2. 590
Progress Reviews
Intracerebral Hemorrhage: Non-Hypertensive Causes
CARLOS S. KASE, M.D.
SPONTANEOUS, non-traumatic intracerebral hemor- small and medium-size arteries of the cerebral hemi-
rhage (ICH) in the adult is most commonly secondary spheres. The arteries affected are preferentially located
to hypertensive cerebrovascular disease. In 70-90% of in the superficial layers of the cerebral cortex and the
cases of spontaneous ICH, arterial hypertension is the leptomeninges,6-7 cerebral amyloid angiopathy (CAA)
presumed cause.1 These hypertensive hemorrhages being virtually absent in the deep grey nuclei, where
show a predilection for certain anatomic locations: the ICH of hypertensive origin predominates. CAA is a
basal ganglia (the putamen in particular), the subcorti- condition restricted to the cerebral vasculature, since it
cal white matter, and the thalamus account for 35%, is not associated with systemic vascular amyloidosis,8
25%, and 20% of ICHs, respectively, whereas the and it is almost always of sporadic occurrence, al-
posterior fossa locations, the cerebellum and pons, are though familial incidence has been documented in Ice-
responsible for only 10% and 5% of the cases.2 Irre- land9 and in the Netherlands.10 In these the condition is
spective of their anatomic locations, the common inherited as an autosomal dominant trait, and leads to
pathogenesis in these hemorrhages involves the hyper- ICH early in life, especially in the Icelandic families,
tension-induced degeneration of the media of small in which the majority of the ICHs occurred in the third
(50-200 micra) arteries called "lypo-hyalinosis,"3 and/ and fourth decades. In a group of these patients it has
or the formation of "microaneurysms", both of which beenrecentlyshown that the basic abnormality appears
occur preferentially in perforating or small sub-cortical to be in the metabolism of an alkaline microprotein
arteries, thus resulting in the characteristic anatomic called "y-trace", which is found in abnormally low
distribution of hypertensive ICH. concentrations in the CSF of these patients, as com-
A significant number of cases of ICH are due to pared with controls." An abnormality in the catabo-
causes other than hypertension. The frequency of such lism of this microprotein is thought to cause the vascu-
cases varies from 25% to 50% among series, and de- lar deposition of amyloid fibrils.
pends on the anatomic type of ICH being considered: The common sporadic variety of CAA typically af-
non-hypertensive mechanisms may account for as fects elderly individuals. Its incidence in autopsy se-
many as 55% of "lobar" ICH cases,4'3 while pontine ries increases with age, from a mere 8% in the seventh
hemorrhage is virtually always (90%) hypertensive, decade, to close to 60% in individuals older than 90.7
and the deep ganglionic (putaminal, thalamic) and An association with Alzheimer's disease has long been
cerebellar varieties occupy an intermediate position recognized, particularly in regards to a high frequency
(35%, 25%, and 38% being of non-hypertensive of some of its histopathologic features in brains of
mechanism, respectively). Among the many possible patients with CAA. In at least 50% ofreportedcases of
mechanisms of ICH not primarily related to hyperten- CAA, neuritic plaques have been documented, with a
sion, there are several that occur with high enough lower frequency of detection of neurofibrillary tan-
frequency to warrant separate discussion. These gles. The presence of clinical progressive dementia of
causes of ICH are related to special types of arterial the Alzheimer type has been reported with differing
pathology, to the presence of brain tumors, or to the frequency in various series, affecting from 10% to
use of medications, and include: cerebral amyloid an- 30% of the patients.6-12 Pathologically, CAA is charac-
giopathy, small vascular malformations, primary or terized by deposits of Congo-red positive material in
metastatic brain tumors, oral anticoagulants, and am- the media and adventitia of cortical and leptomenin-
phetamines and a variety of other sympathomimetic geal arteries. These Congo-red stained vessels show
drugs. characteristic bi-refringence under polarized light, and
also exhibit fluorescence with thioflavin T staining.
Cerebral Amyloid Angiopathy Electron microscopic studies demonstrate the typical
This condition is a unique form of angiopathy due to non-ramified 90-110 A diameter amyloid fibrils with-
deposits of amyloid in the media and adventitia of in the vessel wall. These lesions often lead to thicken-
ing of the vascular wall, with stenosis or obstruction of
the lumen,resultingin small foci of necrosis (infarcts),
which are rarely if ever symptomatic. The only consis-
From the Department of Neurology, Boston University School of
MediciDe, Boston, Massachusetts. tent clinical result of CAA is ICH following rupture of
Address correspondence to: Carlos S. Kase, M.D., Department of an affected artery, due to either "weakening" of the
Neurology, Boston University School of Medicine, 720 Harrison Ave- wall by the amyloid deposits, or to rupture of a secon-
nue, Suite 604, Boston, Massachusetts 02118. dary "microaneurysm" developed at sites of amyloid
Received September 10, 1985; revision #1 accepted December 24, deposition.6'l2
1985.
