Most of the site will reflect the ongoing surgical activity of Prof. Munir Elias MD., PhD. with brief slides and weekly activity. For reference to the academic and theoretical part, you are welcome to visit  neurosurgery.me



Spontaneous Intracranial Hemorrhage

A spontaneous intracerebral hematoma (SICH) is a clot that arises without immediately preceding trauma. It may be primary not due to other specific disease except indirectly as in the case of hypertension), or secondary (caused by a variety of congenital and acquired conditions). Interest in so-called apoplexy goes back to the dawn of written history. Apoplexy is a Greek word meaning "struck with violence as if by a thunderstorm" from which derives stroke, meaning a sudden loss of the senses with paralysis and the secondary fear it engenders. Indeed, there are said to be reference to sanguineous apoplexy in the Hippocratic writings. John Abercrombie categorized apoplexy into infarction, intracerebral hemorrhage, and subarachnoid hemorrhage. Durer 1874 believed that the lenticulostriate artery was responsible for cerebral hemorrhage. Charcot and Bouchard (1868) described miliary aneurysm in hypertension, although the significance of these structure, was argued for many years. Many cases of SICH in various anatomic locations were described in the late nineteenth and early twentieth centuries including hemorrhages in the cerebellum and pons. Westphal (1925) suggested that hypertension caused SICH. Sir Charles Symonds (1931) emphasized that intracerebral hemorrhage should be distinguished from subarachnoid hemorrhage. Russell and Cole and Yates discussed the relationships of microaneurysms to hypertension and hemorrhage, including their common distribution. Fisher expanded this discussion and also discovered that disruption of nearby vessels contributed to the enlargement of the hemorrhage.

Macewen (l888) reported the first successful surgery for SICH (1883). By 1891 Chamboniere had reported 31 cases of SICH. Likewise, Lucas also had reported several cases. Cushing in 1903 successfully evacuated an SICH. In 1906 Ballance described the first successful evacuation of a cerebellar hemorrhage. Bagley in 1932 differentiated infiltrating deep hemorrhages from confluent superficial hemorrhages and suggested that the latter could be removed. Penfield in 1933 stated that solid clots could not be successfully aspirated. A number of papers were published on surgical evacuation as well as aspiration. However, the reports of McKissock and colleagues around 1960 suggested that surgical evacuation produced worse outcomes than nonoperative treatment.  Neurosurgeons seemed to accept that little could be done to alter the natural history of SICH and consequently lost interest in this problem.

However, interest slowly increased, possibly because of the high incidence in such places as Japan, and new experimental and clinical studies were begun a little over three decade ago. Much has been learned about the origin of SICH and its manifestations through human autopsies. The pathophysiology of SICH is being better understood through the use of animal studies. It has been often difficult to clinically differentiate infarctions from clots: however, clots can now be accurately diagnosed by computed tomography (CT) or magnetic resonance imaging (MRI), which also has permitted clinical/anatomic correlation of their clinical manifestations. Many changes in the population as well as changes in health care, ranging from improved treatment of hypertension to the increased use of cocaine and an aging population with a higher incidence of amyloid angiopathy, have changed the total incidence of SICH and relative incidence from different etiologies. The epidemiology of SICH cannot be fully understood, however, until all patients in large populations suffering strokes are studied by neuroimaging and/or autopsy. Since a large percent of victims of SICH die immediately or soon after their hemorrhage, prevention is the ultimate solution of the problem. Given the current situation, the question is how to optimize care. There is increased interest in treatment, both by controlling intracranial pressure and by evacuating the hematomas either with open surgery or by aspiration.


The epidemiology of SICH is problematic, although there have been many epidemiologic studies of stroke in general and SICH in particular. Stroke is the third leading cause of death in the United States, accounting for 2 to 4 percent of all deaths. SICH is the cause of 8 to 13 percent of strokes and 15 to 20 percent of death, due to strokes. There are about 37,000 incidences of SICH annually in the United States. Hypertension is a contributing factor to at least 50 percent of SICH, with amyloid angiopathy beginning to assume more significance as the population ages. Risk factor, include age, race, hypertension, prior cerebral infarction, coronary artery disease, diabetes mellitus, and a variety of diseases. The incidence of strokes, including SICH, declined in the 1970s, at least partly due to the more frequent detection and effective treatment of hypertension.

Attempts to be more precise in detection are problematic for many reasons. In order to properly understand the incidence of SICH, we would need national figures to eliminate population and all cases would have to be diagnosed and all contributing factors would have to be identified. Current information does not meet these standards, however. Some hemorrhages are minimally symptomatic and not reported; small ones may be diagnosed as infarcts if neuroimaging is not done . Large hematomas may be diagnosed but their cause not identified (e.g., hemorrhages into the basal ganglia from aneurysms). Patients may die suddenly from SICH but since they are not autopsied their deaths are attributed to other causes (it is thought that intracranial bleeding causes 10 to 15 percent of sudden deaths). On the other hand, patients may die of conditions such as cardiac disease but be presumed to have suffered SICH. Regional statistics, where populations are atypical with regard to race, age, and socioeconomic factors, may not be representative of national statistics. Additionally, medical knowledge and practices are changing. For example, drug therapy for hypertension may be less aggressive, more hemorrhages may be detected because more stroke patients have CT scans, and less patients may be dying of other diseases and thus develop strokes. Therefore, older statistics may not reflect current trends. Thus, when considering epidemiology, it is important to be aware of these problems and the limits of our current knowledge.


A unified theory for the etiology of SICH has been proposed based on acute increase in blood flow in areas of normal or ischemic arterioles or capillaries (or other vessels), or damage to penetrating blood vessels by chronic arterial hypertension. This can be applied to most specific etiologies. Larger vessels may also be subject to weakening, or insignificant acute injuries may not seal in the face of impaired hemostatic mechanisms.

Table-1 outlines the etiologies of SICH. There are, however, some obvious overlaps: the bleeding diathesis in disseminated intravascular coagulation (DIC) contributes to delayed posttraumatic SICH. anticoagulants contribute to hemorrhages after cardiac surgery that are related to emboli, and trauma and infection lead to aneurysms. Etiologies listed after cerebral amyloid angiopathy in Table-1 are less common. Part of the  "unknown" group may include many patients in whom bleeding was caused by acute rises of blood pressure due to a variety of causes.

As mentioned, chronic hypertension (i.e., known hypertension or left ventricular hypertrophy) is the most common cause of SICH, and the incidence is related to the degree and duration of elevation of the blood pressure. Racial (e.g., Japanese, African Americans) predisposition to hypertension and socioeconomic problems that have limited detection and treatment may explain the high incidence of SICH in certain racial groups. The incidence of fatal intracranial hemorrhage is actually many times greater in Japan than in the United States, which probably explains the high level of interest, aggressive approach, and number of studies coming from that country. Although some series implicate hypertension in as many as 90 percent of patients with SICH. this may not be completely accurate and may be changing dramatically. To put these issues in perspective, an approximation of frequency is indicated in Table-1. The future trends in the incidence of these problems are also projected based on assumptions that are fairly apparent (i.e., increased use of street drugs, aging of the population so that amyloid angiopathy will be more common, less use and better control of anticoagulants, and treatment of DIC to avoid delayed traumatic SICH). SICH has traditionally been considered in a bipartite fashion.

TABLE-1 Etiologies of Spontaneous Intracerebral Hematoma
  Incidence Trend
  Chronic (Acute) 1/2 +
Not hypertensive    
  Congenital vascular anomalies 1/4  
    Arteriovenous malformations    
  Coagulopathy   -
  Vasculopathy. vasculitis    
    Cerebral amyloid angiopathy   +
  Drug related    
    Sympathomimetics   +
    Anticoagulants   -
    Fibrinolytics   +
    Intracranial   -
    Carotid   -
    Arterial infarction   -
    Venous occlusion    
  Delayed post-traumatic    
  Mycotic aneurysm    
  Neonatal intraventricular    
Secondary brain stem    

Because hematomas due to hypertension occur in typical locations (Table-2), this group will be considered first by site. The remaining hematomas often originate in the subcortical white matter: each etiology will be considered separately. Although it is true that subcortical clots may have a variety of specific etiologies, a large proportion (45 percent) may still be related to hypertension. On the other hand, clots in locations typical for hypertensive hemorrhage may often be due to other causes. For example, clots in the cerebellum and pons may result from such problems as cryptic vascular malformations. Of course, hypertensive patients may have clots caused by other etiologies such as aneurysms.

