Secondary Logo

Journal Logo

Cerebral Microbleeds and the Safety of Anticoagulation in Ischemic Stroke Patients

A Systematic Review and Meta-Analysis

Liang, Yifan, MD*; Song, Quhong, MD; Jiao, Yang, MD; Lin, Jueying, MD§; Qu, Huiling, MD, PhD; Zhao, Shanshan, MD, PhD*; Xu, Junjie, MD, PhD; Zhao, Chuansheng, MD, PhD*; Zhao, Mei, MD, PhD#

doi: 10.1097/WNF.0000000000000306
Review Articles

Objectives The objective of this study was to investigate the safety of anticoagulation in ischemic stroke (IS) patients with cerebral microbleeds (CMBs).

Methods PubMed, Web of Science, Elsevier Clinical Key, Google Scholar, and Cochrane Library from 1996 to July 2018 were searched to identify relevant studies that included IS patients, underwent T2*-weighted gradient recalled echo, or susceptibility-weighted imaging for detection CMBs and used anticoagulants during follow-up. Primary outcome of interest was intracerebral hemorrhage (ICH). Secondary outcomes were hemorrhage transformation, IS, total mortality, and new developed CMBs. We critically appraised studies and conducted a systematic review and meta-analysis following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidance.

Results We included 7 observational studies. Cerebral microbleeds were associated with a significantly elevated risk of anticoagulation-related ICH (odds ratio, 4.01; 95% confidence interval, 1.82–8.81; P = 0.001). It was significant for warfarin (odds ratio, 8.02; 95% confidence interval, 1.51–42.62; P = 0.015). New developed CMBs in patients on warfarin treatment were associated with baseline CMBs, and the appearance of hemorrhagic transformation did not have a significant relationship with baseline CMBs.

Conclusions The presence of CMBs increases the risk of ICH during anticoagulant treatment (especially warfarin) in IS patients. Further studies with larger numbers of patients are needed to confirm our conclusions.

*Department of Neurology, The First Hospital of China Medical University, Shenyang;

Department of Neurology, West China Hospital of Sichuan University, Chengdu;

Department of Neurology, The First Affiliated Hospital of Dalian Medical University, Dalian;

§Emergency Department, Zhongshan Hospital Xiamen University, Xiamen;

Department of Neurology, The People's Hospital of Liaoning Province;

Department of Laboratory Medicine, The First Hospital of China Medical University; and

#Department of Cardiology, The Shengjing Affiliated Hospital, China Medical University, Shenyang, China.

Address correspondence and reprint requests to Mei Zhao, MD, PhD, Department of Cardiology, The Shengjing Affiliated Hospital, China Medical University, Shenyang 110004, Liaoning, China;; or Chuansheng Zhao, MD, PhD, Department of Neurology, The First Hospital of China Medical University, Shenyang 110001, Liaoning, China; E-mail:

Yifan Liang and Quhong Song contributed to this work equally.

Conflicts of Interest and Source of Funding: The authors have no conflicts of interest to declare.

This work was supported by National Natural Science Foundation of China (no. 81372104, C.Z.), the Shenyang Population and Health Technical Critical Special Project (no. F16-206-9-01, C.Z.), and the Program of the Distinguished Professor of Liaoning Province, Neurology (C.Z.).

Supplemental digital contents are available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (

Cerebral microbleeds (CMBs) refer to small, round, or ovoid hypointense lesions, less than 10 mm in diameter, detected on T2*-weighted gradient recalled echo (T2*-GRE) or susceptibility-weighted imaging (SWI) magnetic resonance imaging (MRI) sequences.1 They often occur in the setting of the damage of small vessel integrity.2 As a marker of underlying hemorrhage-prone microangiopathies, the presence and burden of CMBs appear to predict the future risk of spontaneous intracerebral hemorrhage (ICH), and it has been claimed that patients with more than 5 CMBs may exhibit a higher risk of mortality.3 Because CMBs are also related to the occlusive features of small vessel cerebrovascular disease (eg, lacunar infarcts), their presence might also represent an increased risk of future ischemic stroke (IS) in some populations.4 Several studies have demonstrated that individuals with CMBs are at a greater risk of suffering incident stroke.5,6 A meta-analysis of 15 studies, focusing on patients with IS or transient ischemic attack, revealed that CMBs are associated with increased stroke risk, and with increasing CMBs burden, the risk of ICH increases more steeply than that of IS.7

