Avascular Necrosis of the Hip

Trauma. A femoral neck fracture displaces the head and stretches or tears the retinacular vessels: 30% go on to AVN. A posterior hip dislocation stretches the branches of the medial circumflex artery: 10-25% develop AVN, more if reduction is delayed. Surgery can do the same damage, through hip pinning, osteotomy or excessive retractor pressure.

Radiation. Vessel fibrosis and obliteration after cancer treatment.

Why it matters. The mechanism is the argument for urgent reduction of a dislocated hip and anatomic reduction of a femoral neck fracture.

Sickle cell disease. Rigid sickled cells occlude the small vessels during a crisis. The disease is often bilateral and multifocal, with the shoulders and knees involved as well as the hips.

Coagulopathy. Thrombophilia (Factor V Leiden, protein C or S deficiency, antithrombin deficiency), SLE, antiphospholipid syndrome and lupus anticoagulant, and hypofibrinolysis. This is why a young patient with apparently idiopathic AVN is screened for a clotting disorder.

Fat emboli. Gaucher disease, where glucocerebroside accumulation causes marrow fat embolisation; pancreatitis, where released lipase causes fat necrosis and emboli; and alcohol, where hepatic dysfunction mobilises fat.

Dysbaric. Nitrogen bubbles in divers, caisson disease.

Steroids. Adipocytes hypertrophy and the marrow fat expands, so intraosseous pressure rises above 30mmHg (normal under 10mmHg), venous outflow is obstructed and the marrow becomes ischaemic. The risk is dose-dependent: over 2000mg cumulative prednisone, or over 20mg/day for more than 3 months.

Alcohol. Fat emboli, adipocyte hypertrophy and hepatic dysfunction together, again dose-dependent, at over 400mL of ethanol per week.

Gaucher disease. Glucocerebroside-laden macrophages expand the marrow space and compress the sinusoidal vessels.

Why it matters. Raised intraosseous pressure is the rationale for core decompression.

Original 1985, modified by Steinberg in 1995. The stage is read from the radiograph and the MRI, and it sets the treatment:

View larger

The University of Pennsylvania system adds lesion size to the Ficat stages, and size is what carries the prognosis:

Steinberg also splits the collapsed head into Stage V (flattening without acetabular involvement) and Stage VI (advanced degenerative change). Its prognostic value: a lesion over 30% of the head (Steinberg C) predicts progression even with core decompression, with a 70-80% failure rate, which is why arthroplasty is considered in a young patient with a large lesion even at Stage II.

The Association Research Circulation Osseous system is the most comprehensive, combining stage, location and extent. Location is by anatomical zone on the head:

• Medial - the medial third, not weight-bearing, the best prognosis
• Central - the middle third, the weight-bearing dome
• Lateral - the lateral third, under maximal load, the worst prognosis

Extent is graded as under 15% (A), 15-30% (B) or over 30% (C) of the head surface.

Why location matters. Lateral lesions collapse faster because they carry the weight-bearing load. That is the logic of the valgus osteotomy, which rotates a lateral necrotic segment out of the weight-bearing zone and therefore needs an intact medial column to rotate into it. Knowing the ARCO zone tells you whether an osteotomy is feasible and how fast the lesion is likely to progress.

The 2019 revised ARCO. This is the version an up-to-date examiner may ask for, and the one now used in research; everything above is the original 1993 system:

• Stage I - radiograph normal, MRI (or bone scan) positive; the only abnormality is on advanced imaging
• Stage II - radiograph abnormal (sclerosis, osteolysis or focal osteoporosis) but no subchondral fracture or flattening; still pre-collapse
• Stage III - subchondral fracture or collapse, subdivided into IIIA early (head depression 2mm or less) and IIIB late (greater than 2mm)
• Stage IV - secondary osteoarthritis: joint-space narrowing, acetabular change, head destruction

What changed and why. Stage 0 was removed, because a normal-imaging stage is unverifiable and clinically unused; each stage was tied to a clear imaging step; and Stage III was subdivided on a 2mm depression threshold that carries prognostic and surgical weight. The consensus deliberately kept the staging simple and recommended that lesion size and location be reported separately rather than embedded as sub-stages, because the older quantitative sub-classes reproduced poorly. That is the direct counterpoint to Steinberg's embedded A/B/C size modifiers.

The goal is to prevent collapse, promote revascularisation and keep the native joint.

