Correlation between antiplatelet therapy in secondary prevention of acute cerebral infarction and cerebral microbleeds: A susceptibility-weighted imaging (SWI) study

Abstract

OBJECTIVE:

To investigate the clinical significance of antiplatelet aggregation therapy for patients diagnosed with acute cerebral infarction (ACI) complicated with the cerebral microbleeds (CMBs).

METHODS:

Thirty patients with ACI and 36 patients with intracerebral hemorrhage (ICH) were included in this research. Two groups, studied by susceptibility-weighted imaging (SWI), were compared in terms of the number, location, and severity of CMBs. Then, 30 cases of ACI patients were divided into CMBs sub-group and non-CMBs sub-group. Univariate analysis between these two sub-groups was performed to determine the risk factors regarding the incidence of CMBs. For ACI patients, the number of CMBs before and after applying anti-platelet treatment were compared to examine the impacts of anti-platelet treatment on hemorrhagic transformation.

RESULTS:

CMBs were found to be more prevalent and severe in ICH patients than in ACI patients. CMBs in patients with ICH were more severe than in patients with ischemic stroke (IS), which indicates that CMBs closely relate to ICH. Hypertension and leukoaraiosis were found to have significant effects on the incidence of CMBs. After anti-platelet treatment, patients with CMBs (≥5) increased the number of CMB, whereas there was no obvious effect on patients with the CMBs less than 5 or no CMBs.

CONCLUSIONS:

The number of CMBs increased significantly among ACI patients with 5 or more CMBs before the anti-platelet treatment. CMBs are more frequently found in patients with hemorrhagic stroke than in patients with ischemic stroke, and more severe than the latter, which suggests that the clinical impact of higher association between the increase of the number of the CMBs and the hemorrhagic stroke.

Keywords

Susceptibility-weighted imaging cerebral microbleeds anti-platelet treatment leukoaraiosis hypertension

1 Introduction

Many previous clinical trials and meta-analyses have confirmed the benefits of anti-platelet treatment in the prevention of ischemic stroke [1 –3]. However, the complication of anti-platelet treatment such as intra-cerebral hemorrhage (ICH) is also quite alarming. Among the 60,000 reported cases of ICH in the United States, 4000 were found to be related to aspirin treatment [4]. The meta-analysis conducted by Antithrombotic Trialists’ Collaboration also suggested that the risk of ICH in the aspirin group was 1/4 higher than in the placebo group [5]. Therefore, the best treatment and prevention strategy for ischemic stroke lies in finding a balance between maximizing the benefits of anti-platelet treatment and minimizing the risk of post-treatment ICH.

Cerebral microbleeds (CMBs) are small hemosiderin deposits caused by leakages on small cerebral vessels [6]. These deposits can be identified as small, rounded, homogeneous, and hypointense lesions by T2-weighted gradient echo or susceptibility-weighted imaging (SWI) [7 –9]. CMBs have been found to be closely associated with aspirin-related ICH after acute ischemic stroke [10]. Robinson and Bhuta [11] suggested that there is a tendency for ICH to occur at the locations of CMBs. Several studies also supported this conclusion [3, 12]. Imaizumi et al. [13] found the ischemic and hemorrhagic locations of ICH patients through MRI and concluded that the presence of hemosiderin deposits is an important precursor for either symptomatic or asymptomatic ICH after cerebral infarction. Therefore, the detection of CMBs is of great clinical significance in planning the treatment regimen for cerebral infarction patients and evaluating their prognosis.

Previous researches on CMBs have mostly focused on the correlation between CMBs and ischemic stroke [7]. The dynamic change process of CMBs are still understudied. In order to address these limitations, objective of this research is to study the dynamic change process of intracerebral CMBs before and after anti-platelet treatment so as to evaluate the risk of anti-platelet treatment. Our study will be clinically important for preventing hemorrhagic transformation and optimizing the individualized treatment for ischemic stroke. Our study also helps to the prognosis evaluation.

2 Material and methods

2.1 Study subjects

The study recruited 30 cases of patients with acute non-disabling cerebral infarction (ACI) and 36 cases of patients with cerebral hemorrhage (ICH) from the Department of Neurology at Heibei General Hospital from 2011 to 2013. The study was approved by the ethical committee of Hebei General Hospital and was conducted in accordance with the Helsinki Declaration of 1975. All ICHs were spontaneous, not caused by surgeries. Patients with aneurysm rupture, arteriovenous malformation, non-cerebral parenchymal hemorrhage, and claustrophobia were excluded from this research.

