Investigation of a Prevascularized Bone Graft for Large Defects in the Ovine Tibia

Abstract

In vivo bioreactors are a promising approach for engineering vascularized autologous bone grafts to repair large bone defects. In this pilot parametric study, we first developed a three-dimensional (3D) printed scaffold uniquely designed to accommodate inclusion of a vascular bundle and facilitate growth factor delivery for accelerated vascular invasion and ectopic bone formation. Second, we established a new sheep deep circumflex iliac artery (DCIA) model as an in vivo bioreactor for engineering a vascularized bone graft and evaluated the effect of implantation duration on ectopic bone formation. Third, after 8 weeks of implantation around the DCIA, we transplanted the prevascularized bone graft to a 5 cm segmental bone defect in the sheep tibia, using the custom 3D printed bone morphogenic protein 2 (BMP-2) loaded scaffold without prior in vivo bioreactor maturation as a control. Analysis by micro-computed tomography and histomorphometry found ectopic bone formation in BMP-2 loaded scaffolds implanted for 8 and 12 weeks in the iliac pouch, with greater bone formation occurring after 12 weeks. Grafts transplanted to the tibial defect supported bone growth, mainly on the periphery of the graft, but greater bone growth and less soft tissue invasion was observed in the avascular BMP-2 loaded scaffold implanted directly into the tibia without prior in vivo maturation. Histopathological evaluation noted considerably greater vascularity in the bone grafts that underwent in vivo maturation with an inserted vascular bundle compared with the avascular BMP-2 loaded graft. Our findings indicate that the use of an initial DCIA in vivo bioreactor maturation step is a promising approach to developing vascularized autologous bone grafts, although scaffolds with greater osteoinductivity should be further studied. Impact statement This translational pilot study aims at combining a tissue engineering scaffold strategy, in vivo prevascularization, and a modified transplantation technique to accelerate large segmental bone defect repair. First, we three-dimensional (3D) printed a 5 cm scaffold with a unique design to facilitate vascular bundle inclusion and osteoinductive growth factor delivery. Second, we established a new sheep deep circumflex iliac artery model as an in vivo bioreactor for prevascularizing the novel 3D printed osteoinductive scaffold. Subsequently, we transplanted the prevascularized bone graft to a clinically relevant 5 cm segmental bone defect in the sheep tibia for bone regeneration.

Keywords: 3D printed scaffolds; in vivo bioreactor; large bone defect; prevascularization; sheep model.

FIG. 1.

Schematic of engineered vascularized bone graft development in an in vivo bioreactor and subsequent transplantation to tibial defect. In Phase 1, the osteoinductive growth factor-impregnated 3D printed scaffold was implanted into the vascular enriched muscle pouch to allow for microvascular invasion, whereas the osteoinductive factor in the scaffold promoted ectopic bone formation. In Phase 2, after maturation, the vascularized bone graft was transplanted into a critical segmental tibial defect to accelerate bone healing and restoration of function. 3D, three-dimensional.

FIG. 2.

Sheep tibia and 3D printed polycaprolactone/β-tricalcium phosphate scaffold with side-hooks. (A) A representative micro-computed tomography image of a sheep tibia; (B) computer model of the scaffold geometry; (C) top view of the scaffold after manufacturing; (D) a scaffold contains two plastic tubes to illustrate vascular bundle placement; and (E) the scaffold with deep circumflex iliac artery and accompanying vein during implantation.

FIG. 3.

Representative hematoxylin and eosin staining histology slide images from the seven scaffold treatments. Red indicates mineralized bone.

FIG. 4.

Representative photomicrographs of ISO-10993-6 histological scoring parameters. (A–C) Osteogenic response and fibrosis scoring. (A) Animal SY05-left. Photomicrograph demonstrating an osteogenic response score of 2 (mild; 25–50% bone filling of implant/site) and a fibrosis score of 2 (mild), characterized by thin anastomosing trabeculae of new bone being laid down directly on a fibrous tissue intermediate that has permeated the implant. Dense fibrous connective tissue comprises the remaining implant associated new tissue. (B) Animal SY05-right. Photomicrograph demonstrating an osteogenic response score of 0 (none), and a fibrosis score of 3 (moderate). New host tissue growing into the implant is composed entirely of dense fibrosis (black asterisks), whereas the only visible bone is pre-existing tibial bone of the implant site. (C) Animal SY07-right. Photomicrograph demonstrating an osteogenic response score of 2 (mild; 25–50% bone filling of implant/site) and a fibrosis score of 1 (minimal). Abundant new bone composed of densely packed trabeculae (blue asterisks) infiltrates the implant material/void space left by the implant. The remaining space between the new bone is filled with increased amounts of adipose tissue, with only minimal fibrous tissue (black asterisk). (A–C) are 1.25 × magnification, scale bar = 1 mm. (D–F) Inflammation. (D) Animal SY05-left. Photomicrograph demonstrating a cumulative inflammation score 6. Multifocally, low to moderate numbers of lymphocytes (arrowhead) and multi-nucleate giant cells (arrows) infiltrate the fibrous tissue immediately surrounding the implant-associated void space. (E) Animal SY05-right. Photomicrograph demonstrating a cumulative inflammation score of 3/20. Inflammation is predominately composed of a few giant cells (arrows) infiltrating the fibrous tissue immediately adjacent to implant void spaces. (F) Animal SY07-right. Photomicrograph demonstrating a cumulative inflammation score of 2/20. Similar to SY05-right, inflammation in this animal was characterized primarily by few scattered giant cells (arrows) present throughout the fibrous tissue surrounding the implant/void space. Images (D–F) are 20 × magnification, scale bar = 50 μm. (G–I) Osteoblast activity and neovascularization. (G) Animal SY05-left. Photomicrograph demonstrating an osteoblast activity score of 2 and a neovascularization score of 1. A single layer of plump reactive osteoblasts (arrow) segmentally line endosteal surfaces of new bone. There is a minimal degree (∼1–2 vessels per 20 × magnification field) of new blood vessel formation (arrowheads). (H) Animal SY05-right. Photomicrograph demonstrating an osteoblast activity score of 0 and a neovascularization score of 2. No new bone and associated osteoblast activity was observed either extending from the margin of the tibial defect or being laid down directly within the implant. There is a mild degree (∼3–5 vessels per 20 × magnification field) of new blood vessel formation (black arrows) throughout the fibrous connective tissue surrounding the implant. (I) Animal SY07-right. Photomicrograph demonstrating an osteoblast activity score of 1 and a neovascularization score of 1. Osteoblast activity (arrow) was rarely observed along endosteal surfaces of new bone. Images (G–I) are 10 × magnification, scale bar = 100 μm.