Aims

The standard of surgical treatment for lower limb neoplasms had been characterized by highly interventional techniques, leading to severe kinetic impairment of the patients and incidences of phantom pain. Rotationplasty had arisen as a potent limb salvage treatment option for young cancer patients with lower limb bone tumours, but its impact on the gait through comparative studies still remains unclear several years after the introduction of the procedure. The aim of this study is to assess the effect of rotationplasty on gait parameters measured by gait analysis compared to healthy individuals.

Aims

Endoprosthetic reconstruction following distal femur tumour resection has been widely advocated. In this paper, we present the design of an uncemented endoprosthesis system featuring a short, curved stem, with the goal of enhancing long-term survivorship and functional outcomes.

Aims. This study aimed to analyze the accuracy and errors associated with 3D-printed, patient-specific resection guides (3DP-PSRGs) used for bone tumour resection. Methods. We retrospectively reviewed 29 bone tumour resections that used 3DP-PSRGs based on 3D CT and 3D MRI. We evaluated the resection amount errors and resection margin errors relative to the preoperative plans. Guide-fitting errors and guide distortion were evaluated intraoperatively and one month postoperatively, respectively. We categorized each of these error types into three grades (grade 1, < 1 mm; grade 2, 1 to 3 mm; and grade 3, > 3 mm) to evaluate the overall accuracy. Results. The maximum resection amount error was 2 mm. Out of 29 resection amount errors, 15 (51.7%) were grade 1 errors and 14 (48.3%) were grade 2 errors. Complex resections were associated with higher-grade resection amount errors (p < 0.001). The actual resection margins correlated significantly with the planned margins; however, there were some discrepancies. The maximum guide-fitting error was 3 mm. There were 22 (75.9%), five (17.2%), and two (6.9%) grade 1, 2, and 3 guide-fitting errors, respectively. There was no significant association between complex resection and fitting error grades. The guide distortion after one month in all patients was rated as grade 1. Conclusion. In terms of the accurate resection amount according to the preoperative planning, 3DP-PSRGs can be a viable option for bone tumour resection. However, 3DP-PSRG use may be associated with resection margin length discrepancies relative to the planned margins. Such discrepancies should be considered when determining surgical margins. Therefore, a thorough evaluation of the preoperative imaging and surgical planning is still required, even if 3DP-PSRGs are to be used. Cite this article: Bone Joint J 2023;105-B(2):190–197

Objectives. We evaluated the accuracy of augmented reality (AR)-based navigation assistance through simulation of bone tumours in a pig femur model. Methods. We developed an AR-based navigation system for bone tumour resection, which could be used on a tablet PC. To simulate a bone tumour in the pig femur, a cortical window was made in the diaphysis and bone cement was inserted. A total of 133 pig femurs were used and tumour resection was simulated with AR-assisted resection (164 resection in 82 femurs, half by an orthropaedic oncology expert and half by an orthopaedic resident) and resection with the conventional method (82 resection in 41 femurs). In the conventional group, resection was performed after measuring the distance from the edge of the condyle to the expected resection margin with a ruler as per routine clinical practice. Results. The mean error of 164 resections in 82 femurs in the AR group was 1.71 mm (0 to 6). The mean error of 82 resections in 41 femurs in the conventional resection group was 2.64 mm (0 to 11) (p < 0.05, one-way analysis of variance). The probabilities of a surgeon obtaining a 10 mm surgical margin with a 3 mm tolerance were 90.2% in AR-assisted resections, and 70.7% in conventional resections. Conclusion. We demonstrated that the accuracy of tumour resection was satisfactory with the help of the AR navigation system, with the tumour shown as a virtual template. In addition, this concept made the navigation system simple and available without additional cost or time. Cite this article: H. S. Cho, Y. K. Park, S. Gupta, C. Yoon, I. Han, H-S. Kim, H. Choi, J. Hong. Augmented reality in bone tumour resection: An experimental study. Bone Joint Res 2017;6:137–143

