Early Surgical Treatment for Supratentorial Intracerebral Hemorrhage. A Randomized Feasibility Study

Transcription

1 Early Surgical Treatment for Supratentorial Intracerebral Hemorrhage A Randomized Feasibility Study Mario Zuccarello, MD; Thomas Brott, MD; Laurent Derex, MD; Rashmi Kothari, MD; Laura Sauerbeck, RN, BSN; John Tew, MD; Harry Van Loveren, MD; Hwa-Shain Yeh, MD; Thomas Tomsick, MD; Arthur Pancioli, MD; Jane Khoury, MS; Joseph Broderick, MD Background and Purpose The safety and the effectiveness of the surgical treatment of spontaneous intracerebral hemorrhage (ICH) remain controversial. To investigate the feasibility of urgent surgical evacuation of ICH, we conducted a small, randomized feasibility study of early surgical treatment versus current nonoperative management in patients with spontaneous supratentorial ICH. Methods Patients with spontaneous supratentorial ICH who presented to 1 university and 2 community hospitals were randomized to surgical treatment or best medical treatment. Principal eligibility criteria were ICH volume 10 cm 3 on baseline CT scan with a focal neurological deficit, Glasgow Coma Scale score 4 at the time of enrollment, randomization and therapy within 24 hours of symptom onset, surgery within 3 hours of randomization, and no evidence for ruptured aneurysm or arteriovenous malformation. The primary end point was the 3-month Glasgow Outcome Scale (GOS). A good outcome was defined as a 3-month GOS score 3. Results Twenty patients were randomized over 24 months, 9 to surgical intervention and 11 to medical treatment. The median time from onset of symptoms to presentation at the treating hospitals was 3 hours and 17 minutes, the time from randomization to surgery was 1 hour and 20 minutes, and the time from onset of symptoms to surgery was 8 hours and 35 minutes. The likelihood of a good outcome (primary outcome measure: GOS score 3) for the surgical treatment group (56%) did not differ significantly from the medical treatment group (36%). There was no significant difference in mortality at 3 months. Analysis of the secondary 3-month outcome measures showed a nonsignificant trend toward a better outcome in the surgical treatment group versus the medical treatment group for the median GOS, Barthel Index, and Rankin Scale and a significant difference in the National Institutes of Health Stroke Scale score (4 versus 14; P 0.04). Conclusions Very early surgical treatment for acute ICH is difficult to achieve but feasible at academic medical centers and community hospitals. The trend toward less 3-month morbidity with surgical intervention in patients with spontaneous supratentorial ICH warrants further investigation of very early clot removal in larger randomized clinical trials. (Stroke. 1999;30: ) Key Words: craniotomy intracerebral hemorrhage medical management stereotaxic aspiration surgical treatment Despite major advances in early CT diagnosis, improvements in neurosurgical critical care, and refinements in microsurgical techniques, there is no treatment that has been shown to be effective for the estimated patients who have an intracerebral hemorrhage (ICH) in the United States each year. 1 In the CT era, the 30-day mortality for ICH has been reported in the range of 35% to 52% in population studies. 2 5 One fifth of the survivors are independent at 6 months. 4 Only 9 randomized surgical 6 10 or medical trials for ICH have been reported. The 4 randomized medical trials for ICH, 2 testing steroids, 12,13 1 glycerol, 14 and 1 hemodilution, 11 did not show significant benefit for any therapy. The efficacy of surgical treatment for primary ICH is still controversial. Consequently, the management of ICH varies throughout the world. 15,16 This randomized feasibility study compares the outcome of patients with spontaneous parenchymal supratentorial ICH who are treated surgically with those who are given the best medical treatment. Received February 9, 1999; final revision received June 8, 1999; accepted June 16, From the Departments of Neurosurgery (M.Z., J.T., H. Van L., H-S.Y.), Neurology (L.S., J.B.), Emergency Medicine (R.K., A.P.), Radiology (T.T.), and Environmental Health (J.K.), University of Cincinnati Medical Center (Ohio); Department of Neurology, Mayo Clinic, Jacksonville, Fla (T.B.); and Service de Neurologie, Hopitaux de Lyon, Lyon, France (L.D.). Correspondence to Joseph Broderick, MD, University of Cincinnati Medical Center, Department of Neurology, 231 Bethesda Ave, Cincinnati, OH broderjp@ucsm 1999 American Heart Association, Inc. Stroke is available at

