3.1 Full details of all clinical outcomes considered by the Committee are available in the assessment report overview. Reference to EXOGEN in the context of non-union should be taken to mean the EXOGEN 4000+ device and in the context of delayed healing to be the EXOGEN Express device, unless otherwise stated.
3.2 The key clinical outcomes for EXOGEN for long bone fractures with non-union or delayed healing presented in the decision problem were:
evidence of bridging on radiograph (3 out of 4 cortices bridged)
fracture healing time
return to painless weight bearing
avoidance of further surgery
device-related adverse events.
3.3 The sponsor presented clinical evidence from 18 studies in its submission. Of these, the External Assessment Centre excluded 1 because it did not report any outcomes defined in the scope. The clinical evidence for EXOGEN is therefore based on 17 clinical studies (total of 1710 patients), including 3 randomised controlled trials, 13 case series and 1 prospective comparison. Of these, 13 studies reported on non-union fractures, 2 reported on delayed healing and 2 reported on both types of fracture. There were no controlled or randomised studies in which EXOGEN and surgery were compared directly in the treatment of either non-union or delayed fractures. However, independent estimates of healing rates for EXOGEN and surgery were available from non-comparative case series for non-union fractures. The age of study participants ranged from 13 to 92 years and follow-up across the studies ranged from 2 months to 6 years. None of the studies were carried out in the UK.
3.4 Mayr et al. (2000) described 256 patients with non-union fractures (failure to heal 9 months after fracture) from an international register of patients treated with EXOGEN. Healing was defined as 3 cortices bridged in 3 X-ray planes or trabecular bridging of at least 80% of the fracture in the case of cancellous fractures. The mean healing rate across all long bone fractures (humerus, radius/ulna, femur, tibia-fibula) was 84% (216/256), with a mean healing time of 5.3 months.
3.5 Gebauer et al. (2005) described a case series of 51 patients with non-union fractures (defined as minimum fracture age 8 months, radiographic indication that the healing process had stopped for at least 3 months, and a minimum of 4 months without surgical intervention before EXOGEN). A healing rate (healing defined as no pain or motion upon gentle stress and weight bearing if applicable, and radiographic healing defined as 3 of 4 cortices bridged) of 90% (46/51) for all long bone fractures (not otherwise described) was reported with a mean healing time of 178 days (range 86–375 days).
3.6 In a case series of 32 patients with non-union fractures (defined on the basis that surgery was otherwise deemed to be indicated), Jingushi et al. (2007) reported a healing rate (defined as clinical and radiographic healing as determined by experienced orthopaedic surgeons) of 66% (21/32); analyses by individual long bone were not included. A mean healing time of 219 days (range 56–588 days) was reported for a mixed group of 72 patients with non-union and delayed healing fractures. When treatment with EXOGEN was started within 6 months of the most recent operation, the union rate was approximately 90%. When treatment was started after 12 months, the union rate was less than 65% (follow-up not reported).
3.7 Nolte et al. (2001) evaluated a case series of 22 patients with non-union fractures (defined as failure of the fracture to unite at a minimum of 6 months from fracture, no progression towards radiographic healing or healing had stopped for a minimum period of 3 months before EXOGEN). Healing rates (healing defined as absence of pain, weight bearing without pain or normal function of the limb, 3 or 4 cortices bridged on radiograph) of 100% (10/10) for tibia-tibia/fibula (mean healing time 144 days), 80% (4/5) for femur (mean healing time 185 days), 80% (4/5) for radius-radius/ulna (mean healing time 139 days) and 100% (2/2) for other long bone fractures (mean healing time 153 days) were reported.
3.8 Romano et al. (1999) studied 13 patients with non-union fractures of long bones (tibia, humerus and femur) and septic pseudoarthrosis. Healing was reported in 62% (8/13) of patients (no further details reported).
3.9 Data were identified by the sponsor on the rates of healing for non-union long bone fractures treated by surgery. Healing rates ranged from 62% to 100%, and healing time ranged from 9 weeks (Livani et al. 2010) to 24 weeks (Ring et al. 1997). Across 3 case series and 1 cohort study, including a total of 166 patients with non-union fractures treated by surgery, 10 patients needed further surgery (follow-up not reported; Birjandinejad et al. 2009, Khalil et al. 2010, Lin et al. 2010 and Ring et al. 1997). These studies reported on fractures of different long bones, including distal femur, femur, tibia and ulna/radius.
3.10 Schofer et al. (2010) carried out a randomised controlled trial of 101 patients with delayed healing of tibial shaft fractures (defined as lack of clinical and radiologic evidence of union, bony continuity or bone reaction at the fracture site no less than 16 weeks from the index injury or the most recent intervention) treated by EXOGEN (n=51) or placebo (n=50). No significant difference was reported between the groups in healing rate (judged by clinician, not otherwise described) over a 4 month follow-up period (65% [33/51] for EXOGEN compared with 46% [23/50] for placebo, HR 1.69, p=0.07).
