AUTHOR AND EDITOR INFORMATION

Section 1 of 10  

• Authors and Editors
• Introduction
• Clinical
• Differentials
• Workup
• Treatment
• Medication
• Follow-up
• Miscellaneous
• References

INTRODUCTION

Section 2 of 10  

• Authors and Editors
• Introduction
• Clinical
• Differentials
• Workup
• Treatment
• Medication
• Follow-up
• Miscellaneous
• References

Background

The term heterotopic ossification (HO) describes bone formation at an abnormal anatomical site, usually in soft tissue. HO can be classified into the following 3 types:

• Myositis ossificans progressiva (fibrodysplasia ossificans progressiva) - This disorder is among the rarest genetic conditions, with an incidence of 1 case per 2 million persons. Transmission is autosomal dominant with variable expression. The condition is characterized by (a) recurrent, painful soft-tissue swelling that leads to HO and (b) congenital malformation of the great toe. There is no treatment for this form of HO. Limited benefits have been reported using corticosteroids and etidronate. Most patients die early from restricted lung disease and pneumonia; however some patients live productive lives.¹
• Traumatic myositis ossificans - In this condition, a painful area develops in muscle or soft tissue following a single blow to the area, a muscle tear, or repeated minor trauma. The painful area gradually develops masses with a cartilaginous consistency; within 4-7 weeks, a solid mass of bone can be felt. Common sites include the pectoralis major, the biceps, and thigh muscles. A nontraumatic type of myositis ossificans also may exist.
• Neurogenic heterotopic ossification - This condition is the one that comes to mind when the generic phrase heterotopic ossification is used. This type of HO is the subject of this article. The various terms mentioned at the outset all refer to this type of HO.

Pathophysiology

Despite many investigations, the etiology and pathogenesis of neurogenic heterotopic ossification remain unknown.³ An extensive review of the problem in 1973 by Rossier and colleagues attempted to address the question of pathogenesis by investigating the following parameters⁴:

• Radiographs
• Lower limb angiography
• Venous and arterial blood gas analyses
• Serial serum calcium
• Phosphorus
• Creatine kinase (CK) and alkaline phosphatase (AlkP)
• Urine calcium and hydroxyproline
• Skin temperature
• Bone scans
• Biopsy

None of these observations explain the factors responsible for the development of HO. Although the etiology of HO remains unknown, clinical and experimental evidence supports the hypothesis that trauma is one of the most important initiating factors. In the studies in which HO was induced experimentally, 2 factors were found to be prerequisites for ectopic ossification: (1) traumatic ischemic degeneration of involved muscle and (2) tissue expression of bone morphogenic proteins (BMPs).

It also has been shown that expression of many genes, including BMP, is regulated by mechanical stress. The target cells in the muscle for BMP are mesenchymal stem cells, also called satellite cells. These cells are precursors capable of differentiating into many cell types, including osteoblasts. Thus, BMP may play a role as a paracrine factor in the differentiation of satellite cells into bone-forming cells.

Most likely, other factors also are involved in the etiology of HO. Studies have shown that AlkP may have an important role in ectopic calcification and ossification of soft tissues. The major role of AlkP in soft tissue is to remove inhibitors of mineralization. An increased expression of AlkP was found in vascular smooth muscle cells in the presence of macrophages and inflammatory cytokines. These observations may have clinical importance, because inflammation and trauma have long been suggested by many investigators as possible etiologic factors.

Clinically, muscle trauma has been reported as a cause of HO after SCI by numerous investigators, including Bodley and colleagues,⁵ as well as Snoecx and co-investigators.⁶ The types of muscle trauma proposed as initiating HO are muscle tears, ruptures, edema, and bleeding.

It has also been suggested that factors such as intensive rehabilitation, transfer activities, and repeated minor trauma during activities of daily living can cause superimposed mechanical stress and initiate HO. The hypothesis that trauma is an important factor in HO formation after SCI also has been documented by ultrasonographic and histologic studies. Various degrees of muscular damage with evidence of tissue bleeding have been found in the early stage of HO.