2.2 ACI patients newly diagnosed patients with acute non-disability cerebral infarction

Inclusion Criteria: 1) All patients are between 40 and 80 years old, NIHSS ≤3 (National Institutes of Health Stroke Scale); 2) Recent onset occurs were within 48 hours before their admission and lasted for more than 1 hour without remission; 3) All patients took antiplatelet drugs for the first time; 4) Patients with infarcts larger than 1.5 cm were included, so as to exclude lacunar infarction.

Exclusion Criteria: 1) Patients were diagnosed with vascular malformation, tumor, abscess, non-cerebral vascular diseases and other pathological disorders by cranial CT or MRI; 2) Patients were with a clear indication for anticoagulation therapy (suspicion of cardiac embolism, such as atrial fibrillation, known artificial heart valves, suspected endocarditis); 3) There are only separate sensory symptoms, vision changes, dizziness or vertigo alone and no evidence of CT or MRI to acute infarction on; 4) Patients were with contraindications for anti-platelet treatment, such as allergic history, severe liver and kidney dysfunction, blood disease, gastrointestinal bleeding, intracranial space, pregnancy and as lactating women.

2.3 Drug administration

Thirty cerebral infarction patients were divided into two subgroups. One subgroup was given enteric aspirin tablet 100 mg once per night, the other was given clopidogrel 75 mg once per night. Continuous administration for 12 months (All patients were administered with anti-platelet treatment in compliance with individualized treatment principles). All the patients with the main study endpoints were gastrointestinal hemorrhage and ICH.

2.4 Definition of risk factors

Following are the common definition of the risk factors, which include: 1) Smoking: smoking history ≥ 6 months. 2) Hypertension: previously confirmed hypertension by definite diagnosis and taking anti-hypertensive drugs currently. 3) CMBs: rounded, homogenous foci with the signal loss under the SWI view, but with distinct borders and diameters between 2-10 mm. The foci with hypointense signals located symmetrically on both sides of the globus pallidus, which represent calcified areas and the flowing-void spots located in the distal branches of cerebral arteries are all excluded. All identified CMBs were confirmed by imaging professionals. All patients were then allocated into either the CMB subgroup or the non-CMB subgroup basing upon the detection results. The locations of CMBs were classified into three areas: cortico-subcortical (CSC) area, DGM (deep gray matter/basal ganglia and thalamus and infratentional (IT) area/brainstem and cerebellum. Basing upon the criteria established by Lee et al. [14], patients were classified into three groups according to the number CMBs identified: Mild (1–4 CMBs), intermediate (5–9 CMBs), and severe (≥10 CMBs). 4) Severity of cerebral white matter lesions (Severity of Leukoaraiosis): according to evaluation method of Aharon-Ptretz leukoaraiosis according to the severity can be devided into five grade from 0–4 [15]. 5) Carotid artery plaques: Color Doppler Ultrasound was used to measure the initial portion and bifurcation portion of the common carotid arteries, the posterior wall of the initial portion of the internal and external carotid arteries. IMT (Intima-media thickness) of carotid arteries on both sides was also measured. The larger IMT was taken as the indicator of the severity of carotid atherosclerosis: normal (IMT <0.9 mm), vascular intimal thickening (0.9 mm ≤ IMT ≤ 1.3 mm), plaque formation (protruding surface or IMT >1.3 mm).

2.5 MRI protocol

All scans were performed with a 3.0-T super-conductive MRI scanner (General Electric Company (GE) Co. the USA) with an 8-channel head/neck coil. Before the scan, the patient was placed in the supine position with the head in the center of the coil. Three-plane positioning scanning was conducted at the very beginning. Then transverse T1 weighted imaging (T1WI), T2 weighted imaging (T2WI), sagittal T1WI, diffusion-weighted imaging (DWI), 3D TOF magnetic resonance angiography (MRA) and susceptibility-weighted imaging (SWI) were performed. The scanning parameters were as follows

T1WI with spin-echo (SE) sequence, repetition time (TR)=1600.0 ms, echo time (TE)=26.2 ms, slice thickness = 5.0 mm, slice gap = 1.5 mm, field of vision (FOV)=24 cm×19.2 cm, matrix = 288×224, the excitation frequency (NE) is 1.0. T2WI with fast spin echo (FSE) sequence, TR = 5000.0 ms, TE = 102.5 ms, echo train length (ETL)=22, slice thickness = 5.0 mm, slice gap = 1.5 mm, FOV = 24 cm×24 cm, matrix = 352×352, NE = 1.5.

