What bone marrow lesions actually are

Finding the phrase 'bone marrow lesion' on an MRI report can sound alarming. The name is misleading — these are not lesions in the tumour sense, and the term replaced an earlier label, 'bone marrow oedema', precisely because the underlying process is more complex than simple fluid build-up.

A bone marrow lesion (BML) appears as a bright area on MRI sequences designed to suppress fat signal — typically STIR or fat-saturated T2 sequences — in the subchondral bone: the dense layer of bone that sits directly beneath the articular cartilage. What the signal reflects is a zone of trabecular micro-damage, inflammatory change, and abnormal bone remodelling within that layer, rather than straightforward swelling. Later research established that inflammatory responses, rather than fluid influx alone, drive the MRI signal — which is why the terminology shifted.

BMLs are common in knee osteoarthritis, though they also arise after acute joint trauma. Crucially, the size of a BML on imaging does not map neatly onto pain severity: some people with sizeable lesions report modest discomfort, while others with smaller findings experience significant symptoms.

Because BMLs sit in the bone beneath the cartilage surface rather than in the cartilage itself, they represent a structurally distinct problem. An MRI finding of a BML is one piece of clinical information — it needs to be weighed alongside symptoms, physical examination, and the wider joint picture before any conclusions are drawn about progression or treatment.

The link between BMLs and cartilage loss in knee OA

The subchondral bone and articular cartilage are not independent structures — they form a functional unit, and what happens in one layer influences what happens in the other. This matters because it shifts how BMLs should be interpreted: not as isolated bone findings, but as signals of a process that may be accelerating damage to the cartilage above.

The relationship appears to work in both directions. Abnormal bone remodelling — the increased turnover, microfracture, and structural change characteristic of progressive osteoarthritis — contributes to BML formation. At the same time, established BMLs are associated with measurable acceleration of overlying cartilage volume loss in longitudinal OA studies, making them a structural progression marker rather than merely a pain signal. The precise mechanism is not fully resolved, but abnormal load transmission is thought to play a central role: when forces concentrate unevenly across a compartment, the subchondral bone reflects that mechanical stress, and the inflammatory cascade sustaining BMLs may interfere directly with the cartilage maintenance processes immediately above.

At a cellular level, mesenchymal stem cells (MSCs) — which contribute to both bone and cartilage repair — show documented alterations within bone marrow lesions in patients with osteoarthritis (Campbell et al., 2016). Whether this represents a cause of impaired cartilage maintenance or a consequence of existing tissue damage remains an open question, but the finding reinforces the view that BMLs are an active pathological environment rather than an inert MRI finding.

The clinical implication is straightforward: if BML burden can be reduced, there is a plausible, biologically grounded rationale for expecting benefit to the overlying cartilage — though the evidence chain from BML reduction to measurable cartilage preservation has not yet been fully established in prospective studies.

What standard intra-articular injections do — and don't do — for BMLs

Corticosteroids, hyaluronic acid (viscosupplementation), and platelet-rich plasma (PRP) are the most widely used intra-articular agents in knee osteoarthritis, and each carries a reasonable evidence base for reducing pain and improving function over the short to medium term.

Their mechanism of action, however, operates in a different compartment from BMLs. Corticosteroids dampen intra-articular inflammation and synovial fluid volume; hyaluronic acid restores some of the viscous properties that synovial fluid loses in OA; PRP delivers growth factors into the joint space to modulate the local inflammatory environment. All three act primarily at the level of the joint cavity and synovium — not within the subchondral bone where BMLs reside.

None of these agents were designed or specifically tested with BML volume reduction as a primary outcome, and the current evidence does not document a structural effect on subchondral lesions (Jones et al., Nat Rev Rheumatol, 2019). That is not a criticism of their clinical value — it is a statement about therapeutic scope. Where pain relief and short-term functional improvement are the priority, standard intra-articular injections remain a reasonable pathway. The distinction becomes important when the question shifts to whether the bone pathology itself, rather than its downstream symptoms, can be structurally modified.

