- Email: [email protected]
Location of acute brain hemorrhage in patients undergoing antithrombotic therapy
Journal of the Neurological Sciences 280 (2009) 87–89
Contents lists available at ScienceDirect
Journal of the Neurological Sciences j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / j n s
Location of acute brain hemorrhage in patients undergoing antithrombotic therapy Ryo Itabashi ⁎, Masahiro Yasaka, Takahiro Kuwashiro, Hideaki Nakagaki, Fumio Miyashita, Hiroaki Naritomi, Kazuo Minematsu Cerebrovascular Division, Department of Medicine, National Cardiovascular Center, 5-7-1 Fujishirodai, Suita, Osaka, Japan
a r t i c l e
i n f o
Article history: Received 13 December 2008 Received in revised form 5 February 2009 Accepted 6 February 2009 Available online 28 February 2009 Keywords: Antithrombotic agents Brain hemorrhage Hematoma Thalamus Cerebral cortex Cerebellum
a b s t r a c t Introduction: The relationship between antithrombotic therapy and the anatomical location of acute brain hematoma remains disputed. The current study was therefore designed to address this issue. Methods: The medical records and CT images were retrospectively reviewed in 484 consecutive patients with an acute brain hemorrhage (291 men, 193 women; mean age, 67.2 ± 12.3 years) who were admitted to the hospital within 7 days of stroke onset from January 1999 through October 2003. Antithrombotic therapy had been performed in 116 patients (AT Group): warfarin (n = 38), antiplatelet therapy (n = 70), or both (n = 8). The other 368 patients had not received antithrombotic therapy (non-AT Group). The hematoma location was compared among the groups. Results: The location of the hematoma was significantly different between the two groups (p b 0.0001). The following locations were seen more frequently in the AT Group than in the non-AT Group: thalamic hemorrhage (44.8% vs. 30.7%), cerebellar hemorrhage (7.8% vs. 2.7%), and lobar hemorrhage (18.1% vs. 11.4%). The clinical characteristics in patients with thalamic, cerebellar, or lobar hemorrhage were compared with those with putaminal hemorrhage. A multivariate analysis using the logistic regression model showed that antithrombotic therapy was an independent factor for cerebellar hemorrhage (OR 3.66, 95%CI 1.31–10.18), lobar hemorrhage (OR 2.27, 95%CI 1.12–4.57), and thalamic hemorrhage (OR 2.20, 95%CI 1.06–4.54) in comparison to putaminal hemorrhage. Conclusions: It therefore appears that antithrombotic therapy is independently associated with thalamic, cerebellar, and lobar hemorrhage. © 2009 Elsevier B.V. All rights reserved.
1. Introduction The use of oral anticoagulant and antiplatelet therapy is well known to increase the risk of brain hemorrhage [1–3] and it may also cause hematoma enlargement during the acute stage [4–8]. Several studies have discussed the association between antithrombotic therapy and the hematoma location [6,9–16], and previous studies have also reported that antithrombotic therapy is related to cerebellar [6,10], lobar [11–15], or thalamic hemorrhage [16]. Hart et al. reviewed previous reports and concluded that the relative frequency of lobar hemorrhage is similar in anticoagulated and non-anticoagulated patients [9]. Therefore, no consensus regarding the site of bleeding during antithrombotic therapy has been established. The current study was conducted to determine whether antithrombotic therapy was a factor for other sites of bleeding in comparison to putaminal hemorrhage, which is the most common site of bleeding in patients Abbreviations: CT, computed tomography; AT, antithrombotic therapy; DM, diabetes mellitus; BP, blood pressure; OR, odds ratio; CI, confidence interval; PT, prothrombin time; INR, international ratio; MB, microbleeds. ⁎ Corresponding author. Department of Stroke Neurology, Kohnan Hospital, 4-20-1 Nagamachi-minami, Taihaku-ku, Sendai, 982-8523, Japan. Tel.: +81 22 248 2131; fax: +81 22 248 1966. E-mail address: [email protected] (R. Itabashi). 0022-510X/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2009.02.304
with spontaneous brain hemorrhage. A large number of patients with acute brain hemorrhage at a single institute were herein studied. 2. Methods The medical records and computed tomography (CT) images of the brain were retrospectively reviewed in 493 patients with acute brain hemorrhage who were admitted to the stroke care unit within 7 days of stroke onset, from January 1999 through October 2003. This institute is a comprehensive stroke center in the northern part of Osaka prefecture, Japan. The diagnosis of brain hemorrhage was confirmed by the presence of intraparenchymal hematoma demonstrated on a brain CT. Nine patients with hemorrhages secondary to hemorrhagic transformation of ischemic strokes, trauma, brain tumor, intracranial aneurysm, dural arterio-venous fistula, or arterio-venous malformation were excluded. The remaining 484 patients (291 men, 193 women; mean age, 67.2 ± 12.3 years) served as subjects for the current study. The regional ethics and hospital management committees approved the study. Antithrombotic therapy had been performed in 116 patients (AT Group): warfarin (n = 38), antiplatelet therapy (n = 70), or both (n = 8). The other 368 patients had not received antithrombotic therapy (non-AT Group). Antiplatelet therapy included aspirin in 52