2 1834 Stroke September 1999 Subjects and Methods Selection of Patients Patients with spontaneous supratentorial ICH who presented to 1 university and 2 community hospitals from March 29, 1994, through April 9, 1996, were randomized to surgical treatment or best medical treatment. A history and examination, including an assessment of the baseline level of consciousness by the Glasgow Coma Scale (GCS) 17 and of the neurological status by the National Institutes of Health Stroke Scale (NIHSS), 18 were obtained at the time of admission. The time of onset of symptoms was determined on the basis of interview of the patient, family, and witnesses. If the onset was unobserved, it was considered to be at the last time the patient was definitely normal. The diagnosis of spontaneous ICH was made on acute onset of neurological symptoms and signs in the absence of trauma and confirmed by CT scan. Supratentorial ICH was classified according to the location of the hemorrhage into lobar, putaminal, thalamic, or basal ganglia (larger ICH in deep areas of the brain that involved 1 structure and did not clearly originate from the putamen or the thalamus). The baseline volume of the ICH was measured according to a bedside method of measuring CT ICH volume. The formula ABC/2 was used, where A is the greatest hemorrhage diameter by CT, B is the diameter 90 degrees to A, and C is the approximate number of 10-mm CT slices with hemorrhage. The measurements by the ABC/2 method have been shown to correlate highly with the volumes calculated by planimetric methods for all ICH locations in a recent study, with excellent interrater and intrarater reliability. 19 Inclusion criteria were supratentorial ICH diagnosed by CT scan, ICH volume 10 cm 3 with a focal neurological deficit, age 18 years, and GCS score 4 at the time of enrollment; in addition, diagnosis, enrollment and randomization, and therapy were to be instituted within 24 hours of onset of clinical symptoms, and initiation of surgery within 3 hours of randomization was required. An ICH volume of 10 cm 3 was chosen because of previous work by the authors demonstrating that many persons with a 10- to 20-cm 3 ICH have substantial morbidity at 1 month after onset. 20 Patients were excluded from the study for the following reasons: lack of neurological deficit, infratentorial ICH, CT scan suggestive of underlying structural vascular abnormality such as arteriovenous malformation or aneurysm (presence of subarachnoid hemorrhage), terminal medical illness, coagulopathy accounting for the hemorrhage, traumatic ICH, pregnancy, and lack of informed consent. The operating room and anesthesia staff were notified of possible emergent ICH evacuation at the time of initial diagnosis. A list of random sequential assignment of patients to medical or surgical treatment was generated by the study statistician. Each patient assignment was placed in an individual sequentially numbered opaque envelope that was to be opened by the nurse investigator after she was contacted by the physician investigator and informed that signed informed consent had been obtained. This study was approved by the institutional review board at each of the participating study hospitals. Informed consent was obtained from the patient or, if the patient could not provide informed consent, from the patient s legally designated representative. Surgical Treatment After the informed consent process and randomization, those patients selected for surgical therapy were taken to the operating room as soon as possible, where the ICH was removed and bleeding was controlled with the use of standard neurosurgical techniques. The surgical approach was individualized on the basis of the site and size of the ICH. Allowed techniques included open craniotomy and CT-guided stereotaxic placement of a catheter for evacuation of supratentorial ICH. In a deepseated ICH, stereotaxic evacuation was the procedure of choice. The intention of surgical treatment was complete removal of the clot. The protocol for stereotaxic aspiration included the following steps, in order of completion: (1) Placement of stereotaxic frame was performed; (2) the patient had stereotaxic CT to check placement of frame; (3) the patient was taken to operating room; (4) in presence of basal ganglia hemorrhage, a large frontal burr hole was used; (5) in