THA Heterotopic Ossification

Heterotopic ossification (HO) is the formation of mature lamellar bone in the soft tissues around the hip after total hip arthroplasty. It represents abnormal differentiation of pluripotent mesenchymal stem cells into osteoblasts in response to local and systemic factors.

Why it matters. Mild HO (Brooker I-II) is often asymptomatic and clinically insignificant. Severe HO (Brooker III-IV) occurs in 5-20% of patients without prophylaxis and causes substantial morbidity: restricted range of motion, especially internal rotation and flexion, pain with movement, and difficulty with sitting and personal hygiene. In extreme cases (Brooker IV) the hip ankyloses completely and needs revision surgery.

What a Brooker III-IV hip costs the patient.

• Loss of 20-50 degrees of flexion and rotation
• Pain in 30-40% of Brooker III patients
• Difficulty with sitting, shoe tying and toileting
• 1-3% require HO excision to restore function, and prophylaxis is cost-effective compared with revision surgery

Who gets it. The incidence runs from 3% to 90% depending on risk factors:

• Primary osteoarthritis: 3-7%
• Ankylosing spondylitis, with its HLA-B27 association: 50%
• Post-traumatic arthritis: 50-90%, especially after acetabular fracture, where clinically significant HO is over-represented
• Previous HO: 90% recurrence without prophylaxis

Other risk factors. Men carry 2-3 times the risk of women, consistently across cohorts, and older patients a slightly higher risk. DISH, hypertrophic osteoarthritis with large osteophytes, spinal cord injury and burns all predispose, as do revision surgery and difficult exposures that damage soft tissue extensively. The bearing surface (ceramic or polyethylene) is not an established driver of HO risk.

Three conditions. HO needs osteogenic precursor cells (mesenchymal stem cells in traumatised tissue), inducing factors (BMPs and inflammatory cytokines) and a permissive environment (adequate vascularisation and local hypoxia). All three must be present. Prophylaxis works by interrupting the cellular differentiation pathway.

The cascade. Heterotopic bone forms in a predictable sequence that begins at the moment of surgical trauma.

Phase 1, 0-48 hours: inflammation and induction. Surgical trauma releases inflammatory mediators (IL-1, IL-6, TNF-alpha), and the tissue becomes hypoxic around a local haematoma. BMPs are released from bone debris and damaged soft tissue, and mesenchymal stem cells are recruited to the site.

Phase 2, 2-7 days: differentiation. Prostaglandin E2 (PGE2) drives the stem cells to differentiate into osteoblasts. This is the critical window in which prophylaxis is effective. NSAIDs block the COX-2 enzyme and so PGE2 synthesis; radiation causes DNA damage in the rapidly dividing mesenchymal cells, inhibiting their proliferation and osteoblast function.

Phase 3, 1-6 weeks: proliferation. Osteoblasts proliferate, produce matrix and form primitive woven bone. Continuing the NSAID prevents ongoing differentiation, but prophylaxis started after the osteoblasts are established is ineffective.

Phase 4, 6-18 months: maturation. Woven bone remodels to mature lamellar bone and vascular channels develop. Bone-scan activity decreases and alkaline phosphatase, the marker of maturation, normalises.

Beyond NSAIDs and radiation. BMP inhibitors, which would block osteogenic signalling in the early post-operative period, are experimental and not yet clinically available.

Brooker. The Brooker classification is the gold standard for grading HO after THA and the most widely used. It is read on the AP pelvis radiograph and grades how far the bone has progressed between the pelvis and the proximal femur.

View larger

View larger

Reliability. The grade is observer-dependent, most of all at the II/III boundary where treatment thresholds are drawn, and it correlates only loosely with symptoms (see the evidence and controversies below). Della Valle and colleagues proposed a simplified three-grade system that is more reproducible for research and audit. It has not shown superior clinical utility and has not displaced Brooker grading in routine practice.

