Mr S. Mojtaba Golzan, Stuart L. Graham, Miok Kim, Alberto Avolio
To compare non-invasive estimates of cere-brospinal fluid pressure (CSFp) using spontaneous retinal venous pulsations (SRVP) and blood flow veloc-ity measurements (Doppler) of the middle cerebral artery (MCA).
10 normal subjects (37 ± 12 yrs) were included. Baseline IOP and SRVP were recorded using Dynamic Vessel Analyzer (DVA, Imedos). IOP was lowered using Iopidine 0.5% and measured every 15 mins followed by a 20sec SRVP recording. Pulse amplitude of SRVP was measured and plotted against different IOP levels. Linear regression determined the IOP at which SRVP cessation would occur, and the intercept estimated CSFp. Blood flow velocity in MCA and arterial blood pressure (ABP) measurements were performed simultaneously on the same subjects to extract CSFp (ICM+, Cambridge) and to compare with estimated values from SRVP approach. Central blood pressure waveform was derived from ABP waveforms using applanation tonometry. The cardiac component was removed to extract CSFp waveform. Finally esti-mated CSFp value added to create a continuous CSFp waveform.
CSFp values ranged from 3.2 ± 1.2 to 12.9 ± 2.3 mmHg for SRVP estimates and from 2.4 ± 1.0 to 13.9 ± 0.8 mmHg using ICM+. A high correlation between CSFp measurements using the two different methods was observed (r2 = 0.62, p < 0.05). Average CSFp estimations by the two methods showed no significant difference (student t-test, p > 0.05). Waveforms demonstrated CSF pulse pressures of 3 ± 1 mmHg.
CSFp can be estimated non-invasively by our technique using SRVP amplitudes, and corre-lates with estimates made using Doppler. A non- invasive analysis of ABP waveform can also provide details of the CSFp waveform and pulse amplitude.