Hypertensive Hemorrhage


The vasculopathy of chronic hypertension affects the perforating arteries, 100 to 400 µm diameter, which arise directly from much larger trunks to enter the brain at right angles and which are end arteries. These vessels are subjected directly to changes in blood pressure, unlike cortical vessels which are protected by a series of bifurcations and have collaterals for run off. These small arteries accumulate lipid and proteinaceous material in their walls ( lipohyalinosis) that in turn can cause a scarring (hyalinosis) or, alternatively, focal necrosis and even Charcot-Bouchard or miliary "aneurysms." Target arteries include the lenticulostriate arteries, the thalamoperforating arteries, and the paramedian branches of the basilar artery as well as the superior and anterior inferior cerebellar arteries, in whose distributions SICH may develop. This process is more common in the proximal part of the artery (explaining why putaminal hemorrhages are more common than those in the caudate). In addition, autoregulation of blood flow is altered in chronic hypertension, and the vessels are less able to compensate for increased blood pressure, predisposing to bleeding as well as impaired compensation after.

TABLE-2 Relation of Hypertension to Location
  Hypertension 75 to 50% + N0 Hypertension 25 to 50% -
Basal ganglia 65% 35%
Subcortical white matter 45% 55%
Thalamus 75% 25%
Cerebellum 62% 38%
Pons 90% 10%

Many other factors can contribute to bleeding. Damage to the parenchyma may compromise support to the vessels: the significance of this is not clear, however. Acute increases in blood pressure and flow may also be important, particularly where autoregulation may be compromised (as in trauma), or where pressure may be above the limits of autoregulation, as in toxemia. SICH generally occurs during the morning or early afternoon when a patient is active. Therefore, it has been postulated that the trigger for bleeding may be a diurnal rise or an acute increase in blood pressure from whatever cause. Last, compromise of hemostatic mechanisms may playa role, as in delayed traumatic SICH, a bleeding diathesis, or anticoagulant usage.

It had been thought that the bleeding event is relatively acute. However, angiograms have shown bleeding for several hours and, sometimes, even days from onset. In one systematic study, it was shown that six of eight patients with serial CT scans had an increase of the volume of their clot of over 40 percent. In another series, late deterioration was seen in a small proportion. It seems that most bleeding takes place within 6 h, and clots larger than 5 cm in diameter are most likely to expand, which fits with the theory of expansion due to tearing, which would be more likely to occur in a larger clot. Many authors believe that this secondary bleeding is a very important mechanism in clot development. Small satellite hemorrhages, the marginal hemorrhages of Stemmer, may be due to similar disruption of more distant small vessels. The pathologic evidence of such bleeding is the fibrin-platelet masses found within and at the margin of the clots. Blood pressure (systemic and local), size and rigidity of the vessel involved, state of autoregulation, state of the hemostatic system and physical condition of the surrounding parenchyma probably all play a role in determining the size of the hematoma. A small number of patients will develop new clots, usually in a different location.

The ultimate clinical manifestations of the clot relate to the speed and volume of the hemorrhage as well as its location. The patterns of spread for each location have been described, as have the clinical manifestations related to location and extension. A small hemorrhage may dissect along tissue planes (e.g., a lobar hemorrhage), splitting the tissue apart rather than destroying it, with limited compromise and/or with restitution of function when the blood is absorbed. A very large hemorrhage may explode into the brain substance, destroying large amounts of tissue, raising intracranial pressure to the level of the blood pressure before the bleeding is tamponaded, and causing herniation of that part of the brain from its normal position under the falx, through the tentorial incisura, or through the foramen magnum, depending on the location of the bleeding. Even if there is not acute herniation, the brain is plastic and can further deform or "creep" due to pressure from the original mass. Local pressure and edema, disconnection and more distant changes in metabolism and blood flow, are also important.

Blood may rupture into a ventricle (especially with caudate, thalamic, cerebellar and pontine hemorrhages) and even cause hydrocephalus. On the other hand, the rupture may actually decompress the clot. Blood may also find its way into the subarachnoid space causing irritation and hydrocephalus as well. Distortion of the upper brain stem may also lead to hydrocephalus.

Death is due to distortion or compression of the brain stem, development of secondary brain stem hemorrhages, or direct extension of the clot into the brain stem. With posterior fossa hemorrhage, there may be direct compression of the brain stem. Basal ganglia clots of more than 85 ml or more than 6 percent of the volume of the brain, or cerebellar clots more than 3 cm in diameter have a poor prognosis if left untreated.

If the patient lives, the clot will eventually be broken down and reabsorbed. A six-phase process based on evolution of the clot has been described. It includes invasion by macrophages, development of surrounding edema, and development of microvessels at the margin of the clot, followed by quieting of these processes and development of gliosis (Table-3 ). In the case of a larger clot, this will take many months. A small number of patients will develop new clots, usually in a different location.

TABLE-3 Evolution of a Spontaneous Intracerebral Hematoma
  Stages in Development and Resolution


Source Parameter Hyperacute Acute Early Late Early Late Ancient

Kirkpatrick and Hayman

Histology <6h Evolution of clot 7 h-3 days Lysis of clot: entry of rnacrophages: brain edema

4-10 days Microvessels at margin

11days- 6 weeks Resorption of edema 7 weeks-6 months Processes quiet: lesion contracts: gliosis  develops >6 months Contracted glial scar: stained by hemosiderin
Chaney et al. Histology 0-24 h 1-7 days 1-2 weeks 2-4 weeks

>1 month

Williams ct al. Changes in hemoglobin  

< I week Intracellular deoxyhemoglobin

1-2 weeks Intracellular methemoglobin

 2-4 weeks extracellular methemoglobin 1-6 months Intracellular methemoglobin : hemosiderin >6 months Hemosiderin  

Models have been used to study various aspects of SICH. In vivo models mimic the natural dynamic milieu of human hematomas. But, animals are expensive, their hemostatic systems may be very different from those of humans, and their brains are too small to accept clots large enough to use to evaluate new surgical devices. However, animal studies have revealed many details about pathologic and physiologic changes after SICH. They have demonstrated that blood is irritating to the parenchyma, causing a progressive hemorrhagic necrosis with edema at the margin of the clot. This process is fixed by 6 h. Animal studies have also demonstrated changes in local and distant blood flow and metabolism. And, in animal studies, early evacuation of the clot has been shown to improve outcome. In vitro models using human blood have been helpful in studying lytic drugs and aspiration devices, but they lack the dynamic setting of an animal model.

Supratentorial Hematomas


Supratentorial hematomas constitute about 80 percent of SICHs. Perhaps half of these hematomas are related to hypertension. Their highest incidence is in the fifth and sixth decades of life; males may predominate. Table-4 lists their distribution sites. They may be divided into gangliobasal and lobar. Gangliobasal hematomas may occur in the basal ganglia or thalamus. Those in the basal ganglia may be internal or deep (two-thirds) or external or superficial (one-third), depending on their relationship to the internal capsule. This classification may have considerable surgical significance. Lobar hematomas tend to be seen in younger patients. One-third are due to hypertension. Aneurysms and arteriovenous malformations (AVMs) are frequent causes, as are tumors and coagulopathies. There is no obvious etiology at the time of presentation in almost one-quarter of these hematomas.

TABLE-4 Distribution of Hypertensive Hemorrhages


Percentage Trends of Variation
Putamen 35-50 +
Subcortical white matter 30 -
Cerebellum 16 -
Thalamus 10-15 -
Pons 5-12 +

Symptoms and Signs

Presentation is abrupt or acute with an altered level of consciousness and progression to death within hours to days in one­third to one-half of cases (although this is not a certain figure and reports vary considerably. On the other hand, there may be only focal signs with preservation of consciousness with small hematomas. There are subgroupings for each of the primary sites and degree of hematoma extension. An excellent grading scheme based on level of consciousness has been developed (Table-5 ). Initial symptoms may include headache, nausea and vomiting. Seizures may be present, especially in lobar hematomas (50 percent or more) and may occur at onset or later: they may be an ongoing problem. Clinical manifestations relate to site of origin, direction and extent of further bleeding, secondary effects and herniation, as well as to ventricular and subarachnoid extension with hydrocephalus and meningeal irritation.