Observation. Reserved for very small medial lesions, because observation alone carries an 80% progression rate in Stage I-II. The criteria:

• Small lesion, under 15% of the head (Steinberg A)
• Medial location, out of the weight-bearing zone
• Asymptomatic or minimal symptoms
• Patient declines surgery

The asymptomatic contralateral hip is the case where this matters most, and stage alone will not decide it. A systematic review of 16 studies and 664 untreated asymptomatic hips found 59% progressed to symptoms or collapse, but the risk is not evenly spread: small, medially located lesions collapsed in under 10%, while medium and especially large lesions progressed in a substantial majority. Sickle cell disease carried the highest rate of progression and systemic lupus the most benign course (PMID 20844158; systematic review of prognostic studies, 664 hips, level 4 evidence pooled).

So a Stage I lesion found on the screening MRI of a painless hip is not automatically an indication to drill it. Measure the lesion and locate it before offering surgery: a small medial lesion in a lupus patient is watched, a large lateral lesion in sickle cell disease is the one where prophylactic decompression is defensible. There is still no trial comparing observation with intervention in asymptomatic disease, so this is a risk-stratified judgement rather than an evidence-based rule, and the patient should be told which of the two they are getting. If both hips need decompression, the operations are staggered, for the infection and anaesthetic risk.

What observation involves. Reduce impact loading, with crutches if symptomatic; NSAIDs for pain; bisphosphonates (alendronate) may slow progression, though the later multicentre trial did not confirm it (see Evidence Base); stop alcohol and reduce or taper steroids where possible; MRI every 6 months for 2 years, then annually if stable. The patient is told that 80% progress without surgery: observation is a decision, not benign neglect.

Core decompression. The standard operation for Stage I-II. It lowers intraosseous pressure from over 30mmHg to under 10mmHg, opens a drill track along which granulation tissue and new vessels grow, removes necrotic marrow and relieves venous congestion, and stimulates remodelling. The quoted success rates are Ficat-era conventions: 80% in Stage I, 60-70% in Stage II depending on lesion size, under 30% once the head has collapsed. Modern trial evidence is less certain: the network meta-analysis found no significant difference between joint-preserving surgery and non-operative care for conversion to arthroplasty (see Evidence Base), so the operation is offered with realistic counselling rather than as a cure.

Success means no progression of collapse, resolution of pain and revascularisation on MRI, checked on radiographs at 6 weeks, 3 months and 6 months, then annually for 2 years. Failure declares itself as a crescent sign, increasing pain or progression to Stage III, and at that point the conversation turns to arthroplasty. The technique is in Surgical Technique.

Adding bone graft. For a large Stage II lesion (over 30% of the head, Steinberg C), significant cystic change, or where the weakened subchondral bone needs structural support, a non-vascularised graft goes in through the tract: a cortical strut from the fibula or iliac crest, passed down the drill tract to contact the subchondral plate. A porous tantalum rod is the alternative, inert and promoting bone ingrowth. The strut is quoted at 80-90% success in Stage II B-C against 60% with core decompression alone.

Biologic Augmentation: Bone Marrow Aspirate Concentrate (Cell Therapy)

A major refinement of core decompression is to add bone marrow aspirate concentrate (BMAC) to the drill tract. The rationale is that osteonecrosis, particularly steroid- and alcohol-related disease, may involve a depleted or impaired pool of marrow osteoprogenitor cells; concentrating autologous iliac-crest marrow re-delivers mesenchymal stem cells, osteoprogenitors and growth factors to the necrotic segment to support repair and revascularisation. The underlying cell biology is covered in the dedicated mesenchymal stem cell topic.

Technique. Aspirate marrow from the iliac crest, concentrate it by centrifugation, and deliver the concentrate down the core-decompression channel into the necrotic zone, often combined with small or multiple percutaneous drill channels rather than a single large tract.

Evidence. Core decompression plus bone marrow concentrate reduced femoral-head collapse compared with core decompression alone (relative risk roughly 0.55 over short-term follow-up in the ARCO systematic review), and core decompression plus cell therapy ranked best for limiting radiographic progression in the network meta-analysis (SUCRA 96.4%). The certainty of that evidence remains low, so counsel honestly rather than overpromise: the decompression relieves pressure and opens a channel, the concentrate adds the biology, and the combination appears to slow collapse in pre-collapse disease more than decompression alone.

