RESEARCH—HUMAN—CLINICAL STUDIES RESEARCH—HUMAN—CLINICAL STUDIES
TOPIC Sang-Beom Jeon, MD, PhD*‡§ Gunjan Parikh, MD‡k H. Alex Choi, MD, MS‡§ Neeraj Badjatia, MD, MS‡k
Cerebral Microbleeds in Patients With Acute Subarachnoid Hemorrhage
Kiwon Lee, MD‡§ J. Michael Schmidt, PhD‡ Hector Lantigua, MD‡ E. Sander Connolly, MD¶ Stephan A. Mayer, MD‡ Jan Claassen, MD, PhD‡ *Department of Neurology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea, Departments of ‡Neurology and ¶Neurosurgery, Columbia University College of Physicians and Surgeons, New York, New York; §Departments of Neurology and Neurosurgery, University of Texas Medical School at Houston, Houston, Texas; kDepartment of Neurology, University of Maryland School of Medicine, Baltimore, Maryland Correspondence: Sang-Beom Jeon, MD, PhD, Department of Neurology, Asan Medical Center, University of Ulsan College of Medicine, 88, Olympic-ro 43-gil, Songpa-gu, Seoul, Republic of Korea. E-mail: [email protected] Received, July 2, 2013. Accepted, October 28, 2013. Published Online, October 30, 2013. Copyright © 2013 by the Congress of Neurological Surgeons
WHAT IS THIS BOX? A QR Code is a matrix barcode readable by QR scanners, mobile phones with cameras, and smartphones. The QR Code above links to Supplemental Digital Content from this article.
BACKGROUND: Cerebral microbleeds (CMBs) are commonly found after stroke but have not previously been studied in patients with subarachnoid hemorrhage (SAH). OBJECTIVE: To study the prevalence, radiographic patterns, predictors, and impact on outcome of CMBs in patients with SAH. METHODS: We analyzed retrospectively 39 consecutive patients who underwent T2*weighted gradient-echo imaging within 7 days after onset of spontaneous SAH. We report the frequency and location of CMBs and show their association with demographics, vascular risk factors, the Hunt-Hess grade, the modified Fisher Scale, the Acute Physiological and Chronic Health Evaluation II, magnetic resonance imaging findings including diffusion-weighted imaging lesions, and laboratory data, as well as data on rebleeding, global cerebral edema, delayed cerebral ischemia, seizures, the Telephone Interview for Cognitive Status, and the modified Rankin Scale. RESULTS: Eighteen patients (46%) had CMBs. Of these patients, 9 had multiple CMBs, and overall a total of 50 CMBs were identified. The most common locations of CMBs were lobar (n = 23), followed by deep (n = 15) and infratentorial (n = 12). After adjustment for age and history of hypertension, CMBs were related to the presence of diffusion-weighted imaging lesions (odds ratio, 5.24; 95% confidence interval, 1.14-24.00; P = .03). Three months after SAH, patients with CMBs had nonsignificantly higher modified Rankin Scale scores (odds ratio, 2.50; 95% confidence interval, 0.67-9.39; P = .18). CONCLUSION: This study suggests that CMBs are commonly observed and associated with diffusion-weighted imaging lesions in patients with SAH. Our findings may represent a new mechanism of tissue injury in SAH. Further studies are needed to investigate the clinical implications of CMBs. KEY WORDS: Magnetic resonance imaging, Microbleeds, Subarachnoid hemorrhage Neurosurgery 74:176–181, 2014
DOI: 10.1227/NEU.0000000000000244
C
erebral microbleeds (CMBs), commonly found in patients with ischemic stroke, intracerebral hemorrhage, and cerebral amyloid angiopathy, are recognized as radiologic surrogates for bleeding-prone microangiopathy.1 The presence of CMBs has not previously been reported in acute spontaneous subarachnoid hemorrhage (SAH). Finding CMBs in patients ABBREVIATIONS: CI, confidence interval; CMB, cerebral microbleed; DWI, diffusion-weighted imaging; DWIL, diffusion-weighted imaging lesion; OR, odds ratio; SAH, subarachnoid hemorrhage Supplemental digital content is 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 (www.neurosurgery-online.com).
176 | VOLUME 74 | NUMBER 2 | FEBRUARY 2014
www.neurosurgery-online.com
with SAH may shed light on the pathomechanism of CMBs and SAH. It may also help explain neurological sequela after SAH because CMBs have been associated with cognitive impairment, depression, gait instability, and poor quality of life in other populations.2-5 Here, we report prevalence, radiographic patterns, predictors, and impact on outcome of CMBs in patients with acute SAH.
PATIENTS AND METHODS This retrospective analysis of a prospectively collected data set considered all patients enrolled in the Columbia University SAH Outcomes Database Project between June 1, 2006, and April 30, 2011.6 We included patients who had a diagnosis of SAH confirmed by computed tomography scan and underwent 3.0-T magnetic resonance imaging (MRI),
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
MICROBLEEDS IN SUBARACHNOID HEMORRHAGE
TABLE 1. Baseline Characteristics According to the Presence of Cerebral Microbleedsa CMBs Absent (n = 21) Demographics Age, y Male sex, n (%) White, n (%) Vascular risk factors, n (%) Hypertension Diabetes mellitus Smoking Alcohol Previous stroke Antithrombotic agents before SAH Initial characteristics Hunt-Hess grade 3 to 5, n (%) Modified Fisher Scale score 3 or 4, n (%) Hijdra Scale score Global cerebral edema, initial, n (%) Aneurysm Presence, n (%) Size, mm Location, anterior (vs posterior), n (%) APACHE II score Systolic blood pressure, mm Hg Diastolic blood pressure, mm Hg Body temperature, F Leukocyte, ·1000/mL Platelet count, ·1000/mL Partial thromboplastin time, s Prothrombin time, INR GFR, mLmin211.73 m22 MRI findings, concurrent with CMB Interval from symptom onset to MRI, d Leukoaraiosis, n (%) DWILs present, n (%) Punctate pattern Territorial pattern Mixed pattern (punctate1territorial)
52 (38-64) 5 (24) 11 (52) 7 2 6 1 1 3
Present (n = 18) 60 (45-66) 6 (33) 8 (44)
(33) (10) (29) (5) (5) (14)
3 2 9 2 0 3
OR (95% CI) 1.03 (0.98-1.07) 1.60 (0.39-6.51) 0.73 (0.21-2.57)
(17) (11) (50) (11) (0) (17)
0.40 1.19 2.50 2.50 0 1.2
(0.09-1.86) (0.15-9.41) (0.67-9.39) (0.21-30.1) (0) (0.21-6.84)
11 14 13.5 7
(52) (67) (5-18) (33)
11 11 15.5 2
(61) (61) (5.5-21.5) (11)
1.43 0.79 1.03 0.25
(0.40-5.12) (0.21-2.92) (0.95-1.11) (0.04-1.41)
16 6.0 14 14 148 83 98.2 11.6 243 25.8 1.03 88
(76) (3.0-7.0) (88) (9-20) (128-180) (66-96) (97.4-98.6) (9.7-18.6) (208-333) (23.8-28.3) (0.96-1.09) (71-98)
14 8.0 12 16 174 91 98.7 10.8 232 26.1 1.02 86
(78) (5.0-10.0) (86) (8-22) (147-180) (80-100) (97.4-99.5) (9.5-14.95) (168-282) (23.7-28.4) (0.95-1.07) (70-110)
1.09 1.11 0.86 1.03 1.01 1.02 1.49 0.90 0.99 0.95 0.09 1.01
(0.25-4.89) (0.92-1.35) (0.10-7.04) (0.94-1.13) (0.99-1.03) (0.99-1.06) (0.78-2.85) (0.77-1.04) (0.99-1.00) (0.80-1.12) (0-139.4) (0.98-1.03)
(1-5) (29) (83) (28) (17) (39)
1.24 2.50 5.50 0.96 4.00 3.82
(0.92-1.68) (0.50-12.47) (1.22-24.81)b (0.24-3.90) (0.38-42.37) (0.81-17.93)
1 3 10 6 1 3
(1-2.5) (14) (48) (29) (5) (14)
2 5 15 5 3 7
a
APACHE II, Acute Physiology and Chronic Health Evaluation II; CI, confidence interval; CMB, cerebral microbleed; DWIL, diffusion-weighted imaging lesion; GFR, glomerular filtration rate; INR, international normalized ratio; MRI, magnetic resonance imaging; OR, odds ratio; SAH, subarachnoid hemorrhage. Values are number (column %) or median (interquartile range) as appropriate. b P , .05.
