Avascular Necrosis (Osteonecrosis)

Avascular necrosis (AVN), also called osteonecrosis, is death of bone tissue that follows an interruption to its blood supply. Once perfusion is lost, osteocytes and marrow elements die; the dead bone initially retains its shape and mechanical strength, but over time it loses structural integrity, and — without effective intervention — the overlying joint surface can collapse, leading to secondary osteoarthritis and joint destruction. The femoral head is by far the most commonly affected site. Its blood supply is almost entirely dependent on a small number of terminal, largely retrograde vessels (chiefly the deep branch of the medial femoral circumflex artery, via the lateral epiphyseal vessels), so unlike most bones it has very little collateral circulation to fall back on when flow is disrupted.^[1] Osteonecrosis can also affect the humeral head, the knee (most often the medial femoral condyle), the talus, and the carpal bones (the lunate — Kienböck disease — and the scaphoid), but this chapter focuses on nontraumatic osteonecrosis of the femoral head (ONFH), the presentation a community pharmacist is most likely to encounter and where the international evidence base is best developed.

ONFH classically presents in the 3rd to 5th decade of life — noticeably younger than the typical osteoarthritis population — and this age mismatch is itself a useful clue: new, progressive hip or groin pain in an adult under 50 should raise suspicion for ONFH rather than being assumed to be routine osteoarthritis. This is genuinely a medication-safety topic as much as an orthopedic one. Chronic or high-dose corticosteroid exposure and heavy alcohol use are two of the leading identifiable causes, and a pharmacist is frequently the health professional best positioned to notice cumulative steroid exposure across prescribers, ask about alcohol use, and counsel on minimizing avoidable risk. There is currently no proven disease-modifying pharmacotherapy for established ONFH — management centres on early detection through staging, symptom control while a joint-preservation window still exists, and timely surgical referral, since irreversible collapse can occur within one to a few years of diagnosis once a hip reaches the subchondral-fracture stage.

Typical presentation

The dominant early symptom is groin pain on the affected side, sometimes radiating to the buttock, anterior thigh, or knee — a patient describing only knee pain occasionally has referred hip pathology, so examine the hip when knee findings do not fully explain the symptoms. Pain is typically activity- and weight-bearing-related early in the disease course, then progresses to pain at rest and at night as the disease advances toward and through subchondral collapse. Onset is often insidious over weeks to months, though some patients recall a distinct point at which pain suddenly worsened, corresponding to a subchondral fracture event. On examination, internal rotation and abduction of the hip are typically the first and most affected movements, often before pain is present at rest; a logroll test (passive internal/external rotation of the fully extended leg) reproducing groin pain is a useful bedside finding. An antalgic gait and a positive Trendelenburg sign can develop as disease progresses.

Natural history and bilaterality

Nontraumatic ONFH is frequently bilateral, particularly when the cause is systemic (corticosteroid- or alcohol-associated, sickle cell disease) rather than a single local insult — evaluate, and consider imaging, the contralateral hip even when only one side is symptomatic, since the second hip is often already affected before it becomes painful. Left untreated, disease at an early (pre-collapse) stage frequently progresses to subchondral fracture and femoral head collapse; the time course varies by patient and underlying cause but can occur within one to a few years of diagnosis, which is why staging and referral should not be delayed once ONFH is suspected.