During the formation of HO, initially immature connective tissue, fibroblasts, ground substance, and collagen fibers are seen. Eventually, usually within 7-14 days, osteoblasts are noted, located irregularly in osteoid. New bone formation may start in multiple foci within osteoid. As mineralization progresses, amorphous calcium phosphate is gradually replaced by hydroxyapatite crystals. Commonly, after approximately 6 months, the appearance of true bone is noted. Rossier noted that after approximately 30 months, the pattern in HO approached that of normal young adult bone.⁴ Anatomically, HO is always extra-articular, but it may attach to the joint capsule without disrupting it. Occasionally, HO may attach to the cortex of adjacent bone, with or without cortical disruption.

Frequency

United States

The reported incidence of HO following SCI varies greatly from study to study. Incidence varies from a low of 3.4% to a high of 47% reported by Hassard, who found HO around the hips of 62 of 131 patients with SCI who were admitted to the Hot Springs Rehabilitation Center over a 2-year period.⁷ Most studies cite a range between these 2 extremes. Peak incidence is noted from 4-12 weeks postinjury and can occur up to 5 months following trauma. Later onset has been reported but is very rare.

In 1954, Irving and LeBrun first documented HO in patients with hemiplegia. Roberts reported 6 cases of HO following intracranial lesions (3 traumatic, 2 vascular, and 1 neoplastic) in 1968, and subsequent studies cited the incidence of clinically significant HO following severe closed head injury (CHI) as being 11-76%. Incidence of HO following other neurologic disorders has not been delineated yet, but it appears to be lower than the incidence following SCI or head injury.

The following reasons may be postulated for the large variability in incidence seen in different studies:

• Different authors use different criteria to define HO. For example, using a radiologic screening survey yields higher incidence figures than does reporting only clinically significant HO detected on physical examination.
• Although neurogenic heterotopic ossification (NHO) generally occurs periarticularly (as noted in the European term paraosteoarthropathy), some authors include patients with calcification or bone formation, possibly secondary to other causes (eg, decubitus ulcers, septic arthritis, trauma, surgery), in their studies.
• Various authors study different populations (eg, early vs late cord injured patients).

Mortality/Morbidity

Approximately 10-35% of all patients with HO secondary to SCI have significant reduction of range of motion (ROM) at the affected joint or joints.³ Wharton and Morgan found that 3% of patients with SCI have an ankylosed joint caused by HO. Effects on activities of daily living (ADL) and functional mobility (eg, transfers) are not difficult to imagine. In addition, abnormal weight distribution may lead to increased frequency of decubitus ulceration, as Hassard noted in 1975.⁷

Race

No known correlation exists between race and incidence of HO.

Sex

No known correlation exists between sex and incidence of HO.

Age

Age has no significant correlation with HO formation, although the condition is somewhat less frequent in pediatric and geriatric patients with SCI.

CLINICAL

Section 3 of 10  

• Authors and Editors
• Introduction
• Clinical
• Differentials
• Workup
• Treatment
• Medication
• Follow-up
• Miscellaneous
• References

