Introduction

   Spontaneous nonaneurysmal intraventricular hemorrhage (IVH) is one of the most serious types of stroke. The majority of cases are associated with arterial hypertension and/or elderly age, diabetes mellitus, bleeding diatheses. The mortality rate for IVH is related to the amount of intraventicular blood and increases from 32.2% in cases with mild hemorrhage to 91% in cases with all ventricular chamber and most survivors are typically left severely disabled.^{1)2)6)16)18)28)}
   A large part of the complications seen after IVH is related to intracranial hypertension from hydrocephalus that cannot be adequately treated with standard external ventricular drainage. 
   The failure of ventriculostomy alone to clear IVH is frequently related to clots within or around the catheter, which obstruct attempts at therapeutic cerebrospinal fluid(CSF) drainage. Anatomic correlates of this impaired CSF outflow include compression of periventricular structures and brain stem injury. Patients initially may not have significant parechymal injury, and relief of persistent IVH may prevent subsequent significant brain damage.⁴⁾
   The authors present experience with consecutive cases of IVH treated by stereotactic computed tomography guided aspiration and thrombolysis with urokinase. In this study, we attempted to assess the feasibility and safety of this treatment modality. This study was undertaken to assess the influence that adjunctive thrombolysis with urokinase has on speeding hemorrhage resolution and on outcome. The hypothesis was that this relatively simple therapy would improve neurological prognosis in patients presenting with IVH.⁴⁾

Materials and Methods

   From January 1999 to December 2003, 53 patients of non-aneurysmal IVH with thrombolysis of urokinase were treated and are reviewed in this report. The patients were treated according a standardized protocol as illustrated in Fig. 1. Eligibility criteria for this protocol consisted of IVH without brain stem extension, clinical onset <48 hours before in-tervention, age >20 years, hemorrhage volume >15 ML, GCS score >5 at admission, no signs of transtentorial herniation, no suspected underlying structural etiology to account for the hemorrhage, no systemic bleeding diathesis, and no severe concurrent illness with life expectancy <6 months.
   Ventriculostomy was performed in cases, who had hydrocephalus from ventricular extension of hemorrhage. Ventricular drainage was weaned while intracranial pressure was monitored. 
   A baseline CT scan was obtained in all patients with axial images at 0.5-1.0 cm slice thickness, and the dimensions of the hematoma were assessed. Volume of the IVH in milliliters was estimated on the basis of approximate ellipse volume with the A×B×C/2 formula, where A represents the largest diameter of the hematoma on axial CT cuts in centimeters, B the diameter of hematoma perpendicular to A on the same cut, and C the number of CT slices in which hematoma is visible multiplied by the slice thickness in centimeters. For the purpose of this calculation we did not count the highest or lowest CT slices in which hematoma was first and last visualized. Intravenous contrast was administered to assess for any enhancement that would be suspicious for an underlying structural lesion. Patients aged <60 years or with abnormal contrast enhancement on CT scan underwent digital subtraction angiography before hematoma aspiration and thrombolysis to exclude an underlying vascular anomaly.