Downloaded from http://stroke.ahajournals.org/ by guest on October 15, 2012
3. INTRACEREBRAL HEMORRHAGE: NON-HYPERTENSIVE CAUSES/Kase 591
The ICHs from CAA characteristically occur in su- hematoma. The histologic type of vascular malforma-
perficial, subcortical or "lobar" locations, since the tions leading to ICH most frequently corresponds to
angiopathy selectively affects arteries of the cortical AVMs or venous angiomas, the cavernous angioma
surface and leptomeninges. Furthermore, these sub- variety being rarely associated with bleeding. AVMs
cortical ICHs have shown, in some series, a predilec- usually become symptomatic at a relatively early age,
tion for the posterior aspects of the cerebral hemi- most commonly in the third and fourth decades, thus
spheres,reflectinga generally heavier concentration of representing a potential source of ICH in non-hyper-
CAA-affected arteries in the parietal and occipital tensive young populations.l8 A striking female prepon-
lobes.6'13 In some instances, subarachnoid or subdural derance of these small vascular malformations has
hemorrhages have coexisted, again reflecting the su- been reported in several series.17'l8
perficial location of the vascular lesions. An additional Their clinical presentation is that of ICH in any
feature of ICH related to CAA has been a tendency to location, deep or superficial, without specific clinical
producerecurrenthemorrhages over periods of months features, other than perhaps a relatively slower course
or years,14 occasionally even leading to simultaneous of development, as compared with the more abrupt and
acute intracerebral hematomas in two different brain rapidly evolving course of hypertensive ICH. The
locations. In a number of published reports a history of hemorrhages produced by these lesions tend to be more
head trauma or, less commonly, a prior neurosurgical often at the level of the cerebral convexity (subcortical
procedure has preceded the onset of CAA-related ICH. white matter) than in the deep portions of the hemi-
This raises the possibility that some of these hemor- sphere, reflecting their usually more superficial loca-
rhages may be related to trauma, and the potential for tion, a feature also documented for non-ruptured small
such complication should be recognized when a neuro- cerebral vascular malformations incidentally found at
surgical procedure (such as ventricular shunt insertion, autopsy. Due to their small size, the usual absence of
brain biopsy) is indicated for elderly demented pa- symptoms prior to the onset of ICH, and the difficulty
tients. Furthermore, the need for meticulous hemosta- in diagnosing them in life, these malformations have
sis following surgical evacuation of ICHs in the elderly been called "cryptic" by some authors.16 On occasions
cannot be overemphasized. Some authors12 have con- there is a family history of such vascular malforma-
sidered ICH from CAA a condition with poor vital tions and ICH," but they occur more commonly on a
prognosis, on account of a generally larger size of the sporadic basis. Since their documentation can result in
hemorrhages, and their tendency to recur. It is possible successful surgical therapy and prevention of recurrent
that systematic search for this etiology in specimens ICH, angiography should be considered part of the
from surgically-evacuated intracerebral hematomas routine evaluation of non-hypertensive patients pre-
will provide an accurate estimate of the true frequency senting with spontaneous ICH, in particular if the hem-
of CAA in cases of spontaneous ICH. Only then will orrhage is located in the subcortical white matter of a
its prognosis and tendency to post-operative local re- cerebral hemisphere. In a number of instances, pre-
currence be documented. operative angiography has failed to document a vas-
In conclusion, CAA is probably an important etiolo- cular malformation that was diagnosed histologically
gic factor for ICH in non-hypertensive elderly individ- following surgical evacuation of an intracerebral he-
uals who present with single or recurrent hemorrhages matoma. l 7 '" These patients have usually had small
of subcortical "lobar" location. The actual frequency AVMs or venous angiomas. The difficulties in demon-
of this disorder is unknown, but is probably quite high strating the latter lesion angiographically probably
in elderly populations." It is expected that increased stem from their small size and lack of multiple tortuous
awareness of this condition will lead to more accurate vascular channels, as well as the absence of arterioven-
estimates of its frequency in surgical and autopsy ous shunting of blood leading to marked dilatation of
specimens, and this effort may eventually result in venous structures. In the case of small AVMs, lack of
effective measures for its treatment and prevention. angiographic demonstration may at times reflect com-
pression by an adjacent hematoma, since repeat angio-
grams following resorption of the hematoma have oc-
Small Vascular Malformations casionally disclosed mem. Spontaneous thrombosis
These correspond to previously unsuspected small may be an additional reason for lack of angiographic
intracerebral arteriovenous malformations (AVMs), visualization of malformations. Due to the potentially
cavernous angiomas, or venous angiomas, that present negative angiograms in the setting of an acute ICH due
as spontaneous ICH in adult life. Their reported fre- to one of these malformations, some authors have rec-
quency in series of ICH has varied between 4 and 8% ommended surgical evacuation of intracerebral hema-
of the cases.16' " The diagnosis is usually established tomas in young non-hypertensive patients,17 since the
by cerebral angiography or, more commonly, follow- resection of this type of lesion will likely prevent ICH
ing histologic study of biopsy specimens taken at the recurrence.
