Objectives. To assess the structure and extracellular matrix molecule expression of osteogenic cell sheets created via culture in medium with both dexamethasone (Dex) and ascorbic acid phosphate (AscP) compared either Dex or AscP alone. Methods. Osteogenic cell sheets were prepared by culturing rat bone marrow stromal cells in a minimal essential medium (MEM), MEM with AscP, MEM with Dex, and MEM with Dex and AscP (Dex/AscP). The cell number and messenger (m)RNA expression were assessed in vitro, and the appearance of the cell sheets was observed after mechanical retrieval using a scraper. β-tricalcium phosphate (β-TCP) was then wrapped with the cell sheets from the four different groups and subcutaneously implanted into rats. Results. After mechanical retrieval, the osteogenic cell sheets from the MEM, MEM with AscP, and MEM with Dex groups appeared to be fragmented or incomplete structures. The cell sheets cultured with Dex/AscP remained intact after mechanical retrieval, without any identifiable tears. Culture with Dex/AscP increased the mRNA and protein expression of extracellular matrix proteins and cell number compared with those of the other three groups. More bridging bone formation was observed after transplantation of the β-TCP scaffold wrapped with cell sheets cultured with Dex/AscP, than in the other groups. Conclusions. These results suggest that culture with Dex/AscP improves the mechanical integrity of the osteogenic cell sheets, allowing retrieval of the confluent cells in a single cell sheet structure. This method may be beneficial when applied in cases of difficult tissue reconstruction, such as nonunion, bone defects, and osteonecrosis. Cite this article: M. Akahane, T. Shimizu, T. Kira, T. Onishi, Y. Uchihara, T. Imamura, Y. Tanaka. Culturing bone marrow cells with dexamethasone and ascorbic acid improves osteogenic cell sheet structure. Bone Joint Res 2016;5:569–576. DOI: 10.1302/2046-3758.511.BJR-2016-0013.R1

Summary. Four highly cross-linked UHWMPEs except vitamin E-stabilised explants. Introduction. The development of both first and second generation highly cross-linked material focused on stabilizing radiation-induced free radicals as the sole precursor to oxidative degradation; however, secondary in vivo oxidation mechanisms have been identified in both conventional and highly cross-linked UHMWPE, induced by absorbed lipids and cyclic mechanical load. Retrieval studies are reporting in vivo oxidation highly cross-linked retrievals with up to ten year in vivo durations. Preclinical aging tests did not predict these in vivo material changes. With only a decade of these materials in clinical use, retrieval studies are limited to mid-term follow-up. In vitro studies face a challenge in effectively replicating the precise in vivo conditions that lead to this loss of oxidation resistance. In this study, we bypass replicating these in vivo variables by examining surgically-retrieved components, thereby testing material that has been affectively “pre-conditioned” by their in vivo service. After a preliminary post-operative analysis, we subjected retrievals to accelerated aging tests in order to predict the extent to which their oxidative stability had been uniquely compromised in vivo. Patients & Methods. Twenty-four highly cross-linked retrievals of four manufacturing methods (n=6 each of Longevity™, Prolong™, X3™ and E1™) and in vivo durations (1–4 years) were analyzed post-operatively and after accelerated aging (70°C, 5atm O. 2. for 2 weeks; ASTM F2003). Never-implanted components (n=1) of each material type were also aged. Infrared microscopy was used to evaluate lipid absorption, oxidation (per ASTM F2102-01ε1) and hydroperoxide levels after 16 hrs of nitric oxide staining for oxidation potential, and gravimetric swelling analysis assessed cross-link density (ASTM F2214). Results. All retrievals contained absorbed lipids penetrating below both loaded (penetration depth=1.3 ± 0.5 mm) and unloaded (0.6 ± 0.2 mm) surfaces. Each material type subset contained retrievals with and without detectable oxidation after in vivo service (Max OI=0.01–0.94). After aging, all post-irradiation thermally-treated, highly cross-linked retrievals, regardless of initial lipid levels or oxidation, showed oxidative degradation, demonstrated by subsurface oxidative peaks (MOI=0.30–2.63), increased hydroperoxides (3–5X), and decreased cross-link density (−34–90%). In contrast, vitamin E-stabilised retrievals showed below MOI<0.2 with no significant loss of cross-link density. Never-implanted controls for each material type showed no oxidative changes after accelerated aging. Discussion/Conclusion. Accelerating aging after in vivo service has shown oxidative instability characterised by high oxidation and material property loss in the three highly cross-linked materials without an incorporated antioxidant. This oxidative degradation took place regardless of post-operative oxidation levels, indicating that even without detectable oxidation the material had undergone changes during in vivo service, as compared to the lack of oxidative response in never-implanted controls. These findings also suggest that the presence of an antioxidant may be able to slow down and/or stabilise in vivo mechanisms compromising long-term oxidative stability and increase the longevity of highly cross-linked UHMWPE materials