What happens to cartilage when knee OA is left alone

Doing nothing about knee osteoarthritis does not leave the joint in a holding pattern — it sets in motion a structural cascade that becomes progressively harder to reverse.

In the early stages, the smooth surface of articular cartilage begins to soften and fray. By the time changes become visible on X-ray — captured using the Kellgren-Lawrence (K-L) grading system — the joint space is already narrowing. At K-L Stage 1, roughly 10% of cartilage volume is lost and symptoms may be mild or absent. Stage 2 brings greater matrix breakdown, with definite osteophytes forming at the joint margins but the bones still not touching. By Stage 3, widespread erosion exposes patches of underlying bone and early subchondral sclerosis begins. At Stage 4, ≥60% of cartilage is gone: the femur and tibia articulate directly, the joint's cushioning function is essentially absent, and chronic pain and stiffness typically become constant.

Each structural change accelerates the next. Cartilage debris entering the joint space inflames the synovial lining; osteophytes grow larger and more painful; bone remodels in ways that reduce shock absorption — creating conditions that further stress whatever cartilage remains.

One important caveat: X-ray grade and lived experience do not always align. Some people carry Stage 3 or 4 changes on imaging with surprisingly manageable symptoms, while others report severe pain at Stage 2. That mismatch matters when thinking about next steps.

Why cartilage cannot repair itself

Unlike skin or bone, articular cartilage has no blood vessels running through it. That single structural fact explains much of what makes knee OA so difficult to arrest without intervention. Most tissues recruit fresh blood supply when damaged — delivering oxygen, nutrients, and the cells needed to rebuild. Cartilage cannot do this. Its resident cells, chondrocytes, are sparse, slow to divide, and rely entirely on diffusion from the surrounding synovial fluid for sustenance. Once the superficial layer is disrupted, there is simply no repair crew to call in.

This limitation matters because OA does not just passively wear cartilage thin — it actively dissolves it. Two enzymes, MMP-13 and ADAMTS5, are markedly upregulated in osteoarthritic joints and together break down the main structural proteins of the cartilage matrix: type II collagen and aggrecan. The rate of this enzymatic destruction can outpace any intrinsic repair that chondrocytes might attempt.

The process becomes self-amplifying. MMP-13 cleaves intact collagen fibres; fibroblast activation protein (FAP), which is elevated in the osteoarthritic synovium, then degrades the resulting fragments — clearing the ground for further attack. Meanwhile, the inflammatory cytokine IL-1β drives this enzyme activity higher still, meaning inflammation is not merely a by-product of cartilage loss but an active accelerant of it.

The result is a tissue with no internal plumbing for repair, being dismantled at a rate it cannot match.

Stage by stage: how the cartilage changes

The pace at which a joint moves through those structural changes is where the picture becomes more complex — and where data from large prospective studies start to matter.

The OA Initiative cohort (n=1,317, mean follow-up 4.9 years) quantifies this directly. In knees without full-thickness tibiofemoral cartilage defects, the annual rate of radiographic OA progression was 3.8%. In knees with medial full-thickness defects, that rate nearly doubled to 6.7%. The driver behind the acceleration is the so-called 'kissing' lesion: when both opposing cartilage surfaces in the medial compartment are eroded through to bone, the direct contact between them removes remaining cartilage far more rapidly than surface-level wear alone. The OA Initiative identified these medial bipolar lesions as an independent risk factor for compartment-level progression.

This makes full-thickness erosion a structural inflection point. Once subchondral bone is exposed and in direct contact with the opposing surface — the transition that occurs across Stages 3 and 4 — the joint loses the mechanism that had been slowing further damage. Bone-on-bone loading also drives the subchondral sclerosis and osteoclast-mediated remodelling described earlier, stiffening the bone plate and reducing its capacity to absorb load, which in turn places more stress on any cartilage that remains.

What this dataset cannot tell any individual is how quickly they personally will move through these changes. Some people carry full-thickness defects for years with limited radiographic deterioration; others progress rapidly over months. The stages describe structural status — they are not a fixed timeline, and they should not be read as a prognosis in isolation.

What happens to the rest of the joint

The inflammatory environment that builds as cartilage erodes produces visible structural changes well beyond the cartilage itself. Synovitis — irritation of the joint lining — generates effusion, the painful swelling patients often notice as a flare. But the more consequential long-term changes are mechanical.

As the joint senses instability, the body lays down new bone at the joint margins — osteophytes, or bone spurs. The intention, if such language applies to biology, is to broaden the load-bearing surface and limit unwanted movement. In practice, osteophytes grow progressively larger, restrict range of motion, and become a source of pain in their own right, particularly in Stage 3 and 4 disease.