including T2*-weighted gradient-echo imaging, within 7 days after symptom onset. We excluded patients whose origin of SAH was trauma, arteriovenous malformation, vasculitis, or coagulopathy, as well as those whose imaging quality was poor. This study was approved by our institutional review board. Written informed consents were obtained from the patients or surrogates. We obtained demographics and medical history by interviewing patients or their families. Demographics, vascular risk factors, the Hunt-Hess grade, the Acute Physiologic and Chronic Health Evaluation II, the modified Fisher Scale, the Hijdra Scale, and routine laboratory or physiological data, as well as data on global cerebral edema on initial
NEUROSURGERY
and/or follow-up computed tomography, rebleeding of SAH, delayed cerebral ischemia, and seizures, were assessed as previously described in detail.6 The modified Rankin Scale score and Telephone Interview for Cognitive Status were evaluated at discharge from the intensive care unit and 3 months after SAH.7 Imaging was performed with a 3-T MRI unit (Signa HDxt, GE Healthcare, Milwaukee, Wisconsin) and an 8-channel phased-array head coil. MRIs were ordered to help assess prognosis and to determine the extent of the underlying brain injury. The decision to send a patient to MRI was made by consensus between the neurosurgery and neurocritical care attending physicians. The MRI protocol included T2*-weighted
VOLUME 74 | NUMBER 2 | FEBRUARY 2014 | 177
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
JEON ET AL
TABLE 2. Blood Pressure From Admission to Magnetic Resonance Imaging According to the Presence of Cerebral Microbleedsa CMBs Absent (n=21) Initial SBP, mm Hg Initial DBP, mm Hg Highest SBP, mm Hg Highest DBP, mm Hg DSBP (highest2lowest), mm Hg DDBP (highest2lowest), mm Hg
152.7 81.0 156.7 83.4 41.8 27.0
6 6 6 6 6 6
Present (n=18)
32.7 161.9 6 27.8 24.0 90.6 6 16.9 29.6 162.3 6 25.4 24.3 85.5 6 12.6 26.2 53.9 6 33.8 26.9 31.1 6 15.0
P Value .35 .17 .52 .74 .21 .57
a
CMB, cerebral microbleeds; DBP, diastolic blood pressure; SBP, systolic blood pressure.
gradient-echo imaging and diffusion-weighted imaging (DWI). The gradient-echo imaging parameters were 1400-millisecond repetition time, 40-millisecond echo time, 15 flip angle, 5-mm slice thickness, 5-mm interslice gap, 512 · 512 matrix, and 220-mm field of view. The DWI parameters were 7000-millisecond repetition time, 73-millisecond echo time, 5-mm slice thickness, 5-mm interslice gap, 256 · 256 matrix, 220-mm field of view, and 2 b values of 0 and 1000 s/mm2. MRIs were reviewed jointly by 2 neurologists (S.-B.J. and G.P.), who were blinded to the clinical data. A third investigator (J.C.) was consulted in cases of disagreement. CMBs were defined as unambiguous homogeneous round signal loss lesions with diameters of 2 to 10 mm.8 Hypointense lesions in subarachnoid space, ventricles, intracerebral hematoma, and DWI lesions (DWILs) were not regarded as CMBs. The number of CMBs was counted according to locational classification (lobar, deep, and infratentorial) by the Microbleed Anatomic Rating Scale.5 DWIL was defined as a hyperintense lesion on DWI (b = 1000 s/mm2) with a corresponding signal reduction in the apparent diffusion coefficient map. The presence of DWIL was interpreted in each location (lobar, deep, and infratentorial). The number of DWILs was not counted because the size and shape of DWILs were variable, and it was uncertain whether multiple tiny DWILs were more detrimental than a single large DWIL. Instead, the patterns of DWILs were classified into punctate (,10 mm in diameter) and territorial ($10 mm in diameter) lesions.9 Leukoaraiosis was defined as a Fazekas score $2 in either periventricular or deep white matter on the fluid-attenuated inversion recovery imaging.10 We analyzed continuous and categorical variables using the t and Fisher exact tests and logistic regression analysis, as appropriate. A 2-tailed value of P , .05 was considered significant. SPSS for Windows (version 17.0, SPSS Inc, Chicago, Illinois) was used for all statistical analyses.
RESULTS During the study period, 408 patients with SAH were enrolled in our database. Of those, 39 met inclusion criteria. Compared with the 369 excluded patients, 39 included patients were less likely to have a history of hypertension (see Table, Supplemental Digital Content 1, http://links.lww.com/NEU/A591, which
178 | VOLUME 74 | NUMBER 2 | FEBRUARY 2014
illustrates baseline characteristics according to study inclusion or exclusion). Of the 39 included patients, the median age was 56 years (interquartile range, 44-65 years), and 28 (72%) were female (Table 1). Blood pressure variables are shown in Table 2. The median time interval from symptom onset to the acquisition of MRI was 40 hours (interquartile range, 28-101 hours). In 30 cases (77%), the causes of SAH were ruptured aneurysms, with perimesencephalic hemorrhage being the second most common. CMBs were seen in 18 patients (46%). Of these, 9 patients had multiple ($2) CMBs. DWILs were found in 25 patients (64%; Figure 1). The presence of DWILs was associated with the presence of CMBs (odds ratio [OR], 5.50; 95% confidence interval [CI], 1.22-24.81; P = .03; Table 1). Logistic regression analysis controlling for age and hypertension showed that DWILs were independently associated with CMBs (OR = 5.24; 95% CI, 1.14-24.00; P = .03; and OR, 5.61; 95% CI, 1.21-25.98; P = .03). Of 50 CMBs, 23 were found in the lobar (8 frontal, 4 parietal, 9 temporal, 1 occipital, and 1 insular), 15 in the deep (3 basal ganglia, 7 thalamus, 4 internal capsule, and 1 corpus callosum), and 12 in the infratentorial (2 brainstem and 10 cerebellum) location. Patients with CMBs in the deep location had a larger amount of subarachnoid blood in the basal sylvian fissure and suprasellar cistern (median, 6.0; interquartile range, 2.0-8.0) than those without CMBs in this location (median, 11.5; interquartile range, 8.8-12.0; P = .006; see Table, Supplemental Digital Content 2, http://links.lww.com/NEU/A592, which illustrates the locational relationship between the presence of CMBs and the amount of subarachnoid blood). The patterns of the DWILs were punctate in 11 patients, territorial in 4 patients, and mixed (both punctate and territorial) in 10 patients. These patterns were not associated with the presence of CMBs (punctate, P = .96; territorial, P = .25; and mixed, P = .09). Among 15 patients who had both CMBs and DWILs, locational (lobar, deep, or infratentorial) associations between CMBs and DWILs were not significant. Patients with CMBs were more likely to have a higher modified Rankin Scale score on discharge (OR, 2.10) and at 3 months after SAH (OR, 2.50) and lower Telephone Interview for Cognitive Status score on discharge (OR, 5.06) and at 3 months (OR, 1.50) after SAH, but the associations were not statistically significant (Table 3).