DWI with En sequence, TR = 5000.0 ms, TE = 74.9 ms, FOV = 24 cm×24 cm, matrix = 160×160, NE = 1.0, b = 1000.

3D TOF MRA: TOF fast field echo (FFE), axial acquisition, TR = 20.0 ms, TE = 3.4 ms, flip angle = 15°, slice thickness = 1.2 mm, slice gap = 0.6 mm, FOV = 20 cm×20 cm, matrix = 224×320, NE = 1.0.

SWI with 3DT1-FFE sequence, TR = 78.6 ms, TE = 48.04 ms, flip angle = 15°, slice thickness = 2 mm, slice gap = 1 mm, FOV = 24 cm×24 cm, matrix = 256×384, intensity contrast images and phase contrast images were both obtained.

2.6 Image post-processing and analysis of data

All original image were uploaded to the GE ADW4.3 workstation, and the Functool software was applied for post-processing of SWI images. The Minimum Intensity Projection (MinIP) image of SWI was performed using the 3D MIP software package (GE ADW4.3 workstation) (slice thickness = 2 mm). All the MR images were analyzed blindly by two experienced neuro-radiologists which ensured the accuracy. The qualitative results were being descriptive and concordant. Images were analyzed by the combination of CT and SWI (including phase images and MinIP images) images. The diagnosis of HT images (hemorrhagic infarction) were reviewed and recorded. During the process, the signal characteristics and display of HT in each sequence were recorded, and the imaging features of calcification and micro-hemorrhage on the SWI images and the appearance and morphology of the vessels in the infarct and surrounding areas were observed.

2.7 Follow up

1) The first day after being grouped, patients were examined through routine blood test, routine urine test, biochemical tests, blood coagulation tests (prothrombin time (PT), activated partial thromboplastin time (APTT) thrombin time (TT), fibrinogen (FIB), cranial MRI, DWI, MRA, SWI, cranial CT, carotid artery ultrasonography. 2) One month after being grouped, adverse events, medication compliance, and drug combination were recorded. The situation of blood pressure control and blood glucose control were also enquired. 3) 6 months after being grouped, adverse events, medication compliance, drug combinations were again recorded, and the control of blood pressure and blood glucose were also enquired. 4) 12 months after being grouped, patients were re-examined through a routine blood test, cranial SWI, carotid artery ultrasonography and the control of blood pressure and blood glucose were also recorded. Two imaging professionals evaluated the SWI results before and after the treatment. Patients were divided into CMB and non-CMB subgroup according to the presence of CMBs. The number of cases with CMBs, the location of CMBs, the number of CMBs in each case, the presence of lacunar cerebral infarction, leukoaraiosis, blood pressure, blood lipid, blood glucose and other test results was also recorded.

2.8 Statistical analysis

Statistical analysis was performed using SPSS 13. (IBM, Armonk, NY). Relationships between the two groups were assessed with the Chi-square test for nominal variables. The normal distribution of the measurement data was expressed by means±standard deviation (SD). The t-test was used for continuous variables. If the P value was less than 0.05, it was accepted as statistically significant.