Anticoagulation is an effective means to prevent recurrent IS due to cardiogenic sources. However, for IS patients with atrial fibrillation, it is often the fear of ICH rather than the recognition of the benefits of anticoagulation that governs clinical decision-making for anticoagulant treatment in this condition.7–9 Long-term anticoagulation increases the risk of ICH, and the presence of CMBs may further increase the risk of warfarin-related ICH.10

The presence of CMBs combined with the need for anticoagulation to prevent recurrent IS creates a combination of ischemic and hemorrhagic cerebrovascular processes that can be challenging for clinicians. Although there has been a meta-analysis to justify treatment decisions on anticoagulation based on the presence of CMBs, it includes too few subjects and is limited to stroke patients with atrial fibrillation.3 Given the new evidence in The Clinical Relevance of Microbleeds in Stroke study (CROMIS-2),11 it is necessary to conduct a new meta-analysis. In view of these considerations, we conducted a systematic review of published literature and sought to investigate the relationship between CMBs and the safety of anticoagulation in IS patients, by concentrating on the existence of CMBs at baseline.

Back to Top | Article Outline


Study Eligibility Criteria

Studies were considered eligible for inclusion if they fulfilled the following selection criteria: (1) included patients with acute IS, using anticoagulants during follow-up; (2) performed paramagnetic-sensitive MRI sequences (T2*-GRE or SWI) to detect CMBs and accessed baseline CMBs and association with the predefined outcome events; and (3) primary outcome of interest was ICH, and secondary outcomes were hemorrhage transformation (HT), IS, total mortality, and new developed CMBs. We included prospective and retrospective studies and the search was conducted on July 2016, repeated on July 2018, and included studies since 1996. Only English language studies were included in this review because translational services were not available.

Back to Top | Article Outline

Search Methods

PubMed, Web of Science, Elsevier Clinical Key, Google Scholar, and Cochrane Library were searched in the time from 1996 (the year CMBs first reported) to July 2018. The searches mainly focused on English-language sources and studies in human subjects. Search items included Cerebral Microbleed*, CMB, cerebral microh?emorr*, stroke*, Cerebrovascular Accident*, CVA*, Apoplexy, Brain Vascular Accident*, Cerebrovascular Disorder*, infarction*, cerebral, brain, intracerebral, intracranial, infarct*, ischemic, thrombo, emboli, haemorrhage, hemorrhage, and bleed. The complete search algorithm that was used in PubMed and Web of Science search was available in Supplemental Digital Content 1, available at Reference lists of all included articles and relevant review articles were examined to identify studies that may have been missed in the initial database search.

Back to Top | Article Outline

Study Selection

All retrieved studies were scanned independently by 2 reviewers (Y.J. and J.L.). In case of disagreement regarding the literature search results, disagreements were resolved via consensus after a discussion. If no consensus could be agreed, we consulted a third colleague. We also screened the references of recent meta-analysis and potentially eligible studies.

Back to Top | Article Outline

Data Collection

A data collection form was developed and used by both authors for data extraction from the studies. This was based on previously published data extraction methods by the Agency for Healthcare and Research Quality.12 It was used by both authors for data extraction.

Back to Top | Article Outline

Data Items

The main data we extracted included the author, publication time, location and timing of study, type of study, outcome events of interest, method of detecting CMBs, definition of CMBs size, use of anticoagulants, follow-up time, population characteristics, and main results. Separate data extraction was done for each outcome. When several articles had originated from the same group, these were carefully reviewed to ensure that the study populations did not overlap. When it was not possible to ensure that this was the case, then data from the largest available study were used.

