Tendon versus ligament: two collagen tissues, two very different fates

The most striking piece of tendon research is a carbon dating study. Heinemeier and colleagues used the atmospheric carbon-14 spike from mid-century nuclear testing as a biological timestamp and found that the core of the human Achilles tendon contains collagen laid down in adolescence. After roughly seventeen years of age, the tendon core essentially stops turning over. The tissue you compete on is, chemically, the tissue you grew.

That does not mean tendon is inert. The outer layers, the peritendinous tissue, and the interfascicular matrix all remain metabolically active, and tendon absolutely adapts: it stiffens, thickens, and changes its material properties in response to training. It just does so slowly, and it does so through cells that are outnumbered by matrix at a ratio that makes muscle look positively crowded.

For the clinical side, the pivotal work is Alfredson's 1998 heavy eccentric calf protocol, which took chronic Achilles tendinopathy patients who were waiting for surgery and returned most of them to running. Since then, dozens of trials across the Achilles and the patellar tendon have compared loading protocols to rest, to ultrasound, to injections, and to each other. Loading wins, essentially every time. What has not been established is which loading flavor is best: heavy slow resistance, eccentric-only, and isometric-first protocols all work, with differences smaller than the between-patient variance.

Ligaments split into two stories. Extra-articular ligaments with decent blood supply, the medial collateral of the knee being the classic example, heal reliably through scar formation and usually do well non-operatively. The anterior cruciate is the opposite case: intra-articular, poorly vascularized, and continuously washed by synovial fluid that disperses the fibrin clot a healing response would need to organize around. The traditional teaching has been that a fully ruptured ACL does not heal.

That teaching is being revised, and honesty requires saying so. Secondary analyses of the KANON trial reported MRI evidence of ACL healing in a meaningful minority of patients managed non-operatively, some of whom had good functional outcomes. This is not permission to skip surgical consultation. It is a reminder that the categorical claim is softer than it was taught, and it is being actively studied.

Tendon is a hierarchical rope. Tropocollagen triple helices assemble into fibrils, fibrils into fibers, fibers into fascicles, fascicles into the tendon, with an interfascicular matrix between them that allows sliding. The strength comes from covalent crosslinks between collagen molecules, many of them formed by lysyl oxidase, and those crosslinks are why mature collagen is both strong and slow to replace: to remodel the core you have to break chemistry that was built to not break.

The resting tendon holds a crimp, a visible waviness in the collagen. The first few percent of strain simply straightens the crimp, which is the toe region of the stress-strain curve and costs almost nothing. Past that you are in the linear elastic region, where the tendon behaves like a spring and stores energy. Somewhere past roughly four to eight percent strain, individual fibrils start failing, and past that the whole structure ruptures. Tendinopathy lives in the repeated-microdamage zone below rupture.

Tenocytes read strain through integrins, through primary cilia, and through stretch-activated ion channels, and they respond by upregulating IGF-1, TGF-beta, and procollagen expression. This is mechanotransduction, and it is the entire therapeutic lever: an unloaded tenocyte down-regulates matrix synthesis and up-regulates matrix-degrading enzymes. Rest is not neutral for tendon. Rest is a signal to disassemble.

Now the ligament. Ligament injury follows the ordinary wound-healing sequence: hemorrhage and clot, inflammation, proliferation with fibroblasts laying down a disorganized matrix rich in type III collagen, then remodeling in which type III is gradually replaced by type I and the fibers reorient along lines of strain. The problem is the finish line. Scar ligament in animal models plateaus somewhere around fifty to seventy percent of the original tissue's tensile strength, with different crosslink profiles and a smaller mean fibril diameter. It is functionally adequate for the MCL. It would be inadequate for the ACL even if it formed.

The ACL's specific problem is environmental. It sits inside the joint capsule, so a clot cannot organize into a scaffold: synovial fluid contains plasmin activators that lyse fibrin, and the torn ends retract. Its blood supply comes largely from the middle geniculate artery through the synovial sheath, and when that sheath tears the supply goes with it. So you get resorption rather than repair, and the stump you see on arthroscopy months later is often just that.

One more mechanism worth knowing because it explains a common frustration. A tendon graft or a healing tendon goes through ligamentization or remodeling in which it is initially revascularized and repopulated by host cells, and during that phase its mechanical properties are temporarily worse than they were on the day of surgery. The tissue gets weaker before it gets stronger. That is the physiological reason rehab timelines have a middle section that feels like an unnecessary delay and is not one.