AIMS
Slowing the bioresorption of polymeric stent coatings can reduce tissue exposure to potentially proinflammatory erosion by-products, yet the requirements of preclinical evaluation down to trace polymer levels add development delays and costs. The aim of this study was to demonstrate the ability of using mechanistic computational modeling to predict absorption down to trace levels based on fits of early polymer bioresorption data.

METHODS AND RESULTS
Thirty-two Yucatan mini swine underwent a single interventional procedure in which up to 3 coronary arteries were implanted with cobalt chromium stents (balloon:artery=1.05-15:1) coated with coating comprised 1.4 μg/mm2 sirolimus and 3 bioresorbable polymers (Supraflex Cruz, Sahajanand Medical Technologies, India). Animals survived to their scheduled timepoint and subjected to euthanasia and comprehensive necropsy. Stented arteries were excised post sacrifice and either analyzed for polymer molecular weight (MW) and polymer quantity using Gel Permeation Chromatography (GPC) with refractive index detection (RI), or histopathological evaluation, including scoring for the presence of polymer coating on a scale of 0 (absence) to 4 (high incidence). Hearts and representative tissue samples were collected for histopathological evaluation. A mechanistic computational model was validated against GPC results of polymer MW and quantity through 12-months and then used to predict terminal bioresorption and then determine later timepoints. GPC-RI of stented artery extracts effectively separated and quantified polymer chains at MW of ~5000-150,000 Daltons (Da) and exhibited polymer peaks that were clearly distinguishable from biological material peak through 9 months. The polymer peak at 5000 Da began to co-elute with biological materials at 12 months, becoming indistinguishable in signal intensity from biological material at later times. At 12 months the co-eluting biological material was estimated at up to 10% of the overall 5000 Da peak height. As such, the 12-month timepoint was the last analytical value used in fitting a mechanistic kinetic model of polymer scission and mass bioresorption. Extrapolation of the best-fit bioresorption model predicted a decline of residual polymer from 29.4% at 12 months to 10.8% at 18 months, 0.2% at 24 months and 0.00% at 27 months. Median histopathology scores for coating presence showed good concordance with GPC and modeling, dropping from 2 at 1-3 months to 1 between 6-19 months and to 0 at 24-30 months. Histologically there was progressive endothelialization of implanted vessels across timepoints, with minimal acute endothelial loss and complete coverage by 3 months.  Neointima was minimal and mature at 1 month and all subsequent timepoints. Histomorphometric analyses showed comparable arterial parameters across timepoints with nominal (14-25%) area stenosis, thin neointima, and broadly patent lumens. There were no histologic findings in downstream myocardium.

CONCLUSIONS
Overall, treatment in swine coronary arteries with the Supraflex Cruz Sirolimus Eluting Coronary Stent System, which utilized a bioresorbable coating matrix to regulate release, resulted in mild neointimal responses across all timepoints, and no adverse safety events. Model-based extrapolation of GPC measured polymer masses showed high concordance with histopathology scores, offering a novel means for dynamically determining the terminal timepoints in preclinical studies.