Osteonecrosis (Avascular Necrosis)

Overview

Osteonecrosis (ON), also called avascular necrosis (AVN) or ischemic necrosis of bone, is the death of bone tissue resulting from interruption of its blood supply. Unlike osteomyelitis — in which bone dies because of bacterial infection and the accompanying inflammatory destruction — the cause of osteonecrosis is purely vascular: ischemia deprives osteocytes and marrow stromal cells of oxygen and nutrients, triggering cell death that the bone cannot reverse on its own. Once bone cells die in sufficient numbers, the structural integrity of that skeletal region is fundamentally compromised, and if the affected area is weight-bearing — especially the femoral head — progressive mechanical collapse of the articular surface is the almost inevitable endpoint without timely intervention.

The hip is by far the most clinically significant site. Approximately 10,000–20,000 new cases of femoral head osteonecrosis are diagnosed in the United States each year. The condition accounts for 5–12% of all total hip arthroplasties (THA) performed, making it a major driver of hip replacement surgery. Because the disease disproportionately strikes people in their 30s and 40s — younger than typical osteoarthritis patients — a THA placed at age 40 would need to function for 30–40+ years or require revision, raising the long-term stakes considerably. Osteonecrosis is a leading cause of hip disability in young adults, and because many risk factors (corticosteroids, sickle cell disease, systemic lupus erythematosus, alcohol misuse) are common, the condition is not rare.

Early diagnosis matters because outcome is strongly stage-dependent. Disease discovered before subchondral collapse (Ficat Stage I–II) can often be managed with joint-preserving procedures. Once the characteristic crescent sign of impending collapse appears (Stage III), joint-preserving options become much less reliable, and most patients progress to total hip arthroplasty. The treatment goal is therefore to diagnose early — magnetic resonance imaging (MRI) is the key tool — and intervene before the femoral head collapses.

Back to Table of Contents

Blood Supply to Bone and Mechanism of Ischemia

Understanding why the femoral head is so uniquely vulnerable to osteonecrosis requires understanding its peculiar vascular anatomy. The femoral head receives its dominant blood supply from the medial circumflex femoral artery (MCFA), a branch of the profunda femoris (deep femoral artery). The MCFA gives rise to the superior retinacular (posterosuperior) vessels, which course along the posterior surface of the femoral neck beneath the joint capsule — running in a vulnerable extracapsular and then intracapsular position — before penetrating the bone at the head-neck junction to supply the superior and anterolateral portions of the femoral head. This is the portion of the head that bears weight during normal gait, and it is precisely the area where osteonecrosis most often begins.

A minor contribution comes from the lateral circumflex femoral artery via the anterior retinacular vessels, and a small intraosseous supply via vessels in the femoral shaft. The ligamentum teres carries the foveal artery, which is functionally important only in infants and young children; in adults it contributes little. Critically, the femoral head has very limited collateral circulation compared with other bones that have robust periosteal and medullary supply networks. This means that disruption of the retinacular vessels produces a "watershed ischemia" with no meaningful backup route.

The mechanism of ischemia varies by etiology but the downstream pathology is consistent. Loss of blood supply to the osteocytes — the mature bone cells embedded in lacunae within mineralized matrix — leads to cell death within 12–48 hours of complete ischemia. Hematopoietic marrow cells are even more sensitive and die within 6–12 hours. The dead bone is structurally normal in appearance initially, but the repair process that normally replaces dead bone (creeping substitution) is slow and mechanically incomplete when the zone of necrosis is large. The interface between dead and living bone becomes a zone of resorption and weakened structure, setting the stage for the fatigue fracture — the subchondral crescent — that heralds collapse.

In the non-traumatic causes, the ischemia is typically intravascular rather than from external vessel disruption: fat emboli obstruct sinusoidal vessels, sickled red blood cells cause vaso-occlusion, nitrogen bubbles mechanically occlude capillaries, or thrombus forms in a hypercoagulable state. In each case the end result is a region of bone without perfusion, and the subsequent progression toward collapse follows the same pathological script regardless of the initiating event.

Back to Table of Contents

Causes — Traumatic

Femoral Neck Fracture is the classic traumatic cause of femoral head osteonecrosis. Intracapsular fractures — those within the hip joint capsule proximal to the intertrochanteric line — are the dangerous ones because the retinacular vessels run along the femoral neck inside the capsule and are stretched, kinked, or torn at the moment of fracture. Garden classification grades III and IV (displaced fractures) carry an AVN incidence of approximately 15–30% for Garden III and up to 35% for Garden IV. The risk increases substantially with delayed reduction: the compressed or stretched retinacular vessels may still be patent immediately after fracture, but delay beyond 6–12 hours before surgical reduction and fixation dramatically increases ischemia time. This is why femoral neck fractures are treated as urgent orthopaedic injuries. Impacted valgus fractures (Garden I–II) have much lower AVN risk (~5–10%) because displacement is minimal and vessels are less likely to be disrupted.

Hip Dislocation is the second major traumatic cause. Posterior dislocation — far more common than anterior, typically caused by dashboard-type mechanisms in motor vehicle collisions — drives the femoral head posteriorly and superiorly out of the acetabulum, tearing the posterior retinacular vessels and the joint capsule. The AVN risk following posterior dislocation is approximately 10–25%, with the critical determinant being time to reduction: reduction within 6 hours is strongly associated with lower AVN rates, while delays beyond 12–24 hours dramatically increase risk. Associated femoral head or acetabular fractures further increase AVN incidence. Anterior dislocation is less common and carries lower AVN risk because the posterior retinacular supply is less disrupted.

