Purpose. Previous experimental work suggests that convection may be importa
nt in determining the biodistribution of drugs implanted or injected in the
vitreous humor. To develop accurate biodistribution models, the relative i
mportance of diffusion and convection in intravitreal transport must be ass
essed. This requires knowledge of both the diffusivity of candidate drugs a
nd the hydraulic conductivity of the vitreous humor.
Methods. Hydraulic conductivity of cadaveric bovine vitreous humor was meas
ured by confined compression tests at constant loads of 0.15 and 0.2 N and
analyzed numerically using a two-phase model. Diffusion coefficient of acid
orange 8, a model compound, in the same medium was measured in a custom-bu
ilt diffusion cell.
Results. Acid orange 8 diffusivity within vitreous humor is about half that
in free solution. When viscous effects are properly accounted for, the hyd
raulic conductivity of bovine vitreous humor is 8.4 +/- 4.5 x 10(-7) cm(2)/
Pa . s.
Conclusions. We predict that convection does not contribute significantly t
o transport in the mouse eye, particularly for low-molecular-weight compoun
ds. For delivery to larger animals, such as humans we conclude that convect
ion accounts for roughly 30% of the total intravitreal drug transport. This
effect should be magnified for higher-molecular-weight compounds, which di
ffuse more slowly, and in glaucoma, which involves higher intraocular press
ure and thus potentially faster convective flow. Thus, caution should be ex
ercised in the extrapolation of small-animal-model biodistribution data to
human scale.