Introduction: Location on iliac bone account for 20% to 30% of sarcomas. Gold standard of local treatment is wide resection but till now few papers tried to evaluate the long term results of reconstructive procedures when chemotherapy and/or radiotherapy are used.

Patients: 44 patients (25 males and 19 females aged 9 to 66 years) with bone sarcoma of innominate bone in Zone 1, 2 or 4 (without involvement of acetabulum) were treated and/or followed up by the same team in 23 years. Histology was: chondrosarcoma (28), Ewing (13), osteosarcoma (2), MH (1)

Preoperative screening of patients included standard X rays, CT and bone technetium scan in all cases and MRI in 15 cases. Diagnosis was made by open biopsy except for 4 cases of chondrosarcoma for these preoperative screening was sufficient (and diagnosis confirmed by postoperative histological examination).

Following limb salvage using reconstruction of pelvis was performed with methyl metacrylate without prosthesis Titanium screws were inserted in remaining bone before moulding of acrylic cement (2 to 3 packs of antibiotic loaded cement).

Results: With a median follow-up of 15 years (minimal 2- maximal 22). 11 patients died from disease after local recurrence (6) and/or metastases (7). One disease free survivor has been lost for follow after 3 years,1 patient is alive with disease. The 31 others are disease free survivors.

Prognostic value: in our patients the prognosis was directly correlated with the histological grading (low grade chondrosarcoma have a 85% DFS) and for high grade tumours with the efficacy of the chemotherapy protocol. For primary metastatic patients, when chemotherapy is suboptimal or margins contaminated, the prognosis is dismal. With our most effective protocols and free margins, metastatic lesions did not affect the disease free survival of our patients.

Orthopaedic results: weight bearing was immediate in all cases. We observed 3 deep infections (2 compelled to make resection of the cement) and 2 late mobilisations of cement. In all other patients, the reconstructive procedure gave a good and stable functional result even in very long follow up.

Conclusion: Acrylic reconstruction is an easy and reliable reconstructive procedure after en bloc resection of iliac bone for malignant tumours in zone 1, 2 or 4. It is more reliable than bone graft when chemotherapy or radiotherapy are necessary.

Desmoid tumour is an histological benign tumour. Nevertheless, peri-scapular relapses can decrease the function and intra thoracic progression threaten life. To prevent these complications, damaging treatment (radiotherapy, amputation) are sometimes proposed. To precise the optimal indications of treatment, we reviewed our cases.

Patients: from 1984 to 2008, we treated 11 patients with peri-scapula fibromatosis (mean age 42 (13–58)). Only 4 patients were seen at first hand, 7 for relapses (3 of them after radiotherapy).

Treatment was adapted to each patient, in function of age, history of illness, and risks of spontaneous evolution. En bloc extratumoral resection was performed each time, when it didn’t expose to heavy functional risk (8). The other patients were treated by contaminated resection, but never invaliding. 4 patients received pre or/and post-operative chemotherapy. 1 received Interferon alpha, and 7 tamoxifen.

Results: with a median follow-up 15 years 3 months, 7 patients suffered of recurrence. No patient died from disease (thoracic complications) or therapeutic complication9 patients are in complete remission and 2 in stable disease. Following radiotherapy, local relapses (7 cases) and repeated surgery, functional sequellaes are numerous: 2 circumflex nerve palsies, 3 articular stiffness. Major functional sequellaes came from radiotherapy (limb discrepancy, lung and thorax deformity, skin and muscle atrophy.

Conclusion: in this non predictable illness, therapeutic indications should individually balance risks of spontaneous evolution and of complications of treatment. Besides surgery, needed in fast all cases, but often insufficient, it must be considered the value of interferon, tamoxifen and/or chemotherapy. The most important concept is the necessity to treatment avoiding late sequellaes and particularly radiotherapy or mutilating surgery.

Congenital fibrosarcoma (CFS) is a rare tumor most often affecting extremities of babies. Considering age, surgery of primary is preferred. Nevertheless amputation rate remains high. Preoperative chemotherapy (CT) role must be emphasised. We present 3 cases receiving preoperative CT.