• Chronic or high-dose corticosteroid use — the leading identifiable cause of nontraumatic ONFH, and the one where a pharmacist can have the most direct impact. Risk rises with both the daily dose and the cumulative duration of exposure; a daily dose above roughly 20 mg prednisone-equivalent has been associated with a roughly 9-fold increase in risk compared with lower daily doses, and each additional 10 mg/day has been associated with a further incremental rise in risk.^[2] Most corticosteroid-associated ONFH develops within the first several months of high-dose therapy. This is a genuine, actionable counselling point: review the ongoing need for, and lowest effective dose of, any chronic corticosteroid regimen, and flag cumulative exposure across multiple prescribers (e.g., repeated short "steroid burst" courses, or an oral course layered onto a chronic inhaled, topical, or injected corticosteroid) rather than treating each course in isolation.
• Excessive alcohol use — heavy, sustained alcohol consumption is the second leading identifiable cause, likely via a combination of fat embolism, altered lipid metabolism, and direct effects on bone marrow stromal cells. Risk correlates with the amount and duration of consumption.
• Trauma — femoral neck fracture and traumatic hip dislocation directly disrupt the retinacular vessels supplying the femoral head and are a well-recognized cause of ONFH, mechanistically distinct from the atraumatic causes that make up most of this chapter; risk is highest with a displaced fracture or dislocation and increases with delayed reduction/fixation.
• Sickle cell disease — recurrent vaso-occlusion causes a distinctly high rate of ONFH, which is often bilateral and can involve multiple joints; see the Sickle Cell Disease chapter for disease-specific management. Hydroxyurea therapy for sickle cell disease has not been shown to clearly reduce the risk of developing ONFH.
• Systemic lupus erythematosus and other connective tissue disease — an independent risk factor beyond the corticosteroid exposure often used to treat these conditions; both the underlying disease (via vasculitis/vasculopathy and antiphospholipid antibodies) and its corticosteroid treatment contribute.
• Coagulopathies and thrombophilic states — including antiphospholipid antibody syndrome, protein C or S deficiency, factor V Leiden, and other hypercoagulable or hypofibrinolytic disorders, thought to contribute via intraosseous microvascular thrombosis.
• Solid organ transplantation — a recognized high-risk population owing to the combination of corticosteroids and calcineurin inhibitors used for immunosuppression.
• HIV infection — an independent association exists, compounded when corticosteroids or certain antiretroviral regimens are also used.
• Dysbaric/decompression conditions ("the bends") — nitrogen bubble formation in divers or compressed-air workers can cause osteonecrosis, classically at multiple sites.
• Radiation therapy or chemotherapy involving the affected bone.
• Gaucher disease, chronic pancreatitis, hyperlipidemia, and tobacco use — each reported as a contributing or independent risk factor.
• Idiopathic — in a meaningful proportion of patients, no identifiable cause is found despite a thorough workup; this remains a valid diagnosis after other causes are reasonably excluded, not a sign the workup was incomplete.

History

• Onset, location (groin most classic), and pattern of pain (activity-related, progressing to rest/night pain).
• Full risk-factor review: current and past corticosteroid exposure (dose, duration, route, indication, and all prescribers), alcohol use, prior hip trauma, sickle cell disease or another hemoglobinopathy, connective tissue disease, personal or family history of a clotting disorder, organ transplantation, HIV status, and occupational/diving history.
• Symptoms in the contralateral hip or other joints (shoulder, knee), given the frequency of bilateral/multifocal disease.

Physical examination

• Range of motion — internal rotation and abduction are typically limited and painful earliest.
• A positive logroll test (pain with passive internal/external rotation of the extended leg) is a useful bedside finding.
• Gait assessment for an antalgic pattern or a positive Trendelenburg sign.
• Examine the contralateral hip and other commonly affected joints when a systemic risk factor is present.

Imaging — MRI is the key early test

• MRI is the most sensitive and specific test and the imaging modality of choice, particularly for early (pre-radiographic) disease. The classic finding is a "double line sign" on T2-weighted images — a low-signal-intensity outer rim (the sclerotic reactive interface) alongside a high-signal-intensity inner line (the granulation-tissue interface) — marking the border between necrotic and viable bone. Bone marrow edema and a moderate-to-large joint effusion on MRI are associated with symptomatic, rather than incidental asymptomatic, disease.^[3]
• Plain radiography is appropriate as the initial test but is insensitive early. As disease progresses, look for a crescent sign (a subchondral radiolucent line indicating fracture), sclerosis, cyst formation, femoral head flattening, and — in late disease — joint space narrowing and secondary osteoarthritic change.
• CT detects a subchondral fracture more reliably than plain radiography and can help with surgical planning, but is not more sensitive than MRI for early disease and adds radiation exposure MRI does not.^[3]
• Bone scan (technetium-99m) can be used when MRI is contraindicated (e.g., certain implants) but is less sensitive and specific than MRI and does not provide the anatomic detail needed for staging.

Coronal MRI showing avascular necrosis of the left femoral head. (Jmarchn, CC BY-SA 3.0, Wikimedia Commons)

Staging — the 2019 revised ARCO system

The Association Research Circulation Osseous (ARCO) system is the internationally used staging system for ONFH and is what determines the treatment pathway below. Confirm staging with an orthopedic surgeon before finalizing a treatment-pathway decision, since accurate staging — particularly distinguishing stage IIIA from IIIB — directly affects surgical candidacy.^[4]

Laboratory testing

No laboratory test diagnoses ONFH itself. Order targeted bloodwork when the presentation is atypical (age under 40, bilateral or multifocal disease, or no identifiable corticosteroid/alcohol/trauma history) to look for an underlying secondary cause — hemoglobin electrophoresis for sickle cell disease/trait, a coagulation and thrombophilia workup (antiphospholipid antibodies, protein C/S, factor V Leiden) if a clotting disorder is suspected, a fasting lipid panel, and an autoimmune screen (ANA and related testing) if connective tissue disease is suspected on history or examination.