time of surgical evacuation of the ICH. With the intro-
duction of CT scanning, an increasing number of these
lesions are being diagnosed by the demonstration of Brain Tumors
calcium deposits in their vicinity or, more commonly, Hemorrhage into a brain tumor is a relatively rare,
by the post-contrast opacification of abnormal serpigi- but well documented, cause of non-traumatic, non-
nous vascular channels adjacent to an intracerebral hypertensive ICH. In series of brain tumors, ICH oc-
Downloaded from http://stroke.ahajournals.org/ by guest on October 15, 2012
4. 592 STROKE VOL 17, No 4, JULY-AUGUST 1986
curs in less than 1% of these cases, whereas underlying acute effects of the ICH leading to uncal herniation and
tumors in ICH series are found in 2 to 6% of the brainstem compression or, less commonly, to the pro-
cases,2021 occasionally in as many as 10% of the gression of the underlying malignant tumor.23
cases.22 These are instances of ICH in either a pre-
viously unknown cerebral neoplasm or, more com- Oral Anticoagulants
monly, as a complication of a known brain tumor. As a Warfarin sodium, a widely used oral anticoagulant
result, ICH produces the acute onset of either a new for the prevention of venous and arterial origin embo-
focal neurologic deficit, or worsening of pre-existing lism, is associated with bleeding complications in ap-
focal deficits, in both instances commonly associated proximately 7-8% of patients.2*-27 Intracranial hem-
with deterioration in the level of consciousness. The orrhage accounts for a small fraction of these
brain tumors likely to present as ICH are largely malig- hemorrhages, amounting to only 0.5-1.5% of all
nant, either primary astrocytoma (glioblastoma multi- bleeding events related to warfarin. However, these
forme) or metastatic, most commonly bronchogenic intracranial hemorrhages have a generally dismal
carcinoma, melanoma, choriocarcinoma, or renal-cell prognosis, thus resulting in a significant contribution
carcinoma.2124 Benign brain tumors such as meningio- to the fatal complications of oral anticoagulant
mas or oligodendrogliomas have rarely been reported therapy.
as presenting with episodes of non-traumatic ICH. The The common intracranial sites of bleeding in orally
bleeding potential of malignant tumors is thought to be anticoagulated patients are, in decreasing order of fre-
related to their tendency toward spontaneous necrosis quency, the subdural space, brain parenchyma, and
and to the richness and neoplastic character of their subarachnoid space. ICH in this setting represents a
vasculature, as well as the biologic tendency of some condition with some distinct clinical characteristics
tumors such as choriocarcinoma to invade the walls of and course, as well as high mortality and serious long-
blood vessels. Metastases from this tumor are note- term sequelae in survivors. Oral anticoagulation has
worthy for their tendency to become hemorrhagic in at been estimated to increase between 8-fold28 and 11-
least 50% of the cases.24 fold29 the risk of ICH, as compared with non-anti-
The sites of ICH relate in some degree to the type of coagulated individuals with similar risk factors for
underlying tumor, since deeply-seated white matter intracranial hemorrhage. A number of possibly con-
tumors such as glioblastoma multiforme will produce tributing factors to ICH have been suggested in this
deep hemispheric hemorrhages, while those resulting patient population: A) Age. Several studies have indi-
from metastatic tumors are more often cortico-subcor- cated a low frequency of ICH in patients younger than
tical, reflecting the predilection of secondary tumors 50, and an increased risk for ICH as a function of
for the superficial portions of the cerebral hemi- increasing age,29 some suggesting age 65 as a point of
spheres. Most commonly, these hemorrhages originate sharp rise in the risk.28 Although other series have
at the margins of the tumor, or between the tumor and failed to show a clear relationship between increasing
the adjacent edematous brain parenchyma.21 Instances age and higher risk of bleeding in anticoagulated pa-
in which an underlying tumor is suspected in a patient tients,2* we believe that this form of therapy should be
with ICH include: (a) History of preceding chronic given to patients who are older than 70 years of age
headache and/or focal neurologic deficit and/or per- only after close scrutiny of its indications and the
sonality change for days or weeks prior to the onset of subject's ability to comply with its proper use and
ICH, or the finding of papilledema on initial presenta- monitoring. B) Hypertension. The contribution of this
