To identify the incidence of a cortical breech on the initial presentation X-rays of patients with distal femoral GCTs, and whether this lead to a higher rate of local recurrence of tumour, a prospective database is kept of all patients seen in the unit. Initial presentation X-rays on 54 patients with distal femroal GCTs were reviewed. The size of the tumour was estimated by measuring the largest dimensions of the tumour (depth, breadth & height). The volume of the distal femur was estimated using the same X-ray and computer programme. The X-rays were then carefully studied for evidence of a cortical breach. The records were also checked for evidence of subsequent locally recurrent disease and subsequent surgery.

X-rays were reviewed on 54 patients (29 male, 25 female), range of 18–72 years. All patients had a biopsy-proven GCT of the distal femur, X-rays (prior to biopsy) were reviewed. 34 (63%) patients with a cortical breech on X-ray. The mean tumour volume: distal femoral volumes (TV:DFV) was statistically greater between those patients with a cortical breach and those without, using ANOVA (p< 0.0001). There were 13 patients with local recurrent disease but no statistical difference in subsequent local recurrence rates between the two patient groups. There was also no statistical differences between the number of operations for those who presented with a cortical breach or without. There was no evidence that more radical surgery was required if a patient presented with a cortical breach.

The risk of cortical breech in patients with GCTs of the distal femur is dependant upon the tumour volume to distal femur volume ratio. If the ratio is above 54% then present with a cortical breech on X-ray is likely (95% confidence interval).There is no evidence those patients with a cortical breach have a higher rate of local recurrence, an increased number of operations or more radical surgery.

Conclusion: The risk of cortical breech in patients with GCTs of the distal femur is dependent upon the tumour volume to distal femur volume ratio.

The unusual phenomenon of histological grade change in locally recurrent soft tissue sarcomas is examined by retrospective review of a large sarcoma database. Increased histological grade was found to occur in 20% of recurrent tumours. Several possible factors predisposing to grade change were examined, and only the histologic diagnosis of myxoid malignant fibrous histiocytoma was found to be significant. Despite increased histologic grade, these tumours do not appear to have a worse prognosis in terms of developing systemic disease.

Soft tissue sarcomas (STS) have a reported local recurrence rate of between five and thirty percent. Recurrent tumours are often similar histologically to the initial tumour, however they are occasionally of higher histological grade than the original lesion. Factors that predispose to this change in grade are not known.

We sought to identify the frequency at which locally recurrent STS demonstrate a change in histological grade, and to investigate the possible factors leading to this change. We also investigate whether a change in grade is associated with a poorer prognosis.

We identified one hundred and seventy-three patients who developed locally recurrent STS, one hundred and twenty-four of which met inclusion criteria and who will form the basis of this study. Ninety-two patients (74%) had no change in histological grade, twenty-four (19%) demonstrated an increase in histological grade and eight (7%) a decreased histological grade. Univariate analysis of time to local recurrence, histological diagnosis and use of radiotherapy and chemotherapy did not reveal significant differences between the groups who did and did not undergo change in grade. When the diagnosis of myxoid MFH was looked at separately, there was a higher proportion in the group that developed increased histological grade. Development of a change in grade was not associated with a poorer survival rate.

Increase in histological grade occurs in approximately 20% of locally recurrent STS, but this phenomenon is not associated with a poorer prognosis than if the grade remains the same. A histological diagnosis of myxoid MFH predicts for an increase in histological grade.

Non-union of long bone fractures can be a challenging problem. There are several methods of treatment and they depend upon various patient factors, biology of non-union, and presence of infection. When faced with failure of treatment with biological reconstructive procedures patients have little choice. At our institute we have treated 10 such patients with radical excision and reconstruction using tumour endoprostheses as a last attempt to save the limb.

Median age of the patients was 71 years (25–85). 2 patients were male and 8 were female. Median follow-up was 49 months (8–229). 5 had infected non-union. Resection and massive endoprosthetic reconstruction involved the distal femur in 4 patients, proximal femur 3, distal humerus 2 and total Humerus in 1 patient. Time from diagnosis of non-union to treatment was 0 to 96 months (median 11 months) and patients had had 0 to 6 (median 3) previous operations 5 infected non-unions were operated as 2 stage procedures and received long term antibiotics. 4 out of 5 infected non-unions were salvaged. There were 5 complications, namely periprosthetic fracture, infection, a dislocated shoulder, radial nerve palsy, suture of bosing.

