Abstract
Introduction:
Covert brain infarcts (CBI) are frequent incidental findings on MRI and associated with future stroke risk in patients without a history of clinically evident cerebrovascular events. However, the prognostic value of CBI in first-ever ischemic stroke patients is unclear and previous studies did not report on different etiological stroke subtypes. We aimed to test CBI phenotypes and their association with stroke recurrence in first-ever ischemic stroke patients according to stroke etiology.
Patients and methods:
This study is a pooled data analysis of two prospectively collected cohorts of consecutive first-ever ischemic stroke patients admitted to the comprehensive stroke centers of Bern (Switzerland) and Graz (Austria). CBI phenotypes were identified on brain MRI within 72 h after admission. All patients underwent a routine follow-up (median: 12 months) to identify stroke recurrence.
Results:
Of 1577 consecutive ischemic stroke patients (median age: 71 years), 691 patients showed CBI on brain MRI (44%) and 88 patients had a recurrent ischemic stroke (6%). Baseline CBI were associated with stroke recurrence in multivariable analysis (HR 1.9, 95% CI 1.1–3.3). CBI phenotypes with the highest risk for stroke recurrence were cavitatory CBI in small vessel disease (SVD)-related stroke (HR 7.1, 95% CI 1.6–12.6) and cortical CBI in patients with atrial fibrillation (HR 3.0, 95% CI 1.1–8.1).
Discussion and conclusion:
This study reports a ≈ 2-fold increased risk for stroke recurrence in first-ever ischemic stroke patients with CBI. The risk of recurrent stroke was highest in patients with cavitatory CBI in SVD-related stroke and cortical CBI in patients with atrial fibrillation.
Subject terms: Covert brain infarcts, stroke
Keywords: Covert brain infarction, stroke recurrence, stroke etiology, infarct phenotypes
Graphical abstract.
Introduction
Covert brain infarcts (CBI) are a frequent incidental finding on brain MRI and are defined as focal cerebral lesions of presumed ischemic origin in patients without a fitting history of stroke or transient ischemic attack (TIA).1,2 CBI are highly prevalent in the general elderly population (20%–35%) and are associated with neurocognitive decline, gait imbalances and depression.3,4 Moreover, an increased risk for future clinically evident ischemic stroke was identified in patients with CBI on brain imaging for routine non-stroke indications (e.g. headache, vertigo). 1 However, studies on the value of CBI detected on neuroimaging work-up in first-ever ischemic stroke patients are scarce and showed conflicting results.5–8 While a substudy of the large PRoFESS trial could not identify an effect of CBI on the risk for stroke recurrence in first-ever noncardioembolic ischemic stroke patients,5,9 other studies in the past decade reported a moderate association of (multiple) CBI on recurrent stroke risk.6–8 Those studies were limited by low event rates, observing <20 recurrent ischemic strokes in their CBI subgroups.6–8 In contrast, a recently published large Korean population-based study, including more than 500 CBI patients, reported a nearly 3-fold increased risk of recurrent ischemic stroke. 10 However, this study only included first-ever ischemic stroke patients with an underlying atrial fibrillation and could therefore not report on the prognostic value of CBI in other etiological subtypes. 10 Furthermore, it is yet unknown whether CBI phenotypes or location affect the time interval to a recurrent stroke.
We hypothesized that CBI would be associated with stroke recurrence and that this would differ according to CBI phenotype and stroke etiology. We therefore aimed (1) to evaluate the association of CBI including its phenotypes for stroke recurrence in first-ever ischemic stroke patients with (2) an analysis according to etiological subtypes and (3) to report on the time intervals to a recurrent stroke according to different CBI phenotypes in a European study population.
Materials and methods
Data pooling was approved by the ethics committee of the Medical University of Graz (EK 29–285 ex 16/17; informed consent was obtained by all included subjects) and Bern (ID 2020-01696, requirement for informed consent was waived according to Swiss law) and performed according to the standards of Declaration of Helsinki. 11 Reporting was performed according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement. Data from this study are available from the corresponding author upon reasonable request.
Selection of study participants and data collection
This study is based on two cohorts from prospectively collected stroke registries from the university stroke centers Bern (Switzerland) and Graz (Austria). All consecutive patients with first-ever manifest ischemic stroke, who were admitted at the stroke units of both centers between January 2015 and December 2017 (Bern) and April 2017 and August 2019 (Graz) were included. To ensure that the selected patients truly had their first-ever stroke, medical history was reviewed electronically via the hospital information system and clinically including the Verifying Stroke Free Status Questionnaire. 12 Ischemic stroke was confirmed by brain MRI in all patients. Both registries include data on demographics, vascular risk factors, stroke severity assessed by the National Institutes of Health Stroke Scale (NIHSS) upon admission and stroke etiology.
In general, all patients underwent a thorough etiological stroke work-up at the stroke unit including medical history, brain imaging via MRI, laboratory tests, ECG at admission, continuous ECG monitoring at the stroke unit, MRI/computed tomography (CT)-based angiography or sonography of the extra- and intracranial vessels, and echocardiography in the first days after stroke.
