HONOLULU -- Recent cocaine use raised the risk of death nearly threefold in a cohort of more than 1,000 patients who were hospitalized for an acute aneurysmal subarachnoid hemorrhage.
The significantly elevated risk of death was observed despite a lack of difference in severity of initial presentation and remained after exclusion of deaths due to rebleeding, which was higher in cocaine users, Dr. Neeraj Naval said at the International Stroke Conference. The study will also be presented at the annual meeting of the American Academy of Neurology in San Diego on March 19.
Patrice Wendling/IMNG Medical Media
Dr. Neeraj Naval
"Patients with acute subarachnoid hemorrhage following cocaine use warrant very close monitoring," said Dr. Naval, director of Neurosciences Critical Care at Johns Hopkins Bayview Medical Center in Baltimore.
Although cocaine use and aneurysmal subarachnoid hemorrhage (SAH) is not controversial, data have been scattered on how cocaine affects presentation and outcomes, he observed. In the largest series prior to this, cocaine use had no significant effect on symptomatic vasospasm or neurologic outcome among 600 patients with SAH (World Neurosurg. 2010;73:357-60). An earlier study, however, showed a 2.8-fold higher risk of vasospasm and 3.3-fold higher risk of poor outcome among cocaine users with SAH (J. Neurosurg. 2003;99:271-5).
Dr. Naval and his colleagues reviewed 1,134 patients admitted to one of two Johns Hopkins University hospitals for ruptured brain aneurysm between 1991 and 2009. The cohort included 142 patients (12.5%) who had a history of cocaine use in the 72 hours prior to admission based on self-report or urine toxicology, and 992 with no cocaine use. Cocaine users were more likely to be younger (49 years vs. 53 years), but had similar rates of poor grade 4/5 Hunt & Hess scores (21% vs. 26%) and associated intraventricular hemorrhage (IVH, 56% vs. 51%). Their mean Glasgow Coma Scale scores at admission were also similar (15 in users vs. 14 in nonusers).
In all, 26% of cocaine users and 17% of nonusers died in the hospital (P = .01).
Significant independent predictors of in-hospital death were cocaine use (adjusted odds ratio 2.85), admission Hunt & Hess score (OR 2.33) and higher age (OR 1.03; all P less than .001), Dr. Naval said.
Cocaine users had higher rates of aneurysm re-rupture (7.7% vs. 2.7%; P = .004). Unfortunately, admission mean arterial pressure (MAP) data were unreliable from 1991 to 2005, but data available from 2006 to 2009 showed higher MAP in cocaine users, he said.
Cocaine users were more likely to have delayed cerebral ischemia (22% vs. 16%; P = .041), but the association was not statistically significant after correction for other confounding factors, including age and IVH. Delayed cerebral ischemia was defined as new clinical deterioration more than 48 hours post-SAH and more than 24 hours after surgical clipping or endovascular coiling, radiologic confirmation of cerebral infarction, and/or angiographic confirmation of vasospasm and/or clinical responsiveness (transient or sustained) to hemodynamic augmentation.
Dr. Naval suggested that it is controversial to include IVH in the model because data are available suggesting that cocaine use is independently associated with a higher rate of IVH in patients with intracerebral hemorrhage.
"If there really is a cause-effect relationship between cocaine use and IVH, one wonders whether using IVH in the multivariate analysis may mask the true impact of cocaine exposure on vasospasm-mediated cerebral infarction," he said.
The investigators did not demonstrate any difference between groups in functional outcomes at discharge or post discharge. Dr. Naval said this was not surprising given the significant difference in age between the groups, with younger patients much more likely to recover from neurological injury.
When asked during a discussion of the study to speculate on the mechanism of elevated mortality in cocaine users, Dr. Naval said subsequent analyses found no difference in rates of withdrawal of care between groups and no impact with frequency of cocaine use or delayed cerebral ischemia. Regional wall motion abnormalities have been identified in cocaine users, but electrocardiograms were not performed to tease out the impact of myocardial stunning on mortality.
In terms of its implications for management, a case could be made to use antifibrinolytic therapy in patients with SAH who are cocaine users because of the higher risk of rebleeding, but Dr. Naval said that a trial would be needed to evaluate this.
The conference was sponsored by the American Heart Association. Dr. Naval reported honoraria from EKR Therapeutics. His coauthors made no disclosures.