All the patients achieved immediate mobility and stability. Extendible prosthesis allowed partial correction of limb shortening.

Conclusion: Resection of established non-union and reconstruction with endoprostheses is a good salvage operation for elderly and low demand patients in whom time consuming biological reconstruction is not desirable.

Pathological fracture occurs in 5–10% of all primary malignant bone tumours. It is thought that they unfavourably influence survival, because the fracture haema-toma may contaminate adjacent tissues. Management is often more aggressive and one is less inclined to consider limb saving surgery.

Aim of this study was to determine whether the presence of pathological fracture had an effect on rate of limb salvage surgery, role of adjuvant treatment and survival.

A retrospective study was done on all patients with a pathological fracture through localised Ewing’s sarcoma, treated between 1979 and 2001. Of 289 patients with localised Ewing’s sarcoma, 27 had a pathological fracture. Eighteen presented with fracture, in 9 fracture occurred after biopsy. All were treated with chemotherapy according to protocol. Two fractures were already treated by osteosynthesis elsewhere, the rest healed with conservative treatment. After chemotherapy, 20 patients were treated surgically: 19 with limb saving surgery, 1 with amputation. Apart from chemotherapy, treatment was surgery alone in 15, surgery and radiotherapy in 5, and radiotherapy alone in 7 patients. Indications for radiotherapy were close margins, poor chemotherapy response, or pelvic tumours. Surgical margins were wide in 16 patients, marginal in 2, and intralesional in 1 patient. Local recurrence occurred in 2 patients, primarily treated with chemotherapy and radiotherapy alone. Five year survival was 60%, metastasis free survival 59%, both comparable with rates reported in literature.

Conclusion: Chemotherapy allows fractures to consolidate with conservative treatment. Adequate surgical margins can be achieved in the majority of patients with limb saving surgery. Adjuvant radiotherapy does not seem necessary if margins are wide. Survival is not negatively influenced by pathological fracture. The survival rate following limb saving surgery in these patients is similar to that of patients in literature where amputation is done. Limb saving surgery seems a safe option.

Aim: To identify prognostic for patients who develop local recurrence after initial attempted curative treatment for a soft tissue sarcoma (STS).

Method: All patients who developed a local recurrence (LR) after initial primary treatment of a STS were identified from a prospective database. Their management and outcome were analysed to find prognostic factors.

Results: 178 patients were identified. They had a median age at original diagnosis of 53 and 102 of the patients had high grade tumours, 50 intermediate grade and 23 low grade. The median time to LR was 14 months but extended up to 11.5 years. 47 of the patients developed metastases either before or synchronously with the LR. In these patients the median survival was 20 months with only 4% surviving to 5 years. In the 131 patients who did not have identifiable metastases at the time of diagnosis, 74 subsequently developed metastases at a median time of 12 months following the development of LR.

The median survival for patients without metastases at the time of LR was 3 years with a 31% survival at 10 years. The most important prognostic factor in this group was grade with low grade tumours having a much better outlook (70% survival at 10 years) than intermediate or high grade tumours (24% at 10 years). Complete control of the first local recurrence could not be shown to be a prognostic factor.

Conclusion: Local recurrence has a poor prognosis but this is because it frequently arises in patients who have other bad prognostic factors. Whilst obtaining local control is important, overall survival is poor, but not as bad for those patients who develop metastases.

Background The exclusion of infection at the site of a painful or failed prosthetic joint replacement is important for pre-operative planning and counselling. A variety of investigations can be used to assist in the diagnosis or exclusion of infection.

An ESR and CRP are widely used as the initial screening investigation to differentiate between aseptic and septic loosening of prosthetic joint replacements¹. Propionobacteria are organisms of low virulence, although they do cause deep peri-prosthetic infections². We believe that Propionobacteria do not always cause a significant rise in ESR and CRP.

