Aim

Periprosthetic joint infections are a devastating complication after modular endoprosthetic reconstruction following resection of a musculoskeletal tumour. Due to long operating times, soft tissue dissection and immunosuppression, the infection rate after limb salvage is high and ranges between 8% and 15%. The aim of this retrospective single centre study was to assess the reinfection and re-reinfection rate after septic complications of megaprostheses.

Introduction

The current investigation includes a retrospective review of the experience of five Institutions with distal femur megaprostheses for tumor over a twenty year period, to analyze the incidence and etiology of failure, using a new classification system based upon the failure modes.

Background

Modular endoprostheses today represent a standard treatment option in the management of musculoskeletal tumors of the lower extremities. Long-term results of these reconstructions, however, are often limited by the course of the underlying disease. We therefore report our experiences in cancer patients with megaprostheses of the lower limb after a minimum of 15 years.

Aim

Preoperative serum CRP has been identified as an independent predictor in various malignancies. For osteosarcoma, however, the value of serological markers is unreliable. Aim of this study was to evaluate the prognostic power of preoperative CRP in patients with osteosarcoma.

Resection of the upper limb together with the shoulder girdle is known as forequarter amputation. In selected patients, this type of resection may have to be extended to a resection including parts of the thoracic wall in order to achieve local tumour control in advanced stages of neoplasms. Although, forequarter amputation alone is well described, reports on its combination with parts of the chest wall only consist of few case reports, while larger experiences and systematic reviews of its obstacles and complications are missing.

A series of six patients, five females and one male with, average age 58 years (range 41 to74 years), undergoing forequarter amputation with partial chest wall resection has been treated for malignancies around the shoulder at the Medical University of Vienna since 1993, thereby, to our best knowledge, presenting the largest series published to date. Patient data of all departments involved were retrospectively analysed and patients were followed-up according to standard oncological protocols. The underlying disease was malignant schwannoma in two, undifferentiated sarcoma in two, osteosarcoma in one and squamous cell sarcoma in one. A vast tumour of the shoulder and axillar space, uncontrollable pain, lymphedema, partial loss of function of the affected limb and ulceration indicated treatment.

All patients underwent radical amputation of the upper limb and all structures of the shoulder girdle including two to seven ribs. Chest wall reconstruction was achieved by a PTFE patch alone or in combination with a Stratos^®-implant. Myo-cutaneous reconstruction was performed by a pedicled (n=3) or a free myo-cutaneous forearm flap (n=3). There were no cases of peri-operative mortality. In two patients full forearm flap necrosis occurred and indicated two revisions in each with coverage by a free flap. One patient suffered vascular complications and underwent five revisions. Average follow-up was 14 months (range 5 to 35 months). Median survival was 8.5 months, with three patients still alive at the time of this investigation. Three patients died of systematic metastatic disease between 5 and 35 months after surgery. Two patients were successfully supplied with myo-electrical prostheses.

The key issues for surgical management of the resulting defect were

the restoration of mechanical chest wall stability,

Although forequarter amputation is a mutilating operation and in advanced stages of disease certainly can only temporarily limit disease progression, in selected cases, it may remain an option in vast tumours of the shoulder girdle.

Background: Endoprosthetic reconstruction has become the gold standard of treatment after the resection of tumors around the proximal femur, however, the rate of complications linked to megaprostheses is clearly higher than with standard implants. Aim of this study was to investigate the incidence and type of complications related to modular proximal femur prostheses.

Patients and Methods: By retrospective database analysis of the Vienna Tumor Registry, we evaluated the incidence of complications in 170 consecutive patients who have received a proximal femur KMFTR/GMRS at our institution between 1982 and 2007. 71 patients with an average age of 41.7 years (range 18.2–79.9 years) received the implant following the resection of a malignant bone tumor, 95 patients with an average age of 61.7 years (range 5.9–84.2 years) due to metastatic disease. The average time of follow-up was 3.5 years (SD ±4.9 years).

Results: Overall patient survival at five years was 32%. For patients with metastatic disease the overall survival was 10% at five years. Patients being treated for a primary bone tumor had an overall survival of 55% at 5 years. The overall survival of the prosthesis was 90% at two years and 72% at five years. Twenty-one patients (12.65%) suffered from dislocation after a mean time of 6.5 month (range 0.3–33 months) after surgery. Out of these, nine had to be treated by open reduction. Nine patients suffered recurrent dislocation after their first event. Patients who underwent extensive pelvic reconstruction had a significantly higher dislocation rate (33.3%) compared to patients with no or standard acetabular components (11.2%). Deep infection occurred in twelve patients (7.3%) after a mean of 39 months (range 1–166 months) after surgery. Treatment of infection was one-stage revision in eight and hip disarticulation in one patient. Two patients were successfully treated by local wound revision. One patient died of septic shock four days after surgery. Re-infection occurred in three patients. Aseptic loosing occurred in 13 patients (12.8%) after a mean time of 75.6 months (range 1–223 months) after surgery. Revision surgery was necessary in 27 patients (15.8%) with a mean time to first revision of 32 months (range 0.3–116 months). Prosthetic fracture was found in two patients. Local recurrence occurred in 14 patients (8.4%). In seven patients (4.2%) amputation was necessary.

Conclusion: Modular endoprosthesis allow excellent reconstruction of the proximal femur following tumor resection. However, the main complications, dislocation and infection, still remain considerable drawbacks.

In order to understand the role and efficacy of vascularized fibular graft and massive allograft in reconstruction of the knee, we have analyzed and review 25 patients of primary malignant bone tumours within 5 cm around the knee, that were managed primarily by this technique. In 4 patients the distal femur was affected while the proximal tibia was affected in 21 patients. There were 16 male and 9 female with an average age at the time of surgery of 19.7 years (range; 5 to 52), 17 patients (68%) were skeletally immature. The pathology was mostly represented by Osteosarcoma and Ewing sarcoma (15) and 18 patients (72%) received pre-operative chemotherapy. The resection of the tumor was transepiphyseal in 13 patients (52%) and intercalary in 12 patients (48%). The method of reconstruction was mainly concentric (allograft and fibula inside in 22 patients, 88%), while fixation was done principally by diaphyseal plate and metaphyseal screws (14 patients, 56%). Only three flaps failed (12%) detected by postoperative bone scan and confirmed by the clinical follow up. Twelve patients (48%) had 17 local complications (68%). Management of these complications succeeded to control them in nine patients (75%). The average time of union of fibula was 5.6 month (range: 3–10). The average time of union of allograft was 19.6 month (range: 10–34). All allograft united primarily (92%) except two cases; one case required bone graft and re-platting at 13 month postoperative after implant failure to achieve union 2 months later ; the other had infected non-union of allograft and amputation was done. Functional results were evaluated using the modified 30-points Musculoskeletal Tumor Society rating score (MTSRS) at final follow up of average 143 month (range; 28–213): the average total score was 27.4 (range; 18 to 30). All patients had good functional range of motion of the knee with stable knee at final follow up and were able to perform sport. Long term results of this study clearly indicates that allograft and vascularized fibular graft is a useful limb salvage procedure providing a biological long-term solution especially in skeletally immature. This technique provides single stage life long reconstruction. The allograft shell provide early stability and fixation to support a small epiphyseal fragment to preserve the articular surface and the vascularized fibula provides revascularization and osteointgration with the allograft to finally offer a long lasting durable reconstruction with full rang of motion of the knee.

Besides the femur and the tibia, the humerus is the third most common localisation of osteosarcoma. 78 patients with osteosarcoma of the humerus have been treated at our institution since 1934. Among these, 7 patients have been admitted before implementation of the Vienna Tumour Registry in 1968, additionally 4 patients had undergone primary surgical resection at another institution. This left 67 patients for follow-up after multi-modal therapy of humeral osteosarcoma comprising neo-adjuvant and adjuvant chemotherapy and surgical resection. (38 males and 29 females with an average age of 21.8 years, range 3.6 to 73.2 years) The subtypes of tumours observed were classic osteosarcoma in 56 patients, parostal sarcoma in 4, teleangiectatic sarcoma in 3, secondary sarcoma in 2 (one in Morbus Paget and one after radiation of a hemangioendothelioma), high-grade surface sarcoma in 1 and a humeral lesion within a multifocal osteosarcoma in 1. The localisation was foremost the proximal humerus (61) and rarely affecting the distal (5) or total bone (1). 11 patients suffered from pulmonary metastases upon primary diagnosis. In 9 cases resection alone was indicated. 9 patients underwent a resection-replantation-plasty, and in 2 patients primary amputation was performed. 46 patients were treated by resection and endoprosthetic reconstruction using ceramic prostheses (7), custom-made endoprostheses (13) or humeral HMRS modular prostheses (26). Before 1980 a non-standardised neo-adjuvant and adjuvant chemotherapy was administered in 12 patients, all patients thereafter received a chemotherapeutic regimen according to the COSS or EURAMOS-1 protocol. In 3 patients with parostal sarcoma no adjuvant therapy was indicated. The patient with multifocal osteosarcoma was treated conservatively by chemotherapy, radiation and immunotherapy by dendritic cell vaccination. The overall survival was 58% at 5 years. 23 patients died of their disease at an average of 25 months after operation (range 2 to 135 months). Average follow-up of the remaining patients was 91 months. (range 1 to 389 months). 4 patients treated before 1982 developed local tumour recurrence leading to secondary amputation in all cases, and death of disease within 12 months in 3 cases, respectively. 16 patients had to undergo one or more revisions, including secondary amputation in 2. Pulmonary metastases were observed in 15 patients, 2 patients developed skeletal metastates. After resection of metastatses, 14 patients died of disease, among them 9 patients died within 12 months after operation. Upon latest follow-up, 3 patients were alive disease-free, the patient with multifocal osteosarcoma was alive with disease 22 months after primary diagnosis. The multimodal treatment of osteosarcoma shows satisfactory oncological results. The implementation of standardised chemotherapeutic protocols has improved overall outcome.