DISCUSSION CMBs were found in approximately half of patients with acute SAH. The most common locations were lobar, followed by deep and infratentorial regions. The presence of CMBs was significantly associated with the occurrence of DWILs. CMBs were not significantly related to acute complications of SAH such as rebleeding, global cerebral edema, delayed cerebral ischemia, or seizures. In patients with CMBs, poor functional outcome measured by modified Rankin Scale score and cognitive impairment measured by Telephone Interview for Cognitive Status
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
MICROBLEEDS IN SUBARACHNOID HEMORRHAGE
FIGURE 1. Examples of cerebral microbleeds (CMBs) and diffusion-weighted imaging lesions (DWILs). A, a previously healthy 66-year-old woman presented with headache and loss of consciousness. Neurological examination and computed tomography (CT) with angiography suggested subarachnoid hemorrhage (SAH; Hunt-Hess grade 5, and modified Fisher Scale score, 4) resulting from a ruptured aneurysm in the anterior cerebral artery. Magnetic resonance imaging (MRI) performed 38 hours after symptom onset revealed multiple CMBs and DWILs. B, a 58-year-old woman with a 30–pack-year smoking history developed a sudden headache and altered mental status. Initial examination and CT with angiography revealed SAH (Hunt-Hess grade 4 and modified Fisher Scale score, 4) caused by a ruptured aneurysm in the right distal internal carotid artery. MRI performed 37 hours after symptom onset showed multiple CMBs and DWILs. White arrows indicate CMBs; black arrows, DWILs.
tended to be high at discharge and at 3 months after SAH, but this association was not statistically significant. In other populations, CMBs have been independently associated with variable neurological deficits: cognitive impairment, depression, gait instability, poor quality of life, and future hemorrhagic risk.1-5,11,12 Given that SAH survivors also have the aforementioned neurological complications, the contribution of CMBs to such morbidities after SAH needs to be further investigated.13 The high (64%) frequency of DWILs in our SAH patients (a median Hunt-Hess grade of 3) is in agreement with previous reports: up to 82% in high-grade SAH.13 DWILs after SAH have been postulated to be acute infarcts caused by vasospasm of large or small vessels, microemboli, global ischemia, cortical spreading
NEUROSURGERY
depression, etc.9 Interestingly, our study parallels a recent report that found that both CMBs and DWILs were common after intracerebral hemorrhage. 14 These reports may suggest some common pathomechanism after hemorrhagic and ischemic stroke. Old age and chronic hypertension are important risk factors for CMBs and leukoaraiosis, which are known imaging markers of chronic microangiopathy and are strongly linked to each other.1,14 However, these factors were not associated with CMBs in our population. Recent studies have shown that CMBs may develop rapidly in patients with acute ischemic stroke and refractory status epilepticus.15,16 Mechanisms related to disruption of the blood-brain barrier, endothelial dysfunction, and sudden elevation of blood pressure may underlie the formation of CMBs after
VOLUME 74 | NUMBER 2 | FEBRUARY 2014 | 179
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
JEON ET AL
TABLE 3. Clinical Findings According to the Presence of Cerebral Microbleedsa CMBs Absent (n = 21) During admission, n (%) Rebleeding of SAH Global cerebral edema Delayed cerebral ischemia Seizure On discharge, n (%) Modified Rankin Scale score 4-6 Cognitive declineb At 3 mo after symptom onset, n (%) Modified Rankin Scale score 4-6 Cognitive declinec
0 10 7 3
(0) (48) (33) (14)
Present (n = 18) 2 5 7 4
(11) (28) (39) (22)
OR (95% CI) 0 0.42 1.27 1.71
(0) (0.11-1.62) (0.34-4.73) (0.33-8.94)
P Value .99 .21 .72 .52
9 (43) 6 (40)
11 (61) 11 (73)
2.10 (0.58-7.56) 5.06 (0.88-19.27)
.26 .07
6 (29) 7 (54)
9 (50) 7 (64)
2.50 (0.67-9.39) 1.50 (0.29-7.75)
.18 .63
a
CI, confidence interval; CMB, cerebral microbleeds; OR, odds ratio; SAH, subarachnoid hemorrhage. Values are number (column %) or median (interquartile range) as appropriate. Telephone Interview for Cognitive Status #30 (data available for 30 patients). c Telephone Interview for Cognitive Status #30 (data available for 24 patients). b
SAH.15,16 Additionally, the development of CMBs may be triggered by regional blood load.
reveal the natural history and potential prognostic or therapeutic implications of CMBs in patients with SAH.
Limitations Our study has some limitations. First, this is a retrospective study without a comparison group, and it has a potential risk of selection bias because only a subpopulation of SAH patients underwent MRI. Therefore, our findings should be interpreted with caution. However, approximately half of our patients had CMBs, whereas the prevalence of CMBs in a similar age group in a healthy population is ,20%, making this difference unlikely to be just a coincidence.17 Second, because of the retrospective design, serial MRIs were not obtained. Thus, it is uncertain when CMBs developed, before or after SAH. The association of CMBs with acute infarcts (DWILs) supports the hypothesis that CMBs may have developed during the active cerebrovascular process rather than preexisting. CMBs have been shown to develop rapidly during the first week in patients with acute ischemic stroke.15 Alternatively, it is possible that patients who had CMBs before SAH are more likely to have acute infarction after SAH. Future prospective studies are needed to understand the natural history of CMBs in patients with SAH. Third, because of the small numbers of patients, comprehensive multivariable analysis could not be performed. Instead, associations between DWILs and CMBs were adjusted for age and hypertension, well-known risk factors of CMBs in other populations.1 This small size may contribute to the lack of statistical power to show the association between CMBs and clinical outcomes.
Disclosure
CONCLUSION CMBs are common findings in patients with acute SAH and are associated with DWILs. Our findings may represent a new mechanism of tissue injury in SAH, presumably relating to an active microangiopathy. Prospective imaging studies are needed to
180 | VOLUME 74 | NUMBER 2 | FEBRUARY 2014
The authors have no personal financial or institutional interest in any of the drugs, materials, or devices described in this article.