TABLE-5 Level of Consciousness
Grade Criteria
1 Alertness or confusion
2 Somnolence
3 Stupor
4a Semicoma without herniation
4b Semicoma with herniation
5 Deep coma

In putaminal hemorrhages, motor deficits predominate over sensory abnormalities. Depending on the extent of the hemorrhage, other symptoms may include frontal gaze paresis, homonymous hemianopsia, aphasia if the dominant hemisphere is involved, and hemineglect if the nondominant hemisphere is involved . Caudate hematomas are less common and tend to be more benign. They do often extend into the lateral ventricle and cause hydrocephalus. However, some spread into the adjacent brain structures, which becomes problematic.

Specific symptoms of thalamic hemorrhages include hemiparesis, sensory deficits, oculomotor and pupillary disturbances due to extension into the brain stem or hydrocephalus, a dysphasia characterized by fluctuation and paraphasia if the dominant hemisphere is involved. and neglect if the nondominant hemisphere is involved. Thalamic pain syndromes and hemisensory strokes may be seen. Specific syndromes have been described for small hemorrhages. Ventricular extension (including that from paramedian and dorsal hematomas, size> 10 cm3) warrants a poorer prognosis.

Lobar hematomas commonly result from occult vascular malformations, microaneurysms (some not related to hypertension), cerebral amyloid angiopathy or occult tumors (although many patients also have hypertension. These causes might be anticipated based on previous hemorrhages, enhancement on CT, oval or round shape (malformations) or subarachnoid bleeding (amyloid angiopathy). The clinical picture of lobar hematomas depends on their location and extent. The location of a headache may indicate the site. Seizures are more common and coma is less common than with deep clots. The clinical pictures of all these lesions depend on site of origin, direction of spread and size: prognosis is better than for deep clots. An outcome grading scheme has been developed (Table-6).

TABLE-6 Postoperative Evaluation of Patients

Activities of Daily Living

1 Well (full work)
2 Minimal disability (work, self-sufficient)
3 Partial disability (semi-self-sufficient)
4 Total disability (bedridden)
5 Vegetative
6 Dead

Diagnostic Studies

The general laboratory evaluation indicated for SICH may be extensive. Besides, routine admission studies, there should be evaluation of the heart, peripheral vessels and kidneys. The cause of hypertension might be investigated in patients with elevated blood pressure, It may be useful to screen for hematologic abnormalities: infectious processes. and vasculitides.

The most critical tests for the investigation of SICH are CT or MRI, both for initial diagnosis and for surgical planning. The presence of primary intracranial lesion, including tumor and congenital vascular abnormalities must be kept in mind.

Because of the high density of blood, hematomas just a few mm in diameter can be seen on CT. Indeed, recent studies have shown that many strokes formerly believed to be due to infarction are really due to hemorrhages. In addition, details about the hemorrhage, including exact location, size, associated brain shifts, ventricular extension and secondary hydrocephalus aid in surgical planning and may provide the means for improving prognostication and understanding of the pathophysiology involved. A CT grading scheme has been developed for basal ganglionic hemorrhage (Table-7). If the medial edge of the hematoma is less than 28 mm from the pineal, the posterior limb of the internal capsule is involved and the prognosis is worse. A CT grading scheme has also been developed for thalamic hemorrhage (Table-8). If the lateral edge of the hematoma is more than 32 mm from the pineal, the posterior limb of the internal capsule is involved and the prognosis is worse. Contrast infusion may provide additional information about primary lesions and may be indicated in patients (1) less than 40 years of age. (2) without hypertension. (3) with neurological impairment increasing for more than 4h. (4) with history of neoplasm, blood dyscrasia, vasculitis, or bacterial endocarditis, or (5) with blood in the subarachnoid space or an atypical location or appearance of the clot.

TABLE-7 CT Classification of Basal Ganglionic Hemorrhage
Class Type Criteria
I External capsule Localized at of internal capsule
II Capsular (C)* Extends to anterior limb of internal capsule
IIIa Cp without V* Extends to posterior limb of internal capsule
IIIb Cp* with V*  
IVa Ca* + p without V Extends to anterior and posterior limbs of internal capsule
IVb Ca + p with V Th*  
V   Extends to thalamus and subthalamus

* V. massive ventricular hemorrhage. C: capsule. a: anterior. p: posterior. Th: thalamus.


TABLE-8 CT Classification of Thalamic Hemorrhage
Class Criteria
Ia Localized in thalamus  without V*
Ib Localized in thalamus  with V
IIa Extends to internal capsule without V
IIb Extends to internal capsule with V
IIIa Extends to hypothalamus or midbrain without V
IIIb Extends to hypothalamus or midbrain with V
* V. massive ventricular hemorrhage

The change in the CT appearance of the hematoma has been studied extensively. Within hours, the clot becomes more dense and a ring of low density develops around it which may represent edema or fluid squeezed out as it retracts. Initially hyperdense because or high protein content, acute clots are better seen with CT than with MRI. With time the clot becomes isodense with liquefaction and resorption. Small clots (<2 cm) are absorbed especially rapidly. Edema dissipates more slowly than clot reabsorb, but this is difficult to study in detail because the clot itself becomes more radiologically isodense. Only a small proportion leave typical slit-like lesions, and in a number there may be no residual abnormalities. More work remains to fully understand the time course of these changes and how they relate to the CT appearance. The high-field and midfield MRI appearance of SICH has been studied extensively. We now know that there are many factors involving clotting and breakdown of the hematoma as well as the sequences used that relate to its appearance. Because of the chemical and physical alterations within and around the clot, characteristic changes in it, appearance in different sequences also permit its approximate dating. The appearance of the hematoma center, hematoma periphery, and adjacent nearby brain hale been looked at systematically. Acute clots have magnetic characteristics similar to brain on T1 and T2 sequences, so gradient echo sequences should be used. They are better seen after a few days, however MRI changes reflect lysis of erythrocytes, which occurs from the center outward, and chemical changes in the hemoglobin molecule (oxyhemoglobin, 0 to 12 h: deoxyhemoglobin. 1 to 7 days methemoglobin. 5 days to months: and hemosiderin, 1 week to years). An area of "ring enhancement" develops around the margins of the clot, probably related to edema, which is maximal by 4 to 5 day, and whose duration is between 3 and 64 days, and then inflammation occurs between 48 and 84 days. There is eventual resorption of the hematoma (weeks) and resolution of edema (months). One early schema based theoretically on changes in hemoglobin uses five time intervals (acute < I week: early subacute 1 to 2 weeks: late subacute 2 to 4 weeks: early chronic 1 to 6 months: Late chronic >6 months).

  Another schema, derived from an extensive review of the literature and correlated with histologic changes uses, five slightly different intervals (hyperacute 1 to 24 h: acute 1 to 7 days: subacute 1 to 2 weeks and 2 to 4 weeks: chronic > 1 month). Table-3 describes the time interval, and MRI appearances. It has been noted, however, that there may be considerable variability in the appearance of the clot, particularly early, because of differences in the many complex processes that contribute to the rapidly changing appearance (Fig-1). One great advantage of MRI is that lesions such as AVMs or tumors are visualized better than on CT, particularly after enhancement with gadolinium. Hemosiderin remains in the brain after the blood is absorbed and provides evidence of prior bleeding. Also, the clots can be visualized in all planes.

Fig-1: The MRI appearance  of hematoma on T1 and T2 weighted  images.
Note: mixtures indicate that either intensity has been reported.

Angiography may also be indicated if a primary lesion is suspected. It may be positive in 50 percent of younger patients. Angiography provides evidence of mass effect and confirms the diagnosis of a primary lesion such as tumor, aneurysm, or AVM. Because edema as well as clot can contribute to mass effect, the volume of the clot may be overestimated by angiography. Conversely, where the brain is split, the angiographic changes may not fully reflect the size of the clot.

