Vascular and Nonvascular Mimics of the CT Angiography Spot Sign in Patients With Secondary Intracerebral Hemorrhage

Transcription

1 Vascular and Nonvascular Mimics of the CT Angiography Spot Sign in Patients With Secondary Intracerebral Hemorrhage Steve Gazzola, MD; Richard I. Aviv, MBChB, FRCR; David J. Gladstone, MD, FRCPC; Gabriella Mallia, BSc(C); Vivian Li, BA(C); Allan J. Fox, MD, FRCPC; Sean P. Symons, MD, MSc, FRCPC Background and Purpose The newly-described computed tomography angiography (CTA) Spot Sign is present in about one third of patients with acute primary intracerebral hemorrhage (PICH) and predicts hematoma expansion. This sign has not been systematically evaluated in patients with secondary causes of ICH, and mimics have not been characterized. The purpose of this study was to assess for the presence of the Spot Sign in secondary ICH and to document potential mimics of the Spot Sign and their distinguishing features. Methods We performed a retrospective chart review of consecutive patients presenting with ICH to our regional stroke center between January 2002 and May Ninety-six ICH patients underwent a CT stroke protocol including CTA. CTA documented a secondary cause for hemorrhage in 30 patients (31%). Each patient was assessed for the presence or absence of the CTA Spot Sign or a mimic by 2 blinded neuroradiologists. Clinical and radiological features of PICH and secondary ICH were compared. Results No patients with secondary ICH had a true CTA Spot Sign, but several Spot Sign mimics were identified including: micro AVM, posterior communicating artery aneurysm, Moya Moya, and neoplasm-associated calcification. The secondary ICH group was younger (P ) and less likely to be hypertensive at presentation (P ). Significant hematoma expansion ( 33% increase from baseline volume) occurred in 20% of secondary ICH patients and 28% of PICH patients (P ). Conclusion This study describes mimics of the CTA Spot Sign and classifies them as vascular (microavm, aneurysm, Moya Moya) or nonvascular (tumor and choroid plexus calcification). Evaluation of the noncontrast CT together with the CTA source images is an essential part of the evaluation for the Spot Sign. Vessels entering the hematoma from the periphery are indicative of an underlying vascular lesion. Our findings suggest that the Spot Sign may be rare in secondary ICH and most specific for PICH. (Stroke. 2008;39: ) Key Words: computed tomography angiography intracerebral hemorrhage hematoma expansion Spot Sign secondary ICH factor VIIa The Computed Tomography Angiography (CTA) Spot Sign has recently been shown to predict hematoma expansion in patients presenting acutely with primary intracerebral hemorrhage (PICH). 1 The ability to accurately predict which patients will undergo hematoma expansion is of major clinical importance, not only to predict morbidity and mortality well known to be associated with hematoma size and expansion, 1 5 but also to potentially guide acute therapy. However, the prevalence of the sign in the context of secondary causes of intracerebral hemorrhage is unknown. Accurate identification of the Spot Sign and distinction from potential mimics is essential for the success of studies that may use this sign to stratify or select patients for treatment interventions. In this study, we assess the prevalence of the CTA Spot Sign in secondary ICH patients and report on several spot mimics. We hypothesize that the CTA Spot Sign is specific for PICH and may be less common in SICH. Patients and Methods Study Group A retrospective chart review of consecutive patients presenting with ICH between January 2002 and May 2007 was performed. Ninetysix patients investigated with a CT stroke protocol including CTA were reviewed. Six patients with only subarachnoid blood were excluded. Fifty-nine patients with PICH and 31 patients with CTA documented secondary cause for ICH formed the patient cohort. One Received July 16, 2007; final revision received August 23, 2007; accepted August 30, From the Division of Neuroradiology (R.A., S.S., A.J.F., V.L., G.M.), Department of Medical Imaging (S.G.), and Regional Stroke Centre (D.G.), Sunnybrook Health Sciences Centre, University of Toronto, Ontario, Canada. Correspondence to Dr R. Aviv, MBChB, Department of Medical Imaging, Division of Neuroradiology, Room AG 31, Sunnybrook Health Sciences Centre, 2075 Bayview Avenue, Toronto, Ontario, M4N 3M5, Canada. Richard.aviv@sunnybrook.ca 2008 American Heart Association, Inc. Stroke is available at DOI: /STROKEAHA