Slipped Capital Femoral Epiphysis (SCFE) is an adolescent condition in which the femoral head epiphysis displaces posteriorly and inferiorly relative to the femoral neck through the proximal femoral physis. In severe unstable SCFE (where the patient cannot bear weight even before manipulation), acute displacement may tear the retinacular vessels, producing AVN of the displaced epiphysis. AVN rates in unstable SCFE range from 20–50% depending on severity, and it represents one of the most feared complications in pediatric orthopaedics. Stable SCFE treated with in-situ pinning has very low AVN risk.

Surgical complications following acetabular fracture open reduction and internal fixation, proximal femoral osteotomy, or hip arthroscopy can also produce iatrogenic AVN if the retinacular vessels are inadvertently injured during surgical approach or hardware placement.

Back to Table of Contents

Causes — Non-Traumatic

Corticosteroids represent the most common non-traumatic cause of osteonecrosis, accounting for approximately 30–40% of non-traumatic AVN cases. The risk is dose-dependent: cumulative doses exceeding 2 grams of prednisolone equivalent are most strongly associated, but even shorter high-dose courses (e.g., pulse intravenous methylprednisolone) carry risk. The pathogenesis is multifactorial. Steroids induce hyperlipidemia and fat embolism — lipid droplets occlude the small sinusoidal vessels of the femoral head. Steroids also promote adipogenic differentiation of bone marrow mesenchymal stem cells at the expense of osteogenic differentiation, increasing intraosseous fat volume and elevating intraosseous pressure, which impairs venous outflow. Direct cytotoxic effects on osteocytes and impaired angiogenesis complete the picture. Steroid-induced AVN typically affects multiple sites simultaneously: bilateral femoral heads (in up to 70% of steroid cases), humeral heads, femoral condyles, and occasionally talus and carpal bones. Patients receiving steroids for systemic lupus erythematosus (SLE), organ transplantation, inflammatory bowel disease, lymphoma, or asthma are at highest risk.

Alcohol — chronic heavy alcohol consumption is the second most common non-traumatic cause, accounting for roughly 20–25% of cases. The mechanism overlaps with steroid-induced AVN: alcohol causes hyperlipidemia and fat embolism, promotes adipogenesis in bone marrow, and may directly impair bone cell differentiation. Fatty liver and alcohol-induced hyperlipidemia are consistent risk factors. The threshold at which alcohol intake becomes a meaningful risk is generally cited as more than 400 mL of pure alcohol per week (approximately 40 standard drinks), though risk increases with any heavy use. Alcohol-induced AVN also tends to be bilateral and multifocal.

Sickle Cell Disease — vasoocclusion from polymerized sickled hemoglobin in bone marrow sinusoids produces ischemia of the femoral head during or even between crises. AVN affects 25–45% of patients with HbSS disease (homozygous sickle cell anemia), making it one of the most common musculoskeletal complications of the disease. It is bilateral in the majority of affected patients. The humeral head is also commonly affected. Pain crises and AVN can be clinically difficult to distinguish early, and magnetic resonance imaging is required for definitive diagnosis. The presence of AVN strongly predicts progression and eventual THA requirement — often by the patient's 30s or 40s.

Decompression Sickness (Caisson Disease) — rapid ascent from compressed-air environments (deep-sea diving, tunnel and caisson construction work) causes dissolved nitrogen to come out of solution as bubbles in blood and tissues. Bubbles forming within bone marrow sinusoids cause mechanical occlusion of microvascular flow, producing osteonecrosis. Dysbaric osteonecrosis characteristically affects the humeral head, distal femur, and proximal tibia in addition to the femoral head. It is a recognized occupational hazard in commercial divers. Diving and compressed-air work rules use staged decompression protocols to prevent it.

Systemic Lupus Erythematosus (SLE) — AVN complicates systemic lupus erythematosus through two intersecting mechanisms: the immunosuppressive corticosteroids used to treat the disease carry their own independent AVN risk, and antiphospholipid antibodies (present in approximately 30–40% of systemic lupus erythematosus patients) cause a hypercoagulable state with thrombotic occlusion of small vessels including the retinacular arteries. The combination makes systemic lupus erythematosus patients among the highest-risk groups; studies report AVN in 5–15% of systemic lupus erythematosus cohorts, with bilateral disease common.

Gaucher Disease — the most common lysosomal storage disorder, caused by deficiency of glucocerebrosidase, leads to accumulation of glucocerebroside-laden macrophages (Gaucher cells) throughout the reticuloendothelial system, including bone marrow. Massive infiltration of the femoral head marrow by Gaucher cells causes mechanical obstruction of sinusoidal blood flow, producing AVN. Bone crises (acute severe bone pain) and chronic AVN are the dominant musculoskeletal manifestations of Gaucher disease. Enzyme replacement therapy with imiglucerase can reduce but not eliminate AVN risk.

