Heterotopic Ossification

What it is. Heterotopic ossification (HO) is the formation of mature lamellar bone in extra-skeletal soft tissues, typically muscle and periarticular connective tissue. It is not dystrophic calcification, and it is not myositis ossificans traumatica, which involves muscle injury specifically.

How common. The figure depends on the setting, and males are more commonly affected:

• Total hip arthroplasty: 10-50% radiographic, 3-5% symptomatic
• Acetabular fractures: 20-40% overall
• Elbow surgery: 3-20%
• Spinal cord injury: 20-30%
• Traumatic brain injury: 10-20%

After an acetabular fracture, the approach is the variable. The surgical approach is the dominant variable, not the fracture pattern. The Kocher-Langenbeck and extended iliofemoral approaches strip and devitalise the abductors and carry the great majority of the risk; the ilioinguinal approach does not violate that muscle envelope and carries very little. Debriding necrotic gluteus minimus at closure is the one intraoperative step shown to reduce it.

In the randomised acetabular trial in the evidence section, HO prevalence was not associated with fracture type or with posterior dislocation. Risk-stratify on the exposure you used, not on the injury film.

Who is at high risk. Previous HO is the strongest predictor, with a 50% recurrence rate. The other high-risk groups:

• Traumatic brain injury
• Spinal cord injury
• Burns greater than 20% TBSA
• Ankylosing spondylitis, a systemic tendency to ossification
• Diffuse idiopathic skeletal hyperostosis (DISH)
• Hypertrophic osteoarthritis
• Polytrauma

A head-injured patient with an acetabular fracture is at very high risk; the HO is often bilateral and may be severe.

View larger

Three things have to be present. HO requires osteogenic precursor cells, an inducing signal and a permissive environment. Understanding the triad is what guides prevention.

The cells. Mesenchymal stem cells in muscle and connective tissue, circulating osteoprogenitor cells, local fibroblasts with osteogenic potential and endothelial cells via endothelial-to-mesenchymal transition. Under the right signals they differentiate into osteoblasts and lay down osteoid, which then mineralises to form mature bone.

The signal. Bone morphogenetic proteins are key: BMP-2, BMP-4 and BMP-7 are released from damaged tissue, and hypoxia upregulates their expression. Inflammatory cytokines (IL-6, TGF-beta) promote differentiation, and prostaglandins stimulate osteogenesis. That last step is why NSAIDs prevent HO: they inhibit prostaglandin synthesis and so block one of the key osteoinductive signals.

View larger

The environment. Tissue hypoxia from trauma or surgery, a haematoma to provide scaffolding, denervation, immobilisation and local inflammation. Burns cause a systemic inflammatory response.

Why head injury in particular. Four mechanisms operate at once: loss of inhibitory neural signals to bone turnover, release of brain-derived osteogenic factors into the circulation, a systemic inflammatory response and prolonged immobilisation. Spinal cord injury shares the loss of inhibitory neural signals and the systemic release of osteogenic factors, which is why both carry such a high incidence. Head-injured patients can develop HO at sites remote from any surgery or local trauma.

View larger

Fibrodysplasia ossificans progressiva. FOP is a rare autosomal-dominant disorder caused by a gain-of-function mutation in ACVR1/ALK2, a BMP type I receptor, in which connective tissue progressively turns into bone. The pathognomonic clue is present at birth, before any ossification: a congenital malformation of the great toes (short, valgus, monophalangic hallux). Affected children then suffer episodic painful soft-tissue "flares" that ossify, progressing axial to appendicular, proximal to distal and cranial to caudal, and eventually causing widespread ankylosis.

The cardinal rule is that trauma triggers ossification, so surgery and biopsy are contraindicated. Each of these provokes catastrophic flares:

• Biopsy
• Intramuscular injections, including some immunisations
• Dental blocks
• Surgical excision

Recognise FOP clinically, great-toe sign plus flares, avoid the knife, and manage with flare protection and emerging ACVR1-targeted therapy such as palovarotene. The NSAID, radiation and excision pathway used for acquired HO is never applied to it.

Brooker (1973) is the most commonly used system and is a classification of hip HO. Grades I-II are usually observed; grades III-IV typically require surgical consideration, which is why "Brooker III-IV" is shorthand for clinically significant disease.

View larger

Della Valle (2002) is the alternative that grades functional impact: Type A, HO does not affect function; Type B, HO affects function; Type C, HO prevents function (ankylosis). It is more clinically relevant but less commonly used than Brooker.

Hastings-Graham is the elbow classification, and it separates the two arcs of elbow motion.