REFERENCES 1. Cordonnier C, Al-Shahi Salman R, Wardlaw J. Spontaneous brain microbleeds: systematic review, subgroup analyses and standards for study design and reporting. Brain. 2007;130(pt 8):1988-2003. 2. Choi P, Ren M, Phan TG, et al. Silent infarcts and cerebral microbleeds modify the associations of white matter lesions with gait and postural stability: populationbased study. Stroke. 2012;43(6):1505-1510. 3. Choi-Kwon S, Han K, Choi S, et al. Poststroke depression and emotional incontinence: factors related to acute and subacute stages. Neurology. 2012;78(15):1130-1137. 4. Fan YH, Zhang L, Lam WW, Mok VC, Wong KS. Cerebral microbleeds as a risk factor for subsequent intracerebral hemorrhages among patients with acute ischemic stroke. Stroke. 2003;34(10):2459-2462. 5. Tang WK, Chen YK, Lu J, et al. Cerebral microbleeds and quality of life in acute ischemic stroke. Neurol Sci. 2011;32(3):449-454. 6. Lord AS, Fernandez L, Schmidt JM, et al. Effect of rebleeding on the course and incidence of vasospasm after subarachnoid hemorrhage. Neurology. 2012;78(1):31-37. 7. Frontera JA, Fernandez A, Schmidt JM, et al. Defining vasospasm after subarachnoid hemorrhage: what is the most clinically relevant definition? Stroke. 2009;40(6):1963-1968. 8. Gregoire SM, Chaudhary UJ, Brown MM, et al. The Microbleed Anatomical Rating Scale (MARS): reliability of a tool to map brain microbleeds. Neurology. 2009;73(21):1759-1766. 9. Sato K, Shimizu H, Fujimura M, Inoue T, Matsumoto Y, Tominaga T. Acutestage diffusion-weighted magnetic resonance imaging for predicting outcome of poor-grade aneurysmal subarachnoid hemorrhage. J Cereb Blood Flow Metab. 2010;30(6):1110-1120. 10. Fazekas F, Chawluk JB, Alavi A, Hurtig HI, Zimmerman RA. MR signal abnormalities at 1.5 T in Alzheimer’s dementia and normal aging. AJR Am J Roentgenol. 1987;149(2):351-356. 11. Charidimou A, Werring DJ. Cerebral microbleeds and cognition in cerebrovascular disease: an update. J Neurol Sci. 2012;322(1-2):50-55. 12. Jeon SB, Kang DW, Cho AH, et al. Initial microbleeds at MR imaging can predict recurrent intracerebral hemorrhage. J Neurol. 2007;254(4):508-512.
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
MICROBLEEDS IN SUBARACHNOID HEMORRHAGE
13. van Gijn J, Kerr RS, Rinkel GJ. Subarachnoid haemorrhage. Lancet. 2007;369 (9558):306-318. 14. Kang DW, Han MK, Kim HJ, et al. New ischemic lesions coexisting with acute intracerebral hemorrhage. Neurology. 2012;79(9):848-855. 15. Jeon SB, Kwon SU, Cho AH, Yun SC, Kim JS, Kang DW. Rapid appearance of new cerebral microbleeds after acute ischemic stroke. Neurology. 2009;73(20):1638-1644. 16. Jeon SB, Parikh G, Choi HA, et al. Acute cerebral microbleeds in refractory status epilepticus. Epilepsia. 2013;54(5):e66-e68. 17. Poels MM, Vernooij MW, Ikram MA, et al. Prevalence and risk factors of cerebral microbleeds: an update of the Rotterdam scan study. Stroke. 2010;41(10 suppl): S103-S106.
Supplemental Digital Content is 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 (www.neurosurgery-online.com).
COMMENTS
C
erebral microbleeds (CMBs) detected with blood-sensitive sequences on magnetic resonance imaging have been found to be associated with a wide variety of neurovascular entities like ischemic and hemorrhagic stroke, cerebral amyloid angiopathy, chronic hypertension, and antithrombotic treatment, as well as other conditions like Alzheimer disease and diabetes mellitus.1,2 Even among healthy adults, the prevalence of CMB is about 5% and increases with age.1 The present study is the first to report on the phenomenon of CMB in patients with subarachnoid hemorrhage. In a retrospective series of 39 consecutive patients with acute spontaneous and mainly aneurysmal (77%) subarachnoid hemorrhage who underwent 3-T magnetic resonance imaging soon after their hemorrhage, T2*-weighted gradient-echo imaging revealed CMB in almost half (46%) of the cohort. After controlling for age and hypertension, the detection of CMB was significantly associated with acute cerebral ischemic lesions on diffusion-weighted imaging. The authors speculate that disruption of the blood-brain barrier, endothelial dysfunction, and sudden increases in blood pressure may have contributed to this association. It is debatable whether CMBs directly lead to specific clinical findings, are an indicator of an underlying pathology of clinical significance, or both. As a result of the design of the present study, it remains unclear whether CMB occurred before or after subarachnoid hemorrhage. This, however, might be of interest given the hypothesis that CMB results from some “instability” of the vessel wall.3 Although, of course, different types of arterial vessels are involved, it would be interesting to learn if patients with unruptured aneurysms are more likely to harbor CMBs and whether such a “CMB load” predisposes to future aneurysmal rupture. Hans-Christian Koennecke Berlin, Germany
1. Cordonnier C, Salman RAS, Wardlaw J. Spontaneous brain microbleeds: systematic review, subgroup analyses and standards for study design and reporting. Brain. 2007;130(pt 8):1988-2003.
NEUROSURGERY
2. Koennecke HC. Cerebral microbleeds on MRI: prevalence, associations, and potential clinical implications. Neurology. 2006;66(2):165-171. 3. Roob G, Fazekas F. Magnetic resonance imaging of cerebral microbleeds. Curr Opin Neurol. 2000;13(1):69-73.
I
n this interesting article on the magnetic resonance imaging findings in acute subarachnoid hemorrhage (SAH), the authors included 39 patients and performed both blood-sensitive (T2* gradient-echo imaging) and ischemia-sensitive (diffusion-weighted imaging [DWI]) imaging within 1 week of acute SAH. About three-quarters of the patients (77%) had aneurysmal SAH; the remainder had nonaneurysmal (perimesencephalic) SAH. They found cerebral microbleeds (CMBs) in 18 patients (46%). Microbleeds were related to DWI lesions after adjustment for age and hypertension. In this small sample, CMBs were not related to clinical outcomes, although there were nonsignificant trends for higher dependency and worse cognitive outcomes in patients with CMBs. The authors conclude that CMBs are common in acute SAH, are associated with acute ischemia, and may add to our understanding of the mechanisms of brain injury after SAH. The observation of CMBs in acute SAH has not been previously reported and could represent a new mechanism of tissue injury in SAH, presumably relating to an active microangiopathy. However, there are some weaknesses. First, the sample is very small, and many patients were excluded; those who were included underwent magnetic resonance imaging for clinical reasons (to assess prognosis or to assess brain injury), potentially introducing a selection bias and reducing the generalizability of the findings. Second, it is not known whether these CMBs were preexisting: the prevalence of CMBs in this age group in a healthy population might be up to about 20%, so the prevalence observed is still high, making this unlikely to be a complete explanation. The association with DWI lesions also supports the hypothesis that the CMBs are due to active cerebrovascular process. On the other hand, the authors found no evidence that CMB prevalence was higher at the earlier time points, which does not support a clear relationship to the acute ictus. Future studies with prospective imaging at serial acute time points are needed to see whether CMBs do in fact form in the acute phase of SAH. A comparison group, perhaps patients with suspected SAH but no bleeding shown, or age-matched control subjects would help to determine the specificity of the findings to SAH. The authors speculate that acutely high blood pressure in the context of disordered cerebral autoregulation may have contributed to CMBs, and although there was a trend supporting this, larger samples are needed to properly test this hypothesis. In summary, this study provides new evidence of acute microbleeding, suggestive of an active widespread acute microangiopathic process in SAH, and represents a possible new mechanism of brain injury with potential for contributing to our understanding of the underlying pathophysiology in SAH, as well as in assessing prognosis and monitoring. David Werring London, United Kingdom
VOLUME 74 | NUMBER 2 | FEBRUARY 2014 | 181
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
TOPIC Sang-Beom Jeon, MD, PhD*‡§ Gunjan Parikh, MD‡k H. Alex Choi, MD, MS‡§ Neeraj Badjatia, MD, MS‡k
Cerebral Microbleeds in Patients With Acute Subarachnoid Hemorrhage
Kiwon Lee, MD‡§ J. Michael Schmidt, PhD‡ Hector Lantigua, MD‡ E. Sander Connolly, MD¶ Stephan A. Mayer, MD‡ Jan Claassen, MD, PhD‡ *Department of Neurology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea, Departments of ‡Neurology and ¶Neurosurgery, Columbia University College of Physicians and Surgeons, New York, New York; §Departments of Neurology and Neurosurgery, University of Texas Medical School at Houston, Houston, Texas; kDepartment of Neurology, University of Maryland School of Medicine, Baltimore, Maryland Correspondence: Sang-Beom Jeon, MD, PhD, Department of Neurology, Asan Medical Center, University of Ulsan College of Medicine, 88, Olympic-ro 43-gil, Songpa-gu, Seoul, Republic of Korea. E-mail: [email protected] Received, July 2, 2013. Accepted, October 28, 2013. Published Online, October 30, 2013. Copyright © 2013 by the Congress of Neurological Surgeons
WHAT IS THIS BOX? A QR Code is a matrix barcode readable by QR scanners, mobile phones with cameras, and smartphones. The QR Code above links to Supplemental Digital Content from this article.