3 Results

3.1 Observation result of CMBs

As shown in Table 1, all the 66 patients in both the ACI and the ICH group were examined through SWI, and the occurrence of CMBs occurred in both groups (Fig. 1). Among the patients with ACI, 16 cases were positive, indicating a positive rate of 53.3%. In the ICH group, 28 patients were found CMBs positive, indicating a positive rate of 77.8%. A total of 38 CMBs were identified in the 16 CMB-positive ACI patients. The number of CMBs among these patients ranges from 1 to 12. According to the number of CMBs, 4 patients were identified as mild, 8 as intermediate and another 4 as severe. 15 of these CMBs were located in the CSC area, 22 in the DGM area, and only 1 in the IT area. Therefore, CSC and DGM were found to be the predilection sites for CMBs among CMB-positive ACI patients. A total of 159 CMBs were identified in the 28 CMB-positive ICH patients. The number of CMBs among these patients ranges from 1–47. According to the number of CMBs, 4 patients were identified as mild, 8 as intermediate and another 16 as severe. 60 of these CMBs were found in the CSC area, 62 in the DGM area and 28 in the IT area (Table 1). The predilection sites for CMBs are mainly the DGM area and the CSC area (Fig. 2). CMBs were less likely to be found in the IT area.

Fig.1

CMBs were mostly found in the basal ganglia and thalamus region in ACI patients.

Fig.2

Patient in the ACI group, SWI reminds multiple CMBs in the CSC and DGM areas. MRI reminded severe demyelination of the cerebral white matter MRI.

Table 1

The observation result of CMBs

Characteristic	ACI (n = 30)	ICH (n = 36)	P-value
CMB-positive patients	16/30 (53.3%)	28/36 (77.8%)	0.852
Predilection sites of CMBs (%)			0.917
CSC	15/38 (39.5%)	60/159 (37.7%)
DGM	22/38(57.9%)	62/36 (39.0%)
IT	1/38 (2.6%)	28/159 (17.6%)
Severity of CMBs (%)			0.349
Mild	4/30 (13.3%)	4/36 (11.1%)
Intermediate	8/30 (26.7%)	8/36 (22.2%)
Severe	4/30 (13.3%)	16/36(53.3%)

CMB, cerebral microbleed; CSC, corticosubcortical; DGM, deep gray matter; IT, infratentional.

3.2 Univariate analysis of the risk factors for CMBs

Univariate analysis was only conducted for ACI patients. ACI patients were divided into the CMB subgroup and the non-CMB subgroup, basing upon the presence of the CMBs before the anti-platelet treatment.

The univariate analysis results showed that there were not statistically significant differences between the CMB subgroup and the non-CMB subgroup in factors including gender, age, blood TC, TG, HDL, LDL. However, Table 2 showed the degree of cerebral white matter lesions and hypertension had the difference (P≤0.05).

Table 2
Univariate Analysis of CBM–severity of Leukoaraiosis and hypertension

Factors Non-CMB group (14 cases) CMB group (16 case) t/χ² P-value

Hypertension

Systolic Pressure (mmHg) 138.57±17.728 161.25±18.077 –2.446 0.029^*

Severity of Leukoaraiosis (degree/case) 11.987 0.017^*

0 6 0

1 6 0

2 2 6

4 0 6

5 0 4

Factors	Non-CMB group (14 cases)	CMB group (16 case)	t/χ²	P-value
Hypertension
Systolic Pressure (mmHg)	138.57±17.728	161.25±18.077	–2.446	0.029^*
Severity of Leukoaraiosis (degree/case)			11.987	0.017^*
0	6	0
1	6	0
2	2	6
4	0	6
5	0	4

CMB, cerebral microbleed.

3.3 Changes in CMBs before and after taking anti-platelet drugs

Table 3 showed that the number of CMBs before and after anti-platelet treatment which indicated that there was significant difference (P = 0.005, Fig. 3). The Table 3 also showed the number of CMBs were significantly increased in patients with more than 5 microbleeds (P = 0.027, statistically significant) after administration of antiplatelet agents, as compared to pretreatment levels. There was no significant difference regarding the number of CMBs in patients with less than 5 microbleeds between after the pre-and post-anti-platelet treatment group (P = 0.085 >0.05, Table 3). It was suggested that patients with CMB <5 in the skull were safe to take antiplatelet aggregation therapy. Patients with CMB ≥5 were need to monitor the changes of SWI dynamically. We will collected more patients with CMBs to study the association between the CMBs and final clinical outcome in our future study.

Fig.3

Image A and B were for ACI patients prior to the anti-platelet treatment. Image C and D showed the increase of CMBs in the basal ganglia region of ACI after anti-platelet treatment. Image E and F showed an increase in CMB counts in the cortex/subcortex region of ACI patients after anti-platelet treatment.

