(Stroke. 1998;29:2517-2521.)
© 1998 American Heart Association, Inc.
Original Contributions |
Value of Repeat Angiography in Patients With Spontaneous Subcortical Hemorrhage
From the Departments of Neurosurgery (A.H., M.F., T.Y., Y.I.) and Radiology (T.K.), Saiseikai Shigaken Hospital, Shiga, Japan.
Correspondence to Akihiko Hino, MD, Department of Neurosurgery, Saiseikai Shigaken Hospital, Ohashi 2-4-1, Ritto, Shiga 520-30, Japan.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Background and Purpose—Neuroradiological investigations do not disclose a source of bleeding in some patients with spontaneous subcortical hemorrhage. These patients may harbor undetected vascular malformations and may be at risk of rebleeding in the future. We investigated patients with subcortical hemorrhage with use of repeat angiography and MRI to determine the incidence of occult vascular malformations and the risk of bleeding during follow-up.
Methods—We reviewed a consecutive series of 137 patients with subcortical hemorrhage during a 10-year period (June 1987 through June 1997). If the patient was <65 years old and the first angiogram and/or MRI did not show a source of bleeding, repeat angiography was recommended. All angiographic and MRI studies were reviewed. The relationship between the identified bleeding source and clinical variables such as patient age, sex, and history of hypertension and the size and location of the hematoma were examined.
Results—One hundred seven patients (78%) underwent angiography on admission, 10 (7%) had immediate surgery for hematoma without angiography, and 20 (15%) had neither angiography nor surgery. Overall, an etiology for the hemorrhage was found in 55 cases (40%). Vascular malformations were common in young patients without preexisting hypertension. A second angiogram was obtained in 22 patients, and 4 arteriovenous malformations were demonstrated. Rebleeding at the site of the initial hemorrhage was not observed after a mean follow-up of 68 months.
Conclusions—Angiography performed acutely after hemorrhage may not demonstrate vascular malformations. Consideration should be given to repeat angiography in patients who do not have a specific cause for hemorrhage.
Key Words: angiography • cerebral arteriovenous malformations • intracerebral hemorrhage • vascular malformations
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Clinical and neuroradiological investigations do not disclose a source of bleeding in some patients with spontaneous subcortical hemorrhage. The etiology of hemorrhage in these patients is usually classified as unknown, but these patients may harbor undetected vascular malformations and may be at risk of rebleeding.1 2 3 4 5 6 7 8 9 10 11 Several scenarios are possible. If angiograms do not show a vascular malformation, the pathologically identified lesion (usually discovered at surgery) has been called an angiographically occult vascular malformation (AOVM).1 2 3 4 5 6 7 8 9 10 11 12 Most of these lesions are cavernous malformations,1 2 3 4 and some authors have recommended surgical inspection of the hematoma wall.1 5 6 7 These lesions, however, may include true arteriovenous malformations (AVMs),8 9 10 11 12 which may be best left alone in situations where detailed information of vascular anatomy is not available.
In another subset of patients, the initial angiogram does not show a vascular malformation, but repeat angiography several weeks later may reveal one.7 8 9 13 14 There have been no detailed studies, however, to support the usefulness of follow-up angiography in patients with subcortical hemorrhage who have had one angiogram that did not show a vascular malformation. Therefore, the aims of this study were to document the source of bleeding in patients with spontaneous subcortical hemorrhage, to determine the frequency of identification of occult vascular lesions in relation to clinical and radiological features, and to determine whether repeat angiography to identify occult vascular lesions is justified.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We reviewed a consecutive series of 137 patients (84 males and 53 females, aged 8 to 91 years) who were hospitalized between June 1, 1987, and June 30, 1997. All had CT scans showing single, spontaneous hematomas in the supratentorial subcortical white matter. We excluded patients with hemorrhage judged to be primarily in the cortical gray matter or the subarachnoid space. The institutional policy was for patients to be investigated with high-resolution digital subtraction angiography on admission unless there was a need for immediate surgery or it was not believed that the patient could be saved. Patients <65 years of age in whom the source of bleeding was undetermined were recommended for repeat angiography after 3 weeks and/or repeat MRI examination after discharge. Patients even <65 years of age did not undergo repeat angiography if they were judged to be at high surgical risk; if they had severe neurological disability that would advise against aggressive intervention, even if underlying vascular lesions were found; or if a coagulopathy was identified that could account for the hemorrhage. Some patients refused repeated angiographic examination. Hematoma size was estimated from CT scans based on the maximum diameter.
