Nonunion of the humerus with bone loss and shortening due to osteomyelitis is rare but difficult to treat. We describe our experience with a callus distraction technique using a monolateral external fixator for the treatment of this condition.

Between October 1994 and January 2004, 11 patients were treated. There were seven males and four females, with a mean age of 14 years (10 to 17). The mean bone loss was 1.9 cm (1 to 2.7) and the mean length discrepancy in the upper limb was 5.6 cm (3.5 to 8.0).

The mean follow-up was for 106 months (54 to 166). The mean external fixation index was 34.8 days/cm (29.8 to 40.5). The mean lengthening was 9.5 cm (5.5 to 13.4). There were seven excellent results, three good and one poor. There were nine excellent functional results and two good.

The treatment of humeral nonunion with bone loss and shortening due to osteomyelitis by callus distraction is a safe and effective means of improving function and cosmesis.

We have previously described the mid- to long-term results of conventional simple varus intertrochanteric osteotomy for osteonecrosis of the femoral head, showing that 19 of the 26 hips had good or excellent results. We extended the follow-up to a mean of 18.1 years (10.5 to 26) including a total of 34 hips in 28 patients, with a mean age at surgery of 33 years (19 to 53). There were 18 men and ten women and 25 hips (74%) had a satisfactory result with a Harris hip score ≥ 80. In all, six hips needed total hip replacement (THR) or hemiarthroplasty. The collapse of the femoral head or narrowing of the joint space was found to have progressed in nine hips (26%). Leg shortening after osteotomy was a mean of 19 mm (8 to 36). With conversion to THR or hemiarthroplasty as the endpoint, the ten-year survival rate was 88.2% (95% confidence interval (CI) 82.7 to 93.7) and the 20-year survival rate was 79.7% (95% CI 72.1 to 87.3); four hips were converted at ten years and other two hips were converted at 20 years.

Shortening of the leg after osteotomy remains a concern; however, the conventional varus half-wedge osteotomy provides favourable long-term results in hips with less than two-thirds of the medial part of the femoral head affected by necrotic bone and with normal bone superolaterally.

Aims. The epiphyseal approach to a chondroblastoma of the intercondylar notch of a child’s distal femur does not provide adequate exposure, thereby necessitating the removal of a substantial amount of unaffected bone to expose the lesion. In this study, we compared the functional outcomes, local recurrence, and surgical complications of treating a chondroblastoma of the distal femoral epiphysis by either an intercondylar or an epiphyseal approach. Methods. A total of 30 children with a chondroblastoma of the distal femur who had been treated by intraregional curettage and bone grafting were retrospectively reviewed. An intercondylar approach was used in 16 patients (group A) and an epiphyseal approach in 14 (group B). Limb function was assessed using the Musculoskeletal Tumor Society (MSTS) scoring system and Sailhan’s functional criteria. Results. At final follow-up, the mean MSTS score was 29.1 (SD 0.9) in group A and 26.7 (SD 1.5) in group B (p = 0.006). According to Sailhan’s criteria, the knee function was good and fair in 14 (87.5%) and two (12.5%) patients of group A, and eight (57.1%) and six (42.9%) patients of group B, respectively (p = 0.062). The lesion had recurred in one patient (6.2%) in group A and four patients (28.6%) in group B. Limb shortening > 1 cm was recorded in one patient (6.2%) from group A and six patients (42.8%) from group B. Joint degeneration was noted in one patient from group A and three patients from group B. Conclusion. An intercondylar approach to a chondroblastoma of the middle two-quarters of the distal femoral epiphysis results in better outcomes than a medial or lateral epiphyseal approach: specifically, better limb function, a lower rate of recurrence, and a lower rate of physeal damage and joint degeneration. Cite this article: Bone Joint J 2024;106-B(2):195–202