presence of lobar hemorrhage, the burr hole was placed over the affected lobe; (6) local or general anesthesia was administered, and introducer cannula was placed stereotaxically into the center of the clot; (7) initial aspiration and evacuation of hemorrhage were performed; (8) a catheter was placed via an introducer cannula; (9) the patient was taken for repeated CT scan to check catheter placement; (10) 6000 U of urokinase was injected through the catheter to facilitate aspiration, and the catheter was clamped; (11) the CT scan was repeated 12 hours later, and in the presence of residual clot, aspiration through the catheter was repeated, followed by injection of 6000 U of urokinase; (12) this process was repeated every 12 hours until 80% of clot was evacuated or at the discretion of the neurovascular team. Neurological deterioration at any point prompted a safety CT scan. Medical Treatment The surgical and medical patients were managed by the same team of physicians to ensure comparable management in both groups at each center. At the discretion of the primary team, patients in both groups might have additional diagnostic tests, such as brain MRI and cerebral angiography. All patients, medical and surgical, were cared for in an intensive care unit setting for 24 hours or until they were considered stable enough to move to an intermediate care or general unit. Neurological status was monitored in the intensive care unit by daily NIHSS and hourly neurological evaluation, which included limb strength, level of consciousness, and vital signs. Treatment of ICH was delivered according to current practices at a university hospital, yet treatment was not rigidly regimented, and the primary attending neurosurgeon/neurologist was allowed to use his or her best medical judgment. Therapy included pneumatic compression boots for deep vein thrombosis prophylaxis, physical therapy, intravenous fluids, H 2 blockers, maintenance of normoglycemia, and early nutritional support. Intubation was performed in patients as needed for respiratory depression,

3 Zuccarello et al Early Surgical Treatment for Supratentorial ICH 1835 airway protection, and/or control of intracranial pressure (ICP). Hyperventilation was used if ICP remained 20 mm Hg despite ventricular drainage, mannitol, and sedation/paralysis. If hyperventilation was used, PCO 2 was maintained between 25 and 30 mm Hg. Hyperventilation was weaned when ICP was 15 and was continued for no longer than 24 hours at any one interval. Steroids were discouraged, but their use by an individual physician did not exclude patients from the study. Hypertension was regulated early in the course of therapy. Mean arterial blood pressure (MAP) was maintained between 100 and 130 mm Hg by appropriate antihypertensive therapy. Treatment included invasive monitoring of ICP when indicated. In the presence of an ICP monitor, cerebral perfusion pressure (MAP ICP) was maintained at 70 to 100 mm Hg. Crossover and Reoperation Crossover from the medical treatment group to the surgical treatment group was discouraged, and repeated surgery was also discouraged. Ventricular catheterization of hydrocephalus was allowed and did not constitute a crossover in medical treatment group patients or reoperation in surgical treatment group patients. Adverse Events Patients were monitored for adverse events and particularly serious adverse events, which included intraoperative death, all deaths, crossover or reoperation, rebleeding, surgical wound infection, and intracranial infection (meningitis, ventriculitis, abscess). Assessments and End Points Outcome was measured at the time of discharge from the hospital to home or rehabilitation facility and at 3 months. The 3-month Glasgow Outcome Scale (GOS) 21 was the major end point. A good outcome was defined prospectively as a 3-month GOS score 3. Primary intent-to-treat analysis compared all patients randomized to medical or surgical treatment. Neurological status was assessed by the NIHSS. Other outcome measurements included mortality, GOS, 21 Barthel Index (BI), 22 and modified Rankin Scale (RS). 