Surgical evacuation of the hematoma was performed in patients with mass
effect associated with focal or global neurological deficit that was
judged to be caused by the hematoma; it was carried out in 49 cases.
Extensive surgical inspection of the hematoma cavity was not
undertaken. The patients were divided into subgroups according to sex,
age, size and location of hematoma, and history of hypertension, and
the identified source of bleeding was cataloged in each subgroup. We
treated chronic hypertension not as a specific cause of bleeding but as
a possible risk factor for hemorrhage, although many previous
reports have considered it a major cause for even a subcortical
hemorrhage.15 16 17 18 19 20 21 22 The frequency of abnormal
findings on angiography and MRI were also evaluated. Statistical
comparisons between subgroups were made using the
2 test or Fisher's exact probability test,
and values of P
0.05 were considered significant.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Table 1
shows the final
classification of source of bleeding, taking into account the results
of initial and repeat angiographic studies. There were 30 AVMs (22%),
5 cavernous malformations (4%), 2 aneurysms (1.5%), and 4
brain tumors (3%). Severe coagulopathy was present in 8 cases
(6%). The underlying etiology was unknown in the remaining 82 patients
(60%). However, 20 of these patients underwent neither angiography nor
surgery, mainly because it was not believed that they could be saved.
Of the remaining 62 patients, 23 exhibited preexisting chronic
hypertension. Thus, the 23 patients with chronic hypertension (11%)
and 39 without hypertension (33%) of the 117 patients undergoing
angiographic or surgical evaluation were considered to have
hypertensive intracerebral hematoma15 16 17
and hemorrhage of unknown etiology, respectively.
|
Table 2 shows the identified bleeding
sources according to clinical and radiological subgroup. Preexisting
chronic hypertension was present in 45 patients, 9 of whom were
found to have a specific vascular lesion accounting for the
hemorrhage. A specific bleeding source was more frequent in
younger patients (P<0.05) and in those without preexisting
hypertension (P<0.005). It was uncommon for angiography to
disclose a source of hemorrhage in patients with hypertension.
The likelihood of an angiographic abnormality also decreased with
increasing age. In the 36 patients in whom vascular malformations were
identified, cavernous malformations were common in those with small
hematomas whereas AVMs were frequently identified in those with larger
hematomas (P<0.005).
|
Angiography was performed on admission in 107 patients; 22 AVMs, 3
aneurysms (including 1 unruptured aneurysm), 2 cases of
moyamoya disease, 1 venous malformation, and 1 sagittal sinus
thrombosis were identified. Thirty patients (28%) did not undergo
angiography either because there was a need for immediate surgery (10
cases) or it was not believed that the patient would survive (20
cases). Of the 79 patients who had negative initial angiograms, 5
cavernous malformations, 3 AVMs, 2 cases of amyloidosis, and 2 brain
tumors (1 lung cancer and 1 melanoma metastasis) were identified as
bleeding source on MRI or surgical specimens; severe coagulopathy was
present in 2 cases. Of the remaining 65 patients, 22 had repeat
angiography after 3 weeks, and 4 AVMs were identified (Table 3 and the
Figure). The clinical features of the 4
angiographically positive cases and the 18 negative ones are shown in
Table 3: all 4 AVMs were found in females (P<0.05), but no
significant difference in other characteristics was observed between
the 2 groups. The remaining 43 patients did not undergo repeat
angiography, since 25 patients were over 65 years of age; 14 had high
surgical risk or severe neurological disability that indicated no
aggressive intervention, even if underlying vascular lesions were
found; and 4 refused the repeat study.
|
|
Forty-nine patients underwent emergent surgery for decompression of life-threatening hematomas. Ten of these patients did not undergo angiography due to a need for immediate surgery: severe coagulopathy was present in 6 cases and an unexpected AVM was discovered during the operation in 1 case, but no specific etiology was found in the remaining 3 cases. Seven AVMs were diagnosed preoperatively, but 4 AVMs were unexpectedly identified during the operation: 3 were negative on preoperative angiography, and 1 was found in a patient who underwent immediate surgery without angiography.
