 0.001 | 138 (25.5) | 11 (11.6) | 0.003 |
| Permanent AF | 889 (74.4) | 120 (79.5) | 0.175 | 377 (75.0) | 70 (79.5) | 0.354 |
| Bleeding history | 43 (3.3) | 11 (6.6) | 0.036 | 12 (2.2) | 7 (7.3) | 0.015 |
| Mechanical valve prosthesis | 67 (5.2) | 3 (1.8) | 0.053 | 24 (4.4) | 1 (1.0) | 0.154 |
| Biovalve prosthesis | 26 (2.1) | 5 (3.0) | 0.399 | 13 (2.5) | 2 (2.1) | 1.000 |
| CHA2DS2‐VASc score | 4 (3‐5) | 3 (3‐5) | 0.001 | 4 (3‐5) | 4 (3‐5) | 0.146 |
| ≥2 | 1197 (92.7) | 155 (92.8) | 0.964 | 502 (92.6) | 91 (94.8) | 0.444 |
| HAS‐BLED score | 2 (1‐2) | 2 (2‐3) | 0.812 | 2 (1‐2) | 2 (1‐2) | 0.936 |
| Modified HAS‐BLED ≥3 | 221 (17.1) | 34 (20.5) | 0.283 | 82 (15.1) | 18 (18.8) | 0.368 |
| Time from AF to CE, d | 1118 (356‐2070) | 1173 (468‐1819) | 0.580 | 1107 (415‐2024) | 1174 (527‐1834) | 0.681 |
| eGFR, mL/min/1.73 m2 | 67 (53‐82) | 76 (56‐91) | 0.001 | 68 (55‐82) | 80 (58‐94) | 0.001 |
| eGFR subgroups | 0.014 | 0.066 | ||||
| >60 | 783 (61.8)a | 121 (73.3)b | 350 (65.3)c | 73 (76.8)d | ||
| 30–60 | 441 (34.8) | 39 (23.6) | 171 (31.9) | 19 (20.0) | ||
| <30 | 44 (3.5) | 5 (3.0) | 15 (2.8) | 3 (3.2) | ||
| Hb, g/L | 135 (±19) | 137 (±16) | 0.254 | 135 (±18) | 139 (±16) | 0.264 |
| INR | 2.0 (1.6‐2.4) | 2.7 (2.3‐3.1) | 0.001 | |||
| 2.0–3.0e | 553 (43.4) | 96 (58.2) | 0.001 | — | — | |
| <2.0f | 608 (47.8) | 16 (9.7) | 0.001 | — | — | |
| >3.0g | 109 (8.6) | 53 (32.1) | 0.001 | — | — | |
| Medications | ||||||
| Warfarin | 1284 (99.5) | 167 (100) | 1.000 | — | — | |
| Dabigatran | 6 (0.5) | 0 (0) | 1.000 | — | — | |
| ASA | 122 (9.5) | 23 (13.8) | 0.080 | 55 (10.1) | 12 (12.5) | 0.488 |
| Clopidogrel | 13 (1.0) | 2 (1.2) | 0.687 | 6 (1.1) | 1 (1.0) | 1.000 |
| NSAIDs | 26 (2.0) | 0 (0) | 0.064 | 13 (2.4) | 0 (0) | 0.234 |
| SSRIs | 37 (2.9) | 7 (4.2) | 0.569 | 15 (2.8) | 4 (4.2) | 0.700 |
Abbreviations: AF, atrial fibrillation; ASA, acetylsalicylic acid (aspirin); CAD, coronary artery disease; CE, cerebral event; CHA2DS2‐VASc, CHF, HTN, age > 75 y, DM, stroke/TIA, vascular disease, age 65–74 y, sex category (female); CHF, congestive heart failure; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; HAS‐BLED, HTN, abnormal renal and liver function, stroke, bleeding history or predisposition, labile INR, elderly age (>65 years); Hg, hemoglobin; HTN, hypertension; ICH, intracerebral hemorrhage; INR, international normalized ratio; IQR, interquartile range; MI, myocardial infarction; NSAID, nonsteroidal anti‐inflammatory drug; SD, standard deviation; SSRI, selective serotonin reuptake inhibitor; TIA, transient ischemic attack.
