LOW INTENSITY ULTRASOUND (EXOGEN) THERAPY FOR NON UNION OF FRACTURES

BACKGROUND

The treatment of non union or delayed union of fractures includes a growing number of approaches and techniques, one of which is bone growth stimulationusing low-intensity ultrasound (LIUS). Fracture non-union is likely to occur when there is limited blood supply to the specific bone, or if there is severe trauma. Bones such as the femur head and neck and the scaphoid, have a limited blood supply, which can be destroyed if the bones are broken. The tibia has a moderate blood supply, however, severe trauma and injury can destroy the internal blood supply or the external supply from overlying skin and muscle. Fracture of the fifth metatarsal (i.e., Jones fracture) also frequently results in delayed healing and non-union, despite surgical treatment, generally due to poor blood supply of the proximal metaphyseal diaphyseal region. [1]

THE TECHNOLOGY

Ultrasound bone growth stimulation is a non-invasive intervention, designed to transmit low-density, pulsed, high-frequency acoustic pressure waves to accelerate healing of fresh fractures and to promote healing of delayed unions and non-unions that are refractory to standard treatment. Low-intensity ultrasound has also been suggested to enhance healing of fractures that occur in patients with diseases such as diabetes, vascular insufficiency, and osteoporosis, and those taking medications such as steroids, non-steroidal anti-inflammatory drugs (NSAIDs), or calcium channel blockers.[2]

Currently, Exogen, a technology from the Smith and Nephew Corporation, is the only device marketed worldwide that uses LIUS to influence the fracture healing process. The Exogen bone growth stimulator uses ultrasound,the energy level of which is close to that of diagnostic ultrasound machines. The device is intended to be used by the patient at home and is applied 20–30 minutes daily until healing occurs

According to the manufacturer, the safety and effectiveness of LIUS has only been established for fracturesof the distal radius or tibial diaphysis; fractures with post-reduction displacements of more than 50%; fractures that are open Grade II or III; fractures that require surgical intervention or external fixation; or for fractures that are not sufficiently stable for closed reduction and cast immobilization. Individuals who are not skeletally mature, or who are pregnant/nursing are not candidates for this therapy. LIUS is also not indicated for fractures related to bone pathology or malignancy.[3]

THE PATIENT GROUP

A considerable number of treatments are provided annually to adults in Wales forthe treatment of fresh fractures of the tibia, radius,scaphoid and fifth metatarsal. Treatments include castimmobilisation and closed or open reduction, with or without internal fixation. Closedand Grade I open fractures, for which LIUS is indicated, are most often treated with castimmobilisation, although the use of intramedullary rods is relatively common for tibialfractures. Failure to respond to treatment can result in non-union with implications forthe patient’s quality of life and functional capacity, together with financial costs to bothpatient and the NHS. The patient factors that may inhibit bone healing are the presence of diabetes,smoking or nicotine in any form, older age, severe anaemia, diabetes and drugs such as NSAIDs. [4]

TREATMENT ALTERNATIVES

Currently, a variety of invasive and noninvasive interventions are used to treat non-unions, includingimmobilisation/casting, open or closed reduction, pins, screw fixation, intramedullary rods and bone grafting. Immobilisation is considered the primary treatment for any non-union. Bone growth stimulators (noninvasive or invasive), may be used instead of or in addition to other interventions to promote bone healing. Implantable devices may be used as an adjunct to planned surgical treatment (e.g., bone grafts, internal/external fixation) of an established non-union. [5]

RESEARCH EVIDENCE

See Table 1 for search terms and results

Table 1

DATABASE / SEARCH TERM/S / NUMBER OF HITS/NUMBER RELEVANT
OVID Medline 1966-2006
Cochrane DSR
ACP Journal Club
DARE
CCTR
EMBASE
CINAHL / Ultrasound, low intensity, Exogen, fractures, orthop$edics, pulsed ultrasound, bone growth stimulator / 168/9
Sumsearch / Ultrasound, low intensity, Exogen, fractures, orthop$edics, pulsed ultrasound, bone growth stimulator / 10/4
Google scholar / Exogen, low intensity ultrasound, fractures / 16,100/14
Google / Exogen, low intensity ultrasound, fractures / 46,232/ 24
HMIC
TRIP / Exogen, low intensity ultrasound, fractures
UpToDate / low intensity ultrasound, fractures / 10/1
National Electronic Library for Health / low intensity ultrasound, fractures / 4/2
INAHTA
NICE
National Horizon Scanning Centre
National Research Register
Current Controlled Trials / low intensity ultrasound, fractures / 2/2
0
0
0
5/2