Osteonecrosis of the femoral head after femoral neck fracture (ONFHpoFNFx) poses challenges in children, particularly at Ficat III stage. Limited effective treatments are available. This study explores basicervical femoral neck rotational osteotomy (BFNRO) for ONFHpoFNFx in children and adolescents and evaluates its outcomes. Children and adolescents with ONFHpoFNFx (Ficat stage III) underwent BFNRO at our center from June 2017 to September 2022 were included. Follow-up exceeded 1 year, with data on modified-Harris-hip-score (mHHS), range of motion (ROM), patient satisfaction, femoral head collapse, necrotic area repair, leg-length, and osteoarthritis progression recorded. This study included 15 cases (15 hips), with 8 males and 7 females, averaging 12.9 years in age (range: 10–17 years). Nine cases had BFNRO alone, and six had combined PAO. Rotation angles varied from 70° to 90° for anterior rotation and 110° to 135° for posterior rotation. Nine patients had femoral neck fixation in a varus position (10° to 30°). The postoperative contour of the weight-bearing area of the femoral head has significantly improved in all patients. With an average follow-up of 28.6 months (range: 12.2–72.7 months), mHHS significantly improved (65.2 to 90.2, P<0.001). Only one patient showed femoral head collapse. Patients experienced no/mild hip pain (VAS=0-3), slight restriction in range of motion, and mild limb shortening. Two patients showed osteoarthritis progression. No infections, joint replacements, or nerve injuries were observed. Even in cases of ONFHpoFNFx in the late stage, BFNRO in children and adolescents can still yield positive early to mid-term results by relocating the necrotic area and restoring the integrity of the anterior-lateral column of the femoral head, thereby preventing femoral head collapse and delaying the onset of severe osteoarthritis

The results of total hip arthroplasty (THA) revisions to correct leg length discrepancy (LLD) are not clear, with only two former limited series (< 25 patients). Therefore, we conducted a retrospective study of THA revisions for LLD to determine: 1) the change in LLD, 2) the function outcomes and whether obtaining equal leg lengths influenced function, 3) the complication and survival rates. This multicenter study included 57 patients: 42 THA revisions for limb shortening and 15 revisions for limb lengthening. LLD was measured on conventional radiographs and EOS. The Oxford-12 and FJS outcome scores were collected and the number of patients achieving the Oxford-12 MCID. The revisions were carried out a mean of 2.8 years after the index THA. The median LLD decreased from 7.5 mm (IQR: [5;12]) to 1 mm (IQR: [0.5;2.5]) at follow-up (p=0.0002). Overall, 55 of 57 patients (96%) had < 5 mm LLD at follow-up and 12 patients (21%) had equal leg lengths. The complication rate was 25%: 12 mechanical complications (8 periprosthetic femoral fractures, 2 stem loosening and 1 cup loosening, 1 dislocation) and 1 periprosthetic infection. The patient satisfaction was high with a median FJS of 79.2/100 and 77% of patients reached the Oxford-12 MCID. Lengthening procedures had significantly worst function than shortening (38% vs 91% of patients achieving the Oxford-12 MCID (p=0.0004)). Survivorship was 85% (95% CI: 77.9 – 92.5) at 2 years and 77% (95% CI: 66.3 – 87.1) at 4.6 years when using re-revision for any reason as the endpoint. When LLD after THA does not respond to conservative management, revision THA should be considered. Although revision THA for LLD improved medium-term functional outcomes with a high patient satisfaction rate, especially for shortening procedures, the complication rate was high, particularly related to periprosthetic femoral fracture

We aimed to use data from a randomized controlled trial (RCT) comparing the sliding hip screw vs. intramedullary nailing (IMN) for trochanteric fractures to examine complication rates between those managed with a short vs. long IMN. This is a secondary analysis using one arm of an RCT of patients ≥18 years with trochanteric fractures. We examined differences in fracture-related (femoral shaft fracture, implant failure, surgical site infection (SSI), nonunion, limb shortening, and pain) and medical (organ failure, respiratory distress, stroke, deep vein thrombosis [DVT] gastrointestinal upset, pneumonia, myocardial infarction, sepsis, or urinary tract infection) adverse events (AE), and readmission between short vs. long IMNs. We included 412 trochanteric fracture patients, 339 (82.2%) of whom received a short (170mm–200mm) nail, while 73 (17.7%) received a long (260mm–460 mm) nail. Patients in the long group were more likely to be admitted from home (vs. an institution), and have comorbidities, or more complex fracture types. Patients in the long group had higher rates of fracture-related AE (12.3%) vs. the short group (3.5%). Specifically, SSI (5.5% vs. 0.3%) and pain (2.7% vs. 0.0%) were significantly higher in the long group. Patients in the long group were also more likely to develop DVT (2.7% vs. 0.3%), and be readmitted to the hospital (28.8% vs. 20.7%). Following covariable adjustment, long nails remained associated with a higher odds of fracture-related AE (5.11, 1.96–13.33) compared to short nails. We found no association between the adjusted odds of readmission and nail length (1.00, 0.52–1.94). Our analyses revealed that trochanteric fracture patients managed with long IMN nails may have a higher odds of fracture-related AE compared to short nails. Future research is required to validate these findings with larger event rates, and further optimize IMN for trochanteric fracture patients