1. Operative technique
   All operations were performed under local anesthesia and intravenous sedation unless the patient was already intubated for medical or neurological indications independent of the procedure. In this series, initial localization of the hematoma and catheter placement was performed with the aid of a Fisher stereotactic system(Leibinger, Germany). An ipsilateral frontal standard burr hole location(3 cm lateral to mid-line and just anterior to the coronal suture) was typically used for capsular and thalamic hemorrhages. If the hematoma was lobar in location or extended to the cortical surface, the burr hole was localized over the hematoma. A 3F to 5F rigid metal catheter was placed with the introducer cannula into the clot via CT guidance. Careful manual hematoma aspiration was attempted using a syringe. The rigid cannula was removed and replaced by a soft ventriculostomy catheter(15 cm long and 1 to 2 mm internal diameter) with perforations spanning the center of the clot. Catheter placement was adjusted under CT guidance as necessary. After satisfactory placement within the hematoma, the catheter was tunneled subcutaneously and the exit site was covered with antibiotic ointment. The catheter was connected to a single port and capped, and a sterile dressing was applied. The patient was maintained on intravenous antibiotic prophylaxis until the brain catheter was removed.
   All patients were managed in a dedicated neurovascular intensive care unit, where subsequent thrombolysis and clot aspiration were performed using sterile technique. Urokinase 6000 IU(Green Cross Biotech., South Korea) in 3 ML of normal saline was injected into the catheter if the CT scan revealed a residual hematoma volume of >25 ML. The catheter was flushed with 2 mL of normal saline. After clamping of catheter for 1 hour, manual aspiration of lysed clot was attempted, and the aspirated volume was recorded. A CT scan was repeated at least every second aspiration. If the volume of residual hematoma remained >25 mL, catheter instillation of urokinase was repeated. The protocol of aspiration, CT scan, and urokinase instillation was repeated as necessary until less than half of its initial volume, or arbitrarily after 4 catheter aspirations of urokinase instillations per day. The catheter was removed under sterile technique, and a single suture was placed at its exit site and covered with an occlusive dressing. 

2. Evaluation of outcome 
   Follow-up clinical information was obtained on all patients 6 months after the procedure. Clinical outcomes were graded according to the GOS, ranging from grade 5(good recovery) to grade 1(dead), by a single investigator not involved in the patients' clinical management. For analysis, we classified two groups；good prognosis group(GOS grade 3, 4, 5) and bad prognosis(GOS grade 1, 2) at the time of discharge and 6 months' follow-up. We performed comparative analysis between two groups in aspect of all possible relating factors of this procedure. 
Unpaired Student t tests were used for the statistical analysis. The value of statistically significance means more than 0.05.

Results

1. Clinical outcome assessment 
   The Table 1 summarizes clinical and radiographic data in the 53 cases treated during the course of 3 years. The mean age of treated patients was 60.2 years(range 38 to 82 years), and there were 28 males and 25 females. There were 28 right side lesions, and 25 left side lesions. Twenty patients (37.7%) had a prior history of arterial hypertension, and twenty-three patients(43.3%) had a diabetes mellitus. And, nine patients (16.9%) had both arterial hypertension and diabetes mellitus. All patients had spontaneous, nontraumatic ICH. Median initial GCS score was 10(range 5 to 15. All patients had some degree of neurological deficit such as contralateral hemiparesis, hemiplegia, and dysphasia. 
   The mean initial hematoma volume was 35 mL(ranging from 16 to 72 mL). Hematoma aspiration via the inserted catheter was easily achieved in 50 patients. In 3 patients, uncomplicated repositioning of the catheter was necessary after initial placement for optimal positioning within the hematoma before thrombolysis. 
   The average time from symptom onset until first aspiration was 3.9hours(ranging from 2 to 7 hours). The hematoma catheter was in place for a median duration of 5 days(range 2 to 8 days). During this time the average number of urokinase instillations was 13(range 5 to 26 times). Initial ICH volume was reduced by an average of 74.2%(range 65% to 88%) and the average final hematoma volume was 4.5 mL (range 2 to 12 mL). 
   Ventriculitis developed during procedure in 2 patients, and this complication was probably related to catheter instillation for long periods. And, there were instance of local rebleeding in originally presented sites developed in 3 patients. Repeat brain CT image performed after the first urokinase instillation revealed hematoma size increased, and the patients became less responsive. They remained severely disabled (GOS 3, 1 patient), 1 patient remained vegetative(GOS 2), and 1 patient died(GOS 1) at 6 months' follow-up. However, systemic hemorrhage related to procedure was not encountered in any patient. There were no instances of late clinical deterioration from mass effect or edema associated with residual hematoma. 
   Nine patients(16.9%) died before hospital discharge(1 from cardiac problems and 8 from respiratory failure). At discharge, 25 patients(47%) had achieved good recovery(17 patients GOS 3, 6 patient GOS 4, and 2 patients GOS 5), and 19 patient(35.8%) remained vegetative(GOS 2). And, at 6 months' follow up, 29 patients(55%) had achieved good recovery(17 patients GOS 3, 8 patient GOS 4, and 4 patients GOS 5), and 15 patient(28.3%) remained vegetative(GOS 2).