2 1178 Stroke April 2008 Table 1. Demographic Characteristics of Patients With PICH and SICH Primary ICH Secondary ICH P Value Median age, y 64 (31 86) 45.5 (16 82) Male gender 26 (44.1%) 16 (53%) No. of patients with CTA under 6 hours 6 (20%) 39 (100%) Volume of hematoma at presentation 33.5 (32.75) 20.9 (19.43) Volume of hematoma at follow-up, mean (SD) 36.7 (34.42) 22.3 (18.30) No. with expansion 11 (18.6%) 20 (20%) ICH location: lobar 34 (58.6%) 18 (60%) ICH location: basal ganglia 22 (37.2%) 6 (20%) ICH location: cerebellar 3 (5.1%) 6 (20%) Presentation mean arterial pressure, mean (SD) (17.16) (21.75) MAP (25.4%) 4 (13.3%) History of hypertension 18 (30.1%) 5 (16.7%) Anticoagulation at presentation 11 (18.6%) 1 (3.3%) Presentation INR, mean (SD) 1.3 (0.59) 1.07 (0.19) Presentation APTT, mean (SD) 30.2 (7.27) 28.4 (3.18) Presentation Glucose, mean (SD) 8 (3.13) 7.7 (2.28) Glucose 8.3 mmol/l at presentation 10 (16.9%) 8 (26.6%) Median mrs at 3 months 3 (0 6) 2.5 (0 6) Presence of hydrocephalus 5 (8.5%) 17 (62.9%) Median hospital length of stay 11 (2 76) 10 (3 210) All values represent No. of patients (percent) unless otherwise indicated. SICH patient was excluded because of surgical decompression before CTA. Thirty SICH patients were therefore available for review. The study was institutional review board approved. Materials/Image Acquisition A hemorrhagic stroke protocol was performed on a 4-slice ( ) or a 64-slice ( ) CT (GE Lightspeed plus and VCT; GE). Noncontrast head imaging is followed by a CTA study and a postcontrast study 1 to 2 minutes after contrast injection. Pre- and postcontrast head imaging is acquired from the skull base to the vertex with the following parameters: 120 kvp; 340mA; 4 5 mm collimation; 1 second/rotation; and a table speed of 15 mm/rotation. CTA studies are obtained from the level of the C6 vertebra to the vertex in the helical HS mode. CTA parameters are 0.7 ml/kg contrast (to a maximum of 90 ml through an antecubital vein via at least an 18 or 20 gauge angiocatheter); 5 to 10 second delay; 120 kvp; 270 ma; 1 second/rotation; 1.25 mm slice thickness reconstructed to mm. Table speed 3.75 mm/rotation. CT technologists perform all postprocessing including multiplanar reformats at the CT operator s console. Coronal and sagittal multiplanar reformat images are created as 10.0-mm-thick images, spaced by 3 mm. Bilateral rotational multiplanar reformat images are created at each carotid terminus with a thickness of 7 mm and a spacing of 3 mm. All images are viewed on AGFA Impax 4.5 PACS workstation. Cerebral digital subtraction angiography (DSA) was used as the reference standard for confirmation of vascular lesion presence where the diagnosis was in doubt or occasionally to confirm a diagnosis or before therapeutic intervention. DSA was performed on a Phillips uniplane neuroangiographic unit. Selective angiography was performed for confirmation of CTA-detected lesions or for therapeutic procedures. Four vessel angiography was performed where no abnormality was detected on CTA. Imaging Analysis/Interpretation All studies were prospectively evaluated by 2 neuroradiologists for the presence or absence of the CTA spot sign. The sign was defined as a 1- to 2-mm focus of enhancement within a hematoma on the CT angiographic axial source and multiplanar reformatted (MPR) images as assessed by simple visual inspection. 