88
R. Itabashi et al. / Journal of the Neurological Sciences 280 (2009) 87–89
patients (median 81 mg/day), ticlopidine in 12 (median 200 mg/ day), cilostazol in 1, and aspirin and ticlopidine in 5. The antiplatelet agent in patients receiving both warfarin and antiplatelet therapy was aspirin in three patients, ticlopidine in four, and sarpogrelate in one. The locations of the hematomas were divided into the thalamic, putaminal, cerebellar, brainstem, lobar, multiple type (more than two hematomas detected), and mixed type (located in the deep ganglionic hemorrhage and unable to be differentiated as thalamus or putamen). The patients' backgrounds, including age, gender, history of hypertension, diabetes mellitus (DM), hypercholesterolemia, hypocholesterolemia, cardiac disease (including atrial fibrillation, valve disease, cardiomyopathy, and ischemic heart disease), liver disease (liver cirrhosis, active hepatitis), a past history of symptomatic stroke (ischemic or hemorrhagic), smoking (including past habit of smoking), and drinking habits (currently more than 46 g of alcohol per day) were investigated. Hypertension was defined as systolic blood pressure (BP) ≧ 140 or diastolic BP ≧ 90 mm Hg before stroke onset or taking regular antihypertensive drugs; DM as having a fasting plasma glucose level more than 126 mg/dl, a random plasma glucose level of more than 200 mg/dl, a hemoglobin A1c level more than 6.5%, or taking antidiabetic medication; hypercholesterolemia as having a total cholesterol level more than 220 mg/dl or taking antihyperlipidemic drugs; and hypocholesterolemia as having a total cholesterol level less than 130 mg/dl without any antihyperlipidemic drugs. The data were analyzed using Stat View Version 5.0 statistical software program (SAS Institute Inc., Cary, NC, United States). The ttest and the chi square test were used to compare the clinical characteristics and medical histories between the AT and non-AT Groups. The hematoma location was compared between the two groups by absolute rate difference, relative rate ratio, and post hoc cell contribution, and also performed the chi square test. Differences were considered to be significant at a value of p b 0.05. Univariate analyses were performed to compare the backgrounds, including history of taking antithrombotic agents, among patients indicated to have frequent sites of bleeding in the AT Group and in those with putaminal hemorrhage, which is the most common spontaneous brain hemorrhage. Multivariate analyses with a logistic regression model were used to analyze values of p b 0.10 in univariate analyses. Heart disease and previous strokes were excluded from the multivariate analysis because of the strong correlation between the administration of antithrombotics and previous cardiovascular disease. 3. Results Table 1 shows the clinical characteristics of the 484 patients. No significant differences were observed in the frequency of hypertension, DM, hypercholesterolemia, hypocholesterolemia, liver disease, or Table 1 Clinical characteristics of the study population. Characteristics Age (years) Men (%) Smokers (%) Regular drinkers (%) Hypertension (%) Heart disease (%) Diabetes mellitus (%) Hypercholesterolemia (%) Hypocholesterolemia (%) Liver disease (%) Past history of symptomatic ischemic stroke (%) Past history of symptomatic brain hemorrhage (%)
AT Group (n = 116)
Non-AT Group (n = 368)
71.3±11.0 72.4 61.8 4.5 87.8 52.2 29.3 37.7 5.5 9.0 63.8
65.9 ± 12.5 56.3 44.4 16.5 92.6 8.8 23.8 30.7 3.1 13.4 8.2
10.3
11.8
Based on the Mann–Whitney U-test or the χ2 test.
p value b0.0001 0.019 0.004 0.016 0.238 b0.0001 0.082 0.203 0.579 0.569 b0.0001 0.933
Table 2 Comparison of the hematoma location between the non-AT and AT Groups. Hematoma location
All patients (n=484)
Non-AT Group (n=368)
Relative Absolute AT Post hoc cell rate Group differenceb contributions (n = 116) ratioa
Thalamic 165(34.1) 113(30.7) 52(44.8) Putaminal 171(35.3) 144(39.1) 27(23.3) Cerebellar 19(3.9) 10(2.7) 9(7.8) Brainstem 38(7.9) 36(9.8) 2(1.7) Lobar 63(13.0) 42(11.4) 21(18.1) Mixed type 14(2.9) 11(3.0) 3(2.6) Multiple type 11(2.3) 10(2.7) 1(0.9) Miscellaneous 3(0.6) 2(0.5) 1(0.9)
1.459 0.596 2.889 0.173 1.588 0.867 0.333 1.8
14.1 − 15.8 5.1 − 8.1 6.7 − 0.4 − 1.8 0.4
2.798 − 3.115 2.438 − 2.814 1.867 − 0.226 − 1.169 0.381
a
Relative rate ratio (AT Group/non-AT Group) No. (%). Absolute difference (AT Group–non-AT Group).