After diagnostic work-up, patients’ stroke etiology was classified into one of the following: (A) large artery atherosclerosis (LAA)-related stroke defined by a symptomatic extra- or intracranial stenosis of brain supplying vessels of >50%, (B) cerebral small vessel disease (SVD; acute small subcortical infarct <20mm in diameter and moderate/severe white matter hyperintensities according to age-related white matter changes rating scale scores ⩾2), 13 (C) AF-related stroke or (D) other specific stroke etiologies (e.g. cardiac high-risk source other than AF, arterial dissection, cancer-associated hypercoagulopathy). 14 If there were several possible stroke etiologies the most probable cause of stroke was documented at the discretion of the treating physician. If no distinct etiology was found, the ischemic stroke was classified as cryptogenic.
Brain imaging
All patients underwent brain MRI within the first 72 h after stroke (1.5 or 3 -Tesla MRI, Magnetom Scanner, Siemens Healthineers, Munich, Germany).
Imaging protocol routinely included standard axial T1- and T2-weighted pulse sequences (slice thickness = 3–5 mm), Fluid Attenuated Inversion Recovery (FLAIR) and axial diffusion and susceptibility weighted sequences.
CBI phenotypes and locations were identified in accordance with recent studies in this field,10,11 utilizing a classification proposed by Vynckier et al. 13 : CBI were classified as cavitatory (lesions ⩾3mm in diameter with central cavity and adjacent gliosis), combined gray and white matter lesions, isolated gray matter lesions and large subcortical (noncavitatory) infarcts (Supplemental Figure 1). 13 The cavitatory phenotype corresponds to the STRIVE definition of lacune of presumed vascular origin, but also includes cerebellar lesions. 15 For this purpose, cerebellar defects ⩽15 mm in diameter were rated as cavitatory CBI whereas larger defects were classified as combined gray and white matter lesions. For infratentorial CBI rating T2-weighted sequences were used.
All images were rated by two in neuroimaging experienced vascular neurologists (JV and TRM; Bern cohort) and by two trained neuroradiologists (SL and MJ; Graz cohort). In case of uncertainties, images were discussed with a third rater with ample experience in neuroimaging (MK; Graz cohort). All raters were blinded to clinical data.
Follow-up
All patients were followed-up for the occurrence of recurrent ischemic stroke for a minimum period of 3 months after the index stroke incorporating clinical and imaging information. Follow-up consisted of a questionnaire and a clinical neurological follow-up (Bern cohort). For the Graz cohort, all patients underwent at least one clinical follow-up visit (mostly at 3 months post stroke) and were additionally followed via the medical and nursing documentation and communication network of Styria (MEDOCS). MEDOCS includes medical information collected in all public hospitals in the federal state of Styria, which cover all neurological treatment facilities. 16 To minimize loss to follow-up, further data were retrieved from the Austrian electronic health record (ELGA) database, an electronic information system documenting medical records from public and non-public hospitals and healthcare facilities in Austria.
Statistics
Nominal data were calculated in count and percentage, while median and range were used to present ordinal and continuous variables. Comparison was performed between patients with recurrent ischemic stroke and patients with no stroke recurrence. Pearson’s chi-square test was utilized for comparison of nominal variables.
Distribution testing was done in two steps: First, quantitative continuous variables were tested for normal distribution using histograms, the Q-Q plot and Kolmogorov-Smirnov test. Afterward, in case of normal distribution, comparison of variables was performed with parametric analysis using the unpaired Student’s test. In case of non-normal distribution continuous variables were tested with the nonparametric Mann-Whitney-U test.
To account for the wide range of follow-up investigations, a multivariable Cox regression model was calculated for stroke recurrence during follow-up as the target variable and adjusted for age, sex, mortality and variables that were associated with recurrent ischemic stroke in univariable analysis (p < 0.1; i.e. arterial hypertension, admission NIHSS and any CBI). To test for the association of CBI locations and phenotypes on stroke recurrence, the model was also calculated including different CBI phenotypes instead of any CBI. In a second step, the model was run targeting recurrent stroke in different etiological subtypes (LAA, SVD, AF-related stroke and cryptogenic stroke) to identify the value of CBI in patients with different stroke mechanisms. Of note, there was no missing data in the variables included in the models. To report on CBI-based interrater agreement, Cohen’s kappa statistics were used.
A p-value of <0.05 was set to identify significant differences without adjustments for multiple testing. Assuming a stroke recurrence risk of 5% in patients without CBI and 10% in patients with CBI, we estimated a sample size of 1242 patients. To account for missing data, we aimed to include 1500 patients.
We used Cohen’s kappa statistics to report interrater agreement.
Statistical analyses were conducted using the software IBM SPSS Statistics, version 28 (IBM Corp, Armonk, NY, USA).
Results
A total of 1577 patients with first-ever clinically evident ischemic stroke were included (details are shown on Figure 1). In the final study cohort (median age 71 years; interquartile range [IQR]: 59–83 years; female: 49%), arterial hypertension was the most prevalent vascular risk factor (68%). Most frequent stroke etiologies were atrial fibrillation (21%) followed by LAA-related stroke (20%). In more than one third of all patients (39%) stroke etiology could not be determined after the initial work-up (cryptogenic).