Patients and methods in 1985, we treated a 3 months old girl for CFS of the thigh. To avoid amputation, preoperative CT (3 Ifosfamide- Vincristine- Actinomycine D) was performed leading to complete radiological and histological response. She benefited of conservative surgery She is in first complete remission 23 years later.

In September 1999, a 3 ½ y old boy with recurrent l buttock CFS operated elsewhere twice (6 months old, 2 years old), received preoperative chemotherapy with good clinical and radiological response. “En-bloc” extra tumoral resection was performed. Histology showed viable tumoral cells. We completed treatment by chemotherapy. In 01/ 2003 bilateral pulmonary metastases occurred leading to surgery and chemotherapy. In 09/ 2003 a new local recurrence appeared treated by surgery and postoperative chemotherapy. From this time, he received Alpha interferon. He is in complete remission for 6 years.

In 12/2005, a 14 y old girl, with local recurrence of CFS, treated elsewhere at the age of 5 months by partial surgery and chemotherapy (remained in remission for 13 years)was admitted. Since this time, she recurred locally despite resections and multiple lines of chemotherapy, but without metastasising. She was amputated in 2008.

Conclusion: preoperative chemotherapy is feasible despite low age of the patients, can allow conservative surgery and avoid late metastases.

The stiffness of the shoulder can result of many illness. Nevertheless, we observed a severe stiffness of the scapulo thoracic space only in fibromatosis. To precise the real diagnostic value of this symptom, we examined patients with different diseases of shoulder (tumoral and non tumoral).

The passive mobility of the shoulder of 11 patients with peri-scapular fibromatosis was compared to the mobility of those in 50 patients with non tumoral diseases of shoulder (arthritis and rotator cuff pathology), 50 peri-scapular soft tissues tumours, and 100 patients with primitive or secondary malignancies of humerus or scapula.

Results: in 10 of 11 patients with peri-scapular fibromatosis, the passive mobility of the scapulo thoracic spacewas severely impaired (less than 20°). In non tumoral pathology of shoulder, the passive mobility of the shoulder is frequently impaired but the stiffness hangs only on scapulo humeral articulation. In metastases, sarcoma and soft tissue tumour (except fibromatosis) the passive mobility of the scapulo-humeral joint is usually preserved and the mobility of the scapulothoracic space is always normal even in very huge tumours.

After treatment of fibromatosis, 9/11 patients are in complete remission and the mobility of their scapulo thoracic space restored. 2 patients are in stable disease and one suffers of a residual stiffness of the scapulothoracic space.

We conclude that the frozen scapulo thoracic is a specific symptom of peri scapulo thoracic fibromatosis. The restoration of the mobility of the scapulo thoracic after cure of the desmoid tumour confirms its specific role and represents a good marker of the tumoral evolution.

Purpose: In our records on bone tumours, secondary chondrosarcomas account for slightly less than 15% of all chondrosarcomas (20/150). The presentation is quite variable making diagnosis relatively difficult. We reviewed our experience to evaluate diagnosis, frequency, and prognosis.

Material and methods: From 1981 to January 2002, we had 20 chondrosarcomas which developed on pre-existing lesions: solitary exostoses (n=11), solitary chondroma (n=1), multiple exostosis (n=6), multiple enchondromatosis (n=2). Localisations were: pelvis (n=9), femur (n=3), humerus (n=2), tibia (n=3), spine (n=2), scapula (n=1). Histological classification was: grade I (n=7), grade II (n=9), grade III (n=1), and dedifferentiated sarcoma (n=3). Surgery was performed in all patients, alone for grade I and II chondrosarcoma, in association with chemotherapy (n=3) and radiotherapy (n=1) in three patients with dedifferentiated sarcoma.