Natural History

Only one-third of patients present with an abrupt onset. The remaining patients deteriorate and progression is usually maximum within hours. Decreased level of consciousness is seen in 60 percent, with coma in 10 percent. Most who die do so within a few days .The patient's subsequent course may be one of deterioration, improvement, or even improvement with subsequent deterioration. Comatose patients with large clots can be expected to die. Over­all figures suggest more than 50 percent of hospitalized victims now survive, which may be attributed both to more frequent identification of small clots and to improved treatment. The level of consciousness, the size of the clot, the presence and degree of shift and evidence of ventricular rupture are the most important prognostic indicators. Thalamic clots have the worst prognosis. Unfortunately, older patients fare worse. Any delay in treatment is harmful. Patients with marked focal neurological deficits and moderate­sized clots will survive with significant deficits. It is thought that most survivors are left with deficits, many of which may be incapacitating. Those with mild deficits and small clots will recover completely.

General Treatment

For severely affected patients, comprehensive management in an intensive care unit (lCU) seems warranted, especially to prevent the cardiac and pulmonary complications that often contribute to death. Hypertension should be controlled. There is the possibility of increasing edema if the blood pressure is too high, and the risk of compromising cerebral blood flow if the blood pressure is too low in the face of increased intracranial pressure. The difficulty in patients with chronic hypertension, however, is that autoregulation may be altered with regard to the blood pressure required to sustain flow. It is not yet possible to individualize the blood pressure required to optimize cerebral blood flow given generally available technology.

Anticonvulsants, should be used for lobar hemorrhages: indications for their use in deep clots is not clear. Corticosteroids are contraindicated since they do not improve the patient but do cause increased complications.

Intracranial Pressure Monitoring and Treatment

Several studies demonstrated that patients with poor neurological status had a high ICP. However, some patients without high ICP did die, presumably from local damage. Patients with intermediate neurological status did or did not have increased ICP. Early surgery seemed to help reduce ICP and improve outcome, but delayed surgery did not. Patients in good clinical condition had low ICP. ICP monitoring permitted optimal medical management of the patients, as well, as helped to successfully guide decision-making regarding whether surgery was necessary. Ventricular drainage can be beneficial in treating the hydrocephalus seen in thalamic hemorrhage with ventricular extension.

Stereotactic Aspiration with Fibrinolytic and Mechanical Assistance

There are two purposes for actually removing hematomas: (1) to preserve life, and (2) maximize recovery of function. Both of these reasons may be threatened by the mass effect of the clot and progressive edema and tissue damage. The optimal approach for removal of an SICH would be a rapid simple method that combines a high success rate with low risk at minimal cost. One technique that may prove to have such characteristics is stereotactic aspiration.

A number of features of clots make them suitable for stereotactic aspiration: (1) they can be easily detected by CT or MRI. (2) they can be localized using stereotactic frames compatible with CT or MRI. (3) their physical properties make them susceptible to aspiration with special devices and this can be facilitated by instillation of thrombolytic substances, and (4) their removal may be accomplished without high risk of rebleeding, or under circumstances where bleeding can be detected (including by intraoperative CT or ultrasonography) and treated.

Although the biophysical characteristics of clots and how these change over time have not been described in detail, early attempts at aspiration of fresh hematomas were recognized as being only partially successful because of difficulties in removing the more solid components of the clots. On the other hand, the use of a large (5 mm) cannula for aspiration with transventricular irrigation of deep clots may at times be successful. In Japan, aspiration often yielded one-half to two-thirds of the clot volume. To fully understand the meaning of this information would require more knowledge about the inner diameter of the catheters, the size of the ports, and the amount of vacuum applied as well as the age of the clot and its appearance on CT or MRI, and the hematocrit and clotting status of the patient. Simple aspiration does appear helpful for medium-sized (22 to 30 mm in diameter) pontine hematomas.

To facilitate aspiration, a number of devices to physically morcellate hematoma material have been developed and used in both experimental and clinical investigations. The first such instrument was a 4mm cannula in which there was an Archimedes screw. Suction was applied to bring the clot up into the cannula where it could be broken up by rotating the screw. The device was used successfully for subtotal removal. It has been modified by other surgeons and used with some success, but was never adopted widely. Another device involves high-pressure fluid irrigation to facilitate suction-aspiration of hematomas. The authors who described the device suggested restricting its use to hematomas more than 24h old, due to fears of rebleeding in operations done earlier. Other sophisticated mechanical approaches, namely breaking down the clot with ultrasonic aspirators, have been used. With a vacuum of 150 mmHg, it could be  aspirated 75 percent of a 4-h old clot in 15 min.

Another approach has been to try to liquefy clots chemically to make them more amenable to aspiration. There is a great deal of activity in the development of thrombolytic drugs. In experimental studies of subarachnoid, intracerebral and intraventricular injections, urokinase appeared safe and indeed promoted clot reabsorption. Since 1980. several Japanese groups have had extensive experience using urokinase in spontaneous intracerebral hematomas in humans, including posterior fossa clots and intraventricular clots. They have indicated that it can be helpful, although it has been associated with rebleeding in 4 percent of cases. Tissue plasminogen activator is safe when injected into the brain of rats and the CSF of rabbits. It also seems to promote clot absorption and has been used to dissolve intraventricular clots.

An endoscope with irrigation, suction, and a laser for hemostasis was employed in a randomized series of 100 patients and was thought to be useful in removing subcortical as well as putaminal and thalamic hematomas. Although in this report the endoscope was inserted using ultrasonography for guidance, a similar technique has been reported using stereotactic positioning of the endoscope.

The aforementioned information suggests that eventually some form of stereotactic aspiration will be developed that will provide an optimum method for evacuating intracerebral hematomas. It is obvious that more needs to be known about the properties of clots, particularly regarding their susceptibility to morcellation and lysis at various intervals after formation, and the coagulation status of the patient. The capabilities of the various mechanical devices need to be studied in more detail, as do the fibrinolytic drugs now available and the new and improved drugs that certainly will be developed in the future.

Open Evacuation

Surgery is not indicated in the face of irreversible neurological damage suggested by great depressed level of consciousness, rapid clinical deterioration or massive size of hematoma and is generally not needed in the case of patient, who are alert and have hematomas less than 2 cm in diameter. Some patients with clots between 2 and 3 cm may benefit from surgery. Critical size may also be 85 ml. One suggestion is that surgery is not needed if the clot occupies <4 percent of the intracranial space, should be based on the clinical status if 4 to 8 percent, should be done for 8 to 12 percent, and will not help if > 12 percent. Not all agree on these guidelines, and there are certainly exceptions: for example, small clots in critical area can be life threatening and large clots can be surprisingly well tolerated. Some authorities advocate immediate surgery (<6h) to minimize ongoing bleeding, irritation of the brain, and edema. Others suggest waiting at least 6 h to minimize the possibility of rebleeding. If patients have not needed surgery by day 10, deterioration is infrequent.

Open surgery has been used for lobar hematomas with considerable success although hematomas due to amyloid angiopathy are more problematic. Because of the risk of brain stem compression, temporal clots as small as 30 ml should be considered for evacuation. Approaches should be made where the clot extends toward the surface or through silent areas of the brain. Surgery for deep clots has been facilitated by the development of transtemporal and trans-sylvian approaches.

The general principles regarding skin, bone, and dura incisions should be followed. Ultrasonography can be used to confirm localization of the clot. Modern technique includes the standard use of magnification and good illumination as well as gentle retraction to minimize difficult-to-control intraoperative bleeding. Most authors recommend that small amounts of adherent clot be left undisturbed, although some suggest that all clot should be removed, which would allow examination of the entire cavity for evidence of an AVM or tumor. If there is any question of the etiology of the hematoma, surgery should include biopsy of the wall of the cavity. An interesting observation since CT has been available is that, with early surgery, the mass effect may actually increase after evacuation.

A number of developments from the 1970s make only recent series relevant when trying to understand the potential role of surgery. With modern neuroimaging, diagnosis is rapid and the anatomy clear. Medical care in an lCU setting can optimize cardiac and pulmonary function. ICP monitoring and control (although this has not been used frequently in SICH patients) is available. Clinical and CT grading scales as well as outcome scales (Tables-5 to 8) have been developed so that patients can be compared from series to series. Experimental design uses randomized clinical trials at best or closely matched controls in contemporary prospective trials. Even in current reports, however, information about these factors is not always available.