Back to Top | Article Outline

Risk of Bias

All included studies are observational studies. Risk of bias was assessed for each predefined outcome in each study using the Newcastle Ottawa Scale (NOS). Data relevant to study quality and design were extracted independently by both authors and consensus reached regarding final assessment. Publication bias was not able to be formally assessed by funnel plot owing to the small number of studies and heterogeneity of outcomes.

Back to Top | Article Outline

Data Synthesis

The review adopted and was written according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines.13 A narrative synthesis of the data was used in this review with a discussion of limitations due to bias and due to the small number of studies. When the data were available, a meta-analysis for the corresponding outcome would be conducted. We first conducted a meta-analysis for all included studies with a random-effects model and then conducted a subgroup analysis based on the type of anticoagulants. All analyses were performed with STATA 13.1 (StataCorp LP, College Station, Tex), using a package of “metan.”

Back to Top | Article Outline


Study Selection

A total of 2816 potential records were identified from the data base during the primary research. Title and abstract were screened against inclusion and exclusion criteria after which 32 articles were identified for full text review (Fig. 1). Thirty-two articles underwent full text review and 25 were excluded, 4 of which were abstract and review; 8 were excluded for not being the predefined population, 2 for not describing outcome separately according to the baseline CMBs status, 9 for not referencing an outcome of interest, and 1 for being a protocol that were in progress; and 1 study was excluded for other reasons. This left 7 original studies that were included in the review, 6 of which considered ICH, 1 considered new developed CMBs, and 1 considered HT.



Back to Top | Article Outline

Study Characteristics

Seven studies11,14–19 met our predefined inclusion criteria and were included in this review, 4 of which recruited patients from Asian populations (China, Japan),14,16–18 211,15 from European, and the other19 from North American populations. Six studies used T2*-GRE MRI to detect CMBs, whereas 1 study had used GRE or SWI MRI to identify CMBs,19 although magnet strengths, echo times, and slice thicknesses varied. The definition of CMBs size also varied. The mean follow-up time in the 7 studies was different, 311,14,15 of which had a mean follow-up of 1 to 2 years, 316,18,19 had a mean follow-up more than 2 but less than 3 years, and 117 investigated the hemorrhagic transformation of the patient treated with anticoagulants in 36 hours after stroke onset. Two15,16 studies used warfarin, 311,14,17 used vitamin K antagonist or non–vitamin K oral anticoagulants (NOACs) during follow-up, and the anticoagulants used in the other 2 studies18,19 were not mentioned. Time to start anticoagulation treatment was also not mentioned in four studies.15–18 The characteristics and methodology of the studies included in this review are summarized in Table 1.



Back to Top | Article Outline

Quality Appraisal

All included studies were critically appraised against a checklist of key quality indicators of NOS. All studies had a quality score of greater than or equal to 6/9. Because of the small number of identified studies, all those that met inclusion were considered in the narrative review below, with their weaknesses addressed within the review. The generally low-quality highlights the need for further well-designed studies in this area to address these weaknesses. The sum of positive adjudications estimated by NOS is given in Table 1 (for details, see Supplemental Digital Content 2, available at

Back to Top | Article Outline

ICH and CMBs

Six identified studies11,14–16,18,19 had ICH as an outcome. The indication for anticoagulants during follow-up was mainly cardiogenic reasons in 5 studies,11,14–16,19 whereas the reason why anticoagulation therapy used in the other one study18 was not mentioned in detail. Five11,14–16,18 of these 6 studies used T2*-GRE, the field strength of which were all 1.5 T or 3.0 T, to detect CMBs, and the other study19 used both T2*-GRE and SWI while the field strength varied among patients. The outcome of ICH was confirmed on MRI or computed tomography (CT) scans.

