Predictors of Outcome in Patients With Cerebral Venous Thrombosis and Intracerebral Hemorrhage
Background and Purpose— Although intracerebral hemorrhages are frequent in patients with cerebral venous thrombosis, and lead to worse outcome, predictors of outcome in cerebral venous thrombosis patients with intracerebral hemorrhages have never been evaluated in adequately powered studies.
Methods— This study was conducted as a part of the International Study on Cerebral Vein and Dural Sinus Thrombosis. We evaluated predictors of outcome in cerebral venous thrombosis patients who had an “early intracerebral hemorrhage,” ie, intracerebral hemorrhages already present at time of diagnosis of cerebral venous thrombosis by a logistic regression analysis, with a modified Rankin scale 3 to 6 at month 6 as dependent variable. The same analysis was performed with “delayed intracerebral hemorrhages,” ie, intracerebral hemorrhages that occurred after the diagnosis of cerebral venous thrombosis, as dependent variable.
Results— Of 624 patients recruited in International Study on Cerebral Vein and Dural Sinus Thrombosis, 245 (39%) had an early intracerebral hemorrhage: at month 6, 51 (21%) of them had a modified Rankin Scale 3 to 6. Independent predictors of having modified Rankin scale 3 to 6 at month 6 were older age (adjusted odds ratio for 1-year increase in age, 1.05; 95% CI, 1.02 to 1.08); male gender (adjusted odds ratio, 3.25; 95% CI, 1.29 to 8.16); having a deep cerebral venous system thrombosis (adjusted odds ratio, 5.43; 95% CI, 1.67 to 17.61) or a right lateral sinus thrombosis (adjusted odds ratio, 2.56; 95% CI, 1.03 to 6.40); and having a motor deficit (adjusted odds ratio, 2.94; 95% CI, 1.21 to 7.10). Of the 36 patients who had a delayed intracerebral hemorrhage, those who had a modified Rankin scale 3 to 6 at month 6 were less likely to have received heparin at the acute stage, and more likely to have had early intracerebral hemorrhage.
Conclusion— Among patients with early intracerebral hemorrhage, those who were older, men, had a thrombosis of the deep cerebral venous system or of the right lateral sinus, and a motor deficit were at higher risk for death or dependency at month 6. This subgroup of patients with predictors of poor outcome can be the target for new therapeutic strategies.
Intracerebral hemorrhages (ICH) occur in approximately one-third of patients with cerebral venous thrombosis (CVT), and are usually associated with a more severe clinical presentation at onset1–3 and a worse outcome.4 However, predictors of outcome in CVT patients with ICH have never been evaluated in properly powered studies.
The aim of this study was to identify predictors of outcome in patients with ICH who were included in the International Study on Cerebral Vein and Dural Sinus Thrombosis (ISCVT).
Materials and Methods
The design of ISCVT, the list of participating centers and investigators, and the main results, have already been reported.4
Organization of the Study and Case Ascertainment
ISCVT was a prospective multinational observational study that included consecutive patients with symptomatic CVT. All investigators provided data on all consecutive cases diagnosed in their institutions during the study period, and performed at least a 6-month follow-up assessment. Case report forms with inclusion and follow-up data were centralized in the coordinating center in Lisbon. Patients were recruited over a 3-year period (May 1998 to May 2001).
Evaluation at Baseline
The diagnosis of CVT had to be confirmed by conventional angiography, CT venography, and MRI including venous sequences, surgery, or autopsy, following established diagnostic criteria.5
We recorded the following information: demographic data (age, gender); date of symptom onset; date of admission; date of confirmation of the diagnosis of CVT; symptoms and signs from clinical onset to diagnosis; Glasgow coma score at admission; imaging methods used (CT, MRI, angiography); location of the thrombus and number of sinus occluded; location and size. We defined: (1) early ICH (E-ICH) as any ICH present on CT or MRI scan at time of diagnosis of CVT; and (2) delayed ICH (D-ICH) as any ICH that was not present on CT or MRI scan at time of diagnosis but occurred later, with or without clinical worsening. D-ICH included E-ICH with a new ICH located elsewhere, or a “de novo D-ICH,” which was a new ICH that occurred in a patient without ICH on the first CT or MRI scan. Hemorrhagic transformation within an infarct was considered as an ICH.