Figure 1.
Flow diagram of selection of study participants.
Of all included patients, 691 (44%) had at least one CBI on brain MRI. Compared to patients with no CBI, they were older (median age: 75 vs 67 years, p < 0.001), and more often had vascular risk factors (arterial hypertension: 77% vs 63%, p < 0.001; diabetes 23% vs 14%, p < 0.001). CBI were found more often in LAA- and SVD-related stroke (29% vs 15%, p < 0.001 and 14% vs 10%, p < 0.005 respectively) (Table 1).
Table 1.
Demographics, clinical parameters and follow-up data of first-ever ischemic stroke patients with/without chronic covert brain infarcts on brain MRI within 72 h after admission.
| Total cohort (n = 1577) | ⩾1 CBI n = 691 | No CBI n = 886 | p-Value | |
|---|---|---|---|---|
| Demographics | ||||
| Age, years (median, IQR) | 71 (59–80) | 75 (63–84) | 67 (55–76) | <0.001 |
| Female (n, %) | 770 (48.8) | 317 (49.4) | 453 (48.4) | 0.717 |
| Vascular risk factors at index event (n, %) | ||||
| Hypertension | 1069 (67.8) | 489 (77.4) | 580 (62.7) | <0.001 |
| Dyslipidemia | 878 (56.4) | 375 (59.3) | 503 (54.4) | 0.053 |
| Diabetes mellitus | 275 (17.4) | 148 (23.4) | 127 (13.7) | <0.001 |
| Smoking | 407 (26.6) | 154 (24.9) | 253 (27.7) | 0.218 |
| Peripheral artery disease | 49 (3.1) | 20 (4.4) | 29 (3.8) | 0.580 |
| NIHSS score at presentation (median, IQR) | 4 (2–8) | 4 (2–8) | 4 (2–8) | 0.316 |
| Stroke etiology | ||||
| Small vessel disease | 183 (11.6) | 92 (14.3) | 91 (9.7) | 0.005 |
| Large artery atherosclerosis | 322 (20.4) | 185 (28.8) | 137 (14.7) | <0.001 |
| Atrial fibrillation | 334 (21.2) | 133 (20.7) | 201 (21.5) | 0.709 |
| Cryptogenic | 616 (39.1) | 240 (37.4) | 376 (40.2) | 0.258 |
| Other | 122 (7.7) | 41 (5.9) | 81 (9.1) | 0.106 |
| Moderate/severe WMH (ARWMC ⩾ 2) | 382 (24.2) | 231 (36.0) | 151 (16.1) | <0.001 |
| Secondary stroke prevention* | ||||
| Antiplatelets | 430 (74.6) | 189 (74.3) | 241 (75.0) | 0.882 |
| Anticoagulation | 143 (24.8) | 65 (24.9) | 78 (24.7) | 0.910 |
| Follow-up | ||||
| Time of follow-up (median, IQR) | 12 (12–29) | 12 (13–29) | 11 (11–27) | 0.850 |
| Recurrent ischemic stroke (n, %) | 88 (5.6) | 49 (7.6) | 39 (4.2) | 0.003 |
| Time to recurrent ischemic stroke (in days; median, IQR) | 85 (29–194) | 80 (22–171) | 98 (41–239) | 0.080 |
| Death during follow-up (n, %) | 339 (21.5) | 167 (26.0) | 172 (18.4) | <0.001 |
MRI: magnetic resonance imaging; CBI: covert brain infarcts; IQR: interquartile range; NIHSS: National Institutes of Health Stroke Scale, WMH: white matter hyperintensities, ARWMC: age-related white matter changes.
Available in 576 patients.
Cavitatory lesions were the most frequent CBI phenotype (n = 420, 62%) and 212 patients (33%) had >1 (=multiple) CBI (see Table 2 for details on location and phenotypes). The majority of patients with multiple CBI had >1 CBI of the same phenotype (59%, mostly cavitatory).
Table 2.
Demographics, clinical parameters and chronic covert brain infarct phenotypes in patients with/without ischemic stroke during the follow-up period.