The question. Age and expectations. Joint preservation after collapse is salvage with poor success rates; a patient under 40 who is motivated may still choose it, and in a patient over 40 arthroplasty is preferred. The same principles explain to a patient why joint preservation may not be offered in Stage III with extensive collapse.

Indications. Stage IV, failed joint preservation, severe pain limiting function, a collapsed head over 2mm, and a patient who accepts an arthroplasty.

Counselling the young patient. Under 40, the conversation is about revision. The conventional counselling figure is 15-20% revision at 10 years, against 5-10% in osteoarthritis. With highly cross-linked polyethylene the matched cohort in the Evidence Base found cumulative revision of 6.6% at 15 years against 4.5% for osteoarthritis (HR 1.8, p=0.09), with reoperation still higher (HR 2.2) and strongly aetiology-dependent: fixation is durable, and the risk remains higher than for osteoarthritis.

Planning. Necrotic bone may be sclerotic or cystic, so bone stock is assessed on CT. Chronic collapse shortens the leg; plan the lengthening but beware the sciatic nerve. Bilateral disease is staged rather than done at one sitting, for the infection and anaesthetic risk. Sickle cell disease needs exchange transfusion before surgery; steroid users need DVT prophylaxis and attention to wound healing. Implant selection and the technical points are in Surgical Technique.

Consent. This is a joint preservation attempt, not a guarantee against collapse. The stage-specific success rates are in Management; the specific risks:

• Femoral neck fracture under 1%, the reason for protected weight-bearing for 6-8 weeks
• Infection under 1%, with prophylactic antibiotics
• Neurovascular injury under 0.5% (sciatic nerve, femoral vessels)
• Progression despite surgery: 20-40% come to arthroplasty in the future
• DVT and PE, the standard surgical risk, with LMWH prophylaxis

Equipment. A radiolucent table (fracture table or standard operating table), C-arm fluoroscopy for AP and lateral views, an 8-10mm cannulated drill system with guidewire, a variable-speed power drill, a standard hip tray, and a bone graft harvesting set if a non-vascularised graft is planned. Optional adjuncts are a tantalum rod, iliac crest bone graft instruments and fibular strut instruments.

Positioning. Supine on a radiolucent table; a fracture table is not essential but helps with traction control. The hip in neutral or slight internal rotation opens the lateral cortex; the knee flexed 10-20 degrees relaxes iliopsoas and improves the fluoroscopy; the contralateral leg is abducted and externally rotated for C-arm access; the arms are across the chest or on arm boards, out of the field. Check the full range of C-arm movement for AP and lateral views before draping.

Preparation. Chlorhexidine or iodine from umbilicus to knee and lateral to the midline; standard hip draping exposing the proximal femur to mid-thigh; a sterile cover on the intensifier. Palpate the greater trochanter and mark the entry point 2-3cm distal to it on the lateral thigh. Confirm an AP pelvis showing both hips and a perfect lateral of the affected hip before the incision.

Who. The selection criteria:

• Age under 40, wanting joint preservation
• Stage IIIA (early collapse under 2mm) or Stage IIB-C (a large pre-collapse lesion)
• A lateral necrotic lesion with an intact medial column, checked on CT or MRI
• An arc of intact bone over 90 degrees, enough viable bone to rotate into the weight-bearing zone
• A motivated patient willing to accept a 6-12 month recovery

Contraindications. Medial column involvement, because there is no intact bone to rotate into the load; collapse over 2mm, which is too late for salvage; an intact arc under 90 degrees; and a patient unwilling to accept complex surgery and a 30-40% failure rate, which persists even with ideal selection.

Planning. CT or MRI to map the extent of the necrosis, confirm the intact medial column and measure the arc of viable bone; calculate the valgus correction from the femoral shaft axis; template against the contralateral normal hip and plan the fixation, blade plate or locking plate.

How it differs from osteoarthritis. The patients are younger (mean age 50 against 70), so more active and with more revision ahead of them; the bone is sclerotic or cystic, so reaming is difficult and subsidence and periprosthetic fracture are risks; the other hip is often diseased, so a second arthroplasty may follow; and the underlying disease (steroids, SLE, sickle cell) raises the risk of infection and wound problems.

Sickle cell disease. Exchange transfusion to an HbS under 30% reduces the risk of crisis. Maintain oxygenation and hydration through the perioperative period, and avoid hypothermia and acidosis, which trigger sickling.