The clinical presentations at time of diagnosis of CVT were divided into isolated intracranial hypertension, and any other clinical presentation. The presence of clinical worsening during hospitalization was recorded, together with its date and mechanisms. We also recorded occurrence of worsening of a previous focal deficit, new focal deficit, new seizures, and visual loss. We also recorded the results of brain CT or MRI scan when repeated.
A list of potential risk factors for CVT was attached to the inclusion form to assist investigators with the diagnostic work-up. We pooled the different causes and predisposing conditions together, as detailed in the appendix.
The choice of the treatment was left to the opinion of the treating physician, but all treatments and date of onset were systematically recorded. We studied the influence of any treatment that could cause bleeding: (1) therapeutic heparin when patients were treated by heparin within the first 24 hours after the diagnosis of CVT and used at therapeutic dosages, including intravenous or subcutaneous nonfractionated heparin and low-molecular-weight heparin; (2) any treatment by heparin whatever the dosage and the day of introduction; (3) low-molecular-weight heparin used at preventive dosage; and (4) antiplatelet drugs.
Follow-Up and Outcome
Follow-up visits were performed at month 6, then at month 12, and yearly thereafter. For the purpose of this study, the 6-month follow-up was taken into account. When a visit with the investigator was not possible, it was replaced by a telephone interview with a relative or the general practitioner. For patients who were lost to follow-up, the condition on the day of discharge was regarded as the final follow-up. The outcome at month 6 was assessed with the modified Rankin Scale (mRS),6 with patients with mRS 0 to 2 being classified as independent survivors, and patients with mRS scores 3 to 6 being classified as dependent or dead. In patients who had a telephone follow-up, the mRS score was assessed by three previously validated questions.7 In patients who missed the 6-month evaluation but had the 1year follow-up visit, we adopted the “worst mRS” scenario: we used either the mRS score at discharge, or at 1-year follow-up, whichever was worst, to estimate disability at month 6. The study was performed according to the Helsinki declaration and the local legislation in each country.
The first step of the statistical analysis was performed in the whole ISCVT database. It consisted of a bivariate comparison between patients with and without E-ICH, using t test, χ2 test with Yates correction, or Fisher exact test when appropriate.
The second step of the statistical analysis was performed only in the subgroup of patients with E-ICH. It consisted of a bivariate comparison of baseline characteristics between patients with mRS 0 to 2 and patients with mRS 3 to 6 at month 6, using t test, χ2 test, with Yates correction, or Fisher exact test when appropriate.
The third step of the statistical analysis, performed in the same group of patients with E-ICH, consisted of a logistic regression analysis with a stepwise procedure, with death or dependency (mRS score 3 to 6) at month 6 as dependent variable (scored 1 when present, and 0 when absent). We included in the model all variables associated with P<0.15 in the bivariate analysis, after having checked for the absence of colinearity within variables.8,9 We excluded variables with >10% missing data, variables with frequency <5% and treatments except those that can cause bleeding, and we chose heparin use at therapeutic dosage to be forced into the model. We considered P<0.05 as statistically significant.
The last step of the statistical analysis was performed in the subgroup of patients with D-ICH. It consisted of a bivariate comparison of demographic characteristics, first radiological findings (infarct or E-ICH), and heparin use at time of diagnosis of CVT, between patients with mRS 0 to 2 and mRS 3 to 6 at month 6, using t test, χ2 test with Yates correction, or Fisher exact test when appropriate.
We performed statistics with SPSS 11.0 for windows.
Of 624 patients included in ISCVT, 245 (39%) had an E-ICH. Six-month follow-up information was obtained in 241 (98.4%) of them. Details on the main characteristics of E-ICH and de novo D-ICH are provided in Table 1.