Introduction. Orthopaedic rehabilitation of adolescences and young adults with high dislocation of the femur is rather challenging. The role of palliative salvage procedures is controversial enough in the cohort of patients. Materials and Methods. Treatment outcomes of 10 patients with congenital hip dislocation were analyzed. Mean age at intervention was 17,8 years (15–22). The grade of dislocation were assessed according to Eftekhar: type C − 2, type D − 8. The mean baseline shortening was 4.7±0.36 cm. All subjects underwent PSO with the Ilizarov method. Another osteotomy for lengthening and realignment was produced at the boundary of the upper and middle third of the femur. The mean time in the Ilizarov frame was 5.3 months. Results. The mean follow-up was 2.6±.1 years (range, 15 to 32 years). Limb shortening of 1 cm to 1. 1. /. 2. cm was observed in four cases. Functional outcomes according to d'Aubigne-Postel were: Pain 4,4±0,15 points. ROM − 4,1±0,3 points. Walking ability − 4,5±0,2 points. Two cases had good results (15–17 points), and seven patients had fair outcomes (12–14 points). A poor result (7 points) was recorded in one female patient 28 years after PSO followed by THA. Conclusions. Hip reconstruction with the Ilizarov method can be used in specific clinical situations as an alternative salvage procedure to delay THR in young patients with high dislocation of the femur

Introduction. Congenital posteromedial bowing of tibia (CPMBT) is characterized by a decreasing deformity and an increasing limb shortening. Our series compares the duration and complications of lengthening in younger vs older children. Materials and Methods. We studied 28 tibial lengthenings in 23 patients, divided into two equal groups of 14 segments: group-A ≤5 years (preschool) and Group-B >5 years. We measured preoperative (bo) and postoperative (po) sagittal, coronal, and oblique plane deformities, limb length discrepancy (LLD), amount of lengthening (AmtL), percentage lengthening (%L), external fixator duration (EFD) and external fixator index (EFI). Complications were graded by Lascombes’ criteria, results by ASAMI Bone score. Results. Mean age= 8.8 ± 7.1 years; mean follow-up= 7.9 years. Group-A had significantly greater bo-sagittal, coronal, and oblique plane deformities. LLD (3.4 cm in group-A vs. 4.1 cm in group-B) was similar in both. LLD at maturity (LLD. M. ) by multiplier method ranged 4.4–9.5 cm in group-A and 2.5–9.7 cm in group-B. Though AmtL (3.5 cm and 4.1 cm) was similar in both, %L was 24% in group-A and 15.7% in group-B (p=0.002). EFD (116.6 days) and EFI (33.7 days/cm) were lesser in group-A vs group-B (200.3 days, p=0.001; 50.2 days/cm, p=0.01). Lascombes’ criteria were fulfilled in 11/14 lengthenings in group-A vs. 3/14 in group-B. ASAMI bone score was good and excellent in both groups (p=0.44). Conclusions. Younger children with large deformities and LLD. M. could be safely lengthened with lesser external fixator duration and complications than in older children