6 Presence of potential CTA spot sign mimics was recorded. The modality securing the final diagnosis was recorded. The location and size of hematoma was noted. Hematoma volume at presentation and follow-up study was calculated using the previously validated ABC/2 method. 7 Intraventricular extension was graded as mild, moderate, or severe. Mild IVH constituted a small amount of dependant blood in any ventricle. Moderate IVH was scored when more than one ventricle contained blood. Blood filling all ventricles was considered severe IVH. An increase in hematoma volume of 33% was considered significant. 8 Statistical Methods For each patient baseline data that was collected included: the presence of the CTA Spot Sign or mimic, time to CTA from presentation to first ER, time to CTA from symptom onset, age at presentation, sex, anticoagulant/antiplatelet use, history of hypertension, blood pressure, clotting profile, and glucose at presentation. The size of the hematoma at presentation and on follow-up was documented. Thresholds of glucose 8.3 mmol/l and mean arterial pressure (MAP) 120 mm Hg were selected based on previous studies. 2,3 Modified Rankin scores at 3 months were recorded. The primary and secondary ICH groups were compared for age, gender, MAP, glucose, hematoma volume at presentation and follow-up, and clinical outcome. The Student t test and Fisher test were used for group comparisons for continuous and categorical data, respectively. Data were analyzed using SPSS for Windows (Version 14; SPSS Inc). A probability value 0.05 is considered significant. Results A secondary cause of ICH was present in 31 patients, representing 34% of consecutive ICHs presenting to our institution investigated with CT angiography. One patient was excluded because of surgical decompression before CTA. Cerebral angiography was performed in 25/31 patients (80%). Patients who did not undergo angiography included 4

3 Gazzola et al CTA Spot Sign in Secondary Intracranial Hemorrhage 1179 Table 2. Summary of Cases Demonstrating CTA Spot Sign Mimics Classified by Etiology, Mimic Type, and Presence of Expansion Patient Actual Diagnosis Type of Mimic Explanation Expansion 1 Tumoral hemorrhage Nonvascular Left thalamic hematoma calcifications in a patient yes with metastasis 2 Tumoral hemorrhage Nonvascular Right frontal lobe hematoma calcification in a no patient with oligodendroglioma 3 Tumoral hemorrhage Nonvascular Choroid plexus calcification no 4 Moya Moya Vascular Pseudoaneurysm arising from lenticulostriate no artery 5 Micro AVM Vascular Micro AVM draining to a cortical vein no 6 Micro AVM Vascular Micro AVM with pseudoaneurysm draining to no lateral ventricular vein 7 PCOM aneurysm Vascular Partially thrombosed PCOM aneurysm yes patients with tumor and 2 patients with large CTA identified aneurysms which were surgically treated. Characteristics of patients with primary and secondary ICH are presented in Table 1. Patients with secondary ICH were younger (P ) with a median age of 45 years (range, 16 to 82 years). There was a trend for initial and follow-up hematoma volumes to be smaller than PICH patients (P ). There was a trend to more posterior fossa lesions in the secondary ICH patients (P ). History of hypertension, MAP, and number of patients with MAP elevation 120 were lower (P , P , P , respectively). Fewer patients were on anticoagulants (P ), and there was a trend to a lower mean INR (P ). Hydrocephalus was more often seen in SICH than PICH (P ). Secondary causes of ICH in the study patients were: arteriovenous malformation (AVM) (n 12), dural arteriovenous fistula (AVF) (n 2), tumor (n 6), aneurysm (n 9), and Moya Moya (n 1). Sixty-three percent of SICH patients presented between the 4th to 6th decades. Aneurysms and tumor were more common beyond the 4th decade, and AVMs presented predominately in the first 4 decades. Sixteen percent of patients over 60 had a secondary cause for ICH. CTA revealed a secondary cause for basal ganglia bleeds in 