Results and prognostic factors: At last follow-up (mean 9 years 10 months), five patients had died after local recurrence (n=3) or metastatic dissemination (n=2). The other fifteen patients were living (mean follow-up 155 months). The main prognostic factor was histological grade of chondrosarcoma. All patients with grade I chondrosarcoma (n=7) survived versus only two-thirds of those with grade II chondrosarcoma and half (2/4) of those with grade III or dedifferentiated chondrosarcoma. The second prognostic factor was initial management. Inadequate care initially led to misdiagnosis or delayed diagnosis (n=4), local recurrence (n=3) and loss of chance of survival (n=3). Grade I chondrosarcoma was occasionally taken for benign exostosis despite a cartilage cuff measuring more than 1 cm, normally a sign of chondrosarcoma.

Conclusion: 1. Because of the severity of secondary dedifferentiated chondrosarcoma, resection should be performed in adults presenting exostosis with a large residual cartilage cuff, particularly in high-risk locations (pelvis). 2. Because of the difficulty in recognising the histological features of grade I chondrosarcoma, the diagnosis of degeneration should be retained in adults if the cartilage cuff exceeds 1 cm. Lesions are suspicious if the cartilage cuff exceeds 5 mm.

Purpose: Massive allografts have been used in orthopaedic surgery for many years. Early results have generally been encouraging, but the risk of long-term resorption remains to be evaluated. We analysed resorption of 39 allografts implanted more than 10 years earlier.

Material and methods: Thirty-nine massive allografts implanted to sheath hip, knee or shoulder prostheses were studied a mean 13 years (range 10 – 17) after implantation. These allografts had been prepared in the same manner: preservaton at −30°C, sterilisation by 25000 Gy before implantation. A radiological scale was used to class graft resorption as minor (less than 5 mm in depth and less than 1 cm in length), moderate, or major (concerning the entire thickness of the allograft over three quarters of the circumference, irrespective of the length involved. These allografts measured more than 15 cm and were used for revision of loosened total hip arthroplasties, or reconstruction after tumour resection (tibia, femur, humerus). Among the 39 patients, 21 had received postoperative chemotherapy alone or in combination with radiotherapy.

Results: The allograft healed within the first or second postoperative year. Peripheral resorption was only observed at last follow-up. Among the 39 massive allografts assessed more than 10 years after implantation, 28 exhibited radiologically visible resorption. Among these 28, resorption was minimal for 15, moderate for nine and major for four. The first factor favouring resorption was complementary treatment, particularly local radiotherapy after carcinological resection. To a lesser degree, chemotherapy, given alone, also favoured resorption. None of the allografts implanted during a revision procedure for loosened total hip arthroplasty exhibited signs of absorption 10 years later. Two patients who had undergone resection of a chondrosarcoma with no complementary treatment exhibited only minor allograft resorption. Minor and moderate resorption was observed preferentially in the upper portions of the tibia and humerus, compromising the implant (three loosenings at more than 10 years. Peripheral resorption of the allograft appeared generally between the third and seventh postoperative year. Beyond that time, resorption did not appear to occur or increase. Evident polyethyl-ene wear observed in several patients operated on more than 10 years earlier was not associated with endosteal osteolysis of the subjacent allograft.

Discussion: Resorption of massive allografts is preferentially observed in the upper portions of the humerus and tibia, and generally occurs in patients who have undergone carcinological resection of a tumour followed by complementary treatment. Resorption is exceptional for allografts inserted during total prosthesis revision.

Purpose: Villonodular synovitis is a highly proteiform disease. The classical localisation is in the hand joints, but we focused on pseudomalignant forms which must be recognised to avoid unnecessarily aggressive treatment.

Material and methods: This series included ten files of patients seen for soft tissue tumour of the knee (n=6) or the hip (n=4). Time from first clinical sign to first consultation was two years on the average. The clinical presentation was dominated by major tumefaction of the soft tissues in all patients. There were no signs of joint disease in about half the patients. Radiological signs were noted in six patients. When obtained, MRI demonstrated, in all cases, a tumour formation with ferric deposits and thus contributed considerably to diagnosis.

Results: All patients underwent surgery. Five developed recurrence, as many as four times. One of these recurrent episodes was noted 17 years after the initial intervention that was considered satisfactory (wide monobloc resection). Joint involvement was minimal in these patients with a strong discordance between the minimally troublesome clinical presentation and the major anatomic modifications revealed radiographically.