Aims. Iliac wing (Type I) and iliosacral (Type I/IV) pelvic resections for a primary bone tumour create a large segmental defect in the pelvic ring. The management of this defect is controversial as the surgeon may choose to reconstruct it or not. When no reconstruction is undertaken, the residual ilium collapses back onto the remaining sacrum forming an iliosacral pseudarthrosis. The aim of this study was to evaluate the long-term oncological outcome, complications, and functional outcome after pelvic resection without reconstruction. Methods. Between 1989 and 2015, 32 patients underwent a Type I or Type I/IV pelvic resection without reconstruction for a primary bone tumour. There were 21 men and 11 women with a mean age of 35 years (15 to 85). The most common diagnosis was chondrosarcoma (50%, n = 16). Local recurrence-free, metastasis-free, and overall survival were assessed using the Kaplan-Meier method. Patient function was evaluated using the Musculoskeletal Tumour Society (MSTS) and Toronto Extremity Salvage Score (TESS). Results. At a mean follow-up of 159 months (1 to 207), 23 patients were alive without disease, one was alive with lung metastases, one was alive following local recurrence, four were dead of disease, and three had died from other causes. The overall ten-year survival was 77%. There was only one (3%) local recurrence, which occurred at 26 months. There were 18 complications in 17 patients; 13 wound healing complications/infections, three fractures, one pulmonary embolism, and one dislocation of the hip. Most complications occurred early. The mean functional scores were 21.1 (SD 8.1) for MSTS-87, 67.3 (SD 23.9) for MSTS-93 and 76.2 (SD 20.6) for TESS. Conclusion. Patients requiring Type I or Type I/IV pelvic resections can expect a good oncological outcome and a high rate of local control. Complications are generally acute in nature and are easily manageable. These patients achieved a good functional outcome without the need for bony reconstruction. Cite this article: Bone Joint J 2020;102-B(6):779–787

INTRODUCTION. Bone tumour resection and subsequent reconstruction remains challenging for the surgeon. Obtaining adequate margins is mandatory to decrease the risk of local recurrence. Improving surgical margins quality without excessive resection, reducing surgical time and increasing the quality of the reconstruction are the main goals of today's research in bone tumour surgical management. With the outstanding improvements in imaging and computerised planning, it is now a standard. However, surgical accuracy is essential in orthopaedic oncologic surgery (Grimmer 2005). Patient specific instruments (PSI) may greatly improve the surgeon's ability to achieve the targeted resection. Thanks to its physical support, PSI can physically guide the blade yielding to a better control over the cutting process (Wong, 2014). Surgical time might significantly be reduced as well when compared to conventional method or navigated procedure. Finally, reconstruction may gain in rapidity and quality especially when allograft is the preferred solution as PSI can be designed as well for allograft cutting (Bellanova, 2013). Since 2011, PSI have systematically been used in our institution for bone tumour resection and when applicable allograft reconstruction. This paper reports the mid- to long-term medical outcomes on a large series. MATERIALS AND METHODS. Between 2011 and 2016, we systematically used PSI to remove bone tumours in 30 patients. The pre-operative planning involved the tumour delineation drawn on MRI by the surgeon. The MRI and obtained tumour volume were transferred to the CT-scan by image fusion (co- registration). Cutting planes were positioned around the tumour including a safe margin. The PSI were designed to ensure a sufficient stability but kept thin enough to limit the bone exposure. The PSI was manufactured by 3D-printing in a biocompatible and sterilisable material. PSI has been intraoperatively to cut the bone with predetermined margins. Medical files were reviewed for large data collection: type, size and site of the tumour, pre-and post-operative metastatic status, bone and soft tissues resection margins, local recurrence, use of an allograft and a PSI for graft adjustment or not for the reconstruction, the fusion of the allograft when applicable, the follow-up time and early/late complications. RESULTS. Over a period of 5 years, 30 patients were operated on with PSI (10 osteosarcomas, 4 chondrosarcomas, 10 Ewing sarcomas and 6 other types of bone tumours). Mean follow-up was 27±20 months. 18 cases out of 30 have more than 2 years follow-up and 13 out of 30 have more than 3 years of follow-up. Mean operating time was 6h02±3h44. Mean size of the tumours was 8,4±4,7cm and location was the upper limb in 5 cases, inferior limb in 15 cases and the pelvis in 10 occurrences. Metastatic disease developed postoperatively in 5 patients. Surgical margins in the bone were R0 in all cases but one case where a R1 surgery was planned to preserve a nerve root. We did not observe any local recurrence in the bone. Within soft tissues, margins were classified as R0 in 28 patients and R1 in 2 patients. In 26 cases, an allograft was used to reconstruct the bone defect. In 23 of those patients, the allograft was selected by CT scan and cut using a PSI. In the 3 allografts cut free-handily, only one demonstrated a fusion. Of the 23 cut with a guide, 12 fused completely, 2 demonstrated a partial fusion and 9 were not fused at the last follow-up. At the last follow-up, 2 patients were dead of disease, 5 were alive with metastatic disease and 23 were alive without disease. DISCUSSION. Oncology is probably the field where PSI can bring the largest advantage when compared to the conventional procedure. Several papers have reported the use of PSI for bone tumour resection. All of them have shown very promising results on in-vitro experiments (Cartiaux 2014), cadaver experiment (Wong 2012) or small clinical series (Bellanova 2013, Gouin, 2014). None of these papers report a large patient series associated with a clinically relevant follow-up. This series is the first mid- to long-term follow-up series involving PSI tumour surgery. These results are showing strong evidences of clinical improvements. It comes into contradiction with PSI for total knee arthroplasty where controversial results on the patient's outcome has been reported (Thienpont 2014). R0 margin has been systematically obtained for all bone cuttings, and local recurrence has been strongly decreased (3%) when compared to the usual recurrence rates published in the literature (from 15% to 35% according to the location). Allograft fusion seems improved as well thanks to the shape-matching of the selected allograft and a close contact between host and allograft at bony junctions. With a longer follow-up, these evidences should be stronger to definitely make PSI the best option for bone tumour resection