Radiation Therapy — ionizing radiation damages vascular endothelium, producing a progressive obliterative endarteritis that slowly reduces perfusion of the irradiated bone over months to years. Radiation-induced osteonecrosis of the jaw (RONJ) is the best-known example, occurring in patients who receive high-dose radiation to the head and neck for cancer. Femoral head osteonecrosis can also occur in patients irradiated for pelvic malignancies (cervical cancer, prostate cancer, rectal cancer). The onset is typically delayed years after radiation delivery, making the relationship to treatment non-obvious without careful history.

Legg-Calvé-Perthes Disease — idiopathic osteonecrosis of the femoral head in children aged 4–10 years (peak incidence 5–7 years). The cause of the vascular disruption is unknown in most cases; theories include transient synovitis causing intracapsular tamponade of the retinacular vessels, hypercoagulability, and repeated trauma. The disease is self-limiting — the femoral head gradually revascularizes — but the quality of the resulting femoral head shape depends on the patient's age and the extent of head involvement at onset. Younger children and those with small lesions tend to remodel well; older children (>8 years) with large lesions are at risk of residual coxa magna (enlarged flattened head) and premature osteoarthritis (OA).

Other causes include organ transplantation (independent of steroids, possibly from cyclosporine effects on lipid metabolism), hypercoagulable states (Factor V Leiden, antiphospholipid syndrome without systemic lupus erythematosus), pregnancy (rare, possible hormonal and coagulation changes), human immunodeficiency virus (HIV) infection and its treatments, and bisphosphonate-related osteonecrosis of the jaw (MRONJ — a separate entity from AVN, caused by impaired bone remodeling and turnover rather than ischemia). A significant proportion of cases — approximately 20–30% — are classified as idiopathic when no risk factor is identified after thorough evaluation.

Back to Table of Contents

Anatomical Sites

Femoral Head is overwhelmingly the most clinically important and most commonly affected site. The anterosuperior segment of the femoral head — the portion that sustains peak loading during the stance phase of gait — is where the necrotic lesion typically originates, corresponding to the terminal distribution of the posterosuperior retinacular vessels. It is critical to image both hips at initial presentation, because approximately 10% of patients with unilateral symptoms already have contralateral disease on magnetic resonance imaging; this rate climbs to 40–80% in steroid-induced, sickle cell, and alcohol-related cases, where bilateral disease is the rule rather than the exception.

Humeral Head — the second most commonly affected site, particularly in steroid-induced and decompression-sickness AVN. The blood supply to the humeral head, like that of the femoral head, depends on a few vessels (anterior and posterior circumflex humeral arteries) with limited collateralization. Clinical significance is lower than femoral head AVN because the shoulder is a non-weight-bearing joint; patients tolerate early humeral head collapse much better than femoral head collapse. Pain, restricted elevation, and difficulty with overhead activities are the typical complaints. Hemiarthroplasty or reverse total shoulder arthroplasty is reserved for advanced collapse with arthritis.

Femoral Condyles — the distal femur, particularly the medial femoral condyle, is the site of spontaneous osteonecrosis of the knee (SONK). This entity predominantly affects elderly women after trivial trauma or spontaneously, is usually unilateral, and carries a different demographic and clinical profile than classic AVN. MRI shows subchondral T2 signal change with or without a definable necrotic lesion. Secondary osteonecrosis of the femoral condyle also occurs in younger patients on steroids.

Talus — the talar body receives its blood supply primarily through vessels entering the talar neck from the dorsal side; the body itself has a retrograde "end-artery" supply that makes it vulnerable when the neck is fractured. Hawkins type II, III, and IV talar neck fractures (with subtalar and/or tibiotalar dislocation) carry AVN rates of 30–40%, 50–100%, and approaching 100% respectively. The Hawkins sign (subchondral osteopenia on anteroposterior radiograph at 6–8 weeks) indicates intact vascular supply and predicts against AVN. AVN of the talar body ultimately causes tibiotalar arthritis requiring ankle fusion or total ankle arthroplasty.

Carpal Bones — Kienböck's disease is AVN of the lunate, thought to be related to repetitive microtrauma and possibly anatomical variants in the radius-ulna relationship (negative ulnar variance). It typically affects young manual workers (30–40s). Progressive lunate collapse causes carpal instability and wrist arthritis. The scaphoid proximal pole has a similar watershed blood supply problem after scaphoid waist fractures: the proximal fragment receives blood only from the distal fragment, making displaced waist fractures at high risk for proximal pole AVN if not fixed promptly.

Capitellum (Elbow) — Panner's disease is a self-limiting osteochondrosis of the capitellum in children (ages 5–10), analogous to Perthes in the hip. In older adolescent overhead throwing athletes, osteochondritis dissecans (OCD) of the capitellum represents a distinct condition with focal subchondral and cartilage fragmentation.

Vertebral Body — Kümmel's disease is a rare delayed post-traumatic collapse of a vertebral body, typically in the thoracolumbar spine, that presents months to years after an apparently minor compression fracture. Osteonecrosis of the fractured vertebra produces a "vacuum cleft sign" on plain radiographs (intradiscal nitrogen accumulation visible as a radiolucency within the vertebra on extension views or computed tomography (CT)). Vertebroplasty or kyphoplasty can stabilize these lesions.

Back to Table of Contents

Pathological Stages (Ficat Classification)

The Ficat and Arlet classification, originally described in 1964 and modified in 1985, remains the most widely used system in clinical practice for staging femoral head osteonecrosis. It is based primarily on plain radiographic findings, with magnetic resonance imaging incorporated into Stage I characterization in contemporary usage.

