SELECTIVE CLOSURE OF THE VASCULAR BED OF AN EXPERIMENTAL GLIOMA MODELDURING IN-SITU SALINE PERFUSION

Flow through the vasculature of an experimental rat glioma has been in vestigated during in situ perfusion of the brain, via the ascending ao rta, with a simple saline solution. Using such a system, it has been s hown previously that the blood-brain barrier will remain intact with a n adequate cerebral perfusate flow rate for at least 10 min, providing that the metabolic inhibitor 2,4-dinitrophenol is present. Cerebral p erfusate flow rate was measured in both tumour and non-tumour areas us ing [C-14] iodoantipyrine. The perfusion pump rate was set between 4.8 and 84 ml/min in different animals and the mean flow rate for cerebra l hemisphere remote from the tumour was 1.03 +/- 0.67 ml/min/g (mean /- sd; n = 17) whereas that for intracerebral tumour was considerably lower at 0.060 +/- 0.11 ml/min/g (mean +/- sd; n = 17). Linear regress ion of tumour now on cerebrum now showed a highly significant correlat ion (r = 0.88). Light and electron microscope examination of the tumou r vessels revealed no luminal obstruction or perivascular oedema to ex plain the low flow. We suggest that perfusion with a low viscosity med ium, at now rates that result in a low intraluminal pressure, probably causes glioma vessels to close preferentially because they require a higher intraluminal pressure to remain patent than do normal cerebral vessels.