1986 Volume 26 Issue 2 Pages 153-159
It is important for adequate treatment after subarachnoid hemorrhage (SAH) to estimate objectively the disturbed cerebrospinal fluid (CSF) circulation and the intracranial spatial compensatory capacity. This paper reports the results of a preliminary trial to apply infusion studies as a method of quantifying the amount of CSF circulatory disturbance and the elasticity of the brain. The infusion studies (bolus injection, bolus withdrawal and steady-state infusion) were performed in 14 adult SAH patients for a total of 19 times. In 12 patients, studies were carried out under two spinal taps (one for CSF manipulation and the other for a pressure transducer). Two other patients received the tests through ventricular drainage. Pressure changes were recorded by a polygraph and a data recorder, and were later analysed by the methods of Marmarou, et al. and Katzman, et al. to calculate three parameters, namely, pressure-volume index (PVI), CSF outflow resistance (Ro) and CSF formation rate. These parameters were compared with computed tomography (CT) findings, radioisotope cisternographic findings (6 cases) and the effectiveness of therapeutic procedures (external decompression in 4 cases and CSF shunt in 9 cases). PVI values showed statistically significant differences between the decompressed group (44.62±30.33 ml) and nondecompressed group (22.23±9.02 ml) (p<0.001, F test). Ro values recorded their peak from 10 to 20 days after SAH, and tended to be higher among the patients who showed a reflux-stasis pattern in cisternography and in whom the shunt procedure was effective. Patients who developed periventricular lucency in the CT scan tended to show higher Ro values. However, there was one case which did not develop hydrocephalic symptoms despite a high Ro value. In some cases Ro remained low despite the fact that other examinations suggested advanced disturbance of CSF absorption. These discrepancies were thought mainly to have occurred because Marmarou's theory did not consider interactions between another two intracranial components (the vascular bed and the brain parenchyme) on volume loading or withdrawal into CSF space. If this problem were solved, the bolus infusion test could be a much more useful and practical method than steady-state infusion.
