What the MRI report actually means

Receiving an MRI report that mentions 'bone marrow edema' or a 'bone marrow lesion' can feel alarming — neither phrase sounds straightforward, and neither is explained in the covering letter. Both terms describe the same thing: an area of abnormal signal in the subchondral bone, the dense layer of bone that sits directly beneath the joint cartilage and acts as the knee's primary shock absorber and load distributor.

The older label, bone marrow edema, has largely been replaced in research by bone marrow lesion (BML), and the reason matters. Early imaging studies assumed the abnormal signal represented fluid that had leaked into the bone — hence 'edema'. Later work established that the dominant process is more often inflammatory, involving changes at the trabecular (spongy) bone level rather than simple fluid accumulation. The term BML reflects this more accurately, though radiology departments still use both, sometimes within the same report.

Importantly, a BML on MRI is a finding — not a standalone diagnosis. These lesions exist on a spectrum: small BMLs are relatively common in middle-aged adults who exercise regularly, in people with early-stage knee osteoarthritis, and after high-load sporting activity. The presence of a BML does not automatically explain every symptom, nor does its size reliably predict pain severity. Clinical interpretation requires the lesion to be considered alongside a patient's history and examination findings. Some BMLs cause significant deep, aching pain; others are incidental.

Why BMLs cause a distinctive type of pain

The subchondral bone is densely innervated — nerve endings run throughout it, responding to pressure, chemical change, and mechanical stress. When inflammation takes hold in this layer, the result is a pain that feels qualitatively different from the surface ache associated with worn cartilage: deeper, duller, and harder to locate precisely. Patients frequently describe it as a 'bone pain' rather than a joint pain, often worse after prolonged standing, walking on hard surfaces, or impact activity, and easing — though not always fully — with rest. This load-sensitive pattern reflects the mechanical stress component: the subchondral bone is being asked to absorb forces it can no longer distribute normally.

One of the most clinically important features of BML-driven symptoms is the potential mismatch between pain severity and what the MRI shows at the cartilage surface. A scan may reveal only mild cartilage thinning — findings a radiologist might describe as early or minor — yet the patient reports pain that is disproportionately disabling. The explanation lies beneath the cartilage: the trabecular bone itself is inflamed, and that inflammation sensitises bone pain receptors independently of what the overlying cartilage looks like. Put simply, the bone is hurting, and the cartilage scan does not capture this fully.

This distinction has a direct bearing on treatment logic. Many common approaches — corticosteroid injections, for instance, which primarily act on synovial inflammation within the joint space — may not adequately reach the subchondral compartment where the pain is originating. Addressing the symptoms without addressing their source often produces limited or short-lived relief.

Conditions that look similar on imaging

Not every BML-like signal on MRI points to the same problem — and the distinction matters because each condition follows a different path.

The closest look-alike is spontaneous osteonecrosis of the knee, commonly abbreviated to SONK. On MRI, SONK can produce a large, intense signal that resembles an extensive BML; under the microscope, however, the underlying process is a subchondral insufficiency fracture — a small collapse in the bone plate that then triggers secondary bone death. SONK tends to present in older women after minimal or low-energy activity, and the clinical significance is higher than a typical BML: if the subchondral plate is at risk of further collapse, weight-bearing restrictions or surgical input may be needed. Yamamoto and Bullough (J Bone Joint Surg Am, 2000) established this fracture-first sequence as the likely mechanism. Distinguishing SONK from a large OA-related BML depends on the clinical story, the lesion's position and morphology, and sometimes a repeat scan after a few weeks.

In younger, high-load athletes, a third pattern appears: stress reactions in bone. These also generate BML-like signal — think of the bone equivalent of a deep bruise from repetitive impact — but they frequently resolve with load reduction and structured rehabilitation, without any specific procedure.

All three scenarios share the same MRI language but demand different responses. That is why a scan finding alone is never sufficient: the patient's age, activity level, symptom pattern, and lesion characteristics all feed into a diagnosis that determines what happens next.

Starting with conservative care

For most patients, conservative management is the right place to start — and in many cases, it produces meaningful recovery without any procedure at all. BMLs that develop after a sports injury or a period of high mechanical loading have a recognised tendency to regress spontaneously over weeks to months, provided the load driving them is reduced. This is not a passive process: structured, supervised care gives that natural repair the best chance of completing.

The first priority is reducing stress on the affected compartment. That may mean a temporary switch from running to cycling or swimming, footwear adjustments, or the addition of offloading orthotics — the aim is to lower the repetitive impact reaching the subchondral bone without leaving the patient deconditioned.