History. The questions that change the plan:

• Previous surgery or trauma to the area
• Risk factors: head injury, spinal cord injury, burns
• Timeline of symptom development
• Previous HO at any site
• Current medications, especially NSAIDs
• Functional limitations

Examination. Range of motion is the key functional assessment. In the early, active phase the signs are local warmth and swelling; a palpable firm mass is a late finding. Record pain on movement, especially at end-range, any skin change over the lesion, and the neurovascular status (compression is rare).

Natural history. The lesion passes through recognisable phases, and the imaging and biochemistry follow them.

Early HO mimics other things. In the inflammatory phase the X-ray is usually still negative, so the picture of swelling, warmth and a mass overlaps with DVT, infection, stress fracture and tumour. Bone scan and serial X-rays help differentiate them. Do not miss the DVT: consider duplex if the limb is a lower limb.

Myositis ossificans traumatica and the zonal phenomenon. This is the post-traumatic, muscle-specific form of heterotopic ossification, classically in the quadriceps or brachialis after a contusion or dislocation. Its defining feature is the zonal phenomenon: the lesion matures from the outside in, so a mature, well-organised, mineralised bony rim surrounds an immature, cellular centre. That is the exact reverse of a soft-tissue sarcoma, which is most cellular and aggressive at its periphery.

On radiographs, peripheral mineralisation appears at about 3 to 6 weeks and a corticated rim defines the lesion by 6 to 8 weeks; serial films showing maturing, sharply marginated peripheral bone are reassuring. The trap is the early biopsy: before zoning, the tissue is hypercellular, mitotically active osteoid that is histologically indistinguishable from osteosarcoma. Wait, image serially, and biopsy only if the appearance is atypical or progressive.

View larger

Imaging. Plain radiographs are the primary study and the basis of classification, but they are usually negative early, with calcification appearing from 3-6 weeks. The bone scan may be positive earlier, at 1-2 weeks, and its role at the other end of the process is to judge activity before surgery. CT is the pre-operative planning study once the HO is mature. MRI is rarely needed; it shows soft-tissue oedema and early changes.

View larger

Blood tests. Alkaline phosphatase rises while HO is active, to 2-3 times normal, and returns to normal as it matures, so it is followed as a marker of activity. CRP and ESR may be elevated in active HO, and their real job is to rule out infection. Calcium and phosphate are usually normal and are checked to exclude a metabolic disorder.

Judging maturity. Maturity rests on a stable, corticated, trabeculated mass on serial radiographs with a quiescent bone scan. Before excision, confirm all four:

1. Normal ALP for at least 2-3 months
2. Cold or inactive bone scan, with no increased uptake or minimal activity
3. Cortical margins on X-ray
4. Stable radiographic appearance on serial imaging

The conventional timing rule is to wait until ALP has returned to baseline and the bone scan is cold. Keep ALP in its place, though: it is a supporting number only, since it rises with any healing fracture and with liver disease, and it should never be the criterion on its own. Operating on immature HO dramatically increases recurrence.

Algorithm

View larger

The decision. Three questions settle it: what grade, how much function is lost, and whether the HO is mature. Brooker I-II or minimally limiting disease is observed; grade III-IV with functional limitation is considered for excision, but only once it has matured.

Conservative management is for Brooker grade I-II, minimal functional limitation, and HO that is still maturing. It consists of pain management, physiotherapy to maintain range of motion (gentle, not forced), serial X-rays, and a wait-and-watch stance, since some HO resorbs partially. Asymptomatic Brooker I-II needs no further prophylaxis if primary prevention was given, and rarely progresses if stable at 6 months.

Aggressive physiotherapy does not cause HO; that is an old myth. Forced range of motion through established HO, however, can cause fracture or bleeding.

Surgical excision is indicated for:

• Brooker grade III-IV with functional limitation
• Established maturity (ALP normal, bone scan cold)
• Failed conservative management
• Ankylosis preventing function

Timing. Wait a minimum of 6 months after HO formation, and ideally 12 months, especially for neurogenic HO after head injury or spinal cord injury, confirming maturity with ALP and bone scan. There is no single waiting time because the cause sets the pace: Garland's 1991 review recommended traumatic HO at 6 months, spinal cord injury at 1 year and traumatic brain injury at 1.5 years. Establish the cause before quoting a date to the patient.

Know what kind of rule that is. It is an expert recommendation from a review article, never tested against earlier or later excision, so treat it as the accepted convention rather than a validated threshold.

Principles of the operation. Complete excision of the HO bridge, capsule release as needed, gentle manipulation after excision, meticulous haemostasis (haematoma promotes recurrence), a drain, and immediate prophylaxis post-operatively.

Recurrence. Quote it by site rather than using one number. The conventional hip-derived figures are 20-30% with prophylaxis and over 50% without. The largest pooled series of elbow excisions (Lee, 384 patients across 24 studies) reports 11.9%, so do not carry the hip numbers to the elbow. Whichever site, recurrence is the commonest complication of excision.