The association between ICH and CMBs suggested by each study was different (Table 2). The incidence rate of ICH in the cohort of Charidimou et al14 was too low (occurred in only 1 of 22 patients with CMBs and 2 of 80 patients without CMBs). In the study of Fan et al,18 1 of 4 patients with CMBs and 1 of 3 patients without CMBs developed ICH during follow-up. Similarly, the small number of patients and the low incidence rate of ICH may affect the robustness of the results, and it was impossible to separate the confounders using logistic regression analysis. Haji et al19 found that 1 of 22 patients with CMBs had an ICH compared with none out of 66 patients without CMBs (P = 0.2741) during anticoagulation, and the presence of CMBs did not predict future ICH. Orken et al15 followed a cohort of IS patients who had been receiving warfarin therapy for 2 years and found that 1 of 35 patients with CMBs and 1 of 169 patients without CMBs developed asymptomatic ICH and the baseline CMBs predicted asymptomatic ICH in patients on warfarin treatment. Imaizumi et al16 found that warfarin did not significantly increase the incidence of deep ICH in patients with 1 or more deep CMBs (P = 0.19) whereas elevated the occurrence of deep ICH in patients with 3 or more deep CMBs (P = 0.05). Given the small sample size of the previous studies and the low incidence rate of ICH, the results were underpowered to precisely answer the question of whether the patients with CMBs should use anticoagulants. A large prospective study by Wilson et al11 found that 7 of 300 patients with CMBs and 7 of 1136 patients without CMBs developed symptomatic ICH.



For better understanding the association between ICH and CMBs, we undertook a meta-analysis based on the available data in the 6 included cohorts. The results were shown in Figure 2. In all patients with anticoagulants therapy, the total ICH incidence rate was 27/1942 (1.4%). In those patients with CMBs, 15 (3.6%) of 418 experienced an ICH during the follow-up, compared with 12 (0.8%) of 1524 patients without CMBs; thus, CMBs confer an absolute risk increase of 2.8% for ICH. The pooled odds ratio (OR) of ICH was 4.01 (95% confidence interval [CI], 1.82–8.81; P = 0.001] for those patients with CMBs in comparison with those without CMBs (Fig. 2). The results were consistent from study to study (test for heterogeneity, P = 0.775). Warfarin was the only anticoagulant being administered in 2 studies15,16 during the follow-up. We explored the influence of warfarin, by stratifying the meta-analysis according to the type of anticoagulants. Because there were incomplete details of which anticoagulant was actually used in 4 studies, we were compelled to pool the data only from those patients who were administered warfarin. In the comparison between patients with CMBs versus patients without CMBs, the pooled OR of ICH during follow-up was 8.02 (95% CI, 1.51–42.67; P = 0.015, Fig. 3) if the patients had been treated only with warfarin (n = 309, CMBs prevalence, 22.6%), and the absolute risk increase was 6.3% for ICH for CMBs versus no CMBs.





Back to Top | Article Outline

New Developed CMBs and CMBs

One cohort15 reported the association between new developed CMBs between baseline CMBs in the setting of anticoagulation. Orken et al15 assessed the link between new developed CMBs and baseline CMBs in Turkey using the similar method to its ICH comparison. It found that the presence of CMBs at baseline increased the incidence rate of new developed CMBs in patients on warfarin therapy (26% in CMBs positive patients vs 12% in CMB negative patients, P = 0.03) and did not find any association with the duration of warfarin therapy. Their results confirmed that the administration of warfarin may be harmful to patients with CMBs, especially when the patient was at an older age and had leukoaraiosis or lobar CMBs related with cerebral amyloid angiopathy (CAA). The small sample size was a limitation that we were unable to generalize the results to all warfarin users.

Back to Top | Article Outline

HT and CMBs

The relationship between HT and baseline CMBs in IS patients treated with anticoagulants was examined in just 1 study.16,17 Takahashi et al17 consecutively selected acute IS patients (98 atherothrombotic and 89 cardiogenic stroke patients) who were conducted CT during day 2 to check the presence or absence of HT, and they found that the appearance of HT did not have a significant relationship with the presence of CMBs in the subgroup analysis of patients treated with anticoagulants (9% in CMBs positive patients vs 21% in CMB negative patients, P = 0.161). Hemorrhage transformation mainly happened within 4 weeks after index IS, especially in the first week, but the follow-up CT (mean 36 hours after onset) of this study was conducted in the early stage, so the finding may not reflect actual incidence rate of HT that occurred on day 3 or later of an IS.