18% of cases. Hemorrhage from aneurysm and AVM accounted for the majority of causes and greatest severity of hydrocephalus. Significant hematoma expansion occurred in 6/30 (20%) SICH and 11/59 (19%) of PICH patients, respectively. One patient taken to surgery emergently for hematoma evacuation because of sudden clinical deterioration without interval scan was included. This patient s initial CTA, done at 3.7 hours from symptom onset, showed no hydrocephalus or intraventricular extension of a temporal lobe hematoma secondary to a dural AVF. The median time from presentation to CTA was 7 hours (0.5 to 116 hours). Eighty-four percent (26/31) of SICH patients had CTA within 48 hours of symptom onset and 77% (24/31) within 24 hours. Twenty percent presented within 6 hours. All SICH expansion occurred within 24 hours of symptom onset, half occurring within 12 hours. The CTA Spot Sign was not present on any study, but several mimics were identified. We divided the causes into nonvascular and vascular (Table 2). Nonvascular causes were attributable to punctuate calcification that simulated contrast on a postcontrast study. Distinction may not have been possible without the presence of a noncontrast study for comparison (Figure 1). In 2 cases punctuate perihematoma calcification was present. Hematoma induced compressed, irregularly calcified, choroid plexus accounted for the other mimic. Three vascular mimics were seen Figure year-old male with a right frontal oligodendroglioma (patient 2, Table 2). A, NCCT demonstrates punctuate peripheral calcification, (white arrow) within the hematoma (white asterisk). B, Hyperdensity apparent on CTA axial source images and post contrast CT images.

4 1180 Stroke April 2008 Figure year-old female with a micro AVM (patient 6, Table 2). A, NCCT demonstrates centrum semiovale hematoma (white asterisk). B, A punctate focus of enhancement (white arrow) representing a pseudoaneurysm (white arrow) on CTA axial source images mimics a CTA spot sign. C, A vessel is clearly seen extending from the lateral ventricular surface to the hematoma periphery and communicates with the focal spot of enhancement on coronal MPR image (black arrow). D, AP Oblique conventional angiographic image demonstrates a micro AVM with pseudoaneurysm (black arrow) fed by branches of the anterior cerebral artery and draining to the lateral ventricular vein. in patients with micro AVM (Figure 2) and Moya Moya (Figure 3). These cases demonstrated pseudoaneurysms arising from discrete CTA discernable vessels. This differs from the CTA Spot Sign where contrast foci are suspended within the hematoma with no vascular connection to surrounding structures. The other mimic occurred in a patient with a partially thrombosed posterior communicating artery aneurysm (PCOM). However, in this unusual case the punctuate foci of enhancement occurred in the subarachnoid space and were slightly distanced from the hematoma edge (Figure 4). Discussion This is the first study to evaluate the CTA Spot Sign in secondary ICH cases and to describe spot sign mimics. The CTA Spot Sign is defined as 1 to 2 mm foci of enhancement within a parenchymal hematoma. 6 These densities may be spot-like or linear, single or multiple, and are only visualized on CT angiographic axial source and multiplanar reformatted images images. Previous studies suggest that the sign may have important implications for stratification of hematoma expansion risk. If the sign is to be used to select a target group Figure year-old male with moya moya (patient 4, Table 2). An enhancing focus is seen in the inferior Basal Ganglia/Nucleus accumbens consistent with a pseudoaneurysm (white arrow) arising from a lenticulostriate artery (white arrowhead) on (A) CTA source, (B) coronal MPR, and (C) saggital MPR images. The lenticulostriate vessel lies peripheral to the hematoma only contacting it at the site of pseudoaneurysm. Intraventricular hemorrhage and hydrocephalus incidentally noted.

