Aims

Dislocation rates are reportedly lower in patients requiring proximal femoral hemiarthroplasty than for patients undergoing hip arthroplasty for neoplasia. Without acetabular replacement, pain due to acetabular wear necessitating revision surgery has been described. We aimed to determine whether wear of the native acetabulum following hemiarthroplasty necessitates revision surgery with secondary replacement of the acetabulum after proximal femoral replacement (PFR) for tumour reconstruction.

We are currently facing an epidemic of periprosthetic fractures around the hip. They may occur either during surgery or post-operatively. Although the acetabulum may be involved, the femur is most commonly affected. We are being presented with new, difficult fracture patterns around cemented and cementless implants, and we face the challenge of an elderly population who may have grossly deficient bone and may struggle to rehabilitate after such injuries. The correct surgical management of these fractures is challenging. This article will review the current choices of implants and techniques available to deal with periprosthetic fractures of the femur.

Cite this article: Bone Joint J 2014;96-B(11 Suppl A):48–55.

The goals of revision arthroplasty of the hip are to restore the anatomy and achieve stable fixation for new acetabular and femoral components. It is important to restore bone stock, thereby creating an environment for stable fixation for the new components. The bone defects encountered in revision arthroplasty of the hip can be classified either as contained (cavitary) or uncontained (segmental). Contained defects on both the acetabular and femoral sides can be addressed by morselised bone graft that is compacted into the defect. Severe uncontained defects are more of a problem particularly on the acetabular side where bypass fixation such as distal fixation on the femoral side is not really an alternative. Most authors agree that the use of morselised allograft bone for contained defects is the treatment of choice as long as stable fixation of the acetabular component can be achieved and there is a reasonable amount of contact with bleeding host bone for eventual ingrowth and stabilisation of the cup. On the femoral side, contained defects can be addressed with impaction grafting for very young patients or bypass fixation in the diaphysis of the femur using more extensively coated femoral components or taper devices. Segmental defects on the acetabular side have been addressed with structural allografts for the past 15 to 20 years. These are indicated in younger individuals with Type 3A defects. Structural grafts are unsuccessful in Type 3B defects. Alternatives to the structural allografts are now being utilised with shorter but encouraging results in most multiply operated hips with bone loss. New porous metals such as trabecular metal (tantalum), which has a high porosity similar to trabecular bone and also has a high coefficient of friction, provide excellent initial stability. The porosity provides a very favorable environment for bone ingrowth and bone graft remodeling. Porous metal acetabular components are now more commonly used when there is limited contact with bleeding host bone. Porous metal augments of all sizes are being used instead of structural allografts in most situations. On the femoral side, metaphyseal bone loss, whether contained or uncontained, is most often addressed by diaphyseal fixation with long porous or tapered implants, modular if necessary. Distal fixation requires at least 4 centimeters of diaphyseal bone and in Type IV femurs, a choice must be made between a mega prosthesis or a proximal femoral allograft. The proximal femoral allograft can restore bone stock for future surgery in younger patients. The mega prosthesis which is more appropriate in the older population may require total femoral replacement if there is not enough diaphyseal bone for distal fixation with cement. Cortical struts are used for circumferential diaphyseal bone defects to stabilise proximal femoral allografts, to bypass stress risers and to serve as a biological plate for stabilising peri-prosthetic fractures

A seventy-five-year-old female patient presented with pain and deformity of her left leg of three days duration. Hybrid THRA has been done 11 years ago at her left hip for the treatment of osteoarthritis. Massive osteolysis and pathologic fracture were observed on plain radiograph (Fig. 1). Revision THRA using an allograft prosthesis composite (APC) was planned for solution of extensive bone loss of the proximal femur. Surgical exposure was performed through extended trochanteric osteotomy with the patient supine. Step-cut osteotomy was done at the remained proximal part of host femur to make match with the distal part of APC. Meticulous removal of granulation tissues and remaining cement was done. As Acetabular cup was stable, 60 mm sized high-walled polyethylene liner was exchanged. Calcar reconstruction prosthesis was cemented into a proximal femoral allograft measuring 15 cm and cement at the vicinity of the step-cut osteotomy was removed for later bony union at interface. After solid fixation of APC with cement, the distal half of APC was cemented with the host femur. Step-cut osteotomy was wired and autogenous bone grafts from the greater trochanter were added at the interface. Leg length and stability were rechecked using a standard necked 28 mm metal head and reduction was done stably. Greater trochanter was fixed over the trimmed proximal allograft with multiple wiring and paper-thin host femur was enveloped around the femoral allograft using absorbable sutures. Following insertion of the closed suction drainage drains, closure was done as routine fashion and healing of the wound was uneventful (Fig. 2). An abduction brace was applied post operatively for a period of four weeks. Crutch walking with partial weight bearing was started at four weeks and crutch protection was applied for a period of six months. Incorporation of allograft with the host bone was observed on two-year follow-up radiographs. At seven-year follow-up, the patient walks well with a mild limp, and Harris score is 90. We report on a seven-year follow-up case of revision THRA with APC with references (Fig. 3)

