Cartilage healing

Why does a broken bone knit itself back together while damaged knee cartilage never does?

Snap a femur — the largest, densest, most heavily loaded bone in the body — and in a few months it is whole, often stronger at the repair than the original. Scuff the cartilage on the end of that same femur, a layer a few millimetres thick, and it stays scuffed for life.

Intuition says the bigger injury to the harder tissue should be the harder repair. It is the other way round. So the question is not why cartilage is fragile. It is what bone has that cartilage lacks.

Follow what happens in the first hours after a fracture, because the sequence is the answer. The break tears vessels in the marrow and in the periosteum, the sleeve around the bone. Blood fills the gap and clots. That clot is not debris — it is the delivery system. Platelets caught in it release growth factors; inflammatory cells follow the chemical gradient in; and behind them come mesenchymal progenitor cells from marrow and periosteum, which lay down a soft callus, convert it to woven bone, and then spend a year or two remodelling that to aligned lamellar bone under load.

Now count how many steps in that sequence require blood. The clot: blood. The signalling: blood-borne cells. The builders: cells arriving from marrow through vessels. Essentially all of it.

Articular cartilage has none. It is avascular, and also without nerves or lymphatics; its cells are fed by diffusion from the joint fluid, helped along by walking compressing the tissue and pumping fluid through it. So a defect confined to the cartilage tears no vessel. No bleeding, therefore no clot, therefore no signal, therefore no arriving cells. The injury is not merely slow to heal; the machinery of healing is never summoned, because the summons travels by a route that does not exist here.

Could the cartilage repair itself from within? Chondrocytes are sparse — occupying something on the order of one to two per cent of the tissue's volume in an adult (a recalled figure) — and each sits sealed in a small cavity, a lacuna, inside a dense mesh of type II collagen and aggrecan it cannot crawl through. A skin cell at a wound edge migrates into the gap; a chondrocyte cannot. So the tissue has neither an external delivery route nor an internal one.

A third constraint explains why even a partial repair disappoints. Estimates from the racemisation of aspartate put the half-life of the collagen network in adult articular cartilage in the region of a century, with the proteoglycan between it renewed over years rather than decades (recalled, and the softest number here — one indirect method, and the quoted figures vary). Read plainly: the scaffold is built once during growth and thereafter maintained, not rebuilt.

Now a test. If avascularity is the limiting step, deliberately breaching the bone beneath the cartilage should produce repair. Surgeons do exactly this, drilling through the subchondral plate into the marrow — and it works as the theory predicts: a clot forms, marrow cells arrive, the defect fills. But what fills it is fibrocartilage, dominated by type I collagen, mechanically inferior and prone to wearing out over some years. The prediction is confirmed in direction and refined in substance: blood access is necessary, and not sufficient.

Two refinements, and then the observation that undoes the simple version.

The first refinement is that "cartilage never heals" is too strong. A defect penetrating into the bone below does heal, badly. And the knee's meniscus makes the point from the other side: its outer rim carries a blood supply and can be stitched with reasonable expectation, while its inner portion is avascular and is trimmed away instead, because stitching it achieves nothing. Same tissue type, same joint, different vascularity, different outcome.

The second is that this piece sits beside three neighbours in the collection. Wound healing traces the clot-to-scar sequence in soft tissue; Bone remodelling closes by asking how a fracture heals; Limits of regeneration asks why mammals scar where a salamander rebuilds. The fixed point of difference here is narrower: not why repair produces scar rather than the original, but why in one tissue the repair process is never triggered at all.

Now the failure. If avascularity alone explained it, every avascular tissue should behave this way. The cornea does not. It is avascular by necessity — vessels would blind it — and it nonetheless resurfaces an epithelial injury within days, from a resident population of stem cells at its rim that divide and migrate across the surface. So avascularity is not sufficient as an explanation. What covers the gap is that the real requirement is cell delivery by some route, and the cornea has an internal one that cartilage lacks. The load-bearing claim is therefore a conjunction: no vascular access, no mobile resident cells, and a matrix that cannot be rebuilt on any physiological timescale. Remove any one of the three — as the cornea does, and as the surgeon partly does — and the picture changes.

• Why cell-based repairs — implanting cultured chondrocytes, or a scaffold seeded with them — have proved so much harder than the reasoning above would suggest.
• How joint loading acts as cartilage's circulation, and what happens to the tissue during prolonged immobilisation.

Every term the collection defines is gathered in the glossary.