Bivariate Comparison Between Patients With and Without E-ICH
The main characteristics of patients with and without E-ICH are detailed in Table 2⇓. Patients with E-ICH were older, less likely to have isolated intracranial hypertension, subacute onset, right lateral sinus occlusion, any infection and vasculitis, and more likely to have a left lateral sinus thrombosis, an acute onset, Glasgow coma score <9, aphasia, monoparesis or hemiparesis, seizures, a cerebral infarct on CT or MRI, one or more lesions >5 cm; bilateral lesions, acquired thrombophilia, to have had neurological worsening, to be dead within 30 days after onset, to have a D-ICH, and to have mRS 3 to 6 at month 6 and at the last follow-up.
Bivariate Comparison in the 241 Patients With E-IC of Baseline Characteristics Between Patients With mRS 3 to 6 and Those With mRS 0 to 2 at Month 6
At month 6, 51 (21%) patients were dead or dependent (mRS 3 to 6) and 190 (79%) were independent (mRS 0 to 2). The main characteristics of patients with mRS 0 to 2 and mRS 3 to 6 at month 6 are detailed in Table 3⇓. Patients with mRS 3 to 6 were older, and more likely to be men, to have superior longitudinal sinus thrombosis or deep cerebral veins thrombosis, one or more lesion >5 cm, Glasgow Coma Scale <9, monoparesis or hemiparesis, puerperium as predisposing factor, to have neurological worsening, and D-ICH. They did not differ for the use of heparin at therapeutic dosage or not, low-molecular-weight heparin at preventive dosage, and antiplatelet drugs at the acute stage.
Logistic Regression Analysis in Patients With E-ICH With mRS 3 to 6 as Dependent Variable
We found as independent predictors of being dead or dependent (mRS 3 to 6) at month 6: being older (adjusted odds ratio for 1-year increase of age 1.05; 95% CI, 1.02 to 1.08); male gender (adjusted odds ratio, 3.25; 95% CI, 1.29 to 8.16); having a thrombosis of the deep cerebral venous system (adjusted odds ratio, 5.43; 95% CI, 1.67 to 17.61) or a right lateral sinus thrombosis (adjusted odds ratio, 2.56; 95% CI, 1.03 to 6.40), and having a motor deficit (adjusted odds ratio, 2.94; 95% CI, 1.21 to 7.10).
Bivariate Comparison of Baseline Characteristics, First Radiological Examination, and Heparin Use at Time of Diagnosis Between Patients With mRS 3 to 6 and Patients With mRS 0 to 2 at Month 6 in CVT Patients with D-ICH
Of the 624 patients included in ISCVT, 272 (44%) patients had a control CT/MRI. Thirty-six patients (13%) had D-ICH, with 23 of them having a de novo D-ICH. Nineteen of these 36 patients had a mRS 3 to 6 at month 6, and 17 had a mRS 0 to 2. Patients with mRS 3 to 6 at month 6 were more likely to have E-ICH at first radiological examination, and they were less likely to have received heparin in therapeutic doses at the acute phase. The comparison of demographic characteristics, initial imaging features, and heparin use on the outcome in patients with late hemorrhage are detailed in Table 4. We did not perform a logistic regression analysis because of the small sample size.
This study was conducted in the largest prospective cohort ever published of CVT patients. It has shown that: (1) the proportion of CVT revealed by ICH is high; (2) among CVT patients, those with E-ICH were older, and more likely to have a severe clinical presentation and a worse 6-month outcome that CVT patients without ICH; (3) in patients with E-ICH, independent predictors of death or dependency were older age, male gender, having a thrombosis of the deep cerebral venous system or of the right lateral sinus, and having a motor deficit; (4) in patients with D-ICH, those who had a worse outcome were less likely to have been treated by heparin, and to have had an E-ICH.
