Does a negative correlation of heme oxygenase-1 with hematoma thickness in chronic subdural hematomas affect neovascularization and microvascular leakage? A retrospective study with preliminary validation.

Objective: Chronic subdural hematoma (CSDH) is a common neurological disease among elderly adults. The progression of CSDH is an angiogenic process, involving inflammatory mediators that affect vascular permeability, microvascular leakage, and hematoma thickness. The authors aimed to identify biomarkers associated with angiogenesis and vascular permeability that might influence midline shift and hematoma thickness.

Methods: Medical records and laboratory data of consecutive patients who underwent surgery for CSDH were analyzed. Collected data were basic demographic data, CSDH classification, CSDH thickness, midline shift, heme oxygenase-1 (HO-1) levels in hematomas, and common laboratory markers. Linear regression analysis was used to evaluate the relationship of CSDH thickness with characteristic variables. The chick chorioallantoic membrane (CAM) assay was used to test the angiogenic potency of identified variables in ex ovo culture of chick embryos.

Results: In total, 93 patients with CSDH (71.0% male) with a mean age of 71.0 years were included. The mean CSDH thickness and midline shift were 19.7 and 9.8 mm, respectively. The mean levels of HO-1, ferritin, total bilirubin, white blood cells, segmented neutrophils, lymphocytes, platelets, international normalized ratio, and partial thromboplastin time were 36 ng/mL, 14.8 μg/mL, 10.5 mg/dL, 10.3 × 103 cells/μL, 69%, 21.7%, 221.1 × 109 cells/μL, 1.0, and 27.8 seconds, respectively. Pearson correlation analysis revealed that CSDH thickness was positively correlated with midline shift distance (r = 0.218, p < 0.05) but negatively correlated with HO-1 concentration (r = -0.364, p < 0.01) and ferritin level (r = -0.222, p < 0.05). Multivariate linear regression analysis revealed that HO-1 was an independent predictor of CSDH thickness (β = -0.084, p = 0.006). The angiogenic potency of HO-1 in hematoma fluid was tested with the chick CAM assay; topical addition of CSDH fluid with low HO-1 levels promoted neovascularization and microvascular leakage. Addition of HO-1 in a rescue experiment inhibited CSDH fluid-mediated angiogenesis and microvascular leakage.

Conclusions: HO-1 is an independent risk factor in CSDH hematomas and is negatively correlated with CSDH thickness. HO-1 may play a role in the pathophysiology and development of CSDH, possibly by preventing neovascularization and reducing capillary fragility and hyperpermeability.