23 Patients who died were assigned the worst possible score on each outcome measure (eg, GOS of 1). The nurse coordinator, who was not involved in the treatment of patients, performed the 3-month outcome evaluation. All baseline, postoperative (in surgical patients), 24-hour, and 3-month CT scans were evaluated by the study neuroradiologist for ICH location, ICH volume, peri-ich edema, and intraventricular extension. Statistical Analysis Data were collected and entered onto forms by the study nurse and then entered into DataEase, where the data were checked by predefined criteria. Data were subsequently managed and analyzed with the use of SAS (SAS Institute). The primary hypothesis to be tested was that surgical patients have a significantly higher rate of a good outcome (GOS 3) than medical patients. Wilcoxon rank sum test and 2 or Fisher s exact test were used for univariate comparisons as appropriate for continuous or categorical variables because of sample size and distribution of the data. A value of P 0.05 was considered statistically significant. Median values are used to present data because the data for most variables were not normally distributed. Power calculations are presented to provide information regarding further study. Results Twenty eligible consenting patients were randomized over 24 months, 9 to surgical intervention (craniotomy [n 5] or stereotaxic evacuation [n 4]) and 11 to best medical therapy. In 24 months the 3 participating hospitals were required to enroll the 20 study patients. We did not prospectively collect data concerning patients with ICH who presented to these hospitals but were not randomized. However, the annual number of patients with ICH seen at these 3 hospitals is approximately 100 per year on the basis of our earlier population-based studies of ICH. Accordingly, only 10% of patients presenting with ICH were enrolled in the study. Most patients were excluded because of failure to meet study inclusion/exclusion criteria, most often because the patient was in extremis, was too mildly affected, or had an ICH 10 cm 3. In a few instances, a patient was not entered into the study because of the reluctance of the treating neurosurgeon to randomize the patient. One crossover from the medical group to craniotomy occurred at day 4. The clinical status of the patient had deteriorated, and the attending neurosurgical physician performed a craniotomy and clot evacuation. As the patient was randomized to the medical treatment group, he was analyzed as such. One reoperation was performed 13 days after craniotomy for an infected scalp incision and cerebral abscess. The patient recovered well initially but died at day 55 from intracerebral rebleeding. Ventricular catheters were placed in 3 of 11 (27%) of the medical patients. The surgical and the medical groups were comparable with respect to baseline characteristics (age, sex, race, NIHSS score, GCS score, median time from onset to admission, ICH volume). The median (25th percentile, 75th percentile) baseline NIHSS score was 20 (19, 22) in the surgical treatment group and 21 (13, 26) in the medical treatment group (P 1.0) (Table 1). The median baseline GCS score was 13 (11, 14) in the surgical treatment group and 11 (6, 13) in the medical treatment group (P 0.16). The median baseline ICH volume was 35 cm 3 (19, 63 cm 3 ) in the surgical treatment group and 30 cm 3 (19, 50 cm 3 ) in the medical treatment group (P 0.79) (Table 2). The median time from onset of symptoms to presentation at the treating hospitals was 3 hours and 17 minutes, the time from admission to randomization was 3 hours and 10 minutes, and the time from randomization to surgery was 1 hour and 20 minutes. The median time from admission to surgery was 4 hours and 3 minutes, and the time from onset of symptoms to surgery was 8 hours and 35 minutes. Surgery was begun 12 hours from symptom onset in only 2 patients. One patient s family could not be reached and interviewed for informed consent for many hours. The other patient s surgery was delayed because of waiting for an available operating room.

4 1836 Stroke September 1999 TABLE 1. Three-Month Outcome by Treatment Groups, ICH Location, and Baseline NIHSS Patient No. ICH Day 0 NIHSS NIHSS BI RS GOS Surgical treatment (n 9) 203 (S) L putamen (S) R putamen (S) R putamen (S) R lobar (C) R basal ganglia 22 42* 0* 5* 1* 207 (C) R lobar 34 42* 0* 5* 1* 214 (C) L lobar (C) R lobar (C) L lobar Median (25%, 75%) NA 20 (19, 22) 4 (3, 8) 60 (80, 90) 2 (2, 5) 4 (3, 4) Medical treatment (n 11) 201 L lobar R thalamus 26 42* 0* 5* 1* 206 R thalamus L thalamus 29 42* 0* 5* 1* 210 L lobar 6 PU R putamen R basal ganglia 29 42* 0* 5* 1* 216 R lobar L putamen 18 PU L lobar L lobar Median (25%, 75%) NA 21 (13, 26) 14 (7, 420) 25 (0, 90) 4 (3, 5) 2 (1, 4) S indicates