b
a past history of symptomatic brain hemorrhage between the AT and non-AT Groups. The patients in the AT Group were older, with a higher proportion of male subjects and a past history of smoking, heart disease, and symptomatic ischemic stroke; they were less likely to be regular drinkers in comparison to those in the non-AT Group. As shown in Table 2, putaminal hemorrhage was the most common hematoma location in all patients; the location was significantly different between the two groups (chi square test, p b 0.0001). The following hemorrhage locations were more frequent in the AT Group than the non-AT Group, with positive values of post hoc cell contribution: thalamic (44.8% vs. 30.7%), cerebellar (7.8% vs. 2.7%), and lobar (18.1% vs. 11.4%). In the AT Group, there was no significant difference between the patients that were treated with warfarin and those with antiplatelet therapy (chi square test, p = 0.2815): thalamic (44.7% vs. 42.9%), putaminal (15.8% vs. 28.6%), cerebellar (5.3% vs. 10%), brainstem (2.6% vs. 1.4%), and lobar (23.7% vs. 17.1%). In patients with both antithrombotics (n = 8), there were five thalamic, one putaminal, and two mixed type hemorrhages. The absolute rate difference was calculated between the AT and non-AT Groups by subtracting the proportion of each site of hematoma in the non-AT Group from that of the AT Group and the relative rate ratio of the AT Group was compared to the non-AT Group by dividing the proportion of the AT Group by that of the non-AT Group (Table 2). In comparison to patients with putaminal hemorrhage, those with thalamic hemorrhage were older (69.2 ± 10.2 vs. 64.2 ± 12.7, p b 0.0001) and more commonly had heart disease (23.2% vs. 13.5%, p = 0.0226), diabetes mellitus (33.3% vs. 20.5%, p = 0.0078), symptomatic ischemic stroke (28.7% vs. 11.2%, p b 0.0001), symptomatic brain hemorrhage (14.6 vs. 5.9, p = 0.0086), and antithrombotic therapy(31.5% vs. 15.8%, p = 0.0007); those with cerebellar hemorrhage were older(72.1 ± 9.0 vs. 64.2± 12.7, p b 0.0097) and more commonly had symptomatic ischemic stroke (26.9% vs. 11.2%, p = 0.0022)and antithrombotic
Table 3 Multiple logistic regression models. Odds ratio
95% Confidence interval
A) Thalamic vs. putaminal hemorrhage Age/10 DM Past history of symptomatic brain hemorrhage Antithrombotics
1.04 2.75 3.69 2.20
1.01–1.07 1.38–5.46 1.38–9.85 1.06–4.54
B) Cerebellar vs. putaminal hemorrhage Age/10 Antithrombotics
1.51 3.66
0.98–2.31 1.31–10.18
C) Lobar vs. putaminal hemorrhage Age/10 DM Past history of symptomatic brain hemorrhage Antithrombotics
1.03 0.42 3.02 2.27
1.01–1.06 0.16–1.09 1.14–7.97 1.12–4.57
R. Itabashi et al. / Journal of the Neurological Sciences 280 (2009) 87–89
89
therapy (47.4% vs. 15.8%, p b 0.0001), and those with lobar hemorrhage were older(70.8 ± 13.8 vs. 64.2 ± 12.7, p = 0.00008), more frequently had heart disease(26.2% vs. 13.5%, p = 0.0231), diabetes mellitus(9.5% vs. 20.5%, p = 0.0508), symptomatic ischemic stroke (27.0% vs. 11.2%, p = 0.0032), symptomatic brain hemorrhage (15.9% vs. 5.9%, p = 0.0163) and antithrombotic therapy (33.3% vs. 15.8%, p b 0.0001). A multivariate logistic regression analysis revealed that antithrombotic therapy was an independent factor for thalamic (OR 2.20, 95%CI 1.06–4.54), cerebellar (OR 3.66, 95%CI 1.31–10.18), and lobar hemorrhage (OR 2.27, 95%CI 1.12–4.57) for backgrounds with univariate analysis of p b 0.1 (Table 3). Heart disease and a past history of stroke were not selected as independent factors for the multivariate analysis because many patients with heart disease or past history of ischemic stroke received antithrombotic therapy (70%). Other than antithrombotic therapy, both age and the occurrence of a symptomatic brain hemorrhage were independent factors for thalamic or lobar hemorrhage, while diabetes mellitus was an independent factor for thalamic hemorrhage.
Several reports have suggested an association between brain hemorrhage associated with AT and microbleeds (MBs) detected on T2⁎-weighted gradient echo magnetic resonance imaging as rounded areas of signal loss. MBs appear to represent cerebral amyloid angiopathy or the vasculopathy that complicates long-standing hypertension [19], and might be a risk factor in brain hemorrhage associated with antiplatelets [20]. Further study is thus required to elucidate the association between MBs and subsequent hemorrhage during anticoagulant, antiplatelet, or both therapies combined. Some limitations associated with this study include: the study was a retrospective study and performed in a single-center; the differences in hemorrhage location might be driven by the indication for antithrombotics rather than antithrombotic use: the dose of antiplatelets and the international normalized ratio were not studied. The proportion of patients treated with anticoagulant was small (only 38 patients). In conclusion, the above findings indicate that antithrombotic therapy may be independently associated with thalamic, cerebellar, and lobar hemorrhage.
4. Discussion
References
A significant difference was observed in the location of hematomas between the AT Group and the non-AT Group, primarily reflecting the higher incidence rate of thalamic, cerebellar, and lobar hematoma among the subjects in the AT Group. A multivariate logistic regression analysis demonstrated antithrombotic therapy to be an independent factor for cerebellar, lobar, and thalamic hemorrhage. Kase et al. first reported a high percentage of cerebellar hemorrhage (37%) among 24 patients with acute brain hemorrhage treated with anticoagulants [10]. Toyoda et al. recently analyzed 327 consecutive patients with intracerebral hemorrhage and found warfarin therapy with PTINR N 2.5 and high blood glucose levels on admission to be independently predictive of cerebellar hemorrhage in comparison to other intracerebral hemorrhage [6]. Although numerous studies report that the proportion of lobar hemorrhage is high among brain hemorrhage patients with AT [11–15], Hart et al. reviewed reports regarding the site of bleeding in anticoagulant-related brain hemorrhage and suggested that the relative frequency of lobar hemorrhage was similar in anticoagulated and non-anticoagulated patients [9]. In Japan, Karibe et al. investigated the location of bleeding in patients with intracerebral hemorrhage and reported a high frequency of thalamic hemorrhage in 6 of 11 patients (55%) with brain hemorrhage during anticoagulant therapy in comparison to 96 of 313 patients (32%) with spontaneous brain hemorrhage (not significant) [16]. This hematoma site predilection is possibly caused by the anatomical factors of the cerebrovascular circulation. Yamaguchi et al. hypothesized a different progression of degeneration associated with hypertension and aging in perforators between the anterior circulation and posterior circulation, because older patients have a higher percentage of thalamic hemorrhage [17]. It is suggested that the sparse innervation of the sympathetic nerves in the posterior circulation might be related to the pathogenesis of reversible posterior leukoencephalopathy syndrome. Although the posterior brain regions might be susceptible to an elevated system pressure in the posterior circulation, the relationship between the cerebrovascular autoregulation system and the vascular fragility associated with antithromboticrelated brain hemorrhage is still poorly understood [18]. The present study demonstrated an older age and a history of symptomatic brain hemorrhage to be independently associated with a high frequency of thalamic and lobar hemorrhage. The high frequency of lobar hemorrhage was attributed to cerebral amyloid angiopathy and thalamic location might be related to vasculopathy that complicates different progressions of hypertensive degeneration [17].
