Purpose:
To visualize and predict the delivery of ocular drugs from an intravitreal injection to the retina using micro-computed tomography imaging and computational fluid dynamics (CFD). This allows for a better understanding of the mechanisms involved in the transport and fate of drugs in the vitreous.
Methods:
Clinically relevant mid-vitreal injections of 30 µl of a contrast agent (Omnipaque) acting as a drug mimic were delivered to our model system, cadaveric porcine eyes (n = 18). Three dimensional micro-computed tomography images were acquired in 8-16 seconds over a period of several hours post-injection. Closed form mathematical solutions for diffusive transport were used to calculate the diffusion coefficients used in the computational fluid dynamics model. Model parameters that could not be found through the acquired images were taken from literature.
Results:
Location was realized to be a very important factor in intravitreal injections and one that can be highly variable. The imaging allowed for precise determination of the injection location and contrast agent concentration, flow patterns and fate. By creating a diffusion based fluid dynamics model, it was realized that to most accurately represent the experimental injections other transport phenomena (e.g. gravity, retardation and boundary interaction) must also be implemented. The model provided quantitative agreement with the concentration values found in the experimental study (Figure).
Conclusions:
By making simple changes to the geometry of the model, properties of the materials, boundary conditions and domain conditions, the model can be translated into other systems, such as a human eye. The unique capability to visualize a drug mimic flow through the eye and use the data to derive a computer model provides a very accurate predictive tool to better understand and optimize ocular drug delivery systems.
Keywords: computational modeling • imaging/image analysis: non-clinical • injection