Seventy patients were discharged without identification of any specific bleeding source,but follow-up MRI after discharge identified 2 gliomas. Amyloid angiopathy was assumed to be the cause of hemorrhage in 4 cases, but this was confirmed by histopathology only in 2 cases. Rebleeding at the site of the initial hemorrhage was not observed after a mean follow-up of 68 months.
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We have herein reviewed 137 consecutive patients with spontaneous subcortical hemorrhage and have identified 41 (30%) as being caused by specific underlying vascular lesions, 8 (6%) secondary to coagulopathy, 4 from brain tumor, and 2 from amyloid angiopathy (Table 1
). The so-called AOVMs, found in 12 cases (9%), included 5 cavernous
malformations, 3 AVMs that were unexpectedly encountered during
emergent operation for hematoma despite negative preoperative
angiography, and 4 AVMs that were occult on initial angiography but
were later angiographically visible. Vascular malformations were common
in younger patients and those without hypertension. Cavernous
malformations were common in patients with small hematomas whereas AVMs
were frequent in those with larger hematomas. Thirty-nine of the 117
patients (33%) undergoing angiographic or surgical evaluation
demonstrated no specific etiology or risk factors. They were considered
to have "hemorrhage of unknown etiology." The low incidence
of AOVMs (12 cases; 9%) and amyloid angiopathy (2 cases; 1.5%) and
the high frequency of "unknown etiology" may reflect the low rate
of extensive surgical and/or histological inspection of
the hematoma cavity in this series.6 10 19 Previous
studies have estimated that between 27% and 53% of patients with
lobar hemorrhage have AOVMs.5 Several authors recommended surgical exploration of the hematoma wall to detect occult vascular lesions.1 5 6 7 The rationale for early identification would be to obliterate them at initial surgery and thus prevent subsequent bleeding. Recent reports have noted that the majority of AOVMs are cavernous malformations and that they may not be as benign an entity as they were previously thought to be.1 2 3 4 11 12 However, these reports deal with many deep-seated and posterior fossa lesions, which have been reported to be associated with an increased propensity to bleed.12 23 24 25 Furthermore, even a small hemorrhage in these areas may cause a debilitating neurological deficit.1 12 24 This may have increased the cumulative morbidity of AOVMs in previously reported series. In the current series, no rebleeding occurred over a mean follow-up of 68 months, suggesting that the rebleed rate of AOVMs may be overestimated.
Is it safe to inspect the hematoma cavity under the operating microscope after hematoma evacuation? The operative risks may be low for most cavernous malformations, but these lesions tend to be very small and the vessels may not be compact, thus making it difficult to be certain of total removal in the acute stage. Overly vigorous exploration under sometimes less-than-optimal emergency conditions may cause additional brain damage in the swollen, acutely injured brain.1 It is also more difficult to obtain hemostasis in patients with hemorrhage due to amyloid angiopathy6 26 or true AVMs due to the fragile vascular walls. Empirical manipulation may cause catastrophic intraoperative bleeding in patients with AVMs.
Several studies have reported the usefulness of intraoperative angiography after emergent decompression of an acute intracerebral hemorrhage.27 28 29 If aneurysmal rupture is highly suspected in a moribund patient with a huge hematoma, this may be the best option, considering both the high risk of rebleeding and the need for prompt decompression.29 In this series, no patient underwent intraoperative angiography, but this procedure appears to be more attractive than the empirical exploration of the hematoma cavity. Once the characteristics of the underlying vascular lesions are detected, the surgeon can make a better decision about whether it is safe to proceed with resection or obliterations at that time. If high-resolution digital subtraction angiography is available in the operating room, this procedure may be a better choice to treat hematomas requiring emergent surgery.