Back to Top | Article Outline


Summary of Evidence

This systematic review concentrated on the effect of anticoagulation for bleeding complications in patients with CMBs. For patients without baseline CMBs, new developed CMBs are more frequent in patients with prior ICH (OR, 9.7; 95% CI, 1.06–90.9) and in those pretreated antiplatelets (OR, 1.66; 95% CI, 1.09–2.53),20 but not in those pretreated NOACs.21 However, for patients with baseline CMBs, we found evidence that there is a significant association between the baseline CMBs and the risk of ICH in those who receive anticoagulants during the follow-up period. The pooled estimates suggested that among all included subjects, the risk of subsequent spontaneous ICH in CMBs positive patients is higher than those with negative CMBs.

Our systematic review found that, in patients on anticoagulants, presence of CMBs increased the risk of future ICH (OR, 4.01; 95% CI, 1.82–8.81; P = 0.001), especially those on warfarin (OR, 8.02; 95% CI, 1.51–42.67; P = 0.015), which led to an absolute risk increase of 6.3% for ICH between CMBs positive group and CMBs negative group. Recent studies also indicated that the risk of ICH occurs less frequently in patients taking NOACs than in those treated with warfarin.22 When the risk of bleeding and stroke are both high, NOACs may have a greater clinical benefit than warfarin.23 At present, the direct evidence to clarify the benefits of different anticoagulants on patients with CMBs is still rare. A number of large observational studies exploring the association between CMBs and oral anticoagulants are underway: CMB-NOW (Protocol for Cerebral Microbleeds during the Non–Vitamin K Antagonist Oral Anticoagulants or Warfarin Therapy in Stroke Patients with Non-valvular Atrial Fibrillation study),24 IPAAC (Risk of Intracerebral Hemorrhage in Patients Taking Oral Anticoagulants for Atrial Fibrillation With Cerebral Microbleeds [IPAAC] — Warfarin study in Hong Kong; the parallel Novel Oral Anticoagulants [IPAAC-NOAC] study in Hong Kong assesses CMB-related future ICH in patients with stroke taking novel oral anticoagulants), and so on.25

At present, some studies have shown that CMBs develop dynamically over time not only in a considerable proportion of patients but also in healthy elderly individuals.6,26 Most new CMBs developed in those patients with baseline CMBs. Thus, it is possible that in some conditions (eg, with the use of antithrombotic drugs), CMBs may transform into a symptomatic ICH.18 This hypothesis has been supported by some studies that have detected incident symptomatic ICH at the site of a previous CMBs.18,27 An alternative hypothesis would be that CMBs may simply reflect the increased fragility and vulnerability to bleeding of small blood vessels. Despite of the small number of studies, our review suggests that baseline CMBs increase the incidence rate of new developed CMBs in patients treated with warfarin. It is conceivable that the new developed CMBs during the follow-up may increase the risk of IS or ICH. However, because of insufficient data, our study is not capable of evaluating the impact of the number of new CMBs on bleeding complications.

Hemorrhage transformation is a commonly seen complication in acute phases of IS, which refers to concomitant ICH after acute IS.28 Several previous studies have demonstrated the relationship between CMBs and HT in patients who received thrombolysis, some of which found that baseline CMBs could be a risk factor for symptomatic,25 whereas some others indicated that the presence, number, and location of CMBs were not related to the incidence of HT.29–31 What is more, Wang et al31 found that CMBs seem to reduce the risk of HT (OR, <1). In this review, the results suggest that the presence of CMBs might not be related to the development of HT in patients who received anticoagulants. This might be owing to the reason that the pathophysiological mechanism of HT formation after acute IS is different from that of ICH.