Methods Between May 2001 and May 2004, we identified 78 patients with prosthetic joint replacements colonised with Propionobacteria. There were 48 hip joint replacements, 27 knee joint replacements, 2 endoprosthetic replacements of the femur and 1 shoulder joint replacement. There were 48 males and 30 females. The preoperative values of ESR and CRP were recorded. For the purposes of this study, an ESR rate of 30mm/hr or higher and a CRP level of 10mg/lt or higher were considered to be suggestive of infection and were deemed a positive result.

Results All of the 78 patients had both ESR and CRP measured preoperatively. In only 17 patients (22%) both ESR and CRP were higher than 30mm/hr and 10mg/l respectively. In 33 patients (42%) with prosthetic joint replacements colonised with Propionobacteria, the preoperative values of both ESR and CRP were normal.

Conclusion In our study we have shown that 33 out of 78 patients (42%) with prosthetic joint replacements colonised with Propionobacteria had normal preoperative of both ESR and CRP values. This is to suggest that normal preoperative values of ESR and CRP in suspected failed prosthetic joint replacements might not exclude infection, if the causative organism is of low virulence such as Propionobacteria.

Aims: To establish the frequency and demographics of soft tissue sarcomas (STS) presenting in the lower limb.

Methods: Patients presenting to a tertiary referral orthopaedic oncology unit over a 10-year period were prospectively entered into a computerised database. The site of primary STS and demographic details were also recorded.

Results: 1519 STS in all body regions were treated. 1067 (70.2%) within the lower limb. 57.0% thigh, 13.0% calf, 8.2% foot and ankle, 7.7% buttock, 5.7% knee, 4.6% pelvis and 3.8% in the groin. There was a male predominance (56.2%). M:F ratio was 2.5:1 for the groin and 1.3:1 for the thigh with the other body regions approximately equal.

Conclusion: The majority of STS are found in the lower limb. In this large series there was a male predominance most marked in groin presentations.

Aims: Soft tissue sarcomas (STS) of the foot and ankle are rare tumours. The aims of this study were to examine the presenting features and highlight those associated with a delay in diagnosis.

Methods: Patients presenting during a 10-year period were identified using a computerised database within the Orthopaedic Oncology Unit at the Royal Orthopaedic Hospital, Birmingham, UK. Additional information was obtained from a systematic case note review.

Results: 1519 patients were treated for STS of which 87 (8.2%) had tumours sited in the foot and ankle. Of these, 75 (86.2%) had presented with a discrete lump (42 (56%) of them having an inadvertent “whoops” excision biopsy), 3 (3.4%) with ulceration and the remaining 9 (10.3%) with symptoms more commonly associated with other benign foot and ankle pathology. Within the group of 9 patients they had previously been treated as plantar fasciitis (3), tarsal tunnel syndrome (2), Morton’s neuroma (1) and none specific hind foot pain (3). Median delay from onset of symptoms to diagnosis as STS was 26 months for this group (mean 50; range 6–180 months) compared to 12 months (mean 32; range 3–240) for the “whoops” biopsy group and 10 months (mean 16; range 2–60 months) for the unbiopsied discrete lump group.

Conclusion: Soft tissue sarcoma in the foot and ankle may present insidiously and with symptoms of other benign pathologies. Failure to respond to initial treatment of suspected common benign pathology should be promptly investigated with further imaging e.g. MRI scan or high resolution ultrasound, or with specialist consultation.

To establish whether Patients or Medical Professionals are the main source of delay for patients referred to a Specialist Centre for Soft Tissue Sarcoma.

Methods: Patients were recruited from both outpatient clinics and from the surgical ward. A semi-structured interview was used to take a detailed history of the patients’ treatment pathway, before arriving at the Specialist Centre. Results: The average time for patient to present to a SC from the onset of symptoms was 110 weeks, (min 3 days, max 1089 weeks), with a median of 40 weeks. Average delay to presentation to a medical professional (patient delay) was 24.5 weeks (min 0, max 530), median 2 weeks. Average delay in referral to a SC (service delay) was 84 weeks (min 0 max 1083), median 25weeks.

Discussion: Medical professionals rather than patients contribute the greatest source of delay in patients reaching a Specialist Centre for treatment of Soft Tissue Sarcoma. Adherence to previously published guidelines could decrease this delay. Medical professional awareness of these guidelines and their contents needs to be increased.

Purpose: To identify the risk of metastases at the time of diagnosis in patients with soft tissue sarcomas and to estimate the cost effectiveness of identifying these.

Methods: A retrospective database review was used to identify all new soft tissue sarcoma patients referred to our unit and to find those identified to have metastases at diagnosis. Data of tumour size, depth, grade, age, type of tumours, Chest x-ray (CXR)/CT chest results were available in all patients. We estimated the efficacy of CXR in identifying metastases and the costs of various staging strategies.

Patients: 1170 with newly diagnosed STS in 7.5 years (1996–2004) were included.

Results: The incidence of metastases at diagnosis was 10% (116 patients), 8.25% (96 patients) had lung metastases and 20 had metastases elsewhere. The risk of having lung metastases at diagnosis was 11.8% in high grade tumours, 6.95% in intermediate grade and 1.2% in low grade tumours. The risk increased almost linearly with size at presentation and was higher in deep tumours and older patients. CXR alone detected 2/3 of all lung metastases. The positive predictive value of the CXR was 93.7%, the negative predictive value was 96.7%, the sensitivity 62.5% and the specificity 99.6%.

The accuracy was 96.9%. CT overestimated metastases in 4%.

Discussion: We recommend that all patients with a newly diagnosed STS should have a CXR and only those with an abnormality or who have large, deep high grade tumours should have a CT chest. This strategy will save £7500 per 100 new patients with STS and will detect 93% of all chest metastases, missing 1 patient with metastases per 166 patients.

Limb preserving surgery following segmental resection of the distal end of the radius and its articular surface presents a major challenge. We have studied 11 consecutive patients with aggressive tumours located in the distal radius that required segmental resection of the distal radius and its articular surface to evaluate the clinical and functional outcome of reconstruction of such defects.

The mean age at the time of diagnosis was 33 years (7–60). Follow up ranged from 12 to 306 months (median 56). Histological diagnosis was osteosarcoma in 4 patients, chondrosarcoma in 2, giant cell tumour in 5 and meta-static carcinoma in 1 patient. Four patients received chemotherapy. The length of excised bone ranged from 6 to 14cm. Reconstruction was performed with non-vascularised proximal fibula strut graft in 6 patients, ulna transposition in 3 and custom made endoprosthesis in 2 patients. The wrist joint was arthrodesed in 5 patients.

At the time of review 2 patients had died of disease, one was alive with disease and 8 were alive and free of disease. Non-union of the graft occurred in one patient, reflex sympathetic dystrophy in 2 and prosthetic dislocation in one. One patient had local recurrence. Four patients required further surgery including one patient who needed an amputation for severe reflex sympathetic dystrophy, one graft revision for non-union, one secondary wrist arthrodesis and one closed reduction of dislocated endoprosthesis. Patients without arthrodesis often had clinical and radiological signs of wrist instability. The majority of the patients achieved satisfactory function with little or no discomfort and ability to perform activities of daily living.

We conclude that limb salvage surgery is worthwhile in patients with resectable tumours of the distal radius.

Purpose: The purpose of this study was to determine how patients with soft tissue sarcoma are followed up in the United Kingdom to inform the development of a prospective clinical trial.

Methods: A list of clinicians (surgeons and oncologists) treating patients with soft tissue sarcomas in the United Kingdom was compiled and a postal survey performed. Reminders were sent to non-responders. The survey included questions about the specialty of the clinician, the grade, membership of specialist societies, perceptions about risk factors for recurrence and the value of follow up and asked specifically about three clinical scenarios.