BACKGROUND: Cerebral microbleeds (CMBs) are commonly found after stroke but have not previously been studied in patients with subarachnoid hemorrhage (SAH). OBJECTIVE: To study the prevalence, radiographic patterns, predictors, and impact on outcome of CMBs in patients with SAH. METHODS: We analyzed retrospectively 39 consecutive patients who underwent T2*weighted gradient-echo imaging within 7 days after onset of spontaneous SAH. We report the frequency and location of CMBs and show their association with demographics, vascular risk factors, the Hunt-Hess grade, the modified Fisher Scale, the Acute Physiological and Chronic Health Evaluation II, magnetic resonance imaging findings including diffusion-weighted imaging lesions, and laboratory data, as well as data on rebleeding, global cerebral edema, delayed cerebral ischemia, seizures, the Telephone Interview for Cognitive Status, and the modified Rankin Scale. RESULTS: Eighteen patients (46%) had CMBs. Of these patients, 9 had multiple CMBs, and overall a total of 50 CMBs were identified. The most common locations of CMBs were lobar (n = 23), followed by deep (n = 15) and infratentorial (n = 12). After adjustment for age and history of hypertension, CMBs were related to the presence of diffusion-weighted imaging lesions (odds ratio, 5.24; 95% confidence interval, 1.14-24.00; P = .03). Three months after SAH, patients with CMBs had nonsignificantly higher modified Rankin Scale scores (odds ratio, 2.50; 95% confidence interval, 0.67-9.39; P = .18). CONCLUSION: This study suggests that CMBs are commonly observed and associated with diffusion-weighted imaging lesions in patients with SAH. Our findings may represent a new mechanism of tissue injury in SAH. Further studies are needed to investigate the clinical implications of CMBs. KEY WORDS: Magnetic resonance imaging, Microbleeds, Subarachnoid hemorrhage Neurosurgery 74:176–181, 2014
DOI: 10.1227/NEU.0000000000000244
C
erebral microbleeds (CMBs), commonly found in patients with ischemic stroke, intracerebral hemorrhage, and cerebral amyloid angiopathy, are recognized as radiologic surrogates for bleeding-prone microangiopathy.1 The presence of CMBs has not previously been reported in acute spontaneous subarachnoid hemorrhage (SAH). Finding CMBs in patients ABBREVIATIONS: CI, confidence interval; CMB, cerebral microbleed; DWI, diffusion-weighted imaging; DWIL, diffusion-weighted imaging lesion; OR, odds ratio; SAH, subarachnoid hemorrhage Supplemental digital content is 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 (www.neurosurgery-online.com).
176 | VOLUME 74 | NUMBER 2 | FEBRUARY 2014
www.neurosurgery-online.com
with SAH may shed light on the pathomechanism of CMBs and SAH. It may also help explain neurological sequela after SAH because CMBs have been associated with cognitive impairment, depression, gait instability, and poor quality of life in other populations.2-5 Here, we report prevalence, radiographic patterns, predictors, and impact on outcome of CMBs in patients with acute SAH.
PATIENTS AND METHODS This retrospective analysis of a prospectively collected data set considered all patients enrolled in the Columbia University SAH Outcomes Database Project between June 1, 2006, and April 30, 2011.6 We included patients who had a diagnosis of SAH confirmed by computed tomography scan and underwent 3.0-T magnetic resonance imaging (MRI),
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
MICROBLEEDS IN SUBARACHNOID HEMORRHAGE
TABLE 1. Baseline Characteristics According to the Presence of Cerebral Microbleedsa CMBs Absent (n = 21) Demographics Age, y Male sex, n (%) White, n (%) Vascular risk factors, n (%) Hypertension Diabetes mellitus Smoking Alcohol Previous stroke Antithrombotic agents before SAH Initial characteristics Hunt-Hess grade 3 to 5, n (%) Modified Fisher Scale score 3 or 4, n (%) Hijdra Scale score Global cerebral edema, initial, n (%) Aneurysm Presence, n (%) Size, mm Location, anterior (vs posterior), n (%) APACHE II score Systolic blood pressure, mm Hg Diastolic blood pressure, mm Hg Body temperature, F Leukocyte, ·1000/mL Platelet count, ·1000/mL Partial thromboplastin time, s Prothrombin time, INR GFR, mLmin211.73 m22 MRI findings, concurrent with CMB Interval from symptom onset to MRI, d Leukoaraiosis, n (%) DWILs present, n (%) Punctate pattern Territorial pattern Mixed pattern (punctate1territorial)
52 (38-64) 5 (24) 11 (52) 7 2 6 1 1 3
Present (n = 18) 60 (45-66) 6 (33) 8 (44)
(33) (10) (29) (5) (5) (14)
3 2 9 2 0 3
OR (95% CI) 1.03 (0.98-1.07) 1.60 (0.39-6.51) 0.73 (0.21-2.57)
(17) (11) (50) (11) (0) (17)
0.40 1.19 2.50 2.50 0 1.2
(0.09-1.86) (0.15-9.41) (0.67-9.39) (0.21-30.1) (0) (0.21-6.84)
11 14 13.5 7
(52) (67) (5-18) (33)
11 11 15.5 2
(61) (61) (5.5-21.5) (11)
1.43 0.79 1.03 0.25
(0.40-5.12) (0.21-2.92) (0.95-1.11) (0.04-1.41)
16 6.0 14 14 148 83 98.2 11.6 243 25.8 1.03 88
(76) (3.0-7.0) (88) (9-20) (128-180) (66-96) (97.4-98.6) (9.7-18.6) (208-333) (23.8-28.3) (0.96-1.09) (71-98)
14 8.0 12 16 174 91 98.7 10.8 232 26.1 1.02 86
(78) (5.0-10.0) (86) (8-22) (147-180) (80-100) (97.4-99.5) (9.5-14.95) (168-282) (23.7-28.4) (0.95-1.07) (70-110)
1.09 1.11 0.86 1.03 1.01 1.02 1.49 0.90 0.99 0.95 0.09 1.01
(0.25-4.89) (0.92-1.35) (0.10-7.04) (0.94-1.13) (0.99-1.03) (0.99-1.06) (0.78-2.85) (0.77-1.04) (0.99-1.00) (0.80-1.12) (0-139.4) (0.98-1.03)
(1-5) (29) (83) (28) (17) (39)
1.24 2.50 5.50 0.96 4.00 3.82
(0.92-1.68) (0.50-12.47) (1.22-24.81)b (0.24-3.90) (0.38-42.37) (0.81-17.93)
1 3 10 6 1 3
(1-2.5) (14) (48) (29) (5) (14)
2 5 15 5 3 7
a
APACHE II, Acute Physiology and Chronic Health Evaluation II; CI, confidence interval; CMB, cerebral microbleed; DWIL, diffusion-weighted imaging lesion; GFR, glomerular filtration rate; INR, international normalized ratio; MRI, magnetic resonance imaging; OR, odds ratio; SAH, subarachnoid hemorrhage. Values are number (column %) or median (interquartile range) as appropriate. b P , .05.