tion with ICH; (b) The presence of multiple separated factor to warfarin-related ICH has also been controver-
foci of ICH occurring simultaneously; (c) An area of sial, some series strongly suggesting a relationship,29
"ring-like" hemorrhage with a low-density center in while others have failed to document it.28 As a result,
non-contrast CT scan; (d) An ICH that on CT scan there are no defined guidelines for the indications or
appears as irregular, mottled high-density, and affect- contraindications for oral anticoagulants in hyperten-
ing structures that are rarely involved in hypertensive sives. It is our policy to exclude from chronic warfarin
ICH, such as the corpus callosum, which on the other therapy patients who continue to have severe and labile
hand is frequently affected in glioblastoma multi- hypertension in the face of full compliance with maxi-
forme; (e) A disproportionate degree of surrounding mal anti-hypertensive therapy. C) Preceding cerebral
edema and mass effect associated with the hematoma; infarction has been considered by some to be important
(f) Presence of post-contrast enhancement in the vi- in the pathogenesis of anticoagulant-related ICH.30
cinities of the acute high-density ICH. An angiogram However, recent large series29-3I have failed to support
can be useful in demonstrating a mass lesion with the such an association: in a group of 24 patients with ICH
classical "tumor blush" characteristic of highly vascu- in the course of warfarin therapy,31 only 1 bled into the
lar primary or metastatic brain tumors. In many in- area of the brain that had 3 weeks before been affected
stances the diagnosis is only suspected by the finding by an embolic infarct; all others with prior embolic
of clinical or radiologic signs of a systemic malignan- infarcts (8 patients) had ICH in a vascular territory
cy, or it is established by biopsy of the hematoma different from that involved previously. Although an
cavity following its surgical evacuation. The prognosis embolic infarct can acutely become complicated by the
in this form of ICH is poor, with short-term (days to formation of a hematoma in the setting of heparin or
weeks) mortality in the 90% range, mostly due to the warfarin anticoagulation, this phenomenon is uncom-
Downloaded from http://stroke.ahajournals.org/ by guest on October 15, 2012
5. INTRACEREBRAL HEMORRHAGE: NON-HYPERTENSIVE CAVSES/Kase 593
mon, and appears to be limited to large size embolic noted in one series,31 and that location was also over-
infarcts.32 D) Excessive prolongation of the prothrom- represented in a group of 13 cases gathered from the
bin time. With very few exceptions,13 this factor is literature.30-3739 This anatomical feature of warfarin-
considered to be a consistent feature in most hemorrha- related hemorrhages has no clear explanation, as it is
gic complications in orally anticoagulated patients. also unclear what the actual pathogenesis of these hem-
ICH is no exception to this rule: fully 80% of patients orrhages is, in terms of the size and type of the ruptured
with ICH have excessively prolonged prothrombin vessels leading to the hemorrhage. It is possible that
times (PT) at the onset of hemorrhage. One major these hemorrhages arerelatedto bleeding from vessels
difficulty in clearly establishing this fact from the lit- different from thoseresponsiblefor hypertensive ICH.
erature is the lack of agreement on what the "therapeu- Serial microscopic sections of pathologic specimens
tic" PT range is: some authors recommend a prolonga- will be required to clarify these points.
tion of the PT to 1 Vi to 2Vi times control,28' "• M while The overall prognosis in these ICHs is poor, with
others use strictly I1/- times control.233 In studies of mortalities in the range of 65% of the cases.29'3I This
warfarin therapy for prevention of venous thromboem- high mortality correlates with generally large size he-
bolism, there is evidence that a "low-dose" schedule of matomas, probably reflecting the slow but eventually
warfarin (enough to prolong the PT to 1VA control) is massive extravasation of blood into the parenchyma as
associated with equal protection, but significantly few- a result of the drug-induced coagulation defect. De-
er bleeding complications, than a "conventional" spite the routine use of vitamin K and fresh frozen
schedule with PT prolonged to 1 Vi to 2 times control.36 plasma for the rapidreversalof the coagulation defect,
Similar comparisons of different schedules of oral anti- neurologic deterioration and fatal outcome is the ex-
coagulation are not available for the prevention of arte- pected course in two-thirds of the cases.