Background. In poliomyelitis; hand to knee gait is the sum of quadriceps weakness and fixed knee flexion deformity. Limb shortening is another added problem. Usually, each problem is attacked separately; with variable end results and complication rates for each procedure. Patients and methods. 22 patients (16–46 y.); with poliomyelitis with hand to knee gait due to fixed knee flexion deformity of mild to moderate degree (10–400); and limb shortening of 4.5–9.5 cm., were managed simultaneously by a single operation. A distal femoral metaphyseal anterior closing wedge (recurvatum) corrective osteotomy was done to treat acutely the fixed knee flexion deformity(and subsequently hand to knee gait). A modified Wagner or Orthofix frame was applied as a mono-plane mono-axial lengthening device to stabilize the osteotomy and to lengthen the short limb. Lengthening was started in all cases two weeks post-operatively. Results. All cases showed full correction of the deformity with elimination of the hand to knee gait problem and all showed the desired limb length equalization. The femora were lengthened to an average of 6.4 cm.(= 21 % of original femoral bone length), range of lengthening was: 4–9 cm. (= 11–26 %). The average healing index was 37 days per centimeter (with a range of 31 to 53 days per centimeter.). Conclusion. Simultaneous management of hand to knee and limb shortening could be achieved by a single osteotomy which is used to extend the knee acutely and for distraction two weeks later using a mono-lateral frame. Results were good, complications were few and accepted and the procedure is well tolerated by the patients

Background. In poliomyelitis; hand to knee gait is the sum of quadriceps weakness and fixed knee flexion deformity. Limb shortening is another added problem. Usually, each problem is attacked separately; with variable end results and complication rates for each procedure. Patients and methods. 22 patients (16–46 y.); with poliomyelitis with hand to knee gait due to fixed knee flexion deformity of mild to moderate degree (10–400); and limb shortening of 4.5–9.5 cm., were managed simultaneously by a single operation. A distal femoral metaphyseal anterior closing wedge (recurvatum) corrective osteotomy was done to treat acutely the fixed knee flexion deformity(and subsequently hand to knee gait). A modified Wagner or Orthofix frame was applied as a mono-plane mono-axial lengthening device to stabilize the osteotomy and to lengthen the short limb. Lengthening was started in all cases two weeks post-operatively. Results. All cases showed full correction of the deformity with elimination of the hand to knee gait problem and all showed - the desired - limb length equalization. The femora were lengthened to an average of 6.4 cm.(= 21% of original femoral bone length), range of lengthening was: 4–9 cm. (= 11–26%). The average healing index was 37 days per centimeter (with a range of 31 to 53 days per centimeter.). Conclusion. Simultaneous management of hand to knee and limb shortening could be achieved by a single osteotomy which is used to extend the knee acutely and for distraction two weeks later using a mono-lateral frame. Results were good, complications were few and accepted and the procedure is well tolerated by the patients

A variety of surgical procedures are reported for the management of large volumetric bone loss about the ankle. Although the success rates of these various methods are generally adequate for fusion, they commonly utilize autogenous bone graft and usually result in limb shortening. In seven patients a titanium spinal cage was utilized as a structural support augmented with bone graft for complex ankle arthrodesis. This technique offers immediate structural support, maintenance of limb length, and limits autogenous graft morbidity. Early results of ankle arthrodesis with this cage are encouraging with regard to fusion rates, aesthetic attribute and functional outcome. To present a novel method of tibio-talar and tibio-calcaneal arthrodesis utilizing a titanium spinal cage for the management of bone loss. Complex ankle arthrodesis management with a spinal cage offers reliable fusion rates (comparable with existing techniques) without limb shortening. A variety of procedures are reported for the challenging management of bone loss in ankle arthrodesis. Although the success rates of these various methods are generally good, they commonly utilize large autogenous bone grafts and are subject to donor site morbidities. Most methods also require significant shortening with commensurate functional and aesthetic deficits. Early results of ankle arthrodesis with a titanium spinal cage are encouraging with regard to fusion rates and functional outcome. A cohort of seven patients treated by this technique has resulted in a 100% fusion rate without loss of limb length. Hindfoot and SMFA scores have revealed good functional results for a usually devastating problem. A titanium cage was utilized as a structural support for complex ankle arthrodesis with large volumetric bone loss. The cage was contoured at the time of each operation to fit in situ. This technique offers immediate structural support and reliable fusion while limiting autogenous graft morbidity and limb shortening

Our clinical experience in treating lower limb deformities by external fixation started in 1982 by applying an Ilizarov external fixation frame. Correcting lower limb deformity by gaining the proper length and the optimal mechanical performance is of the outmost importance and essential for valid restoration of the articular function. We treated 145 patients: 25 patients affected by post-traumatic deformities; 12 patients affected by axial deviation of the knee; 82 patients affected by limb shortening and associated limb deformities; eight patients with lower limb deformities and shortening following pathological hip alterations; and 18 patients affected by severe foot deformities. It is important to stress that in congenital lower limb deformities both axial deviation and limb shortening contribute to the final picture of the deformity. Combined (hybrid) external fixation provides an adequate correction of the mechanical axis and a proper lengthening of the shorter limb by a corticotomy followed by a gradual distraction

Aims: The aim of our study was to assess results of treatment of trochanteric fractures using a speciþc algorythm Methods: From January 2000 to March 2002, 215 patients affected by fractures of trochanteric region were treated. 130 were female and 85 male, the average age was 72,5 (5–96). Author used an algorythm based on the analysis of stability of the bone-device system considering three groups: GROUP 1: stable fractures (A1 and A2.1 according to AO classiþcation) in which dynamic hip screw was employed, GROUP 2: unstable fractures (A2.2, A2.3, A3.1) in which dynamic hip screw and trochanteric stabilizing plate were employed, GROUP 3:unstable fractures (A3.2 and A3.3) in which intramedullary nailing was performed. Results: The average healing time was 6 weeks for GROUP 1 and GROUP 2 and 7,2 weeks for GROUP 3. According to the HHS at 12 months the average score for GROUP 1 was 83,1, with 72% of excellent results, 20% good, 4% fair and 4% poor; for GROUP 2 the average score was 79,7 with 56% of excellent results, 25% good, 10% fair and 9% poor; for GROUP 3 the average score was 77,2 with 40% of excellent results, 29% good, 18% fair and 13% poor. In GROUP 2 we observed 2 cases of limb shortening more then 2 cm and 9 cases of pain due to trochanteric plate. In 2 cases of cut-out of the lag screw cemented endoprosthesis was required. In GROUP 3 we observed 5 cases of limb shortening more then 2 cm and 4 cases of persistent trochanteric pain. Conclusions: Following this algorythm a low rate of complication were observed specially in GROUP 2

The rate of periprosthetic joint infections (PJI) after primary total hip arthroplasty (THA) is approximately 1%. As the number of THAs performed each year continue to increase (550,000 by 2030), a corresponding increase in the number of hip PJI cases is likely to occur. A chronic deep infection may be treated by either chronic suppression, irrigation and debridement, single-stage exchange, or two-stage exchange. In the United States, the gold standard for chronic PJI continues to be a two-stage exchange. The benefit of an antibiotic impregnated cement is that they produce higher local concentrations of antibiotics than systemic intravenous administration. Hip spacers may be either static or articulating. Static spacers are reserved for cases of massive acetabular bone loss in which an articulating spacer is not feasible. A static spacer consists of a block of antibiotic cement in the native acetabulum and antibiotic coated rod in the femoral canal. Limb shortening, loss of soft tissue planes, and disuse osteopenia and muscle atrophy are all limitations of static spacers. In contrast, articulating spacers fulfill the goals of the interim construct during two-stage exchange which is to enhance eradication of the infecting organism through drug elution, to maintain limb length, to facilitate exposure during revision surgery, and to improve functional mobilization. Articulating spacers may be divided into three general categories based on method of spacer creation: Handmade custom spacers, prefabricated spacers, custom molded spacers (hemiarthroplasty molds and molded stem with cemented all-polyethylene cup). Handmade custom spacers are usually created with K-wire or rush rods coated with antibiotic cement. Handmade spacers are relatively simple to make, they are economical, and the amount and type of antibiotics incorporated can be customised for the infecting organism. Commercially available hemiarthroplasty spacers can be either prefabricated (Spacer G, Exactech, Gainesville. FL) or made intraoperatively (Stage One, Zimmer Biomet, Warsaw, IN) are available in several head and stem sizes. The advantage of prefabricated spacers is that they do not require additional time to mold in the operating room. The downside of prefabricated spacers is that the antibiotic concentration and type is predetermined. A custom molded stem with cemented all-polyethylene cup can be made with off the shelf implants or used as part of a commercially available spacer (PROSTALAC, DePuy Synthes, Warsaw, IN). A common antibiotic/cement combination includes Tobramycin (3.6 g/40 g of cement) and vancomycin (1.5 g/40 g of cement). In all of these spacer constructs, the principles of using a high-elution cement mixed without a vacuum and with high doses of heat stable antibiotics are consistent. Tobramycin works synergistically to improve Vancomycin elution properties and is usually added in higher doses. Overall infection eradication is similar between all categories of spacers and range between 90–97%. Complications after placement of an articulating spacer are often specific to the type of spacer used. Handmade spaces that have K-wires for support are at risk for spacer cement fracture. Spacer dislocation is also a common complication in up to 15% of cases with all types of spacers. In addition, periprosthetic fractures can occur postoperatively in up to 10% of patients. Overall, despite this complication profile, articulating antibiotic spacers have excellent rates of infection eradication and offer improved mobilization in the interim two stage period and reduce operative time, complexity, and morbidity during reimplantation

