Predictors of clinical failure of decompressive hemicraniectomy for malignant hemispheric infarction

Object The aim of this study is to identify pre-operative clinical and/or radiological predictors of clinical failure of decompressive hemicraniectomy (DH) in the setting of malignant hemispheric infarction. These predictors could guide the decision for adjunctive internal brain decompression (e.g. strokectomy) at the time of the initial DH. Methods Retrospective chart review of all patients with malignant hemispheric infarction who underwent DH at our institution, from November 2008 to January 2013. Demographics, pre- and post-operative clinical characteristics and neuroimaging data were reviewed. The surgical outcome after DH was evaluated and clinical failure was defined as follows: lack of post-operative resolution of basal cistern effacement, and/or failure to achieve a post-operative decrease in midline shift by at least 50%, and/or post-operative neurological deterioration felt to be due to persistent mass effect, with or without a second, salvage operation (strokectomy). Results Out of 26 patients included in the study, 7 were considered to have clinical failure of their DH. Preoperative clinical and imaging variables were similar in the two groups, except that the presence of a nonreactive pupil immediately before surgery was associated clinical failure of the DH (p = 0.0015). Patients in the clinical failure group had a lower postoperative GCS motor score and a strong but not statistically significant trend towards less favorable functional outcome (GOS 1-3). Conclusions The presence of a nonreactive pupil before surgery is associated with clinical failure of DH, and should be taken into account when deciding whether to perform strokectomy at the time of DH.