Data are presented as n (%), mean ± SD, or median (IQR).
Data missing for 22 patients.
Data missing for 2 patients.
Data missing for 17 patients.
Data missing for 1 patient.
INR = 2.5–3.5 in patients with mechanical prosthetic valve.
INR <2.5 in patients with mechanical prosthetic valve.
INR >3.5 in patients with mechanical prosthetic valve.
Figure 2.
Admission INR levels in patients with ischemic stroke/TIA and ICH. Abbreviations: ICH, intracerebral hemorrhage; INR, international normalized ratio; TIA, transient ischemic attack
In multivariable analysis, CHF (odds ratio [OR]: 2.30, 95% confidence interval [CI]: 1.39–3.81, P = 0.001), hypercholesterolemia (OR: 1.94, 95% CI: 1.30–2.88, P = 0.001), impaired renal function (OR: 1.86, 95% CI: 1.23–2.80, P = 0.003), and mechanical valve prosthesis (OR: 4.41, 95% CI: 1.32–14.7, P = 0.016) were overrepresented in patients with ischemic stroke/TIA relative to patients with ICH (Table 2). On the contrary, INR (OR: 0.40, 95% CI: 0.33–0.48, P < 0.001) and aspirin use (OR: 0.52, 95% CI: 0.30–0.91, P = 0.022) were associated with a reduced risk of stroke/TIA compared with ICH (Table 2).
Table 2.
Multivariable analysis of the risk factors for ischemic stroke/TIA compared with ICH in anticoagulated patients with AF
| Variable | Whole Cohort, N = 1457 | INR in Therapeutic Range, n = 649 | ||
|---|---|---|---|---|
| OR (95% CI) | P Value | OR (95% CI) | P Value | |
| CHF | 2.30 (1.39–3.81) | 0.001 | 2.33 (1.18–4.58) | 0.014 |
| Treatment for hypercholesterolemia | 1.94 (1.30–2.88) | 0.001 | 2.52 (1.51–4.19) | <0.001 |
| eGFR <60 mL/min/1.73 m2 | 1.86 (1.23–2.80) | 0.003 | 1.56 (0.91–2.66) | 0.106 |
| INR | 0.40 (0.33–0.48) | <0.001 | — | — |
| Mechanical valve prosthesis | 4.41 (1.32–14.7) | 0.016 | — | — |
| Other vascular disease | 1.60 (0.72–3.55) | 0.248 | 1.91 (0.65–5.59) | 0.237 |
| Bleeding history | 0.54 (0.25–1.17) | 0.118 | 0.30 (0.11–0.82) | 0.019 |
| ASA use | 0.52 (0.30–0.91) | 0.022 | — | — |
| NSAIDs | 7.02 (0.43–115.7) | 0.998 | — | — |
Abbreviations: AF, atrial fibrillation; ASA, acetylsalicylic acid (aspirin); CHF, congestive heart failure; CI, confidence interval; eGFR, estimated glomerular filtration rate; ICH, intracerebral hemorrhage; INR, international normalized ratio; NSAID, nonsteroidal anti‐inflammatory drug; OR, odds ratio; TIA, transient ischemic attack.
3.1. Patients with INR within therapeutic range
A total of 553 (42.9%) stroke/TIA patients and 96 (57.4%) ICH patients had INR within the target therapeutic range at the time of admission (P = 0.001; Table 1). In line with the overall study cohort, the stroke/TIA patients presented more frequently with a history of CHF (P = 0.003), hypercholesterolemia (P = 0.001), and lower eGFR (P = 0.001; Table 1). A history of bleeding was more frequent in patients with ICH than in patients with ischemic stroke/TIA (P = 0.015). There were no differences between the patient groups with regard to CHA2DS2‐VASc or modified HAS‐BLED risk scores (Table 1). In multivariable analysis, CHF (OR: 2.33, 95% CI: 1.18–4.58, P = 0.014) and hypercholesterolemia (OR: 2.52, 95% CI: 1.51–4.19, P < 0.001) were more common in ischemic stroke/TIA compared with ICH, whereas a history of bleeding (OR: 0.30, 95% CI: 0.11–0.82, P = 0.019) was more frequent in ICH than in ischemic stroke/TIA (Table 2).