The National Public Health Service produced policy advice on electrical and electromagnetic field treatment for non-union of bones in 2005, which was updated in 2008; this advice considered non-invasive electrical bone growth stimulators such as the Exogen and concluded that the efficacy was still uncertain but some studies are beginning to show statistically significant effects. [6]

There have been several good quality systematic reviews and meta-analyses on the evidence for effectiveness of LIUS in different fracture situations. One performed in 2001 in Australia5 concluded that it is not possible to stateirrefutably that LIUS is more effective than other treatments for fresh fractures. The authors identified only 2 randomised controlled trials (RCTs)concerned with distal radius and tibial fractures and the results were contradictory.[7][8] There was no RCT data for non-union fractures and only poorly controlled registry data and case series evidence was available. The authors considered however, that this represented minimally acceptable low-level evidence to support the efficacy of LIUS for treatment of non-unions. This conclusion was restricted to patients with radiologically confirmed fracture non-union who have failed previous treatment. Importantly this conclusion is valid only for comparison with no further treatment, which is an inappropriate comparator for usual orthopaedic practice.

A subsequentAmerican technology assessment [9] considered the data for non-union of fractures. Three separate studies with a total of 1446 patients reported data concerning ultrasound treatment of non-unions. All studies were case series using the Exogen system and examined the responses of multiple bone types. Nolte et al.[10] examined 29 patients and included both radiographic evidence andclinical assessment in determining healing. No differences in healing rates were found for the 8 patients treated by casting, 12 patients treated by osteosynthesis, 6 with intramedullary rods, and 3 with external fixators. Even if this analysis were adequately powered to detect a difference, it would not rule out a role for fixation and stabilisation inthe healing process, only that each method of fixation worked equally well with ultrasound. Not including the 12 patients excluded from their analysis, all 10 tibias healed and 4/5 of the femurs, radii, and scaphoids healed. Response to treatment was monitored with clinical and radiographic examinations at 6 to 8-week intervals, and treatment was continued until the treating surgeon declared the non-union healed. This study was not of good quality because it did not mask patient assessment.Results for 12 patients enrolled but excluded from the analysis were also presented in the article. It was clear that some of the patients who had undergone surgery within the 90-day period prior to use of ultrasound were still considered to be ultrasound “successes.” And this could inflate the effect of ultrasound, if theprior treatment effect was still ongoing. The healing rate as reported in the study was 86% (25 of 29 non-unions) healed excluding the 12 patients described above. It is not clear whether the stratified analyses presented were planned before the study began or were post hocanalyses and could therefore be considered as data dredging. Among the patients that completed the study, 18 had no treatment other than the initial procedures used to treat the fracture. Fourteen of these patients (78%) healed during the study. All 11 of the patients who received a secondary procedure prior to ultrasound therapy healed.10

Mayr et al. examined data on 1317 patients from a registry maintained by Smith and Nephew, themanufacturer of the ultrasound device.[11] Patients were described as having “delayed union” (951 cases) if a fracture remained unhealed for 3 to 9 months following fracture, and “non-union” (366 cases) if the time since initial fracture was greater than 9 months. Thepercentage of patients healed and the time to healing for delayed union and non-union in this prescription-based registry were compared with results obtained for 42 patients in the authors’ clinic, but no other validation of the registry is mentioned in the article. The authors found very similar rates of healing and time to healing in their clinic population as in the registry population, based on lack of a statistically significant difference. It cannot be concludedhowever, that the two groups are the same based on non-significant results. Among the non-unions in this study, 314 of 366 (86%) fractures healed, including 105 of 120 (88%) tibias and 57 of 66 (86%) femurs. Among the 951 “delayed unions,” 862 (91%) healed, including 350 of 380 (92%) tibias and 85 of 98 (87%) femurs.(10) Given the requirement that non-unions be diagnosed at 9 or more months following injury (mean 24.9 months), these results suggest that the ultrasound therapy contributed tohealing. Of note, the mean time to healing in the non-union patients was 152 (S.E.: 5.3) days versus 129 (S.E.: 2.7) days in the delayed union group.11