History

• The onset of HO usually is 1-4 months after injury in SCI patients, although it may occur as early as 19 days or as late as 1 year following injury.
• The condition may occur later with other precipitating circumstances (eg, fracture, surgery, severe systemic illness).^{8, 9}
• Not uncommonly, incidental HO that was not noted clinically may be detected much later on radiographs.
• HO always occurs below the level of injury in SCI patients, and most authors agree that there is no relation to presence or absence of spasticity in SCI patients.
• HO tends to occur more frequently with complete injuries.
• In SCI patients with HO, the hips are most commonly involved.
• • At the hip, the flexors and abductors tend to be involved more frequently than are the extensors or adductors.
  • At the knee, the medial aspect is most commonly affected by HO.
  • Shoulders and elbows are the most commonly affected upper extremity joints.
  • One report in the literature notes involvement of the metacarpophalangeal joints of the hand.
  • The lumbar paravertebral region also has been mentioned as an infrequent site.
• In patients who have sustained head injury or stroke, the story is a bit different. HO almost always occurs on the affected side, and most authors have noted that HO is more frequent in patients with spasticity than in those without it.
• • Garland and colleagues studied 496 patients with severe head injuries.¹⁰ Clinically significant HO, causing pain and decreased ROM, was noted in 100 joints in 57 patients. Of the 100 involved joints, 89 were in spastic extremities. The frequency of involvement of different joints was slightly different than it was in patients with SCI; the hips were most commonly involved (44), followed by the shoulders (27) and elbows (26). HO was detected in only 3 knee joints.
  • Spielman also looked at the occurrence of HO in patients with head injuries. In that study, the inclusion criteria were (1) initial Glasgow Coma Scale score of 8 or less and (2) coma lasting more than 2 weeks. All patients had passive range of motion (PROM) of unknown frequency. Once again, HO was more common in the limbs of patients with severe spasticity. Prolonged coma also appeared to increase the likelihood of HO development.
• In patients with neurologic deficits, increased limb spasticity, decreased joint ROM, and inflammatory signs near a joint strongly suggest the possibility of HO.

Physical

• A diagnosis of HO can be made clinically if localized inflammatory reaction, palpable mass, or limited ROM is observed.
• Clinically, the onset of larger masses of HO is often characteristic of any inflammatory reaction.
• Fairly suddenly, a warm and swollen extremity becomes obvious, and fever is present.
• If sensation is intact, the area of swelling is painful.
• • The swelling usually is localized more than it is in thrombophlebitis, and within several days, a more circumscribed, firmer mass is palpable within the edematous area.
  • If the mass is adjacent to a joint, gradual loss of PROM may follow.
• With the development of early HO at the hip or knee, effusion may be noted at the knee.

Causes

See Pathophysiology.

DIFFERENTIALS

Section 4 of 10  

• Authors and Editors
• Introduction
• Clinical
• Differentials
• Workup
• Treatment
• Medication
• Follow-up
• Miscellaneous
• References

Other Problems to Be Considered

Joint sepsis
Fracture
Hematoma
Early pressure sore (before skin breakdown is evident)
Local trauma

WORKUP

Section 5 of 10  

• Authors and Editors
• Introduction
• Clinical
• Differentials
• Workup
• Treatment
• Medication
• Follow-up
• Miscellaneous
• References

Lab Studies

• Creatine kinase
• • This test is not specific for HO but is of value in determining the severity of muscle involvement and may be helpful in planning treatment of HO. Rossier and colleagues showed in 1973 that patients with an acute form of HO after SCI have elevated CK levels that correlate with histologic involvement of muscle.⁴ Two subsequent studies found CK to be useful in the diagnosis and management of HO. Singh and coauthors reported significantly higher CK levels in patients with HO.¹¹ Data published by Sherman and colleagues indicated that a higher level of CK ultimately correlates with a more severe form of HO, suggesting more widespread involvement of surrounding muscle.¹²
  • These results are promising, because they indicate that CK may reliably predict a higher risk of HO development, can help to predict the severity of a patient's HO, and can be used to follow treatment success.
• C-reactive protein
• • The initial stage of HO is manifested by a prominent inflammatory response. This acute reaction is accompanied by changes in levels of cytokines that stimulate the production of acute-phase proteins, one of these being C-reactive protein (CRP).
  • A study by Estrores and colleagues indicated that the serum concentration of CRP correlates better than does the erythrocyte sedimentation rate with the inflammatory activity of HO after SCI.¹³ In their study, the normalization of CRP in serum was accompanied by a resolution of the inflammation of soft tissue. It seems that administering nonsteroidal anti-inflammatory drugs (NSAIDs) in the early phase of HO, as well as monitoring the serum CRP level, may provide added benefit in reducing the inflammatory reaction that is proposed to be an important factor in HO's genesis.
• Alkaline phosphatase
• • The AlkP level, once a commonly used test, is not often employed today.
  • In many patients, serum AlkP levels are not elevated in acute HO.
  • The elevation can be nonspecific because of associated skeletal injuries or the surgical treatment of fractures.
  • The serum AlkP level is of little value in determining the maturity of HO prior to surgical removal.