Discussion: The risk of misdiagnosis is illustrated in our series which included two patients proposed for amputation with one which was actually performed due to a doubtful diagnosis of synovialosarcoma. This observation points out the review of the Swedish registry of synovialosarcomas: of the 81 files enrolled in the registry, 12 were found to be an aggressive form of villonodular synovitis.

Conclusion: The diagnosis of synovialosarcome, considered in patients who present with major tumefaction of soft tissues, signs of joint involvement, or ferric deposits identified on the MRI, should be evaluated with excessive care to rule out possible villonodular synovitis which responds to minimally aggressive treatment to be adapted to each individual case.

The purpose of this study was to investigate the importance of the timing of surgery for disease-free survival (DFS). The increasing efficacy of neo-adjuvant chemotherapy in Ewing’s sarcoma modifies the prognostic factors. In a recent monocentric study the classical prognostic value of size and location of the primary disappeared (Delepine G, Alkallaf S. J. Chem.1997;9:352–63.). This study confirmed the value of histologic response and pointed out the importance of dose intensity of VCR and ACTD. However, the role of local treatment could not be significantly demonstrated because the number of patients was too small.

Seventy-five patients with an average age of 19 years (range 4 to 40) years with Ewing’s sarcoma of bone fulfilled the inclusion criteria for this study: localised tumour at first screening (CT of lungs + bone scan) and location of the tumour in resectional bones (limb, scapula, innominate, rib, maxilla, skull). Metastatic patients and vertebral locations were excluded. The patients received multi-drug chemotherapy and were treated by surgery and radiotherapy in cases of bad responders and/or marginal surgery. The histologic response was evaluated according to Picci’s criteria (Picci, A. J Clin Oncol.1993;11:1793–99). The duration of local treatment was calculated from biopsy to surgery in weeks.

After a mean follow-up of 54 months, 41 patients were in first complete remission. Patients operated before the tenth week had a higher chance (68%) of first complete remission than patients operated later (DFS: 43%). The difference is significant (p< 0.03). Further analysis shows that the difference is due to late local control, which causes a dismal prognosis for bad responders.

Local treatment must be performed early, especially when histologic response is incomplete or uncertain. Preoperative chemotherapy that is too long increases the risk of metastases in bad responders. These factors must be taken into account when analysing multicentre protocols.

Nowadays 80% of patients with bone sarcomas can benefit from limb salvage. Their disease-free life expectancy is not jeopardised by conservative surgery as long as safe margins are obtained. For this reason, the oncological result relies on the accuracy of pre-operative and per-operative surgical measurements. Pre-operative evaluation of tumours is now quite accurate with digital margins (computed tomography, MNR, digital angiography). However, surgeons are still using centimeters or conventional radiographs with their own technical limitations for per-operative evaluation. A more accurate technique is needed.

The system is composed of three components: 1) a color, graphic computer workstation with software to calculate and present the location of the surgical instrument on a three-dimensional, reconstructed bone image, 2) a complete set of hand-held instruments containing infrared emitters, 3) an infrared receiver linked to the work station. This measuring system enables determination of the position and incidence of a surgical instrument in real time during surgery, with an accuracy of less than one mm.

The system requires four steps: 1) recording data with C.T., N.M.R. or angiography, 2) creating a three-dimensional image displayed on the computer screen for preoperative simulation of a virtual operation, 3) recording the very important anatomical points of the patient and optimal incidences of the surgical instruments, 4) preoperative location of surgical instruments and control of their location on bone.

This system is very useful for resection of bone tumours when the conventional location is uncertain (innonimate bone, rib), when very sharp accuracy is needed to preserve the growth plate of the distal femur in young children, and to avoid medullary damage in a spinal tumour.

The frameless stereotactic device is also very accurate in the reconstructive phase of limb salvage. After an internal hemipelvectomy, the device permits localisation of the acetabular prosthesis in the precise location before resection.

In our practice, the accuracy of the video guiding system is always within two mm as compared to conventional measurements usually between one or two cm for long bones and three to five cm for innominate bone.

The use of a video guidance system is very beneficial for limb salvage surgery for pelvic bone tumours.