Results: Of 192 clinicians who were sent the questionnaire, responses were obtained from 155 (81%). 128 of these met the criteria for analysis. In the given clinical scenarios, length of follow up varied from 1 year to lifelong. The total number of clinic visits in 5 years varied from 5 to 30, of chest radiographs from 0 to 24, of chest CT scans from 0 to 10, and of local site imaging from 0 to 13. 88 (84%) agreed that follow up is of benefit. 57 (59%) agreed that it would be reasonable to follow up selected patients in the community. 96 (93%) agreed that a study of follow up protocols would be of value.

Discussion: There is significant variation in follow-up protocols amongst clinicians in the United Kingdom. A prospective study of follow-up protocols is likely to be supported.

Myxoid liposarcoma (MLS) is an unusual type of soft tissue sarcoma as it tends to metastasize frequently to sites other than the lungs. This study was aimed to investigate the natural history of patients with MLS to try and identify prognostic factors which could help predict outcome and aid earlier detection of metastases.

Data was prospectively collected from patient notes and analysed retrospectively. Prognostic factors and metastatic pattern were examined using Kaplan-Meier curves. There were 124 patients with MLS, aged between 28 and 93, the median size of the tumours was 12cm and the most common site was the thigh. Following treatment with excision and radiotherapy the 5yr survival was 65%. Survival was related to younger age (p=0.010) and proximal site (p=0.003) and was also related to the % round cell component of the tumour but was not related to either size or depth of the tumour. Site and margins of excision were significant prognostic factors for local recurrence of disease. 32% of patients developed metastases, of whom 18 cases (46.2%) developed pulmonary metastases and 21 (53.8%) developed extra pulmonary metastases. The sites of these varied hugely and was not significantly related to the site or size of the primary tumour. There was no difference in time to develop metastases or in overall survival between the two groups. Median survival following metastases was 24 months.

Although MLS has an unusual pattern of metastases the site of metastases does not predict a better or worse outcome. Intensive follow up for extraskeletal metastases is probably not justified until they become symptomatic.

Purpose: The aim of this study was to understand the effect of endoprosthetic reconstruction in treatment of solitary bone plasmacytoma threatening structural integrity of bone.

Materials and methods: We retrospectively studied 11 patients who underwent endoprosthetic reconstruction for solitary bone plasmacytoma between 1988 and 2003 with more than 1 year follow up. Most had radiotherapy and those who sustained structural damage to a joint or thought to be salvageable were treated with endoprosthetic replacement.

Results: There were 7 males and 4 females, with M: F ratio of 1.75:1, the median age at diagnosis was 53.61years (35–74). Average duration of symptoms prior to presentation at oncology unit was 7.27 months. We had 4 proximal femoral, 2 pelvic, 4 humeral and one tibial Plasmacytomas that were treated with endoprosthetic replacements.8/11 had preoperative radiotherapy for at least 4weeks and 3 did not. Two had postoperative radiotherapy and one adjuvant chemotherapy. Average follow up is 5.45 years (range 1–16years). We had one death due to unrelated causes, one progression to Myeloma treated with adjuvant chemotherapy, two revisions and one dislocation which was reduced by open method. The cumulative overall survival for all patients was 91% at 5 years. The cumulative risk of failure of reconstruction including; infection, dislocation, local recurrence/progression to Myeloma was 27% at 5 years.

Conclusion: Literature review shows that nearly 53% of SBP progress despite radiotherapy to Myeloma at a median time of 1.8 years (2–4 years). But despite average follow up of 5.45 years, progression to Myeloma after endoprosthetic replacement at our unit is 9.09%. We concluded that the use of endoprostheses for reconstruction after excision of solitary bone plasmacytomas threatening structural integrity of bone combined with radiotherapy decreases the disease progression to Myeloma than radiotherapy alone and offers a reasonable but not absolute chance of cure.

Large benign lytic lesions of the proximal femur present a significant risk of pathological fractures. We report our experience of treating 9 consecutive patients with such defects treated with curettage and fibula strut grafting without supplementary osteosynthesis to evaluate the outcome of this type of reconstruction..

The mean age at the time of diagnosis was 13 years (8–21). Follow up ranged from 2 to 215 months (median 15). Histological diagnosis was fibrous dysplasia in 10 patients and unicameral cyst in 2. All the patients were at risk of pathological fracture. None of the patients developed pathological fracture after surgery and the lesions consolidated fully within one year. Local recurrence occurred in one patient (8%). Minor donor site complications occurred four patients.

