What a bone marrow lesion actually is

If your MRI report mentions a bone marrow lesion, the phrase can feel alarming — especially when you were expecting news about your cartilage. The two findings are related but distinct, and understanding the difference matters for how your knee pain is assessed and managed.

A bone marrow lesion (BML) is not damage to the smooth cartilage lining the joint surface. It sits deeper — in the subchondral bone, the dense layer of bone immediately beneath the cartilage. On an MRI scan, a BML appears as a bright area of signal within that bone, reflecting a localised zone of microfracture, low-grade inflammation, and fluid-like change (often called bone oedema). Think of it as an area where the bone structure has been disrupted and is showing signs of stress.

That distinction has a direct bearing on pain. Articular cartilage contains no nerves and no blood vessels, so cartilage damage itself cannot generate pain directly. Subchondral bone, by contrast, is richly innervated. When a BML develops, the affected bone tissue can produce its own pain signal — one that is sometimes disproportionately intense relative to what the cartilage images show.

BMLs are more common than many patients realise. They are found in up to 66% of people with symptomatic arthritic knees, so a positive finding is a familiar clinical observation rather than an unusual or alarming one. Equally, an MRI BML is a finding, not a self-contained diagnosis — its clinical significance depends on where it sits in the joint, how large it is, and whether it corresponds to the location of the patient's symptoms.

Why BMLs cause pain out of proportion to cartilage wear

Several mechanisms converge in a BML zone to produce pain that can feel completely out of step with what a cartilage scan shows. Because the bone beneath the cartilage is densely wired with sensory nerve fibres, the inflammation, raised intraosseous pressure, and tiny microfractures concentrated within a BML all feed directly into the pain pathway — without any cartilage wear being required. Within an active lesion, new perivascular sensory nerve fibres grow into the affected tissue, amplifying the signal further and generating the deep, aching quality that patients often describe as coming from inside the joint rather than the surface.

The clinical upshot is that two patients with broadly similar cartilage scans can have very different pain experiences depending on whether BMLs are present. In knee OA populations, people who have BMLs are two to five times more likely to report knee pain than those without — a gap too large to attribute to coincidence.

Crucially, BMLs are not fixed scars. Lesion volume can expand and contract over time, and that fluctuation has a measurable effect on symptoms. Data from the Osteoarthritis Initiative, drawn from 1,412 knees with varus alignment, showed that changes in BML score over 24 months tracked closely with corresponding changes in weight-bearing pain severity: as lesion volume grew, pain tended to worsen; as it shrank, pain tended to ease.

That bidirectionality reframes BMLs as an active target rather than a passive imaging finding. It raises the possibility that treatments aimed at reducing lesion volume may translate into meaningful pain relief — something standard cartilage-surface injections are not designed to achieve.

How BML location shapes the specific activities that hurt

Many patients notice that their knee pain is sharply selective — bad on the stairs, manageable on a flat walk, or vice versa — and wonder whether that inconsistency means something is being missed. It is not inconsistency; it is anatomy.

Research using OAI data found that higher BML scores in the medial femorotibial compartment were specifically associated with greater pain during walking and standing, while higher scores in the lateral patellofemoral compartment were associated with stair-climbing pain. The pattern reflects the mechanical loading each compartment absorbs during those tasks: medial tibiofemoral contact forces dominate level-ground walking, whereas patellofemoral joint forces increase substantially when the knee bends under load — on stairs, slopes, or from a chair.

For rehabilitation, this distinction is clinically useful. It points toward modifying the specific activities that load the affected compartment rather than prescribing general rest, which offers little benefit and risks deconditioning. Knowing that a medial BML is driving walking pain, for example, guides which movement patterns to temporarily adapt and which to preserve — a more targeted approach than blanket activity avoidance. Task-specific pain is, in this sense, information rather than a mystery.

BMLs as early drivers of cartilage loss — not just downstream markers

The question that follows most imaging conversations — "will this get worse?" — has a more direct answer when BMLs are involved than many patients expect.

Both the presence and size of a BML predict future cartilage loss, and that deterioration concentrates in the same joint compartment as the lesion rather than spreading diffusely across the knee. A medial BML signals elevated risk to medial cartilage specifically — making it a locatable, targetable finding rather than a generalised warning sign.

More significant is an emerging reframing of where BMLs fit in the disease sequence. Conventional thinking placed them downstream — a consequence of cartilage wear already under way. Research published in 2023 (Shi et al.) challenges that ordering, finding that subchondral bone degeneration, including BML formation, may precede cartilage breakdown rather than merely follow it. This positions BMLs earlier in the OA cascade than traditionally assumed, at a stage where intervention may still influence the trajectory — though this reframing, while well-supported, continues to be refined by ongoing research.

Biological evidence reinforces the idea that BMLs are active rather than inert. Short-term rises in bone-resorption markers — CTX-I in blood and CTX-Iβ in urine — are associated with increased odds of BML worsening over 24 months, reflecting ongoing bone turnover beneath what a static scan captures at a single point in time.

