Does Immediate Brain Hypothermia Improve Outcome of Patients with Poor-grade Subarachnoid Hemorrhage?

To evaluate the protective effect of hypothermia in severe subarachnoid hemorrhage (SAH), we induced mild brain hypothermia by surface cooling in 16 patients with WFNS Grade V SAH immediately after resuscitation and diagnosis by computed tomography (CT). Between April 1999 and September 2000, 10 men and 6 women between 50 and 72 years old (average 63.1), were studied. Those over 75 years and with no brain stem reaction, obvious hypoxemia, hypotension, or huge hematoma in the dominant cerebral hemisphere were excluded. The level of consciousness assessed by the Glasgow Coma Scale (GCS) was 3 in 8 (including 2 resuscitated from cardiopulmonary arrest), 4 in 1, 5 in 3, and 6 in 4. Four patients showed massive intracerebral or subdural hematoma associated with deviation of midline structure and 4 showed intraventricular hematoma with cast formation. All but 1 showed pupillary abnormality. Blood glucose exceeded 160mg/dl in 13. Median time from onset to arrival was 33min, cerebral angiography 86min, and surgery 174min. Core patient temperature reached 35.1°C just before surgery, 34.1°C at the beginning of microsurgery, and 33.6°C before leaving the operating room. After complete obliteration of an aneurysm by clipping and external decompression, patients were maintained under mild hypothermia at 33-34°C for 48 hours or more. Patient outcome as assessed by the Glasgow Outcome Scale at 6 months was as follows: good recovery (GR) in 2, moderate disability (MD) in 2, severe disability (SD) in 8, persistent vegetative state (VS) in 2, and death (D) in 1. No mortality related to hypothermia was encountered. Poor outcome was mostly related to primary brain damage, 3 caused by intracerebral hematoma, 1 by brain herniation, 1 by hypoxic damage, 3 by rerupture of the aneurysm before surgery, and 1 by delayed ischemia due to cerebral vasospasm. Overall outcome of consecutive 22 patients with Grade V SAH, including 6 without immediate brain hypothermia, was favorable (GR+MD) in 22.5% and unfavorable (SD+VS+D) in 77.3% of patients. In the comparison with a historical cohort of 24 patients with Grade V SAH between January 1998 and March 1999, decrease of VS+D and increase of SD were significant. Immediate induction of mild hypothermia possibly reduces primary brain damage and improves outcome in patients with WFNS Grade V SAH.