Early reports on revision total hip arthroplasty (RTHA) suggested that outcomes of this procedure are as good as those of primary total hip arthroplasty (THA). However, RTHA is associated with longer surgery time, greater blood loss and increased risk of complications (thromboembolism, nerve injury, periprosthetic fractures, recurrent hip dislocations and infections). Aseptic loosening after RTHA was reported in 36% of patients aged over 55 years within 4 years after revision. Infections were reported in 32% and complications during surgery in 23% of patients respectively. Unsatisfactory results of RTHA stimulate the search for alternative procedures. Girdlestone excision arthroplasty (GA) seems to be a good solution for older patients with high risk of complications related to a poor general condition. Material and method Between 2000 and 2003 we operated 39 patients, 10 for septic (26%) and 29 aseptic (74%) loosening of their THA. All patients complained of painful limb aggravated by weight bearing and the severity of pain was the main indication for the surgery. Average survival time of previous THA was 9 year (range: 1 to 20). We assessed pain, walking distance and the need to use walking aids. The outcomes were measured according to the Harris Hip Scale. The patients had the GA performed. The procedure involved removing implant and bone cement and placing the major trochanter into bone acetabulum. If an infection was present, an antibiotic irrigation system was introduced. No cast or braces were used and walking was started 2–7 days after surgery, depending on patients general condition. Results Good pain control was reported by 33 (85%) patients. The average Harris Hip Score changed from 25 points preoperatively to 53 at latest follow-up. Average limb shortening was 4 centimetres (range: 2 to 8). Walking aids (one or two crutches) were required by all patients. Eighteen (46%) patients walked more than 500 m, 12 (31%) patients walked 200–500 m and 9 (23%) patients walked less than 200 m, of whom one patient was wheelchair bound. Infection ceased in 9 cases, 1 patient died because of complications related to chronic infection. Discussion GA yields satisfactory results in patients who have to have their prostheses removed. It provides a mobile, painless joint. The disadvantages of GA are: limb shortening and unstable gait which requires the use of crutches. This procedure should be indicated for patients with high risk of complications due to poor general health, infection and/or massive loss of bone stock which render more invasive procedures impossible. GA is also advisable in patients with weak hip abductor muscles, when RTHA is associated with a high risk of recurrent hip dislocation. The Girdlestone arthroplasty is a satisfactory salvage procedure in most cases of failed THA, when the choice of reimplantation exposes the patient to a high risk of further failure

Aims

The ulna is an extremely rare location for primary bone tumours of the elbow in paediatrics. Although several reconstruction options are available, the optimal reconstruction method is still unknown due to the rarity of proximal ulna tumours. In this study, we report the outcomes of osteoarticular ulna allograft for the reconstruction of proximal ulna tumours.