rial thromboembolism. However, extrapolation from
the venous thromboembolism data suggest that it is Use of Amphetamines and Other
prudent to recommend adherence to "conservative" Sympathomimetic Drugs
levels of warfarin anticoagulation, in the range of 1 Vi
times control, for the prevention of bleeding complica- A number of examples of ICH secondary to the use
tions. E) Duration of anticoagulant treatment. This and abuse of amphetamines and related drugs have
factor has not shown a clear association with risk of been documented. The most commonly implicated
ICH, since in some series the duration of therapy has preparation has been methamphetamine by the intrave-
most often (in 65% of the cases) exceeded 1 year,29 nous route,40 but examples of ICH following its intra-
while in others as many as 70% of the events occurred nasal and oral use are also in record. Less frequently,
within the first year of treatment.31 In a group of 12 the responsible substances have been amphetamine
cases gathered from several reports, the cases were and pseudoephedrine. ICH has generally occurred
evenly distributed below and above 1 year of therapy shortly after use of the drug, within minutes to a few
when the ICH occurred.3037-39 F) Head trauma does hours after exposure, and the affected individuals have
not appear to play a role in ICH in the setting of oral been in general chronic users, although occasional ex-
anticoagulation: only 4 of 24 patients (16%) had a amples have followed alleged first-time use. The ma-
preceding history of trauma in a series of anticoagu- jority of the hematomas have been located in the sub-
lant-related ICH.31 In all 4 instances the traumatic epi- cortical white matter of the cerebral hemispheres, only
sode was considered mild in nature, and was not asso- occasional ones occurring at the level of the basal
ciated with loss of consciousness. ganglia.4142
An association with transiently elevated blood pres-
The clinical presentation of anticoagulant-related sure has been noted in approximately 50% of the cases,
ICH has some distinctive features. A gradual and slow and thisrepresentsa likely etiologic mechanism. How-
progression of the focal signs was recorded in 58% of ever, angiographic changes suggestive of vasculitis
the cases in one series, and one-half of these pro- ("arteritis") have also been documented, raising the
gressed over exceedingly long periods of time, of 24, possibility of a drug-induced angiopathy as the etiolo-
48, and even 72 hours.31 In some instances, this slowly gic factor.43 This view is further supported by the re-
progressive course could be correlated with CT-detect- ported disappearance of such angiographic changes
ed increase in size of the ICH. This feature contrasts following drug discontinuation or the use of steroids.
with the usual course of hypertensive ICH, in which This angiopathy, also called "speed arteritis", is char-
such a protracted initial course is rarely observed, and acterized angiographically by multiple focal areas of
coincident enlargement of the hematoma by CT occurs stenosis or constriction of medium-size intracranial
exceptionally following admission (observed in only 2 arteries. Pathologic examination of these vessels has
of a personal consecutive series of 100 cases of ICH). shown a necrotizing angiitis similar to periarteritis no-
This extended initial course suggests a process of slow dosa characterized by fibrinoid degeneration and ne-
bleeding into the parenchyma, different from the usu- crosis of the media and intima of medium-size and
ally catastrophic course in patients with hypertensive small arteries and arterioles, associated with variable
ICH. An additional feature in anticoagulant-related degrees of inflammatory leukocytic infiltration of the
hemorrhages has been an apparent difference in their vessel walls. At a later, reparative phase of the angiop-
topographic distribution, as compared with hyperten- athy, collagen replacement of muscular and elastic
sive ICH: a relative predilection for the cerebellum was tissue can follow, at times resulting in the formation of
Downloaded from http://stroke.ahajournals.org/ by guest on October 15, 2012
6. 594 STROKE VOL 17, No 4, JULY-AUGUST 1986
aneurysmal dilatations of the arterial wall.43 These vas- 6. Okazaki H, Reagan TJ, Campbell RJ: Clinicopathologic studies of
cular changes are considered to be secondary to either primary cerebral amyloid angiopathy. Mayo Clin Proc 54: 22-31,
1979
a direct "toxic" effect of the drug or a hypersensitivity 7. Vinters HV, Gilbert JJ: Cerebral amyloid angiopathy: Incidence
reaction to the drug or its vehicle. In a few isolated and complications in the aging brain II. The distribution of amyloid
instances, angiography has revealed a coincidental vascular changes. Stroke 14: 924-928, 1983
AVM or aneurysm as the source of hemorrhage, which 8. Gilbert JJ, Vinters HV: Cerebral amyloid angiopathy: Incidence
has been parenchymatous or subarachnoid, respective- and complications in the aging brain I. Cerebral hemorrhage.