Stage I — Silent Ischemia / Bone Marrow Edema. The patient has symptoms (groin pain, reduced range of motion) but plain radiographs are entirely normal. Magnetic resonance imaging at this stage shows a bone marrow edema pattern: diffuse T2 signal increase in the femoral head and neck, representing the hyperemic reactive response at the periphery of the early necrotic zone. A necrotic lesion itself may not yet be clearly defined. Bone scintigraphy may show increased uptake. This is the ideal stage for diagnosis because the femoral head architecture remains completely intact and joint-preserving treatment has the highest success rate. The challenge is that symptoms may be subtle and the condition is not on the radar of primary care physicians; magnetic resonance imaging is often not ordered until symptoms are more established.

Stage II — Sclerosis and Cyst Formation. Plain radiographs now show abnormalities in the femoral head: areas of increased bone density (reactive sclerosis at the border of necrotic and viable bone, as new bone is deposited by the body in an attempt at repair) interspersed with areas of radiolucency (cystic change from resorption of dead trabecular bone). Importantly, the femoral head sphericity and subchondral surface remain completely intact — there is no fracture line and no collapse. Magnetic resonance imaging at this stage often shows the pathognomonic "double line sign" on T2-weighted images: an inner bright band of granulation tissue (vascularized repair tissue at the necrotic margin) surrounded by an outer dark band (sclerotic reactive bone). The double line sign is highly specific for osteonecrosis. Joint-preserving surgery (core decompression ± cell therapy) remains effective at this stage in many patients.

Stage III — Subchondral Fracture and the Crescent Sign. This is the critical watershed stage. The dead subchondral bone, unable to withstand repetitive weight-bearing loads, has developed a fatigue fracture immediately below the articular cartilage. On anteroposterior pelvis radiograph, this appears as a thin curvilinear radiolucency paralleling the articular surface of the femoral head — the crescent sign — representing the cleavage plane between the necrotic subchondral bone and the overlying (still viable, cartilage-covered) articular surface. The articular cartilage itself may still be intact at this point — it receives nutrition from synovial fluid rather than subchondral bone vascularity — but the structural support underneath it is failing. The crescent sign indicates imminent or early collapse. Once identified, joint-preserving procedures are much less reliable, and most patients with Stage III disease ultimately progress to THA, though younger patients may still be offered more heroic joint-preservation attempts (vascularized fibula graft).

Stage IV — Femoral Head Collapse and Secondary Osteoarthritis. The femoral head contour is lost: the articular surface has flattened or fragmented, the head-neck junction may show step-off deformity, and the joint space has narrowed. The acetabulum develops secondary degenerative changes — sclerosis, osteophyte formation, subchondral cysts — reflecting years of contact with a mechanically abnormal femoral head. Pain is constant, range of motion is severely restricted, and the patient's functional impairment is comparable to advanced hip OA. Total hip arthroplasty is the definitive treatment. Plain radiographs in Stage IV may resemble severe OA from any cause, and the history of risk factors (prior steroid use, sickle cell disease, hip fracture) helps establish the diagnosis.

Later systems — the University of Pennsylvania (Steinberg) classification and the 1994 ARCO (Association Research Circulation Osseous) staging system — added magnetic resonance imaging-based quantification of lesion size and specific sub-stages. The Steinberg system, for example, grades Stage II and III into A/B/C sub-stages based on percentage of femoral head involvement (mild <15%, moderate 15–30%, severe >30%). Larger lesions at any stage carry worse prognosis for joint preservation. However, for most clinical communication and operative decision-making, Ficat staging remains the common language.

The 2019 revised ARCO staging. In 2019 a 29-expert ARCO task force revised the 1994 ARCO system into four stages: stage I (X-ray normal, but magnetic resonance imaging or a bone scan abnormal), stage II (X-ray changes such as patchy hardening or cysts in the femoral head, but no fracture under the joint surface and no flattening), stage III (a fracture under the joint surface or in the dead zone on X-ray or computed tomography, split into IIIA with the head pressed in 2 mm or less and IIIB with more than 2 mm), and stage IV (arthritis with narrowing of the joint space). The revised system dropped the older lesion-size sub-grades, and ARCO approved and recommends it as the universal staging system for osteonecrosis of the femoral head.

Back to Table of Contents

Clinical Presentation and Symptoms

Early Stage (Stages I–II): Many patients are asymptomatic at Stage I, and the diagnosis is made incidentally when magnetic resonance imaging is performed for another reason or when screening is performed in high-risk patients (e.g., a patient who has received high-dose steroids for systemic lupus erythematosus). When symptoms are present, they are typically intermittent groin pain — a deep, aching discomfort in the anterior hip region that worsens with activity and settles with rest. The pain may be exacerbated by end-range hip movements, particularly internal rotation and adduction. It is common for patients at this stage to describe weeks or months of mild symptoms that they initially attributed to a pulled muscle, hip flexor strain, or general arthritic change, delaying presentation. Physical examination in early stages is often unremarkable except for mildly reduced and/or painful internal rotation of the hip.

