How scaffolds, stem cells and bioreactors are combined to build living bone grafts, and why the field is harder than it looks.
Bone is one of the few tissues in the body that heals without leaving a scar. A clean fracture, held still, rebuilds itself into bone that is mechanically indistinguishable from the original. That built-in capacity is exactly what tissue engineers try to borrow. When a defect is too large to bridge on its own – after trauma, tumour removal or a congenital condition – surgeons need to supply material that the body can colonise and eventually replace with its own tissue.
The traditional answer is the autograft: bone harvested from the patient’s own hip or another donor site. It integrates well because it is alive and immunologically identical, but it means a second operation, a second wound and a limited supply. Allografts from tissue banks and synthetic substitutes such as calcium phosphate ceramics avoid the harvest, yet they are slower to remodel and do not carry living cells. Engineered bone sits between these options: the goal is a graft that behaves like an autograft without the cost of taking one.
Almost every bone engineering strategy combines the same three elements. The first is a scaffold – a porous structure that gives cells something to attach to and defines the final shape. Scaffolds are made from ceramics, biodegradable polymers, or decellularised bone in which the original cells have been washed away and only the mineralised matrix remains. Pore size matters: openings of a few hundred micrometres let blood vessels grow in, while smaller pores mostly trap cells near the surface.
The second element is cells. Mesenchymal stem cells, which can be isolated from bone marrow or from fat tissue, are able to become bone-forming osteoblasts under the right signals. Using the patient’s own cells avoids immune rejection, although it means every graft is a small, individual manufacturing run. The third element is the environment: growth factors, mechanical loading and nutrient flow that tell the cells what to become and keep them alive while they do it.
A thin layer of cells in a dish gets all the oxygen it needs by diffusion. A graft the size of a jaw segment does not: cells more than a fraction of a millimetre from fresh medium starve. Perfusion bioreactors solve part of the problem by pumping culture medium through the pores of the scaffold, so cells deep inside the construct are fed as well as those on the surface. Flow also produces shear stress, which is itself a signal that pushes stem cells towards the bone lineage.
Once implanted, the graft faces the same challenge again. Until the body grows new blood vessels into it, the centre depends on diffusion. Much current research therefore focuses on pre-forming vascular channels, on co-culturing endothelial cells with bone cells, and on surgical techniques that bring a blood supply to the graft site early.
Craniofacial defects are a recurring example in the literature because the geometry is so irregular. A CT scan of the defect can be converted into a three-dimensional model, and the scaffold machined or printed to match it. The idea is that the surgeon receives a part that already fits, rather than carving a graft to shape in the operating theatre. The same digital pipeline makes it easier to document what was made and to repeat the process consistently.
None of these problems is unique to bone, which is why progress here tends to inform cartilage, tendon and other connective tissues as well.

Heart muscle does not regrow after a heart attack. Engineered cardiac tissue is being developed both as a future repair material and, already today, as a model for testing drugs.
Read more
Many cells in the body respond to electrical signals, and the body itself generates electricity every second. A sketch of how an introductory course on the subject can be structured.
Read more
Cells as a construction material: from bioprinted tissue to grown leather and fungal bricks, and the questions this raises for design, industry and society.
Read more