Recently, several preliminary reports have been issued on the application of computer assistance to bone tumour surgery. Surgical navigation systems can apply three-dimensional images such as CT and MR images to intraoperative visualization. Although CT is better at describing cortical bone status, MRI is considered the best method for defining the extent of marrow involvement for bone tumours and for planning surgical resection in bone tumour surgery. There have been a few reports on the application of MR imaging to navigation-assisted bone tumour surgery through CT–MR image fusion. However, the CT–MRI fusion technique requires additional costs and exposure of the patient to radiation from the preoperative CT, as well as additional time for image fusion. Above all, the image fusion process is a kind of registration (image to image registration) that inevitably leads to registration error. Herein we describe a new method for the direct application of MR images to navigation-assisted bone tumour surgery as an alternative to CT–MRI fusion. Six patients with an orthopaedic malignancy were employed for this method during navigation-assisted tumour resection. Resorbable pin placement and rapid 3-dimensional spoiled gradient echo sequences made the direct application of MR images to computer-assisted bone tumour surgery without CT–MR image fusion possible. A paired-point registration technique was employed for patient-image registration in all patients. It took 20 min on average to set up the navigation (range 15 to 25 minutes). The mean registration error was 0.98 mm (range 0.4 to 1.7 mm). On histologic examination, distances from tumours to resection margins were in accord with preoperative plans. Mean duration of follow-up was 25.8 months (range 18 to 32 months). No patient had a local recurrence or distant metastasis at the last follow-up. Direct patient-to-MRI registration is a very useful method for bone tumour surgery, permitting the application of MR images to intraoperative visualization without any additional costs or exposure of the patient to radiation from the preoperative CT scan

Aims. We present a retrospective review of patients treated with extracorporeally irradiated allografts for primary and secondary bone tumours with the mid- and long-term survivorship and the functional and radiographic outcomes. Patients and Methods. A total of 113 of 116 (97.4%) patients who were treated with extracorporeally irradiated allografts between 1996 and 2014 were followed up. Forms of treatment included reconstructions, prostheses and composite reconstructions, both with and without vascularised grafts. Survivorship was determined by the Kaplan-Meier method. Clinical outcomes were assessed using the Musculoskeletal Tumor Society (MSTS) scoring system, the Toronto Extremity Salvage Score (TESS) and Quality of Life-C30 (QLQ-30) measures. Radiographic outcomes were assessed using the International Society of Limb Salvage (ISOLS) radiographic scoring system. Results. There were 61 (54%) men with a mean age of 22 years (6 to 70) and 52 (46%) women with a mean age of 26 years (3 to 85). There were 23 deaths. The five-year patient survivorship was 82.3% and the ten-year patient survivorship was 79.6%. The mean follow-up of the 90 surviving patients was 80.3 months (2 to 207). At the last follow-up, 105 allografts (92.9%) were still in place or had been at the time of death; eight (7%) had failed due to infection, local recurrence or fracture. Outcome scores were comparable with or superior to those in previous studies. The mean outcome scores were: MSTS 79% (. sd. 8); TESS 83% (. sd. 19); QLQ 82% (. sd. 16); ISOLS 80.5% (. sd. 19). . Pearson correlation analysis showed a strong relationship between the MSTS and ISOLS scores (r = 0.71, p < 0.001). Conclusion. This study shows that extracorporeal irradiation is a versatile reconstructive technique for dealing with large defects after the resection of bone tumours with good functional and radiographic outcomes. Functional outcomes as measured by MSTS, TESS and QLQ-30 were strongly correlated to radiographic outcomes. Cite this article: Bone Joint J 2017;99-B:1681–8