Stage III — Subchondral Fracture: The onset of the crescent sign is frequently associated with a sudden change in symptom pattern. The patient may recall a specific moment when hip pain became acutely worse — corresponding to the subchondral fatigue fracture event. Thereafter, groin pain is more constant, clearly worsened by weight bearing and activity, and present at rest and with sleep. The patient begins to limp (antalgic gait) — unconsciously shortening stance time on the affected side to minimize loading. Pain may radiate from the groin to the anterior thigh and occasionally to the knee, which can mislead clinicians into focusing on the knee before recognizing that hip pathology refers pain distally. Night pain — aching in bed that disturbs sleep — is a classic feature of progressive hip disease at this stage.

Stage IV — Advanced Collapse: Pain becomes constant and severe, limiting even basic activities of daily living. Rest pain, night pain, and inability to sleep on the affected side are universal. Range of motion is severely curtailed in all planes — flexion, abduction, internal and external rotation are all restricted — producing the stiff, shortened gait of advanced hip arthritis. Patients at this stage often need walking aids. The functional picture is clinically indistinguishable from severe OA of any etiology.

Physical Examination Signs: The log roll test — performed with the patient supine and the hip extended, gently rotating the limb medially and laterally — reproduces pain by stressing the hip capsule without loading the articular surface; it is one of the most sensitive but non-specific tests for hip joint pathology. The FADIR maneuver (Flexion, ADduction, Internal Rotation) compresses the anterior hip capsule and consistently reproduces pain in femoral head AVN. Groin tenderness on direct palpation is present in many patients with active synovitis. True leg length discrepancy (shortened limb) may develop late as the femoral head collapses. Trendelenburg sign (contralateral pelvic drop during single-leg stance) reflects abductor muscle weakness or inhibition from pain.

A critically important clinical point: because steroid-induced and sickle cell-related AVN are bilateral in the majority of cases, the other hip is usually evaluated as well — both clinically and with magnetic resonance imaging — even when the patient has symptoms on one side only. Missing an asymptomatic Stage I contralateral lesion delays the opportunity for joint-preserving intervention on that side.

Back to Table of Contents

Diagnosis — Magnetic Resonance Imaging and Imaging

Magnetic resonance imaging (Gold Standard): Magnetic resonance imaging is the most sensitive and specific diagnostic modality for osteonecrosis, particularly at the early stages when plain radiographs are still normal. The sensitivity of magnetic resonance imaging for Stage II–IV disease approaches 100%; for Stage I (bone marrow edema only), sensitivity is approximately 75–90% depending on field strength and imaging protocol. Dedicated hip magnetic resonance imaging protocols use coronal and axial T1-weighted sequences (showing the low-signal necrotic zone as a replacement of normal fatty marrow) and T2-weighted sequences (demonstrating the double line sign and marrow edema).

The double line sign is pathognomonic for osteonecrosis on magnetic resonance imaging: on T2-weighted fat-suppressed images, the margin of the necrotic zone shows an inner band of high signal (granulation tissue with vascular ingrowth) and an outer band of low signal (reactive sclerotic bone). When this sign is present — and it is present in approximately 80% of magnetic resonance imaging-confirmed cases — the diagnosis is established without the need for biopsy. The necrotic zone itself appears as a geographic area of low T1 and variable T2 signal, often with a characteristic "geographic pattern" distinct from the infiltrative patterns of bone marrow infiltration from tumor or infection.

Magnetic resonance imaging also allows precise quantification of lesion size — the percentage of the femoral head volume occupied by the necrotic zone — which is one of the strongest predictors of collapse risk and treatment outcome. Lesions occupying more than 30% of the femoral head volume are at high risk of collapse regardless of treatment, while small lesions (<15%) have reasonable rates of spontaneous quiescence. The Steinberg system and the 1994 ARCO system use this magnetic resonance imaging-derived information for sub-staging; the 2019 ARCO revision does not.

A 2026 systematic review written for the ARCO clinical practice guideline (36 studies, 18 of them on test accuracy) found that magnetic resonance imaging had the highest pooled accuracy for diagnosing osteonecrosis of the femoral head: sensitivity 0.91 (it picked up about 91% of cases) and specificity 0.96 (about 96% of people without the disease were correctly read as negative). Bone marrow edema plus a grade 2 or higher joint effusion on magnetic resonance imaging separated hips with symptoms from hips without, and computed tomography and magnetic resonance imaging detected the fracture under the joint surface better than plain X-ray.

Plain Radiographs: Anteroposterior pelvis and lateral hip films are the standard initial imaging. Normal in Stage I. Stage II shows a heterogeneous mixture of sclerosis and lucency within the femoral head, with preserved articular surface. Stage III is identified by the crescent sign — a thin radiolucent line paralleling the articular surface of the femoral head best seen on the lateral (frog-leg) view; it represents the subchondral fracture cleavage plane. Stage IV shows frank femoral head deformity: flattening of the superior articular surface, sclerosis, osteophyte formation, and joint space narrowing. Both hips are usually imaged on a single anteroposterior pelvis film.

Computed tomography Scanning: computed tomography provides better characterization than plain radiography of early sclerosis patterns, cystic change, and the crescent sign. It is particularly useful for surgical planning (quantifying lesion volume, assessing available bone stock for core decompression drill placement) and for detecting the crescent sign when plain radiographs are equivocal. Computed tomography is less sensitive than magnetic resonance imaging for Stage I disease. Low-dose computed tomography protocols reduce radiation burden while maintaining adequate detail.

