What is avascular necrosis?

Avascular necrosis (AVN), also called osteonecrosis of the femoral head (ONFH), occurs when the blood supply to part of the femoral head — the "ball" of the hip joint — is disrupted. Bone is living tissue that depends on a continuous blood supply; when that supply is cut off or significantly reduced, the bone cells in the affected area die. The overlying cartilage can survive for a time because it draws nutrition from joint fluid rather than blood vessels directly, which is part of why early AVN is often silent on plain X-rays even though damage is already underway underneath.

The femoral head is particularly vulnerable because its blood supply runs largely through a small number of vessels traveling along the femoral neck, with limited alternative routes. Once a significant volume of bone loses its blood supply, the dead bone gradually loses structural strength. Without intervention, the subchondral bone can collapse under normal body weight, distorting the round shape of the femoral head and, over time, damaging the cartilage and the socket (acetabulum) as well — at which point the process effectively becomes secondary osteoarthritis of the hip.

Who gets AVN, and why

AVN is not a single disease with one cause — it's a final common pathway that a number of different risk factors can lead to. In non-traumatic cases, corticosteroid use and heavy alcohol intake together account for the large majority of cases. Both appear to disrupt blood flow to bone through overlapping mechanisms, including fat cell overgrowth in the bone marrow, small blood vessel changes, and disturbed clotting, and the risk with both factors tends to be dose- and duration-dependent.

Trauma is the other major category: a femoral neck fracture or hip dislocation can directly tear the vessels supplying the femoral head, causing AVN sometimes years after the injury has otherwise healed. Less common causes include sickle cell disease, decompression sickness ("the bends," seen in divers and caisson workers), radiation exposure, certain clotting disorders, and — in a meaningful minority of cases — no identifiable cause at all, termed idiopathic osteonecrosis.

Common risk factors

AVN typically affects people between 30 and 50 years old — younger than the typical age for hip osteoarthritis — which is part of why it matters as a distinct diagnosis: joint-preserving treatment, when it works, can spare a patient in their 30s from needing a hip replacement decades earlier than they otherwise would.

How AVN is diagnosed and staged

Early AVN often presents as groin pain that is worse with weight-bearing, sometimes with referred pain to the buttock, thigh, or knee. Plain X-rays are frequently normal in the earliest stage, because bone density changes visible on X-ray only appear after the dead bone begins to be resorbed and repaired around its edges. MRI is the most sensitive test and can detect osteonecrosis well before it shows up on X-ray, which is why MRI is the standard next step whenever AVN is suspected clinically in a patient with a relevant risk factor.

Once diagnosed, AVN is staged — most commonly using the ARCO (Association Research Circulation Osseous) system — because stage is the single biggest factor in deciding what treatment is appropriate and what outcome to expect.

ARCO StageFindingsFemoral head shape
Stage 0–INormal or near-normal X-ray; abnormal MRI/bone scan onlyRound, no collapse
Stage IIX-ray changes (sclerosis, cysts) without collapseRound, no collapse
Stage IIISubchondral fracture ("crescent sign") and/or early flatteningEarly collapse
Stage IVJoint space narrowing and secondary osteoarthritic changeCollapsed, arthritic

Stages 0–II are described as "precollapse" and stages III–IV as "postcollapse." This distinction matters enormously: joint-preserving treatments are aimed at precollapse disease, where there is still a round femoral head worth saving. Once meaningful collapse and secondary arthritis have set in, the evidence for joint preservation weakens considerably, and the discussion shifts toward joint replacement.

Non-surgical treatment: what the evidence actually shows

For early, small-volume lesions, some clinicians recommend protected weight-bearing (crutches) and modification of activity, on the theory that reducing mechanical load may lower the risk of collapse while the bone attempts to heal itself. The evidence for activity restriction alone is limited and largely observational, and a 2024 systematic review of conservative treatment for AVN concluded that non-operative measures alone are generally insufficient to prevent progression in most patients, particularly with larger lesions, and are more reasonably framed as an adjunct or a bridge while a definitive treatment plan is worked out rather than a stand-alone cure.

Bisphosphonates (the same class of drug used for osteoporosis) have been studied specifically because they slow bone resorption, and resorption of dead bone is part of what weakens the femoral head structurally before collapse. The trial evidence here is genuinely mixed, which is worth being upfront about rather than overselling.