Table 3

Analysis on the number of CMBs before and after anti-platelet treatment

Group	Cases	No. of CMBs $\bar{x} \pm S$	t	P-value
Pre-treatment	30	2.82±2.96	3.249	0.005
Post-treatment	30	4.47±4.52
No. of CMBs <5			1.91	0.085
Pre-treatment	22	1.00±1.61
Post-treatment	22	1.91±2.81
No. of CMBs≥5			3.105	0.027
Pre-treatment	8	6.17±1.47
Post-treatment	8	9.17±2.93

CMB, cerebral microbleed.

4 Discussion

CMBs is a cerebral parenchymal lesion characterized by micro-hemorrhage. It is sometimes alternatively referred to as chronic CMBs, resting CMBs, or lacunar hemorrhage. The major pathological change is the deposition of hemosiderin around small cerebral vessels [16, 17]. Most of the clinical symptoms without being recognized, their diagnosis is often based upon the use of imaging techniques. The CMBs as general maker for cerebral microangiopathy were not known to many people as the widespread application of MRI [9]. CMBs are commonly identified as punctuate, rounded, homogeneous, signal-loss foci with distinct borders and diameters between 2–10 mm on SWI sequences.

All 66 patients in this study were examined by SWI. CMBs were found in both the ACI group and the ICH group. In earlier studies [18, 19], the incidence rate of CMBs was found to be about 21%–47% among ischemic patients, 46%–52% among ACI patients and 33%–80% among ICH patients. The findings of this research show a 53.3% incidence rate for ACI patients and a 77.8% incidence rate for ICH patients, which are quite consistent with findings of previous studies [10]. CMBs were also found to be more severe in the ICH group than in the ACI group, suggesting that CMBs is closely related with ICH.

In this study, a total of 38 CMBs were identified in the 16 CMB-positive ACI patients, 15 of these CMBs were located in the CSC area, 22 in the DGM area, and only 1 in the IT area. A total of 159 CMBs were identified in the 28 CMB-positive ICH patients. Therefore, CSC and DGM were found to be the predilection sites for CMBs among. 60 of these CMBs were found in the CSC area, 62 in the DGM area and 28 in the IT area. Similarly, CSC and DGM were found to be the predilection sites of CMB among CMB-positive ACI patients and ICH patients. The predilection sites of CMBs is consistent with the findings of extant researches [18 , 20–22]. Blood supplying arteries for the DGM area are deep perforating branches. These branches are the predilection sites for cavity infarction and ICH. This makes the DGM area a predilection site for CMBs. The severity of CMBs was significantly higher in the hemorrhage group than in the infarction group, suggesting that CMBs is closely related to the cerebral hemorrhage. ICH is often accompanied by hypertension. Under the impact of hypertension, the arteries of ICH patients would become more fragile as they get older. Besides, hypertension could also lead to the hyaloid degeneration of small arteries, Charcot-Bouchard micro-aneurysm, and the degeneration of arterial media. Charcot-Bouchard micro-aneurysm has a very thin wall, lacks elastic fiber and smooth muscle layer. It could easily rupture and lead to ICH.

Histopathological studies have found that most CMBs occur around these small arteries or microaneurysms, suggesting a relationship between CMBs and microangiopathy caused by hypertension. These lesions on micro-vessels are the pathological basis for ICH under hypertension. CMBs could alternatively be understood as a type of intracerebral target organ damage in patients with hypertension. Leukoencephalopathy is often thought to be related to cerebral microangiopathy [23]. It is pathologically manifested through the formation of amyloid deposition on the media and external layer of small and medium-sized blood vessels on the surface of the cerebral cortex, and pia mater. This would subsequently lead to the loss of vascular elasticity and rise of fragility, and ultimately to CMBs around these small vessels. These microangiopathies could cause hemorrhage, infarction, and leukoencephalopathy. The severity of leukoencephalopathy would develop with the worsening of microangiopathy. SWI on ACI patients in this research reminds that cerebrovascular amyloidosis and CMBs are often present in the cortical, subcortical and subarachnoid areas, but seldom in the area of basal ganglia, thalamus, and brain stem. The findings of this research show that there is a significant relationship between CMBs and the severity of leukoaraiosis. The positive correlation between CMBs and leukoaraiosis suggests leukoaraiosis and CMBs may share the same pathogenesis.