Bone Scintigraphy: Technetium-99m bone scan shows a characteristic "cold-in-hot" pattern in established osteonecrosis: a photopenic ("cold") zone corresponding to absent vascularity in the necrotic bone, surrounded by a rim of increased uptake ("hot") reflecting the reactive hyperemia and osteoblastic response at the necrotic margin. This pattern is fairly specific for osteonecrosis when present, but the cold-in-hot appearance requires a lesion large enough to produce a detectable photopenic zone; very early Stage I disease may show only non-specific increased uptake. In current practice, bone scan has been largely replaced by magnetic resonance imaging, which is more sensitive and provides better anatomical detail for surgical planning.

Laboratory Evaluation: No serum biomarker is specific for osteonecrosis. However, laboratory testing is used to identify underlying risk factors and treatable causes: complete blood count (sickle cell screening, anemia), hemoglobin electrophoresis if sickle cell disease is suspected, lipid panel, coagulation studies (antiphospholipid antibodies, Factor V Leiden, protein C/S levels), cortisol if Cushing's is considered, and glucocerebrosidase assay if Gaucher disease is in the differential. Identifying the underlying etiology affects both treatment and the need to image other potentially affected sites.

Back to Table of Contents

Treatment — Early Stage (Pre-Collapse)

The therapeutic goal in pre-collapse osteonecrosis (Ficat Stages I–II) is to prevent subchondral fracture and femoral head collapse, thereby preserving the native hip joint and avoiding or indefinitely deferring total hip arthroplasty. All joint-preservation strategies work from the principle that if the necrotic bone can be supported, revascularized, or if the repair process can be augmented, the femoral head may remain structurally intact. Success depends heavily on lesion size: small lesions have reasonable natural histories, while large lesions (more than 30% of the femoral head) have high collapse rates even with intervention.

Non-Operative Management: Protected weight bearing with crutches is used to reduce mechanical stress on the femoral head during the period of subchondral vulnerability. However, non-weight bearing alone has not been shown to reliably prevent collapse in large lesions; it is an adjunct to, not a substitute for, other treatment. Analgesic management (nonsteroidal anti-inflammatory drugs (NSAIDs), tramadol) addresses pain but has no disease-modifying effect. Removing or reducing the triggering risk factor — tapering steroids where medically possible, stopping alcohol — is standard practice, but it does not reverse established osteonecrosis once it has occurred.

Core Decompression: Core decompression is the most widely used joint-preserving surgical procedure and the standard of care for symptomatic Stages I–II disease. The technique involves drilling one or more channels through the femoral neck cortex into the necrotic zone under fluoroscopic guidance. The original rationale was reduction of elevated intraosseous pressure — osteonecrosis is associated with markedly elevated marrow pressure, which impairs venous outflow and perpetuates the ischemic cycle. Decompression relieves this pressure and, by entering the necrotic zone, also stimulates osteoclast and osteoblast activity and vascular ingrowth. Multiple small-diameter drill holes (2.5–3 mm, "multiple drilling") have largely replaced the original large-diameter (10–13 mm) core in many centers, as multiple drilling may better stimulate repair while reducing the structural weakening of the femoral neck and lowering fracture risk. Clinical results: approximately 65–80% of Stage I hips avoid collapse at 5 years after core decompression; Stage II results are more variable, with approximately 40–60% avoiding collapse depending on lesion size. Stages I and II with small lesions have the best outcomes.

Bisphosphonates: Bisphosphonates inhibit osteoclast-mediated bone resorption. In osteonecrosis, the theoretical benefit is that by slowing the resorption of dead trabecular bone — which weakens the subchondral plate and contributes to collapse — bisphosphonates may allow more time for new bone deposition (the slow anabolic arm of bone remodeling) to replace dead trabeculae and restore structural integrity. This is the rationale for their use rather than a proven effect. Multiple randomized and non-randomized studies have evaluated alendronate (70 mg weekly) and zoledronic acid (5 mg intravenous annually). Bisphosphonates have been associated with a reduction in collapse risk and need for THA compared with placebo or observation alone in early-stage disease. Evidence is strongest for steroid-induced and idiopathic etiologies; less data exists for sickle cell and alcohol-related AVN. A 2016 meta-analysis of 5 randomized controlled trials (329 people, about 921 patient-years of follow-up) found that bisphosphonates did not significantly reduce progression to collapse (risk ratio 0.71, p = 0.23) or the need for total hip arthroplasty (risk ratio 0.61, p = 0.13) compared with placebo, and Harris hip scores did not improve significantly either; the result held when the one trial that used a bisphosphonate alone was left out. The authors concluded that the analysis does not support the use of bisphosphonates for osteonecrosis of the femoral head.

Bone Marrow Aspirate Concentrate (BMAC) / Cell Therapy: Concentrated bone marrow aspirate containing mesenchymal stem cells, osteoprogenitor cells, and growth factors can be injected into the decompression channel at the time of core decompression to augment the biological repair response. An early prospective case series (Hernigou and Beaujean, 2002; 189 hips in 116 patients, followed 5–10 years) reported that only 9 of 145 hips treated before collapse (Stages I–II) needed hip replacement — about 94% avoided THA — compared with 25 of 44 hips treated after collapse; hips that received more progenitor cells did better. Subsequent controlled studies have confirmed a benefit over core decompression alone in early-stage disease, particularly for steroid-induced AVN, where osteoprogenitor cell numbers and function are impaired by the steroids themselves. BMAC-augmented core decompression is increasingly used at high-volume centers, though it remains technically demanding and not universally available.

