Barbiturate coma for vasospasm following SAH
- Richard Venn
© Current Science Ltd 1999
Published: 19 May 1999
KeywordsBarbiturate papaverine subarachnoid haemorrhage vasospasm
The protective effects of burst suppression barbiturate therapy has been proven in animal models of cerebral ischaemia but unfortunately dismal results have been shown in patients presumably because irreversible damage has occurred prior to therapy commencing. This study in a small number of patients shows optimism possibly because identification of vasospasm and therefore potential ischaemia was prompt allowing early initiation of treatment. However much larger studies are required prior to embarking on a treatment which requires complicated intensive care supportive therapy due to potential serious adverse events.
Cerebral artery vasospasm accounts for nearly a quarter of the mortality following subarachnoid haemorrhage (SAH). Treatments include calcium antagonists, triple H therapy and angioplasty (chemical and physical) if available, although significant morbidity and mortality remain. Experimentally EEG burst suppression induced by barbiturates protects against cerebral ischaemia. Unfortunately studies in 1980 utilising this technique in patients with ischaemia produced by vasospasm produced grim results and further investigation was halted.
To investigate the role of barbiturate coma in the management of vasospasm resistant to other treatments including angioplasty.
Patients following urgent aneurysmal repair are managed according to a previously published protocol on the intensive care unit (ICU). Essentially vasospasm detected electively by angiography at day 5-7 following SAH (maximum risk) or earlier if clinically indicated is treated with papaverine injected into the appropriate artery. Arterial systolic pressure is increased to 160-180 mmHg and papaverine angioplasty is repeated each day until vasospasm resolves. Balloon angioplasty is attempted if vasospasm persists and then if this fails barbiturate induced burst suppression is implemented. Eleven patients who had attained 6 months of follow up were managed with barbiturate induced burst suppression. The Glasgow Outcome Scale was used to rate neurological outcome and mortality was compared to predicted APACHE II mortality, and previous studies utilising nimodipine to treat patients with ischaemic deficit secondary to vasospasm.
One hundred and sixty-four consecutive patients treated according to protocol were reviewed. Eleven patients with symptomatic vasospasm (varying from mild cognitive to severe motor deficits and decreased conscious level) received barbiturate coma therapy where chemical angioplasty had failed on a number of occasions.Thiopentone 10 mg/kg was infused intravenously over 20 min and barbiturate coma was maintained with an infusion rate adjusted to maintain a burst suppression EEG pattern. Intended duration of barbiturate therapy was 3 days but three patients required earlier discontinuation because of infectious complications and two patients received longer infusions because of persistent angiography proven vasospasm. Following termination of barbiturate infusion median times for return of pupillary light reflex and motor response to painful stimuli were 18.5 and 80 h, respectively. Predicted mortality from APACHE II was 31% but all 11 patients survived to hospital discharge. Literature search revealed survival rates of 19% from previous reports of patients receiving barbiturate therapy for vasospasm - none of these survivors made a good neurological recovery. Ten of 11 patients (91%) had a good neurological outcome when rated by the Glasgow Outcome Scale at 6 months following discharge. This compared to 51% for historical controls in patients receiving nimodipine for delayed ischaemic deficit from vasospasm.
Calcium antagonists may improve prognosis for vasospasm following SAH, but evidence is limited for triple H therapy and angioplasty has yet to be validated in controlled trials. This study suggests that the small number of patients with vasospasm despite angioplasty may benefit from barbiturate coma. Protective effects are thought to be due to a reduction in cerebral oxygen consumption and intracranial pressure, and via free radical and fatty acid metabolism. However immune and cardiovascular depression and lengthy artificial ventilation and intensive care therapy are major drawbacks of this therapy. Previous studies utilising barbiturate coma therapy have shown dismal results and the authors question whether improved intensive care therapy, earlier implementation of therapy or chance explain their very encouraging results which are comparable with present therapies for vasospasm following SAH.