2. Evaluation of prognostic factors 
   Significant good prognostic factors for this procedure of ICH were found in our study to be young age, small pre-and-postoperative volume, low pre-and-postoperative GCS, absence of rebleeding, absence of underlying disease, and absence of complication of pneumonia after procedure. The other factors such as sex, time to procedure, total instillation volume, and total number of instillation were not meaningful prognostic factors in our study(Table 2). 

Discussion

1. Causes and natural history of IVH
   Forty six to 62% of IVH result from penetration of a hypertensive or arteriosclerotic intracerebral hemorrhage into ventricle. Nineteen to 29% of IVHs are caused by rupture of cerebral aneurysm, predominantly arising from anterior communicating artery and anterior cerebral artery. Less frequent causes of IVH are periventricular arteriovenous malformations, head trauma, and tumors.⁴⁾²⁸⁾
   Intraventricular hemorrhage carries a poor prognosis. The mortality rate for IVH is related to the amount of intraventricular blood and increases from 32.2% in cases with mild hemorrhage to 91% in cases with all ventricular chambers.^4)22)28)

2. Pathophysiology of IVH
   The clinical course in patients with IVH is determined by different pathogenetic mechanisms which must be considered when establishing an effective therapeutic strategy. At the time of bleeding, a sudden increase of intracranial pressure(ICP) occurs, which may lead to a significant decrease of cerebral perfusion and ischemic brain damage. A further raise of ICP and ventricular dilatation may occur due to obstruction of the flow of CSF at the foramen of Monro, aqueduct of Sylvius, or the fourth ventricle, depending on the amount and location of blood clots. Furthermore, intraventricular clotted blood, as well as accompanying intracerebral hemorrhage exert a direct mass effect upon adjacent brain structures. It has been shown that the prognosis in patients with IVH is directly related to the amount of intraventricular blood and the degree of ventricular dilatation on early image study. Thus, the aims of any specific therapy in severe IVH must be rapid elimination of intraventricular blood, diversion of ventricular CSF, reduction of ventricular dilatation and normalization of ICP. Obviously, this cannot be achieved by conservative medical therapy alone. Surgical blood removal through a conventional craniotomy has been advocated namely in cases with accompanying intracerebral hemorrhage. However, complete surgical removal of intraventricular blood may be hazardous or even impossible in cases in which all ventricular chambers are filled with blood. External ventricular drainage usually fails in the aim of blood elimination and normalization of ICP and ventricular size, as the catheters quickly become obstructed by clotted blood. Furthermore, ventricular drains have no effect on solid clots within the ventricles.^{7)8)9)18)19)21)}
   In the last decade, recognition of the proinflammatory role that certain blood components have on neuronal tissue led to a growing interest in inflammation as a mechanism of secondary brain injury. A blood component identified to play a role in the development of acute and chronic brain injury as well as degeneration is thrombin.²²⁾³¹⁾ After first developing an animal model of IVH, Pang et al have shown that blood and its product produce inflammation and fibrosis of ependymal lining.^{6)10)25)26)27)28)30)} Other previous experimental studies have shown that infusion of urokinase promotes clot lysis and restoration without producing neurotoxicity, histopathological alterations, or recurrent bleeding^{2)8)20)23)25)26)27)32)} And, preterm neonatal IVH differ from adult IVH although treatment strategies aim towards the same end-points. Neonatal IVH originates from the germinal matrix and despite efforts with intraventricular fibrinolytics, still develop major neurodevelopmental handicaps and approximately 50% require CSF shunting for permanent hydrocephalus.¹⁰⁾ Pang et al^{6)25)26)27)30)} speculated that intraventricular fibrinolytic treatment might prove useful for infants with perinatal IVH and older patients with IVH whose specific lesions have been definitively treated. 