Revision total hip arthroplasty (THA) is projected to increase by 137% from the years 2005 to 2030. Reconstruction of the femur with massive bone loss can be a formidable undertaking. The goals of revision surgery are to create a stable construct, preserve bone and soft tissues, augment deficient host bone, improve function, provide a foundation for future surgery, and create a biomechanically restored hip. Options for treatment of the compromised femur include: resection arthroplasty, allograft prosthetic composite (APC), proximal femoral replacement, cementless fixation with a modular tapered fluted stem, and impaction grafting. The purpose of this article is to review the treatment options along with their associated outcomes in the more severe femoral defects (Paprosky types IIIb and IV) in revision THA.

Introduction. Revision hip arthroplasty with massive proximal femoral bone loss remains challenging. Whilst several surgical techniques have been described, few have reported long term supporting data. A proximal femoral allograft (PFA) may be used to reconstitute bone stock in the multiply revised femur with segmental bone loss of greater than 8 cm. This study reports the outcome of largest case series of PFA used in revision hip arthroplasty. Methods. Data was prospectively collected from a consecutive series of 69 revision hip cases incorporating PFA and retrospective analyzed. Allografts of greater than 8 cm in length (average 14cm) implanted to replace deficient bone stock during revision hip surgery between 1984 and 2000 were included. The average age at surgery was 56 years (range 32–84) with a minimum follow up of 10 years and a mean of 15.8 years (range). Results. From the original cohort four patients had died with the original PFA, 21 (30.4%) patients required further surgery with 14 (20.3%) of these needing revisions of the femoral component. The mean time to femoral revision was 9.5 years and Kaplan-Meier survivorship analysis demonstrates a 79.9% PFA survivorship at 20 years. Discussion. Proximal femoral allograft affords long lasting reconstruction of the femoral component in revision hip surgery. We advocate PFA as an attractive option in the reconstruction of the hip in the presence of significant segmental bone loss in younger patients

Massive uncontained glenoid defects are a difficult surgical problem requiring reconstruction in the setting of either primary or revision total shoulder arthroplasty. Our aim is to present a new one-stage technique that has been developed in our institution for glenoid reconstruction in the setting of massive uncontained glenoid bone loss. We utilise a modified delto-pectoral approach to perform our dual biology allograft autograft glenoid reconstruction. The native glenoid and proximal femoral allograft are prepared and shaped to create a precisely matched contact surface, which permits axial compression to secure fixation. The surface of the glenoid is lateralised to at least the level of the coracoid. The central cancellous femoral allograft is removed and impaction autografting is performed prior to implantation of a glenoid base plate with 25-mm long centre peg. Two screws are inserted into the best quality native scapular bone available to ensure compression. A reverse shoulder arthroplasty is implanted. We have performed our dual-biology reconstruction of the glenoid in combination with reverse total shoulder arthroplasty in 8 patients to date. The technique has been performed in the setting of massive uncontained glenoid defects without prostheses as well as in revisions from failed hemiarthroplasties and total shoulder arthroplasties. Our post-operative follow-up is now up to 32 months. CT scanning as early as 6 months demonstrates incorporation of the graft. There has been no evidence of loosening. None of our cases have been complicated by infection or peri-prosthetic fracture and there have been no dislocations. One patient sustained an acromial stress fracture at 9 months post-operatively after lifting a 100-pound gas cylinder. This was diagnosed on bone scan, had no impact on the construct and was managed in a sling for comfort. Another patient has developed Nerot grade I notching which substantially in all patients, with an average improvement of 6.6 on a 10-point scale. Our dual biology allograft-autograft reconstruction is a useful and elegant technique in the setting of massive uncontained defects of the glenoid, which permits the implantation of a reverse total shoulder arthroplasty. We believe this technique to be reproducible and uses materials that are both readily available and familiar

The treatment of substantial proximal femoral bone loss in young patients with developmental dysplasia of the hip (DDH) is challenging. We retrospectively analysed the outcome of 28 patients (30 hips) with DDH who underwent revision total hip replacement (THR) in the presence of a deficient proximal femur, which was reconstructed with an allograft prosthetic composite. The mean follow-up was 15 years (8.5 to 25.5). The mean number of previous THRs was three (1 to 8). The mean age at primary THR and at the index reconstruction was 41 years (18 to 61) and 58.1 years (32 to 72), respectively. The indication for revision included mechanical loosening in 24 hips, infection in three and peri-prosthetic fracture in three.

Six patients required removal and replacement of the allograft prosthetic composite, five for mechanical loosening and one for infection. The survivorship at ten, 15 and 20 years was 93% (95% confidence interval (CI) 91 to 100), 75.5% (95% CI 60 to 95) and 75.5% (95% CI 60 to 95), respectively, with 25, eight, and four patients at risk, respectively. Additionally, two junctional nonunions between the allograft and host femur required bone grafting and plating.

An allograft prosthetic composite affords a good long-term outcome in the management of proximal femoral bone loss in revision THR in patients with DDH, while preserving distal host bone.

We have followed a consecutive series of revision hip arthroplasties, performed for severe femoral bone loss using anatomic specific proximal femoral allografts. Forty-nine revision hip arthroplasties, using anatomic specific proximal femoral allografts longer than five centimetres were followed for a mean of 10.4 years. The mean preoperative HHS improved from 42.9 points to 76.9 points postoperatively. Six hips (12.2%) were further revised, four for non-union and aseptic failure of the implant (8.2%), one for infection (2%), and one for host step-cut fracture (2%). Junctional union was observed in 44 hips (90%). Three hips underwent re-attachment of the greater trochanter for trochanteric escape (6.1%). Asymptomatic non-union of the greater trochanter was noticed in three hips (6.1%). Moderate allograft resorption was observed in five hips (10.2%). Two fractures of the host step-cut occurred (4.1%). There were four dislocations (8.2%), two of them developed in conjunction with trochanteric escape. By definition of success as increase of HHS by 20 points or more, and no need for any subsequent re-operation related to the allograft and/or the implant, a 75.5% rate of success was found. Kaplan-Meier survivorship analysis predicted 73% rate of survival at 12 years, with the need for further revision of the allograft and/or implant as the end point. We conclude that the good medium-term results with the use of large anatomic- specific femoral allografts justify their continued use in cases of revision hip arthroplasty with severe bone stock loss

Introduction: Chronic hip arthroplasty infection is a difficult situation to solve. The use of uncemented stem is questionable, as no antibiotic loaded cement is used to fix the implant. Bone deficiencies are often enlarged by the chronic infection process and the multiple previous surgeries. We report our experience in two stage revisions with massive allograft and uncemented distal locked stem. Material and method: We have retrospectively reviewed 15 patients operated on from chronic hip arthroplasty infection and secondary large femoral defects. Two stages revision with a temporarily antibiotic loaded cement spacer and 6-weeks second look reconstruction were performed with massive proximal femoral allograft and a long distally locked stem. The uncemented stem was cemented into the allograft and remained uncemented in the recipient distal femur. Outcome was assessed regarding infection control, allograft complication and function according to the Merle d’Aubigné score. Results: Patients mean age was 62 years (SD 11). Patients were already operated from the same hip 5.9 +/− 2.3 times before the actual procedure. Isolated microorganisms were MRSE (7), staph Aureus (4), MRSA (1), VRE (1), enterococcus (2). Muscle flaps were performed in 2 cases and discuss for 2 more. Allograft length was 22 +/− 9 cm. Antibiotic treatment was maintained for 4.6 +/− 2.5 months (median 3 months). Mean follow-up was 29 months (SD 26; median 23). Clinical outcome based on Merle d’Aubigné score is 5.3 +/− 3 before treatment and 13.8 +/− 2.5 at the last visit. Complications were registered in 4 cases. Infection recurrence was observed in 3 cases (2,7,20 months) but the function was preserved in all cases. One of them needed a two stage procedure with a new allograft after 19 months due to acute staph Aureus infection. One allograft was broken after 80 months and need one stage revision. Discussion: While the treatment of these cases was particularly demanding as bone deficiencies and microorganisms antibiotic resistance were high, functional result obtained was good. When infection recurrence was observed, function remains preserved. The use of uncemented distal locked stem with a cemented massive allograft did not lead to mechanical or hardware problem and allow us to restore the bone stock and the gluteus tendons insertion. The use of a modular stem was very helpful in solving these complex situations. Massive allografts can be used in case of infection as antibiotic delivery system

