Aneurysm & AVM Center
Hemorrhage of Unknown Etiology
Etiology of angiogram-negative subarachnoid hemorrhage
| Recent advances
Figure | Definition:
The diagnosis of subarachnoid hemorrhage | Other sources
to twenty percent of patients with a spontaneous subarachnoid hemorrhage will
have a normal cerebral angiogram and no source of hemorrhage apparent on other
neuroimaging studies. Subarachnoid hemorrhage with an initially negative cerebral
angiogram has a more benign prognosis than untreated aneurysmal subarachnoid hemorrhage.
Nonetheless, a small percentage of such patients experience morbidity or mortality
because of recurrent hemorrhage presumably from an undetected aneurysm. This has
lead to the common practice of repeating angiography in patients whose initial
imaging studies reveal no vascular malformations.
At the M.G.H. Aneurysm/AVM Center a second angiogram is usually obtained approximately
two weeks after subarachnoid hemorrhage in patients whose initial studies do not
reveal a source of bleeding. A third angiogram is obtained several months later
if the clinical suspicion is high or if vasospasm was observed in one or both
of the initial studies. Magnetic resonance imaging of the entire neuraxis is often
performed to rule out other sources of subarachnoid bleeding including tumors
and occult vascular malformations of the brain or spine. If the pattern of blood
on computed tomography is consistent with a spinal source, myelography and spinal
angiography are performed in an effort to detect a treatable source of future
Patients with so-called angiogram-negative subarachnoid hemorrhage are at risk
for any of the potential complications seen in aneurysmal subarachnoid hemorrhage.
These include the development of delayed neurologic deficits (stroke) due to cerebral
vasospasm, hydrocephalus, deep venous thrombosis, pulmonary embolism, and respiratory
depression. In addition there is a small but real risk of re-hemorrhage. Thus,
these patients usually require several weeks of careful monitoring to assure adequate
blood pressure control and prompt treatment of vasospasm. They also require long
term follow-up to rule out the late onset of hydrocephalus if the amount of initial
hemorrhage is significant.
as well as early mobilization with adequate control of blood pressure are the
best current methods of prophylaxis against deep venous thrombosis and resultant
pulmonary embolism. In patients whose neurologic or medical status does not allow
mobilization, pneumatic compression stockings are an alternative. Stool softeners
have also been used traditionally, in an effort to prevent transient elevations
of arterial pressure in association with straining.
Most patients with angiogram-negative subarachnoid hemorrhage are treated with
the calcium channel antagonist Nimodipine® for 21 days to lessen the risk of a
permanent delayed neurologic deficit. The anti-fibrinolytic, epsilon amino-caproic
acid (Amicar®), is avoided in this population as most authorities agree that the
potential risks of thrombotic complications outweigh the potential benefits of
reduced re-hemorrhage rate.
A number of causes of angiographically-occult subarachnoid hemorrhage have been
observed or proposed. These include: venous or sinus thrombosis; tumors in the
brain, spine, or subarachnoid space; infections; blood dyscrasias including anticoagulants
and sporadic responses to drugs such as Prozac®; trauma; spinal or cerebral dural
or parenchymal vascular malformation; nonvisualized aneurysm-due to vasospasm
or thrombosis; and rupture of a small superficial artery.
As the majority of these sources are detected by modern imaging techniques current
consensus favors the latter two causes as the most frequent sources of angiogram
negative hemorrhage. It is likely that the two have different prognostic implications.
Rupture of a small superficial artery is presumed to result in obliteration of
the source of hemorrhage with little risk of recurrent bleeding. In distinction,
an undetected aneurysm poses a significant risk of re- hemorrhage with potentially
The pattern of subarachnoid
blood on computed tomography may give clues to the source of hemorrhage as shown
in the figure. Recent evidence indicates that blood limited to the perimesencephalic
and chiasmatic cisterns without significant extension into the Sylvian fissure
or interhemispheric fissure is unlikely to represent an aneurysmal hemorrhage.
This pattern of blood likely represents rupture and subsequent thrombosis of a
small artery which arises directly from the basilar artery or the Circle of Willis.
In distinction, diffuse subarachnoid hemorrhage with blood extending into the
Sylvian and interhemispheric fissures is more likely to represent aneurysmal hemorrhage.
Several mechanisms have been proposed
for the occasional failure of angiography to detect such aneurysms, including:
thrombosis of the aneurysm, local vasospasm, and obliteration of the aneurysm
by the rupture. In each of these cases the patient remains at risk of re-rupture
due to resumption of filling of the vascular defect by arterial blood.
Given the small risks associated with surgery
and the potentially catastrophic results of aneurysmal rupture patients with an
aneurysmal pattern of blood are frequently surgically explored. Patients who do
not undergo exploration continue to be followed periodically by neuroimaging in
an effort to detect an aneurysm. Recent experience at the Aneurysm/AVM Center
indicates that the majority of patients who undergo surgery after having an aneurysmal-
pattern of blood on computed tomography but negative angiography have an aneurysm
discovered and obliterated at surgery. Patient's undergoing these procedures have
returned to their normal activities with increased confidence in being free from
future subarachnoid hemorrhage
Figure legend. Angiogram-negative
subarachnoid hemorrhage classically displays the pattern of blood demonstrated
in the images above (top panels). Occasionally, the blood will extend beyond
the peri-mesencephalic cisterns as demonstrated on the lower two panels. Our recent
experience suggests that in many such cases exploration will reveal a small aneurysm
as the source of hemorrhage. At surgery an anterior communicating artery aneurysm
was discovered in the patient whose CT is shown in the bottom panels.
Further information, including citations to the medical literature can be found
in Tatter SB , Crowell RM, Ogilvy
CS (1995). Aneurysmal and microaneurysmal angiogram-negative subarachnoid
hemorrhage. Neurosurgery . in press.
A high percentage of patients with subarachnoid hemorrhage present acutely complaining
of the sudden onset of the worst headache of their life. This is frequently followed
by photophobia, nausea, and vomiting. Meningismus is very frequently the only
abnormality on examination. In this population current generation computed tomographic
scanners detect almost all subarachnoid hemorrhages. If the clinical suspicion
of subarachnoid hemorrhage remains high a lumbar puncture may then be performed.
Frequently the results of lumbar puncture are equivocal because of a so-called
traumatic tap, during which blood is introduced into the lumbar subarachnoid space.
The presence of xanthochromia in the spun
sample often confirms the diagnosis of subarachnoid hemorrhage in such cases.
Xanthochromia, however, only becomes apparent twelve or more hours after subarachnoid
hemorrhage. Clearing of the blood as more samples are collected often allows subarachnoid
hemorrhage to be excluded.
In cases where
the CSF RBC count does not trend toward zero and when there is no xanthochromia,
acute subarachnoid hemorrhage remains a diagnostic possibility. Currently, such
patients are often managed as if they have had a subarachnoid hemorrhage. In the
future, it is likely that sensitive biochemical tests-such as determination of
CSF D-dimer-will allow more patients to be spared the risks of cerebral angiography
and the expense of hospitalization and treatment with calcium channel blockers.