Am Fam Physician. 2005;72(9):1868-1870
Each year, about 20 out of every 100,000 persons has a spontaneous supratentorial intracerebral hemorrhage. Hemorrhage accounts for approximately 20 percent of all cases of sudden stroke-related neurologic deficits. Intracerebral hemorrhage has the highest mortality and morbidity rate of any kind of stroke; more than 40 percent of these patients die, and the majority of survivors are significantly disabled. The biggest controversy in management of this condition involves removal of the hematoma to reduce brain edema and ischemia in the penumbra of impaired, but potentially viable, surrounding tissue. In 1961, a clinical trial reported worse outcomes in patients treated surgically compared with those who received conservative treatment. Trials since then have given conflicting results, but many have had small numbers of patients or methodologic difficulties. Mendelow and colleagues led a team to compare early surgical intervention with conservative therapy—the International Surgical Trial in Intracerebral Hemorrhage.
The trial randomized 1,033 patients in 27 countries to early surgery or conservative treatment. Eligible patients had computed tomographic evidence of intracerebral hemorrhage within 72 hours, with a minimum hematoma diameter of at least 2 cm and a Glasgow Coma Scale (GCS) score of at least 5. Excluded were patients with evidence of hemorrhage likely caused by aneurysm, hemorrhage secondary to tumor or trauma, cerebellar or brainstem involvement, severe preexisting conditions that could inf luence outcomes, or delay of surgery beyond 24 hours. In all cases, the neurosurgeon had to be uncertain about the relative benefits of neurosurgery or conservative treatment. Patients allocated to surgery underwent the procedure deemed most appropriate by the local neurosurgeon within 24 hours. Patients randomized to conservative treatment were provided with the best medical treatment, in the judgment of the local physicians. The primary outcome was death or disability, measured using the Glasgow Outcome Scale, at six months. Data on secondary outcome measures were gathered using standardized validated outcome measures such as the Barthel index and modified Rankin scale.
The 503 patients allocated to early surgery were similar in all significant variables to the 530 allocated to conservative management. The median age was 62 years (range, 19 to 93 years), 57 percent were men, and 41 percent had GCS scores of 13 or higher. The median hematoma volume (measured using the Broderick method) was 38 mL (range, 4 to 210 mL), and the median depth was 1 cm from the cortical surface. At two weeks, 496 patients were analyzed from the surgery group and 529 patients were analyzed from the conservative treatment group. Six months after the event data were analyzed, 477 patients were randomized to surgery, and 505 patients were randomized to medical care (51 patients were lost to follow-up). In the conservative treatment group, 140 patients (26 percent) underwent surgery after initial therapy. The most common indication for surgery in this group was neurologic deterioration (82 patients), followed by clinical deterioration (20 patients) and rebleeding (17 patients). Data were not available on nine patients in the surgical group and eight patients in the conservative treatment group who were known to be alive at the six-month analysis. Overall, a favorable outcome at six months was reported in 26 percent of surgical patients and 24 percent of conservative treatment patients (see accompanying table). Mortality rates did not differ significantly between the groups. Patients with hematomas located 1 cm or less from the cortical surface were more likely to have favorable outcomes from early surgery, but comatose patients were more likely to have worse outcomes from early surgery.
|Early surgery, n = 468 (%)||Initial conservative treatment, n = 497 (%)||Absolute benefit (95% CI)|
|Favorable||122 (26)||118 (24)||2.3 (−3.2 to 7.7)|
|Unfavorable||346 (74)||378 (76)||—|
|Alive*||304 (64)||316 (63)||1.2 (−4.9 to 7.2)|
|Dead||173 (36)||189 (37)||—|
|Prognosis-based modified Rankin index|
|Favorable||153 (33)||137 (28)||4.7 (−1.2 to 10.5)|
|Unfavorable||312 (67)||351 (72)||—|
|Prognosis-based Barthel index|
|Favorable||124 (27)||110 (23)||4.1 (−1.4 to 9.5)|
|Unfavorable||341 (73)||377 (77)||—|
The authors conclude that outcomes in patients with intracerebral hemorrhage do not differ significantly with early neurosurgical intervention compared with conservative management. This finding concurs with meta-analyses conducted on all prior trials. Despite the many technical and other problems in conducting clinical trials in this topic, the authors call for further studies to investigate new treatments and to identify subgroups of patients who could benefit from specific interventions. Overall, they do not find sufficient evidence to support a general policy of early operative intervention in patients with spontaneous supratentorial intracerebral hemorrhage.