Summary Statement. Demineralised bone matrix augmented tendon-bone fixations in the animal model show less scar tissue and an enthesis morphology closer to the physiologic one which may lead to a more resistant repair construct. Introduction. Rotator cuff repair is one of the most common operative procedures in the shoulder. Yet despite its prevalence recurrent tear rates of up to 94% have been reported in the literature. High failure rates have been associated with tendon detachment from bone at the tendon – bone interface. Exogenous agents as biological strategies to augment tendon – bone healing in the shoulder represent a new area of focus to improve patient outcomes. Demineralised bone matrix (DBM) contains matrix bound proteins, exposed through acid demineralization step of DBM manufacture, and has long been recognised for its osteoinductive and osteoconductive properties. We hypothesised that DBM administered to the bone bed prior to the reattachment of the tendon, will upregulate healing and result in enhanced tissue morphology that more closely resembles that of a normal enthesis. An established ovine transosseous equivalent rotator cuff model was used. Methods. Following ethics approval, 10 adult wethers (18 months) were randomly allocated to control, n=4 (without DBM) or DBM, n=6 (DBM administered to bone bed) groups. The infraspinatus tendon was detached from its insertion and repaired in a transosseous equivalent fashion using PEEK suture anchors. In treatment animals 0.25cc of ovine DBM, previously prepared using a modified Urist protocol, was injected into two drill holes within the bony tendon footprint. Animals were culled at 4 weeks following surgery and processed for tissue histology and microcomputed tomography (μCT) endpoints. Results. No infection or tendon detachment following repair was noted in either group. 3D reconstructed images of μCT scans verified correct DBM and suture anchor placement. Histological images demonstrated distinct differences in tissue morphology between the two groups; however there was no evidence of the four – zoned structure characteristic of a healthy tendon bone insertion, in any specimens. In the control group specimens, the tendon midsubstance was highly disorganised with randomly arranged collagen fibres and diminutive areas of fibrocartilage. In the treatment group, large regions between tendon and bone were occupied by fibrocartilage. Within the fibrocartilage region, insertional collagen fibres appeared organised and chondrocytes were orientated in the direction of the insertional collagen fibres. Organised collagen fibre orientation within the tendon midsubstance was observed, though this was not consistent throughout all the specimens. DBM particles were resorbed and trabecular bone occupied the DBM holes. The PEEK anchors were all in direct contact with the ongrowing bone indicating good quality integration and fixation. Discussion. This study showed that DBM augmented tendon to bone repair leads to an upregulated cellular activity resulting in increased amounts of fibrocartilage between the repaired tendon and underlying bone. The upshot of this is an improved tissue organization which more closely resembles the morphology of the normal enthesis. Introduction of osteoinductive DBM at the tendon – bone interface during surgery may reduce failure rates associated with rotator cuff repair and improve clinical outcomes

Background

Re-attachment of tendon to bone is challenging with surgical repair failing in up to 90% of cases. Poor biological healing is common and characterised by the formation of weak scar tissue. Previous work has demonstrated that decellularised allogenic demineralised bone matrix (DBM) regenerates a physiologic enthesis. Xenografts offer a more cost-effective option but concerns over their immunogenicity have been raised. We hypothesised that augmentation of a healing tendon-bone interface with DBM incorporated with autologous mesenchymal stem cells (MSCs) would result in improved function, and restoration of the native enthesis, with no difference between xenogenic and allogenic scaffolds.

Demineralized bone matrix (DBM) is a natural, collagen-based, well-established osteoinductive biomaterial. Nevertheless, there are conflicting reports on the efficacy of this product. The purpose of this study was to evaluate whether DBM collagen structure is affected by particle size and can influence DBM osteoinductivity.

Sheep cortical bone was ground and particles were divided in three fractions with different sizes, defined as large (L, 1–2 mm), medium (M, 0.5–1 mm), and small (S, < 0.5 mm). After demineralization, the three DBM samples were characterized by DTA analysis, XRD, ICP-OES, and FTIR. Data clearly showed a particle size-dependent alteration in collagen structure, with DBM-M being altered but not as much as DBM-S. The in vivo study showed that only DBM-M was able to induce new bone formation in a subcutaneous ectopic mouse model. When sheep MSC were seeded onto DBM particles before implantation, all DBM particles were able to induce new bone formation with the best incidence for DBM-M and DBM-S. Gene expression analysis performed on recovered implants supports the histological results and underlines the supportive role of MSC in DBM osteoinduction through the regulation of host cells. In conclusion, our results show a relation between DBM particle size, structural modification of the collagen and in vivo osteoinductivity. The medium particles represent a good compromise between no modification (largest particles) and excessive modification (smallest particles) of collagen structure, yielding highest osteoinduction. We believe that these results can guide researchers to use DBM particles of 0.5–1 mm size range in applications aimed at inducing new bone formation, obtaining results more comparable and reliable among different research groups. Furthermore, we suggest to carefully analyze the structure of the collagen when a collagen-based biomaterial is used alone or in association with cells to induce new bone formation.