Some, although not all of the literature is encouraging. For example, on the basis of matched controls (410 surgical patients vs. 204 medical controls), surgery was helpful in all patients except those who were alert or only somnolent. Using 165 medical controls for 187 surgical patients, only grade III (i.e., moderately impaired) patients benefited from surgery. Using internal controls (N = 265),others believed that moderate and severe cases who were operated on did slightly better: they did not come to a conclusion with regard to thalamic clots (N = 135 ). Comparing 44 patients who had surgery for putaminal hemorrhage with 130 who did not, it was decided that surgery was actually harmful. Kaneko et al., comparing 100 patients with putaminal hemorrhage who had ultra-early operations with historical controls, believed that surgery was more beneficial than conservative management or delayed operations. On the basis of a combined study in Japan using historical controls. Kanaya et al. believed that surgery for putaminal hemorrhage associated with stupor, semi­coma, and coma (N = 3216) was definitely helpful. They also recommended operation for thalamic hemorrhages (N = 639). Juvela et al. randomized 52 patients to surgery or conservative care. The only patients who did better with surgery had Glasgow Coma Scale scores between 7 and 10. However, although several survived, all were disabled. In a small (N = 21) randomized study of putaminal hematomas (>3 cm) in hypertensive patients utilizing transinsular microsurgery, Batjer et al. concluded that evacuation was not helpful. There were a higher proportion of survivors in those operated on, however, and the early termination of the study may have truly been premature. Fujitsu et al. concluded that the most important factor in determining the need for surgery was the time course in the first 6 h after bleeding. They believed that surgery was not beneficial for those with a fulminating course, but it could help those with a rapidly or slowly progressive course if done before irreversible damage, and was not needed if the patient was stable.

Some authorities suggest delaying surgery 48 to 72 h until the clot is partially liquefied. The surgery is therefore technically easier, and the chance of rebleeding is reduced. Also, there are some patients who either stabilize and then deteriorate, do not improve, or only improve slowly, and who may benefit from surgery even up to 4 weeks later to decompress nearby neurons.

In the largest reported experience, Kanaya and Kuroda continued their work in the fifth all-Japan study in which 339 institutions participated. There were 7010 patients with putaminal hemorrhages studied, of whom 3375 were operated on and 3635 were not. Both aspiration and craniotomy were investigated. Using information from neurological examination, CT grading, clot volume and deformity of cisterns, it was concluded that small hematomas did not require surgery, intermediate ones should be treated with serial aspiration with injection of urokinase, large ones causing "semicoma" or early herniation should have open surgery, and terminal patients should be treated expectantly. They also suggested that thalamic hemorrhages with hydrocephalus should be treated with ventricular drainage and possibly open surgery if they extended to the hypothalamus and midbrain and that lobar hematomas with semicoma benefited from surgery. More formal and rigorous studies are needed to define precisely which patients should be managed aggressively and how to optimize treatment.

Intraventricular Hematoma

Intraventricular hemorrhage may be an isolated (and often benign) problem and may be due to an AVM of the choroid plexus. However, almost 80 percent of intraventricular hematomas (IVHs) are related to intracerebral hematomas, and they are usually caused by hypertension, aneurysms, AVMs and even pituitary apoplexy. They are often accompanied by slightly enlarged ventricles. One-third of SICHs are accompanied by lVH, and these have a higher mortality rate. The primary hematoma and disease process are probably of more significance than the IVH. However, prognosis is also determined by the extent of the hemorrhage. Headache, vomiting, confusion, decreased level of consciousness and, in the case of secondary bleeding, hemiparesis, are common clinical findings. The clots tend to disappear within 2 weeks. Recent CT studies suggest that IVH is more frequent than previously suspected, but it is often not significant clinically. When clots are symptomatic, intraventricular drainage (possibly bilateral) may be useful, but the blood often occludes catheters used for this purpose. Ultimately, a shunt may be required if permanent hydrocephalus develops. Intraventricular thrombolytic therapy has been shown experimentally to be useful and safe and has been used successfully in humans. Direct surgery has not proved useful.



Cerebellar hematomas constitute about 10 percent of SICH, a proportion coincident with the volume of brain in which they occur. These occur more commonly in males. The highest frequency is in the sixth through eighth decades of life. Two-thirds of cerebellar hematomas are related to hypertension. These usually occur in the dentate nucleus which is irrigated by the superior cerebellar artery. There may be a left predominance. A small number originate in the vermis. Hematomas in the younger age group may be related to vascular malformations. Anticoagulants are another common predisposing factor.

Many of these hematomas are extensive and rupture into the fourth ventricle and also into the subarachnoid space. Secondary hydrocephalus may develop in up to 75 percent of patients. Death occurs in 60 to 80 percent of patients and is due to brain stem compression and tonsilar herniation.

Symptomatology is related to the rapidity of bleeding and the size and location of the hematoma as well as to compression of the brain stem, upward cerebellar herniation and tonsilar herniation, hematoma rupture into the fourth ventricle, and the development of hydrocephalus. The onset of symptoms is often abrupt. but may be subacute with progression over various times, or subacute with resolution. Symptoms and signs are protean and include headaches, alterations in level of consciousness, vomiting with or without nausea, dizziness, eye signs. including changes in pupils and gaze abnormalities, dysarthria, and motor signs, both cerebellar and pyramidal. The classic triad of signs includes appendicular ataxia, ipsilateral gaze palsy, and peripheral facial weakness. Two out of three of these findings are seen in 75 percent of patients. A classic three-stage evolution has been described.

The diagnosis can be difficult if the history is not known and the patient is stuporous. The differential diagnosis is again extensive and includes cerebellar infarction, brain stem hematoma or infarction, bleeding from an aneurysm or a tumor in the posterior fossa, as well as acute labyrinthitis. Clinical diagnosis is often difficult. In one series of 33 patients, 13 with cerebellar hemorrhages or infarctions were diagnosed correctly. 10 were not diagnosed initially, and 10 diagnosed as having cerebellar strokes actually had other problems.

Diagnosis can now be readily made using CT, which can also be helpful in surgical planning. The clot can be well visualized and other abnormalities including blood in the fourth ventricle, brain stem distortion, and hydrocephalus can also be seen. If the patient is not too sick and the test is possible, MRI can provide evidence of a vascular malformation and previous bleeding as well as to better define the anatomy. Angiography can demonstrate mass effect and might be employed if an AVM or other specified lesion is suspected, particularly in a young patient without a history of hypertension. (Even if this is negative, because of the high risk of rebleeding from a vascular malformation, surgery should also be considered.)

Treatment includes control of blood pressure and respiratory support as needed. Surgery involves a posterior fossa craniectomy and evacuation of the clot. The indications for surgical therapy are probably better defined in this group of hematoma than in those in other locations. The key indicator, are based on the level of consciousness, clinical course, and size (2 to 3 cm) of the hematoma, unless the patient is seen after doing well for a week. All clots 3 cm or larger and those between 2 and 3 cm (if the patient's level of consciousness is altered), should be considered for surgery, especially if there is deterioration, since some patients may decompensate rapidly. Mortality is 72 percent if the patient is comatose. Most patients without impaired consciousness will improve spontaneously. Indeed, CT scans have shown that clots tend to disappear in 2 to 6 weeks.

In the past, the use of ventricular drainage by itself was discouraged for two reasons: (1) it did not address the major problem, namely brain stem compression, and (2) because of the risk of upward cerebellar herniation. Indeed, it may delay definitive treatment and it has been suggested that it therefore be employed only in conjunction with clot evacuation. But in cases with clots of borderline size, and possibly in conjunction with mannitol administration, this may be an alternate mode of treatment.

Excellent surgical results with relatively low operative mortality have been described in patients with only moderately depressed levels of consciousness. Occasionally, patients with marked alterations in level of consciousness, particularly if they did not have too abrupt an onset and if operation was performed promptly, have improved with surgery. Patients in extremis are beyond help. Some patients with late deterioration or persistent deficits may also be helped by evacuation. Stereotactic aspiration has also been advocated.

