Neuropathological and neurophysiological effects of interstitial white matter autologous and non-autologous protein containing solutions: further evidence for a glioma derived permeability factor.

The feline infusion model of brain edema was used to evaluate the pathophysiological effects of 0.6 ml infusions of autologous serum protein (66%), human serum protein (66%), human glioma cyst fluid and a tissue culture medium (TCM) on the structure and function of the forebrain white matter. These infusions increased local white matter water content by between 10.8 and 12.5 ml/100 g brain and were associated with moderate increases in ICP and CSF outflow resistance and a significant decrease in lumped craniospinal compliance. Cortical somatosensory potentials, motor evoked potentials, EEG and local cerebral blood flow (rCBF) at normocapnia were generally unchanged by the various infusions. All infusates except the 66% autologous serum protein infusion impaired rCBF CO2 reactivity. Histologically all infusates caused marked extracellular edema. The autologous serum protein infusion caused no additional histological changes whereas the glioma cyst infusates caused profound endothelial and astrocytic swelling, focal endothelial necrosis, basement membrane disruption, perivascular microglial reaction and pavementation and perivascular migration of polymorphonuclear leukocytes. Similar but less marked changes were seen after infusion of human serum protein whilst the TCM produced only minimal changes. The intensity and extent of Evans Blue extravasation into the forebrain white matter was greatest with glioma cyst infusates and with all infusions reflected the extent to microvascular changes. These studies show that products derived from gliomas cause additional damage to the blood-brain-barrier than that caused by non-autologous serum proteins. These results add further support for the existence of glioma derived permeability factors (GDPF), but suggest neither serum proteins nor glioma derived compounds in the white matter interstitium significantly influence local electrophysiological function. Some limitations of the infusion edema model when using non-autologous infusions and difficulties quantitating brain dysfunction are emphasised.