Time course of cortical hemorrhages after closed traumatic brain injury: Statistical analysis of posttraumatic histomorphological alterations

We examined 305 autopsied brains for histomorphological alterations to determine the time course of reactions in cortical hemorrhages following traumatic closed brain injury. Eighteen morphological criteria were considered: red blood cells (RBCs), polymorphonuclear leukocytes (PMNs), macrophages (Ms), RBC-containing Ms, hemosiderin, hematoidin, lipid-containing Ms, fibroblasts, endothelial cells, collagenous fibres, gemistocytic astrocytes, fibrillary gliosis, hemosiderin-containing astrocytes, neuronal damage, neuronophagy, axonal swelling (beta-amyloid precursor protein: beta-APP), axonal bulbs (van Gieson stairi), and mineralisation of neurons. The interval between the time of brain injury and death ranged from 1 min to 58 years. Following routine staining and immunohistochemical staining of microglia (CD68), astrocytes (GFAP) and injured axons (beta-APP), paraffin sections were examined by light microscopy for the presence of the selected histomorphological features. For each cytomorphological phenomenon, the time at which it could be demonstrated for the first time and for the last time (observation period) was determined. The relative frequency of each criterion was established for each observation period. The limits of confidence for the respective relative frequencies were estimated with a reliability of 95% according to Clopper and Pearson. An apparent correlation was found between the frequency of a given histomorphological phenomenon and the length of the posttraumatic interval. To check for accuracy of prediction, half of the cases (group 1; n = 153) were used to develop a multistage evaluation model; half (group 2; n = 152) were used to evaluate the validity of the data of group 1. Applying this model, 117 of the 152 control group cases (76.97%) could be correctly classified and further 26 cases (17.11%) being assigned to an interval close to the correct interval. Thus, this model allows classification of the correct posttraumatic interval or an interval close to the correct posttraumatic interval in about 95% of cases. We developed a software program that allows the estimation of survival time of TBI based on the relative frequency of the 18 morphological features. Applying this software will help to estimate the posttraumatic interval of cortical hemorrhages following TBI of unknown survival time (till.walter@t-online.de).