Imaging Studies

• Bone scintigraphy¹⁴
• • Ideally, the use of diagnostic imaging should focus on the detection of nonmineralized HO, because the presently available medication, etidronate, can inhibit early mineralization.¹⁴ In this respect, bone scintigraphy and ultrasonography are recommended imaging studies for the early diagnosis of HO.
  • Bone scintigraphy is highly sensitive in the early diagnosis of HO. This is the most commonly used diagnostic study for HO.
  • Freed and colleagues evaluated the 3-phase bone scan in the detection of HO and found that a marked vascular blush and increased blood pool about the hips preceded the development of clinical HO by 2-4 weeks.¹⁵
  • The 3-phase bone scan using technetium-99m (^99mTc) diphosphonate is used in diagnosing and monitoring HO.
• Ultrasonography - This is also used in the early diagnosis of HO about the hips. However, no data are available on the diagnostic value of ultrasonography in the diagnosis of HO in other joints (eg, knee, shoulder, elbow).
• Radiography
• • While plain radiography is highly specific in the diagnosis of HO, this method lacks sensitivity in early diagnosis. Because soft-tissue calcification must occur for radiographic evidence of HO to be present, radiographs are not helpful in the early stages. Radiologic examinations do not show evidence of HO until a flocculent, patchy appearance develops, as calcium is deposited about 7-10 days after the onset of clinical symptoms.
  • This patchy appearance coalesces and enlarges on subsequent examinations, and by 2-3 months, the boundaries of the HO demarcate with the appearance of mature bone. Radiography, however, is not reliable at assessing the maturity of HO, because more mature areas may hide immature areas.
• Computed tomography (CT) scanning and magnetic resonance imaging (MRI)
• • CT scanning and MRI may be useful in delineating local anatomy prior to resection.
  • The role of CT scanning and MRI in the evaluation of other aspects of HO has not been well established.

Other Tests

• Biopsy
• • Biopsy has no role in the diagnosis of HO, but it has been considered as a means of helping to determine maturity.
  • There is a possible risk of inadequate sampling, because mature and immature HO may be intermixed.

TREATMENT

Section 6 of 10  

• Authors and Editors
• Introduction
• Clinical
• Differentials
• Workup
• Treatment
• Medication
• Follow-up
• Miscellaneous
• References

Rehabilitation Program

Physical Therapy

The use of physical therapy (PT) in HO has long been controversial. Rossier and co-investigators noted occasional transverse microfractures on sections of HO that they thought might be caused by spasticity or by overly aggressive PROM.⁴ Since then, the debate between resting the joint and aggressive PROM has continued. In the literature, however, the developing consensus appears to be that aggressive PROM and continued mobilization, once acute inflammatory signs have subsided, are indicated, because they help to maintain ROM and (in more extensive HO) they may lead to the formation of a pseudarthrosis. Resting the joint appears more likely to lead to decreased ROM or to ankylosis.

During the acute inflammatory stage, the patient should rest the involved joint in a functional position, and the physical therapist should initiate gentle PROM as soon as possible. The role of continuous PROM machines has not been studied in this situation. For patients with incomplete SCI or head injuries, maintaining ROM may be difficult because of pain from ROM exercises. The use of joint manipulation has been reported in patients with HO who, because of limited joint ROM, have functional limitations. However, such manipulation is controversial owing to the risk of the formation of new hematoma and because of the chance that long-bone fracture will occur in patients with secondary osteoporosis.