Concurrently, the subchondral bone plate beneath the eroding cartilage remodels under abnormal load. Osteoclast activity rises whilst osteoblast and osteocyte counts fall, leaving the plate denser and less elastic — a stiffer platform that absorbs less impact and transfers more stress into whatever cartilage remains.

Uneven cartilage loss between compartments shifts the mechanical axis of the knee, producing varus (bow-legged) or valgus (knock-knee) deformity. This malalignment concentrates load precisely on the most damaged area, accelerating localised cartilage loss in a self-reinforcing cycle.

Finally, the gait adaptations that follow — altered stride pattern, subtle trunk shifts, reduced knee flexion — progressively overload the ipsilateral hip and lumbar spine. Hip and lower back pain in long-standing knee OA is often compensatory rather than independent pathology, which is one reason early specialist assessment carries value beyond the knee alone.

How quickly does this progress — and can it be predicted?

The figures already introduced — drawn from a prospective cohort of over 1,300 participants — describe average annual rates across a large, mixed population. That framing is useful, but averages conceal considerable individual spread, and no equivalent dataset yet allows a clinician to reliably forecast how fast any one person's knee will move through the stages.

What the evidence does identify are structural features associated with faster progression. Full-thickness defects that produce direct bone-on-bone contact in the medial compartment nearly double the radiographic annual progression rate compared with knees where the cartilage surface remains intact. Beyond that structural marker, certain sub-phenotypes carry a different risk profile: post-traumatic OA — arising after a prior ligament or meniscal injury — tends to advance more rapidly than primary degenerative disease; higher body mass index concentrates load across the joint and is consistently linked to faster cartilage loss; and cases with prominent synovial inflammation may behave differently from the more purely mechanical form.

Stage 4 disease, once reached, typically converges on joint-replacement assessment. The timeline to that endpoint, however, cannot be given with any precision for an individual — and the condition does not inevitably progress in every patient. The practical implication is that periodic monitoring — tracking both structural change and functional symptoms together over time — carries more prognostic value than a single assessment taken at one moment.

When to get a specialist assessment

Several signals suggest the time for GP-level management has passed. Pain that restricts daily function — stairs, getting up from a chair, walking more than a few hundred metres — for longer than six weeks despite rest and simple analgesia warrants specialist review. So does night pain that disturbs sleep, any episode of rapid functional decline over days to weeks, or persistent joint effusion (swelling that does not settle between flares). These are not emergencies, but they indicate that the joint is under a degree of stress that conservative self-management alone is unlikely to reverse.

What a specialist assessment actually involves is worth understanding before the appointment. A thorough clinical history and physical examination come first — range of motion, alignment, ligament stability, and a sense of which activities provoke symptoms. Imaging follows where needed: a weight-bearing X-ray allows Kellgren-Lawrence grading of joint space narrowing and osteophyte burden; MRI adds soft-tissue and cartilage detail that plain film misses. The imaging is one input into the clinical picture, not a verdict on its own — it is common to see near-normal X-rays in symptomatic patients, and severe radiographic changes in patients still functioning well.

The case for early review rests on the progression evidence covered in previous sections: joint-preservation strategies — whether conservative, biologic, or surgical — carry the most therapeutic latitude in Stages 1 to 3. By Stage 4, the viable options narrow toward replacement rather than preservation. Getting a clear sense of where on that spectrum a patient sits, while more of the pathway remains open, is the practical value of assessment.

For patients in London seeking that clinical picture, the London Cartilage Clinic on Harley Street offers specialist joint-preservation assessment as a structured starting point. An initial consultation can be arranged at londoncartilage.com.

1. [1] Integrated physicochemical and therapeutic assessment of nutraceutical capsules in a rabbit osteochondral defect model with OA-like changes. (2026). https://doi.org/10.1038/s41598-026-49088-9 https://doi.org/10.1038/s41598-026-49088-9
2. [2] Kartogenin prevents cartilage degradation and alleviates osteoarthritis progression via the miR-146a/NRF2 axis. (2021). https://doi.org/10.1038/s41419-021-03765-x https://doi.org/10.1038/s41419-021-03765-x
3. [3] Inhibition of fibroblast activation protein ameliorates cartilage matrix degradation and osteoarthritis progression. (2023). https://doi.org/10.1038/s41413-022-00243-8 https://doi.org/10.1038/s41413-022-00243-8
4. [4] Role of full-thickness cartilage defects in knee OA incidence and progression: Data from the OA Initiative. (2018). https://doi.org/10.1002/jor.24140 https://doi.org/10.1002/jor.24140
5. [5] EGFR signaling is required for maintaining adult cartilage homeostasis and attenuating osteoarthritis progression. (2022). https://doi.org/10.1002/jbmr.4531 https://doi.org/10.1002/jbmr.4531