The size of the registry population does improve the generalisability of the results; however, retrospective data collection and post-hoc analyses of registry data generally raise concerns about the potential for bias in patient selection and analysis. Outcome assessments in the Mayr et al. study 11were not blinded and may not have been consistently applied across all patients. Compliance was not reported, and clinical examination (pain and weight-bearing) was not used as part of the assessment of healing. The study was therefore considered to have low internal validity. The study also failed to consider the effect of concurrent immobilisation or other treatments in theassessment of healing rates. The other study by Mayr describes 100 patients treated with ultrasound.[12] Healing rates were 55/64 (86%) for delayed unions, which healed in an average of 142 days, and 31/36 (86%) for non-unions, which healed in an average of 171 days. Inclusion of the 21 patients who discontinued treatment reduces the overall healing rate to 86 of 121 (71%). The internal validity of the study was considered low because the outcome assessment was not blinded, pain and weight-bearing outcomes were not reported, and statistical analysis was not reported. Although the size of the patient populations and the methods of data collection and analysis were very different, Nolte et al.10 and Mayr et al.11 reported 86% healing innon-unions of all bone types at an average of 152 days and Mayr et al.12reported 86% healing of non-unions in 171 days. Nolte et al.10 reported that all six patients 65 years or older had healed non-unions after treatment, but Mayr et al.11 analysed the registry data for an effect of age and reported that the healing rate for non-unions consistently declined from 97% at 20 years to 71% at 70 years. The evidence on the effect of age on healing was therefore inconsistent.

The authors of the AETMIS 4study considered separately the evidence for acceleration of healing, prevention of non-union and treatment of non-union. The technology brief relies heavily on the evidence presented in the review from Australia5and the subsequent meta-analysis by Busse et al. [13]on the effect of LIUS on time to fracture healing. The latter meta-analysis has been criticised for introducing bias in favour of an effect by excluding one RCT8 that did not find an effect of LIUS. According to AETMIS’s assessment, the availablestudies did not mention any adverse effectsassociated with this treatment modality. Giventhe efficacy and safety evidence AETMIS consideredthat, with regard to the acceleration ofhealing and the prevention of non-union, thelevel of evidence was insufficient to recommendthe use of low-intensity ultrasound. However, inthe case of non-union of tibial fractures however, theprognosis is often so poor that it seems reasonable toconsider the use of LIUSafter failed surgical intervention and after theconsolidation process, as measured by serialradiographs including multiple views, hasceased for several months. Further data has confirmed the utility of LIUS for post traumatic non-unions of the tibia [14][15][16] As for fracture sitesother than the tibia, the uncertainties concerningthe efficacy of Exogen in the treatment of non-union should be assessed in light of the prognosisspecific to these fractures and of the clinicalcontext.

The review of the literature by CIGNA[17]concluded there was sufficient evidence in the peer-reviewed scientific literature to support the safety and efficacy of ultrasound bone growth therapy in patients with fresh fractures of the distal radius and the tibial diaphysis, when the patients have skeletal maturity, and when the therapy is used as an adjunct to closed reduction and cast immobilization, or for non-union of bones other than skull or vertebrae in skeletally mature individuals. There was also some evidence that LIUS may enhance healing of fractures that are high risk for delayed union or non-union, in addition to stress fracture non-union.

The authors concluded that therewas insufficient evidence in the peer-reviewed, published scientific literature to support the clinical utility of bone growth stimulation for the treatment of any of the following non-union conditions:

  • fresh fractures (other than when using ultrasound bone stimulation for the tibia, radius or other high-risk fractures)
  • toe fractures
  • sesamoid fractures
  • avulsion fractures
  • osteochondral lesions
  • displaced fractures with malalignment
  • synovial pseudarthrosis
  • the bone gap is either > 1 cm or > one-half the diameter of the bone

COST EFFECTIVENESS

From the health economic aspect, it is recognised that the longer the delay to union, the greater the total cost for the treatment of this fracture. In a 1997 publication, Heckman[18]et al estimated an overall cost savings for the use of LIUS of approx US$ 13,000-15,000 /case. In the Australian review5the incremental costs per quality adjusted life year (QALY) gained for LIUS treatment of fresh tibial, distal radius and sca[phoid fractures were Au$106,601, Au$501.699 and Au$641,060, respectively. The authors considered that the cost-effectiveness of LIUS in each of the indications reviewed did not compare favourably with a range of other common healthcareinterventions.