3.2. Mortality
A total of 156 stroke/TIA patients (12.1%) and 63 ICH patients (37.7%) died during the 30‐day follow‐up (P < 0.001). Similarly, when only evaluating patients who developed stroke/TIA or ICH while on therapeutic INR range, mortality was higher among the ICH patients (9.4% vs 35.6%, respectively; P < 0.001).
3.3. Patients with renal impairment
Among patients with normal renal function, ICHs accounted for 13.3% of cerebrovascular events; whereas in patients with renal insufficiency, the proportion of ICHs was lower (8.3%; P = 0.004). This was not related to differences in their INR values (median, 2.0 [1.0] vs 2.1 [1.0], respectively; P = 0.654). In patients with INR within the therapeutic range, the proportion of ICH in patients with normal renal function was also higher than in patients with impaired renal function (17.3% vs 10.6%, respectively; P = 0.027). Impairment of renal function was not associated with mortality: The 30‐day mortality was similar in patients with or without renal impairment (in stroke/TIA patients, 13.3% vs 11.2%, respectively, P = 0.257; and in ICH patients, 34.2% vs. 38.3%, respectively, P = 0.64).
4. DISCUSSION
The present retrospective study is one of the largest to compare anticoagulated AF patients with either ischemic stroke/TIA or ICH. Unlike many other earlier studies, we focused on patients with their first lifetime cerebrovascular event. Although the groups resembled each other in many respects, certain differences were also found. In the whole population, CHF, hypercholesterolemia, mechanical heart valve prosthesis, and renal insufficiency were more common in patients with ischemic stroke/TIA when compared with those with ICH, whereas concomitant aspirin use and high INR were overrepresented in ICH patients. When the analysis was restricted to patients with therapeutic INR, patients with ischemic stroke/TIA more frequently had CHF and hypercholesterolemia and less frequently had history of bleeding when compared with patients with ICH.
The challenge in treating patients with AF using OAC is to find the optimal balance between the risks of thromboembolic complications and bleeding. Even though patients were receiving anticoagulation therapy, most of the cerebrovascular events were thromboembolic/thrombotic and the ratio between stroke/TIA and ICH was high (7.7). It remained high (5.8) also in patients with INR within the target range during the index event. These figures are higher than those reported in recent trials in which thromboembolic and bleeding complications have been assessed in AF patients treated with warfarin. In the Randomized Evaluation of Long‐Term Anticoagulation Therapy (RE‐LY), Rivaroxaban Once Daily Oral Direct Factor Xa Inhibition Compared With Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation (ROCKET‐AF), Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation (ARISTOTLE), and Effective Anticoagulation with Factor Xa Next Generation in Atrial Fibrillation–Thrombolysis in Myocardial Infarction 48 (ENGAGE‐AF TIMI 48) trials, the ratio between the incidence of ischemic stroke and ICH has been in the range of 1.4 to 7.5 for patients treated with warfarin.3, 4, 5, 12, 13 The possible explanation for preponderance of ischemic stroke/TIA in our unselected real‐world study compared with the above trials is that our patients were older (mean age, 78 vs 70–73 years) and more often female (50% vs 40%–45%) and patients with mechanical valve prosthesis were included—all factors increasing the risk of thromboembolic complications. Some of the reference studies included also patients with other ICHs (eg, subdural and subarachnoid hemorrhages), as well as traumatic bleedings, whereas in our study only spontaneous ICHs were included.