Although the guidelines for surgical treatment of cerebellar hematomas are probably better defined than for those in other locations, there may still be questions in those patients who are quite ill but not in extremis, or who are doing relatively well but are not improving rapidly.

Brain Stem

Brain stem hemorrhages tend to occur predominantly in the pons, although hemorrhages in the midbrain and medulla have been described. Pontine hemorrhages constitute about 3 to 13 percent of SICH far out of proportion to the volume of brain involved. Males and females are equally affected. The highest frequency is in the fourth and fifth decades of life. Ninety percent are related to hypertension and are believed to be due to vascular disease of penetrating branches of the basilar artery. Those hemorrhages seen in younger patients without hypertension may be related to cryptic vascular malformations, which are especially common in the pons but probably account for less than 10 percent of such hemorrhages.

Hematomas are present unilaterally in the basis pontis (at times with progression into the tegmentum) in 22 percent, in the basis bilaterally in 56 percent, and in the tegmentum in 22 percent, (two-thirds bilaterally). Clots extend upward, even to the thalamus, but infrequently downward. The fourth ventricle is usually distorted. There is rupture into the fourth ventricle in at least 70 percent of cases. Extensive edema is often present, the cause of which is unknown. Local vascular disease is common. as is evidence of other cerebrovascular and cardiovascular disease.

Symptomatology is based on location, size, speed of development and rupture into the fourth ventricle and subarachnoid space. as well as hydrocephalus secondary to ventricular occlusion or compression of the fourth ventricle and aqueduct. In the large post­mortem series, the onset was abrupt in one-half. In 30 percent, the initial symptom was severe headache, usually posterior. Symptoms and signs included alterations in level of consciousness, abnormalities of respiration, pulse and blood pressure, hyperthermia, motor abnormalities that were often bilateral with posturing or paralysis, cranial nerve abnormalities, including pupillary and gaze change, with ocular bobbing, vertigo, vomiting, dysarthria, autonomic dysfunction, and "seizures" believed to arise from the basis pontis. The classic triad of miosis, hyperthermia. and bloody CSF was seldom seen. The diagnosis was suspected in only 25 percent of the cases. Seventy-five percent of patients died within 24h.

The common presentation of coma with neurological devastation involves an extensive differential diagnosis, including massive hemorrhages in other locations as well as posterior fossa infarcts and hypertensive encephalopathy. Definitive diagnosis can be made with CT scanning. The diagnosis may also be made with MRI. Angiography might be employed if a vascular malformation is suspected. Mortality is more than 80 percent in 48 h and more in the first week.

Treatment depends on the patient's condition. Most patients present with an acute onset of devastating symptoms and will die. One group has suggested that ophthalmologic findings as well as the size and location of the clot may be useful in predicting potential survivors. In patients believed to be treatable, there should be immediate attention to respiratory support and control of blood pressure where needed. Ventricular drainage might be used if hydrocephalus is present, but the very presence of hydrocephalus may be a marker of a fatal hemorrhage. A major question concerns the role of direct surgery. On the one hand, hematomas have been followed by CT and have been seen to resorb, occasionally with a good result. On the other hand, several cases, including a few with acute onset, have been thought to have been successfully operated on either through the fourth ventricle or subtemporally. CT might suggest the best route. Biopsy of the wall is thought to lead to deterioration. Stereotactic aspiration has also proved helpful. Other series, however, have suggested that acute surgery does not improve outcome. Patients with persistent symptoms from unresorbed clots might have direct or stereotactic aspiration, and patients with recurrent bleeding due to vascular malformations should be considered for open surgery, Collaborative studies will probably be needed to define those patients who are ill enough to require surgery but not yet beyond hope. Finally, the usefulness of rehabilitation for many stroke victim­ appears well established and must be pursued when the patient is stable.


Aneurysm, AVM

These anomalies are the second leading cause of SICH and there should be a high index of suspicion in young patients and those with superficial hematomas. A history suggesting a sentinel hemorrhage, or in the case of AVMs. seizures, headaches, or focal findings may increase the index of suspicion. It is thought that of those patients with aneurysms that bleed, 40 percent will have SICH, one-half of these >3 cm in diameter. Aneurysms bleed into the brain when the aneurysm is typically imbedded in brain (i.e., internal carotid bifurcation, anterior cerebral artery, distal anterior cerebral artery), when it points into the brain (i.e., posterior cerebral artery), when surrounding structures are scarred from previous bleeding, or when the local brain is already damaged. SICH is more common after the first hemorrhage. Aneurysm should be suspected if the clot is frontal or temporal in location, although even basal ganglia clots may originate from aneurysms. Aneurysms >5 mm in diameter may be seen when enhanced CT scans are compared to unenhanced ones, especially with fine cuts which include the sites of typical aneurysms. Repeat CT may also be helpful to detect lesions not seen initially because of vasospasm and/or compression which could prevent filling, Angiography should be utilized aggressively and certainly for any patient who might be an operative candidate. Repeat angiography may be necessary if no lesion is initially seen. Early surgery is indicated because of the risk of early rebleeding, The aneurysm should be clipped during the initial operation, preferably as the first step using subarachnoid dissection to initially obtain proximal control. If the patient is moribund, there may not even be time for an angiogram, but the CT can provide some information about the aneurysm. Patient outcomes are worse for aneurysm surgery in the presence of an intraparenchymal hematoma.

Angiomatous malformations include AV fistulas, classic AVMs, telangectasias, cavernous angiomas, venous angiomas, and dural AVMs. These lesions should be considered in younger patients without hypertension who have hemorrhages that are superficial, lobar, periventricular, or into the ventricle, clots that have a low density ring around them, or subarachnoid blood. Hemorrhage is the presenting symptom in 30 to 55 percent, and 50 to 66 percent of patients with classic AVMs that bleed have SICH. Ten to twenty percent of AVMs that bleed have aneurysms, which may or may not be associated with the AVM. Bleeding from an AVM occurs most commonly from the draining veins or the nidus near the veins, but can arise from the aneurysm, AVMs with central venous drainage, a periventricular or intraventricular location, and an intranidal aneurysm may be most likely to bleed. Enhanced CT will often show the nidus and the draining veins. In some cases, however, the clot may compress the AVM and prevent its filling: repeat studies may be required, MRI is more sensitive for detecting AVMs. Some AVMs and cavernous angiomas may be occult (i.e., not seen on angiograms) and studies may need to be repeated.

Eighty percent of AVMs can be resected. It is best to delay surgery to allow neurological deficits to resolve. Also, it is easier to operate after the brain is less swollen, the AVM is better seen. and the clot has liquefied. If immediate surgery is required, a flap should be turned that is large enough to permit resection of the AVM and clipping of the feeding vessels. However, if intraoperative bleeding is difficult to control. it is preferable to only remove the hematoma and to resect the AVM at a later time, since the early rebleeding rate is thought to be small. The clot should also be sent for pathologic examination. On the other hand, if during surgery a cavernous angioma is seen, it should be resected if this seems easy. The veins of venous angiomas may drain normal brain, and careful analysis of angiograms is required to determine if they can be safely resected. Dural AVMs are only rarely the cause of SICH, They can be complex entities, and their treatment requires special considerations.

Hematologic Disorders

Hemostasis is governed by complex interactions among blood vessels, platelets, and blood coagulation factors. Defects in hemostasis either exacerbate bleeding from other problems, such as trauma, or they lead to spontaneous bleeding if severe. Thus, spontaneous bleeding is most common if platelets are less than 10.000/ml or activity of a given clotting factor is less than 1 percent of normal. Abnormalities of the two entities can be classified as follows:


Peripheral destruction-immune (i.e., idiopathic thrombocytopenic purpura)

    Decreased production (i.e.. marrow injury or replacement)  
  Disorder, of function    
    Acquired (i.e., von Willebrand's disease)
Coagulation Factors      
  Inherited deficiencies  Hemophilia can be complicated by AIDS, inhibitors
  Acquired disorders involving deficiencies and inhibitors  
    Disseminated intravacular coagulation  
    Liver disease  

It is important to have a consultation regarding treatment of the primary disease process, replacement of clotting factors and platelets, and decisions about surgery based not only on the acute but also on the ultimate prognosis of the patient. With the risk of transmitting various diseases, particularly AIDS, thresholds for prophylactic use of blood products and accepted replacement levels are changing.