Primary bone tumours of the clavicle are rare. Currently the existing literature is limited to a single case series and case reports or cases. Information regarding the patient's demographics and tumour types is therefore limited. The aim of this study was to investigate the and also suggest a management protocol for suspected primary bone tumours of the clavicle. We retrospectively reviewed the Scottish Bone Tumour Register from January 1971 to January 2012 and included all primary bone tumours of the clavicle. We identified only sixteen primary bone tumours over forty one year's highlighting the rarity of these tumours. There were ten benign and six malignant tumours with a mean age of 32 years (Range 4 to 66). The average presentation to orthopaedics after onset of symptoms was two months with five patients presenting following a pathological fracture. Malignant tumour types identified were consistent with previous literature with two cases of Ewing's sarcoma and osteosarcoma and a single case of osteosarcoma post radiotherapy and a single case of chondrosarcoma. Benign tumours were treated effectively with intralesional procedures. Malignant tumours were treated with wide local excision and subtotal or total clavicle excision. We suggest an investigatory and treatment protocol for patients with a suspected primary bone tumour of the clavicle. This is the largest series of primary bone tumours of the clavicle in the literature

Primary bone tumours of the talus are rare. Currently the existing literature is limited to a single case series and case reports or cases described in series of foot tumours. Information regarding the patient's demographics and tumour types is therefore limited. The aim of this study was to investigate these questions and also suggest a management protocol for suspected primary bone tumours of the talus. We retrospectively reviewed the Scottish Bone Tumour Register from January 1954 to May 2010 and included all primary bone tumours of the talus. We identified only twenty three bone tumours over fifty six years highlighting the rarity of these tumours. There were twenty benign and three malignant tumours with a mean age of twenty eight years. A delay in presentation was common with a mean time from onset of symptoms to diagnosis of ten months. Tumour types identified were consistent with previous literature. We identified cases of desmoplastic fibroma and intraosseous lipodystrophy described for the first time. We suggest an investigatory and treatment protocol for patients with a suspected primary bone tumour of the talus. This is the largest series of primary bone tumours of the talus in the literature

Amputation was once widely practised for primary bone tumours of the limbs. Yet this situation has changed with limb salvage surgery becoming increasingly popular in the last 30 years. Many different techniques are now available. These include allografts, autografts, endoprostheses and allograft-prosthesis composites. This article reviews these methods, concentrating on the functional outcomes and complications that have been reported

Needle biopsy is an established technique for the histological diagnosis of bone tumours, usually guided by fluoroscopy or CT. Surface lesions and aggressive tumours which have extended through the cortex are also amenable to imaging with ultrasound (US). We have assessed the diagnostic accuracy of US-guided Trucut needle biopsy in a consecutive series of patients referred to a Bone Tumour Unit with suspected primary bone tumours. Of 144 patients (83 men, 61 women; mean age 34.7 years) referred over a period of two years, 63 were considered suitable for US-guided biopsy. This was based on the presence of a relatively large extraosseous component, seen typically in osteosarcoma and malignant round-cell tumours. The results of needle biopsy were compared with those of surgical biopsy. The diagnostic accuracy was 98.4%, with only a single failed biopsy. Thus, in a selected group of patients, US is a very reliable technique of guidance for percutaneous needle biopsy of bone tumours