There is evidence indicating that the distribution of CMBs in the brain reflects the underlying small vessel disease: strictly, lobar CMBs reflect CAA, whereas deep CMBs represent hypertensive arteriopathy.32,33 Lobar CMBs may be a stronger risk factor for ICH than deep CMBs, but definitive data for this proposal are lacking. There is some recent indirect evidence implicating CAA as a risk for anticoagulation-related ICH,34,35 and thus, the presence of lobar CMBs should deserve particular attention as a means of refining decisions about anticoagulation treatment. Because of the lack of published data about the effect of CMBs distribution on stroke risk, we are not able to investigate this aspect in the current study.

The presence, location, and burden (ie, the number) of CMBs may influence the decision-making on whether to initiate anticoagulation therapy. In view of the insufficient data referring to the burden of CMBs, we are not able to investigate the influence of CMBs on the severity of a future stroke. However, for IS/transient ischemic attack patients with antiplatelet treatment, whereas the association between CMBs burden and risk of IS was owing mainly to nondisabling events (Ptrend = 0.007), the association with ICH was accounted for (Ptrend < 0.0001) by disabling/fatal events (≥5 CMBs: 82% disabling/fatal ICH vs 40% disabling/fatal IS; P = 0.035).36 However, for different anticoagulant-related ICH, the ICH volume, hematoma expansion, 90-day mortality, and functional outcome were similar.37 Therefore, it is necessary to pay attention to the influence of the anticoagulation therapy on the prognosis of IS patients with CMBs. In addition, more studies evaluating the influence of the number of CMBs on the risk of future ICH and IS are needed to determine whether there is a threshold number of CMBs that might tip the risk-benefit balance.

Back to Top | Article Outline


Our systematic review has some limitations. Only 7 studies were included in this review. Some studies had a small sample size, variable follow-up periods, and few outcome events. The kinds of anticoagulants used during follow-up period are varied, and most of included studies used warfarin; however, some used mixed anticoagulants or subcutaneous injection of heparin or intravenous argatroban. The heterogeneity of anticoagulation drugs leads to the inability to evaluate the effect of specific anticoagulant. At the same time, the time of starting application of anticoagulant drugs and the length of application are also key factors affecting the incidence rate of hemorrhagic complications. However, in our review, most information about the application of anticoagulant are not detailed, which also limits the power of our results. There were also differences in the imaging parameters (eg, magnetic field strength, echo time), potentially affecting the detection of CMBs.34 Furthermore, these studies are prone to a selection bias because not all stroke patients undergo T2*-GRE and most investigators have not applied standardized rating scales for CMBs. In addition, some important baseline characteristics (eg, National Institutes of Health Stroke Scale at admission and discharge, the severity of stroke screening, onset-to-treatment time, usage of antiplatelet agents before the application of an anticoagulation drug, and the degree of international normalized ratio) between CMBs positive and CMBs negative group are lacking, which may be potential limitations that can influence the risk of outcomes. Finally, possible confounding is a potentially important limitation because of our inability to adjust for other baseline variables related to the future stroke risk (eg, age, sex, hypertension, white matter change, social background, and standard of medical care in the countries where these studies were completed).

Back to Top | Article Outline


Our findings suggest that the presence of CMBs is a risk factor that can augment the risk of anticoagulant-related (especially warfarin) ICH in IS patients. These observations imply that caution is necessary when contemplating anticoagulation in cardioembolic stroke patients with CMBs. Furthermore, the presence of CMBs may be included in the individual risk stratification system predicting the risk of ICH during anticoagulation therapy. Further studies with larger numbers of patients are needed to confirm our conclusions.