A “hands-on” composite gives a similar functional result as a custom-made prosthesis and has a much better function than alternative techniques. Less expensive and more flexible than custom-made prostheses, it can be used even when no part of the iliac wing remains. The use of cement permits the adjunction of antibiotics needed for these complicated cases.

After peri-acetabular resection for bone sarcoma, a reconstructive procedure is necessary to stabilize the hip, avoid limb discrepancy, and permit full weight bearing. This procedure needs to be easy to perform because resection of the area is time and blood consuming. This leads to the use of a “hands-on” composite prosthesis.

Our reconstructive procedure uses a titanium cup with a long screw that is fixed in the remaining bone (sacrum or spine). When the cup is firmly fixed to the bone, the gap between the cup and bone is filled with cement loaded with antibiotics, and the polyethylene component is cemented on the innominate prosthesis. The femoral component of a usual hip total prosthesis is then implanted.

Since 1990 we have used this reconstructive procedure in 50 patients, 27 with bone sarcomas involving the acetabulum (11 chondrosarcomas, 9 Ewing’s sarcomas and 7 other sarcomas) and 23 for metastatic disease. Thirty of these patients were already metastatic when operated. The average duration of the reconstructive procedure was 45 minutes. Walking started from the fourth to tenth day after operation, but full weight bearing was usually authorised after six weeks.

Postoperative complications were frequent. Seven deep infections occurred, four required ablation of the prosthesis, and one would benefit from a saddle prosthesis. 33% of the patients had postoperative dislocation of the hip prosthesis and 13 patients had to be reoperated. Only two loosenings have been observed – one after deep infection and one after local recurrence in the sacral bone. Oncologic results: With a mean follow-up of five years, 28 patients died of disease and one from an unrelated disease. Four others with disease are still living. Seven local recurrences were observed (four in chondrosarcomas with a contaminated resection). The difficulty in obtaining wide margins explains the high rate of local recurrence (14 %). For patients with localised disease, the five-year overall survival rate is 75% and the five-year disease-free survival rate is 60%.

According to the Society for Musculoskeletal Oncology criteria, orthopaedic results were excellent in 7 patients, good in 30, fair in 6, and bad in 6. The mean functional score of 46 patients who still have their prostheses is 83% with usually no pain, excellent acceptance, length discrepancy of less than 1 cm, average flexion of 100 degrees, and unlimited walking without support.

We conclude that the rapidity and flexibility of this procedure are the positive aspects of this reconstructive technique. However, perfect positioning of the prosthesis remains difficult in a very large peri-acetabular resection. A computed guide is of great help to specify safe margins and prosthesis positioning. Longer follow-up is needed to ensure that the rate of late loosening will not be too high.

Nowadays 80% of patients with bone sarcomas can benefit from limb salvage. Their disease-free life expectancy is not jeopardised by conservative surgery as long as safe margins are obtained. For this reason, the oncological result relies on the accuracy of pre-operative and per-operative surgical measurements. Pre-operative evaluation of tumours is now quite accurate with digital margins (computed tomography, MNR, digital angiography). However, surgeons are still using centimeters or conventional radiographs with their own technical limitations for per-operative evaluation. A more accurate technique is needed.

The system is composed of three components: 1) a color, graphic computer workstation with software to calculate and present the location of the surgical instrument on a three-dimensional, reconstructed bone image, 2) a complete set of hand-held instruments containing infrared emitters, 3) an infrared receiver linked to the work station. This measuring system enables determination of the position and incidence of a surgical instrument in real time during surgery, with an accuracy of less than one mm.

The system requires four steps: 1) recording data with C.T., N.M.R. or angiography, 2) creating a three-dimensional image displayed on the computer screen for preoperative simulation of a virtual operation, 3) recording the very important anatomical points of the patient and optimal incidences of the surgical instruments, 4) preoperative location of surgical instruments and control of their location on bone.

This system is very useful for resection of bone tumours when the conventional location is uncertain (innonimate bone, rib), when very sharp accuracy is needed to preserve the growth plate of the distal femur in young children, and to avoid medullary damage in a spinal tumour.