3. Complications
   Several authors reported the rebleeding rate in CT-guided stereotactic surgery to be 3% to 16%.^12)14)17)24) The factors contributing to recurrent hemorrhage include excessive hematoma aspiration, intraoperative or postoperative hypertension, and a bleeding tendency. Because of the rebleeding risk that could potentially be increased by early aspiration suggest not to do the stereotactic aspiration before 6 to 24 hours after onset.^8)17)24) Hondo et al.¹³⁾ reported a rebleeding risk of only 4% when aspiration had been carried out between 5 and 48 hours after the hemorrhage. However, from our result, rebleeding after procedure seems to be not related to early aspiration. Even if the average time from symptom onset until first aspiration was 4.1 hours(ranging from 2 to 7 hours), the rebleeding risk cannot potentially be increased in our study (9%). Kandel et al.¹⁴⁾ developed a method of preventing recurrent bleeding after hematoma evacuation. After removal of the hematoma, the balloon catheter with a metal shift inside is introduced through the cannula into the cavity. Inflation continues until the pressure inside the balloon equals the pressure in the contralateral ventricle. 
   It is not clear whether the incidence of expanding hematoma in these above series represents any added risk from thrombolytic therapy. During the early period of time after ictus, hematoma may cause neurological deterioration as a result of an increasing mass effect caused by surrounding edema, and this mass effect may last up to 4 weeks after bleeding, even with decreasing density of clot. The risk of hematoma expansion during treatment must be closely monitored in future studies, including any associated untoward clinical sequelae, but this should also be compared with the substantial risk of spontaneous hematoma expansion in the first day among untreated patients.¹⁵⁾
   In our series, three patients of rebleeding and, two patients of ventriculitis due to ventricular drainage developed. These were the only treatment related complication. No intracranial adverse effects were observed during thrombolytic therapy.
   Mortality has been the primary end point of therapeutic studies in most published studies, and it has ranged from 30% to 90% 2,6.^11)15)29) This reflects in part patient inclusion and exclusion criteria, and to a lesser extent the treatment rendered in individual studies. In our series, there was 16.9% mortality with relatively large hematoma volume(>25 mL). These were cardiac and pulmonary compromised, admittedly selected, excluding deeply comatose patients. 
   Disability levels among surviving patients may be more relevant in assessment of management outcome. It is not clear from countless cases in published uncontrolled series whether IVH evacuation in fact enhances functional recovery. Such outcome assessment should be supplemented by documentation of quality of life domains relevant to patient and family, and these should be compared among treated and untreated cases. It may be advantageous to minimize stay in critical care unit and acute hospital settings even if eventual survival or disability level are not significantly altered by treatment. The state of consciousness is the best indicator of survival and that deficit of consciousness are not always a good indicator of functional prognosis. we concluded that stereotactic aspiration should have a definite place in the neurosurgeon's battery of therapeutic technique.

4. Prognosis 
   Volume of IVH is consistently shown to be a powerful predictor of poor outcome regardless of clot location, patient age, and neurological condition.⁵⁾⁶⁾ Most postoperative complications were seen in older patients and in those with severe neurological deficit or chronic disease. Factors contributing to poor outcome are as follows：age over 70 years, large hematoma, bad neurological grading. Stereotactic surgery is effective in regaining higher cerebral function in patients with a small hematoma less than 40 ml in volume, and with minor neurological deficit.¹²⁾ Larger hematomas result in more profound and longer-lasting alterations in adjacent brain parenchyma, attributed in part to mass effect and focal edema. The main prognostic factors affecting the outcome are clinical state of a patient on admissio