Chronic steroids. Stress-dose steroids (100mg hydrocortisone IV), glucose control because steroids cause hyperglycaemia, and extended DVT prophylaxis with LMWH for 28-35 days.

Anticoagulation. Correct a coagulopathy where possible: warfarin stopped 5 days before with bridging if needed, aspirin and clopidogrel stopped 7 days before.

Bone stock. CT of the femur and acetabulum shows the sclerosis (hard bone, drill bit breakage), the cysts (subsidence, possibly impaction grafting), the leg length discrepancy from chronic collapse, and the stem size and fixation to plan for. Sclerotic bone makes cementless fixation difficult, so consider a cemented stem; cystic bone risks subsidence, so consider a calcar-replacing stem or impaction grafting; osteoporotic bone from steroids favours cement.

Implant selection. Cementless fixation is preferred in the young, for bone ingrowth and to avoid cement debris, with cement where the bone is osteoporotic or poor. A porous-coated cementless cup, highly cross-linked polyethylene to limit wear in an active patient, ceramic-on-ceramic or ceramic-on-polyethylene for low wear, and a head of 36mm or larger to reduce dislocation. Avoid metal-on-metal: the AOANJRR shows higher revision rates.

Approach. Posterior or anterolateral by surgeon preference; avoid the lateral approach for its Trendelenburg risk. After a posterior approach, repair the posterior structures, because these patients have a higher dislocation risk. A drain is not routinely needed.

The acetabulum. Chronic collapse leaves a contracted capsule that needs careful release. Ream to bleeding bone, remembering that sclerotic bone may not bleed readily, and restore the hip centre, which chronic collapse may have migrated superiorly. Aim for 40-45 degrees of inclination and 15-20 degrees of anteversion.

The femur. Broach and ream carefully in sclerotic bone to avoid a fracture, and avoid varus stem positioning, which is common in sclerotic bone and causes thigh pain. Maximise offset restoration for stability and to avoid impingement, and consider an extended trochanteric osteotomy if the sclerosis is severe. Sciatic nerve injury is the standard THA risk (under 1%), raised by the lengthening that chronic collapse demands; lengthening over 4cm risks a palsy.

Stability. Trial extensively, checking flexion, adduction and internal rotation, and consider a constrained liner if the hip is unstable.

Core decompression is a slow recovery, 3-6 months to full activity, and the patient is told so at the outset; the reward is a native joint.

Predictors of poor outcome. Large lateral lesions in young patients carry the highest risk of progression. The combination of a lesion over 30% of the head (Steinberg C), a lateral location under maximal weight-bearing load, age under 40 with its high activity, and Stage II or worse at diagnosis predicts 80-90% progression to collapse despite core decompression; a free vascularised fibular graft is considered in this group if the patient refuses arthroplasty. At the other extreme, a small (under 15%) medial lesion caught at Stage I has 80-90% success with core decompression, which is the argument for early diagnosis.

What joint preservation buys. Native anatomy and proprioception, no implant-related complications (dislocation, wear, loosening), no lifelong activity restriction, and an arthroplasty delayed or avoided along with its future revisions. The price is the failure rate that still ends in arthroplasty, a prolonged recovery of 6-12 months to full activity, the risk of collapse during the healing period, and the fact that it is not suitable for every patient, by age, lesion size or stage.

What arthroplasty buys. Predictable pain relief (95% at 2 years), rapid functional recovery to normal activities at 6 weeks, durability at 10 years, and a definitive answer for failed joint preservation. The price is a lifelong ban on high-impact sport, the revision and dislocation risks of the young AVN patient, and the psychological weight of an artificial joint at that age.

Worldwide Burden and Demographics

Major Guidelines, Side by Side

The take-home for any board: MRI is the universally agreed diagnostic standard, joint preservation is reserved for pre-collapse disease with the strongest support for biologically augmented core decompression and vascularised grafting, and arthroplasty is the definitive option once the head has collapsed.

Arthroplasty Registries and Implant Selection

Major joint registries (NJR for England, Wales, Northern Ireland and the Isle of Man; AJRR in the USA; AOANJRR in Australia; the Swedish SHAR; the Norwegian and New Zealand registries) consistently report that THA performed for osteonecrosis carries a higher revision risk than THA for osteoarthritis, driven largely by younger patient age and higher activity.

Where Management Genuinely Differs