Background:The study aimed to analyse the demographic, clinical, and histological features of patients with a malignant primary bone tumour of the femur presenting with a pathological fracture. Methods: Eighty-six patients with primary malignant bone tumours of the femur presenting with pathological fracture were identified from a unique national database that contains original radiographs, casenotes and histology for all patients diagnosed with a primary bone tumour since september 1936 to the present. Demographic data, presenting features, tumour location, histological diagnosis, treatment, local recurrence, metastasis and survival data were gathered. Results: The median age was 63 years (range 4 to 87 years) and 47% were men. Forty-two percent of patients presented with a history of trauma. Forty percent of lesions were in the proximal femur, 34% in the diaphysis and 26% in the distal femur. The most common histological diagnoses were osteosarcoma (13 patients), Paget’s sarcoma (12 patients), myeloma (11 patients), chondrosarcoma and lymphoma (9 patients each). Other diagnoses were fibrosarcoma, Ewing’s sarcoma, spindle cell sarcoma, reticulum cell sarcoma, malignant fibrous hystiocytoma, and malignant giant cell tumour. The local recurrence rate was 31%. The median survival was 12 months (95% confidence interval 6 to 18 months). Overall 5 and 10-year survival were 22.4% and 17.4% respectively. Specifically for osteosarcoma, chondrosarcoma and Paget’s sarcoma the five year survival rates were 15.4%, 11.1% and 19.0% respectively. Those in whom the age at the time of presentation was over 60 years had a significantly worse prognosis (log rank 13.4, p< 0.001). Conclusion: Pathological fracture as a presenting symptom of primary malignant bone tumours is associated with a poor prognosis in nearly all tumour types studied. The prognosis is worse in those who are over 60 years at the time of presentation

Between 1988 and 2006, 18 patients had a custom-made endoprosthetic replacement of the distal humerus for bone tumours at our institution. There were 11 primary malignant neoplasms, six secondary deposits, and one benign aggressive tumour. The mean follow-up was for 4.4 years (1 to 18.2). Complications occurred in nine patients and included aseptic loosening in three (16.6%), local recurrence in two (11%), infection in two (11%), neuropraxia of the radial nerve in one (5.5%) and a peri-prosthetic fracture in one (5.5%). Excision was inadequate in four patients (22%), all of which developed local recurrence and/or metastases. There were seven deaths from the primary disease after a mean of 2.3 years (1 to 5), one of whom had an above-elbow amputation for local recurrence seven months before death. The remaining six had satisfactory elbow function at their last follow-up. The 11 living patients were evaluated using the Musculoskeletal Tumour Society and Toronto Extremity Salvage scoring systems. The mean scores achieved were 76% (67% to 87%) and 73% (59% to 79%), respectively. Overall, 17 of 18 patients had significant improvement in the degree of their pain following operation. Custom-made endoprosthetic reconstruction of the elbow for bone tumours is a viable treatment in carefully selected patients. It maintains satisfactory function and provides good pain relief

Introduction: Only 4–13 % of all spine tumours are primary bone tumours. We report on 180 cases of primary malignant bone tumours of the spine from the Scottish Bone Tumour Registry. The aim of the study was to analyse the incidence, demography, pathology and survival patterns of primary malignant bone tumours of the vertebral column. Materials and Methods: All of the data in the Scottish Bone Tumour Registry is compiled prospectively. This report is based on a retrospective review of all the data from 180 cases of primary malignant bone tumours (excluding metastatic lesions). Results: Of 4,301 registry cases 4.1% were primary malignant lesions (n=180). Seventy two percent of all spinal tumours were primary malignant neoplasms. There were 22 cervical, 72 thoracic, 45 lumbar and 41 sacrococcygeal lesions. There was a male preponderance (103 males & 77 females). The mean age at presentation was 54 years (range, 4–86 years). The top two ranked tumours were myeloma (42) and chordoma (41). Ewing’ sarcoma (15), Leukaemik-Lymphomatous lesions (13), conventional osteosarcoma (10) and Paget’s sarcoma (9) followed thereafter. The predominant presenting symptom was pain. Pathological fracture occurred in 7 patients. The operative treatment consisted of curettage (21), excision (51) and resection (9) with supplemented bone grafting (13). Adjuvant chemo (=61) and radiotherapy (=131) was also used in selective cases. Thirty patients were alive with no evidence of disease at a mean 5 year follow-up. Six were alive with persistent primary disease and/or local recurrence and/or metastases at the time of review. Eighty four patients died with persistent primary disease, 30 patients died of metastatic disease, 9 due to local recurrence and 17 of unrelated causes. Conclusions: Only 4.1% of the musculoskeletal tumours were spinal: 40% involved the thoracic spine. Mean age at presentation was 54 years. Myeloma, chordoma, chondrosarcoma and Ewing’ sarcoma were the most common pathologies. Myeloma predominated: osteosarcoma was much less common in our series compared to previous reports. Early diagnoses resulted into improved outcome