[1] Fogelholm R, Eskola K, Kiminkinen I, Kunnnamo I. Anticoagulant treatment as a risk factor for primary intracerebral hemorrhage. J Neurol Neurosurg Psychiatry 1992;55:1121–4. [2] Franke CL, de Jonge J, van Swieten JC, Op de Coul AAW, van Gijn J. Intracerebral hematomas during anticoagulant treatment. Storke 1990;21:726–30. [3] He J, Whelton PK, Vu B, Klag MJ. Aspirin and risk of hemorrhagic stroke: a metaanalysis of randomized controlled trials. JAMA 1998;280:1930–5. [4] Flibotte JJ, Hagan N, O'Donnell J, Greenberg SM, Rosand J. Warfarin, hematoma expansion, and outcome of intracerebral hemorrhage. Neurology 2004;63:1059–64. [5] Toyoda K, Okada Y, Minematsu K, Kamouchi M, Fujimoto S, Ibayashi S, et al. Antiplatelet therapy contributes to acute deterioration of intracerebral hemorrhage. Neurology 2005;65:1000–4. [6] Toyoda K, Okada Y, Ibayashi S, Inoue T, Yasumori K, Fukui D, et al. Antithrombotics therapy and predilection for cerebellar hemorrhage. Cerebrovasc Dis 2007;23:109–16. [7] Saloheimo P, Ahonen M, Juvela S, Pyhtinen J, Savolainen ER, Hillbom M. Regular aspirin-use preceding the onset of primary intracerebral hemorrhage is an independent predictor for death. Stroke 2006;37:129–33. [8] Yasaka M, Minematsu K, Naritomi H, Sakata T, Yamaguchi T. Predisposing factors for enlargement of intracerebral hemorrhage in patients treated with warfarin. Thromb Haemost 2003;89:278–83. [9] Hart RG, Boop MB, Anderson DC. Oral anticoagulants and intracranial hemorrhage. Facts and hypotheses. Stroke 1995;26:1471–7. [10] Kase CS, Robinson RK, Stein RW, DeWitt LD, Hier DB, Harp DL, et al. Anticoagulantrelated intracerebral hemorrhage. Neurology 1985;35:943–8. [11] Wong KS, Mok V, Lam WWM, Kay R, Tang A, Chan YL, et al. Aspirin-associated intracerebral hemorrhage: clinical and radiologic features. Neurology 2000;54:2298–301. [12] Radberg JA, Olsson JE, Radberg CT. Prognostic parameters in spontaneous intracerebral hematomas with special reference to anticoagulant treatment. Stroke 1991;22:571–6. [13] Mathiesen T, Benediktsdottir K, Johnsson H, Lindqvist M, von Holst H. Intracranial traumatic and non-traumatic haemorrhagic complications of warfarin treatment. Acta Neurol Scand 1995;91:208–14. [14] Gorter JW. for the Stroke prevention in reversible ischemia trial (SPRIT) and European Atrial Fibrillation Trial (EAFT) study groups. Major bleeding during anticoagulation after cerebral ischemia: patterns and risk factors. Neurology 1999;53:1319–27. [15] Sjoblom L, Hardemark HG, Lindgren A, Norrving B, Fahlen M, Samuelsson M, et al. Management and prognostic features of intracerebral hemorrhage during anticoagulant therapy: a Swedish Multicenter Study. Stroke 2001;32:2567–74. [16] Karibe H, Shimizu H, Shamoto H, Takahashi T, Tominaga T. Intracerebral hemorrhage under anticoagulant and/or antiplatelet therapy. Surg Cereb Stroke 2004;32:398–402. [17] Yamaguchi T, Miyashita T, Minematsu K, Yamaguchi S, Moriyasu H. Incidence of thalamic infarction and hemorrhage, and their distribution in the thalamus (in Japanese with English abstract). Jpn J Stroke 1987;9:513–8. [18] Schwartz RB, Feske SK, Polak JF, DeGirolami U, Iaia A, Beckner KM, Rravo SM, et al. Preeclampsia–eclampsia: clinical and neuroradiographic correlates and insights into the pathogenesis of hypertensive encephalopathy. Radiology 2000;217:371–6. [19] Fazekas F, Kleinert R, Roob G, Kleinert G, Kapeller P, Schmidt R, 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 1999;20:637–42. [20] Wong KS, Chan YL, Liu JY, Gao S, Lam WWM. Asymptomatic microbleeds as a risk factor for aspirin-associated intracerebral hemorrhages. Neurology 2003;60:511–3.
Contents lists available at ScienceDirect
Journal of the Neurological Sciences j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / j n s
Location of acute brain hemorrhage in patients undergoing antithrombotic therapy Ryo Itabashi ⁎, Masahiro Yasaka, Takahiro Kuwashiro, Hideaki Nakagaki, Fumio Miyashita, Hiroaki Naritomi, Kazuo Minematsu Cerebrovascular Division, Department of Medicine, National Cardiovascular Center, 5-7-1 Fujishirodai, Suita, Osaka, Japan
a r t i c l e
i n f o
Article history: Received 13 December 2008 Received in revised form 5 February 2009 Accepted 6 February 2009 Available online 28 February 2009 Keywords: Antithrombotic agents Brain hemorrhage Hematoma Thalamus Cerebral cortex Cerebellum
a b s t r a c t Introduction: The relationship between antithrombotic therapy and the anatomical location of acute brain hematoma remains disputed. The current study was therefore designed to address this issue. Methods: The medical records and CT images were retrospectively reviewed in 484 consecutive patients with an acute brain hemorrhage (291 men, 193 women; mean age, 67.2 ± 12.3 years) who were admitted to the hospital within 7 days of stroke onset from January 1999 through October 2003. Antithrombotic therapy had been performed in 116 patients (AT Group): warfarin (n = 38), antiplatelet therapy (n = 70), or both (n = 8). The other 368 patients had not received antithrombotic therapy (non-AT Group). The hematoma location was compared among the groups. Results: The location of the hematoma was significantly different between the two groups (p b 0.0001). The following locations were seen more frequently in the AT Group than in the non-AT Group: thalamic hemorrhage (44.8% vs. 30.7%), cerebellar hemorrhage (7.8% vs. 2.7%), and lobar hemorrhage (18.1% vs. 11.4%). The clinical characteristics in patients with thalamic, cerebellar, or lobar hemorrhage were compared with those with putaminal hemorrhage. A multivariate analysis using the logistic regression model showed that antithrombotic therapy was an independent factor for cerebellar hemorrhage (OR 3.66, 95%CI 1.31–10.18), lobar hemorrhage (OR 2.27, 95%CI 1.12–4.57), and thalamic hemorrhage (OR 2.20, 95%CI 1.06–4.54) in comparison to putaminal hemorrhage. Conclusions: It therefore appears that antithrombotic therapy is independently associated with thalamic, cerebellar, and lobar hemorrhage. © 2009 Elsevier B.V. All rights reserved.
1. Introduction The use of oral anticoagulant and antiplatelet therapy is well known to increase the risk of brain hemorrhage [1–3] and it may also cause hematoma enlargement during the acute stage [4–8]. Several studies have discussed the association between antithrombotic therapy and the hematoma location [6,9–16], and previous studies have also reported that antithrombotic therapy is related to cerebellar [6,10], lobar [11–15], or thalamic hemorrhage [16]. Hart et al. reviewed previous reports and concluded that the relative frequency of lobar hemorrhage is similar in anticoagulated and non-anticoagulated patients [9]. Therefore, no consensus regarding the site of bleeding during antithrombotic therapy has been established. The current study was conducted to determine whether antithrombotic therapy was a factor for other sites of bleeding in comparison to putaminal hemorrhage, which is the most common site of bleeding in patients Abbreviations: CT, computed tomography; AT, antithrombotic therapy; DM, diabetes mellitus; BP, blood pressure; OR, odds ratio; CI, confidence interval; PT, prothrombin time; INR, international ratio; MB, microbleeds. ⁎ Corresponding author. Department of Stroke Neurology, Kohnan Hospital, 4-20-1 Nagamachi-minami, Taihaku-ku, Sendai, 982-8523, Japan. Tel.: +81 22 248 2131; fax: +81 22 248 1966. E-mail address: [email protected] (R. Itabashi). 0022-510X/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2009.02.304
with spontaneous brain hemorrhage. A large number of patients with acute brain hemorrhage at a single institute were herein studied. 2. Methods The medical records and computed tomography (CT) images of the brain were retrospectively reviewed in 493 patients with acute brain hemorrhage who were admitted to the stroke care unit within 7 days of stroke onset, from January 1999 through October 2003. This institute is a comprehensive stroke center in the northern part of Osaka prefecture, Japan. The diagnosis of brain hemorrhage was confirmed by the presence of intraparenchymal hematoma demonstrated on a brain CT. Nine patients with hemorrhages secondary to hemorrhagic transformation of ischemic strokes, trauma, brain tumor, intracranial aneurysm, dural arterio-venous fistula, or arterio-venous malformation were excluded. The remaining 484 patients (291 men, 193 women; mean age, 67.2 ± 12.3 years) served as subjects for the current study. The regional ethics and hospital management committees approved the study. Antithrombotic therapy had been performed in 116 patients (AT Group): warfarin (n = 38), antiplatelet therapy (n = 70), or both (n = 8). The other 368 patients had not received antithrombotic therapy (non-AT Group). Antiplatelet therapy included aspirin in 52