stereotaxic surgery; C, craniotomy; L, left; R, right; NA, not applicable; and PU, permanently unavailable (at 3-month evaluation both patients had moved to or near Chicago, and evaluation was made by phone interview. In-person NIHSS could not be done). *Dead (worst possible 3-month outcome score was given). Antihypertensive therapy (labetalol and/or nitroprusside) was used in all 20 patients. Mannitol was used in 8 patients (40%): 3 patients in the surgical treatment group and 5 patients in the medical treatment group. Steroids were used in 2 patients in the surgical treatment group. Intubation and mechanical ventilation were performed in 16 patients (80%): 7 patients in the surgical treatment group (2 with stereotaxic surgery and 5 with craniotomy) and 9 patients in the medical treatment group. ICP monitoring was performed in 4 patients (20%): 1 patient in the surgical treatment group (craniotomy group) and 3 patients in the medical treatment group. For the primary end point, no statistically significant difference of good outcome (3-month GOS score 3) rate was noted in the surgical treatment group (56%) compared with the medical treatment group (36%). Three-month outcomes by treatment groups, demographic characteristics, baseline NIHSS, ICH location, baseline ICH volume, and ICH volume variation between baseline and 24-hour CT are presented in Tables 1 to 3. There was no significant difference in mortality at 3 months (surgical treatment group, 22% [2/9]; medical treatment group, 27% [3/11]). No intraoperative death was observed. In the surgical treatment group, causes of death were cardiorespiratory failure following intracerebral rebleeding at day 55 and respiratory failure following pneumonia and renal failure (craniotomy had been performed in both patients). In the medical treatment group, causes of death were cardiopulmonary arrest following pneumonia, sepsis, and circulatory collapse; respiratory failure following pneumonia; and brain herniation. Analysis of the outcome measures showed a nonsignificant trend toward a better 3-month outcome (median GOS, BI, and RS) in the surgical treatment group compared with the best medical treatment group (Table 3). The median 3-month NIHSS score was the only statistically significant finding in favor of the surgical treatment (P 0.04). Although the surgical treatment group and the medical treatment group were comparable with respect to baseline NIHSS score, the median (25th percentile, 75th percentile) NIHSS score improvement from baseline to 3-month score was not significantly better in the surgical treatment group ( 12 [ 18, 10]) compared with the medical treatment group ( 2[ 12, 13]), respectively) (P 0.1). The 4 patients who received stereotaxic evacuation were independent at 3 months (BI scores 90, 100, 100, 85) with little or no disability. No significant difference was noted in either baseline character-

5 Zuccarello et al Early Surgical Treatment for Supratentorial ICH 1837 TABLE 2. Three-Month Outcome by Treatment Groups, Demographic Variables, and Baseline and 24-Hour CT Data Patient No. Sex, Race, Age Day 0 GCS Day 0 ICH Volume, cm 3 24-Hour ICH Volume, cm 3 % Change In ICH Volume RS Surgical treatment (n 9) 203 (S) M, W, (S) M, B, (S) F, W, (S) F, B, (C) F, W, (VE) 35 (VE) 44 5* 207 (C) M, B, (VE) 19 (VE) 76 5* 214 (C) F, B, (C) F, W, (C) M, W, (VE) 9 (VE) 91 5 Median (25%, 75%) 63 (59, 66) 35 (19, 63) 16 (9, 16) 44 ( 76, 44) 2 (2, 5) Medical treatment (n 11) 201 M, B, M, W, (VE) 24 (VE) 0 5* 206 M, W, (VE) 30 (VE) M, W, (VE) 25 (VE) 39 5* 210 F, W, F, W, F, B, (VE) 67 (VE) 16 5* 216 F, W, (VE) M, W, (VE) 60 (VE) M, B, (VE) 90 (VE) M, B, (VE) 50 (VE) 0 2 Median (25%, 75%) 65 (59, 71) 30 (19, 50) 43 (25, 60) 0 ( 5, 39) 4 (3, 5) S indicates stereotaxic surgery; C, craniotomy; W, white; B, black; and VE, ventricular extension of ICH. *Dead (worst possible 3-month outcome score was given). istics or 3-month outcome in the subgroup of lobar ICH patients randomized to the surgical treatment group (n 5) and to the medical treatment group (n 5). As expected, clot evacuation resulted in lower ICH volumes. The median ICH volume on the 24-hour CT scan was significantly lower in the surgical treatment group (16 cm 3 ) than in the medical treatment group (43 cm 3 )(P 0.003). The reduction in median ICH volume from the baseline to the 24-hour CT scan was significantly higher in the surgical treatment group ( 26 cm 3 ) compared with the medical treatment group (0 cm 3 )(P ). The percent reduction TABLE 3. Comparison of Outcome Measures In the Surgical and Medical Groups Surgical Group (n 9) Medical Group (n 11) GOS 3 56% (n 5) 36% (n 4) GOS* 4 (3, 4) 3 (1, 4) RS* 2 (2, 5) 4 (3, 5) BI 75 44% (n 4) 27% (n 3) BI* 60 (8, 90) 25 (0, 90) NIHSS* 4 (3, 8) 14 (7, 42) *Median (25%, 75%). of median ICH volume from the baseline to the 24-hour CT scan was 44% in the surgical treatment group. The reduction of median ICH volume from the baseline to the 24-hour CT scan was significantly higher in the craniotomy group ( 29 cm 3 ) compared with the stereotaxic group ( 7.5 cm 3 ) (P 0.04), as was the percent reduction of median ICH volume from the baseline to the 24-hour CT scan ( 76% in the craniotomy group compared with 39% in the stereotaxic group) (P 0.05). Discussion Our study demonstrates the feasibility of a larger randomized clinical trial of early surgical treatment for acute spontaneous ICH at academic medical centers as well as community hospitals. However, the median time from onset of symptoms to surgery in this pilot study was 8 hours and 35 minutes, well beyond the 3-hour window necessary for the successful use of thrombolytic therapy for ischemic stroke. 24 Addressing logistical barriers to very early treatment with thrombolytic therapy was successful in the National Institute of Neurological Disorders and Stroke t-pa Stroke Trial of acute ischemic stroke. Our data suggest that resolving logistical issues could also substantially decrease the time to operation since the median time from admission to surgery was 4 hours and 3

6 1838 Stroke September 1999 TABLE 4. Summary of Randomized Controlled Surgical Trials for Spontaneous Supratentorial ICH Patients Time Window Author Year Surgical Medical From Onset to Surgery, h Open craniotomy McKissock et al Juvela et al Batjer et al Morgenstern et al Stereotaxically guided endoscopic evacuation Auer et al minutes. Very early and enthusiastic involvement of neurosurgeons and rapid trauma-like deployment of the operative team will likely be the critical rate-limiting steps in decreasing the time from symptom onset to surgical treatment. Only 1 of the 5 previous randomized surgical studies focused on an early treatment window despite a good rationale for early intervention (Table 4). For example, ongoing bleeding commonly occurs within the first several hours after the onset of spontaneous ICH In a prospective observational study of 103 patients with ICH imaged by CT within 3 hours of onset, 26% of patients had substantial growth in the volume of parenchymal hemorrhage between the baseline CT and a CT 1 hour later. An additional 12% of patients had substantial growth between the 1- and 20-hour CT scans. Hemorrhage growth between the baseline and 1-hour CT scans was significantly associated with clinical deterioration, as measured by the change between the baseline and 1-hour GCS and NIHSS scores. 28 In addition, it has been demonstrated that serum proteins originating from the intracerebral hematoma itself accumulate in adjacent white matter and result in early (1 hour) and prolonged (8 hours) edema after experimental lobar ICH in a pig model. 29 This interstitial edema likely corresponds to the early CT scan perihematomal hypodensities present in 70% of patients with ICHs studied within 3 hours of symptom onset. 30 Clot removal at 3 hours has been proven to markedly reduce mass effect and perihematomal edema at 24 hours in the same pig model of ICH. 31 Consequently, we believe that randomized trials of surgical clot evacuation should ideally test the efficacy of clot removal within the first several hours after onset of ICH. If very early surgery can remove most of the original hemorrhage with minimal additional brain tissue damage, physicians may be able to reduce edema development and mass effect, prevent white matter injury, and improve clinical outcome in some patients. 32 This small feasibility study was not designed with sufficient statistical power to demonstrate a difference in outcome between surgery and medical therapy. Feasibility of early surgical intervention, not proof of surgical efficacy, was the primary goal. Accordingly, the lack of a significant difference in good outcome in the surgical group compared with the medical group was expected and does not rule out a benefit for surgical clot removal. For example, a nonsignificant trend toward a better 3-month outcome was observed in all the outcome measures (GOS, BI, RS) in the surgical group compared with the medical group, although the NIHSS score was the only significant finding in favor of surgery (P 0.04). In another small study of 34 patients randomized within 12 hours of onset, Morgenstern and colleagues observed that the 6-month mortality for surgically treated patients (19%) was similar to that for those treated medically (24%). 