5 Gazzola et al CTA Spot Sign in Secondary Intracranial Hemorrhage 1181 Figure year-old female with PCOM aneurysm (patient 7, Table 2). A punctuate enhancing focus (white arrow) is present slightly remote from a right PCOM aneurysm (white arrowhead). The focus lies outside of the mesial temporal hematoma (white asterisk) within the subarachnoid space seen on (A) CTA axial source and (B) saggital MPR images. C, Conventional angiographic images reveal filling of a right PCOM aneurysm (white arrow) in a loculated fashion attributable to partial thrombosis. D, Dependant contrast opacification is evident on delayed images consistent with the focal enhancement seen on CTA. for therapy, correct recognition of potential mimics is essential. It might be argued that patients with secondary causes may not be precluded from treatment with Factor Viia to prevent hematoma expansion. However, correct identification of these lesions is important to allow definitive treatment to prevent future hemorrhage risk. Importantly, enhancing densities are isolated within the hematoma and not connected to vascular structures peripheral to the hematoma. The presence of calcification should be excluded on noncontrast CT. We previously reported a prevalence of the Spot Sign in 30% of PICH patients presenting within 3 hours of stoke onset. 6 This study reported extravasation independent of the Spot Sign in half of spot positive cases. Extravasation was defined as an accumulation of contrast on a postcontrast CT. Other studies found contrast extravasation in 50% of ICH patients. 9 Extravasation was reported as late as 5 days after symptom onset. 3 Studies focusing on the hyperacute period (0 to 6 hours) 10,11 found similar rates of contrast extravasation. The CTA Spot Sign has been found to be an independent predictor of hematoma expansion with sensitivity, specificity, positive, and negative predictive value of 91, 89, 77, and 96%, respectively. 6 Contrast extravasation from aneurysms and other vascular lesions is recognized. 11 In our view the detection of extravasation is a later sign in the natural history of a spot-positive patient with less specificity for hematoma expansion. The absence of the Spot Sign in secondary ICH suggests specificity for PICH and reinforces our hypothesis that the Spot Sign may be the cause of PICH bleeding. The presence of multiple Spot Signs in a hematoma does not preclude this notion. Fisher described fibrin globes or pseudoaneurysms up to 600 in size. 12 These are masses of red cells in concentric rings of fibrin adjacent to an arteriolar defect. The lesions are frequently multiple, present at the hematoma margin, and seen in association with focal lipohyalinosis which reflects a more widespread vessel abnormality or fragility. Fisher proposed a domino effect whereby one or more lesions rupture and are responsible for the majority of hematoma growth. The expanding hematoma then results in secondary damage to already fragile vessels with ongoing albeit lesser bleeding. It is unknown what the individual contribution of the multiple lesions to hematoma growth is but we have certainly identified active extravasation from multiple spots on postcontrast CT. This hypothesis could explain the presence of multiple lesions with and without extravasation and an apparent specificity to PICH. Several mimics of the CTA Spot Sign were found in this study, classified as vascular or nonvascular. Nonvascular causes are attributable to calcific deposition. Review of the noncontrast CT before CTA confirmed the presence of calcification in all patients. Three patients had a peripheral hyperdense spot on CTA images. One lesion was deep, situated within the left thalamus, the other lobar, within the right frontal lobe. Both had underlying tumor; a metastasis and oligodendroglioma, respectively. The third had calcification in the choroid plexus, compressed by the hematoma. Although not encountered in this study IVH clot surrounding choroid may be difficult to distinguish from adjacent parenchymal hematoma. The hematoma then appears to engulf the calcification rather than displacing it mimicking the Spot Sign. Careful scrutiny of the choroid plexus on contiguous noncontrast CT slices is sufficient to avoid this pitfall. Physiological and postinfective/inflammatory calcification although not described in this study would similarly be expected to mimic the Spot Sign.