History. Establish what the hip will not do and why it might not:

• Pain: location (anterior groin, lateral) and character (mechanical or inflammatory)
• Restricted movements: flexion, internal rotation, abduction
• Effect on daily life: sitting, shoe tying, toileting, sexual function
• When symptoms developed, early or late
• Previous HO in the contralateral hip or after previous surgery
• Risk factors: ankylosing spondylitis, DISH, post-traumatic arthritis, burns, spinal cord injury

Examination. The end-feel separates bone from soft tissue: HO gives a hard block to motion, sometimes with a firm soft-tissue mass on palpation, especially anterolaterally. Look for an antalgic or Trendelenburg gait, test function (sitting in a chair, tying shoes, crossing the legs), and compare with the other side, since asymmetry suggests HO rather than stiffness.

Measuring range. Record active and passive movement against normal values:

1. Flexion: supine, hip and knee flexed, measuring the angle between thigh and trunk (normal 120 degrees)
2. Internal rotation: prone or sitting, knee flexed 90 degrees, rotating the foot laterally (normal 30-40 degrees)
3. External rotation: prone or sitting, knee flexed 90 degrees, rotating the foot medially (normal 40-50 degrees)
4. Abduction: supine, pelvis stabilised (normal 40-50 degrees)
5. Adduction: supine, crossing the leg over the midline (normal 20-30 degrees)
6. Extension: prone or in the Thomas test position (normal 10-20 degrees)

A Brooker III-IV hip typically flexes to only 70-90 degrees and internally rotates 0-10 degrees, with a hard end-feel and pain on forced motion.

Not all post-THA stiffness is HO. Assess for the alternatives before attributing stiffness to HO alone. Component malposition includes excessive anteversion or retroversion and high offset; abductor deficiency can cause compensatory stiffness; infection brings stiffness with pain and systemic signs; and in instability the patient guards against provocative positions.

In the first few post-operative weeks, developing HO has its own acute presentation, and it is easily mistaken for periprosthetic joint infection (PJI) or deep vein thrombosis (DVT). Those are the two diagnoses that must not be missed, so exclude both before attributing the picture to HO.

Plain radiographs. The AP pelvis is the workhorse for detecting HO, grading it and following it over time. The report should cover:

• Brooker grade (I-IV)
• Location: anterior, lateral, posterior, medial
• Proximity to the components: impinging on the acetabular rim or femoral neck
• Maturation: corticated margins suggest maturity
• Progression against previous radiographs
• The contralateral hip

View larger

When to image. Grading has to wait for the bone to mature. The 6-week film is too early to classify, Brooker classification is made at 3-6 months, and the final grade is recorded on the 6-12 month film.

Other tests. MRI is rarely needed, only when soft-tissue pathology is suspected. Fluoroscopy gives a dynamic assessment of impingement, and serial ALP trends bone-formation activity.

View larger

View larger

Function decides. Management of established HO is guided by whether it actually limits function, not by its radiographic size. When it does, the deciding question is whether the bone is mature, because excising immature bone invites recurrence.

Algorithm

View larger

The complications to counsel for come from the prophylaxis and from excision.

Prophylaxis. NSAID prophylaxis is well tolerated in 90%, and its success depends on compliance for the full 6 weeks, a start within 24 hours and an appropriate dose. Radiation depends on appropriate timing (within 72 hours), the correct dose (700cGy) and an adequate field.

Observation. Mild HO stays stable over time, with minimal symptoms and high satisfaction. Success is predicted by Brooker I-II, an asymptomatic hip, patient reassurance and maintained range of motion.

Excision. Flexion and rotation improve by an average of 30-50 degrees, 70-80% report significant pain relief, and 80% are satisfied when patients are appropriately selected. Haematoma follows in 5%; recurrence, nerve injury and infection rates are in the complications table.

View larger

Five years after excision.

• Range-of-motion gains are maintained in 75%, with some late loss from capsular contracture
• Pain relief is sustained in 80%, up from 70-80% at 1 year
• Recurrence stabilises at 20-30%, most of it within 2 years
• Satisfaction remains high (80%) with appropriate selection
• 5% are revised for recurrent HO, and a second excision has a higher recurrence

Registries and cohorts. Revision specifically for HO is uncommon, since most HO is asymptomatic Brooker I-II.