The emerging signal from polyacrylamide hydrogel (iPAAG)

Polyacrylamide hydrogel (iPAAG, marketed as Arthrosamid®) behaves differently from any of the agents discussed in the previous section. Rather than dispersing through joint fluid, it integrates into the synovial membrane as a non-degradable scaffold — composed of 2.5% cross-linked polyacrylamide and 97.5% water — and remains in place as a mechanical buffer that may alter the way load is distributed across the subchondral plate. That structural integration is what makes it a mechanistically interesting candidate for influencing subchondral pathology, not just symptom management.

The most direct available evidence comes from a 2022 publication in the Journal of Arthritis (Maulana, Cole, and Lee), which reported measurable reductions in patellofemoral bone marrow lesions following a single iPAAG injection in patients with advanced knee OA. The hypothesised pathway is indirect: by cushioning the joint against abnormal load concentrations, iPAAG may attenuate the mechanical and inflammatory drivers that sustain BMLs — a mechanism that fits the bone–cartilage dynamics outlined earlier in this article.

The study is single-centre and non-randomised, which means it establishes proof of concept rather than confirmatory evidence; larger controlled trials with BML volume as a pre-specified outcome are the logical next step. An independent study of 269 patients and 314 knees confirmed that iPAAG produces sustained improvements in pain and function over two years, though systematic BML measurement was not a primary endpoint in that cohort. Mechanistic research is continuing to characterise iPAAG's structural effects in standard care settings.

The evidence is preliminary — that qualification matters — but the Maulana finding remains the clearest available citation linking an intra-articular injection to a reduction in BML burden. That is a meaningful distinction in a field where no comparable data exist for classical agents.

Sub-chondroplasty: injecting directly into the lesion

Sub-chondroplasty occupies a different category from anything discussed so far: its needle enters bone, not the joint cavity.

The procedure is performed under fluoroscopic (X-ray) guidance, with the working needle advanced through the cortical shell and into the trabecular lesion itself. A flowable calcium phosphate bone substitute is then delivered directly into the BML, filling the defect and providing a structural scaffold on which new bone can form. Because the material is deposited within the lesion rather than into the joint space, the intervention targets the subchondral pathology directly, rather than relying on an indirect mechanical or biological pathway to reach it.

In practice, sub-chondroplasty is typically combined with arthroscopy, allowing the surgical team to address cartilage pathology at the joint surface in the same sitting — a recognition that BMLs and cartilage defects commonly coexist and often warrant simultaneous attention. This means it sits at a more advanced point on the decision pathway: it requires surgical access, theatre time, and anaesthesia, making it a materially different class of intervention from an outpatient injection.

The evidence base is still evolving. Published series suggest that patients with symptomatic, imaging-confirmed BMLs that have not responded to conservative management may experience meaningful pain reduction following the procedure, though robust randomised controlled trial data remain limited. It is generally considered once conservative measures and intra-articular options have proved insufficient — not as an early or first-line choice.

What evidence is still needed — and what this means for patients now

No prospective, MRI-controlled trial has yet measured BML volume as a pre-specified primary outcome after any intra-articular injection. The 2022 iPAAG finding is an important signal, but a single-centre, non-randomised observation is not the same as confirmatory evidence. The research chain that clinicians and patients most need — injection to BML reduction to demonstrably slower cartilage loss to delayed joint replacement — has not yet been traced in longitudinal imaging data. The biology supports the investigation, and the mechanistic rationale is sound, but that chain remains incomplete at present.

That gap does not make BMLs irrelevant to clinical decision-making. They represent a structural finding that correlates with pain severity and with the rate at which cartilage deteriorates, which means their presence — particularly when symptomatic — is worth raising with a specialist rather than setting aside until trial data mature.

In practice, not every BML requires active intervention. Size, location, symptom burden, and how the lesion sits relative to areas of cartilage stress all influence whether watchful waiting, targeted injection, or a more direct procedure is appropriate. That judgement cannot be made from an MRI report alone; it requires clinical examination alongside a conversation about the patient's activity demands, symptom severity, and response to conservative measures so far. For patients whose symptoms have persisted despite physiotherapy and load management, a specialist assessment — including MRI review in its clinical context — is a reasonable and well-supported next step.

1. [1] Trabecular oedema (Bone Marrow Lesion). https://en.wikipedia.org/?curid=73122483 https://en.wikipedia.org/?curid=73122483
2. [2] Osteoarthritis. https://en.wikipedia.org/?curid=504841 https://en.wikipedia.org/?curid=504841