10 The major goal of surgical treatment in patients with an ICH is safe and thorough clot evacuation with maximal preservation of neurological function. Regarding surgical technique, the median percent reduction in the volume of ICH from the baseline to the 24-hour CT scan was significantly higher in the craniotomy group than in the stereotaxic group (P 0.05). Aspiration of the jellylike hyperacute clots through a catheter is more difficult but less invasive than a craniotomy, which provides wider exposure and access to the hematoma. The installation of urokinase through the catheter into the clot helps to accelerate clot removal via stereotaxic aspiration. However, a randomized trial of craniotomy versus stereotaxic surgical clot removal in patients with a superficial ICH would be necessary to compare differences in technical and clinical outcomes. Although the surgical and medical groups were relatively well balanced with regard to several baseline variables that are related to outcome (such as age, NIHSS and GCS scores, and ICH volume), our pilot study has limitations. For example, the 2 groups were not comparable with regard to location of ICH. All 3 thalamic ICHs were randomized to the medical group, 2 of whom died, and the third had a 3-month RS score of 5 (Table 1). In addition, there was a trend (P 0.08) toward more intraventricular extension in the medical treatment group (73%) compared with the surgical treatment group (33%). None of the stereotaxic group patients had intraventricular extension of ICH. Intraventricular extension of ICH was a strong predictor of poor outcome in previous studies. 20 In our study, 90% of patients with intraventricular extension of ICH on the 24-hour CT scan (10 of 11) died or had a 3-month RS score of 4 (moderately severe disability) or 5 (severe disability). All of the study patients who died had intraventricular extension of ICH. Future randomized surgical trials should carefully consider whether to include ICH patients with large volumes of intraventricular hemorrhage or to stratify randomization by the presence or absence of intraventricular hemorrhage. A pilot study of intraventricular administration of thrombolysis in patients with intraventricular hemorrhage is currently under development (Daniel Hanley, MD, personal communication, 1998). In summary, this pilot randomized trial of early surgical therapy for spontaneous ICH demonstrates the feasibility of a much larger randomized treatment trial of very early clot removal in patients with an ICH. Our results allow estimation of the sample size needed for a multicenter, randomized trial to test the hypothesis that urgent surgery improves 3-month outcome in spontaneous supratentorial ICH. With power of 80%, a 2-sided significance level of 0.05, and if a 3-month mortality or poor outcome (GOS of 3) of 44% in the surgical treatment group and 64% in the medical treatment group is assumed, a sample size of 107 per group would be

7 Zuccarello et al Early Surgical Treatment for Supratentorial ICH 1839 needed. A major challenge of such a study would be whether surgical treatment can be initiated within 3 to 6 hours of symptom onset, which is the current therapeutic window for treatment of acute ischemic stroke. Acknowledgment This study was supported by National Institute of Neurological Disorders and Stroke grant R-01 NS References 1. Broderick J, Brott T, Tomsick T, Miller, R, Huster G. Intracerebral hemorrhage is more than twice as common as subarachnoid hemorrhage. J Neurosurg. 1993;78: Anderson C, Chakera T, Stewart-Wynne E, Jamrozik K. Spectrum of primary intracerebral haemorrhage in Perth, Western Australia, : incidence and outcome. J Neurol Neurosurg Psychiatry. 1994;57: Broderick J, Brott T, Tomsick T, Huster G, Miller R. The risk of subarachnoid and intracerebral hemorrhages in blacks as compared with whites. N Engl J Med. 1992;326: Counsell C, Boonyakarnkul S, Dennis M, Sandercock P, Bamford J, Burn J, Warlow C. Primary intracerebral haemorrhage in the Oxfordshire Community Stroke Project, II: prognosis. Cerebrovasc Dis. 1995;5: Giroud M, Gras P, Chadan N, Beuriat P, Milan C, Arveux P, Dumas R. Cerebral haemorrhage in a French prospective population study. J Neurol Neurosurg Psychiatry. 