The major strength of this study is the large sample size, and the huge collaboration between many types of hospitals, in different countries and continents, decreasing potential recruitment bias. The study population is not representative of all CVT, but consists in a subgroup of patients who have a higher risk for poor outcome when compared with CVT patients without ICH. Completeness of follow-up was satisfactory, with only 1.6% patients lost to follow-up. However, the methodological limitation was the lack of central review of imaging. Another methodological limitation was the absence of uniform etiological work-up. Nevertheless, an extensive search for risk factors was pursued in most patients, as supported by the small number of patients without identified risk factors. Incomplete case ascertainment is a possible source of bias. Most investigators being neurologists, severe cases admitted in intensive care units may have been missed, especially in CVT with ICH and a more severe clinical presentation. Therefore, we cannot exclude an under-representation of severe cases and, therefore, an underestimation of death and disability rates at month 6. However, investigators had been asked repeatedly to include also patients from other departments, especially intensive care units.4 Other studies found a similar proportion of CVT patients with E-ICH.1,10 However, in most studies, ICH included also small ICH surrounded either by normal brain or by a hypodensity of variable size, more rarely subdural hematoma, subarachnoid hemorrhage, or blood effusion on the tentorium.1,10 Patients with E-ICH, had a better outcome in ISCVT than in previous studies. We found a 6% mortality rate within 30 days, and a 21% rate of death or dependency at month 6, whereas previous studies found mortality rates of 20%,11 and death or dependency rate at month 6 between 31% and 40%.1,12,13 The VENOPORT study was the only one in which the proportion of death and dependency at month 6 was similar in CVT patients with ICH.14 More details on the duration of symptoms of CVT before the occurrence of ICH would have been interesting, but the first symptom is usually found retrospectively, and the value of this information is probably poor in case of coma, delirium, aphasia, or neglect. This is the reason why we classified patients as having an acute, subacute, or chronic onset, and did not record the duration of symptoms before the ICH occurred.
To our knowledge, our study is the first large study, which aimed to identify predictors of outcome in CVT patients with E-ICH. Besides increasing age and male gender, the presence of a deep cerebral venous thrombosis was an independent predictor of poor outcome, as already found in CVT patients without ICH.4 The higher risk of death or dependency in case of thrombosis located at the level of the right lateral sinus might be attributable to the larger size of this sinus, explaining that the drainage cannot be made by the other hypoplastic side,15 but may also just be a chance finding. A future independent study would help to answer this question. The worse outcome in the presence of a motor deficit may be explained by the presence of a cerebral lesion.
We found that the risk of D-ICH was greater in patients with E-ICH, and was not influenced by heparin. Almost 3 of 4 patients with E-ICH were treated with heparin at therapeutic dosages, which was a similar proportion than that found in non-ICH CVT patients, but much higher than in the Venoport study.14 The safety was not as good in Fink’s case series,16 but selection bias cannot be excluded in this study.16 Heparin use was also associated with a better outcome. All patients with D-ICH who had a good outcome were under heparin at therapeutic dosage. This result suggests that a major contribution for D-ICH is progression of CVT, which is limited by heparin. However, because our study was observational and heparin allocation was not randomized, we cannot exclude that those patients who were treated by heparin and those who had the less severe clinical presentation, ie, those who were the most likely to have a good outcome. Patients admitted for CVT revealed by an E-ICH should be considered as being at risk of death and dependency and deserve additional close monitoring. Because at this time other therapeutic options, besides anticoagulation, remain individual,17 and because in the future new options will be proposed,18 it is necessary to identify predictors of outcome in large series of consecutive patients, such as ISCVT, and to try to define target populations for a more aggressive management to be evaluated in randomized trials. Increasing age, male gender, having a thrombosis of the deep cerebral venous system, having a motor deficit, and having a right lateral sinus thrombosis are predictors of poor outcome in CVT patients with ICH, and should therefore be the target for new strategies.