View larger

Hip. Use the same approach as the index surgery if possible; lateral or anterolateral serves most THA HO, and circumferential HO may need dual approaches. Identify and protect the neurovascular structures first. The sciatic nerve is at risk posteriorly, especially with posterior or circumferential HO, and may be encased in bone, so find it early, trace it proximally from known anatomy and protect it throughout; neuromonitoring is advisable for complex cases. Excise the HO to restore the bone-capsule plane, release the contracted capsule, and ensure full range of motion before closure.

View larger

Elbow. The approach follows the HO: lateral for lateral HO and radiocapitellar access, medial for medial HO (protect the ulnar nerve), posterior for posterior or triceps HO, and combined approaches are often required. Transpose the ulnar nerve if it is at risk, excise the anterior capsule separately if it is contracted, and a hinged external fixator may help maintain range of motion.

The complications divide into those of the HO itself and those of removing it.

Surveillance. Serial radiographs after excision are how recurrence is found. The pair below shows a knee early after excision and again two months later.

View larger

View larger

The aim is to keep what was won in theatre. Prophylaxis starts immediately and runs its full course, motion starts on day one, and the radiographs continue to 12 months; most recurrence is evident by 6 months.

Continuous passive motion may help maintain the range gained after excision, especially at the elbow. The evidence is mixed but it is commonly used, started immediately post-operatively, with the goal of holding the range achieved at surgery.

View larger

The prognosis is set by the cause of the HO, the joint's starting point, and how the excision was timed and protected.

Good prognosis

• Isolated HO without a neurological cause
• Good range of motion before the HO developed
• Mature HO at the time of excision
• Prophylaxis given post-excision
• A patient motivated for rehabilitation

Poor prognosis

• Neurogenic HO (TBI, SCI)
• Previous recurrence
• Circumferential HO
• Poor range of motion before the HO
• Non-compliance with physiotherapy

View larger

What prophylaxis actually prevents. Prophylaxis reduces severe (Brooker III-IV) HO by roughly 70-80% in high-risk surgery; in the acetabular-fracture trial in the evidence section, grade III-IV fell from 38% untreated to 7% treated. It does far less to any HO. In the HIPAID trial the relative risk for any ectopic bone was 0.69, a reduction of about a third, and every one of Burd's sixteen untreated patients formed some HO regardless.

Quote the endpoint with the number. "Reduces HO by 70-80%" is true of the ossification that limits movement and false of ossification in general. Note too what the reduction bought: HIPAID found no improvement in pain or function despite the radiographic benefit, so prophylaxis is justified by the severe grades it prevents in patients at real risk, not by a headline percentage applied to everyone.

NSAIDs. They inhibit prostaglandin synthesis and so block an osteogenic signal. Begin within 24-48 hours of surgery, since a delayed start reduces efficacy, and run the course for 6 weeks, the minimum duration for NSAID prophylaxis.

The contraindications are a history of GI ulcer or bleed, renal impairment, anticoagulant use, and concern about fracture healing.

Radiation. It kills the osteoprogenitor cells in the irradiated field. A single dose of 700-800 cGy within 24-72 hours of surgery is the most common protocol and is as effective as fractionation.

Its advantages over NSAIDs are that it has no systemic effects, no GI risk and no effect on fracture healing. Against it are the need for radiotherapy access, cost, a theoretical malignancy risk (minimal), and the prosthesis, which must be shielded.

Who should get it. High-risk patients who should receive prophylaxis:

• Previous HO at any site
• Ankylosing spondylitis
• DISH
• Traumatic brain injury
• Spinal cord injury
• Burns greater than 20% TBSA
• Hypertrophic osteoarthritis
• Extensive surgical approach, since extensive soft-tissue dissection increases risk
• Acetabular fracture with TBI, where prophylaxis is mandatory

Routine THA is the controversy: some give prophylaxis to all, others only to the high-risk. The evidence supports prophylaxis in high-risk patients.

Choosing between them. NSAIDs are the most common choice, convenient and cheap, and are used when there is no contraindication. Radiation is preferred where NSAIDs are contraindicated (GI, renal), where fracture healing is a concern, and where the anatomy is difficult to shield.

On efficacy the two are roughly equivalent, with a tilt toward radiation. The randomised acetabular trial found no significant difference (grade III-IV 4% radiation versus 11% indomethacin, p=0.22), and the meta-analysis found radiation better in relative terms (RR 0.42) but by an absolute margin of only about 1%, too small to justify radiotherapy logistics on efficacy alone. Choose on the contraindication, not the percentage.

Where the two genuinely part company is the fracture: indomethacin trebled long-bone nonunion in patients with concomitant fractures, and radiation does not.

Side-by-Side Guidance on Prophylaxis