In our study, patients with ischemic stroke/TIA had CHF 1.8× to 2.2× more often than did patients with ICH, both in the overall study cohort as well as in the subgroup with therapeutic INR. Heart failure and/or left ventricular dysfunction are established risk factors of thromboembolic complications, both in nonanticoagulated9, 14 and anticoagulated patients15 with chronic AF as well as after cardioversion of recent‐onset AF.16 This supports the view that heart failure and left ventricular dysfunction are prothrombotic. Several factors are involved in this phenomenon, such as decreased flow, endothelial dysfunction, and increased blood level of prothrombotic compounds, all of which predispose to thrombus formation within the cardiac chambers and blood vessels.17
A history of hypercholesterolemia was also more frequently found in patients with stroke/TIA than in patients with ICH, both in the whole study population and in patients with therapeutic INR. Previously, hypercholesterolemia has not been associated with cardiac emboli, though it is known to increase the risk of thrombotic stroke caused by acute rupture of an atherosclerotic plaque in the carotid arteries.18, 19 Consequently, the risk of thrombotic stroke can be reduced with low‐density lipoprotein cholesterol–lowering agents.20 In the light of this finding, it is somewhat surprising that the prevalence of other diseases related to hypercholesterolemia, such as coronary artery disease, myocardial infarction, and other vascular diseases, was similar among the stroke/TIA and ICH groups. McGrath et al. reported that a history of cardiovascular disease (coronary artery bypass grafting, myocardial infarction, percutaneous coronary intervention, peripheral arterial disease) was associated with an increased risk of ischemic stroke, rather than with ICH, in anticoagulated AF patients.21 However, the McGrath study also included patients with a previous stroke, which suggests that their study population included patients with more severe vascular disease. In addition, the impact of hypercholesterolemia was not assessed separately in their study. In many AF patients the etiology of stroke is atherothrombotic rather than embolic. In a previous study, we found that anticoagulated AF patients with coexisting carotid artery stenosis had a 2‐fold higher risk of stroke recurrence and a 4‐fold 30‐day mortality when compared with AF patients without carotid artery stenosis.22 Thus, risk factors for atherosclerosis, also in anticoagulated AF patients, warrant active preventive measures (eg, lipid‐lowering agents for hypercholesterolemia).
High INR increases the risk of ICH 3‐ to 5‐fold.23, 24 On the other hand, it is noteworthy that in our study, the majority (58%) of ICHs occurred in patients within therapeutic INR at the time of the event, with the median INR of 2.6. Previously, Curtze et al. showed that in AF patients on warfarin therapy, approximately 50% of ICHs occur when INR is within the therapeutic range.25 As expected, patients with stroke/TIA had lower INR compared with ICH patients. Nonetheless, more than every second stroke/TIA patient had INR within the therapeutic range during the index event. More plausibly, in some of these cases, the event was not caused by a cardiac embolus but was of atherothrombotic origin. Today many AF patients are treated with novel oral anticoagulants (NOACs). The incidence of ICH during NOAC treatment has been found to be significantly lower than during warfarin treatment, at least in those with moderate or low time in therapeutic range. Thus, our results cannot be applied in patients taking NOACs.
Several studies have shown that bleeding history is a significant predictor of ICH and other major bleedings in patients administered OAC.15, 26, 27 In line with these reports, our ICH patients more often had a history of bleeding than did stroke/TIA patients, and this difference was maintained also when only patients with therapeutic INR were evaluated. Aspirin was used by only 10% of patients in our study; it doubled the number of ICHs in the whole study cohort, but not in patients with INR within the target range. This indirectly suggests that the combination of aspirin and warfarin is harmful in patients with high INR values. Other drugs related to increased risk of bleeding, such as antithrombotic drugs, nonsteroidal anti‐inflammatory drugs, and selective serotonin reuptake inhibitors, were used by only a minority of patients (<5%). Thus, our study was underpowered to address the impact of these medications on the risk of stroke/TIA or bleeding.