All the patients were able to fully weight usually within 3 months of surgery.

At the time of review all but one patient were completely asymptomatic and fully weight bearing. The only symptomatic patient was the patient with local recurrence which has recently been treated.

We conclude that fibula strut graft is a good method of reconstruction of cystic defects in the proximal femut. It prevents pathological fracture, allows mechanical reinforcement of the lesion and delivers biological tissue allowing early consolidation of the defect.

Our study sets out to show whether vascular endothelial growth factor (VEGF) expression in stage 2B osteosarcomas around the knee influences disease-free and overall survival.

Fifty-two such patients treated in out unit were identified and followed-up for for a minimum of 92 months. All were treated according to the current MRC protocol and had resection of their tumour. Tissue from their resected tumours was stained for VEGF using immunohistochemical methods and the percentage of tumour cells staining for VEGF was assessed. The relationship between VEGF expression and survival was assessed using the log-rank test and Kaplan-Meier survival curves.

At follow-up 32 (62%) patients were dead, all from metastatic disease. Twenty-six (50%) tumours showed expression of VEGF. Statistical analysis showed that patients with tumours with VEGF expression in more than 25% of the cells had significantly shorter overall survival (p=0.019) and disease free intervals (p=0.009). Expression of VEGF also correlated with expression of the proteolytic enzyme MMP9 (p=0.02).

VEGF is peptide which acts as a stimulator of new blood vessel growth in normal tissues, as well as in some solid tumours and their metastases. A tumour which is able to induce a blood supply has an increased ability to grow, seed metastases and threaten life. Our study is the first to look at VEGF expression in the tumour cells surviving after chemotherapy. It is this population of cells which is important as it is these cells which may go on to develop into metastatic or locally recurrent tumours. The over-expression of VEGF by osteosarcoma cells is thought to be associated with a worse prognosis due to a number of mechanisms. This study shows that VEGF expression is an important prognostic factor in osteosarcomas and suggests that the mechanisms by which VEGF and MMP9 expression produce a poor prognosis may be linked. Suppression of tumour angiogenesis by inhibition of the action of VEGF has shown promise in animal models as a potential new treatment for osteosarcoma, and warrants further study.

Introduction and Aims: Endoprosthetic replacement (EPR) following Bone Tumor excision is common. A major complication is infection with serious consequences. The aim is to investigate the cause of infection, management and sequalae.

Method: Over 11, 000 patients have been treated in our unit over 35 years. Information collected prospectively on a database, includes demographic data, diagnosis, treatment (including adjuvant), complications, and outcomes. Data was analysed to identify any infection in EPRs, its management and outcome. Factors such as operating time, blood loss, adjuvant therapy, type of prosthesis were investigated. Outcomes of treatment options were evaluated.

Results: Data was analysed on 1265 patients undergoing EPR over 34 years. Giving a total follow-up time of over 6500 patient years. One hundred and thirty-seven (10.8%) patients had deep infection (defined by a positive culture [n=128] or a clinically infected prosthesis with pus in the EPR cavity [n=9]). Forty-nine (34%) required amputations for uncontrollable infection. The commonest organisms were Coagulase Negative Staphylococcus, Staphylococcus aureus and Group D Streptococci. The only satisfactory limb salvaging operation was two-stage revision, with a 71% success in curing infection. Systemic antibiotics, antibiotic cement or beads and surgical debridement had little chance of curing infection. Infection rates were highest in tibial (23.1%) and pelvic (22.9%) EPRs (p< 0.0001). Patients who had pre- or post-operative radiotherapy had significantly higher rates of infection (p< 0.0001), as did patients with extendable EPRs (p=0.007). Patients who had subsequently undergone patella resurfacing and rebushing also had a higher rate of infection (p= 0.019 & p=0.052).

Conclusion: Infection is a serious complication of EPRs. Treatment is difficult and prolonged. Two-stage revision is the only reliable method for limb salvage following deep infection. Prevention must be the key to reducing the incidence of this serious complication.