In a 2018 report from Hernigou's group, 125 adults with steroid-related osteonecrosis and symptoms in both hips (both at stage I or II) had the hip with the larger lesion treated with core decompression plus bone marrow cells and the other hip with core decompression alone. After an average of 25 years (range 20–30), 30 of the cell-treated hips (24%) and 95 of the decompression-only hips (76%) had needed a hip replacement (p < 0.0001). This was a single-centre study from the method's originators.

The 2026 systematic review for the ARCO guideline rated the evidence as very low grade, but it suggested a lower rate of femoral-head collapse when core decompression was combined with bone marrow concentrate than with core decompression alone (pooled relative risk 0.55, 95% CI 0.36–0.83), and with vascularized rather than non-vascularized bone grafts (relative risk 0.35, 95% CI 0.14–0.84), over up to 5 years. No comparative study tested watching-and-waiting against treatment in hips without symptoms, and the reviewers described the evidence for early joint-preserving treatment as limited and mostly observational.

Vascularized Fibular Graft: The free vascularized fibular graft (FVFG) procedure harvests a segment of the fibula with its nutrient blood supply (peroneal artery and vein) and transfers it as a living bone strut into a channel created through the femoral neck and into the necrotic zone. The vascular pedicle is anastomosed to recipient vessels (usually lateral circumflex femoral vessels) to restore blood flow directly to the necrotic area. The fibular strut simultaneously provides structural mechanical support to the subchondral bone. FVFG is the most technically demanding joint-preservation option — it requires microsurgical expertise and a 4–6 hour operative time — and it is reserved for younger patients (<50 years) with large lesions at Stages I–III. Long-term results from specialized centers show 60–80% of Stages I–II hips avoiding THA at 10 years, and even some Stage III hips can be salvaged in young patients. Surgeons weigh FVFG for patients as young as 35, because a THA at that age will almost certainly need revision during their lifetime.

Hyperbaric Oxygen: Hyperbaric oxygen (HBO) therapy increases dissolved oxygen delivery to ischemic tissue and may promote angiogenesis. It is used as adjunctive therapy in decompression sickness-related osteonecrosis with established biological rationale. In non-dysbaric AVN, evidence remains limited. Small studies suggest potential benefit in early-stage disease as an adjunct to core decompression, but HBO is not standard practice outside specialized centers.

Back to Table of Contents

Treatment — Advanced Stage and Total Hip Arthroplasty

Total Hip Arthroplasty (THA) is the definitive treatment for Ficat Stages III–IV osteonecrosis and the ultimate endpoint for the majority of patients with significant femoral head collapse. THA replaces both the femoral head (with a metallic or ceramic femoral stem and head component) and the acetabulum (with a metallic shell and bearing surface — polyethylene, ceramic, or metal) to reconstruct a smooth, pain-free articulation. The results are reliable and durable: THA provides excellent pain relief and functional restoration in 95%+ of patients, and modern implant designs achieve survival rates of 85–95% at 15–20 years in patients with OA. The challenge in osteonecrosis is that patients are younger — often in their 30s and 40s — meaning the hip must last 40 or more years, a substantially greater demand than for a 70-year-old with primary OA.

Cementless press-fit fixation (implant components with porous or hydroxyapatite-coated surfaces that allow bone ingrowth for permanent biologic fixation) is strongly preferred over cemented fixation in young active patients with osteonecrosis, because cemented components inevitably fail at the cement-bone interface with time and high activity levels, and revision of cemented components is more technically demanding. The bearing surface choice in young patients typically favors ceramic-on-ceramic or ceramic-on-highly-crosslinked-polyethylene, both of which demonstrate very low wear rates in laboratory and clinical studies.

The outcomes of THA for osteonecrosis are slightly less favorable than for OA in matched analyses. Younger patients are more active (higher implant loading), and the underlying etiology may compromise bone quality — patients on chronic steroids may have poor bone density and impaired osseointegration; sickle cell disease impairs healing and increases infection risk; alcohol dependence raises perioperative medical risk. Early THA failure rates (5–10 year revision risk) are modestly higher for osteonecrosis than OA in most registry studies. Despite this, THA remains far superior to any alternative for Stage III–IV disease and is the standard treatment when joint preservation has failed or was not feasible.

Hip Resurfacing Arthroplasty preserves the femoral head bone stock by capping the head with a large-diameter metal implant rather than replacing the entire head and neck. In theory this is advantageous in young patients because it conserves bone for future revision surgery and allows near-normal biomechanics. Resurfacing is an option in young patients with Stages III–IV disease whose femoral head, though collapsed in shape, retains adequate bone quality and volume. The major limitation is that osteonecrosis often compromises the very femoral head bone stock on which resurfacing depends for fixation: large necrotic zones, cystic change, and poor bone quality can lead to femoral-component failure. Surgeons plan these operations with pre-operative computed tomography. Large-diameter metal-on-metal resurfacings have largely fallen out of favor due to metal ion concerns; ceramic resurfacing systems are emerging but have limited long-term data.