| Ischemic stroke during follow-up n = 88 | No ischemic stroke during follow-up n = 1489 | p-Value | |
|---|---|---|---|
| Demographics | |||
| Age, years (median, IQR) | 74 (62–83) | 71 (59–80) | 0.029 |
| Female (n, %) | 45 (51.1) | 725 (48.7) | 0.656 |
| Vascular risk factors at index event (n, %) | |||
| Hypertension | 66 (75.9) | 1003 (68.2) | 0.083 |
| Dyslipidemia | 50 (57.5) | 830 (56.5) | 0.854 |
| Diabetes mellitus | 19 (21.8) | 256 (17.4) | 0.181 |
| Smoking | 17 (20.0) | 381 (26.3) | 0.196 |
| Peripheral artery disease | 1 (2.5) | 48 (4.1) | 0.619 |
| NIHSS score at presentation (median, IQR) | 5 (3–9) | 4 (2–8) | 0.011 |
| Chronic covert brain infarcts (CBI; n, %) | |||
| Any | 54 (61.4) | 637 (42.8) | 0.003 |
| Cavitatory | 35 (39.8) | 385 (25.9) | 0.004 |
| Large non-cavitatory subcortical | 3 (3.4) | 65 (4.4) | 0.668 |
| Isolated cortical | 5 (5.7) | 55 (3.7) | 0.344 |
| Combined grey and white matter | 18 (20.5) | 255 (17.1) | 0.251 |
| More than one CBI | 23 (26.1) | 189 (12.7) | <0.001 |
| Moderate/severe WMH (ARWMC ⩾ 2) | 21 (23.9) | 361 (24.2) | 0.935 |
| CBI location (n, %) | |||
| Subcortical | 21 (23.9) | 224 (15.0) | 0.026 |
| Cortical | 21 (23.9) | 281 (18.9) | 0.155 |
| Cerebellar | 24 (27.3) | 260 (17.5) | 0.020 |
| Brainstem | 1 (1.1) | 25 (1.7) | 0.698 |
| Basal ganglia | 13 (14.8) | 107 (7.2) | 0.009 |
| Stroke etiology | |||
| Small vessel disease | 13 (14.8) | 170 (11.4) | 0.340 |
| Large artery atherosclerosis | 19 (21.6) | 303 (20.3) | 0.779 |
| Atrial fibrillation | 18 (20.5) | 316 (21.2) | 0.864 |
| Cryptogenic | 32 (36.4) | 584 (39.2) | 0.593 |
| Other | 11 (12.5) | 111 (7.5) | 0.176 |
| Secondary stroke prevention* | |||
| Antiplatelets | 25 (75.8) | 405 (74.6) | 0.730 |
| Anticoagulation | 8 (24.2) | 135 (24.9) | 0.788 |
IQR: interquartile range, NIHSS: National Institutes of Health Stroke Scale, WMH: white matter hyperintensities, ARWMC: age-related white matter changes.
Available in 576 patients.
The interrater agreement for the presence of any CBI was 86% (κ = 0.72, p < 0.001, Bern cohort) and 89% (κ = 0.77, p < 0.001, Graz cohort).
CBI and stroke recurrence
All included patients were followed for a median duration of 12 months (range, 3–36 months). During the follow-up period, 88 patients had a recurrent ischemic stroke (6%). Apart from age (median 74 vs 71 years, p = 0.029) and stroke severity at admission (median NIHSS 5 vs 4, p = 0.011), CBI were more prevalent in patients with a recurrent ischemic stroke (61% vs 43%, p = 0.003) in univariable analysis (Table 2). After adjusting for age, sex, arterial hypertension and admission NIHSS, CBI remained independently associated with future recurrent stroke (Hazard ratio [HR] 1.9, 95% confidence interval [CI] 1.1–3.3, p = 0.016). The point estimate for the association with stroke recurrence was highest for the cavitatory phenotype (HR 2.3, 95% CI 1.3–3.9, p = 0.003) and cerebellar location (HR 1.9, 95% CI 1.1–3.4, p = 0.030) as well as for multiple CBI (HR 2.4, 95% CI 1.3–4.0, p = 0.001) (Table 3; Figure 2).
Table 3.
Multivariable Cox regression analysis-based hazard ratios for a recurrent ischemic stroke during the follow-up period according to CBI phenotypes and location at baseline.
| Adjusted hazard ratio (95% confidence interval) | p-Value | |
|---|---|---|
| Age (per year) | 1.0 (1.0–1.1) | 0.652 |
| Female | 1.6 (0.9–2.8) | 0.120 |
| Arterial hypertension | 0.7 (0.4–1.5) | 0.375 |
| NIHSS at presentation (per point) | 1.0 (0.9–1.0) | 0.353 |
| Mortality | 1.5 (0.7–2.9) | 0.255 |
| Any CBI | 1.9 (1.1–3.3) | 0.016 |
| Cavitatory* | 2.3 (1.3–3.9) | 0.003 |
| Combined grey and white matter* | 1.4 (0.7–2.7) | 0.308 |
| More than one CBI* | 2.4 (1.3–4.0) | 0.001 |
| Subcortical* | 1.5 (0.8–2.6) | 0.094 |
| Cortical* | 1.4 (0.7–2.6) | 0.308 |
| Cerebellar* | 1.9 (1.1–3.4) | 0.030 |
CBI: chronic covert brain infarcts; NIHSS: National Institutes of Health Stroke Scale.
Cox regression model was recalculated with each CBI phenotype instead of any CBI.
Figure 2.
Cumulative stroke recurrence rates in first-ever ischemic stroke patients according to (a) CBI versus no CBI at baseline, (b) different CBI phenotypes and (c) presents the corresponding percentages at 90 days and 1 year post-stroke.
Stroke etiology
While chronic cavitatory infarcts were the most prevalent CBI phenotype in LAA- and SVD-associated first-ever ischemic stroke patients (41% and 39% respectively), cortical involvement was the most frequently observed CBI location in AF-related stroke patients (26%).