Renal failure increases the risk of thromboembolic events due to alterations in plasma coagulation factors and endothelium, and it increases the risk of bleeding mostly due to platelet dysfunction.6, 21, 28, 29 In our whole study population, renal impairment shifted the type of cerebrovascular event in the direction of ischemic stroke/TIA; the OR favoring stroke/TIA was almost doubled in patients with renal impairment compared with patients with normal renal function. This is in line with a Danish cohort study, which reported that despite the bidirectional risk of thromboembolic and bleeding complications in patients with chronic kidney disease, patients at high risk of thromboembolism (CHA2DS2‐VASc score ≥ 2) benefit from anticoagulation therapy.30
4.1. Study limitations
The limitations of the study need to be acknowledged. One obvious limitation is the retrospective design; in its favor, this type of study protocol allows the enrollment of all consecutive patients, therefore reflecting real‐life practice. Treatment decisions were at the treating physician's discretion, and factors not assessed in the case reports may account for these decisions. Nevertheless, the participating hospitals treat all the patients with stroke/TIA and ICH in their catchment area, indicating a good coverage of data. The data were collected from several sources (eg, hospital database, laboratory database, and electronic patient records). The study personnel were given written structured instructions about interpretation of the clinical data of patient records, similar to those of multicenter prospective clinical trials. Another significant limitation is that although our study allowed us to compare patients with ischemic stroke/TIA and ICH, no data were available for AF patients without these events, and therefore the risk factors for these events could not be assessed. In addition, we were able to record INR values from all patients only at the time of admission. It would have been ideal to assess time in therapeutic range from several months preceding the cerebrovascular event. Finally, we did not have data on plasma cholesterol at the time of the event; the diagnosis of hypercholesterolemia was based on lipid‐lowering therapy. Thus, we might have missed some patients with elevated cholesterol but not taking lipid‐lowering agents.
5. CONCLUSION
CHF and hypercholesterolemia were overrepresented in anticoagulated AF patients within target INR range with ischemic stroke/TIA in comparison with patients with ICH, whereas history of bleeding was more common in patients with ICH. In addition, renal insufficiency was more frequent in stroke/TIA patients than in patients with ICH.
Author contributions
Juha Hartikainen, MD, K.E. Juhani Airaksinen, MD, and Pirjo Mustonen, MD, contributed equally to this work.
Conflicts of interest
Dr Lehtola has received research grants from the Finnish Foundation for Cardiovascular Research. Professor Airaksinen has received research grants from the Finnish Foundation for Cardiovascular Research and has given lectures for Bayer, Cardiome, Pfizer, AstraZeneca, and Boehringer Ingelheim. Dr Kiviniemi has received research grants from the Finnish Medical Foundation, BMS‐Pfizer Finland, and the Finnish Foundation for Cardiovascular Research; has given lectures for Bayer, BMS‐Pfizer, Boehringer Ingelheim, and MSD; and has been a member of advisory board for Boehringer Ingelheim and MSD. Dr Palomäki has given lectures for Bayer and MSD. Prof Dr Mustonen has given lectures for Orion, Boehringer Ingelheim, Bayer, Pfizer, Bristol‐Myers Squibb, Sanofi‐Aventis, and Leo Pharma and has been a member of the advisory boards for Boehringer Ingelheim, Bayer, Pfizer, Bristol‐Myers Squibb, and Leo Pharma. The authors declare no other potential conflicts of interest.
ACKNOWLEDGMENTS
The authors thank Tuija Vasankari for the coordination of the data sampling.
Lehtola H, Hartikainen J, Hartikainen P, et al. How do anticoagulated atrial fibrillation patients who suffer ischemic stroke or spontaneous intracerebral hemorrhage differ? Clin Cardiol. 2018;41:608–614. 10.1002/clc.22935
Funding information This study was supported by the Finnish Foundation of Cardiovascular Research and the Finnish Medical Foundation.