The surgical treatment of bone tumours can result in large perioperative blood loss due to their large sizes and hypervascularity. Preoperative embolisation has been successfully used to downgrade vascularity, thus reducing perioperative blood loss and its associated complications. Prior to embolization era, blood loss as high as 18,500mL have been reported peri-opratively. Twenty-six patients with a variety of bone tumours (average size 10.5×7.5×5.5cm), who underwent pre-operative embolisation between 2005 and 2009, were retrospectively studied. The group comprised of 17 females and 9 males. Their mean age was 38 years old. All patients underwent surgical resection within 48 hours of embolization. Mean blood loss was 796mL and required on average 1.1units of blood. We experienced no complications. Pre-operative arterial embolisation of large, richly vascular bone tumours in anatomically difficult positions, is a safe and effective method of downstaging vascularity and reducing blood loss

Endoprosthetic replacement of the distal tibia and ankle joint for a primary bone tumour is a rarely attempted and technically challenging procedure. We report the outcome of six patients treated between 1981 and 2007. There were four males and two females, with a mean age of 43.5 years (15 to 75), and a mean follow-up of 9.6 years (1 to 27). No patient developed a local recurrence or metastasis. Two of the six went on to have a below-knee amputation for persistent infection after a mean 16 months (1 to 31). The four patients who retained their endoprosthesis had a mean musculoskeletal tumour society score of 70% and a mean Toronto extremity salvage score of 71%. All were pain free and able to perform most activities of daily living in comfort. A custom-made endoprosthetic replacement of the distal tibia and ankle joint is a viable treatment option for carefully selected patients with a primary bone tumour. Patients should, however, be informed of the risk of infection and the potential need for amputation if this cannot be controlled

INTRODUCTION. We conducted a review of the Leeds Regional Bone Tumour Registry for primary bone tumours of the spine since establishment in 1958 until year 2000. AIM. To analyze the incidence of primary tumours of the spine and to record the site of occurrence, sex distribution, survival and pathology of these tumours. Summary of the Background Data. Primary tumours of the spine are particularly rare, accounting for between 4% and 13% of published series of primary bone tumours. METHOD. The Leeds Bone Tumour Registry was reviewed and a total of 2750 cases of bone tumours and tumour-like cases were analyzed. Consultants in orthopaedic surgery, neurosurgery, oncology and pathology in North and West Yorkshire and Humberside contribute to the Registry. RESULTS. Primary bone tumours of the osseous spine constitute only 126 of the 2,750 cases (4.6%). Chordoma was the most frequent tumour in the cervical and sacral regions, while the most common diagnosis overall was multiple myeloma and plasmacytoma. Osteosarcoma ranked third. The mean age of presentation was 42 years and pain was the most common presenting symptom, occurring in 95% of malignant and 76% of benign tumours. Neurological involvement occurred in 52% of malignant tumours and usually meant a poor prognosis,. CONCLUSIONS. The establishment of Bone Tumour Registries is the only way that sufficient data on large numbers of these rare tumours can be accumulated to provide a valuable and otherwise unavailable source of information for research, education and clinical follow-up

Aims. Resection of the proximal humerus for the primary malignant bone tumour sometimes requires en bloc resection of the deltoid. However, there is no information in the literature which helps a surgeon decide whether to preserve the deltoid or not. The aim of this study was to determine whether retaining the deltoid at the time of resection would increase the rate of local recurrence. We also sought to identify the variables that persuade expert surgeons to choose a deltoid sparing rather than deltoid resecting procedure. Patients and Methods. We reviewed 45 patients who had undergone resection of a primary malignant tumour of the proximal humerus. There were 29 in the deltoid sparing group and 16 in the deltoid resecting group. Imaging studies were reviewed to assess tumour extension and soft-tissue involvement. The presence of a fat rim separating the tumour from the deltoid on MRI was particularly noted. The cumulative probability of local recurrence was calculated in a competing risk scenario. Results. There was no significant difference (adjusted p = 0.89) in the cumulative probability of local recurrence between the deltoid sparing (7%, 95% confidence interval (CI) 1 to 20) and the deltoid resecting group (26%, 95% CI 8 to 50). Patients were more likely to be selected for a deltoid sparing procedure if they presented with a small tumour (p = 0.0064) with less bone involvement (p = 0.032) and a continuous fat rim on MRI (p = 0.002) and if the axillary nerve could be identified (p = 0.037). Conclusion. A deltoid sparing procedure can provide good local control after resection of the proximal humerus for a primary malignant bone tumour. A smaller tumour, the presence of a continuous fat rim and the identification of the axillary nerve on pre-operative MRI will persuade surgeons to opt for a deltoid resecting procedure. Cite this article: Bone Joint J 2017;99-B:1244–9