Physiotherapy plays a central role here, though the mechanism is indirect. Strengthening the quadriceps, improving proprioception, and correcting gait patterns reduces the load transmitted to the bone with each step. It does not act on the BML itself, which is why a supervised programme is more effective than simple rest.

In patients who are overweight, weight management has an additional and meaningful effect: studies in knee osteoarthritis suggest that reducing body mass measurably decreases subchondral loading and may reduce BML size over time.

Conservative care warrants a structured review — typically at six to twelve weeks — to assess whether symptoms are tracking in the right direction. If they are not, that is the signal to consider specialist assessment and the options described in the sections that follow.

Injection and pharmacological options — what the evidence shows

When conservative care does not produce sufficient improvement, a range of injection and pharmacological options enter the picture — though it is important to be clear from the outset that most BML-specific evidence is observational, often derived from osteoarthritis trials where BML change was a secondary endpoint rather than the primary focus. No head-to-head trials yet directly compare these interventions for BML resolution.

Intra-articular agents

Corticosteroid, hyaluronic acid, and platelet-rich plasma (PRP) injections work primarily within the joint space — acting on synovial inflammation, joint lubrication, or local tissue signalling. They have no direct mechanism for reaching the subchondral bone, and there is currently no robust evidence that they reduce BML size. That said, they should not be dismissed outright: co-existing synovitis is common in BML-affected joints, and treating the inflammatory component may meaningfully reduce pain even if the deeper lesion persists.

A step further along the evidence spectrum, a 2022 paper published in the Journal of Arthritis reported a reduction in patellofemoral bone marrow lesions following a single intra-articular injection of polyacrylamide hydrogel (Arthrosamid®). This is a notable finding — it suggests some intra-articular agents may exert measurable effects on subchondral pathology, possibly by altering the mechanical and inflammatory environment within the joint. The observation remains to be confirmed in larger, controlled studies.

BMAC (bone marrow aspirate concentrate), combined with PRP, produced significant functional improvements across multiple outcome measures in Kellgren-Lawrence grade 2–3 knee OA compared with exercise therapy alone. The evidence base relates to knee OA broadly rather than BMLs specifically, though the biological rationale for subchondral benefit — given documented mesenchymal stem cell dysfunction within bone marrow lesions — is plausible.

Systemic bone-modifying agents

At the investigational end of the spectrum, strontium ranelate reduced BML scores in a subgroup of knee OA patients who had lesions at baseline, though the authors noted discrepancies between animal-model and human OA responses. Denosumab (60 mg subcutaneously every 12 weeks), a bone-modifying agent, is under active clinical investigation for OA-related BML reduction — trial NCT02771860 is ongoing at sites including Ghent University Hospital. Neither agent is currently standard care for this indication.

Taken together, Arthrosamid® and BMAC combined with PRP represent the most clinically actionable current options with some BML-relevant signal, while strontium ranelate and denosumab remain promising but investigational. Individual suitability depends on OA stage, lesion characteristics, and overall clinical picture — factors a specialist assessment is best placed to weigh.

When to escalate and what specialist assessment involves

Three signals suggest it is time to move beyond self-directed or GP-led care: symptoms that have not meaningfully improved after three to six months of structured conservative management; an MRI finding showing a large lesion, or one whose morphology raises concern about SONK rather than a straightforward BML; or prior injection treatment that produced only partial or short-lived benefit. At any of these points, specialist review is likely to yield clearer answers than another cycle of waiting.

A specialist assessment for suspected BML-driven knee pain typically begins with a detailed clinical history, weight-bearing X-rays (to assess joint-space narrowing and overall alignment), and review of the existing MRI. Where the scan is more than twelve months old, or where a standard sequence leaves the lesion incompletely characterised, a repeat or upgraded study may be requested — advanced quantitative techniques such as T2 mapping can provide tissue detail beyond standard sequences.

When a procedural pathway is under discussion, sub-chondroplasty is the most directly targeted minimally invasive option currently available. Performed as a day-case procedure, it is designed to address the subchondral bone defect underpinning the BML — reaching the deeper pain source that intra-articular treatments do not. Evidence is promising, though the durability of pain relief has not been definitively established in long-term controlled studies; it is best understood as a bone-preserving intervention rather than a definitive cure.

The choice between biologic injections and sub-chondroplasty — or a staged combination — depends on lesion size, OA grade, activity goals, and response to prior treatment. Where BMLs accompany end-stage osteoarthritis with extensive structural loss, joint replacement may be the appropriate pathway; resolving a bone marrow lesion alone cannot reverse advanced joint disease.

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