Who needs prophylaxis. Not every THA does. Prophylaxis reduces radiographic HO, severe grades from roughly 7% to 2%, but the largest RCT (HIPAID, PMID 16885182) found no gain in pain or function and twice the rate of major bleeding. In the standard-risk hip the harm is real and the benefit is a radiograph, so contemporary practice targets prophylaxis to high-risk hips rather than every THA, and risk stratification guides the decision.

Choosing the modality. NSAIDs and radiation are equally effective in the high-risk hip, and combining them adds nothing over either alone. Selective and non-selective NSAIDs are equally effective, so gastrointestinal, renal and cardiovascular risk guides the choice of agent (Migliorini 2022, PMID 35809109), alongside contraindications and local availability. Radiation is reserved for NSAID contraindications, very high-risk hips, or revision with HO excision. Revision THA with HO excision carries the highest recurrence without prophylaxis.

The trauma exception. In patients with concurrent long-bone fractures, radiation is preferred because indomethacin increases non-union risk (Burd 2003, PMID 12892193).

The approach. Direct lateral and anterolateral (transgluteal) exposures are generally reported with higher HO rates than the posterior approach, but no approach, including the muscle-sparing direct anterior, confers immunity. Approach choice and meticulous intra-operative technique are modifiable contributors, worth setting out alongside drug and radiation prophylaxis.

Wash out the local trigger. Pellegrini found pre-operative irradiation as effective as post-operative, which tells us that the osteogenic precursor cells driving HO arise from local operative-field tissue rather than the bloodstream (PMID 8666605). The practical corollary in every THA is to remove the local trigger. These intra-operative measures are pragmatic and supported by observational data rather than level-1 trials.

Whether to give prophylaxis at all in an unselected primary THA This is the central unsettled question, and the honest answer is uncomfortable. Prophylaxis undoubtedly reduces radiographic heterotopic bone, but the largest randomised trial found no improvement in WOMAC pain or function at 6 to 12 months and roughly double the rate of major bleeding, and its authors concluded the data do not support routine prophylaxis in unselected patients. Why it stays unsettled: the trial used a 14-day ibuprofen course and radiographic and WOMAC endpoints rather than range of motion, and it was not powered for the high-risk subgroup - so it argues convincingly against universal prophylaxis without telling us how much benefit the genuinely high-risk hip gains. Practice has therefore moved to risk stratification rather than abandonment.

Does the Brooker grade measure anything the patient feels? Brooker is the universal reporting language, yet the grade correlates poorly with symptoms - many Brooker III hips are asymptomatic and occasional Brooker II hips are stiff and painful. The II/III boundary is also observer-dependent, which matters because that is precisely where most treatment thresholds are drawn. Why it stays unsettled: simplified systems emphasising the clinically meaningful distinction (bridging versus non-bridging bone) perform better on reliability but have not displaced Brooker in registries or trials, so the literature keeps measuring something that is only loosely related to the outcome that matters.

Timing of excision Traditional teaching is to wait for radiographic and biochemical maturity - classically 12 to 18 months, with a normalising alkaline phosphatase and a quiescent bone scan - before excising. Earlier excision is increasingly reported in selected patients with severe functional loss. Why it stays unsettled: the maturity rule derives from an era of higher recurrence without effective prophylaxis, and modern single-dose radiation given around the excision may have changed the calculus; but no randomised comparison of early versus delayed excision exists, and recurrence remains the dominant fear.

How long to continue NSAIDs Regimens of 6 weeks, 2 weeks and even a few days all appear in practice and in trials. Why it stays unsettled: the biological window for stem-cell differentiation is early and probably short, which argues for brief courses, but no trial has directly compared durations with function as the endpoint - so the 6-week convention persists largely by inheritance rather than by evidence.