88
R. Itabashi et al. / Journal of the Neurological Sciences 280 (2009) 87–89
patients (median 81 mg/day), ticlopidine in 12 (median 200 mg/ day), cilostazol in 1, and aspirin and ticlopidine in 5. The antiplatelet agent in patients receiving both warfarin and antiplatelet therapy was aspirin in three patients, ticlopidine in four, and sarpogrelate in one. The locations of the hematomas were divided into the thalamic, putaminal, cerebellar, brainstem, lobar, multiple type (more than two hematomas detected), and mixed type (located in the deep ganglionic hemorrhage and unable to be differentiated as thalamus or putamen). The patients' backgrounds, including age, gender, history of hypertension, diabetes mellitus (DM), hypercholesterolemia, hypocholesterolemia, cardiac disease (including atrial fibrillation, valve disease, cardiomyopathy, and ischemic heart disease), liver disease (liver cirrhosis, active hepatitis), a past history of symptomatic stroke (ischemic or hemorrhagic), smoking (including past habit of smoking), and drinking habits (currently more than 46 g of alcohol per day) were investigated. Hypertension was defined as systolic blood pressure (BP) ≧ 140 or diastolic BP ≧ 90 mm Hg before stroke onset or taking regular antihypertensive drugs; DM as having a fasting plasma glucose level more than 126 mg/dl, a random plasma glucose level of more than 200 mg/dl, a hemoglobin A1c level more than 6.5%, or taking antidiabetic medication; hypercholesterolemia as having a total cholesterol level more than 220 mg/dl or taking antihyperlipidemic drugs; and hypocholesterolemia as having a total cholesterol level less than 130 mg/dl without any antihyperlipidemic drugs. The data were analyzed using Stat View Version 5.0 statistical software program (SAS Institute Inc., Cary, NC, United States). The ttest and the chi square test were used to compare the clinical characteristics and medical histories between the AT and non-AT Groups. The hematoma location was compared between the two groups by absolute rate difference, relative rate ratio, and post hoc cell contribution, and also performed the chi square test. Differences were considered to be significant at a value of p b 0.05. Univariate analyses were performed to compare the backgrounds, including history of taking antithrombotic agents, among patients indicated to have frequent sites of bleeding in the AT Group and in those with putaminal hemorrhage, which is the most common spontaneous brain hemorrhage. Multivariate analyses with a logistic regression model were used to analyze values of p b 0.10 in univariate analyses. Heart disease and previous strokes were excluded from the multivariate analysis because of the strong correlation between the administration of antithrombotics and previous cardiovascular disease. 3. Results Table 1 shows the clinical characteristics of the 484 patients. No significant differences were observed in the frequency of hypertension, DM, hypercholesterolemia, hypocholesterolemia, liver disease, or Table 1 Clinical characteristics of the study population. Characteristics Age (years) Men (%) Smokers (%) Regular drinkers (%) Hypertension (%) Heart disease (%) Diabetes mellitus (%) Hypercholesterolemia (%) Hypocholesterolemia (%) Liver disease (%) Past history of symptomatic ischemic stroke (%) Past history of symptomatic brain hemorrhage (%)
AT Group (n = 116)
Non-AT Group (n = 368)
71.3±11.0 72.4 61.8 4.5 87.8 52.2 29.3 37.7 5.5 9.0 63.8
65.9 ± 12.5 56.3 44.4 16.5 92.6 8.8 23.8 30.7 3.1 13.4 8.2
10.3
11.8
Based on the Mann–Whitney U-test or the χ2 test.
p value b0.0001 0.019 0.004 0.016 0.238 b0.0001 0.082 0.203 0.579 0.569 b0.0001 0.933
Table 2 Comparison of the hematoma location between the non-AT and AT Groups. Hematoma location
All patients (n=484)
Non-AT Group (n=368)
Relative Absolute AT Post hoc cell rate Group differenceb contributions (n = 116) ratioa
Thalamic 165(34.1) 113(30.7) 52(44.8) Putaminal 171(35.3) 144(39.1) 27(23.3) Cerebellar 19(3.9) 10(2.7) 9(7.8) Brainstem 38(7.9) 36(9.8) 2(1.7) Lobar 63(13.0) 42(11.4) 21(18.1) Mixed type 14(2.9) 11(3.0) 3(2.6) Multiple type 11(2.3) 10(2.7) 1(0.9) Miscellaneous 3(0.6) 2(0.5) 1(0.9)
1.459 0.596 2.889 0.173 1.588 0.867 0.333 1.8
14.1 − 15.8 5.1 − 8.1 6.7 − 0.4 − 1.8 0.4
2.798 − 3.115 2.438 − 2.814 1.867 − 0.226 − 1.169 0.381
a
Relative rate ratio (AT Group/non-AT Group) No. (%). Absolute difference (AT Group–non-AT Group).
