Therapy of malignant intracranial hypertension by controlled lumbar cerebrospinal fluid drainage

Objectives: To evaluate the effect of controlled lumbar cerebrospinal fluid drainage in adult patients with refractory intracranial hypertension. Design: Prospective, pre- vs, postintervention study. Setting: Surgical intensive care unit of a university hospital. Patients: Twenty-three patients with severe traumatic brain injury or delay ed ischemia after subarachnoid hemorrhage with intracranial hypertension re fractory to aggressive treatment, including repeated applications of tromet hamine, hypertonic saline solution, barbiturate coma, and decompressive cra niectomy, Patients were considered for controlled lumbar cerebrospinal flui d drainage if basal cisterns on computerized tomography scan were discernib le. Interventions: After institution of a lumbar drain, cerebrospinal fluid was gradually aspirated, and then, continuous cerebrospinal fluid drainage was maintained under control of intracranial pressure (ICP) and pupillary stat us. Measurements and Main Results: ICP and cerebral perfusion pressure before a nd after initiation of lumbar cerebrospinal fluid drainage and related comp lications were documented. The neurologic outcome of the patients was asses sed according to the Glasgow Outcome Scale 6 months after injury. As a resu lt of lumbar cerebrospinal fluid drainage, all patients demonstrated an imm ediate and lasting decrease of ICP and a concomitant increase of cerebral p erfusion pressure. Two patients temporarily showed a unilateral fixed and d ilated pupil 6 and 8 hrs after onset of lumbar cerebrospinal fluid drainage , respectively. Ten patients showed a favorable outcome, four patients surv ived with a severe permanent neurologic deficit, one patient remained in a persistent vegetative state, and eight patients died. Conclusions: Controlled lumbar cerebrospinal fluid drainage significantly r educes refractory intracranial hypertension. The danger of transtentorial o r tonsillar herniation is minimized by considering lumbar drainage in the p resence of discernible basilar cisterns only.