Abstract

Ventriculocisternal perfusion is regarded as a precise method of measuring the rate of formation of cerebrospinal fluid (CSF) but it possesses inherent potential sources of error. Using the technique to measure CSF formation rate in the rhesus monkey, we have observed rate changes when none were expected. Most puzzling has been the steady decline of CSF formation rate at 4 percent each hour during the final five hours of a seven hour perfusion although variables known to affect CSF formation remained stable. In addition, alterations in rate caused by artefacts were observed in experiments in which craniospinal blood volume was changed by sudden changes of either PCO2 or central venous pressure. Mobilisation or sequestration of incompletely equilibrated CSF is believed responsible. In other experiments, a small increase of intracranial pressure produced by increasing outflow resistance was quickly followed by an apparent reduction of CSF formation. We have concluded that to assess accurately the effect a variable has on the rate of CSF formation, one must control perfusion time and craniospinal blood volume as well as intracranial pressure.