Aims

Ilium is the most common site of pelvic Ewing’s sarcoma (ES). Resection of the ilium and iliosacral joint causes pelvic disruption. However, the outcomes of resection and reconstruction are not well described. In this study, we report patients’ outcomes after resection of the ilium and iliosacral ES and reconstruction with a tibial strut allograft.

Autologous bone grafting for bone defect reconstruction is associated with complications including donor site morbidity, infection risk, pain and surgical time. Therefore, bone graft substitutes provide an alternative for distinct indications and different characteristics with regard to their mechanical properties and resorption rates. In order to fill the loss of bone substance and to control the infection, we tried the efficacy of Cerament™G, a new absorbable composite of Calcium Sulphate and Hydroxyapatite with Gentamicin. We present 3 male patients aged between 45 and 68 years affected by post-traumatic severe septic non union of femur, tibia and foot. The first patient with femur fracture was involved in a car accident (mixed flora Acinetobacter Baumanii, MRSA and Klebsiella pneumoniae carbapenemase (KPC)-producing), the second patient with femur and foot fracture falled by height in a work accident (MRSA) and the third one had a chronic tibial osteomyelitis several years after a road accident (Pseudomonas Aeruginosa). All 3 patients had undergone previous surgery. The first patient had several operations including multiple bone resection and debridement with external fixator, occlusion of superficial femoral artery with arterial bypass and finally debridement with implantation of Cerament™ G with external fixator and long term antibiotic therapy. The other 2 patients were subjected to resection of tissue septic with debridement, implantation of Cerament™ G and soft tissue closure and systemic antibiotics. Clinical and radiographic outcome were assessed at final follow-up (mean 8 months; range 8–18). The follow-up was 8–18 months with examining clinical, radiographic, CT scan and laboratory tests. The patients had self-limiting fluid leakage. There was no recurrence of infection during the follow-up period. Bone ingrowth occurred in all cases with limb shortening. Cerament™ G gives good elution of antibiotic and allows bone ingrowth. The implantation of Cerament™ G was associated with good clinical outcomes and satisfactory bone consolidation. We acknowledge Antonella Esposito for septic nursing assistance

To present the results achieved with the use of external fixation techniques in treating 52 cases of post-traumatic and post-surgical septic non-union in a low resources setting. From 2006 to 2014 52 patients were treated for post-traumatic septic non-union of the lower and upper limb bones. Clinical records and radiographs were reviewed; telephone interviews were done for patients unable to reach our institution. There were 39 males and 13 females, with an average age at the time of admission of 29 years (the youngest patient was 8 years old, the oldest 81). Tibia was involved in 43 cases (24 right side, 19 left side), femur in 9 (4 right, 5 left) and left humerus in 1. All the patients, except two, had been treated in other institutions before admission. At presentation, 19 patients had an external fixator in situ, 18 patients had infected osteosynthesis, 15 had exposed necrotic bone, with loss of soft tissues. In 20 cases hardware removal, debridement and sequestrectomy were followed by application of an external fixator. In 31 cases bone transport was done; the fixator was monolateral in 27 cases. In 1 case sequestrectomy and external fixation were followed by a vascularized fibula graft. Bifocal bone transport was utilized in one patient while the bone transport procedure was associated to limb lengthening in 10 patients. Plastic surgery was required in 13 patients. Rotational flaps, vascularized free flaps and extensive skin grafts were all used. The site of non union was cured in all the patients, in an average time of 11 months (from 4 to 32). Two patients required an amputation a few months after the end of the treatment. There was need for fixator adjustment and screws replacement in 21 patients. Non-union at the docking point was observed in 5 patients; it was septic in two of them. There was need of skin-plasty in 2 cases for skin invagination. In 3 cases an extensive skin ulcer was observed during the transport procedure. Limb length discrepancy was corrected in 10 patients. A residual limb shortening was observed in 14 patients. Ankle fusion, knee fusion, foot drop, sensory loss in the foot, reduced range of joint motion were also observed. The external fixator plays a pivotal role in the treatment of septic non unions, especially in low resources setting. The treatment is long and costly. Strict medical supervision is necessary during the entire process