1991;54: Auer LM, Deinsberger W, Niederkorn K, Gell G, Kleinert R, Schneider G, Holzer P, Bone G, Mokry M, Korner E. Endoscopic surgery versus medical treatment for spontaneous intracerebral hematoma: a randomized study. J Neurosurg. 1989;70: Batjer HH, Reisch JS, Allen BC, Plaizier LJ, Su CJ. Failure of surgery to improve outcome in hypertensive putaminal hemorrhage: a prospective randomized trial. Arch Neurol. 1990;47: Juvela S, Heiskanen O, Poranen A, Valtonen S, Kuurne T, Kaste M, Troupp H. The treatment of spontaneous intracerebral hemorrhage: a prospective randomized trial of surgical and conservative treatment. J Neurosurg. 1989;70: McKissock W, Richardson A, Taylor J. Primary intracerebral hemorrhage: a controlled trial of surgical and conservative treatment in 180 unselected cases. Lancet. 1961;2: Morgenstern LB, Frankowski RF, Shedden P, Pasteur W, Grotta JC. Surgical Treatment for Intracerebral Hemorrhage (STICH): a singlecenter, randomized clinical trial. Neurology. 1998;51: Italian Acute Stroke Study Group. Haemodilution in acute stroke: results of the Italian haemodilution trial. Lancet. 1988;1: Poungvarin N, Bhoopat W, Viriyavejakul A, Rodprasert P, Buranasiri P, Sukondhabhant S, Hensley MJ, Strom BL. Effects of dexamethasone in primary supratentorial intracerebral hemorrhage. N Engl J Med. 1987; 316: Tellez H, Bauer R. Dexamethasone as treatment in cerebrovascular disease, I: a controlled study in ICH. Stroke. 1973;4: Yu YL, Kumana CR, Lauder IJ, Cheung YK, Chan FL, Kou M, Chang CM, Cheung RT, Fong KY. Treatment of acute cerebral hemorrhage with intravenous glycerol: a double-blind, placebo-controlled, randomized trial. Stroke. 1992;23: Broderick J, Brott T, Tomsick T, Tew J, Duldner J, Huster G. Management of intracerebral hemorrhage in a large metropolitan population. Neurosurgery. 1994;34: Masdeu J, Rubino F. Management of lobar intracerebral hemorrhage: medical or surgical. Neurology. 1984;34: Teasdale G, Jennett B. Assessment of coma and impaired consciousness: a practical scale. Lancet. 1974;2: Brott T, Adams HJ, Olinger C, Marler J, Barsan W, Biller J, Spilker J, Holleran R, Eberle R, Hertzberg V, Rorick M, Moomaw C, Walker M. Measurements of acute cerebral infarction: a clinical examination scale. Stroke. 1989;20: Kothari R, Brott T, Broderick J, Barsan W, Sauerbeck L, Zuccarello M. The ABCs of measuring intracerebral hemorrhage volume. Stroke. 1996; 27: Broderick JP, Brott TG, Duldner JE, Tomsick T, Huster G. Volume of intracerebral hemorrhage: a powerful and easy-to-use predictor of 30-day mortality. Stroke. 1993;24: Jennett B, Bond M. Assessment of outcome after severe brain injury: a practical scale. Lancet. 1975;1: Mahoney F, Barthel D. Functional evaluation: the Barthel Index. Md State Med J. 1965;14: Rankin J. Cerebral vascular accidents in patients over the age of 60, II: prognosis. Scott Med J. 1957;2: NINDS rt-pa Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. N Engl J Med. 1995;333: Broderick J, Brott T, Tomsick T, Barsan W, Spilker J. Ultra-early evaluation of ICH. J Neurosurg. 1990;72: Fujii Y, Tanaka R, Takeuchi S, Koike T, Minakawa T, Sasaki O. Hematoma enlargement in spontaneous ICH. J Neurosurg. 1994;80: Kelley R, Berger J, Scheinberg P, Stokes N. Active bleeding in hypertensive ICH: computed tomography. Neurology. 1982;32: Brott T, Broderick J, Kothari R, Barsan W, Tomsick T, Sauerbeck L, Spilker J, Duldner J, Khoury J. Early hemorrhage growth in patients with intracerebral hemorrhage. Stroke. 1997;28: Wagner KR, Xi G, Hua Y, Kleinholz M, de Courten-Myers GM, Myers RE, Broderick JP, Brott TG. Lobar intracerebral hemorrhage model in pigs: rapid edema development in perihematomal white matter. Stroke. 1996;27: Broderick J, Brott T, Kothari R. Very early edema growth with ICH. Stroke. 1995;26:184. Abstract. 31. Wagner KR, Xi G, Hua Y, Zuccarello M, Banks A, de Courten-Myers GM, Myers RE, Brott TG, Broderick JP. Clot removal following lysis with tissue plasminogen activator markedly reduces perihematomal edema in an intracerebral hemorrhage model. Stroke. 1996;27:183. Abstract. 32. Tuhrim S, Dambrosia JM, Price TR, Mohr JP, Wolf PA, Hier DB, Kase CS. Intracerebral hemorrhage: external validation and extension of a model for prediction of 30-day survival. Ann Neurol. 1991;29: