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

Resecting bone tumours within the pelvis is highly challenging and requires good cutting accuracy to achieve sufficient margins. Computer-assisted technologies such as intraoperative navigation have been developed for pelvic bone tumour resection. Patient-specific instruments have been transposed to tumour surgery. The present study reports a series of 11 clinical cases of PSI-assisted bone tumour surgery within the pelvis, and assesses how accurately a preoperative resection strategy can be replicated intraoperatively with the PSI.

The patient series consisted in 11 patients eligible for curative surgical resection of primary bone tumor of the pelvis. Eight patients had a bone sarcoma of iliac bone involving the acetabulum, two patients had a sacral tumor, and one patient had a chondrosarcoma of proximal femur with intra-articular hip extension. Resection planning was preoperatively defined including a safe margin defined by the surgeon from 3 up to 15 mm. PSI were designed using a computer-aided design software according to the desired resection strategy and produced by additive manufacturing technology. Intraoperatively, PSI were positioned freehand by the surgeon and fixed on the bone surface using K-wires. The standard surgical approach has been used for each patient. Dissection was in accordance with the routine technique. There was no additional bone exposure to position the PSI. Histopathological analysis of the resected tumor specimens was performed to evaluate the achieved resection margins. Postoperative CT were acquired and matched to the preoperative CT to assess the local control of the tumor. Two parameters were measured: achieved resection margin (minimum distance to the tumor) and location accuracy (maximum distance between achieved and planned cuttings; ISO1101 standard).

PSI were quick and easy to use with a positioning onto the bone surface in less than 5 minutes for all cases. The positioning of the PSI was considered unambiguous for all patients. Histopathological analysis classified all achieved resection margins as R0 (tumor-free), except for two patients : R2 because of a morcelised tumour and R1 in soft tissues. The errors in safe margin averaged −0.8 mm (95% CI: −1.8 mm to 0.1 mm). The location accuracy of the achieved cut planes with respect to the desired cut planes averaged 2.5 mm (95% CI: 1.8 to 3.2 mm).

Results in terms of safe margin or the location accuracy demonstrated how PSI enabled the surgeon to intraoperatively replicate the resection strategies with a very good cutting accuracy. These findings are consistent with the levels of bone-cutting accuracy published in the literature. PSI technology described in this study achieved clear bone margins for all patients. Longer follow-up period is required but it appears that PSI has the potential to provide clinically acceptable margins.

Introduction. Clear operative oncological margins are the main target in malignant bone tumour resections. Novel techniques like patient specific instruments (PSIs) are becoming more popular in orthopaedic oncology surgeries and arthroplasty in general with studies suggesting improved accuracy and reduced operating time using PSIs compared to conventional techniques and computer assisted surgery. Improved accuracy would allow preservation of more natural bone of patients with smaller tumour margin. Novel low-cost technology improving accuracy of surgical cuts, would facilitate highly delicate surgeries such as Joint Preserving Surgery (JPS) that improves quality of life for patients by preserving the tibial plateau and muscle attachments around the knee whilst removing bone tumours with adequate tumour margins. There are no universal guidelines on PSI designs and there are no studies showing how specific design of PSIs would affect accuracy of the surgical cuts. We hypothesised if an increased depth of the cutting slot guide for sawblades on the PSI would improve accuracy of cuts. Methods. A pilot drybone experiment was set up, testing 3 different designs of a PSI with changing cutting slot depth, simulating removal of a tumour on the proximal tibia (figure 1). A handheld 3D scanner (Artec Spider, Luxembourg) was used to scan tibia drybones and Computer Aided Design (CAD) software was used to simulate osteosarcoma position and plan intentioned cuts (figure 1). PSI were designed accordingly to allow sufficient tumour. The only change for the 3 designs is the cutting slot depth (10mm, 15mm & 20mm). 7 orthopaedic surgeons were recruited to participate and perform JPS on the drybones using each design 2 times. Each fragment was then scanned with the 3D scanner and were then matched onto the reference tibia with customized software to calculate how each cut (inferior-superior-vertical) deviated from plan in millimetres and degrees (figure 3). In order to tackle PSI placement error, a dedicated 3D-printed mould was used. Results. Comparing actual cuts to planned cuts, changing the height of the cutting slot guide on the designed PSI did not deviate accuracy enough to interfere with a tumour resection margin set to maximum 10mm. We have obtained very accurate cuts with the mean deviations(error) for the 3 different designs were: [10mm slot: 0.76±0.52mm, 2.37±1.26°], [15mm slot: 0.43±0.40mm, 1.89±1.04°] and [20mm: 0.74±0.65mm, 2.40±1.78°] respectively, with no significant difference between mean error for each design overall, but the inferior cuts deviation in mm did show to be more precise with 15mm cutting slot (p<0.05) (figure 2). Discussion. Simulating a cut to resect an osteosarcoma, none of the proposed designs introduced error that would interfere with the tumour margin set. Though 15mm showed increased precision on only one parameter, we concluded that 10mm cutting slot would be sufficient for the accuracy needed for this specific surgical intervention. Future work would include comparing PSI slot depth with position of knee implants after arthroplasty, and how optimisation of other design parameters of PSIs can continue to improve accuracy of orthopaedic surgery and allow increase of bone and joint preservation. For figures/tables, please contact authors directly.

Aims and methods. We present a review of our use of the Ilizarov apparatus in a non-acute NGO hospital in Cambodia specialising in limb reconstruction. Frames are applied without on table image intensification. A retrospective case-note analysis of Ilizarov apparatus use for all indications was conducted. 53 frames were applied between November 2005 and October 2011. Indications for application were chronic open fracture, osteomyelitis, fracture malunion, infective and non-infective non-union, bone lengthening, primary bone tumour resection, ankle fusion, congenital deformity or pseudarthrosis, chronic hip dislocation, or a combination of the above. Results. Median delay in presentation was 104 weeks for all indications (range 4–864). Median treatment length was 21 weeks (3–76). The most frequent complication was pin-site infection. This occurred in 18 patients (34%). Return to theatre occurred in 21 patients (40%). Indications were frame adjustment, pin addition or removal, addition of bone graft or re-osteotomy. Failure of union occurred in three patients. These rates are comparable with those published in both Asian and Western literature. Conclusions. Our data demonstrate the versatility of the Ilizarov apparatus and its importance in limb reconstruction in a developing world setting. Our centre relies on it as a cost-effective tool for traditional and novel indications. In our centre the apparatus is applied without x-ray control and is maintained without a dedicated outreach pin-site care programme. Despite this our complication rates are comparable with western literature. We therefore recommend it as a safe and cost-effective tool for use in other developing world settings

Introduction. Three-dimensional preoperative planning and bone tumour resection by navigation have been used in the past ten years. According to literature this workflow increases the surgical “accuracy”. However, there are a few and not completely clear reports describing accuracy in preoperative planning and navigation. The objective of this preliminary study was to determine the accuracy of osteotomies planned and guided by navigation in pelvis tumour resection. We assume that the surgical specimen scanned and 3D reconstructed is an acceptable method to determine the accuracy qualitatively and quantitatively of a virtual planning and navigation surgical workflow. Materials and Methods. A total of four patients were evaluated between May 2010 and February 2011, age range: 6–38, 17.4 mean; 2 males and 2 females. There were 4 malignant tumours: Malignant Schwannoma (1), Ewing's tumor (1) and Chondrosarcoma (2). All anatomic regions compromised by the tumour were preoperatively CT scanned (Mutislice 64, Aquilion, Toshiba Medical Systems, Otawara, Japan). Magnetic resonance images (MRI) of the corresponding region were acquired using a 1.5-T unit (Magnetom Avanto, Siemens Medical Solutions, Erlangen, Germany). Image fusion was applied to CT and MRI studies in order to determine the bone cortex and the intra-extraosseous soft tissues tumour extension. Once the fusion was obtained osteotomies were planned, taking into account the tumour extension in a three-dimensional virtual scenario. All patients were planned for two uniplanar osteotomies (intercalary resection). The minimal margin was determined in each plane by measuring the closest distance between malignant tumour and osteotomy plane. These studies allowed the visualisation of the tumour and the application of a virtual osteotomy. The simulation was carried out by using a computer-aided design (CAD) software, Mimics (Materialise, Leuven, Belgium). Three-dimensional preoperative planning was obtained in CAD format. Next, 3D models were exported to CT data sets in Digital Imaging and Communications in Medicine (DICOM) format and uploaded to the navigation system (3D OrthoMap navigation software v1.0, Stryker Navigator, Freiburg, Germany). Using the standard navigation tools (navigated pointer, camera and infrared tracker devices applied to the patient) the surgeon was able to establish a correspondence in a computer monitor between 3D images and real bone. Once osteotomies were performed, the tumour surgical specimen obtained was CT scanned and 3D reconstructed similarly to what was done previously to surgery to CT images acquired with the preoperative protocol in patients. Results. The correlation between the osteotomies preoperatively planned and the osteotomies achieved by navigation was in a global mean of 0.73 millimeters (SD: 3.14) in a total of 6 planes evaluated. Conclusion. According to clinical relevance, this work shows an acceptable accuracy in preoperative planning and navigation. Furthermore, we demonstrate the usefulness of three-dimensional models of surgical specimens when surgeons need to determine quantitative and qualitative accuracy of 3D planning and navigation procedures

Minimally invasive placement of iliosacral screws (SI-screw) is becoming the standard surgical procedure for sacrum fractures. Computer navigation seems to increase screw accuracy and reduce intraoperative radiation compared to conventional radiographic placement. In 2012 an interdisciplinary hybrid operating theatre was installed at the University of Ulm. A floor-based robotic flat panel 3D c-arm (Artis zeego, Siemens, Germany) is linked to a navigation system (BrainLab Curve, BrainLab, Germany). With a single intraoperative 3D scan the whole pelvis can be visualised in CT-like quality. The aim of this study was to analyse the accuracy of SI-screws using this hybrid operating theater. 32 SI-screws (30 patients) were included in this study. Indications ranged from bone tumour resection with consecutive stabilisation to pelvic ring fractures. All screws were implanted using the hybrid operating theatre at the University of Ulm. We analysed the intraoperative 3D scan or postoperative computed tomography and classified the grade of perforation of the screws in the neural foramina and the grade of deviation of the screws to the cranial S1 endplate according to Smith et al. Grade 0 stands for no perforation and a deviation of less than 5 °. Grade 1 implies a perforation of less than 2 mm and a deviation of 5–10°, grade 2 a perforation of 2–4 mm and a deviation of 10–15° and grade 3 a perforation of more than 4 mm and a deviation of more than 15°. All patients were tested for intra- and postoperative neurologic complications and infections. The statistical analysis was executed using Microsoft Excel 2010. 32 SI-screws were implanted in the first 20 months after the hybrid operating theatre had been established in 2012. All 30 patients were included in this study (15 men, 15 women). The mean age was 59 years ±23 (13–95 years). 20 patients received a single screw in S1 (66.7%), 1 patient 2 unilateral screws in S1 and S2 (3.3%), one patient 2 bilateral screws in S1 (3.3%) and 8 patients a single screw stabilising both SI-joints (26.7%). 27 screws showed no perforation (84.4%), 1 screw a grade 1 perforation (3.1%) and 4 screws a grade 2 perforation (12.5%). There was no grade 3 perforation. Furthermore there was no perforation of the neural foramina or the ventral cortex in the axial plane of the SI-screws stabilising one SI-joint (24 screws). Only single SI-screws bridging both SI joints showed a perforation of the neural foramina (37% grade 0, 12.5% grade 1, 50% grade 2, 0% grade 3). In the frontal plane 23 screws (71.9%) showed a deviation of less than 5°. In 5 screws a grade 1 deviation (15.6%) and in 4 screws a grade 2 deviation (12.5%) could be found. There was no grade 3 deviation. There were no infections or neurological complications. The high image quality and large field of view in combination with an advanced navigation system is a great benefit for the surgeon. All SI-screws stabilising only one joint showed completely intraosseous placement. Single SI-screws bridging 2 SI-joints intentionally perforated the neural foramina ventrally in 5 cases because of dysmorphic sacral anatomy. This makes image-guided implantation of SI-screws in a hybrid operating theatre a very safe procedure

CT and MRI scans are complementary preoperative imaging investigations for planning complex musculoskeletal bone tumours resection and reconstruction. Conventionally, tumour surgeons analyse two-dimensional (2-D) imaging information, mentally integrate and formulate a three-dimensional (3-D) surgical plan. Difficulties are anticipated with increase in case complexity and distorted surgical anatomy. Incorporating computer technology to aid in this surgical planning and executing the intended resection may improve precision. Although computer-assisted surgery has been widely used in cranial biopsies and tumour resection, only small case series using CT-based navigation are recently reported in the field of musculoskeletal tumor surgery. We investigated the results of CT/MRI image fusion for Computer Assisted Tumor Surgery (CATS) with the help of a navigation system. We studied 21 patients with 22 musculoskeletal tumours who underwent CATS from March 2006 to July 2009. A commercially available CT-based spine navigation system (Stryker Navigation; CT spine) was used. Of the 22 patients, 10 were males, 11 were females, and the mean age was 32 years at the time of surgery (range, 6–80 years). Five tumours were located in the pelvis, seven sacrum, eight femurs, and two tibia. The primary diagnosis was primary bone tumours in 16 (3 benign, 13 sarcoma) and metastatic carcinoma in four. The minimum follow-up was 17 months (average, 35.5 months; range, 17–52 months). Preoperative CT and MRI scan of each patient were performed. Axial CT slices of 0.0625mm or 1.25mm thickness and various sequences of MR images in Digital Imaging and Communications in Medicine (DICOM) format were obtained. CT and MR images for 22 cases were fused using the navigation software. All the reconstructed 2-D and 3-D images were used for preoperative surgical planning. The plane of tumour resection was defined and marked using multiple virtual screws sited along the margin of the planned resection. We also integrated the computer-aided design (CAD) data of custom-made prostheses in the final navigation resection planning for eight cases. All tumour resections could be carried out as planned under navigation guidance. Navigation software enabled surgeons to examine all fused image datasets (CT/MRI scans) together in two spatial and three spatial dimensions. It allowed easier understanding of the exact anatomical tumor location and relationship with surrounding structures. Intraoperatively, image guidance with the help of fusion images, provided precise visual orientation, easy identification of tumor extent, neural structures and intended resection planes in all cases. The mean time for preoperative navigation planning was 1.85 hours (1 to 3.8). The mean time for intraoperative navigation procedures was 29.6 minutes (13 to 60). The time increased with case complexity but lessened with practice. The mean registration error was 0.47mm (0.31 to 0.8). The virtual preoperative images matched well with the patients' operative anatomy. A postoperative superficial wound infection developed in one patient with sacral chordoma that resolved with antibiotic whereas a wound infection in another with sacral osteosarcoma required surgical debridement and antibiotic. After a mean follow-up of 35.5 months (17–52 months), five patients died of distant metastases. Three out of four patients with local recurrence had tumors at sacral region. Three of them were soft tissue tumour recurrence. The mean functional MSTS score in patients with limb salvage surgery was 28.3 (23 to 30). All patients (except one) with limb sparing surgery and prosthetic reconstruction could walk without aids. Multimodal image fusion yields hybrid images that combine the key characteristics of each image technique. Back conversion of custom prosthesis in CAD to DICOM format allowed fusion with navigation resection planning and prosthesis reconstruction in musculoskeletal tumours. CATS with image fusion offers advanced preoperative 3-D surgical planning and supports surgeons with precise intraoperative visualisation and identification of intended resection for pelvic, sacral tumors. It enables surgeons to reliably perform joint sparing intercalated tumor resection and accurately fit CAD custom-made prostheses for the resulting skeletal defect