3.11 Mayr et al. (2000) reported on a total of 696 patients from the international register for EXOGEN (see section 3.4) who received treatment for fractures with delayed healing (defined as failure to heal 3–9 months after fracture). In this case series, 90% (586/654) of all long bone fractures healed (as defined in section 3.4) in a mean time of 4.4 months. The authors presented healing rates separately for the different types of fracture. Healing rates ranged from 76% (41/54) with a mean healing time of 125 days for fractures of the humerus to 96% (26/27) with a mean healing time of 113 days for fractures of the fibula.
3.12 The case series reported by Jingushi et al. (2007) included 40 patients with delayed healing fractures (defined as union or radiological bone reaction not being observed more than 3 months after the most recent operation). A healing rate (healing defined in section 3.4) for fractures of the femur, tibia, humerus, radius and ulna of 83% (33/40) was reported (follow-up not stated).
3.13 In a case series of 16 patients with delayed healing (defined as no radiological evidence of fracture callus 4–38 months after surgical insertion of an intramedullary nail or the Ilizarov procedure [external fixator]), Lerner et al. (2004) reported a healing rate (as determined by an experienced orthopaedic surgeon) of 94% (15/16) over a mean follow-up of 17 months (fractures included femur, tibia, radius/ulna and humerus).
3.14 No studies that reported healing rates after surgery in patients with delayed healing long bone fractures were presented by the sponsor.
3.15 The US Food and Drug Administration (FDA) Manufacturer and User Facility Device Experience (MAUDE) database reported 3 cases of skin irritation caused by skin sensitivity to the coupling gel (resolved by change of coupling medium) and 1 report of increased chest pain, possibly caused by interference with a cardiac pacemaker, during a 1-year period (the sponsor stated that approximately 55,000 EXOGEN devices were used by patients in the USA over this time period).
3.16 None of the clinical studies reported device-related adverse events and no significant safety concerns were identified by the External Assessment Centre in relation to EXOGEN in its independent search of the literature. In contrast, reports on surgical treatment of non-union and delayed healing fractures documented adverse events including postoperative wound infection, osteomyelitis and pain.
3.17 The Committee considered that although the evidence on using EXOGEN for long bone fractures with non-union was from observational studies and related to a limited number of outcomes (healing defined in various ways including weight-bearing, and radiographic evidence), it suggested good clinical results after treatment with EXOGEN. The Committee judged that the observed healing rates supported the efficacy of EXOGEN in promoting healing of these fractures and that its use meant that many of these patients avoided surgery.
3.18 For long bone fractures with delayed healing, the Committee found the outcomes after treatment with EXOGEN more difficult to interpret. There were uncertainties, including the rate at which healing progresses between 3 and 9 months after fracture, both with and without EXOGEN. There were also uncertainties about the proportion of patients in whom surgery would be avoided, because some of these fractures heal spontaneously.
3.19 The Committee noted that the clinical evidence comparing the efficacy of EXOGEN with surgery was very limited. The Committee recognised the difficulties in conducting comparative studies (and specifically randomised controlled trials) to collect data on healing rates.
3.20 Clinical experts advised the Committee that the efficacy of EXOGEN may differ depending on which long bone is being treated. The experts stated that non-union most commonly occurs in fractures of the tibia.
3.21 The Committee discussed the applicability of the data from Schofer et al. (2010) in the context of delayed healing. The External Assessment Centre stated that 51 of the 101 patients in this trial had sustained fractures 9 months or more before entry to the study. According to the definition used in the scope, these would be classified as non-union fractures. The Committee was advised by the External Assessment Centre that this trial was not powered to detect differences in healing rates (one of the outcomes defined in the decision problem) and so considered the other primary clinical outcomes reported; radiographically-measured bone mineral density and gap at the fracture site (assessed by computed tomography scan). The Committee noted that there were significant improvements in these outcomes in the group treated with EXOGEN compared with placebo (sham treatment).
3.22 The Committee discussed the variation in fracture healing time among patients. It was advised that there is a considerable natural inter-patient variation in healing rates and that this could explain differences in healing rates reported across the studies. In addition, variation in healing time is more pronounced in the early stage of the healing process and that contributes to the greater complexity of interpreting outcomes for fractures with delayed healing compared against those with non-union.
3.23 The Committee recognised that there may be subgroups of patients in whom healing takes place at a slower rate than the general population. However, it considered that neither current evidence nor expert opinions provided sufficient information to model the potential impact of EXOGEN in these groups of patients.