Although intracranial tumors may bleed into a variety of sites, they most commonly bleed into the brain, and even more specifically into the tumor. Depending on the biases of the patient population. tumors may be the third or fourth most common specific cause of SICH. In one literature review, tumors caused 4.6 percent of all SICH, and 3.9 percent of patients with tumors had SICH. Metastatic tumors (especially bronchogenic carcinoma, melanoma, choriocarcinoma or renal tumors) most commonly, but also gliomas, especially more malignant ones  medulloblastoma and even benign tumors (meningiomas, pituitary tumor) have been associated with SICH. Factors leading to bleeding include hypervascularity, abnormal vessels, invasion of vessels and tumor necrosis as well as, related disorders of hemostasis. Bleeding mal occur after needle biopsy, shunting, decompression (even at a distance), and radiation therapy. The bleeding may be related to other factors such as anticoagulation or trauma. In many cases the patient was already known to have a tumor, or there was a prior history of progressive neurological dysfunction or headache. In one-third of cases, however, bleeding may have caused the onset of symptoms, which may have been abrupt or gradual. On the other hand, many hemorrhages, are small and asymptomatic.

CT abnormalities that may suggest a tumor include subcortical site, unusual appearance with abnormally appearing or enhancing tissue within or adjacent to the clot, and excessive edema or mass effect adjacent to the clot and extending even across the midline. Multiple lesions would also be suspicious. An angiogram may demonstrate abnormal vessels but is usually not required.

Surgery may be indicated depending on the clinical significance of the clot and the nature of the underlying disease. The surgeon should remove as much of the tumor as possible, not only to treat the underlying disease but also to minimize the risk of rebleeding. (If surgery is carried out for any clot, any suspicious tissue should he sent for histologic examination). Patient, often, but not invariably, do badly because tumor, that bleed are often very malignant and because the prognosis for a large clot by itself is often poor.

Pituitary hemorrhages occasionally develop from a normal gland or nonadenomatous tumor, but generally arise from adenomas, both active and inactive endocrinologically. Indeed, < 1 to > 12 percent of pituitary adenomas give rise to pituitary apoplexy. Asymptomatic small hemorrhages are more common. and even large asymptomatic hemorrhages are seen. The bleeding may be spontaneous or may be precipitated by trauma, anticoagulant, estrogen, or bromocreptine usage, or radiotherapy. The etiology may not be clear. Clinical presentation includes sudden headache, nausea, stiff neck, decreased vision and field cuts, and impaired eye movements. CT and particularly MRI reveal the diagnosis: most hemorrhages extend above the sella. Steroid replacement and prompt surgery (generally transsphenoidal) are recommended. If some function is preserved, most deficits; (except complete loss of vision) improve or clear even if the patient has been symptomatic for a few days.

Vasculopathy, Vasculitis

Vasculopathy includes conditions, that have proliferative changes or intramural deposits of adventitious materials. Vasculitis includes conditions characterized by inflammation and necrosis of vessel walls. A variety of classification schemes have been used. There are a variety of types which may be of specific or nonspecific etiology which involve the brain only or are generalized. which may involve vessels of different sizes, and which have different histologic appearances. They cause SICH by weakening vessel wall, by occluding vessels leading to infarction into which bleeding occurs, or by causing myocardial infarction, cardiac embolism, and a transforming infarction. The diagnosis is much more obvious when there is typical systemic involvement. The diagnosis is made by angiography and biopsy.

Cerebral amyloid angiopathy (CAA) may be the third leading cause of SICH. It predominates in the elderly population. Amyloid is deposited in the media and adventitia of small- and medium­sized superficial cortical and leptomeningeal arteries that become brittle and rupture and also lose the hemostatic function of their endothelia. CAA is expected to be a more common cause of SICH as our population ages. Seen in 10 percent of those in their 70s, and in over 60 percent over 90, it leads to recurrent and multiple superficial hemorrhages from the weakened vessels. There are familial varieties. It is also associated with a variety of diseases from Alzheimer's disease to dementia pugilistica, and many patients have hypertension. The diagnosis must be made by biopsy or postmortem examination. The prognosis is usually poor, and surgery may be complicated by difficult hemostasis and rebleeding, although it can be done successfully.

Fibromuscular dysplasia may lead to aneurysm formation. Secondary SICH may then occur. Moyamoya disease is a specific condition or a syndrome resulting from various diseases causing occlusion of proximal cerebral vessels. It is characterized by progressive stenosis of the anterior circle of Willis and compensatory transdural or posterior fossa anastomoses and collateral channels in the basal ganglia. Bleeding occurs from microaneurysms in the vessels in the basal ganglia or secondary proximal internal carotid and posterior fossa aneurysms. SICH is the most common cause of death.

There are both multisystem (systemic lupus erythematosus, rheumatoid arthritis, giant cell arteritis) and isolated (granulomatous angiitis) vasculitides that can lead to cerebral vascular weakening and bleeding. The diagnosis can be suspected if the systemic disease is present: it can be confirmed by a picture of vascular stenosis and narrowings on angiography.


A number of sympathomimetic street drugs, including amphetamines and cocaine, as well as over-the-counter drugs, may cause SICH, generally after chronic abuse. This may be due to hypertension and/or vasculitis. At times, angiography will demonstrate vasculitis in the small- and medium-sized arteries. The arteritis will subside with cessation of drug use and the administration of cyclophosphamide and prednisone. However, blood pressure elevations per se may also precipitate rupture of preexisting aneurysms and AVMs. The clots tend to arise in the subcortical white matter.

Anticoagulants can lead to SICH, especially if the clotting studies are especially prolonged (i.e., prothrombin time> 1.5 times normal). The hemorrhage may evolve slowly and become very large. Related to age, hypertension, head injury (even minor), and infarction, it may be the cause of up to 10 percent of SICHs. SICH may be seen in up to 2 percent of patients on anticoagulants. The parenchyma is the second most common site of intracranial bleeding (after the subdural space) that tends to occur in the lobar white matter or cerebellum. Treatment involves normalizing the hemostatic system with vitamin K in the case of oral anticoagulants and protamine sulfate for heparin. Patient outcome is often poor, with two-thirds usually dying.

Thrombolytic drugs, particularly urokinase and tissue plasminogen activator, are now being used more extensively, particularly for treating coronary artery thrombosis. SICH has been identified as a complication of these drugs, but hemorrhagic transformation can occur after myocardial infarction without the drugs, and the increased risk is fairly small. There is growing interest in using thrombolytic drugs to treat cerebral vascular occlusion. Treatment is problematic since reversing the effect of the drug might exacerbate the original thrombosis. Alcohol, if used excessively, can predispose to SICH. This may be related to its causing hypertension or altering coagulation mechanisms.


Bleeding after carotid endarterectomy, although it occurs in well under 1 percent of operations, may be devastating. Usually delayed a few days, it occurs especially after opening a severely stenotic artery (with hypoperfusion) particularly if the artery supplied an area of previous infarction. Postoperative hypertension with hyperperfusion exacerbates the risk, whereas optimal control of blood pressure minimizes it. Postoperative anticoagulants or antiplatelet agents increase the incidence.

Postcraniotomy bleeding probably relates to a number of problems including inadequate hemostasis, low intracranial pressure which minimizes tamponade, local and generalized DIC unrecognized platelet abnormalities (including inhibition by salicylates), breakdown of autoregulation and postoperative hypertension. In one series, such clots were seen in 0.5 percent of -1992 intracranial procedures: of these 24 patients, 8 died and 7 had a poor outcome. Special problems arise in surgery for specific lesions. Surgery for aneurysm may be complicated by bleeding after imperfect clip placement, and surgery for AVMs may be complicated by postoperative circulatory breakthrough. SICH occurs after extracranial to intracranial bypass surgery where there has been a prior infarct. The risk of hematoma formation after stereotactic surgery is 0 to 2.5 percent, but only one-quarter of patients require surgery. Placement of monitoring devices through the brain may lead to direct injury to vessels. particularly in the face of DIC. Diagnostic procedures, including lumbar puncture and angiography, and endovascular techniques such as coiling or embolization for tumor or AVM, occlusion of vessels or an aneurysm itself, and angioplasty for spasm may be complicated by hemorrhage.

SICH after cardiac operations may be related to a number of factors unique to this kind of surgery. These include emboli, arterial hypertension, increased venous pressure and anticoagulant use.


After an ischemic infarct, there may be transformation to a hemorrhagic infarct or even frank parenchymal hemorrhage, presumably due to reopening of the occluded vessel and leakage of blood from the vessels damaged from the ischemic insult. Bleeding has been seen in more than one-half of autopsied patients. Recent MRI studies have shown some hemorrhage in 69 percent: several CT and MRI studies have revealed small hematomas in about 15 percent and large hematomas in 10 percent of patients although they were often asymptomatic. Risk is highest in patients with embolic strokes from carotid or cardiac disease, those who have large infarcts with significant mass effect and herniation, and those who have early hypodense changes or areas of contrast enhancement on CT. Anticoagulants predispose to this problem, and their use in embolic disease of cardiac origin should be individualized. This change is usually not seen in the first day, but often occurs within 4 days, although a certain number occur later. Angiography has revealed that many occlusions reopen within 2-1h, after which reperfusion leads to this bleeding. Later bleeding may relate to the development of collateral circulation. The development of a parenchymatous hemorrhage when accompanied by clinical deterioration has a poorer prognosis. Specific treatment for this hemorrhage has not been discussed in detail. Since heparin can exacerbate the bleeding (but not change the incidence of hemorrhage), it may need to be stopped. Another issue includes the trials of thrombolytic therapy for occlusive stroke. It appears that, this treatment is efficacious with acceptable risk.

Venous and sinus thrombosis, a complication of dehydration or congestive heart failure, hematologic problems, oral contraceptives, pregnancy, trauma, infection or malignancies including leukemia may also cause SICH. Venous thrombosis may involve the sagittal sinus, transverse sinus, cavernous sinus or cortical veins. Clinical manifestations depend on the extent of the thrombosis and collaterals. There may be evidence of elevated intracranial pressure with or without obvious focal signs, depending on the sinus involved and the site of the hemorrhage. Seizures may be a prominent event. Sagittal sinus thrombosis can lead to SICH which is usually in the parasagittal white matter bilaterally. The CT and MRI appearance may be diagnostic for such occlusions because of bilateral clots. There may be a defect in filling of the sinus on contrasted CT or MRI scans. Angiography can be helpful. Venography is not necessary nor worth the risk. Treatment should be aimed at the underlying condition. The use of anticoagulants in the face of a hematoma is problematic. These hemorrhages have a significant mortality rate.


The so-called delayed traumatic intracerebral hematoma (DTICH) is discussed because it does occur spontaneously and differs from other etiologies only in that the primary initiating factor, the injury, occurs at a distinct point in time as opposed to being the result of ongoing or progressive disease. There are actually three groups of such hematomas, depending on the vessel of origin: (1) clots from traumatic aneurysms on larger arteries, (2) classic DTICH from smaller arteries, and (3) clots from venous injuries (see above).

Traumatic aneurysms can be caused by penetrating injuries or closed head injuries (Table-9) and may be, true, false or mixed. At times rupture, often fatal, occurs within days after injury. They cause SICH in 10 percent of cases. The aneurysm may be detected by comparing an uncontrasted CT scan with a contrasted CT scan. Angiography should be performed if missiles or other objects have passed near major arteries. Early prophylactic clipping is suggested. Classic DTICHs occur in 1.3 to 1.7 percent of patients with head injury judged significant enough to perform CT and 2.3 to 8.4 percent of those with Glasgow Coma Scale scores 8 and are generally seen 3 to 4 days after injury. A variety of mechanisms can play a role in their development (Table-10). As noted, decompressive surgery may contribute to their formation by releasing tamponade in areas of contusion. Treatment must be individualized. Prognosis depends on the size and location of the clot and the previous condition of the patient.

TABLE-9 Etiologies of Traumatic Aneurysms

I. Penetrating  
  Depressed fractures  
  Gunshot wounds  
  Knives. etc.  
II. Closed head injury  
    Supraclinoid carotid
  Local injury  
    Anterior cerebral at falx
    Middle cerebral at sphenoid ridge
    Posterior cerebral at tentorium
    Cortical vessels at adhesions or in linear fracture


TABLE-10 Primary and Secondary Factors Leading to Delayed Traumatic Intracerebral Hematoma

Vessel damage
Neuropil damage
Venous back pressure
Hypoxia, hypotension
Medical reduction of intracranial pressure
Surgical reduction of intracranial pressure
Disseminated intravascular coagulation
Effects of alcohol

Mycotic Aneurysm

SICH after infection may be due to disruption of a vessel wall, bleeding into an infarction, or (most commonly) rupture of an aneurysm arising from an infected vessel wall. Aneurysms occur in  perhaps 1.7 percent of patients with bacterial endocarditis, the most common cause, and they are multiple in 20 percent. Neurological complications are seen in up to one-third of patients with bacterial endocarditis, one-half of which are vascular and more than one-half of all victims die. Mycotic aneurysms constitute 2.6 to 6 percent of all aneurysms, but their incidence is thought to be decreasing.

Although called "mycotic" aneurysms, most aneurysms of infectious origin are secondary to bacterial infections, particularly subacute bacterial endocarditis. They are caused by infected emboli that lodge in distal intracranial arteries, particularly middle cerebral branches. Risk factors include subacute endocarditis, intravenous drug abuse, and immunosuppression. The type of infection is changing as the pattern changes in endocarditis. Fungal aneurysms have also been reported. They may also be the result of infections external to a vessel such as septic cavernous sinus thrombophlebitis and meningitis.

Mycotic aneurysms present as SICH or subarachnoid hemorrhage, which have a high mortality, or with just a headache. Bleeding may occur within 1 to 2 days or up to months after infection, but the average is about 17 days. CT without and then with contrast enhancement or MRI may reveal the aneurysm as well as the hemorrhage. The workup should include complete cerebral angiography, which should be repeated until treatment is finished. Initial treatment should include antibiotics and correction of the cardiac lesion if indicated. Decisions about intracranial surgery should be based on the significance of the clot, details about the aneurysm, and the response to antibiotics. Based on review of the literature, it is suggested:

1. If there is one distal middle cerebral artery aneurysm and the patient has bled, excise the aneurysm. The artery of origin will probably need to be sacrificed. An extracranial to intracranial bypass may be required as an adjunct.

2. If there is a proximal or unruptured aneurysm or an aneurysm on a critical branch, treat with antibiotics and obtain serial angiograms, to see if the aneurysm is resolving, stable, or enlarging. Consider excising enlarging aneurysms and follow healing aneurysms until they disappear. The appropriate frequency for angiograms is not well established, but probably every 10 to 14 days is reasonable. A significant proportion of aneurysm will not disappear, but their walls will be stronger after they have had time to develop fibrosis.

3. Individualize if there are multiple aneurysms.

Endovascular occlusion has also been used successfully.

Hemorrhage may occur with encephalitis, specifically Herpes simplex encephalitis and with brain abscess. SICH may also occur in a variety of circumstances in patients who are immunosuppressed, particularly by AIDS.


The most unique hemorrhage in childhood is the intraventricular and periventricular hemorrhage primarily seen in the prernature. Other predisposing conditions, such as vein of Galen malformations, leukemia and idiopathic thrombocytopenic purpura and inherited coagulation disorders may be seen especially in childhood.


Intracranial hemorrhage, including SICH, is the leading nonobstetrical cause of maternal mortality in pregnancy. It may be related to normal changes in cardiovascular physiology, complications of pregnancy including hypertension in toxemia and eclampsia, coagulation disorders and bleeding from preexisting lesions. Routine nonoperative and operative care are indicated, although it should be remembered that mannitol can dehydrate the fetus, hypotension can be detrimental, and anticonvulsants have teratogenic and depressant effects. Method of delivery does not influence bleeding and should be decided on obstetrical grounds


Go back!Back Home!Go next!

Back Up!




© [2005] [CNS CLINIC - NEUROSURGERY - JORDAN]. All rights reserved