Stroke recurrence rates were comparable in patients with different stroke etiologies ranging from 5% to 7%. Of note, the strength of the association between baseline CBI and a recurrent ischemic stroke differed according to stroke etiology:
In SVD-related first-ever stroke patients, CBI (adjusted HR 7.4, 95% CI 1.5–17.6, p = 0.021) were a strong risk factor for stroke recurrence while an association in LAA patients was only observed for a cerebellar CBI phenotype (HR 2.7, 95% CI 1.0–7.3, p = 0.045).
Patients with AF-associated stroke had a higher risk for a recurrent brain infarct if a cortical CBI phenotype (HR 3.0, 95% CI 1.1–8.1, p = 0.029) or multiple CBI (HR 4.1, 95% CI 1.5–11.6, p = 0.007) were present. Multiple CBI were also the most important risk factor for a recurrent stroke in cryptogenic stroke patients (HR 3.2, 1.4–7.5, p = 0.006) (Table 4 and Supplemental Table 1).
Table 4.
Multivariable Cox regression analysis-based hazard ratios for a recurrent ischemic stroke during the follow-up period according to baseline CBI in different etiological stroke subtypes.
| n, % | Adjusted hazard ratio (95% confidence interval)* | p-Value | |
|---|---|---|---|
| Small vessel disease (n = 183) | |||
| Any CBI | 92 (50.5) | 7.4 (1.5–17.6) | 0.021 |
| Large artery atherosclerosis (n = 322) | |||
| Any CBI | 185 (57.5) | 1.2 (0.4–3.2) | 0.589 |
| AF-related stroke (n = 334) | |||
| Any CBI | 133 (39.4) | 2.5 (0.9–6.4) | 0.043 |
| Cryptogenic stroke (n = 616) | |||
| Any CBI | 240 (39.0) | 1.7 (0.8–3.7) | 0.101 |
CBI: chronic covert brain infarcts; NIHSS: National Institutes of Health Stroke Scale.
Adjusted for age, sex, admission NIHSS, hypertension and mortality.
Time interval to stroke recurrence
Median duration until stroke recurrence was 85 days after the index event (IQR: 29–194 days). Recurrent ischemic strokes tended to occur earlier in patients with CBI compared to those with no chronic covert brain lesions (median time to recurrent stroke: 80 vs 98 days, p = 0.080). Cavitatory lesions were the only CBI phenotype that showed a higher risk for earlier stroke recurrence (median time to recurrent stroke: 68 vs 104 days, p = 0.030) (Figure 2(b) and (c)).
Discussion
In this pooled analysis of two large stroke cohorts, we identified CBI in first-ever stroke patients as independent risk factor for (early) recurrent ischemic stroke during a median follow-up period of 1 year. Cavitatory and multiple CBI were the strongest risk factors for future stroke. Importantly, the strength of this association differed by stroke etiology.
The largest existing studies on the value of CBI as a risk factor for stroke recurrence in first-ever ischemic stroke patients only reported on AF-related stroke.10,17 While the results of the EAFT study group published in 1996 must be interpreted cautiously as they used meanwhile outdated neuroimaging techniques, 17 a recent Korean population-based study identified CBI as risk factor for stroke recurrence in AF-related first-ever ischemic stroke patients. 10 We also identified a ≈ 2-fold increased future stroke risk in our AF-related stroke subgroup, which was mainly attributed to multiple CBI and a cortical CBI phenotype. Of interest, cerebellar CBI – which were also included in the “embolic-appearing” CBI subgroup by Kim et al. and were often suspected to be related to a (cardio)embolic origin10,18 – were not associated with a recurrent stroke in the AF-subgroup of our European-population based study sample. Therefore, the nearly 2-fold increased future stroke risk in patients with cerebellar CBI in the total cohort mainly derived from non-cardioembolic stroke subtypes. Our results suggest different underlying mechanisms in cerebellar CBI and encourage future research activities on chronic cerebellar lesion phenotypes and locations (i.e. deep vs cortical cerebellar CBI) to improve the pathophysiological understanding, which could increase their value for stroke risk models and secondary stroke prevention strategies.15,18–20
Previous studies on the association of CBI with stroke recurrence in non-AF related first-ever ischemic stroke patients showed conflicting results. A PRoFESS substudy found no association of CBI with recurrent stroke as compared to matched patients without CBI at baseline during a mean follow-up of 2.5 years (odds ratio 1.4; 95% CI 0.8–2.6), 5 while a Danish registry-based study reported a nearly 3-fold increased risk for future stroke in first-ever stroke patients with multiple CBI. 7 In young stroke patients with longer follow-up duration, multiple CBI (HR 2.5, 95% CI 1.2–5.0), but not single CBI (HR 1.5, 95% CI 0.7–3.2) were associated with recurrent ischemic stroke. 6 This is in contrast to another young ischemic stroke cohort, where a 3-fold increased risk of recurrent ischemic stroke was seen in patients with CBI over a follow-up duration of 25 months. 