Stroke 14: 915-923, 1983
ly.44' 45 In these instances, the apparent role of the drug 9. Gudmundsson G, Hallgrimsson J, Jonasson TA, Bjarnason O:
has been that of the precipitant rather than the actual Hereditary cerebral haemorrhage with amyloidosis. Brain 95:
cause of the hemorrhage. 387^04, 1972
The usual therapy for this variety of ICH has been 10. Wattendorff AR, Bots GTAM, Went LN, Endtz LJ: Familial cere-
bral amyloid angiopathy presenting as recurrent cerebral haemor-
the use of high-dose steroids, occasionally accom- rhage. J Neurol Sci 55: 121-135, 1982
panied by immunosuppressant drugs (cyclophospha- 11. Grugg A, Jensson O, Gudmundsson G, Amason A, Lofberg H,
mide). The clinical picture has followed the expected Malm J: Abnormal metabolism of y-trace alkaline microprotein:
course of slow resolution of the intracerebral hemato- The basic defect in hereditary cerebral hemorrhage with amyloido-
sis. New Engl J Med 311: 1547-1549, 1984
ma, and follow-up angiograms have often shown an
12. Kalyan-Raman UP, Kalyan-Raman K: Cerebral amyloid angiop-
improvement or disappearance of the signs of vasculi- athy causing intracranial hemorrhage. Ann Neurol 16: 321-329,
tis, attesting to its reversible character. 1984
Phenylpropanolamine, a structural analog of am- 13. Tomonaga M: Cerebral amyloid angiopathy in the elderly. J Amer
phetamine contained in over-the-counter appetite con- Geriatr Soc 29: 151-157, 1981
14. Finelli PF, Kessimian N, Bernstein PW: Cerebral amyloid angiop-
trol and decongestant preparations, has been associat- athy manifesting as recurrent intracerebral hemorrhage. Arch
ed with ICH in previously healthy individuals.46-w The Neurol 41: 330-333, 1984
hematomas have occasionally been multiple and si- 15. Drury I, Whisnant JP, Garraway WM: Primary intracerebral hem-
multaneous,47 and a "vasculitic" picture like that seen orrhage: Impact of CT on incidence. Neurology 34: 653-657, 1984
16. Crawford JV, Russell DS: Cryptic arteriovenous and venous ham-
in amphetamine-related hemorrhages has also been artomas of the brain. J Neurol Neurosurg Psychiat 19: 1-11, 1956
documented. **• ** 17. Steiger HJ, Tew JM: Hemorrhage and epilepsy in cryptic cerebro-
Illicit drugs have occasionally been associated with vascular malformations. Arch Neurol 41: 722-724, 1984
episodes of ICH closely following their use. These 18. Becker DH, Townsend JJ, Kramer RA, Newton TH: Occult cere-
include cocaine,49 and the combination Talwin-pyri- brovascular malformations. A series of 18 histologically verified
cases with negative angiography. Brain 102: 249—287, 1979
benzamine ("T's and blues").30 These drugs have at 19. Russell DS, Rubinstein LJ: Pathology of Tumors of the Nervous
times resulted in the production of cerebral infarction System, Fourth edition, Williams and Wilkins Co., Baltimore, pp
instead of ICH, and both lesions are thought to be the 116-145, 1977
result of an angiopathy due to the drug or some of its 20. Russell DS: The pathology of spontaneous intracTanial haemor-
rhages. Proc Roy Soc Med 47: 689-693, 1954
vehicles. 21. Little JR, Dial B, Bellanger G, Carpenter S: Brain hemorrhage
In conclusion, this review of non-hypertensive from intracranial tumor. Stroke 10: 283-288, 1979
causes of ICH suggests that these various pathological 22. Scott M: Spontaneous intracerebral hematoma caused by cerebral
entities account for a significant number of cases of neoplasms. Report of eight verified cases. J Neurosurg 42:
ICH. It is possible that theirrelativecontribution to the 338-342, 1975
23. Mandybur TI: Intracranial hemorrhage caused by metastatic tu-
total group of ICH cases will even increase in the mors. Neurology 27: 650-655, 1977
future, as the frequency of the hypertensive form of 24. Gildersleve N, Koo AH, McDonald CJ: Metastatic tumor present-
ICH is likely to continue to decline, reflecting im- ing as intracerebral hemorrhage. Radiology 124: 109-112, 1977
proved control of hypertension in the population at 25. Bitoh S, Hasegawa H, Ohtsuki H, Obashi J, Fujiwara M, Sakurai
