Quantitative analysis of injured, necrotic, and apoptotic cells in a new experimental model of intracerebral hemorrhage

Objective: To develop a new survival model of intracerebral hemorrhage (ICH) in rabbits and study the patterns of cellular injury in different regions 24 hrs after introduction of hematoma. Quantitation and characterization of injured cells in regions adjacent and distant to the hematoma have not been performed. Design: Prospective case-control study. Subjects: Ten New Zealand rabbits. Intervention: We introduced ICH in six anesthetized New Zealand rabbits by autologous blood injection under arterial pressure in the deep white matter in the frontal lobe. Measurements and Main Results: Hematoxylin and eosin staining was performed in six animals with ICH after 24 hrs to quantify intact, injured, and necrotic cells in regions proximal and distant to the hematoma, and the results were compared with four control animals. Terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) staining was performed to quantify apoptotic cells in specified regions in five animals with ICH, and the results were compared with four control animals, All cell counts were performed by one investigator who used 100x oil emersion microscopy. The presence of localized hematoma was confirmed in all six animals with blood infusion. Compared with controls, animals with ICH had a significantly higher proportion of swollen cells in both the inner (55.9% +/- 3.0% vs. 26.8% +/- 1.7%; p <.05) and the outer (59.8% +/- 4.6% vs. 27.7% +/- 4.5%; p <.05) rim of the perihematoma region. A small proportion of shrunken dark cells were observed in both the inner (4.0% +/- 1.5%) and the outer (3.6% +/- 1.0%) rim of the perihematoma region, The remaining cells were considered morphologically intact, A large proportion of cells trapped within the matrix of the hematoma were either shrunken dark cells (48.8% +/- 16.4%) or swollen (38.8% +/- 15.1%), In the TUNEL-stained sections, a high burden of apoptotic cells was observed in the matrix of the hematoma (17.5 +/- 6.3 cells per high power field) but not in the perihematoma regions (less than two cells per high power field). Conclusions: A reproducible model of ICH in rabbits is described. At 24 hrs, the perihematoma region contains relatively large proportions of morphologically intact or reversibly injured (swollen) cells, suggesting the possibility of an extended window for therapeutic intervention.