8 Of note, those studies were limited by a small sample size6–8 and low CBI rates (≈20%),5–7 which was largely different to most recent studies reporting on CBI prevalence in first-ever ischemic stroke patients (37%–46%) and might be attributed to their young patient cohorts.5–8,10,21 In contrast, we present data of an elderly stroke population (CBI rate: 44%) and were able to report on the effect of CBI for future stroke risk in different etiological stroke subtypes. Although we could not identify an increased risk for stroke recurrence in patients with CBI and LAA-related stroke – possibly influenced by surgical/interventional treatment effects – CBI seem to be strongly associated with recurrent ischemic stroke in patients with cerebral SVD, which was mainly attributed to cavitatory subcortical lesions and multiple CBI. Moreover, cavitatory CBI were associated with reduced time intervals to stroke recurrence. Specific CBI pattern in SVD-related stroke could therefore point toward aggressive cerebral SVD and identify first-ever ischemic stroke patients who are at particular high risk for (early) stroke recurrence. Such patients might benefit from more intense treatment regimens which could be considered in future trials testing novel (pharmacological) secondary stroke prevention strategies for lacunar stroke. 22
Our study included a large subgroup of ischemic stroke patients whose etiology remained cryptogenic after the initial work-up at the stroke unit. As recently presented randomized trials (ATTICUS, ARCADIA) failed to identify subsets of cryptogenic stroke patients that benefit from intensified medical treatment (i.e. oral anticoagulation),23,24 the paradigm in cryptogenic stroke research might evolve from a main focus on stroke etiology to identify those patients at highest risk for stroke recurrence. CBI might contribute to improve both of these approaches. Results from our study support the value of CBI to identify patients at higher risk for stroke recurrence and we identified typical CBI phenotypes in different stroke etiologies such as chronic cortical lesions in AF-related stroke. This could possibly point toward a distinct etiology in patients who remained cryptogenic even after the extensive initial work-up and should be targeted in future studies.
This study has some limitations:
First, we cannot rule out a selection bias, as a substantial number of patients with a missing follow-up had to be excluded, along with patients who only had undergone CT-based imaging. While baseline characteristics did not differ between patients with and without follow-up, except for sex (Supplemental Table 2), our findings may not be generalizable to individuals unable to undergo MRI due to more severe stroke, vomiting, agitation, pacemakers, or frailty.
Second, we did not adjust for multiple testing due to the limited number of recurrent ischemic stroke in the CBI phenotype subgroups in patients with different stroke etiologies. These subgroup findings should therefore be regarded as hypothesis-generating and need replication in a larger study with a higher number of outcome events. Third, although using our recently published classification of CBI phenotyping, 13 certain CBI phenotypes are still incompletely understood (e.g. chronic infratentorial lesions), which could have affected our results and should be a target for future studies in this field. Fourth, since our analysis relies on registry-based data, we lacked information on pre-stroke anticoagulation. Consequently, we were unable to report on the predictive value of CBI in patients with “breakthrough” ischemic strokes under anticoagulation therapy. Finally, data on treatment adherence, control of vascular risk factors during the follow-up period, and information on surgical/interventional secondary stroke prevention strategies were unavailable. These variables might have influenced the risk of stroke recurrence and could potentially elucidate the missing link between CBI and recurrent stroke risk in LAA-related stroke, which could be based on more effective treatment strategies in such patients (i.e. carotid endarterectomy/stenting).
In conclusion, our study reports a ≈ 2-fold increased risk for stroke recurrence in first-ever ischemic stroke patients with CBI and identifies multiple CBI as the most important risk factor besides stroke severity. Of note, CBI-based recurrent stroke risk was affected by stroke etiology (i.e. increased risk of cavitatory lesions in SVD-related stroke and of cortical lesions in stroke caused by AF). Present findings could be considered for etiology-based stroke risk models, which could identify high-risk patient subgroups who would most likely benefit from specific secondary stroke prevention and treatment regimens.
Supplemental Material
Supplemental material, sj-docx-1-eso-10.1177_23969873241229612 for Association of covert brain infarct phenotype with stroke recurrence in first-ever manifest ischemic stroke according to etiology by Thomas Raphael Meinel, Stefan L. Leber, Michael Janisch, Jan Vynckier, Adnan Mujanovic, Anna Boronylo, Johannes Kaesmacher, David Julian Seiffge, Laurent Roten, Marcel Arnold, Christian Enzinger, Thomas Gattringer, Urs Fischer and Markus Kneihsl in European Stroke Journal
Supplemental material, sj-tif-2-eso-10.1177_23969873241229612 for Association of covert brain infarct phenotype with stroke recurrence in first-ever manifest ischemic stroke according to etiology by Thomas Raphael Meinel, Stefan L. Leber, Michael Janisch, Jan Vynckier, Adnan Mujanovic, Anna Boronylo, Johannes Kaesmacher, David Julian Seiffge, Laurent Roten, Marcel Arnold, Christian Enzinger, Thomas Gattringer, Urs Fischer and Markus Kneihsl in European Stroke Journal
Acknowledgments
None.