A 2006 study from the York Health Economics Consortium,[19]evaluated the relative cost-effectiveness of ultrasound stimulation (Exogen) as a complement to conservative therapy ( casting) or surgical fixation in fresh fractures in patients at risk ( e.g smokers or diabetics) of non-union (non-union is defined at six months). A cost-effectiveness model estimated expected outcomes and costs in the 12-months followingfirst presentation for a cohort of patients with fresh fracture of the tibia. The model allowed the probability of healing to be varied to reflect the prognosis of patients with ahigher than average risk of non-union. These groups included current and past smokers andpatients with diabetes. The higher risk of non-union is reflected in a percentage reduction in theprobability of healing. Probabilities are derived from the literature. The analysis reflected costs to the payer – Medicare in the US, the National Health Service in theUK. Healthcare resources (such as physician visits, physiotherapy sessions, X-rays) used to treatpatients with fracture of the tibia were identified through interviews with expert orthopaedicsurgeons in the US and the UK.

For a population at risk of non-union whose probability of healing is 80% of the generalpopulation (or less), Exogen was cost saving irrespective of whether the patient was treatedconservatively or with surgery. The greater the risk of non-union, thegreater the relative advantage of Exogen. In the UK adding Exogen to conservative treatment reduced cost per patient by £1,378 (£2,415versus £3,793) and using Exogen as an adjunct to surgery reduced the cost of treatment by £884per patient (£5,578 versus £6,462). The increase in the number of fractures healed was 7.6% and6.4% respectively.

Including lost productivity in the model the addition of Exogen toconservative treatment was cost-saving, even for the general population of fresh fractures (cost perpatient was reduced by -$2,136). Busse et al [20] suggested that from an economic standpoint, while reamed intramedullary nailingis the treatment of choice for closed and open Grade 1 tibial fractures, treatment with therapeuticultrasound and casting may also be an economically sound intervention (for appropriate patients). The authors concluded that in a population at risk of non-union, ultrasound was both less expensive and led to betteroutcomes. Providing the risk is such that the probability of healing is 80% of the generalpopulation or less, adjunctive ultrasound appeared to be aneffective strategy irrespective of theprimary treatment choice.3-200ource

l., 1988

ONGOING RESEARCH

A search of the controlled trials meta-register and the Cochrane Library revealed the following relevant trials

  1. Trial to evaluate ultrasound in the treatment of tibial fracture- National Institute for Health(NIH), recruiting
  2. Trial to re-evaluate ultrasound in the treatment of tibial fractures (TRUST) – NIH, completed
  3. Pulsed Ultrasound to Speed-up Healing after Intramedullary nailing of Tibia fractures Charite Hospital, Germany - ongoing
  4. Busse JW, Bhandari M, Kulkarni AV, Schünemann HJ. Therapeutic ultrasound for treating fractures in adults. (Protocol) Cochrane Database of Systematic Reviews 2005, Issue 3. Art. No.: CD005464. DOI: 10.1002/14651858.CD005464.

CONCLUSIONS

.

  • There was a lack of consistent high quality evidence from controlled trials for the effectiveness of LIUS (Exogen) in the acceleration of healing of fractures, the prevention of fracture non-union or the treatment of non-union. One often quoted meta-analysis suffered from methodological problems that introduced positive bias into the conclusions. Two reviews did however consider that despite the lack of high level evidence, the data from poorly controlled registry data and case series was adequate to recommend treatment using the Exogen device for non-union of fractures, particularly those of the tibia.
  • While the results of these studies suggest that ultrasound promotes the healing of non-union fractures, they do not rule out a role for other concurrent treatment procedures—such as stabilization of the non-union—contributing to the observed effects. The two studies reporting data for patients over 65 are not in agreement as to the effect of age on response to ultrasound treatment
  • The cost effectiveness data did appear to suggest that adding Exogen to conservative treatment is cost effective providing that the healing of the population undergoing treatment is ≤80%of the general population

.