Back to Top | Article Outline


1. Yates PA, Villemagne VL, Ellis KA, et al. Cerebral microbleeds: a review of clinical, genetic, and neuroimaging associations. Front Neurol 2014;4:205.
2. Shoamanesh A, Kwok CS, Benavente O. Cerebral microbleeds: histopathological correlation of neuroimaging. Cerebrovasc Dis 2011;32(6):528–534.
3. Charidimou A, Karayiannis C, Song TJ, et al. Brain microbleeds, anticoagulation, and hemorrhage risk: meta-analysis in stroke patients with AF. Neurology 2017;89(23):2317–2326.
4. Gregoire SM, Brown MM, Kallis C, et al. MRI detection of new microbleeds in patients with ischemic stroke: five-year cohort follow-up study. Stroke 2010;41(1):184–186.
5. Akoudad S, Portegies ML, Koudstaal PJ, et al. Cerebral microbleeds are associated with an increased risk of stroke: the Rotterdam study. Circulation 2015;132(6):509–516.
6. Bokura H, Saika R, Yamaguchi T, et al. Microbleeds are associated with subsequent hemorrhagic and ischemic stroke in healthy elderly individuals. Stroke 2011;42(7):1867–71.
7. Wilson D, Charidimou A, Ambler G, et al. Recurrent stroke risk and cerebral microbleed burden in ischemic stroke and TIA: a meta-analysis. Neurology 2016;87(14):1501–1510.
8. Fisher M. MRI screening for chronic anticoagulation in atrial fibrillation. Front Neurol 2013;4:137.
9. Gross CP, Vogel EW, Dhond AJ, et al. Factors influencing physicians' reported use of anticoagulation therapy in nonvalvular atrial fibrillation: a cross-sectional survey. Clin Ther 2003;25(6):1750–1764.
10. Lee SH, Ryu WS, Roh JK. Cerebral microbleeds are a risk factor for warfarin-related intracerebral hemorrhage. Neurology 2009;72(2):171–176.
11. Wilson D, Ambler G, Shakeshaft C. Cerebral microbleeds and intracranial haemorrhage risk in patients anticoagulated for atrial fibrillation after acute ischaemic stroke or transient ischaemic attack (CROMIS-2): a multicentre observational cohort study. Lancet Neurol 2018;17(6):539–547.
12. Seida JC, Dryden DM, Hartling L. Observational studies: Empirical evidence of their contributions to comparative effectiveness reviews [Internet]. Rockville, MD: Agency for Healthcare Research and Quality (US); 2013. Available at: Accessed October 23, 2018.
13. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med 2009;151(4):W65–W94.
14. Charidimou A, Inamura S, Nomura T, et al. Cerebral microbleeds and white matter hyperintensities in cardioembolic stroke patients due to atrial fibrillation: single-centre longitudinal study. J Neurol Sci 2016;369:263–267.
15. Orken DN, Uysal E, Timer E, et al. New cerebral microbleeds in ischemic stroke patients on warfarin treatment: two-year follow-up. Clin Neurol Neurosurg 2013;115(9):1682–1685.
16. Imaizumi T, Inamura S, Kohama I, et al. Antithrombotic drug uses and deep intracerebral hemorrhages in stroke patients with deep cerebral microbleeds. J Stroke Cerebrovasc Dis 2013;22(6):869–875.
17. Takahashi W, Moriya Y, Mizuma A, et al. Cerebral microbleeds on T2*-weighted images and hemorrhagic transformation after antithrombotic therapies for ischemic stroke. J Stroke Cerebrovasc Dis 2013;22(8):e528–e532.
18. Fan YH, Zhang L, Lam WW, et al. Cerebral microbleeds as a risk factor for subsequent intracerebral hemorrhages among patients with acute ischemic stroke. Stroke 2003;34(10):2459–2462.
19. Haji S, Planchard R, Zubair A, et al. The clinical relevance of cerebral microbleeds in patients with cerebral ischemia and atrial fibrillation. J Neurol 2016;263(2):238–244.
20. Martí-fàbregas J, Medrano-martorell S, Merino E, et al. Statins do not increase markers of cerebral angiopathies in patients with cardioembolic stroke. Sci Rep 2018;8:1492.