The treatment of deep prosthetic infection in cancer patients with tumour prostheses remains the major complication to be dealt with in this population.

The Vienna Bone Tumour Registry includes information of more than 6500 patients of a period of 36 years. 145 patients with malignant proximal femoral tumors had resection and limb salvage with an uncemented Kotz modular femoral and tibial reconstruction megaprosthesis (KMFTR). There were twenty osteosarcomas, thirteen Ewing’s sarcomas, six chondrosarcomas, six plasmozytomas, three fibrosarcomas, three liposacomas and others. Thirteen patients (7 males, 6 females with an average age of 45 years, range 10 to 75 years) suffered from deep prosthetic infection within an average of 44 months after primary implantation, representing an infection rate of 8,97 percent.

Average follow-up was 109 months, range 7 to 339 months. Two patients with only mild signs of infection were treated by a conservative antibiotic regimen. Nine patients were treated by one-stage revision. One of the remaining two patients with severe infection underwent exarticulation of the hip as primary intervention, the other patient died due to general sepsis on the fourth post-operative day. Six patients showed no further signs of infection. Six patients, however, required one or more reoperations due to recurrent prosthetic infection. Among these four patients have successfully been treated by repeated one-stage revision, in two patients the prosthesis had to be removed permanently.

Deep prosthetic infection around modular tumour prostheses of the proximal femur and hip seems to be less common compared to distal femur, knee or tibia. However, the treatment of this complication has a higher failure rate due to multiply recurrent infection.

Hyperactivation of the epidermal growth factor receptor (EGFR) by gene amplification, mutation as well as overexpression is a hallmark of multiple human carcinomas. However, in recent years data have accumulated that EGFR-mediated signals might also contribute to malignant progression and therapy resistance of human sarcomas. Consequently we have investigated if human osteosarcoma cell lines (n=9) express functional EGFR and its useability as therapeutic target. Osteosarcoma cells expressed distinctly differing level of EGFR reaching in some cases high amounts. However, even low expression levels were sufficient to activate both MAPK and PI3K pathways (determined by phosphorylation of ERK1/2 and S6, respectively) following EGF exposure of serum-starved cells. The EGFR-specific inhibitor gefitinib completely blocked EGF-mediated and attenuated serum-induced downstream signal activation. While gefitinib applied as single agent demonstrated only limited growth inhibiting activity in short term experiments (72h drug exposure), it led to reduced colony formation in long term experiments in the majority of cell lines. Importantly, gefitinb sensitized EGFR-expressing osteosarcoma cell lines against chemotherapy with doxorubicin and methotrexate, while it antagonised cisplatin-induced cell death. Summarizing, our data suggest that EGFR-mediated survival signals protect human osteosarcoma cells against the cytotoxic activity of several antineoplastic drugs. Consequently, combination approaches including EGFR inhibitors in addition to chemotherapy should be evaluated for treatment of high grade osteosarcoma patients.

Introduction: Little data are available about the incidence and the management of hip dislocation following the implantation of megaprosthesis of the proximal femur, which is one of the main complications following this procedure.

Material and Methods: 190 patients, who received a proximal femur KMFTR/GMRS at our institution between 1982 and 2007, were retrospectively reviewed with regard to the incidence of hip dislocation as well as the success rate of the subsequent surgical/non-surgical treatment. A proximal femur tumor endoprosthesis was used in 148 patients following the resection of a malignant tumor and in 43 patients in severe revision cases following total hip arthroplasty. The average age at the time of surgery was 48 [6a to 83a] in the tumor group and 57.3 [45a to 78a] in the revision group. All of the revision cases and 12 patients from the tumor group had additional revision cups, such as the Schoellner pedestal cup.

