Background and aim of the study: Biodegradable polymeric materials or extracellular matrix scaffolds are used in tissue-engineered heart valve designs, with the expectation of replicating the anatomic, histological and biomechanical characteristics of semilunar valves. The study aim was to evaluate the extent of in-vivo recellularization and the explant pathology findings of a prototype anionic, non-denaturing detergent and endonuclease technique used to decellularize allograft (homograft) valve conduits implanted in the right ventricular outflow tract (RVOT) of sheep, and to identify possible risks associated with tissue-engineered heart valve conduits based on decellularized allograft semilunar valve scaffolds.
Methods: Valve conduits were decellularized using a solution of N-lauroylsarcosinate and endonucleases, rinsed in lactated Ringers solution, and stored in an antibiotic solution at 4∞C until implanted. Explanted valves and unimplanted controls were examined macroscopically, radiographically (for calcification) and histologically using immunohistochemistry (IHC), routine and special histological stains, transmission electron microscopy (TEM) and polarized light microscopy (evaluation of collagen crimp).
Results: Cells and cellular remnants were uniformly absent in the decellularized cusps, but occasional focal sites of arterial wall smooth muscle cells and to a greater extent subvalvular cardiac myocytes were variably retained. The trilaminar histological structure of the cusp was preserved. Valve

conduit-related pathology consisted of intracuspal hematoma formation, collagen fraying, thinning of the conduit wall, and inflammatory cells associated with cardiac myocyte remnants. Cuspal calcification was not seen, but elastic fibers in the conduit wall and retained subvalvular cardiac myocyte remnants were liable to calcification. Fibrous sheath formation was present on the luminal surface of the conduit and extended over the cuspal surfaces to a variable extent. Myofibroblast-like cells repopulated the conduit wall and the basal region of the cusp. Re-endothelialization was variably present on the cuspal surfaces.
Conclusion: Explant pathology findings showed that in-vivo recellularization occurred, but was focally limited to regions of the arterial wall and cusp base. Safety concerns related to detergent and endonuclease use were identified. Methods to eliminate the potential for structural deterioration and enhance the rate and extent of recellularization of valve conduit tissue are required. Pathology findings showed implantation of valve conduits in the RVOT of juvenile sheep for 20 weeks to be a reliable animal model for the initial in-vivo assessment of decellularized valves. A 20-week period may be insufficient however to evaluate the long-term safety and effectiveness of a tissue-engineered valve conduit, as these depend on effective and phenotypically appropriate recellularization accompanied by sustained cell viability and function.
The Journal of Heart Valve Disease 2004;13:831-840