M: Cerebral neoplasms initially presenting with massive intracere-
risk. It is hoped that increased awareness, as well as bral hemorrhage. Surg Neurol 22: 57-62, 1984
improved diagnostic methods, willresultin the clinical 26. Coon WW, Willis PW: Hemorrhagic complications of anticoagu-
diagnosis of these conditions, which are currently for lant therapy. Arch Int Med 133: 386-392, 1974
the most part diagnosed only pathologically. This im- 27. Forfar JC: A 7-year analysis of haemorrhage in patients on long-
provement in diagnosis is likely to provide a better term anticoagulant treatment. Brit Heart J 42: 128-132, 1979
28. Furlan AJ, Whisnant JP, Elveback LR: The decreasing incidence of
estimate of their true frequency in the ICH population. primary intracerebral hemorrhage: A population study. Ann Neurol
5: 367-373, 1979
References 29. Wintzen AR, de Jonge H, Loeliger EA, Bots GTAM: The risk of
1. Mohr JP, Caplan LR, Melski JW, et al: The Harvard Cooperative intracerebral hemorrhage during oral anticoagulant treatment: A
Stroke Registry: A prospective registry. Neurology 28: 754-762, population study. Ann Neurol 16: 553-558, 1984
1978 30. Lieberman A, Hass WK, Pinto R, et al: IntracTanial hemorrhage
2. Mohr JP, Kase CS, Adams RD: Cerebrovascular Disorders. In: and infarction in anticoagulated patients with prosthetic heart
Harrison's Principles of Internal Medicine, 10th edition, RG Pe- valves. Stroke 9: 18-24, 1978
tersdorf et al (Eds), Chapter 356, pp 2028-2060. McGraw-Hill, 31. Kase CS, Robinson RK, Stein RW et al: Anticoagulant-related
New York, 1983 intracerebral hemorrhage. Neurology 35: 943-948, 1985
3. Fisher CM: Pathological observations in hypertensive cerebral 32. Hart RG, Lockwood KI, Hakim AM et al: Immediate anticoagula-
hemorrhage. J Neuropath Exper Neurol 30: 536-550, 1971 tion of embolic stroke: Brain hemorrhage and management options.
4. Kase CS, Williams JP, Wyatt DA, Mohr JP: Lobar intracerebral Stroke 15: 779-789, 1984
hematomas: Clinical and CT analysis of 22 cases. Neurology 32: 33. Miale JB: Laboratory Medicine: Hematology, 5th ed, chap 17, p.
1146-1150, 1982 959. CV Mosby, St. Louis, 1977
5. RopperAH, Davis KR: Lobar cerebral hemorrhages: Acute clinical 34. Braunwald E: Heart Disease, 2nd ed, chap 32, p. 1087. WB
syndromes in 26 cases. Ann Neurol 8: 141-147, 1980 Saunders, Philadelphia, 1984
Downloaded from http://stroke.ahajournals.org/ by guest on October 15, 2012
7. THROMBOLYTIC THERAPY IN STROKE/^/ Zoppo et al 595
35. Wessler S, Gitel SN: Warfarin: From bedside to bench. New Engl J 43. Citron BP, Halpern M, McCarron M, et al: Necrotizing angiitis
Med 311: 645-652, 1984 associated with drug abuse. New Engl J Med 283: 1003-1011,
36. HullR, HirshJ, Jay R.etah Different intensities of oral anticoagu- 1970
lant therapy in the treatment of proximal-vein thrombosis. New 44. Lukes SA: Intracerebral hemorrhage from an arteriovenous malfor-
Engl J Med 307: 1676-1681, 1982 mation after amphetamine injection. Arch Neurol 40: 60—61, 1983
37. Barron KD, Fergusson G: Intracranial hemorrhage as a complica- 45. Maticlc H, Anderson D, Brumlik J: Cerebral vasculitis associated
tion of anticoagulant therapy. Neurology 9: 447-455, 1959 with oral amphetamine overdose. Arch Neurol 40: 253-254, 1983
46. Fallis RJ, Fisher M: Cerebral vasculitis and hemorrhage associated
38. Iizuka J: Intracranial and inlraspinal haematomas associated with
with phenylpropanolamine. Neurology 35: 405—407, 1985
anticoagulant therapy. Neurochirurgia 1: 15-25, 1972
47. Kikta DG, Devereaux MW, Chandar K: Intracranial hemorrhages
39. Snyder M, Renaudin J: Intracranial hemorrhage associated with due to phenylpropanolamine. Stroke 16: 510-512, 1985
anticoagulation therapy. Surg Neurol 7: 31-34, 1977 48. Stoessl AJ, Young GB, Feasby TE: Intracerebral haemorrhage and
40. Delaney P, Estes M: Intracranial hemorrhage with amphetamine angiographic beading following ingestion of catecholaminergics.
