At 25 min, the carotid 5-alpha-reductase sections were rinsed in PBS, were digested with Scintigest, and drug content was determined using LSC.18 Effective diffusivity through the artery was estimated from Eq. 2, where M is drug mass measured in tissue, t is time, A is the arterial area exposed to the drugs, C is the drug source concentration, and k is the arterial total binding capacity. This equation, in implicit form, is the analytical representation of 2D drug diffusion using a solely conductive approach. The model was developed to compliment preclinical studies aimed at optimizing drugeluting devices. Thus, arterial geometry was basedon a rat carotid artery, and stent coating dimensions were based on measurements of a 1.25 × 2.5 mm2 stent developed for rodent vessels.31 The laminate arterial structure was simplified by assuming only an intimal and medial layer. The intima was assumed to be denuded at the site of the coated stent strut, fgfr correlating to the area of greatest injury. As shown in Figure 1a, several boundary conditions were applied to this time dependent simulation.
The drug concentration was defined so that at the initial moment, the drug was present only in the polymer coating at the predefined value. A continuity of space wnt pathway condition was applied between the two polymer compartments, indicating that drug flux across that boundary is uninterrupted. Only drug release from the abluminal side of the stent was considered, as drug released from the luminal stent surface was assumed lost to the blood. A constant zero value drug concentration was assumed at the contact surface between the artery and the blood, significantly simplifying the model by excluding fluid flow mediated drug loss. A zero mass flux was imposed as an external boundary condition at the outer arterial wall, preventing drug from leaving the artery via transmural diffusion. A zero concentration boundary was also applied at the lateral ends of the arterial compartment to account for drug that leaves the high throughput screening stented arterial segment via planar diffusion through adjacent uninjured tissue. To prevent a false gradient induced by the lateral boundary condition, the arterial compartment was extended in the planar direction until convergence of the result was achieved. Prior studies have shown that measured drug release from polymer coatings is dependent on initial drug loading.
However, when using a numerical approach, the movement of drug through a single polymer layer will result in identical normalized release curves, independent of initial drug load. This single layer assumption has limited face validity because it wrongly leads to the conclusion that, for any initial drug loading, the same percentage of drug will be released. To address this shortfall, an alternative approach was taken in which the net drug load was distributed between two parallel polymer domains featuring fast and slow diffusion coefficients to describe the burst and sustained drug release, respectively. By creating two theoretical polymer layers, this approach can be applied to nearly all biphasic release profiles that feature an initial burst phase followed by a low level sustained release. This assumption allowed the evaluation of discrete drug distribution within the polymer matrix a phenomenon.