Medical Issues/Complications

Nonarticular complications of HO are rare, but they have been reported. These complications include ulnar nerve compression with HO at the elbow, vascular (predominantly venous) compression with or without associated deep venous thrombosis (DVT), and lymphatic obstruction leading to lymphedema.^{16, 17}

• Prophylaxis
• • Although no effective protocol had previously been developed for preventing HO after SCI, the authors' studies, based on the well-documented beneficial effect of NSAIDs in the prevention of HO after total hip arthroplasty, showed that the following drugs can also be helpful in reducing the incidence and severity of HO after SCI^{9, 18, 19}:
  • • The nonselective NSAID indomethacin SR prescribed for 3 weeks in a dose of 75 mg/d, after SCI, reduced the incidence of HO by 2-3 times.
    • A 25 mg/d prescription of the selective COX-2 inhibitor rofecoxib decreased the risk of HO formation by 2.5 times.
  • These positive results with NSAIDs in the prevention of HO may be an important step forward in the clinical management of this condition.

Surgical Intervention

Once HO has developed to the point that it interferes significantly with the functional capacity of the patient, the only treatment option remaining is surgery, which most commonly is required at the hip.³ Ensure that the HO has reached maturity before resection, because resection of immature HO leads to recurrence rates of nearly 100%. Hemorrhage may be a significant problem at the time of surgery, with an average blood loss of 2100 mL reported. Postsurgical infection may lead to amputation; therefore, great care must be taken at the time of surgery. Initiate a presurgery program to eliminate any possible nidus of bacteremia or infections (eg, decubitus ulcers, urinary tract infections).

The usual surgical technique used on HO occurring anteriorly at the hip is anterior wedge resection. Postoperatively, position the joint properly with foam wedges so that the surgical correction can be maintained and any strain on the incision or pressure sores can be prevented. Start gentle PROM about 72 hours postoperation, and increase therapy intensity gradually to incorporate retraining in functional activities. Patient selection and careful identification of functional goals are critical for successful surgical intervention.

Consultations

Consultation with an orthopedist is necessary for any consideration of surgical management of HO.

Other Treatment

• Radiation therapy^{3, 20}
• • Radiation therapy has been studied mostly in connection with the prevention of HO in patients at high risk for recurrence following hip arthroplasty.
  • The most common use in the rehabilitation setting is for the prevention of postoperative recurrence, but the optimal dosage, frequency, and timing have not been established.
  • Mesenchymal stem cells that may be in muscle and that transform into bone-forming cells are highly radiosensitive. Little is known of radiation therapy's effect on HO after SCI when it is used as a primary treatment. One reason that radiation therapy has not been established as a treatment for HO is a risk of local induction of malignancy. However, radiation has been used in Europe by Sautter-Bihl and colleagues as a primary treatment for early HO after SCI; no adverse effects were noted.²¹

MEDICATION

Section 7 of 10  

• Authors and Editors
• Introduction
• Clinical
• Differentials
• Workup
• Treatment
• Medication
• Follow-up
• Miscellaneous
• References

Today, the medical treatment of HO is directed at early HO. In the later stages of the development of mature bone, medical treatment is ineffective. Etidronate (Didronel) is the only available medication for the treatment of HO after SCI.^{14, 22} Treatment with NSAIDs may be required initially, until the resolution of inflammation and the normalization of CRP levels.³

Drug Category: Nonsteroidal anti-inflammatory agents

Presumed to have direct and indirect effects on the formation of HO. Direct effect refers to the inhibition of the differentiation of mesenchymal cells into osteogenic cells, and indirect effect refers to the inhibition of posttraumatic bone remodeling by suppression of the prostaglandin-mediated inflammatory response.