21. Soo Y, Abrigo J, Leung KT, et al. Correlation of non-vitamin K antagonist oral anticoagulant exposure and cerebral microbleeds in Chinese patients with atrial fibrillation. J Neurol Neurosurg Psychiatry 2018;89(7):680–686. jnnp-2017-317151.
22. Rasmussen LH, Larsen TB, Graungaard T, et al. Primary and secondary prevention with new oral anticoagulant drugs for stroke prevention in atrial fibrillation: indirect comparison analysis. BMJ 2012;345:e7097.
23. Banerjee A, Lane DA, Torp-Pedersen C, et al. Net clinical benefit of new oral anticoagulants (dabigatran, rivaroxaban, apixaban) versus no treatment in a 'real world' atrial fibrillation population: a modelling analysis based on a nationwide cohort study. Thromb Haemost 2012;107(3):584–589.
24. Takizawa S, Tanaka F, Nishiyama K, et al. Protocol for Cerebral Microbleeds during the Non-Vitamin K Antagonist Oral Anticoagulants or Warfarin Therapy in Stroke Patients with Nonvalvular Atrial Fibrillation (CMB-NOW) Study: multisite pilot trial. J Stroke Cerebrovasc Dis 2015;24(9):2143–2148.
25. Microbleeds International Collaborative Network. Worldwide collaboration in the Microbleeds International Collaborative Network. Lancet Neurol 2016;15(11):1113–1114.
26. Goos JD, Henneman WJ, Sluimer JD, et al. Incidence of cerebral microbleeds: a longitudinal study in a memory clinic population. Neurology 2010;74(24):1954–1960.
27. Huang Y, Cheng Y, Wu J, et al. Cilostazol as an alternative to aspirin after ischaemic stroke: a randomised, double-blind, pilot study. Lancet Neurol 2008;7(6):494–499.
28. Knight RA, Barker PB, Fagan SC, et al. Prediction of impending hemorrhagic transformation in ischemic stroke using magnetic resonance imaging in rats. Stroke 1998;29(1):144–151.
29. Lee SH, Kang BS, Kim N, et al. Does microbleed predict haemorrhagic transformation after acute atherothrombotic or cardioembolic stroke? J Neurol Neurosurg Psychiatry 2008;79(8):913–916.
30. Derex L, Nighoghossian N, Hermier M, et al. Thrombolysis for ischemic stroke in patients with old microbleeds on pretreatment MRI. Cerebrovasc Dis 2004;17(2–3):238–241.
31. Wang BG, Yang N, Lin M, et al. Analysis of risk factors of hemorrhagic transformation after acute ischemic stroke: cerebral microbleeds do not correlate with hemorrhagic transformation. Cell Biochem Biophys 2014;70(1):135–142.
32. Greenberg SM, Vernooij MW, Cordonnier C, et al. Cerebral microbleeds: a guide to detection and interpretation. Lancet Neurol 2009;8(2):165–174.
33. Charidimou A, Werring DJ. Cerebral microbleeds: detection, mechanisms and clinical challenges. Future Neurol 2011;6(5):587–611.
34. Rosand J, Hylek EM, O'Donnell HC, et al. Warfarin-associated hemorrhage and cerebral amyloid angiopathy: a genetic and pathologic study. Neurology 2000;55(7):947–951.
35. Charidimou A, Gang Q, Werring DJ. Sporadic cerebral amyloid angiopathy revisited: recent insights into pathophysiology and clinical spectrum. J Neurol Neurosurg Psychiatry 2012;83(2):124–137.
36. Lau KK, Lovelock CE, Li L, et al. Antiplatelet treatment after transient ischemic attack and ischemic stroke in patients with cerebral microbleeds in 2 large cohorts and an updated systematic review. Stroke 2018;49(6):1434–1442.
37. Wilson D, Seiffge DJ, Traenka C, et al. Outcome of intracerebral hemorrhage associated with different oral anticoagulants. Neurology 2017;88(18):1693.

cerebral microbleeds; anticoagulation; intracerebral hemorrhage; systematic review

Supplemental Digital Content

Back to Top | Article Outline
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.