Clinical variables: clinical course (acute if the interval between onset of symptoms and diagnosis was <48 hours; subacute between 48 hours and 30 days; and chronic >30 days); symptoms and signs present at, or before, admission (headache, visual loss, papilledema, diplopia, coma, aphasia, monoparesis or hemiparesis, focal seizure, seizure with generalization, any seizure, sensory symptoms); Glasgow coma scale divided into 2 categories: 9 to 15 and <9; presenting syndrome (isolated intracranial hypertension and other presentations).
Lesions at admission on computed tomographic or MRI scans: infarct or edema (left or right hemisphere, posterior fossa), hemorrhage (left or right, posterior); any cerebral lesion (infarct or hemorrhage); bilateral lesions (infarct or hemorrhage), posterior fossa lesion (infarct or hemorrhage); number of lesions >1 cm and number of lesions >5 cm.
Location of thrombosis (superior sagittal sinus, right lateral sinus, left lateral sinus, straight sinus, deep cerebral venous system, cortical vein, jugular vein, >1 sinus).
Risk factors and causes: any prothrombotic conditions including congenital or acquired thrombophilia; any malignancy (intracranial and extracranial tumors, hematological malignancies); cerebrovascular malformations (dural fistulae, venous anomaly, arteriovenous malformation); hematological condition (anemia, polycythemia, thrombocythemia); vasculitis; any systemic disorders (inflammatory systemic disorders; other systemic disorders); any infections (central nervous system infections, ear, sinus, mouth, face and neck infection, other), puerperium (pregnancy, postpartum), any potential mechanical precipitants (lumbar puncture, cranial trauma, jugular catheter occlusion, neurosurgery), hormonal contraceptive therapy, and other drugs.
Sources of Funding
This study was supported by PRAXIS grant C/SAU/10248/1998 from the “Fundação para a Ciência e Tecnologia,” and grant EA 2691 from the French Ministry of Education, Research, and Technology.
- Received June 17, 2006.
- Revision received September 8, 2006.
- Accepted September 15, 2006.
Ferro JM, Canhao P, Stam J, Bousser MG, Barinagarrementeria F, ISCVT Investigators. Prognosis of cerebral vein and dural sinus thrombosis: results of the International Study on Cerebral Vein and Dural Sinus Thrombosis (ISCVT). Stroke. 2004; 35: 664–670.
Bonita R, Beaglehole R. Recovery of motor function after stroke. Stroke. 1988; 19: 1497–1500.
Glantz S, Slinker B. Primer of applied regression and analysis of variance. New York: McGraw Hill; 1990.
de Bruijn SF, de Haan RJ, Stam J. Clinical features and prognostic factors of cerebral venous sinus thrombosis in a prospective series of 59 patients. For The Cerebral Venous Sinus Thrombosis Study Group. J Neurol Neurosurg Psychiatry. 2001; 70: 105–108.
Cakmak S, Derex L, Berruyer M, Nighoghossian N, Philippeau F, Adeleine P, Hermier M, Froment JC, Trouillas P. Cerebral venous thrombosis: clinical outcome and systematic screening of prothrombotic factors. Neurology. 2003; 60: 1175–1178.
Newton TH, Potts DG. Radiology of the skull and brain/angiography; vol. II Book II Veins. St Louis, MO: Mosby; 1974.
Fink JN, McAuley DL. Safety of anticoagulation for cerebral venous thrombosis associated with intracerebral hematoma. Neurology. 2001; 57: 1138–1139.
Curtin KR, Shaibani A, Resnick SA, Russell EJ, Simuni T. Rheolytic catheter thrombectomy, balloon angioplasty, and direct recombinant tissue plasminogen activator thrombolysis of dural sinus thrombosis with preexisting hemorrhagic infarctions. AJNR Am J Neuroradiol. 2004; 25: 1807–1811.
Kimura R, Nakase H, Tamaki R, Sakaki T. Vascular endothelial growth factor antagonist reduces brain edema formation and venous infarction. Stroke. 2005; 36: 1259–1263.