Footnotes
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Th. Meinel reports research support from the Bangerter Rhyner Foundation, University of Bern, Swiss National Foundation, and the Swiss Heart Foundation; S.L. Leber reports no disclosures relevant to the manuscript; M. Janisch reports no disclosures relevant to the manuscript; J. Vynckier reports no disclosures relevant to the manuscript; A. Mujanovic reports no disclosures relevant to the manuscript; A. Boronylo reports no disclosures relevant to the manuscript; J. Kaesmacher reports no disclosures relevant to the manuscript; D.J. Seiffge reports no disclosures relevant to the manuscript; L. Roten reports speaker honoraria from Abbott/SJM, consulting honoraria from Medtronic and research grant to the institution from Medtronic for an investigator-initiated trial; M. Arnold reports speaker honoraria (Astra Zeneca, Bayer, Covidien, Medtronic) and advisory board involvements (Amgen, Bayer, BMS, Boehringer Ingelheim, Daiichy Sankyo, Medtronic, Novo Nordisk, Novartis, Pfizer); C. Enzinger reports no disclosures relevant to the manuscript; Th. Gattringer reports no disclosures relevant to the manuscript; U. Fischer reports research support of the Swiss National Science Foundation and the Swiss Heart Foundation; PI of the ELAN trial, Co-PI of the DISTAL, TECNO, SWIFT DIRECT and SWITCH trial; research grants from Medtronic (BEYOND SWIFT, SWIFT DIRECT) and from Stryker, Rapid medical, Penumbra and Phenox (DISTAL); consultancies for Medtronic, Stryker, and CSL Behring (fees paid to institution); participation in an advisory board for Alexion/Portola, Boehringer Ingelheim, Biogen and Acthera (fees paid to institution); member of a clinical event committee (CEC) of the COATING study (Phenox) and member of the data and safety monitoring committee (DSMB) of the TITAN, LATE_MT and IN EXTREMIS trials; vice-presidency of the Swiss Neurological Society; M. Kneihsl reports no disclosures relevant to the manuscript.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
Informed consent and ethical approval: Data pooling was approved by the ethics committee of the Medical University of Graz (EK 29-285 ex 16/17; informed consent was obtained by all included subjects) and Bern (ID 2020-01696, requirement for informed consent was waived according to Swiss law).
Guarantor: MK.
Contributorship: TM: data acquisition and interpretation, manuscript preparation. SLL: acquisition of data. MJ: acquisition of data. JV: acquisition of data. AM: critical revision of the manuscript content. AB: acquisition of data. JK: critical revision of the manuscript content. DJS: critical revision of the manuscript content. LR: critical revision of the manuscript content. MA: critical revision of the manuscript content. CE: critical revision of the manuscript content. TG: critical revision of the manuscript content. UF: critical revision of the manuscript content. MK: data acquisition and interpretation, manuscript preparation.
All authors have read and approved the final manuscript and agreed to be accountable for all aspects of the work.
ORCID iDs: Thomas Raphael Meinel 
https://orcid.org/0000-0002-0647-9273
Adnan Mujanovic
https://orcid.org/0000-0002-6839-7134
David Julian Seiffge
https://orcid.org/0000-0003-3890-3849
Markus Kneihsl
https://orcid.org/0000-0002-6334-9432
Supplemental material: Supplemental material for this article is available online.
References
- 1. Gupta A, Giambrone AE, Gialdini G, et al. Silent brain infarction and risk of future stroke: a systematic review and meta-analysis. Stroke 2016; 47: 719–725. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Smith EE, Saposnik G, Biessels GJ, et al. Prevention of stroke in patients with silent cerebrovascular disease: a scientific statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2017; 48: e44–e71. [DOI] [PubMed] [Google Scholar]
- 3. Meinel TR, Kaesmacher J, Roten L, et al. Covert brain infarction: towards precision medicine in research, diagnosis, and therapy for a silent pandemic. Stroke 2020; 51: 2597–2606. [DOI] [PubMed] [Google Scholar]
- 4. Kent DM, Leung LY, Zhou Y, et al. Association of silent cerebrovascular disease identified using natural language processing and future ischemic stroke. Neurology 2021; 97: e1313–e1321. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Weber R, Weimar C, Wanke I, et al. Risk of recurrent stroke in patients with silent brain infarction in the prevention regimen for effectively avoiding second strokes (PRoFESS) imaging substudy. Stroke 2012; 43: 350–355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Putaala J, Haapaniemi E, Kurkinen M, et al. Silent brain infarcts, leukoaraiosis, and long-term prognosis in young ischemic stroke patients. Neurology 2011; 76: 1742–1749. [DOI] [PubMed] [Google Scholar]