• Detect ONFH at the earliest possible (pre-collapse) stage to preserve the widest range of treatment options.
• Control pain adequately while a joint-preservation window still exists, without adding unnecessary drug-related risk.
• Identify and modify reversible risk factors — corticosteroid dose/duration and alcohol use in particular.
• Avoid delay in orthopedic referral, since the opportunity for joint-preserving surgery narrows as disease progresses toward and through subchondral collapse.
• Preserve the native joint and functional independence for as long as safely achievable, then transition promptly to definitive surgical management once preservation is no longer realistic.
• Prevent or minimize development of secondary osteoarthritis and its downstream disability.

• Activity and weight-bearing modification — a cane or crutches and avoidance of high-impact loading for symptomatic pre-collapse disease, primarily to manage symptoms while surgical evaluation proceeds; this is not established to reliably change the natural history of the disease on its own.
• Physical therapy — maintains range of motion, strengthens the surrounding hip and core musculature, and provides gait training with an assistive device; useful both before and after any joint-preserving or joint-replacement procedure.
• Corticosteroid stewardship — reviewing and minimizing corticosteroid dose/duration where clinically possible, always in consultation with the prescriber and never by unilateral patient action.
• Alcohol reduction/cessation counselling and, where appropriate, referral to addiction/counselling support.
• Smoking cessation, given its general association with impaired bone healing and vascular disease.
• Patient education on the time-sensitive nature of ONFH — that delaying follow-up imaging or surgical assessment can mean missing the window for a joint-preserving option — is one of the most important non-drug interventions a pharmacist or other health professional can provide.
• Adjunctive physical modalities — extracorporeal shockwave therapy, hyperbaric oxygen, and pulsed electromagnetic field stimulation have each been studied to reduce pain or slow progression, but evidence is limited and inconsistent; none is standard of care and none should delay a surgical referral.

Pregnancy

Breastfeeding

Acetaminophen is compatible with breastfeeding at standard doses. Oral NSAIDs at standard doses for short courses are generally considered compatible; naproxen is often preferred among the NSAIDs given its long track record. Tramadol passes into breast milk in small amounts — use the lowest effective dose for the shortest duration, or select an alternative where reasonable.

Older adults

ONFH in an older adult more often coexists with, or is mistaken for, primary osteoarthritis; lower functional demand and a higher tolerance for arthroplasty relative to hip preservation often shift the practical treatment pathway directly toward joint replacement once collapse has occurred. Use NSAIDs cautiously given the higher prevalence of renal impairment, cardiovascular disease, and polypharmacy; opioids carry a meaningful fall and fracture risk in this population and should be used at the lowest effective dose for the shortest duration.

Sickle cell disease

ONFH is markedly more common in sickle cell disease, is often bilateral, and can involve multiple joints — see the Sickle Cell Disease chapter for disease-specific management, including acute vaso-occlusive pain-crisis treatment, which follows a different pathway from routine ONFH analgesia. Hydroxyurea therapy, while central to reducing other sickle cell complications, has not been shown to clearly reduce the risk of developing ONFH, so its use does not change the need for routine hip surveillance in this population.

Renal impairment

Avoid NSAIDs at an estimated creatinine clearance under 30 mL/min. Tramadol requires dose reduction and a longer dosing interval in significant renal impairment. Solid organ transplant recipients — a recognized high-risk group for ONFH from combined corticosteroid and calcineurin-inhibitor exposure — frequently have renal impairment as well; coordinate analgesic choice with the transplant team.

Hepatic impairment

Acetaminophen is not contraindicated in liver disease but requires closer monitoring and a lower total daily dose ceiling (2-3 g/day); this is a particularly relevant caution in alcohol-associated ONFH, where underlying liver disease is common. NSAIDs should be used cautiously given the added GI bleeding risk from any associated coagulopathy or portal hypertension.

Pediatrics

Osteonecrosis of the femoral head in a child is usually a distinct entity — Legg-Calvé-Perthes disease, an idiopathic childhood osteonecrosis with its own staging systems, natural history, and management approach — rather than the adult risk-factor-driven ONFH covered in this chapter. Refer a child with hip pain and suspected osteonecrosis to pediatric orthopedics rather than applying adult ONFH staging or treatment pathways.