Purpose: The purpose of this work was to analyse follow-up and prognostic factors in a series of patients treated for soft tissue tumours as a function of the type of facility providing initial care: a supra-regional referral centre (Royal Orthopaedic Hospital, Birmingham), and 38 regional hospitals in the referral area.

Material and methods: This series included 260 patients (111 women and 149 men) treated between 1994 and 1996. Mean age at diagnosis was 61 years. Primary care was given to 96 patients (37%) in the referral centre and 164 (63%) in other centres. Minimum follow-up was five years. The risk of local recurrence and survival prognosis were studied by risk factor: grade, localisation (supra versus infra aponeurotic), tumour size, quality of resection margins.

Results: High-grade tumours were found in 73% of patients with a supra-aponeurotic localisation in 59%. Mean tumour size was 8.6 cm. Tumours in patients treated in the referral centre were larger (10.3 cm versus 7.5 cm) (p< 0.05). Frequency of local recurrence was 20% for the referral centre versus 37% for the other centres. Overall five-year survival rate was 58% and was correlated with grade, tumour size, and localisation (p< à.05). Overall survival of patients given primary care in the referral centre was not statistically different from those treated in the other centres, but for high-grade tumours (UICC grade III), five-year survival was 41% for the referral centre and 14% for the other centres (p< 0.05).

Discussion: Soft tissue sarcomas are rare tumours. For high-grade sarcomas, the rate of recurrence after treatment and the survival rate were better for patients given primary care in the referral centre. The question of centralising patients with this type of disease in referral centres is raised.

Aims: Tumours of the distal humerus are rare but a challenge to treat. Options for treatment are excision and flail elbow, arthrodesis with considerable shortening, allograft replacement or endoprosthetic replacement (EPR). A retrospective analysis of 10 cases of EPR distal humerus was done to assess their success in treating tumours.

Methods: A retrospective analysis of 10 distal humeral tumours operated between 1970 and 2001 was done by retrieving data from notes. No patient was lost to follow up. The Toronto Extremity Salvage Score (TESS) was used to assess function in patients still alive.

Results: There were 4 male and 6 female patients, with ages ranging from 15 to 76 years. The period of follow up ranged from 5 months to 31 years. 8 patients had primary tumours and 2 had secondary tumours. 4 out of 10 patients died of metastatic disease 12 to 71 months after operation. None of the 10 patients had local recurrence, infection, amputation or permanent nerve palsy. There were 3 revisions at 48, 56 and 366 months for aseptic loosening. There were 3 rebushings of the plastic inserts at 62,78 and 113 months. Two of the three rebushings were done after revision of the humeral component at 6 months and 30 months. The average TESS Score for these patients was 72.91 out of 100 (29.2 to 93.33).

Conclusion: Custom-made EPR for distal humeral tumours are an effective way of replacing the diseased bone leading to a reasonable level of function and an acceptable failure rate.

We reviewed our experience with diaphyseal endoprostheses to determine the survival of this type of reconstruction and factors affecting that survival.

Method: We retrospectively studied 44 patients who underwent endoprostheticreconstruction of diaphyseal bone defects after excision of primary sarcomas between 1979 and 2002 with more than 2 years follow up.

Results: There were 27 males and 17 females, the median age at diagnosis was 25 years (8–75) and the median bone defect was 18cm (10–27.6).There were 33 femoral reconstructions, 6 tibial and 5 humeral. The cumulative overall survival for all patients was 67% at 10 years and prosthetic reconstruction using revision surgery as an end point was 62% at 10 years. The cumulative risk offailure of reconstruction including; infection, fracture, aseptic loosening, local recurrence and amputation was 45% at 10 years but for amputation only was 13% at 10 years. The patient age, the type of prosthesis ;whether cemented oruncemented, site of defect (femur, tibia, and humerus) and length of defect did not influence prosthetic survival.

Conclusion: We concluded that the use endoprostheses for reconstruction of diaphyseal bone defects remains a valuable method of reconstruction with predictable results and compares favourably with other forms of reconstruction of massive diaphyseal bone defects.