Thirty patients included this research were given anti-platelet treatment (either aspirin 100 mg, qn or clopidogrel 75 mg, qn). The numbers of CMBs before and after the treatment were found to be statistically different among patients with no less than 5 CMBs prior to the treatment, which consistent with Soo’s [24] study. The risk of secondary ICH rises with the increase of CMBs among patients with both cerebral infarction and CMBs, especially when the number of CMBs is above 5. On the contrary, no statistical difference can be found between pre-treatment CMB count and post-treatment CMB count in patients with less than 5 CMBs prior to the treatment. This means that anti-platelet treatment could result in more CMBs for patients with multiple CMBs, and the study of Vernooij et al. [25] also confirm this result. However, there are also studies showing there is no correlation between anti-platelet treatment, the incidence of CMBs and hemorrhagic transformation among ACI patients [26, 27].

Neither the individualized treatment in the ACI group nor the ICH group has resulted in hemorrhagic transformation. This may be due to the limited observation window, limited sample size, and that patients included in this research are mostly with mild or intermediate cerebral infarction. Moreover, all the hospital-based stroke cases may not represent a community-based group. So far, we still do not have adequate and systematic knowledge about CMBs. The clinical meaning of this research is also limited by its small sample size and research design. Therefore, multi-centered prospective studies with better designs should be conducted in the future to obtain more knowledge about the relationship between CMBs and their hemorrhagic transformation. It will help improve our treatment and secondary prevention strategy for ACI patients.

5 Conclusion

The anti-platelet treatment could be a safe treatment for patients with less than 5 CMBs but should be monitored for patients with 5 or more CMBs. The univariate analysis shows that effective control over blood pressure and the progression of leukoaraiosis might be clinically beneficial for controlling the development of new CMBs. CMBs are more frequently found in patients with hemorrhagic stroke than in patients with ischemic stroke, and be more severe than the latter.

References

Hacke

, Donnan

, Fieschi

et al., Association of outcome with early stroke treatment: Pooled analysis of ATLANTIS, ECASS, and NINDS rt-PA stroke trials, Lancet 363 (2004), 768–774.

Chen

, Xiao

, Lin

et al., The Role of Circulating Platelets Microparticles and Platelet Parameters in Acute Ischemic Stroke Patients, J Stroke Cerebrovasc Dis 24 (2015), 2313–2320.

McCarron

M.O.

and McVerry

, Stroke: Will cerebral microbleeds influence stroke prevention? Nat Rev Neurol 5 (2009), 588–590.

Hart

R.G.

, Tonarelli

S.B.

and Pearce

L.A.

, Avoiding central nervous system bleeding during antithrombotic therapy: Recent data and ideas, Stroke 36 (2005), 1588–1593.

Collaborative meta-analysis of randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients, BMJ 324 (2002), 71–86.

Kleinert

, Fazekas

, Roob

et al., Histopathologic analysis of foci of signal loss on gradient-echo T2*-weighted MR images in patients with spontaneous intracerebral hemorrhage: Evidence of microangiopathy-related cerebral microbleeds, AJNR Am J Neuroradiol 20 (1999), 637–642.

Charidimou

, Kakar

, Fox

and Werring

D.J.

, Cerebral microbleeds and recurrent stroke risk: Systematic review and meta-analysis of prospective ischemic stroke and transient ischemic attack cohorts, Stroke 44 (2013), 995–1001.

Offenbacher

, Fazekas

, Schmidt

et al., MR of cerebral abnormalities concomitant with primary intracerebral hematomas, AJNR Am J Neuroradiol 17 (1996), 573–578.

Greenberg

S.M.

, Vernooij

M.W.

, Cordonnier

et al., Cerebral microbleeds: A guide to detection and interpretation, The Lancet Neurology 8 (2009), 165–174.

10.

Gregoire

S.M.

, Jager

H.R.

, Yousry

T.A.

et al., Brain microbleeds as a potential risk factor for antiplatelet-related intracerebral haemorrhage: Hospital-based, case-control study, Journal of Neurology, Neurosurgery & Psychiatry 81 (2010), 679–684.

11.