6 1182 Stroke April 2008 Vascular mimics were a partially thrombosed posterior communicating aneurysm (PCOM), 2 micro AVMs and a patient with Moya Moya and lenticulostriate pseudoaneurysm. Careful inspection of the PCOM aneurysm case revealed enhancing densities external to the hematoma and brain parenchyma. The extra axial (subarachnoid) location precluded a diagnosis of the Spot Sign. DSA revealed loculated filling of the aneurysm attributable to partial thrombosis. Although not present in our series we speculate that, internal carotid artery terminus, anterior communicating, and middle cerebral artery bifurcation aneurysms that are closely applied to adjacent brain parenchyma may present similar difficulties. In our experience review of CT angiographic images usually confirm the extra axial nature of these vascular lesions. Both patients with micro-avm revealed a continuous linear density extending from vessels on the brain or ventricular surface into the depths of the hematoma. The Moya Moya case had a similar appearance with lenticulostriate vessel arising outside the hematoma and coursing through it to terminate as a pseudoaneurysm. Extension of linear densities beyond the hematoma margin is not seen in any case of CTA Spot Sign. Noncontrast CT (NCCT) has poor sensitivity and specificity for prediction of underlying vascular lesions in ICH patients. 13 Risk of neurological deficit limits the potential use of DSA in certain subgroups where the incidence of a known cause of ICH is low. The American Heart Association guidelines acknowledge that underlying etiology cannot reliably be predicted by hematoma location alone. 2 Apparently typical bleeds have an underlying vascular cause for ICH in 2% to 36%, 14,15 and the incidence of underlying vascular lesions in all ICH cases is 30%. 13,15 We believe that the low complication rates of CTA justify its use even in patients with typical hypertensive bleeds. This is especially so in the face of a potential CTA marker for hematoma expansion and treatment target. The accuracy of CTA for detection of underlying vascular lesions is 90% (Yeung R, unpublished data, 2007). CTA is rapid, accessible, and widely available. Therefore NCCT followed by CTA are the first line of investigation in ICH patients in our institution. DSA is used as a confirmatory test in abnormal cases and to confirm AVM angioarchitecture and drainage pattern. Younger patients without CTA abnormality would also progress to DSA according to the AHA guidelines. The main limitation of this study was that few patients were scanned ultraearly. The delayed onset-to-cta times may therefore have limited our ability to detect Spot Signs in this cohort. The longer delay from presentation to CTA in this study is largely a reflection of patient inclusion before the introduction of a citywide acute stroke prehospital bypass protocol. Our PICH cohort dates predominantly after the implementation of this protocol. We cannot exclude the possibility that a Spot Sign may have been present in some patients if they had been scanned earlier. In our experience, however, half of spot positive patients progress to extravasation. Other studies have confirmed extravasation in half of patients presenting with ICH investigated at a median time of 11 hours (13 to 33 hours) 9 and up to 5 days. 3 Furthermore, the median time to CTA in this and the study by Goldstein et al study is identical. 9 Another limitation of this study is that we did not capture all consecutive patients presenting with ICH. Before the introduction of routine CTA in acute stroke at our institution, the decision to perform a CTA was at the discretion of the referring clinician. This may have resulted in selection bias and resulted in the observed differences in the 2 groups. Larger prospective studies with patients scanned at earlier time points will be important to confirm and extend these findings. Conclusion The Spot Sign is a potentially promising sign for predicting hematoma expansion. This sign may play a role in patient selection for clinical trials of acute hemostatic therapy or surgical interventions aimed