including T2*-weighted gradient-echo imaging, within 7 days after symptom onset. We excluded patients whose origin of SAH was trauma, arteriovenous malformation, vasculitis, or coagulopathy, as well as those whose imaging quality was poor. This study was approved by our institutional review board. Written informed consents were obtained from the patients or surrogates. We obtained demographics and medical history by interviewing patients or their families. Demographics, vascular risk factors, the Hunt-Hess grade, the Acute Physiologic and Chronic Health Evaluation II, the modified Fisher Scale, the Hijdra Scale, and routine laboratory or physiological data, as well as data on global cerebral edema on initial
NEUROSURGERY
and/or follow-up computed tomography, rebleeding of SAH, delayed cerebral ischemia, and seizures, were assessed as previously described in detail.6 The modified Rankin Scale score and Telephone Interview for Cognitive Status were evaluated at discharge from the intensive care unit and 3 months after SAH.7 Imaging was performed with a 3-T MRI unit (Signa HDxt, GE Healthcare, Milwaukee, Wisconsin) and an 8-channel phased-array head coil. MRIs were ordered to help assess prognosis and to determine the extent of the underlying brain injury. The decision to send a patient to MRI was made by consensus between the neurosurgery and neurocritical care attending physicians. The MRI protocol included T2*-weighted
VOLUME 74 | NUMBER 2 | FEBRUARY 2014 | 177
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
JEON ET AL
TABLE 2. Blood Pressure From Admission to Magnetic Resonance Imaging According to the Presence of Cerebral Microbleedsa CMBs Absent (n=21) Initial SBP, mm Hg Initial DBP, mm Hg Highest SBP, mm Hg Highest DBP, mm Hg DSBP (highest2lowest), mm Hg DDBP (highest2lowest), mm Hg
152.7 81.0 156.7 83.4 41.8 27.0
6 6 6 6 6 6
Present (n=18)
32.7 161.9 6 27.8 24.0 90.6 6 16.9 29.6 162.3 6 25.4 24.3 85.5 6 12.6 26.2 53.9 6 33.8 26.9 31.1 6 15.0
P Value .35 .17 .52 .74 .21 .57
a
CMB, cerebral microbleeds; DBP, diastolic blood pressure; SBP, systolic blood pressure.
gradient-echo imaging and diffusion-weighted imaging (DWI). The gradient-echo imaging parameters were 1400-millisecond repetition time, 40-millisecond echo time, 15 flip angle, 5-mm slice thickness, 5-mm interslice gap, 512 · 512 matrix, and 220-mm field of view. The DWI parameters were 7000-millisecond repetition time, 73-millisecond echo time, 5-mm slice thickness, 5-mm interslice gap, 256 · 256 matrix, 220-mm field of view, and 2 b values of 0 and 1000 s/mm2. MRIs were reviewed jointly by 2 neurologists (S.-B.J. and G.P.), who were blinded to the clinical data. A third investigator (J.C.) was consulted in cases of disagreement. CMBs were defined as unambiguous homogeneous round signal loss lesions with diameters of 2 to 10 mm.8 Hypointense lesions in subarachnoid space, ventricles, intracerebral hematoma, and DWI lesions (DWILs) were not regarded as CMBs. The number of CMBs was counted according to locational classification (lobar, deep, and infratentorial) by the Microbleed Anatomic Rating Scale.5 DWIL was defined as a hyperintense lesion on DWI (b = 1000 s/mm2) with a corresponding signal reduction in the apparent diffusion coefficient map. The presence of DWIL was interpreted in each location (lobar, deep, and infratentorial). The number of DWILs was not counted because the size and shape of DWILs were variable, and it was uncertain whether multiple tiny DWILs were more detrimental than a single large DWIL. Instead, the patterns of DWILs were classified into punctate (,10 mm in diameter) and territorial ($10 mm in diameter) lesions.9 Leukoaraiosis was defined as a Fazekas score $2 in either periventricular or deep white matter on the fluid-attenuated inversion recovery imaging.10 We analyzed continuous and categorical variables using the t and Fisher exact tests and logistic regression analysis, as appropriate. A 2-tailed value of P , .05 was considered significant. SPSS for Windows (version 17.0, SPSS Inc, Chicago, Illinois) was used for all statistical analyses.
RESULTS During the study period, 408 patients with SAH were enrolled in our database. Of those, 39 met inclusion criteria. Compared with the 369 excluded patients, 39 included patients were less likely to have a history of hypertension (see Table, Supplemental Digital Content 1, http://links.lww.com/NEU/A591, which
178 | VOLUME 74 | NUMBER 2 | FEBRUARY 2014
illustrates baseline characteristics according to study inclusion or exclusion). Of the 39 included patients, the median age was 56 years (interquartile range, 44-65 years), and 28 (72%) were female (Table 1). Blood pressure variables are shown in Table 2. The median time interval from symptom onset to the acquisition of MRI was 40 hours (interquartile range, 28-101 hours). In 30 cases (77%), the causes of SAH were ruptured aneurysms, with perimesencephalic hemorrhage being the second most common. CMBs were seen in 18 patients (46%). Of these, 9 patients had multiple ($2) CMBs. DWILs were found in 25 patients (64%; Figure 1). The presence of DWILs was associated with the presence of CMBs (odds ratio [OR], 5.50; 95% confidence interval [CI], 1.22-24.81; P = .03; Table 1). Logistic regression analysis controlling for age and hypertension showed that DWILs were independently associated with CMBs (OR = 5.24; 95% CI, 1.14-24.00; P = .03; and OR, 5.61; 95% CI, 1.21-25.98; P = .03). Of 50 CMBs, 23 were found in the lobar (8 frontal, 4 parietal, 9 temporal, 1 occipital, and 1 insular), 15 in the deep (3 basal ganglia, 7 thalamus, 4 internal capsule, and 1 corpus callosum), and 12 in the infratentorial (2 brainstem and 10 cerebellum) location. Patients with CMBs in the deep location had a larger amount of subarachnoid blood in the basal sylvian fissure and suprasellar cistern (median, 6.0; interquartile range, 2.0-8.0) than those without CMBs in this location (median, 11.5; interquartile range, 8.8-12.0; P = .006; see Table, Supplemental Digital Content 2, http://links.lww.com/NEU/A592, which illustrates the locational relationship between the presence of CMBs and the amount of subarachnoid blood). The patterns of the DWILs were punctate in 11 patients, territorial in 4 patients, and mixed (both punctate and territorial) in 10 patients. These patterns were not associated with the presence of CMBs (punctate, P = .96; territorial, P = .25; and mixed, P = .09). Among 15 patients who had both CMBs and DWILs, locational (lobar, deep, or infratentorial) associations between CMBs and DWILs were not significant. Patients with CMBs were more likely to have a higher modified Rankin Scale score on discharge (OR, 2.10) and at 3 months after SAH (OR, 2.50) and lower Telephone Interview for Cognitive Status score on discharge (OR, 5.06) and at 3 months (OR, 1.50) after SAH, but the associations were not statistically significant (Table 3).