What the evidence shows: A small randomised trial (Lai et al., J Bone Joint Surg Am, 2005) of 40 hips with Steinberg stage II–III AVN found that 25 weeks of oral alendronate significantly reduced the rate of early femoral head collapse compared with no treatment. However, a larger, more rigorous two-year multicentre, randomised, double-blind, placebo-controlled trial (Chen et al., Arthritis Rheum, 2012) in 65 hips found no significant difference between alendronate and placebo in preventing collapse, need for joint replacement, or quality of life. Taken together, the current evidence does not clearly establish that bisphosphonates prevent collapse, and their use in AVN remains off-label and individualised rather than a default recommendation.

Joint-preserving surgery for precollapse disease

For symptomatic, precollapse (ARCO 0–II) AVN, the most established joint-preserving surgical option is core decompression: drilling one or more narrow channels from outside the bone into the necrotic area of the femoral head. This is thought to work by lowering the abnormally high pressure inside the bone marrow (which itself may contribute to pain and to restricting blood flow), and by creating a channel for new blood vessels to grow into the dead bone.

Core decompression alone has historically had inconsistent results, with reported failure rates (progression to collapse or need for hip replacement) ranging widely across studies depending on lesion size and stage. This has driven interest in combining core decompression with biological augmentation — injecting concentrated bone marrow cells, platelet-rich plasma, or bone graft material into the decompressed channel — to give the area a better chance of actually healing rather than simply relieving pressure.

What the evidence shows: A 2025 systematic review and meta-analysis (Konarski, Medical Sciences) of 14 studies and over 1,200 hips found that core decompression combined with biological augmentation (cell therapy, PRP, or bone marrow concentrate) produced significantly better functional scores and pain relief at up to 24 months than core decompression alone, whereas adding structural augmentation (bone grafts or fibular struts) did not show a consistent functional advantage. The authors' overall message was that biological augmentation is a reasonable first-line adjunct to core decompression in early-stage disease, while outcomes for structural augmentation are less consistent and need more study.

Other joint-preserving options exist for selected patients, including vascularised or non-vascularised bone grafting and femoral osteotomy to rotate a healthier segment of bone into the weight-bearing zone, but these are more technically demanding, used less often, and generally reserved for younger patients with specific lesion patterns.

When the joint has already collapsed

Once AVN has progressed to ARCO stage III–IV — subchondral fracture, flattening of the femoral head, or secondary osteoarthritis — joint-preserving procedures have a substantially lower chance of success, because there is no longer a round, structurally sound femoral head to preserve. At this stage, total hip replacement is generally the most reliable way to relieve pain and restore function, and long-term outcomes of hip replacement for AVN are broadly comparable to hip replacement for osteoarthritis, though patients with AVN tend to be younger at the time of surgery, which is part of why avoiding or delaying that step through earlier diagnosis and joint-preserving treatment matters so much when it's still an option.

If you have unexplained groin or hip pain and a risk factor such as steroid use, heavy alcohol use, or a past hip injury, the important first step is an MRI-based diagnosis and staging — not any particular treatment — because stage, more than anything else, determines what options are realistically on the table.

References (PubMed)

Mont MA, Salem HS, Piuzzi NS, Goodman SB, Jones LC. Nontraumatic Osteonecrosis of the Femoral Head: Where Do We Stand Today? A 5-Year Update. J Bone Joint Surg Am. 2020;102(12):1084–1099. DOI: 10.2106/JBJS.19.01271

Lai KA, Shen WJ, Yang CY, Shao CJ, Hsu JT, Lin RM. The use of alendronate to prevent early collapse of the femoral head in patients with nontraumatic osteonecrosis: a randomized clinical study. J Bone Joint Surg Am. 2005;87(10):2155–2159. DOI: 10.2106/JBJS.D.02959

Chen CH, Chang JK, Lai KA, Hou SM, Chang CH, Wang GJ. Alendronate in the prevention of collapse of the femoral head in nontraumatic osteonecrosis: a two-year multicenter, prospective, randomized, double-blind, placebo-controlled study. Arthritis Rheum. 2012;64(5):1572–1578. DOI: 10.1002/art.33498

Goncharov EN, Koval OA, Bezuglov EN, et al. Conservative Treatment in Avascular Necrosis of the Femoral Head: A Systematic Review. Med Sci (Basel). 2024;12(3):32. DOI: 10.3390/medsci12030032

Konarski W. Effectiveness of Different Types of Core Decompression in Early-Stage Osteonecrosis of the Femoral Head: A Systematic Review and Meta-Analysis. Med Sci (Basel). 2025;13(4):258. DOI: 10.3390/medsci13040258