Drug Category: Bisphosphonates

The bisphosphonate group of compounds has properties similar to naturally occurring pyrophosphate, which may be a regulator of calcification.² Etidronate disodium is the most extensively studied of this class of drugs for the treatment of HO. Etidronate acts by (1) inhibiting precipitation of calcium phosphate from unsaturated solutions, (2) delaying aggregation of apatite crystals into layers, and (3) blocking conversion of calcium phosphate into hydroxyapatite. Apparently, predisposition to the inflammatory process and mineralization decreases with time, although it is not understood why. This phenomenon may be why there is no massive rebound bone formation after cessation of etidronate. Thus, the effectiveness of etidronate depends entirely on when and how long it is given, and the drug does not affect HO that has already formed.

FOLLOW-UP

Section 8 of 10  

• Authors and Editors
• Introduction
• Clinical
• Differentials
• Workup
• Treatment
• Medication
• Follow-up
• Miscellaneous
• References

Further Inpatient Care

• Patients with advanced HO who undergo surgical intervention generally require hospitalization and further inpatient care. For a discussion of care following anterior hip resection surgery, see Surgical Intervention.

Further Outpatient Care

• Monitoring of HO maturation
• • Monitoring of HO maturation should be performed regularly.
  • In patients with functional limitations for whom surgery is a consideration, radiography should be performed every 4-6 months.
  • CT scanning and MRI may offer more precise delineation of ectopic bone, which may be helpful in preoperative planning.

Deterrence

• As previously stated, the prophylactic use of medications to prevent HO has been studied in individuals with SCI, with promising outcomes.^{18, 19} NSAIDs do have a role in the prevention of HO and in the prevention of postoperative recurrence after the excision of HO. Radiation therapy also may be used to prevent recurrence postoperatively in some patients.³

Complications

• Potential complications associated with surgical intervention in patients with HO include hemorrhage and postsurgical infections (see Surgical Intervention).
• HO in patients with SCI may lead to other complications, such as pressure sores and DVT.^{16, 17}

Prognosis

• Approximately 20-30% of patients with SCI develop clinically evident HO, and 3-8% of them develop severe functional limitations.

Patient Education

• Patient and family education is an important part of the treatment process in individuals with HO. Physical therapists may instruct patients and family members, if needed, to complete ROM exercises as instructed by the physician. Patients should also be taught to watch for signs of other potential complications when dealing with heterotopic ossification (in order, for example, to prevent pressure sores in patients with SCI).

MISCELLANEOUS

Section 9 of 10  

• Authors and Editors
• Introduction
• Clinical
• Differentials
• Workup
• Treatment
• Medication
• Follow-up
• Miscellaneous
• References

Medical/Legal Pitfalls

• Careful consideration of the differential diagnosis of the swollen limb and appropriate imaging studies clarify the diagnosis.
• No general standard of care has been agreed on for prophylaxis of HO except in the postoperative setting.

REFERENCES

Section 10 of 10

• Authors and Editors
• Introduction
• Clinical
• Differentials
• Workup
• Treatment
• Medication
• Follow-up
• Miscellaneous
• References