REFERENCES
- 1. Hart RG, Benavente O, McBride R, et al. Antithrombotic therapy to prevent stroke in patients with atrial fibrillation: a meta‐analysis. Ann Intern Med. 1999;131:492–501. [DOI] [PubMed] [Google Scholar]
- 2. Connolly SJ, Ezekowitz MD, Yusuf S, et al; RE‐LY Steering Committee and Investigators . Dabigatran versus warfarin in patients with atrial fibrillation [published correction appears in N Engl J Med 2010;363:1877]. N Engl J Med. 2009;361:1139–1151. [DOI] [PubMed] [Google Scholar]
- 3. Patel MR, Mahaffey KW, Garg J, et al; ROCKET‐AF Investigators . Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med. 2011;365:883–891. [DOI] [PubMed] [Google Scholar]
- 4. Giugliano RP, Ruff CT, Braunwald E, et al; ENGAGE AF‐TIMI 48 Investigators . Edoxaban versus warfarin in patients with atrial fibrillation. N Engl J Med. 2013;369:2093–2104. [DOI] [PubMed] [Google Scholar]
- 5. Granger CB, Alexander JH, McMurray JJ, et al; ARISTOTLE Committees and Investigators. Apixaban versus warfarin in patients with atrial fibrillation. N Engl J Med. 2011;365:981–992. [DOI] [PubMed] [Google Scholar]
- 6. Pisters R, Lane DA, Nieuwlaat R, et al. A novel user‐friendly score (HAS‐BLED) to assess 1‐year risk of major bleeding in patients with atrial fibrillation: the Euro Heart Survey. Chest. 2010;138:1093–1100. [DOI] [PubMed] [Google Scholar]
- 7. Lip GY, Frison L, Halperin JL, et al. Comparative validation of a novel risk score for predicting bleeding risk in anticoagulated patients with atrial fibrillation: the HAS‐BLED (Hypertension, Abnormal Renal/Liver Function, Stroke, Bleeding History or Predisposition, Labile INR, Elderly, Drugs/Alcohol Concomitantly) score. J Am Coll Cardiol. 2011;57:173–180. [DOI] [PubMed] [Google Scholar]
- 8. Palomäki A, Mustonen P, Hartikainen JE, et al. Underuse of anticoagulation in stroke patients with atrial fibrillation—the FibStroke Study. Eur J Neurol. 2016;23:133–139. [DOI] [PubMed] [Google Scholar]
- 9. Lip GY, Nieuwlaat R, Pisters R, et al. Refining risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor‐based approach: the Euro Heart Survey on Atrial Fibrillation. Chest. 2010;137:263–272. [DOI] [PubMed] [Google Scholar]
- 10. Kirchhoff P, Benussi S, Kotecha D, et al; ESC Scientific Document Group . 2016. ESC Guidelines for the management of atrial fibrillation developed in collaboration with EACTS. Eur Heart J. 37:2893–2962. [DOI] [PubMed] [Google Scholar]
- 11. Vahanian A, Alfieri O, Andreotti F, et al; Joint Task Force onthe Management of Valvular Heart Disease of the European Society of Cardiology (ESC) . Guidelines on the management of valvular heart disease (version 2012). Eur Heart J. 2012;33:2451–2496. [DOI] [PubMed] [Google Scholar]
- 12. Hart RG, Diener HC, Yang S, et al. Intracranial hemorrhage in atrial fibrillation patients during anticoagulation with warfarin or dabigatran: the RE‐LY trial. Stroke. 2012;43:1511–1517. [DOI] [PubMed] [Google Scholar]
- 13. Hylek EM, Held C, Alexander JH, et al. Major bleeding in patients with atrial fibrillation receiving apixaban or warfarin: the ARISTOTLE Trial (Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation): predictors, characteristics and clinical outcomes. J Am Coll Cardiol. 2014;63:2141–2147. [DOI] [PubMed] [Google Scholar]
- 14. Agarwal M, Apostolakis S, Lane DA, et al. The impact of heart failure and left ventricular dysfunction in predicting stroke, thromboembolism, and mortality in atrial fibrillation patients: a systematic review. Clin Ther. 2014;36:1135–1144. [DOI] [PubMed] [Google Scholar]