- 7. Andersen SD, Skjøth F, Yavarian Y, et al. Multiple silent lacunes are associated with recurrent ischemic stroke. Cerebrovasc Dis 2016; 42: 73–80. [DOI] [PubMed] [Google Scholar]
- 8. Gioia LC, Tollard Dubuc V, et al. Silent ischemic lesions in young adults with first stroke are associated with recurrent stroke. Neurology 2012; 79: 1208–1214. [DOI] [PubMed] [Google Scholar]
- 9. Sacco RL, Diener HC, Yusuf S, et al. Aspirin and extended-release dipyridamole versus clopidogrel for recurrent stroke. New Engl J Med 2008; 359: 1238–1251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Kim DY, Han SG, Jeong HG, et al. Covert brain infarction as a risk factor for stroke recurrence in patients with atrial fibrillation. Stroke 2023; 54: 87–95. [DOI] [PubMed] [Google Scholar]
- 11. Epstein A, Schilter M, Vynckier J, et al. Chronic covert brain infarctions and white matter hyperintensities in patients with stroke, transient ischemic attack, and stroke mimic. J Am Heart Assoc 2022; 11: 1–2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Jones WJ, Williams LS, Meschia JF. Validating the uestionnaire for verifying stroke-free status (QVSFS) by neurological history and examination. Stroke 2001; 32: 2232–2236. [DOI] [PubMed] [Google Scholar]
- 13. Vynckier J, Kaesmacher J, Wardlaw JM, et al. Phenotypes of chronic covert brain infarction in patients with first-ever ischemic stroke: a cohort study. Stroke 2022; 53: 558–568. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Kneihsl M, Gattringer T, Bisping E, et al. Blood biomarkers of heart failure and hypercoagulation to identify atrial fibrillation-related stroke. Stroke 2019; 50: 2223–2226. [DOI] [PubMed] [Google Scholar]
- 15. Duering M, Biessels GJ, Brodtmann A, et al. Neuroimaging standards for research into small vessel disease—advances since 2013. Lancet Neurol 2023; 22: 602–618. [DOI] [PubMed] [Google Scholar]
- 16. Gell G, Madjaric M, Leodolter W, et al. HIS purchase projects in public hospitals of Styria, Austria. Int J Med Inform 2000; 58-59: 147–155. [DOI] [PubMed] [Google Scholar]
- 17. Van Latum JC, Van Kooten F, Koudstaal PJ, et al. Silent brain infarction in nonrheumatic atrial fibrillation. EAFT Study Group. European atrial fibrillation trial. Neurology 1996; 46: 159–165. [DOI] [PubMed] [Google Scholar]
- 18. Favilla CG, Ingala E, Jara J, et al. Predictors of finding occult atrial fibrillation after cryptogenic stroke. Stroke 2015; 46: 1210–1215. [DOI] [PubMed] [Google Scholar]
- 19. De Cocker LJ, Kloppenborg RP, van der Graaf Y, et al. Cerebellar cortical infarct cavities: correlation with risk factors and MRI markers of cerebrovascular disease. Stroke 2015; 46: 3154–3160. [DOI] [PubMed] [Google Scholar]
- 20. Park KY, Chung PW, Kim YB, et al. Association between small deep cerebellar ischemic lesion and small-vessel disease. Cerebrovasc Dis 2009; 28: 314–320. [DOI] [PubMed] [Google Scholar]
- 21. Zhang L, Wang S, Qiu L, et al. Effects of silent brain infarction on the hemorrhagic transformation and prognosis in patients with acute ischemic stroke after intravenous thrombolysis. Front Neurol 2023; 14: 4–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Wardlaw JM, Woodhouse LJ, Mhlanga, et al. Isosorbide mononitrate and cilostazol treatment in patients with symptomatic cerebral small vessel disease: the Lacunar Intervention Trial-2 (LACI-2) randomized clinical trial. JAMA Neurol 2023; 80: 682–692. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Geisler T, Poli S, Meisner C, et al. Apixaban for treatment of embolic stroke of undetermined source (ATTICUS randomized trial): rationale and study design. Int J Stroke 2017; 12: 985–990. [DOI] [PubMed] [Google Scholar]
- 24. Kamel H, Longstreth WT, Jr, Tirschwell DL. The AtRial cardiopathy and antithrombotic drugs in prevention after cryptogenic stroke randomized trial: rationale and methods. Int J Stroke 2019; 14: 207–214. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental material, sj-docx-1-eso-10.1177_23969873241229612 for Association of covert brain infarct phenotype with stroke recurrence in first-ever manifest ischemic stroke according to etiology by Thomas Raphael Meinel, Stefan L. Leber, Michael Janisch, Jan Vynckier, Adnan Mujanovic, Anna Boronylo, Johannes Kaesmacher, David Julian Seiffge, Laurent Roten, Marcel Arnold, Christian Enzinger, Thomas Gattringer, Urs Fischer and Markus Kneihsl in European Stroke Journal
Supplemental material, sj-tif-2-eso-10.1177_23969873241229612 for Association of covert brain infarct phenotype with stroke recurrence in first-ever manifest ischemic stroke according to etiology by Thomas Raphael Meinel, Stefan L. Leber, Michael Janisch, Jan Vynckier, Adnan Mujanovic, Anna Boronylo, Johannes Kaesmacher, David Julian Seiffge, Laurent Roten, Marcel Arnold, Christian Enzinger, Thomas Gattringer, Urs Fischer and Markus Kneihsl in European Stroke Journal