We have followed a consecutive series of forty-nine revision hip arthroplasties (45 patients), performed for severe femoral bone loss using anatomic specific proximal femoral allografts longer than five centimetres. The patients were followed for a mean of 10.4 years, with a five year minimum follow-up. The mean preoperative Harris Hip Score improved from 42.9 points to 76.9 points postoperatively, an average improvement of 33.8 points. Six hips were further revised, for a failure rate of 12.2%, four for non-union and aseptic failure of the implant (8.2%), one for infection (2%), and one for host step-cut fracture (2%). Junctional union was observed in 44 hips (89.8%). Three hips underwent re-attachment of the greater trochanter for trochanteric escape (6.1%). Asymptomatic non-union of the greater trochanter were noticed in three hips (6.1%). Moderate allograft resorption was observed in five hips (10.2%), non were full-thickness graft resorption. Two fractures of the host step-cut occurred (4.1%). There were four dislocations (8.2%), two of them developed in conjunction with trochanteric escape. By definition of success as increase of HHS by 20 points or more, and no need for any subsequent re-operation related to the allograft and/or the implant, a 75.5% rate of success was found. Kaplan-Meier survivorship analysis predicted 73% rate of survival at 12 years, with the need for further revision of the allograft and/or implant as the end point. We conclude that the good medium-term results with the use of large anatomic-specific femoral allografts justify their continued use in cases of revision hip arthroplasty with severe bone stock loss

Introduction: Total hip arthroplasty (THA) has proven to be a highly successful procedure, but with its increased use there are an increasing number of joints requiring revision. A number of those patients requiring revision present with a severe loss of femoral bone stock around the failed femoral hip implant, which makes conventional revision techniques difficult or impossible. Materials and Methods: We have followed a consecutive series of forty-nine revisions THA (45 patients), performed for severe femoral bone loss using anatomic specific proximal femoral allografts longer than five centimetres. The patients mean age at the time of the index surgery was 63 (32–86) years. The patients were followed for a mean of 8.4 (5.2–16.6) years, with a five-year minimum follow-up. Results: The mean Harris Hip Score improved from 42.9 points preoperatively to 76.9 points at the last review. Fort-three of the hips (88%) had a successful outcome. Kaplan-Meier survivorship analysis predicted 83% rate of survival at 17 years. Six hips (12.2%) were further revised: four for non-union and aseptic failure of the implant, one for infection, and one for host step-cut fracture. Radiographicly, junctional union was observed in 44 hips (90%). Asymptomatic non-union of the greater trochanter were noticed in three hips (6.1%). Moderate allograft resorption was observed in five hips (10.2%), none were full-thickness graft resorption. The complications include trochanteric escape in three hips, host step-cut fractures in two hips, and four dislocations. Conclusion: We conclude that the good medium-term results with the use of large anatomic-specific femoral allografts justify their continued use in cases of revision hip arthroplasty with severe bone stock loss

We report the use of an allograft prosthetic composite for reconstruction of the skeletal defect in complex revision total hip replacement for severe proximal femoral bone loss. Between 1986 and 1999, 72 patients (20 men, 52 women) with a mean age of 59.9 years (38 to 78) underwent reconstruction using this technique.

At a mean follow-up of 12 years (8 to 20) 57 patients were alive, 14 had died and one was lost to follow-up. Further revision was performed in 19 hips at a mean of 44.5 months (11 to 153) post-operatively. Causes of failure were aseptic loosening in four, allograft resorption in three, allograft nonunion in two, allograft fracture in four, fracture of the stem in one, and deep infection in five. The survivorship of the allograft-prosthesis composite at ten years was 69.0% (95% confidence interval 67.7 to 70.3) with 26 patients remaining at risk. Survivorship was statistically significantly affected by the severity of the pre-operative bone loss (Paprosky type IV; p = 0.019), the number of previous hip revisions exceeding two (p = 0.047), and the length of the allograft used (p = 0.005).