b
a past history of symptomatic brain hemorrhage between the AT and non-AT Groups. The patients in the AT Group were older, with a higher proportion of male subjects and a past history of smoking, heart disease, and symptomatic ischemic stroke; they were less likely to be regular drinkers in comparison to those in the non-AT Group. As shown in Table 2, putaminal hemorrhage was the most common hematoma location in all patients; the location was significantly different between the two groups (chi square test, p b 0.0001). The following hemorrhage locations were more frequent in the AT Group than the non-AT Group, with positive values of post hoc cell contribution: thalamic (44.8% vs. 30.7%), cerebellar (7.8% vs. 2.7%), and lobar (18.1% vs. 11.4%). In the AT Group, there was no significant difference between the patients that were treated with warfarin and those with antiplatelet therapy (chi square test, p = 0.2815): thalamic (44.7% vs. 42.9%), putaminal (15.8% vs. 28.6%), cerebellar (5.3% vs. 10%), brainstem (2.6% vs. 1.4%), and lobar (23.7% vs. 17.1%). In patients with both antithrombotics (n = 8), there were five thalamic, one putaminal, and two mixed type hemorrhages. The absolute rate difference was calculated between the AT and non-AT Groups by subtracting the proportion of each site of hematoma in the non-AT Group from that of the AT Group and the relative rate ratio of the AT Group was compared to the non-AT Group by dividing the proportion of the AT Group by that of the non-AT Group (Table 2). In comparison to patients with putaminal hemorrhage, those with thalamic hemorrhage were older (69.2 ± 10.2 vs. 64.2 ± 12.7, p b 0.0001) and more commonly had heart disease (23.2% vs. 13.5%, p = 0.0226), diabetes mellitus (33.3% vs. 20.5%, p = 0.0078), symptomatic ischemic stroke (28.7% vs. 11.2%, p b 0.0001), symptomatic brain hemorrhage (14.6 vs. 5.9, p = 0.0086), and antithrombotic therapy(31.5% vs. 15.8%, p = 0.0007); those with cerebellar hemorrhage were older(72.1 ± 9.0 vs. 64.2± 12.7, p b 0.0097) and more commonly had symptomatic ischemic stroke (26.9% vs. 11.2%, p = 0.0022)and antithrombotic
Table 3 Multiple logistic regression models. Odds ratio
95% Confidence interval
A) Thalamic vs. putaminal hemorrhage Age/10 DM Past history of symptomatic brain hemorrhage Antithrombotics
1.04 2.75 3.69 2.20
1.01–1.07 1.38–5.46 1.38–9.85 1.06–4.54
B) Cerebellar vs. putaminal hemorrhage Age/10 Antithrombotics
1.51 3.66
0.98–2.31 1.31–10.18
C) Lobar vs. putaminal hemorrhage Age/10 DM Past history of symptomatic brain hemorrhage Antithrombotics
1.03 0.42 3.02 2.27
1.01–1.06 0.16–1.09 1.14–7.97 1.12–4.57
R. Itabashi et al. / Journal of the Neurological Sciences 280 (2009) 87–89
89
therapy (47.4% vs. 15.8%, p b 0.0001), and those with lobar hemorrhage were older(70.8 ± 13.8 vs. 64.2 ± 12.7, p = 0.00008), more frequently had heart disease(26.2% vs. 13.5%, p = 0.0231), diabetes mellitus(9.5% vs. 20.5%, p = 0.0508), symptomatic ischemic stroke (27.0% vs. 11.2%, p = 0.0032), symptomatic brain hemorrhage (15.9% vs. 5.9%, p = 0.0163) and antithrombotic therapy (33.3% vs. 15.8%, p b 0.0001). A multivariate logistic regression analysis revealed that antithrombotic therapy was an independent factor for thalamic (OR 2.20, 95%CI 1.06–4.54), cerebellar (OR 3.66, 95%CI 1.31–10.18), and lobar hemorrhage (OR 2.27, 95%CI 1.12–4.57) for backgrounds with univariate analysis of p b 0.1 (Table 3). Heart disease and a past history of stroke were not selected as independent factors for the multivariate analysis because many patients with heart disease or past history of ischemic stroke received antithrombotic therapy (70%). Other than antithrombotic therapy, both age and the occurrence of a symptomatic brain hemorrhage were independent factors for thalamic or lobar hemorrhage, while diabetes mellitus was an independent factor for thalamic hemorrhage.
Several reports have suggested an association between brain hemorrhage associated with AT and microbleeds (MBs) detected on T2⁎-weighted gradient echo magnetic resonance imaging as rounded areas of signal loss. MBs appear to represent cerebral amyloid angiopathy or the vasculopathy that complicates long-standing hypertension [19], and might be a risk factor in brain hemorrhage associated with antiplatelets [20]. Further study is thus required to elucidate the association between MBs and subsequent hemorrhage during anticoagulant, antiplatelet, or both therapies combined. Some limitations associated with this study include: the study was a retrospective study and performed in a single-center; the differences in hemorrhage location might be driven by the indication for antithrombotics rather than antithrombotic use: the dose of antiplatelets and the international normalized ratio were not studied. The proportion of patients treated with anticoagulant was small (only 38 patients). In conclusion, the above findings indicate that antithrombotic therapy may be independently associated with thalamic, cerebellar, and lobar hemorrhage.
4. Discussion
References
A significant difference was observed in the location of hematomas between the AT Group and the non-AT Group, primarily reflecting the higher incidence rate of thalamic, cerebellar, and lobar hematoma among the subjects in the AT Group. A multivariate logistic regression analysis demonstrated antithrombotic therapy to be an independent factor for cerebellar, lobar, and thalamic hemorrhage. Kase et al. first reported a high percentage of cerebellar hemorrhage (37%) among 24 patients with acute brain hemorrhage treated with anticoagulants [10]. Toyoda et al. recently analyzed 327 consecutive patients with intracerebral hemorrhage and found warfarin therapy with PTINR N 2.5 and high blood glucose levels on admission to be independently predictive of cerebellar hemorrhage in comparison to other intracerebral hemorrhage [6]. Although numerous studies report that the proportion of lobar hemorrhage is high among brain hemorrhage patients with AT [11–15], Hart et al. reviewed reports regarding the site of bleeding in anticoagulant-related brain hemorrhage and suggested that the relative frequency of lobar hemorrhage was similar in anticoagulated and non-anticoagulated patients [9]. In Japan, Karibe et al. investigated the location of bleeding in patients with intracerebral hemorrhage and reported a high frequency of thalamic hemorrhage in 6 of 11 patients (55%) with brain hemorrhage during anticoagulant therapy in comparison to 96 of 313 patients (32%) with spontaneous brain hemorrhage (not significant) [16]. This hematoma site predilection is possibly caused by the anatomical factors of the cerebrovascular circulation. Yamaguchi et al. hypothesized a different progression of degeneration associated with hypertension and aging in perforators between the anterior circulation and posterior circulation, because older patients have a higher percentage of thalamic hemorrhage [17]. It is suggested that the sparse innervation of the sympathetic nerves in the posterior circulation might be related to the pathogenesis of reversible posterior leukoencephalopathy syndrome. Although the posterior brain regions might be susceptible to an elevated system pressure in the posterior circulation, the relationship between the cerebrovascular autoregulation system and the vascular fragility associated with antithromboticrelated brain hemorrhage is still poorly understood [18]. The present study demonstrated an older age and a history of symptomatic brain hemorrhage to be independently associated with a high frequency of thalamic and lobar hemorrhage. The high frequency of lobar hemorrhage was attributed to cerebral amyloid angiopathy and thalamic location might be related to vasculopathy that complicates different progressions of hypertensive degeneration [17].