Results: 12.3 % (18/147) of the tumor patients and 13.9% (6/43) of the revision cases dislocated at least once. 66.7% (12/18) of the first dislocations from the tumor and 50 % (3/6) of the revision group were treated with closed reduction, the rest required surgery. All patients received an abduction cast for at least 8 weeks. 38% (7/18) of the dislocated hips of tumor group (4.8% [7/147] total) and 67% (4/6) of the revision group (9.3% [4/43] total) experienced a second dislocation. 57% (4/7) of the dislocations from the tumor and 100 % (4/4) of the revision group were treated with closed reduction. Three patients from the tumor group (2% [3/147] total) experienced a total of three dislocations and one patient four dislocations (< 1% [1/147] total). The first dislocation occurred in 88% of the cases within 5 months following surgery during activities of daily living. 82% of the second dislocations and all third dislocations occurred within 4 months of the previous dislocation. Interestingly, no significant difference was found in the rate of re-dislocation between surgical and non-surgical treatment in either group.

Discussion: Dislocation of a proximal femur tumor endoprosthesis is an early complication following surgery and continues to be a challenging condition to treat, especially in cases with extensive soft-tissue defects. Since 2000, a polyester ligament is successfully used in our institution as a reinforcement to reduce the risk of hip dislocation in proximal femur tumor endoprosthesis. Surgical and non-surgical methods to reduce the risk of hip dislocation are discussed.

Parosteal osteosarcoma is an uncommon tumour. Different methods of surgical treatment have been reported. Aim of this study was to investigate differences in outcome after biological and prosthetic reconstruction.

Since August 1969, 28 patients have been treated at our institution. Average age was 26 years, range 15 to 59 years. Patient data was retrospectively reviewed within the prospective database of the Vienna Tumour Registry. Average follow-up was 133.9 months, range 8.4 to 382.6 months. Two patients died of disease 8.4 and 81.4 months after operation, respectively, another patient died due to unrelated causes 330.4 months postoperatively. All surviving patients were followed for a minimum of 3.6 months.

Location of the lesion was the distal femur (19), proximal humerus (four), proximal tibia (three), mid-diaphyseal and proximal femur (one each). In 12 patients endoprosthetic reconstruction was indicated. Biological reconstruction was performed in 11 patients. Three patients underwent rotationplasty, two patients were amputated. Eight of 12 patients with endoprostheses have been revised, five have had multiple revisions. Causes for revision were bushing wear (four), aseptic loosening (four), infection (three) and periprosthetic fracture (one). There was no local recurrence in the endoprosthetic group. Two of 11 patients with biological reconstruction underwent revision due to pseudarthrosis and femoral fracture, respectively. There were two cases of local recurrence requiring secondary amputation. Two patients with rotationplasty underwent revision for wound healing disturbance and thrombectomy, respectively. Three patients developed lung metastases, leading to death of disease in two cases of amputation and rotationplasty. One patient with endoprosthetic reconstruction was alive 129.0 months after pulmonary metastasectomy. Functional outcome was satisfactory in all patients; there were no significant differences between patients with endoprosthetic or biological reconstruction.

Biological reconstruction showed less revisions compared to endoprostheses, however, exact preoperative planning is required to obtain clear margins of resection.

Fibula autograft reconstruction, both vascularised (v) and non-vascularised (nv), has been established as a standard method in limb salvage surgery of bone and soft tissue tumours of the extremities. This study retrospectively analyses the results of fibula autograft procedures in general and in relation to vascular reconstruction or simple bone grafting.

Since the implementation of the Vienna Tumour Registry in 1969, 26 vascularised and 27 non-vascularised fibula transfers have been performed at our institution in 53 patients, 26 males and 27 females with an average age of 21 years (range 4 to 62 years). Indications included osteosarcoma in 18, Ewing’s Sarcoma in 15, adamantinoma in 5, leiomyosarcoma in 3 and others in 12. Thirty patients were operated for reconstruction of the tibia (8v/22 nv), 7 for the femur (6v/1nv), 7 for defects of the forearm (4v/3nv), 5 for metarsal defects (all v), 3 for the humerus (1v/2nv) and one patient was treated for a pelvic defect (nv).

Average follow-up was 63 months (range 2 to 259 months). 43 patients showed successful primary bony union of the autograft. In 12 cases pseudarthrosis indicated further surgical revision, 9 of these patients were primarily reconstructed by use of a nv autograft. 4 patients, 2 with v and 2 with nv reconstruction, suffered a fracture of the transplant and were operated for secondary osteosynthesis. 10 patients with v bone graft developed wound healing disturbances which led to surgery, 2 patients with nv grafts suffered such complications. In 2 patients recurrent infection of a nv and a v fibula transfer led to the implantation of a modular tumour prostheses or amputation, retrospectively. Function of all patients with primary bone healing was rated satisfactory.

The use of fibula autograft in limb-salvage surgery under oncological conditions allows biological reconstruction with good functional outcome, especially when primary bone healing is achieved. Vascularised bone grafting seems to have a better outcome in terms of primary bone healing than simple fibula bone grafting, and thus represents a feasible choice in the reconstruction of bone defects from tumour resection.