1. Kaplan FS, Xu M, Glaser DL, et al. Early diagnosis of fibrodysplasia ossificans progressiva. Pediatrics. May 2008;121(5):e1295-300. [Medline].
2. Shafer DM, Bay C, Caruso DM, et al. The use of eidronate disodium in the prevention of heterotopic ossification in burn patients. Burns. May 2008;34(3):355-60. [Medline].
3. Subbarao JV, Garrison SJ. Heterotopic ossification: diagnosis and management, current concepts and controversies. J Spinal Cord Med. Winter 1999;22(4):273-83. [Medline].
4. Rossier AB, Bussat P, Infante F, et al. Current facts of para-osteo-arthropathy (POA). Paraplegia. May 1973;11(1):38-78. [Medline].
5. Bodley R, Jamous A, Short D. Ultrasound in the early diagnosis of heterotopic ossification in patients with spinal injuries. Paraplegia. Aug 1993;31(8):500-6. [Medline].
6. Snoecx M, De Muynck M, Van Laere M. Association between muscle trauma and heterotopic ossification in spinal cord injured patients: reflections on their causal relationship and the diagnostic value of ultrasonography. Paraplegia. Aug 1995;33(8):464-8. [Medline].
7. Hassard GH. Heterotopic bone formation about the hip and unilateral decubitus ulcers in spinal cord injury. Arch Phys Med Rehabil. Aug 1975;56(8):355-8. [Medline].
8. Downing MR, Knox D, Gibson P, et al. Impact of trochanteric heterotopic ossification on measurement of femoral bone density following cemented total hip replacement. J Orthop Res. Apr 10 2008;[Medline].
9. Macfarlane RJ, Ng BH, Gamie Z, et al. Pharmacological treatment of heterotopic ossification following hip and acetabular surgery. Expert Opin Pharmacother. Apr 2008;9(5):767-86. [Medline].
10. Garland DE, Blum CE, Waters RL. Periarticular heterotopic ossification in head-injured adults. Incidence and location. J Bone Joint Surg Am. Oct 1980;62(7):1143-6. [Medline].
11. Singh RS, Craig MC, Katholi CR, et al. The predictive value of creatine phosphokinase and alkaline phosphatase in identification of heterotopic ossification in patients after spinal cord injury. Arch Phys Med Rehabil. Nov 2003;84(11):1584-8. [Medline].
12. Sherman AL, Williams J, Patrick L, et al. The value of serum creatine kinase in early diagnosis of heterotopic ossification. J Spinal Cord Med. 2003;26(3):227-30. [Medline].
13. Estrores IM, Harrington A, Banovac K. C-reactive protein and erythrocyte sedimentation rate in patients with heterotopic ossification after spinal cord injury. J Spinal Cord Med. 2004;27(5):434-7. [Medline].
14. Banovac K. The effect of etidronate on late development of heterotopic ossification after spinal cord injury. J Spinal Cord Med. Spring 2000;23(1):40-4. [Medline].
15. Freed JH, Hahn H, Menter R, et al. The use of the three-phase bone scan in the early diagnosis of heterotopic ossification (HO) and in the evaluation of Didronel therapy. Paraplegia. Aug 1982;20(4):208-16. [Medline].
16. Bradleigh LH, Perkash A, Linder SH, et al. Deep venous thrombosis associated with heterotopic ossification. Arch Phys Med Rehabil. Mar 1992;73(3):293-4. [Medline].
17. Varghese G, Williams K, Desmet A, et al. Nonarticular complication of heterotopic ossification: a clinical review. Arch Phys Med Rehabil. Nov 1991;72(12):1009-13. [Medline].
18. Banovac K, Williams JM, Patrick LD, et al. Prevention of heterotopic ossification after spinal cord injury with indomethacin. Spinal Cord. Jul 2001;39(7):370-4. [Medline].
19. Banovac K, Williams JM, Patrick LD, et al. Prevention of heterotopic ossification after spinal cord injury with COX-2 selective inhibitor (rofecoxib). Spinal Cord. Dec 2004;42(12):707-10. [Medline].
20. Strauss JB, Chen SS, Shah AP, et al. Cost of radiotherapy versus NSAID administration for prevention of heterotopic ossification after total hip arthroplasty. Int J Radiat Oncol Biol Phys. Jan 28 2008;[Epub ahead of print]. [Medline].
21. Sautter-Bihl ML, Liebermeister E, Nanassy A. Radiotherapy as a local treatment option for heterotopic ossifications in patients with spinal cord injury. Spinal Cord. Jan 2000;38(1):33-6.
22. Banovac K, Gonzalez F, Renfree KJ. Treatment of heterotopic ossification after spinal cord injury. J Spinal Cord Med. Jan 1997;20(1):60-5. [Medline].

Article Last Updated: May 22, 2008