- 15. Gallego P, Roldán V, Torregrosa JM, et al. Relation of the HAS‐BLED bleeding risk score to major bleeding, cardiovascular events, and mortality in anticoagulated patients with atrial fibrillation. Circ Arrhythm Electrophysiol. 2012;5:312–318. [DOI] [PubMed] [Google Scholar]
- 16. Airaksinen KE, Grönberg T, Nuotio I, et al. Thromboembolic complications after cardioversion of acute atrial fibrillation: the FinCV (Finnish CardioVersion) study. J Am Coll Cardiol. 2013;62:1187–1192. [DOI] [PubMed] [Google Scholar]
- 17. Shantsila E, Lip GY. Thrombotic complications in heart failure: an underappreciated challenge. Circulation. 2014;130:387–389. [DOI] [PubMed] [Google Scholar]
- 18. Iso H, Jacobs DR Jr, Wentworth D, et al. Serum cholesterol levels and six‐year mortality from stroke in 350 977 men screened for multiple risk factor intervention trial. N Engl J Med. 1989;320:904–910. [DOI] [PubMed] [Google Scholar]
- 19. Mauriello A, Sangiorgi GM, Virmani R, et al. A pathobiologic link between risk factors profile and morphological markers of carotid instability. Atherosclerosis. 2010;208:572–580. [DOI] [PubMed] [Google Scholar]
- 20. Amarenco P, Bougousslavsky J, Callahan A 3rd, et al; Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL) Investigators . High‐dose atorvastatin after stroke or transient ischemic attack. N Engl J Med. 2006;355:549–559. [DOI] [PubMed] [Google Scholar]
- 21. McGrath ER, Kapral MK, Fang J, et al; Investigators of the Registry of the Canadian Stroke Network . Which risk factors are more associated with ischemic stroke than intracerebral hemorrhage in patients with atrial fibrillation? Stroke. 2012;43:2048–2054. [DOI] [PubMed] [Google Scholar]
- 22. Lehtola H, Airaksinen KEJ, Hartikainen P, et al. Stroke recurrence in patients with atrial fibrillation: concomitant carotid artery stenosis doubles the risk. Eur J Neurol. 2017;34:719–725. [DOI] [PubMed] [Google Scholar]
- 23. Fang MC, Chang Y, Hylek EM, et al. Advanced age, anticoagulation intensity, and risk for intracranial hemorrhage among patients taking warfarin for atrial fibrillation. Ann Intern Med. 2004;141:745–752. [DOI] [PubMed] [Google Scholar]
- 24. Hylek EM, Go AS, Chang Y, et al. Effect of intensity of oral anticoagulation on stroke severity and mortality in atrial fibrillation. N Engl J Med. 2003;349:1019–1026. [DOI] [PubMed] [Google Scholar]
- 25. Curtze S, Strbian D, Meretoja A, et al. Higher baseline international normalized ratio value correlates with higher mortality in intracerebral hemorrhage during warfarin use. Eur J Neurol. 2014;21:616–622. [DOI] [PubMed] [Google Scholar]
- 26. Roldán V, Marín F, Manzano‐Fernández S, et al. The HAS‐BLED score has better prediction accuracy for major bleeding than CHADS2 or CHA2DS2‐VASc scores in anticoagulated patients in atrial fibrillation. J Am Coll Cardiol. 2013;62:2199–2204. [DOI] [PubMed] [Google Scholar]
- 27. Shireman TI, Howard PA, Kresowik TF, et al. Combined anticoagulant‐antiplatelet use and major bleeding in elderly atrial fibrillation patients. Stroke. 2004;35:2362–2367. [DOI] [PubMed] [Google Scholar]
- 28. Olesen JB, Lip GY, Kamper AL, et al. Stroke and bleeding in atrial fibrillation with chronic kidney disease [published correction appears in N Engl J Med 2012;367:2262]. N Engl J Med. 2012;367:625–635. [DOI] [PubMed] [Google Scholar]
- 29. Capodanno D, Angiolillo DJ. Antithrombotic therapy in patients with chronic kidney disease. Circulation. 2012;125:2649–2661. [DOI] [PubMed] [Google Scholar]
- 30. Bonde AN, Lip GY, Kamper AL, et al. Net clinical benefit of antithrombotic therapy in patients with atrial fibrillation and chronic kidney disease in nationwide observational cohort study. J Am Coll Cardiol. 2014;64:2471–2482. [DOI] [PubMed] [Google Scholar]