DISCUSSION CMBs were found in approximately half of patients with acute SAH. The most common locations were lobar, followed by deep and infratentorial regions. The presence of CMBs was significantly associated with the occurrence of DWILs. CMBs were not significantly related to acute complications of SAH such as rebleeding, global cerebral edema, delayed cerebral ischemia, or seizures. In patients with CMBs, poor functional outcome measured by modified Rankin Scale score and cognitive impairment measured by Telephone Interview for Cognitive Status
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
MICROBLEEDS IN SUBARACHNOID HEMORRHAGE
FIGURE 1. Examples of cerebral microbleeds (CMBs) and diffusion-weighted imaging lesions (DWILs). A, a previously healthy 66-year-old woman presented with headache and loss of consciousness. Neurological examination and computed tomography (CT) with angiography suggested subarachnoid hemorrhage (SAH; Hunt-Hess grade 5, and modified Fisher Scale score, 4) resulting from a ruptured aneurysm in the anterior cerebral artery. Magnetic resonance imaging (MRI) performed 38 hours after symptom onset revealed multiple CMBs and DWILs. B, a 58-year-old woman with a 30–pack-year smoking history developed a sudden headache and altered mental status. Initial examination and CT with angiography revealed SAH (Hunt-Hess grade 4 and modified Fisher Scale score, 4) caused by a ruptured aneurysm in the right distal internal carotid artery. MRI performed 37 hours after symptom onset showed multiple CMBs and DWILs. White arrows indicate CMBs; black arrows, DWILs.
tended to be high at discharge and at 3 months after SAH, but this association was not statistically significant. In other populations, CMBs have been independently associated with variable neurological deficits: cognitive impairment, depression, gait instability, poor quality of life, and future hemorrhagic risk.1-5,11,12 Given that SAH survivors also have the aforementioned neurological complications, the contribution of CMBs to such morbidities after SAH needs to be further investigated.13 The high (64%) frequency of DWILs in our SAH patients (a median Hunt-Hess grade of 3) is in agreement with previous reports: up to 82% in high-grade SAH.13 DWILs after SAH have been postulated to be acute infarcts caused by vasospasm of large or small vessels, microemboli, global ischemia, cortical spreading
NEUROSURGERY
depression, etc.9 Interestingly, our study parallels a recent report that found that both CMBs and DWILs were common after intracerebral hemorrhage. 14 These reports may suggest some common pathomechanism after hemorrhagic and ischemic stroke. Old age and chronic hypertension are important risk factors for CMBs and leukoaraiosis, which are known imaging markers of chronic microangiopathy and are strongly linked to each other.1,14 However, these factors were not associated with CMBs in our population. Recent studies have shown that CMBs may develop rapidly in patients with acute ischemic stroke and refractory status epilepticus.15,16 Mechanisms related to disruption of the blood-brain barrier, endothelial dysfunction, and sudden elevation of blood pressure may underlie the formation of CMBs after
VOLUME 74 | NUMBER 2 | FEBRUARY 2014 | 179
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
JEON ET AL
TABLE 3. Clinical Findings According to the Presence of Cerebral Microbleedsa CMBs Absent (n = 21) During admission, n (%) Rebleeding of SAH Global cerebral edema Delayed cerebral ischemia Seizure On discharge, n (%) Modified Rankin Scale score 4-6 Cognitive declineb At 3 mo after symptom onset, n (%) Modified Rankin Scale score 4-6 Cognitive declinec
0 10 7 3
(0) (48) (33) (14)
Present (n = 18) 2 5 7 4
(11) (28) (39) (22)
OR (95% CI) 0 0.42 1.27 1.71
(0) (0.11-1.62) (0.34-4.73) (0.33-8.94)
P Value .99 .21 .72 .52
9 (43) 6 (40)
11 (61) 11 (73)
2.10 (0.58-7.56) 5.06 (0.88-19.27)
.26 .07
6 (29) 7 (54)
9 (50) 7 (64)
2.50 (0.67-9.39) 1.50 (0.29-7.75)
.18 .63
a
CI, confidence interval; CMB, cerebral microbleeds; OR, odds ratio; SAH, subarachnoid hemorrhage. Values are number (column %) or median (interquartile range) as appropriate. Telephone Interview for Cognitive Status #30 (data available for 30 patients). c Telephone Interview for Cognitive Status #30 (data available for 24 patients). b
SAH.15,16 Additionally, the development of CMBs may be triggered by regional blood load.
reveal the natural history and potential prognostic or therapeutic implications of CMBs in patients with SAH.
Limitations Our study has some limitations. First, this is a retrospective study without a comparison group, and it has a potential risk of selection bias because only a subpopulation of SAH patients underwent MRI. Therefore, our findings should be interpreted with caution. However, approximately half of our patients had CMBs, whereas the prevalence of CMBs in a similar age group in a healthy population is ,20%, making this difference unlikely to be just a coincidence.17 Second, because of the retrospective design, serial MRIs were not obtained. Thus, it is uncertain when CMBs developed, before or after SAH. The association of CMBs with acute infarcts (DWILs) supports the hypothesis that CMBs may have developed during the active cerebrovascular process rather than preexisting. CMBs have been shown to develop rapidly during the first week in patients with acute ischemic stroke.15 Alternatively, it is possible that patients who had CMBs before SAH are more likely to have acute infarction after SAH. Future prospective studies are needed to understand the natural history of CMBs in patients with SAH. Third, because of the small numbers of patients, comprehensive multivariable analysis could not be performed. Instead, associations between DWILs and CMBs were adjusted for age and hypertension, well-known risk factors of CMBs in other populations.1 This small size may contribute to the lack of statistical power to show the association between CMBs and clinical outcomes.
Disclosure
CONCLUSION CMBs are common findings in patients with acute SAH and are associated with DWILs. Our findings may represent a new mechanism of tissue injury in SAH, presumably relating to an active microangiopathy. Prospective imaging studies are needed to
180 | VOLUME 74 | NUMBER 2 | FEBRUARY 2014
The authors have no personal financial or institutional interest in any of the drugs, materials, or devices described in this article.