Reassess pain, function, and weight-bearing tolerance at each visit. Repeat MRI (preferred over x-ray for detecting early progression) on a schedule set by the treating orthopedic surgeon — the goal is to catch progression toward subchondral fracture while a joint-preserving option is still realistic, not to wait until pain forces the issue. In any patient with ONFH or a major risk factor for it (chronic corticosteroid therapy, sickle cell disease, heavy alcohol use), assess and, where appropriate, image the contralateral hip even if it remains asymptomatic, given how often bilateral disease is present at diagnosis or develops afterward. At every encounter, review cumulative corticosteroid exposure across all prescribers and routes, and flag opportunities to reduce dose or duration without compromising the condition being treated. Reinforce alcohol-reduction counselling and reassess use at follow-up. For a patient on chronic NSAID therapy, periodically check blood pressure, renal function, and ask directly about GI symptoms, as for osteoarthritis. Track functional status and mobility-aid needs over time so that surgical referral timing reflects genuine disease trajectory rather than only the point at which a patient can no longer cope.

1. Confirm the diagnosis and stage are current — if the clinical picture has changed since the last imaging (new rest pain, sudden worsening), repeat MRI or x-ray rather than assuming progression without confirming it.
2. If pain persists despite optimized analgesia and activity modification and the disease remains pre-collapse (ARCO stage 0-II), refer or re-refer for joint-preserving surgical evaluation rather than escalating opioid therapy indefinitely.
3. If subchondral fracture or collapse has occurred (ARCO stage III-IV), or symptoms progress despite a prior joint-preserving procedure, refer for total hip arthroplasty evaluation.
4. Reassess and re-address modifiable risk factors at every escalation step — confirm whether corticosteroid dose has genuinely been minimized with the prescriber, and whether alcohol use has changed.
5. If bilateral or multifocal disease is discovered, ensure every affected or at-risk joint is staged and followed, not only the one that is currently symptomatic.
6. If no clear cause has been identified and the presentation is atypical (young patient, bilateral disease, no corticosteroid/alcohol/trauma history), confirm that a secondary-cause workup (coagulopathy, hemoglobinopathy, connective tissue disease) has actually been completed rather than assuming idiopathic disease by default.

• Assuming a normal plain x-ray rules out ONFH — early-stage disease is frequently radiographically silent; order an MRI when clinical suspicion is genuine.
• Treating new hip pain in a patient under 50 as routine osteoarthritis without considering ONFH, especially when a risk factor (corticosteroid use, heavy alcohol use, sickle cell disease, prior trauma) is present.
• Presenting bisphosphonate therapy to a patient as a proven way to prevent collapse or avoid surgery — randomized evidence does not support a meaningful clinical benefit, and this should not be overstated.
• Allowing chronic or repeated corticosteroid courses to continue across multiple prescribers without anyone reviewing the cumulative exposure or minimization opportunities.
• Advising a patient to stop a prescribed corticosteroid abruptly out of concern for osteonecrosis risk — this can precipitate adrenal insufficiency; any dose reduction must be physician-directed and tapered.
• Delaying orthopedic referral while continuing to escalate analgesia in a patient whose disease is still pre-collapse and who could benefit from a joint-preserving procedure — the treatment window narrows with time regardless of how well pain is controlled.
• Evaluating only the symptomatic hip in a patient with a systemic risk factor, and missing asymptomatic contralateral disease.
• Not asking about alcohol use as a contributing and modifiable factor.
• Missing an underlying secondary cause — sickle cell disease, a coagulopathy, or connective tissue disease — in an atypical presentation, and mislabeling it idiopathic without completing a reasonable workup.
• Confusing adult ONFH with Legg-Calvé-Perthes disease in a pediatric patient — these are distinct entities with different management, and a child with suspected osteonecrosis should go to pediatric orthopedics.

How to use this table: doses are adult doses and require patient-specific confirmation, including renal/hepatic adjustment and individual GI/cardiovascular risk stratification (see Choose Initial Treatment). These agents manage symptoms only — none is a proven disease-modifying therapy for osteonecrosis. Select any drug name to open its complete Pharmasist Drug Database page.

Oral NSAIDs

For mild-to-moderate pain, selected and dosed by individual GI/cardiovascular risk exactly as for osteoarthritis.

Simple analgesic

Opioids — short-term/perioperative use only

Reserve for severe pain not controlled by NSAID/acetaminophen, most often while awaiting or recovering from surgery — not a routine or long-term option, and not a substitute for timely surgical referral.

Investigational agent — no proven disease-modifying benefit

Included here for completeness because patients and prescribers sometimes ask about it, not as a recommended therapy. A meta-analysis of randomized controlled trials found no significant benefit of oral bisphosphonate therapy over placebo/control for preventing progression to total hip arthroplasty in ONFH.^[5] Potential adverse effects (atypical femoral fracture, osteonecrosis of the jaw) are a real consideration with long-term bisphosphonate use generally, which further limits any favourable risk-benefit case for using it here without proven benefit.