Retour à la page précédente

Effect of hemodilution on sub-foveal choroidal blood flow in the human eye with retinal vein occlusion and the contralateral healthy eye


The objective of this prospective study was to evaluate the effect of hemodilution on sub-foveal choroidal blood flow (ChBF) in the human eye with retinal vein occlusion and the contralateral healthy eye.

Isovolemic hemodilution has been proposed to patients with branch (BRVO) or central retinal vein occlusions (CRVO). Some clinical studies have suggested a positive effect of hemodilution on visual recovery. The rationale behind this treatment is to increase blood flow through the veins and capillaries by reducing hematocrit and thus the blood viscosity and red blood cell (RBC) aggregation.

However very little is known about changes of ocular blood flow after hemodilution. Furthermore, the reduction of hematocrit obtained with hemodilution may be an interesting model for the study of the relationship between reduction in red blood cells (RBC) number and blood flow parameters assessed by laser Doppler flowmetry.

The LDF technique is based on the Doppler effect where light scattered from moving RBCs is shifted in frequency by an amount proportional to the velocity of the particles. This scattered light is optically mixed with the light reflected by the tissues and electronically detected by an avalanche photodiode (APD).

The subject is asked to look at a faint point A, so that this light is focussed on his foveola. The light scattered by the RBCs in the choriocapillaris is collected by the optical system and detected by an APD through a fibre bundle (fext = 480 µm, fint = 160 µm) excluding the centre. The signal is amplified, digitalized and analysed by a computer using a PCMCIA card controlled by a LabVIEWprogram.

The following blood flow parameters are derived: volume of RBCs (ChBVol), velocity (ChBVel) and choroidal blood flow (ChBF=ChBVol x ChBVel), which is expected to be proportional to the flow of RBCs. Measurements of one session are validated if the amount of backscattered light (DC) remains stable (±10%).


From these preliminary results, a significant reduction of hematocrit did not affect sub-foveal choroidal blood volume (ChBVol), which is surprising because this flow parameter is a function of the number of moving scatters (i.e. RBCs). Affected and contralateral eyes exhibit a significant opposite behaviour after an hematocrit reduction. Contralateral eyes, considered as healthy, show an increase of ChBF in accordance with literature. The fact that this increase is much smaller and that the affected eye exhibits an opposite behaviour suggests a probable modification of the scattering tissues. A significant decrease of hematocrit induced an opposite change of ChBF and ChBVel between affected and contralateral eye, but no change of ChBVol.

These preliminary results must be confirmed with a larger series of patients. Furthermore, changes of choroidal blood flow parameters may not reflect changes in the retinal circulation.