Robinson

R.J.

and Bhuta

, Susceptibility-weighted imaging of the brain: Current utility andotential applications, J Neuroimaging 21 (2011), e189–204

12.

Wang

D.N.

, Hou

X.W.

, Yang

B.W.

et al., Quantity of Cerebral Microbleeds, Antiplatelet Therapy, and Intracerebral Hemorrhage Outcomes: A Systematic Review and Meta-analysis, J Stroke Cerebrovasc Dis 24 (2015), 2728–2737.

13.

Imaizumi

, Honma

, Horita

et al., Dynamics of dot-like hemosiderin spots on T2*-weighted MRIs associated with stroke recurrence, J Neuroimaging 17 (2007), 204–210.

14.

Lee

S.H.

, Bae

H.J.

, Yoon

B.W.

et al., Low concentration of serum total cholesterol is associated with multifocal signal loss lesions on gradient-echo magnetic resonance imaging: Analysis of risk factors for multifocal signal loss lesions, Stroke 33 (2002), 2845–2849.

15.

Aharon-Peretz

, Cummings

J.L.

and Hill

M.A.

, Vascular dementia and dementia of the Alzheimer type. Cognition, ventricular size, and leuko-araiosis, Arch Neurol 45 (1988), 719–721.

16.

Fazekas

, Kleinert

, Roob

et al., Histopathologic analysis of foci of signal loss on gradient-echo T2*-weighted MR images in patients with spontaneous intracerebral hemorrhage: Evidence of microangiopathy-related microbleeds, AJNR Am J Neuroradiol 20 (1999), 637–642.

17.

Shoamanesh

, Kwok

C.S.

and Benavente

, Cerebral microbleeds: Histopathological correlation of neuroimaging, Cerebrovasc Dis 32 (2011), 528–534.

18.

Schonewille

W.J.

, Singer

M.B.

, Atlas

S.W.

and Tuhrim

, The prevalence of microhemorrhage on gradient-echo magnetic resonance imaging in acute lacunar infarction, J Stroke Cerebrovasc Dis 14 (2005), 141–144.

19.

Offenbacher

, Fazekas

, Schmidt

, Koch

, Fazekas

and Kapeller

, MR of cerebral abnormalities concomitant with primary, AJNR Am J Neuroradiol 17 (1996), 573–578.

20.

Werring

D.J.

, Coward

L.J.

, Losseff

N.A.

et al., Cerebral microbleeds are common in ischemic stroke but rare in TIA, Neurology 65 (2005), 1914–1918.

21.

Martin Dichgans

M.H.

, Herzog

, Peters

, et al., MRI evidence of past cerebral microbleeds in CADASIL: A gradient-echo magnetic resonance imaging and autopsy study, Stroke 33 (2002), 67–71.

22.

Tanaka

U.Y.

, Nakayama

, Takano

et al., Small chronic hemorrhages and ischemic lesions in association with spontaneous intraccrebral hematomas, stroke 30 (1999), 1637–1642.

23.

Loitfelder

S.S.

, Schwingenschuh

et al., Cerebral microbleeds, Panminerva Med 54 (2012), 149–160.

24.

Soo

Y.O.

, Yang

S.R.

, Lam

W.W.

et al., Risk vs benefit of anti-thrombotic therapy in ischaemic stroke patients with cerebral microbleeds, J Neurol 255 (2008), 1679–1686.

25.

Vernooij

M.W.

, Haag

M.D.

, van der Lugt

, Hofman

, Krestin

G.P.

, Stricker

B.H.

and Breteler

M.M.

, Use of antithrombotic drugs and the presence of cerebral microbleeds, Antithrombotic Drug Use 66 (2009), 714–720.

26.

Lee

S.H.

, Kang

B.S.

, Kim

and Roh

J.K.

, Does microbleed predict haemorrhagic transformation after acute atherothrombotic or cardioembolic stroke? J Neurol Neurosurg Psychiatry 79 (2008), 913–916.

27.

Nishikawa

, Ueba

, Kajiwara

et al., Cerebral microbleeds in patients with intracerebral hemorrhage are associated with previous cerebrovascular diseases and white matter hyperintensity, but not with regular use of antiplatelet agents, Neurol Med Chir (Tokyo) 49 (2009), 333–338; discussion 338-339.