It is noted that repeat angiography detected 4 true AVMs in 22 patients
with negative initial angiography results (Table 3 and the Figure). We
do not know whether the term AOVM can be applied to these cases, but we
stress that such lesions indeed exist and that patients without
identified specific bleeding sources should undergo repeated follow-up
studies. The lack of initial angiographic identification may be
explained by compression of the vessel lumens and/or destruction of the
abnormal vessels by hematoma, vascular thrombosis secondary to gross
hemorrhage, and/or posthemorrhagic vascular
spasm.8 9 The fact that 2 of the 4 AVMs initially occult
were detected after evacuation of clot may support the speculation that
the AVMs are compressed by hematomas, making them angiographically
occult on the initial angiogram, and are then "decompressed" on the
follow-up. It is also noted that MRI after discharge identified brain
tumors in 2 patients. Because MRI in the acute stage also often fails
to demonstrate the etiological lesion due to the presence of adjacent
hematoma, long-term follow-up study is mandatory to rule out these
lesions.
Finally, in a subcortical hemorrhage without readily identifiable risk factors, a patient with negative angiographic and MRI findings in whom no etiological lesion is detected during surgery has a significant chance of having a lesion discovered at repeat angiography and/or MRI. Since the likelihood of a specific bleeding source was more frequent in younger patients and in those without preexisting hypertension, repeat examinations should be more strongly recommended to these patients.
|
Acknowledgments |
|---|
We thank Dr R. Loch Macdonald of the section of Neurosurgery at the University of Chicago Medical Center for his helpful advice and support. We are grateful to Drs Yoshinobu Takahashi and Yasuo Inoue for their dedicated contribution to the patient treatment programs.
Received June 1, 1998; revision received September 10, 1998; accepted September 10, 1998.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
1. Tung H, Giannotta SL, Chandrasoma PT, Zee C-S. Recurrent intraparenchymal hemorrhages from angiographically occult vascular malformations. J Neurosurg. 1990;73:174–180.[Medline] [Order article via Infotrieve]
2. Robinson JR Jr, Awad IA, Masaryk TJ, Estes ML. Pathological heterogeneity of angiographically occult vascular malformations of the brain. Neurosurgery. 1993;33:547–555.[Medline] [Order article via Infotrieve]
3.
Becker DH, Townsend JJ, Kramer RA, Newton TH. Occult
vertebrovascular malformations: a series of 18
histologically verified cases with negative
angiography. Brain. 1979;102:249–287.
4. Tomlinson FH, Houser OW, Scheithauer BW, Sundt TM, Okazaki H, Parisi JE. Angiographically occult vascular malformations: a correlative study of features on magnetic resonance imaging and histological examination. Neurosurgery. 1994;5:792–800.
5. Wakai S, Ueda Y, Inoh S, Nagai M. Angiographically occult angiomas: a report of thirteen cases with analysis of the cases documented in the literature. Neurosurgery. 1985;17:549–556.[Medline] [Order article via Infotrieve]
6. Wakai S, Kumakura N, Nagai M. Lobar intracerebral hemorrhage: a clinical, radiographic, and pathological study of 29 consecutive operated cases with negative angiography. J Neurosurg. 1992;76:231–238.[Medline] [Order article via Infotrieve]
7.
Tanaka Y, Furuse M, Iwasa H, Masuzawa T, Saito K, Sato
F, Mizuno Y. Lobar intracerebral hemorrhage:
etiology and a long-term follow-up study of 32 patients.
Stroke. 1986;17:51–57.
8. Ogilvy CS, Heros RC, Ojemann RG, New PF. Angiographically occult arteriovenous malformations. J Neurosurg. 1988;69:350–355.[Medline] [Order article via Infotrieve]
9. Ojemann RG, Heros RC, Crowell RM. Surgical Management of Cerebrovascular Disease. 2nd Ed. Baltimore, Md; Williams & Wilkins; 1987:399–413.
10. Kurata A, Miyasaka Y, Kitahara T, Kan S, Takagi H. Subcortical cerebral hemorrhage with reference to vascular malformations and hypertension as causes of hemorrhage. Neurosurgery. 1993;32:505–511.[Medline] [Order article via Infotrieve]
11. Lobato RD, Perez C, Rivas JJ, Cordobes F. Clinical, radiological, and pathological spectrum of angiographically occult intracranial vascular malformations. J Neurosurg. 1988;68:518–531.[Medline] [Order article via Infotrieve]
12. Lobato RD, Rivas JJ, Gomez PA, Cbrera A, Sarabia R, Lamas E. Comparison of the clinical presentation of symptomatic arteriovenous malformations (angiographically visualized) and occult vascular malformations. Neurosurgery. 1992;30:391–397.[Medline] [Order article via Infotrieve]
13.
