Clinical ReportsDecompressive Craniectomy For Intractable Cerebral Edema: Experience Of a Single CenterZiai, Wendy C.*†‡; Port, John D.§; Cowan, John A.†; Garonzik, Ira M.†; Bhardwaj, Anish*†‡; Rigamonti, Daniele†Author Information From the *Division of Neurosciences Critical Care; and the †Departments of Neurosurgery, ‡Neurology, and §Neuroradiology, Johns Hopkins University School of Medicine, Baltimore, Maryland Address correspondence and reprint requests to Dr. Wendy C. Ziai, Johns Hopkins Hospital, Division of Neurosciences Critical Care, 600 N. Wolfe Street/Meyer 8–140, Baltimore, MD 21287. No financial support has been received in conjunction with the generation of this submission. Dr. Anish Bhardwaj is supported in part by the Established Investigator Grant from the American Heart Association. Accepted for publication on August 7, 2002. Journal of Neurosurgical Anesthesiology: January 2003 - Volume 15 - Issue 1 - p 25-32 Buy Abstract Several case reports and small clinical series have reported benefits of decompressive hemicraniectomy in patients with intractable cerebral edema and early clinical herniation. Specific indications and timing for this intervention remain unclear. We present our experience with this procedure in a subset of 18 patients with massive cerebral edema refractory to medical management, treated with decompressive craniectomy over a 3-year period (1997 to 2000). Computerized tomography (CT) scans were independently analyzed by a neuroradiologist blinded to clinical outcome. Eleven male and seven female patients, ages 20 to 69 years (mean ± SEM, 46 ± 14 years), underwent hemicraniectomy for the following diagnoses: 12 hemispheric infarcts, 3 traumatic intracerebral hemorrhages/contusions, 2 nontraumatic intraparenchymal hemorrhages (ICH), and 1 subdural empyema. This population included four patients with aneurysmal subarachnoid hemorrhage (SAH). Patients were followed for a mean of 10 months. Clinical factors including age, side of lesion, preoperative herniation signs, and early surgery (<12 or <24 hours) were not significantly associated with mortality or Glasgow outcome score (GOS). Preoperative CT evidence of transtentorial herniation (present in 5/17 patients) was associated with mortality (P = 0.04), while preoperative uncal herniation (8/17 patients) was associated with poor outcome (GOS > 1) (P = 0.01). Favorable outcome (GOS > 3) occurred in six patients, three with spontaneous or traumatic focal hematomas. Of four patients with SAH, one died while the others were severely disabled (GOS 3). Seven of nine patients with malignant MCA infarctions unrelated to SAH had poor outcomes. The overall mortality was 4/18 (22%). Patients with refractory cerebral swelling secondary to focal hematomas may have better outcomes following decompressive craniectomy. Patients with preexisting SAH seem to have poor outcomes, possibly related to other neurologic comorbidities. Hemicraniectomy requires definition of proper timing. Preoperative CT findings, especially transtentorial and uncal herniation may be useful in defining when decompressive surgery should not be performed. © 2003 Lippincott Williams & Wilkins, Inc.