use. Neurology 30: 1125-1128, 1980 Stroke 16: 734-736, 1985
41. Harrington H, Heller HA, Dawson D, Caplan L, Rumbaugh C: 49. Caplan LR, Hier DB, Banks G: Stroke and drug abuse. Current
Intracerebral hemorrhage and oral amphetamine. Arch Neurol 40: Concepts of Cerebrovascular Disease/Stroke 17: 9-14, 1982
503-507, 1983 50. Caplan LR, Thomas C, Banks G: Central nervous system compli-
42. Salanova V, Taubner R: Intracerebral haemorrhage and vasculitis cations of addiction to "T's and blues". Neurology 32: 623-628,
secondary to amphetamine use. Postgrad Med J 60:429-430, 1984 1982
Thrombolytic Therapy in Stroke: Possibilities and Hazards
GREGORY J. DEL ZOPPO, M.D., HERMANN ZEUMER, M.D.,* AND LAURENCE A. HARKER, M.D.
AGENTS WHICH MEDIATE THE dissolution of patients with various thrombolytic agents, and to
thrombi are receiving increasingly wide therapeutic weigh the relative risk of intracerebral hemorrhage in
application. Urokinase or streptokinase have been em- patients treated with fibrinolytic agents for stroke and
ployed in the treatment of acute thrombosis of coro- for other thrombotic disorders.
nary1"23 and selected peripheral arteries,15' 23~34 traumat-
ic internal carotid artery occlusion,35 as well as of Mechanism of Thrombolysis
pulmonary embolism36"39 and peripheral deep venous Arterial thrombosis and thrombus extension involve
thrombosis.1340"39 to varying degrees the processes of endothelial injury,
The demonstration that acute stroke is typically an platelet aggregation and release, and thrombin genera-
atherothrombotic or thromboembolic process60"76 pro- tion. Thrombin-mediated fibrinogen cleavage results
vides a theoretical basis for the use of thrombolytic in fibrin formation which is required for thrombus sta-
therapy in the treatment of acute stroke. However, bilization.80 Thrombin-mediated fibrin formation oc-
because of the possibility that intracerebral hemor- curs in direct relation to platelet activation by several
rhage may develop during thrombolytic therapy, use of mechanisms. Platelets promote activation of the early
such agents in stroke treatment has generally been con- stages of intrinsic coagulation by a process that in-
traindicated. Nevertheless, limited recent experience volves a factor XI receptor and high molecular weight
indicates that careful infusion of thrombolytic agents kininogen.81 Also, factors V and VHI interact with
may lead to thrombus dissolution and clinical improve- platelet membrane phospholipids to facilitate the acti-
ment in selected patients presenting with acute vation of factor X to Xa and the conversion of pro-
stroke.77"79 thrombin to thrombin.82 Platelet-bound thrombin-
It is the purpose of this discussion to review the modified factor V (factor Va) serves as a high affinity
molecular basis for the thrombolytic state, clinical platelet receptor for factor Xa.83 Consequently, the rate
experience with systemic and local treatment of stroke of thrombin generation is accelerated 103 fold, provid-
ing a potent positive feedback mechanism for initiation
From The Roon Research Center for Cardiovascular Disease and
of thrombin formation on the platelet surface, fibrin
Thrombosis, Department of Basic and Clinical Research, Scripps Clime network formation in the thrombus, and indirectly,
and Research Foundation, 10666 North Torrey Pines Road, La Jolla, fibrinolysis.
California U.S.A. 92037, and from the Abteilung: Neurologie/Neurora- Thrombus dissolution is, in large part, mediated by
diologie, Rheinische — Westphalische Technische Hochschule, Aa-
chen, Pauwelsstrasse, D5100, Aachen, Federal Republic of Germany.*
fibrinolysis localized within the thrombus.84"86 Fibrin
This work was supported in part by research grant HL 31950 from the (and fibrinogen) degradation is catalyzed by plasmin,
National institutes of Health. the product of plasminogen activation.87 In the consoli-
Address correspondence to: Gregory J. Del Zoppo, M.D., Roon dating thrombus plasminogen binds to fibrin and to
Research Center for Cardiovascular Disease and Thrombosis, Depart- platelets, allowing local release of plasmin within the
ment of Basic and Clinical Research, Scripps Clinic and Research
Foundation, 10666 North Torrey Pines Road, La Jolla, California thrombus. The circulating plasminogen activators, tis-
U.S.A. 92037. sue plasminogen activator (tPA) and single chain uro-
Received February 6, 1986; revision #1 accepted May 12, 1986. kinase plasminogen activator (scuPA), catalyze plas-
Downloaded from http://stroke.ahajournals.org/ by guest on October 15, 2012