Zhn XL, Chann MSY, Poon WS. Spontaneous intracranial
hemorrhage: Which patients need diagnostic cerebral
angiography? Stroke. 1997;28:1406–1409.
14. Isayama Y, Nakagawara J, Takeda R, Wada K, Hyogo T, Sasaki T, Nakamura J, Suematsu K. A case of cerebral arteriovenous malformation revealed at repeated subcortical hematoma with initially normal angiogram [in Japanese]. No Shinkei Geka. 1991;19:1175–1180.[Medline] [Order article via Infotrieve]
15.
Wijdicks EFM, Jack Jr. CR.
Intracerebral hemorrhage after fibrinolytic
therapy for acute myocardial infarction. Stroke. 1993;24:554–557.
16.
Broderick J, Brott T, Tomsick K, Leach A. Lobar
hemorrhages in the elderly: the undiminishing importance of
hypertension. Stroke. 1993;24:49–51.
17.
Kase CS, Williams JP, Wyatt DA, Mohr JP. Lobar
intracerebral hematomas: clinical and CT
analysis of 22 cases. Neurology. 1982;32:1146–1150.
18. Yoshimoto H, Fujita H, Ohta K, Yoshikawa M, Shibata K, Takahashi M, Uozumi T. Clinical study of hypertensive subcortical hemorrhage: surgical indication and long-term, functional prognosis [in Japanese]. No Shinkei Geka. 1988;16:1465–1470.[Medline] [Order article via Infotrieve]
19. Molinari GF. Lobar hemorrhages: Where do they come from? How do they get there? Stroke. 1993; 24:523–526.
20. Ropper, Davis KR. Lobar intracerebral hemorrhages: acute clinical syndrome in 26 cases. Ann Neurol. 1980;8:141–147.[Medline] [Order article via Infotrieve]
21.
Weisberg LA. Subcortical lobar
intracerebral haemorrhage: clinical-computed
tomographic correlations. J Neurol Neurosurg
Psychiatry. 1985;48:1078–1084.
22. Loes DJ, Smoker WR, Biller J, Cornell SH. Nontraumatic lobar intracerebral hemorrhage: CT/angiographic correlation. AJNR Am J Neuroradiol. 1987;8:1027–1030.[Abstract]
23. Chyatte D. Vascular malformations of the brain stem. J Neurosurg. 1989;70:847–852.[Medline] [Order article via Infotrieve]
24. Porter PJ, Willinsky RA, Harper W, Wallace MC. Cerebral cavernous malformations: natural history and prognosis after clinical deterioration with or without hemorrhage. J Neurosurg. 1997;87:190–197.[Medline] [Order article via Infotrieve]
25. Robinson JR, Awad IA, Little JR. Natural history of the cavernous angioma. J Neurosurg. 1991;75:709–714.[Medline] [Order article via Infotrieve]
26. Roosen N, Martin J J, de la Porte C, Van Vyne M. Intracerebral hemorrhage due to cerebral amyloid angiopathy: case report. J Neurosurg. 1985;63:965–969.[Medline] [Order article via Infotrieve]
27. Ayuzawa S, Matsumura A, Nose T. Emergent aneurysmal surgery without preoperative angiography: usefulness of the intraoperative portable digital subtraction angiography. Surg Neurol.. 1993;40:251–254.[Medline] [Order article via Infotrieve]
28. Meguro K, Matsumura A, Tsurushima H, Matsumaru Y, Nose T. Demonstration of a vascular lesion by intraoperative portable DSA in a case of intracerebral hematoma with impending herniation [in Japanese]. No Shinkei Geka. 1990;18:453–456.[Medline] [Order article via Infotrieve]
29. Batjer HH, Samson DS. Emergent aneurysm surgery without cerebral angiography for the comatose patient. Neurosurgery. 1991;28:283–287.[Medline] [Order article via Infotrieve]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||