We report the outcome at ten to 15 years of two-stage revision for hip infection in 99 patients using the Prostalac articulated hip spacer system.

All the patients were contacted to determine their current functional and infection status using the Oxford-12, Short form-12, and Western Ontario and McMaster University Osteoarthritis Index questionnaires. A total of 11 of the 99 patients had a further infection, of whom seven responded to repeat surgery with no further sequelae. The mean interval between the stages was five months (1 to 36). We were able to review 48 living patients, with a mean age of 72 years (46 to 86), 34 (71%) of whom provided health-related quality-of-life outcome scores.

The mean follow-up was 12 years (10 to 15). The long-term success rate was 89% and with additional surgery this rose to 96%. The mean global Western Ontario and McMaster University Osteoarthritis Index score was 80.6 (sd 18.3). The mean Oxford-12 score was 74.0 (sd 22.3), and the mean Short form-12 score was 53.1 (sd 9.4) (mental) and 33.5 (sd 13.5) (physical). The mean satisfaction score was 90.5 (sd 15.3).

Two-stage revision for hip infection using a Prostalac interim spacer offers a predictable and lasting solution for patients with this difficult problem.

Osteogenesis imperfecta is a rare inherited disorder of connective tissue which may present with recurrent fractures which are prone to nonunion and malunion resulting in deformity. Some patients develop osteoarthritis of the hip. Formation of hyperplastic callus after recurrent fractures may deform the shape of the femur and preclude the use of standard implants at joint replacement. Replacement can thus be technically demanding.

We present a case of bilateral hip replacement in a patient with osteogenesis imperfecta and hyperplastic callus which was treated by the use of long femoral allografts and cemented femoral stems.

We reviewed and discussed the results of one hundred and fifty-five proximal femoral allografts for revision total hip replacement at an average of eleven years (range , five to twenty years).We looked at graft survivorship, functional results, radiographic assessment and complications. We concluded that this is an excellent technique for restoration of bone stock in the multiplied revised hip. To review the results of revision hip replacement with use of a proximal femoral allograft (PFA) to restore femoral bone stock. Between 1983 and 1998 a PFA – prosthetic composite was used to revise two hundred and twenty-seven hips in two hundred and ten patients. Of these two hundred and ten revisions involved an allograft longer than 8 cm. The average age of the patients at the time of revision was sixty-six years. The average number of previous total hip replacements was 2.2. Each patient was assigned a SF-12 score and a Short WOMAC Score. Radiographic assessment for trochanteric union, allograft-host union, component stability, graft fracture and resorption was performed. At an average of 11.2 years ( range, five to twenty years) after the revision, one hundred and eighty-five patients were alive, twenty-five patients had died and thirty were unavailable for follow-up. SF-12 and short WOMAC scores are still being calculated at the time of abstract submission. 4.4% of hips failed due to infection with 50% being successfully revised. Aseptic loosening of the construct occurred in 3.9% and 78% of these were revised at the time of follow-up. Nonunion of the allograft-host junction occurred in 4.8% and dislocation in 8.4%. Success was defined as a stable implant and no need for additional surgery at the time of follow-up. Success was noted in 84% of available cases. At an average of eleven years revision hip arthroplasties with a PFA are performing very well. This technique remains an excellent alternative when confronted with femoral bone loss in the revision hip setting

Periprosthetic fracture management after hip arthroplasty is complicated by poor bone stock and loose femoral components. Using a prospective database, thirty-five fractures treated by proximal femoral allograft reconstruction were identified. Patients treated between 1989–2000 with minimum two- year results, were reviewed at a mean of 3.8 years. Twenty-six fractures were acute, and nine had failed previous treatment. Union of the PFA was achieved in all but five cases (83%). In twenty-eight cases (78%) no further surgery was required and patient ambulation was pain free. When conventional treatment is not possible, the use of a PFA provides encouraging results. Periprosthetic femoral fracture treatment is complicated by comminution, bone loss, and potentially loose femoral components. Treatment can include cast-braces, internal fixation, revision arthroplasty or the use of proximal femoral allograft composites (PFA). This study reports on thirty-five fractures treated with a PFA between 1989–2000. Five patients were lost and twelve patients (33%) were deceased. Follow-up averaged 3.8 years (range 0.1–11.3) with minimum two-year results in surviving patients. Six patients had either Rheumatoid arthritis or DDH with very narrow femoral canals. There were twenty-three acute fractures, five failures of non-operative management, four failures of ORIF and two fractured femoral stems. Fractures were classified by the Vancouver system with: B1–7%, B2–30%, B3–43% and C-20%. Prefracture functional scores revealed that 30% of patients had significant functional impairment and were awaiting revision arthroplasty. Patients had had an average of two previous surgical procedures (range 1–4). The mean length of the PFA was 14cm and union between graft and host bone was achieved in all but five cases (83%). Resorption of the graft was seen in eight cases (27%), lucent lines in six cases (20%) and implant migration in four cases (13%). Post revision arthroplasty Harris Hip and SF-36 scores revealed substantial disability in this patient group. Twenty-eight cases (73%) were deemed successful with patients not requiring further surgery and enjoying pain free ambulation. In fractures with unstable femoral components and inadequate bone stock or very narrow femoral canals few options are available. This technique provides encouraging results and a viable option when conventional treatment is not possible

