(T1430-01-04) Analysis and Simulation of In Vivo Nanoparticle Protein Carrier Pharmacokinetics

Purpose: Poly(lactic-co-glycolic acid) (PLGA) remains one of the most studied polymers for controlled release and nanoparticle drug delivery, and numerous studies exist proposing PLGA particles as protein carriers. In this work we report using PLGA nanoparticles modified with biotin (bNPs) to deliver proteins via adsorption in vivo and utilize live tracking to establish and evaluate a pharmacokinetic model of particle and protein trafficking to the lymphatic system.

Methods: Biotin was incorporated into PLGA microparticles using DSPE-PEG-biotin at a ratio of 10:1 PLGA: DSPE-PEG-biotin. bNPs were manufactured using solvent evaporation emulsification with dichloromethane as the organic phase containing PLGA and DSPE-PEG-biotin and 1% polyvinyl alcohol in water as the aqueous phase. bNPs containing a near infrared dye (NIR815) coated with near infrared streptavidin (SA) were injected subcutaneously into C57 mouse footpads and imaged using a Licor Pearl Imager for up to 21 days. In an additional study, mice were injected as above and sacrificed to track the accumulation of bNPs and avidin within the popliteal draining lymph node (dLN). Background subtraction was performed automatic detection was used to quantify fluorescent intensity at 5 standard deviations above background. MATLAB Simbiology module was utilized to simulate the elimination of SA alone (SA [free]), SA complexed to bNPs (SA-bNP) and bNP alone (bNP) from the SC injections. A two-compartment model was built assuming clearance from the footpad either into the dLN or into an undefined central compartment. The dissociation of SA from biotin according to its binding constant (K_D= 10^-14) was included in the model. In vivo footpad bNP and SA concentrations were fitted into the first order decay model to estimate kinetic parameters (FP k_el, dLN k_el, dLN k_abs) for SA [free], bNP and SA-bNP. Evaluation of the model was performed by running simulations with the final fitted parameters by inputting specific doses, either from the observed in vivo data or varying for the purpose of dose ranging and sensitivity analysis was performed.

Results: A two-compartment model was established for the observed in vivo of bNP and streptavidin in the footpad and dLN (Fig. 1). The model provided a good fit for the observed data for each group in both compartments (Fig. 2A-F) and pharmacokinetic parameters were calculated (Table 1). Observed vs predicted plots for the footpad compartment highlighted the fit with average R² values of 0.90, 0.89, 0.83 and 0.94 for SA in SA-bNP, bNP in SA-bNP, bNP [free] and SA [free], respectively (Fig. 3A-D). For SA in SA-bNP and bNP in SA-bNP in the dLN, the observed vs predicted data provided a good overall fit for SA in SA-bNP and bNP in SA-bNP with R² values of 0.94 and 0.91, respectively (Fig. 3E-F). Dosing simulations provided evidence that the exposure of SA in the lymph node would be enhanced through complexation with bNPs when the SA [free] dose (dashed) was increased to 8x above the SA-bNP dose (solid) (Fig. 4A). Sensitivity analysis of the model for SA-bNP showed that SA complexed to bNPs would have elimination rates in the footpad and dLN dictated by the rate constants of the bNPs, while SA dissociated from the complex would be sensitive the elimination rates for both SA [free], SA-bNP and bNP (Fig. 4B). In simulations reducing the binding constant between SA and bNPs the exposure, as area under the curve, reduced drastically when the K_D is reduced below 10^-9. This reduction in exposure is reduced when the ratio of SA: bNP dosed is increased from 1:1 to 10:1 (Fig. 4C).

Conclusion: These data provide evidence that bNPs can enhance exposure of a model protein in the lymph node relative to protein in solution. In addition, the working 2-compartment model can be utilized for evaluating formulation changes and dosing simulations for the bNP system.

Figure 1: In vivo tracking images for SA (700nm) and bNP (800nm) in SA-bNP. 2-compartment model for trafficking of streptavidin and bNPs from footpad to lymph nodes.
Figure 2: Observed vs fit for in vivo tracking data analyzed using the 2-compartment model.

Final pharmacokinetic parameters determined for in tracking data.

Figure 3: Observed vs predicted plots for SA [free], SA-bNP and bNP.
Figure 4: A, Dose ranging of SA [free] (dashed) vs SA-bNP (solid). B, Sensitivity analysis of two-compartment model for SA-bNP and SA dissociated from bNP. C, Dissociation constant vs AUC for different ratios of SA:bNP.