[1] Fogelholm R, Eskola K, Kiminkinen I, Kunnnamo I. Anticoagulant treatment as a risk factor for primary intracerebral hemorrhage. J Neurol Neurosurg Psychiatry 1992;55:1121–4. [2] Franke CL, de Jonge J, van Swieten JC, Op de Coul AAW, van Gijn J. Intracerebral hematomas during anticoagulant treatment. Storke 1990;21:726–30. [3] He J, Whelton PK, Vu B, Klag MJ. Aspirin and risk of hemorrhagic stroke: a metaanalysis of randomized controlled trials. JAMA 1998;280:1930–5. [4] Flibotte JJ, Hagan N, O'Donnell J, Greenberg SM, Rosand J. Warfarin, hematoma expansion, and outcome of intracerebral hemorrhage. Neurology 2004;63:1059–64. [5] Toyoda K, Okada Y, Minematsu K, Kamouchi M, Fujimoto S, Ibayashi S, et al. Antiplatelet therapy contributes to acute deterioration of intracerebral hemorrhage. Neurology 2005;65:1000–4. [6] Toyoda K, Okada Y, Ibayashi S, Inoue T, Yasumori K, Fukui D, et al. Antithrombotics therapy and predilection for cerebellar hemorrhage. Cerebrovasc Dis 2007;23:109–16. [7] Saloheimo P, Ahonen M, Juvela S, Pyhtinen J, Savolainen ER, Hillbom M. Regular aspirin-use preceding the onset of primary intracerebral hemorrhage is an independent predictor for death. Stroke 2006;37:129–33. [8] Yasaka M, Minematsu K, Naritomi H, Sakata T, Yamaguchi T. Predisposing factors for enlargement of intracerebral hemorrhage in patients treated with warfarin. Thromb Haemost 2003;89:278–83. [9] Hart RG, Boop MB, Anderson DC. Oral anticoagulants and intracranial hemorrhage. Facts and hypotheses. Stroke 1995;26:1471–7. [10] Kase CS, Robinson RK, Stein RW, DeWitt LD, Hier DB, Harp DL, et al. Anticoagulantrelated intracerebral hemorrhage. Neurology 1985;35:943–8. [11] Wong KS, Mok V, Lam WWM, Kay R, Tang A, Chan YL, et al. Aspirin-associated intracerebral hemorrhage: clinical and radiologic features. Neurology 2000;54:2298–301. [12] Radberg JA, Olsson JE, Radberg CT. Prognostic parameters in spontaneous intracerebral hematomas with special reference to anticoagulant treatment. Stroke 1991;22:571–6. [13] Mathiesen T, Benediktsdottir K, Johnsson H, Lindqvist M, von Holst H. Intracranial traumatic and non-traumatic haemorrhagic complications of warfarin treatment. Acta Neurol Scand 1995;91:208–14. [14] Gorter JW. for the Stroke prevention in reversible ischemia trial (SPRIT) and European Atrial Fibrillation Trial (EAFT) study groups. Major bleeding during anticoagulation after cerebral ischemia: patterns and risk factors. Neurology 1999;53:1319–27. [15] Sjoblom L, Hardemark HG, Lindgren A, Norrving B, Fahlen M, Samuelsson M, et al. Management and prognostic features of intracerebral hemorrhage during anticoagulant therapy: a Swedish Multicenter Study. Stroke 2001;32:2567–74. [16] Karibe H, Shimizu H, Shamoto H, Takahashi T, Tominaga T. Intracerebral hemorrhage under anticoagulant and/or antiplatelet therapy. Surg Cereb Stroke 2004;32:398–402. [17] Yamaguchi T, Miyashita T, Minematsu K, Yamaguchi S, Moriyasu H. Incidence of thalamic infarction and hemorrhage, and their distribution in the thalamus (in Japanese with English abstract). Jpn J Stroke 1987;9:513–8. [18] Schwartz RB, Feske SK, Polak JF, DeGirolami U, Iaia A, Beckner KM, Rravo SM, et al. Preeclampsia–eclampsia: clinical and neuroradiographic correlates and insights into the pathogenesis of hypertensive encephalopathy. Radiology 2000;217:371–6. [19] Fazekas F, Kleinert R, Roob G, Kleinert G, Kapeller P, Schmidt R, 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 1999;20:637–42. [20] Wong KS, Chan YL, Liu JY, Gao S, Lam WWM. Asymptomatic microbleeds as a risk factor for aspirin-associated intracerebral hemorrhages. Neurology 2003;60:511–3.