Proximal Femoral Osteotomy involves cutting and reorienting the proximal femur to rotate the necrotic segment out of the weight-bearing zone of the femoral head, bringing the remaining healthy cartilage and bone into the area of peak loading. Several osteotomy designs have been used — transtrochanteric anterior rotational osteotomy (ARO), valgus-extension osteotomy, varus osteotomy — all with the same geometric goal. Osteotomy preserves the native joint and avoids implant-related complications. Results from Japanese and Korean centers, where the technique has been most refined, show 60–80% hip survival at 10 years in carefully selected Stage III patients. The procedure requires meticulous case selection (adequate healthy bone to rotate into the weight-bearing zone, no acetabular involvement), is technically demanding, and significantly complicates subsequent THA if the osteotomy fails. In Western countries, osteotomy has largely been supplanted by THA due to the excellent THA outcomes, though it remains a consideration in very young patients (under 30) at specialized centers.

Humeral Head Osteonecrosis is typically managed non-operatively in early stages — the shoulder's non-weight-bearing nature allows many patients to remain functional with activity modification and analgesics for years. When pain and functional limitation become intolerable, glenohumeral arthritis from advanced humeral head collapse is treated with hemiarthroplasty (replacement of the humeral head only) or reverse total shoulder arthroplasty (which inverts the normal ball-and-socket geometry to optimize deltoid mechanics). Results are generally good for pain relief.

Perioperative considerations in sickle cell disease are substantial. Surgical teams typically use exchange transfusion to reduce the sickled hemoglobin (HbS) fraction to below 30% before elective THA, to limit sickling under the low-oxygen and circulatory stress of surgery. Temperature control (low body temperature triggers sickling), hydration, and post-operative deep venous thrombosis prevention are part of standard care. Infection risk is higher in sickle cell patients because the spleen no longer works properly (functional asplenia) and bacteria are cleared less well; extended preventive antibiotics and strict sterile technique are used. THA outcomes in sickle cell disease carry higher complication rates but remain far better than leaving painful Stage IV disease untreated.

Back to Table of Contents

Key Research Papers

1. Mont MA, Jones LC, Hungerford DS. Nontraumatic osteonecrosis of the femoral head: ten years later. J Bone Joint Surg Am. 2006;88(5):1117–32. — PubMed PMID: 16651589
2. Steinberg ME, et al. A quantitative system for staging avascular necrosis. J Bone Joint Surg Br. 1995;77(1):34–41. — PubMed PMID: 7822393
3. Lieberman JR, Berry DJ, Mont MA, et al. Osteonecrosis of the hip: management in the 21st century. Instr Course Lect. 2003;52:337–55. — PubMed PMID: 12690862
4. Assouline-Dayan Y, et al. Pathogenesis and natural history of osteonecrosis. Semin Arthritis Rheum. 2002;32(2):94–124. — PubMed PMID: 12430099
5. Jones LC, Hungerford DS. The pathogenesis of osteonecrosis. Instr Course Lect. 2007;56:179–96. — PubMed PMID: 17472306
6. Hernigou P, Beaujean F. Treatment of osteonecrosis with autologous bone marrow grafting. Clin Orthop Relat Res. 2002;(405):14–23. — PubMed PMID: 12461352
7. Malizos KN, Karantanas AH, Varitimidis SE, et al. Osteonecrosis of the femoral head: etiology, imaging and treatment. Eur J Radiol. 2007;63(1):16–28. — PubMed PMID: 17555906
8. Ficat RP. Idiopathic bone necrosis of the femoral head. J Bone Joint Surg Br. 1985;67(1):3–9. — PubMed PMID: 3155745
9. Yuan HF, Guo CA, Yan ZQ. The use of bisphosphonate in the treatment of osteonecrosis of the femoral head: a meta-analysis of randomized control trials. Osteoporos Int. 2016;27(1):295–9. — PubMed PMID: 26370828
10. Yoon BH, Mont MA, Koo KH, et al. The 2019 Revised Version of Association Research Circulation Osseous Staging System of Osteonecrosis of the Femoral Head. J Arthroplasty. 2020;35(4):933–940. — PubMed PMID: 31866252
11. Parikh RR, Mirzaei A, Butler ME, et al. Diagnosis and Treatment of Nontraumatic Osteonecrosis of the Femoral Head: A Systematic Review and Meta-Analyses for the ARCO Clinical Practice Guideline Development Workgroup. Med Sci (Basel). 2026;14(1):107. — PubMed PMID: 41892822
12. Hernigou P, Dubory A, Homma Y, et al. Cell therapy versus simultaneous contralateral decompression in symptomatic corticosteroid osteonecrosis: a thirty year follow-up prospective randomized study of one hundred and twenty five adult patients. Int Orthop. 2018;42(7):1639–1649. — PubMed PMID: 29744647

Back to Table of Contents

PubMed Research Searches

1. Osteonecrosis femoral head treatment core decompression
2. Avascular necrosis corticosteroid mechanism
3. Avascular necrosis magnetic resonance imaging diagnosis staging
4. Osteonecrosis sickle cell disease

Back to Table of Contents

Connections

• Orthopedics
• Hip Fracture
• Paget's Disease of Bone
• Osteomyelitis
• Compartment Syndrome
• Osteoarthritis
• Sickle Cell Disease
• Calcium
• Vitamin D

Back to Table of Contents

Featured Videos