Periprosthetic fracture management after hip arthroplasty is complicated by poor bone stock and loose femoral components. Using a prospective database, thirty-five fractures treated by proximal femoral allograft reconstruction were identified. Patients treated between 1989–2000 with minimum two- year results, were reviewed at a mean of 3.8 years. Twenty-six fractures were acute, and nine had failed previous treatment. Union of the PFA was achieved in all but five cases (83%). In twenty-eight cases (78%) no further surgery was required and patient ambulation was pain free. When conventional treatment is not possible, the use of a PFA provides encouraging results. Periprosthetic femoral fracture treatment is complicated by comminution, bone loss, and potentially loose femoral components. Treatment can include cast-braces, internal fixation, revision arthroplasty or the use of proximal femoral allograft composites (PFA). This study reports on thirty-five fractures treated with a PFA between 1989–2000. Five patients were lost and twelve patients (33%) were deceased. Follow-up averaged 3.8 years (range 0.1–11.3) with minimum two-year results in surviving patients. Six patients had either Rheumatoid arthritis or DDH with very narrow femoral canals. There were twenty-three acute fractures, five failures of non-operative management, four failures of ORIF and two fractured femoral stems. Fractures were classified by the Vancouver system with: B1–7%, B2–30%, B3–43% and C-20%. Prefracture functional scores revealed that 30% of patients had significant functional impairment and were awaiting revision arthroplasty. Patients had had an average of two previous surgical procedures (range 1–4). The mean length of the PFA was 14cm and union between graft and host bone was achieved in all but five cases (83%). Resorption of the graft was seen in eight cases (27%), lucent lines in six cases (20%) and implant migration in four cases (13%). Post revision arthroplasty Harris Hip and SF-36 scores revealed substantial disability in this patient group. Twenty-eight cases (73%) were deemed successful with patients not requiring further surgery and enjoying pain free ambulation. In fractures with unstable femoral components and inadequate bone stock or very narrow femoral canals few options are available. This technique provides encouraging results and a viable option when conventional treatment is not possible

We investigated 219 revisions of total hip replacement (THR) in 211 patients using a collarless double-taper cemented femoral component. The mean age of the patients was 72 years (30 to 90). The 137 long and 82 standard length stems were analysed separately. The mean follow-up was six years (2 to 18), and no patient was lost to follow-up.

Survival of the long stems to re-revision for aseptic loosening at nine years was 98% (95% confidence interval (CI) 94 to 100), and for the standard stems was 93% (95% CI 85 to 100). At five years, one long stem was definitely loose radiologically and one standard stem was probably loose. Pre-operative femoral bone deficiency did not influence the results for the long stems, and corrective femoral osteotomy was avoided, as were significant subsidence, major stress shielding and persistent thigh pain.

Because of these reliable results, cemented long collarless double-taper femoral components are recommended for routine revision THR in older patients.

Revision arthroplasty after infection can often be complicated by both extensive bone loss and a relatively high rate of re-infection. Using allograft to address the bone loss in such patients is controversial because of the perceived risk of bacterial infection from the use of avascular graft material. We describe 12 two-stage revisions for infection in which segmental allografts were loaded with antibiotics using iontophoresis, a technique using an electrical potential to drive ionised antibiotics into cortical bone.

Iontophoresis produced high levels of antibiotic in the allograft, which eluted into the surrounding tissues. We postulate that this offers protection from infection in the high-risk peri-operative period. None of the 12 patients who had two-stage revision with iontophoresed allografts had further infection after a mean period of 47 months (14 to 78).