Lumbar Drain vs Extraventricular Drain to Prevent Vasospasm in Subarachnoid Hemorrhage

Overview

Vasospasm is a common complication after rupture of intracranial aneurysms causing devastating neurologic deficits and death. Vasospasm has been directly associated with the amount of subarachnoid blood inside the basal cisterns. Prior literature has attempted to refine treatment of ruptured intracranial aneurysms but does not have clear guidelines on the optimal method to drain subarachnoid blood. Two methods, extraventricular drain (EVD) and lumbar drain (LD) have been compared retrospectively yet remain controversial as to which method is optimal in reducing subarachnoid blood and preventing vasospasm. This study would be a prospective randomized trial in which patients would be assigned to EVD or LD and observed to see if one method of intervention is associated with preventing clinical vasospasm, decreasing subarachnoid blood, shortening overall ICU stay, and reducing the need for a permanent ventriculoperitoneal shunt. The conclusions of this study may identify an optimal treatment modality to benefit all future patients with ruptured intracranial aneurysms.

This study will be a prospective randomized controlled trial. Participants will be patients admitted at UCSD Hillcrest or Thornton Hospital for subarachnoid hemorrhage. Patients will be recruited by the Endovascular Neurosurgical treatment team. Since there are no existing guidelines existing for procedural drainage of subarachnoid blood from the intracranial cisterns, patients are currently treated with either lumbar drain or extraventricular drain. This study would randomize the process of assigning patients to either LD or EVD as part of their treatment package for subarachnoid hemorrhage. Both EVD and LD are regularly performed at UCSD. Both procedures are equally likely to take place if patients do not have randomization. There is currently no gold standard that dictates whether EVD should be done over LD and vice versa. Patients will be consented to partake in the study while they are also consented for their usual care in SAH. Briefly, inclusion criteria are patients with Fisher Grade II, III, IV SAH or any aneurysmal SAH patients with radiographic evidence of hydrocephalus. Briefly, exclusion criteria are patients with obstructive hydrocephalus or mass lesion that would preclude lumbar drain placement. Pregnant patients or minors of age \<18 will not be included in this study. There will be two arms of the study: patients randomly assigned to LD and patients randomly assigned to EVD. Randomization will account for Fisher Grade. The two treatment arms will have CSF diversion for 7 days. For patients treated with lumbar drain, patients will have CSF drainage of 10cc/hour for lumbar drain while patients treated with EVD at 15cm above the tragus. Both an extraventricular drain and a lumbar drain are defined as significant risk devices however both devices are FDA approved and will be utilized in the exact manner that they are intended. Both devices are already regularly utilized and a part of the standard of care for sub arachnoid hemorrhage management and will not be utilized in any novel way. LD and EVD assignment will be randomized through the protocol of the study. Data will be collected via electronic medical record and will assess patient age, patient presenting Hess grade and clinical improvement post cerebrospinal fluid diversion, aneurysmal bleed day on presentation, method and technical success of securing aneurysm (clip versus coil), size and location of aneurysm, discharge disposition and clinical exam (NIH Stroke Scale), complications (tract hemorrhage or infection), need for intra-arterial management of vasospasm. Expected Sample size in this study will include approximately 200 patients in which half of patients included in this study will be treated with LD and the other half treated with EVD. The number of estimated patients is calculated from the annual incidence of SAH received at UCSD hospitals. The facilities that are available for this project include UCSD Hillcrest Hospital, UCSD Thornton Hospital, and the UCSD Endovascular Neurosurgery suite. The primary outcome of this study will be time to clearance of acute cisternal blood by surveillance CT (stratified by presenting Fischer Grade), need for endovascular treatment of vasospasm or stroke from vasospasm, time spent in the critical care unit, and need for ventriculoperitoneal shunt. Regarding CT scans; Surveillance CTs are standard of care. They are typically performed on the day of device placement and subsequent imaging takes place pending clinical resolution of the patient. For most patients, they have an additional CT scan 1-2 days after initial device placement and a final scan on the 7th day of device placement. If clinical condition deteriorates, patients will have an additional scan to monitor for intracranial bleeding. The scans are performed for the clinical purpose of monitoring resolution of intracranial injury. This study would collect the data obtained from the scans. The total exposure resulting from these imaging studies is calculated to be approximately 10 mSv. Additional head CT may be performed if clinically indicated with exposure 2.5 mSv per scan. Fluoroscopy is not routinely used, however if clinically indicated, may be utilized during lumbar drainage procedure with radiation exposure 1.5 mSv per minute. This amount is more than one would receive from one year of natural exposure in the San Diego area, which is approximately 1.6 mSv. Cumulative exposure from radiation may increase a person's risk of developing certain types of cancer in the future. This imaging is determined to be part of routine care and management of SAH. In this study, we will be merely collecting data from these images to determine clinical incidence of vasospasm and resolution of cisternal bleeding. DSM criteria were derived from Al Tamimi et al. The initial power calculation was based off of the Al Tamimi et at study 9 that demonstrated that for 85% power, 105 patients were required in each arm of the trial. Interim analysis will be performed after recruiting 40 patients in order to establish adverse effects and an additional reevaluation at 100 patients to determine if one arm of the study is superior to the other. If this is identified, the trial will be stopped at recruitment of 100 patients. Statistical analysis will continue throughout the trial. If statistically significant differences between both arms of the study are identified earlier in the enrollment process, the study will be discontinued. Monitoring will take place internally in which the differences between the two groups will be examined for each additional 5 patients added to both groups. Internal monitoring will take place by (D.R.S.D., A.W., R.R., A.K., S.P.) Statistical methods to evaluate data include chi square test, using independent t tests for normal numerical data, and comparing categorical data with Fisher exact test. P value of .05 will be considered clinically significant in this study.