One modifiable contributor is lower-limb alignment. In early knee OA, greater proximal tibial inclination — associated with varus loading — correlates with BML severity, consistent with medial compartment stress concentrating the load in that zone. Mechanical forces are adjustable in ways that already-lost cartilage is not, which is why identifying alignment as a driver has practical value beyond the diagnostic.

Reducing load as first-line BML management

Acting on the mechanical insight that concentrated load drives BML formation leads directly to the first management step: reducing or redistributing that load.

Subchondral bone has limited capacity to repair itself while the forces that created the lesion remain unchanged. Offloader knee bracing addresses this for medial compartment BMLs by shifting compressive force laterally, giving the overloaded bone a mechanical reprieve. The brace does not treat the lesion directly — it removes the primary stressor long enough for resolution to begin.

Targeted physiotherapy works through similar logic. Quadriceps strengthening, gait re-education, and hip and core stabilisation all reduce peak stress at the tibiofemoral contact point. The aim is to identify which movement patterns are loading the affected compartment and adapt them — a process that benefits from clinical assessment rather than a generic exercise programme.

Where excess body weight is present, reducing overall joint load may be associated with improvements in BML volume. The relationship is mechanically straightforward: each kilogram removed reduces the compressive force crossing the knee with every step.

Importantly, none of this is passive rest. The goal is active load redistribution — removing harmful loading patterns while preserving the movement and conditioning that keep the joint healthy. Inactivity risks deconditioning without addressing the mechanical cause.

These approaches do not guarantee resolution, and some BMLs are shaped by structural alignment factors that require specialist evaluation to manage properly. A clinical assessment to identify the overloaded compartment — and match the appropriate load-redistribution strategy to it — is the right starting point before any escalation is considered.

When conservative care isn't enough — targeted bone interventions

For some patients, load redistribution and physiotherapy do not provide adequate relief — and at that point, the anatomical gap between standard treatments and the actual pain source becomes clinically significant.

Most knee injections — corticosteroids, hyaluronic acid, PRP — are delivered into the joint space, targeting synovial tissue and articular surfaces. The subchondral bone lies beneath that surface, and the deep lesion zones where BML pain originates are not reached by intra-articular delivery. This is a structural mismatch, not a reflection of a patient trying insufficiently with conservative care.

Subchondroplasty addresses this directly. Under imaging guidance, calcium phosphate is injected percutaneously into the BML itself, targeting the lesion site rather than the joint cavity above it. The material fills the structurally compromised zones, aiming to reduce local mechanical stress and the raised intraosseous pressure implicated in pain generation.

A 2022 systematic review covering ten studies and 540 patients found significant pain relief and meaningful functional gains following the procedure, with few reported complications. At two-year follow-up, 14–30% of patients went on to total knee arthroplasty — meaning the majority either avoided or substantially deferred replacement surgery. Long-term radiological outcomes and prospective randomised controlled trial data remain limited, so the evidence base, while sufficiently positive to support recommendation after conservative failure, is not yet definitive.

Subchondroplasty is an escalation option, not a first-line approach. It is appropriate for patients whose BML has been identified as the primary pain driver and who have not responded adequately to conservative management. Establishing that suitability requires specialist assessment — an imaging-supported evaluation of lesion site, size, and structural context. If your symptoms persist despite load-management strategies, a formal assessment can determine whether this targeted intervention is the right next step for you.

1. [1] Fast quantitative BML measurement on knee MRI for OA assessment (deep learning, 2022). (2022). https://doi.org/10.1016/j.ocarto.2022.100234 https://doi.org/10.1016/j.ocarto.2022.100234
2. [2] Subchondroplasty in the treatment of bone marrow lesion in early knee OA: systematic review (2022). (2022). https://doi.org/10.1016/j.knee.2022.10.004 https://doi.org/10.1016/j.knee.2022.10.004
3. [3] Association of subchondral BML localization with weight-bearing pain in knee OA (OAI data). (2020). https://doi.org/10.1186/s13075-021-02422-0 https://doi.org/10.1186/s13075-021-02422-0
4. [4] Association of biochemical markers with BML changes on MRI — FNIH OA Biomarkers Consortium (2024). (2024). https://doi.org/10.1186/s13075-023-03253-x https://doi.org/10.1186/s13075-023-03253-x
5. [5] Effect of Vitamin D supplementation on synovial tissue and BML volume in symptomatic knee OA (VIDEO trial, 2019). (2019). https://doi.org/10.1186/s12891-019-2424-4 https://doi.org/10.1186/s12891-019-2424-4
6. [6] BML severity associated with proximal tibial inclination in early knee OA (2021). (2021). https://doi.org/10.1007/s00167-020-06378-7 https://doi.org/10.1007/s00167-020-06378-7