at minimizing hematoma growth. Correct recognition of the sign and exclusion of secondary ICH causes will be crucial to the success of such studies. Clinicians should be aware of the Spot Sign mimics, classified as either vascular or nonvascular. Nonvascular causes are easily distinguished with NCCT but may be misinterpreted as contrast density in its absence. The presence of a vessel extending from brain or ventricular surface into a hematoma strongly suggests the presence of underlying vascular malformation. A true CTA Spot Sign was not found in any of our patients with secondary ICH. The absence of a Spot Sign in secondary ICH, then, implies that this sign has greater specificity for primary ICH than for secondary ICH, although this will need to be confirmed in larger prospective studies. Acknowledgements We acknowledge our Tech staff without whose dedication and effort our imaging data would be deplete. Sources of Funding Dr Richard Aviv receives grant funding from the Canadian Stroke consortium. Dr Gladstone is supported by the Heart and Stroke Foundation of Ontario, the Heart and Stroke Foundation Centre for Stroke Recovery, and the Department of Medicine at Sunnybrook Health Sciences Centre and University of Toronto. None. Disclosures References 1. Kazui S, Naritomi H, Yamamoto H, Sawada T, Yamaguchi T. Enlargement of spontaneous intracerebral hemorrhage. Incidence and time course. Stroke. 1996;27: Broderick J, Connolly S, Feldmann E, Hanley D, Kase C, Krieger D, Mayberg M, Morgenstern L, Ogilvy CS, Vespa P, Zuccarello M. Guidelines for the management of spontaneous intracerebral hemorrhage in adults update. A guideline from the American Heart Association, American Stroke Association Stroke Council, High Blood Pressure Research Council, and the Quality Of Care And Outcomes In Research Interdisciplinary Working Group. Stroke. 2007;38: Becker KJ, Baxter AB, Bybee HM, Tirschwell DL, Abouelsaad T, Cohen WA. Extravasation of radiographic contrast is an independent predictor of death in primary intracerebral hemorrhage. Stroke. 1999;30: Mayer SA, Brun NC, Begtrup K, Broderick J, Davis S, Diringer MN, Skolnick BE, Steiner T. Recombinant activated factor vii for acute intracerebral hemorrhage. N Engl J Med. 2005;352:

7 Gazzola et al CTA Spot Sign in Secondary Intracranial Hemorrhage Fujii Y, Tanaka R, Takeuchi S, Koike T, Minakawa T, Sasaki O. Hematoma enlargement in spontaneous intracerebral hemorrhage. J Neurosurg. 1994; 80: Wada R, Aviv RI, Fox AJ, Sahlas DJ, Gladstone DJ, Tomlinson G, Symons SP. Ct angiography Spot sign Predicts hematoma expansion in acute intracerebral hemorrhage. Stroke. 2007;38: Kothari RU, Brott T, Broderick JP, Barsan WG, Sauerbeck LR, Zuccarello M, Khoury J. The abcs of measuring intracerebral hemorrhage volumes. Stroke. 1996;27: Brott T, Broderick J, Kothari R, Barsan W, Tomsick T, Sauerbeck L, Spilker J, Duldner J, Khoury J. Early hemorrhage growth in patients with intracerebral hemorrhage. Stroke. 1997;28: Goldstein JN, Fazen LE, Snider R, Schwab K, Greenberg SM, Smith EE, Lev MH, Rosand J. Contrast extravasation on ct angiography predicts hematoma expansion in intracerebral hemorrhage. Neurology. 2007;68: Yamaguchi K, Uemura K, Takahashi H, Kowada M, Kutsuzawa T. Intracerebral leakage of contrast medium in apoplexy. Br J Radiol. 1971;44: Murai Y, Takagi R, Ikeda Y, Yamamoto Y, Teramoto A. Threedimensional computerized tomography angiography in patients with hyperacute intracerebral hemorrhage. J Neurosurg. 1999;91: Fisher CM. Pathological observations in hypertensive cerebral hemorrhage. J Neuropathol Exp Neurol. 1971;30: Halpin SF, Britton JA, Byrne JV, Clifton A, Hart G, Moore A. Prospective evaluation of cerebral angiography and computed tomography in cerebral haematoma. J Neurol Neurosurg Psychiatry. 1994;57: McCormick WF, Rosenfield DB. Massive brain hemorrhage: A review of 144 cases and an examination of their causes. Stroke. 1973;4: Zhu XL, Chan MS, Poon WS. Spontaneous intracranial hemorrhage: Which patients need diagnostic cerebral angiography? A prospective study of 206 cases and review of the literature. Stroke. 1997;28: