Background and Purpose: Hematoma expansion of intracerebral hemorrhages (ICH) is associated with poor patient outcomes. However, the true mechanism remains unclear. Previous studies have investigated the role of admission leukocytosis and neutrophilia in predicting poor outcomes in patients admitted with ICH. In these studies, deep and lobar ICH, which have different pathophysiologies, were not distinguished. In our study, we aim to investigate the value of admission neutrophilia (AN) in predicting hematoma expansion in patients with deep ICH and lobar ICH.
Methods: We performed a retrospective review of a prospective database of patients admitted with ICH between November 1, 2012 and March 31, 2016. We included patients with ICH confirmed by CT imaging admitted for a minimum of 24 hours. Patients were excluded if etiologies included trauma, primary or metastatic brain tumors, evidence of hemorrhagic conversion from ischemic stroke, and vascular malformations. In addition to demographic and clinical data, we collected data on primary ICH location and volume, volume after 24 hours, and admission leukocyte and neutrophil counts. We defined AN as a neutrophil count >7.7 K/μL. Hematoma expansion was defined as an increase of >30% on the 24 hour repeat imaging. Basal ganglia and thalamic ICH were classified as deep ICH.
Results: Of the 573 patients admitted with ICH, 432 who met inclusion criteria were included for our analysis. 153 (41.3%) had deep ICH and 217 (58.6%) had lobar ICH. The presence of AN was significantly different between deep vs lobar ICH (47.7% vs 27.6% p=0.0001). Among these patients, larger initial hematoma volume was found in lobar ICH as compared with deep ICH (67.2ml (SD 51.0) vs 44.6ml (SD 50.5) p=0.003). However, the frequency of hematoma expansion was not significantly different between groups (12.3% vs 15.2% p=0.87). On multivariate analysis, AN was not an independent predictor of hematoma expansion in deep, lobar or both groups combined.
Conclusion: Contrary to previous reports, AN is not an independent predictor of hematoma expansion. This suggests that the pathophysiological process leading to hematoma expansion is likely multifactorial.