REFERENCES 1. Cordonnier C, Al-Shahi Salman R, Wardlaw J. Spontaneous brain microbleeds: systematic review, subgroup analyses and standards for study design and reporting. Brain. 2007;130(pt 8):1988-2003. 2. Choi P, Ren M, Phan TG, et al. Silent infarcts and cerebral microbleeds modify the associations of white matter lesions with gait and postural stability: populationbased study. Stroke. 2012;43(6):1505-1510. 3. Choi-Kwon S, Han K, Choi S, et al. Poststroke depression and emotional incontinence: factors related to acute and subacute stages. Neurology. 2012;78(15):1130-1137. 4. Fan YH, Zhang L, Lam WW, Mok VC, Wong KS. Cerebral microbleeds as a risk factor for subsequent intracerebral hemorrhages among patients with acute ischemic stroke. Stroke. 2003;34(10):2459-2462. 5. Tang WK, Chen YK, Lu J, et al. Cerebral microbleeds and quality of life in acute ischemic stroke. Neurol Sci. 2011;32(3):449-454. 6. Lord AS, Fernandez L, Schmidt JM, et al. Effect of rebleeding on the course and incidence of vasospasm after subarachnoid hemorrhage. Neurology. 2012;78(1):31-37. 7. Frontera JA, Fernandez A, Schmidt JM, et al. Defining vasospasm after subarachnoid hemorrhage: what is the most clinically relevant definition? Stroke. 2009;40(6):1963-1968. 8. Gregoire SM, Chaudhary UJ, Brown MM, et al. The Microbleed Anatomical Rating Scale (MARS): reliability of a tool to map brain microbleeds. Neurology. 2009;73(21):1759-1766. 9. Sato K, Shimizu H, Fujimura M, Inoue T, Matsumoto Y, Tominaga T. Acutestage diffusion-weighted magnetic resonance imaging for predicting outcome of poor-grade aneurysmal subarachnoid hemorrhage. J Cereb Blood Flow Metab. 2010;30(6):1110-1120. 10. Fazekas F, Chawluk JB, Alavi A, Hurtig HI, Zimmerman RA. MR signal abnormalities at 1.5 T in Alzheimer’s dementia and normal aging. AJR Am J Roentgenol. 1987;149(2):351-356. 11. Charidimou A, Werring DJ. Cerebral microbleeds and cognition in cerebrovascular disease: an update. J Neurol Sci. 2012;322(1-2):50-55. 12. Jeon SB, Kang DW, Cho AH, et al. Initial microbleeds at MR imaging can predict recurrent intracerebral hemorrhage. J Neurol. 2007;254(4):508-512.
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
MICROBLEEDS IN SUBARACHNOID HEMORRHAGE
13. van Gijn J, Kerr RS, Rinkel GJ. Subarachnoid haemorrhage. Lancet. 2007;369 (9558):306-318. 14. Kang DW, Han MK, Kim HJ, et al. New ischemic lesions coexisting with acute intracerebral hemorrhage. Neurology. 2012;79(9):848-855. 15. Jeon SB, Kwon SU, Cho AH, Yun SC, Kim JS, Kang DW. Rapid appearance of new cerebral microbleeds after acute ischemic stroke. Neurology. 2009;73(20):1638-1644. 16. Jeon SB, Parikh G, Choi HA, et al. Acute cerebral microbleeds in refractory status epilepticus. Epilepsia. 2013;54(5):e66-e68. 17. Poels MM, Vernooij MW, Ikram MA, et al. Prevalence and risk factors of cerebral microbleeds: an update of the Rotterdam scan study. Stroke. 2010;41(10 suppl): S103-S106.
Supplemental Digital Content is 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 (www.neurosurgery-online.com).
COMMENTS
C
erebral microbleeds (CMBs) detected with blood-sensitive sequences on magnetic resonance imaging have been found to be associated with a wide variety of neurovascular entities like ischemic and hemorrhagic stroke, cerebral amyloid angiopathy, chronic hypertension, and antithrombotic treatment, as well as other conditions like Alzheimer disease and diabetes mellitus.1,2 Even among healthy adults, the prevalence of CMB is about 5% and increases with age.1 The present study is the first to report on the phenomenon of CMB in patients with subarachnoid hemorrhage. In a retrospective series of 39 consecutive patients with acute spontaneous and mainly aneurysmal (77%) subarachnoid hemorrhage who underwent 3-T magnetic resonance imaging soon after their hemorrhage, T2*-weighted gradient-echo imaging revealed CMB in almost half (46%) of the cohort. After controlling for age and hypertension, the detection of CMB was significantly associated with acute cerebral ischemic lesions on diffusion-weighted imaging. The authors speculate that disruption of the blood-brain barrier, endothelial dysfunction, and sudden increases in blood pressure may have contributed to this association. It is debatable whether CMBs directly lead to specific clinical findings, are an indicator of an underlying pathology of clinical significance, or both. As a result of the design of the present study, it remains unclear whether CMB occurred before or after subarachnoid hemorrhage. This, however, might be of interest given the hypothesis that CMB results from some “instability” of the vessel wall.3 Although, of course, different types of arterial vessels are involved, it would be interesting to learn if patients with unruptured aneurysms are more likely to harbor CMBs and whether such a “CMB load” predisposes to future aneurysmal rupture. Hans-Christian Koennecke Berlin, Germany
1. Cordonnier C, Salman RAS, Wardlaw J. Spontaneous brain microbleeds: systematic review, subgroup analyses and standards for study design and reporting. Brain. 2007;130(pt 8):1988-2003.
NEUROSURGERY
2. Koennecke HC. Cerebral microbleeds on MRI: prevalence, associations, and potential clinical implications. Neurology. 2006;66(2):165-171. 3. Roob G, Fazekas F. Magnetic resonance imaging of cerebral microbleeds. Curr Opin Neurol. 2000;13(1):69-73.
I
n this interesting article on the magnetic resonance imaging findings in acute subarachnoid hemorrhage (SAH), the authors included 39 patients and performed both blood-sensitive (T2* gradient-echo imaging) and ischemia-sensitive (diffusion-weighted imaging [DWI]) imaging within 1 week of acute SAH. About three-quarters of the patients (77%) had aneurysmal SAH; the remainder had nonaneurysmal (perimesencephalic) SAH. They found cerebral microbleeds (CMBs) in 18 patients (46%). Microbleeds were related to DWI lesions after adjustment for age and hypertension. In this small sample, CMBs were not related to clinical outcomes, although there were nonsignificant trends for higher dependency and worse cognitive outcomes in patients with CMBs. The authors conclude that CMBs are common in acute SAH, are associated with acute ischemia, and may add to our understanding of the mechanisms of brain injury after SAH. The observation of CMBs in acute SAH has not been previously reported and could represent a new mechanism of tissue injury in SAH, presumably relating to an active microangiopathy. However, there are some weaknesses. First, the sample is very small, and many patients were excluded; those who were included underwent magnetic resonance imaging for clinical reasons (to assess prognosis or to assess brain injury), potentially introducing a selection bias and reducing the generalizability of the findings. Second, it is not known whether these CMBs were preexisting: the prevalence of CMBs in this age group in a healthy population might be up to about 20%, so the prevalence observed is still high, making this unlikely to be a complete explanation. The association with DWI lesions also supports the hypothesis that the CMBs are due to active cerebrovascular process. On the other hand, the authors found no evidence that CMB prevalence was higher at the earlier time points, which does not support a clear relationship to the acute ictus. Future studies with prospective imaging at serial acute time points are needed to see whether CMBs do in fact form in the acute phase of SAH. A comparison group, perhaps patients with suspected SAH but no bleeding shown, or age-matched control subjects would help to determine the specificity of the findings to SAH. The authors speculate that acutely high blood pressure in the context of disordered cerebral autoregulation may have contributed to CMBs, and although there was a trend supporting this, larger samples are needed to properly test this hypothesis. In summary, this study provides new evidence of acute microbleeding, suggestive of an active widespread acute microangiopathic process in SAH, and represents a possible new mechanism of brain injury with potential for contributing to our understanding of the underlying pathophysiology in SAH, as well as in assessing prognosis and monitoring. David Werring London, United Kingdom
VOLUME 74 | NUMBER 2 | FEBRUARY 2014 | 181
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