The frameless stereotactic device is also very accurate in the reconstructive phase of limb salvage. After an internal hemipelvectomy, the device permits localisation of the acetabular prosthesis in the precise location before resection.

In our practice, the accuracy of the video guiding system is always within two mm as compared to conventional measurements usually between one or two cm for long bones and three to five cm for innominate bone.

The use of a video guidance system is very beneficial for limb salvage surgery for pelvic bone tumours.

The purpose of this study was to investigate the importance of the timing of surgery for disease-free survival (DFS). The increasing efficacy of neo-adjuvant chemotherapy in Ewing’s sarcoma modifies the prognostic factors. In a recent monocentric study the classical prognostic value of size and location of the primary disappeared (Delepine G, Alkallaf S. J. Chem.1997;9:352–63.). This study confirmed the value of histologic response and pointed out the importance of dose intensity of VCR and ACTD. However, the role of local treatment could not be significantly demonstrated because the number of patients was too small.

Seventy-five patients with an average age of 19 years (range 4 to 40) years with Ewing’s sarcoma of bone fulfilled the inclusion criteria for this study: localised tumour at first screening (CT of lungs + bone scan) and location of the tumour in resectional bones (limb, scapula, innominate, rib, maxilla, skull). Metastatic patients and vertebral locations were excluded. The patients received multi-drug chemotherapy and were treated by surgery and radiotherapy in cases of bad responders and/or marginal surgery. The histologic response was evaluated according to Picci’s criteria (Picci, A. J Clin Oncol.1993;11:1793–99). The duration of local treatment was calculated from biopsy to surgery in weeks.

After a mean follow-up of 54 months, 41 patients were in first complete remission. Patients operated before the tenth week had a higher chance (68%) of first complete remission than patients operated later (DFS: 43%). The difference is significant (p< 0.03). Further analysis shows that the difference is due to late local control, which causes a dismal prognosis for bad responders.

Local treatment must be performed early, especially when histologic response is incomplete or uncertain. Preoperative chemotherapy that is too long increases the risk of metastases in bad responders. These factors must be taken into account when analysing multicentre protocols.

A “hands-on” composite gives a similar functional result as a custom-made prosthesis and has a much better function than alternative techniques. Less expensive and more flexible than custom-made prostheses, it can be used even when no part of the iliac wing remains. The use of cement permits the adjunction of antibiotics needed for these complicated cases.

After periacetabular resection for bone sarcoma, a reconstructive procedure is necessary to stabilize the hip, avoid limb discrepancy, and permit full weight bearing. This procedure needs to be easy to perform because resection of the area is time and blood consuming. This leads to the use of a “hands-on” composite prosthesis.

Our reconstructive procedure uses a titanium cup with a long screw that is fixed in the remaining bone (sacrum or spine). When the cup is firmly fixed to the bone, the gap between the cup and bone is filled with cement loaded with antibiotics, and the polyethylene component is cemented on the innominate prosthesis. The femoral component of a usual hip total prosthesis is then implanted.

Since 1990 we have used this reconstructive procedure in 50 patients, 27 with bone sarcomas involving the acetabulum (11 chondrosarcomas, 9 Ewing’s sarcomas and 7 other sarcomas) and 23 for metastatic disease. Thirty of these patients were already metastatic when operated. The average duration of the reconstructive procedure was 45 minutes. Walking started from the fourth to tenth day after operation, but full weight bearing was usually authorised after six weeks.

Postoperative complications were frequent. Seven deep infections occurred, four required ablation of the prosthesis, and one would benefit from a saddle prosthesis. 33% of the patients had postoperative dislocation of the hip prosthesis and 13 patients had to be reoperated. Only two loosenings have been observed – one after deep infection and one after local recurrence in the sacral bone. Oncologic results: With a mean follow-up of five years, 28 patients died of disease and one from an unrelated disease. Four others with disease are still living. Seven local recurrences were observed (four in chondrosarcomas with a contaminated resection). The difficulty in obtaining wide margins explains the high rate of local recurrence (14 %). For patients with localised disease, the five-year overall survival rate is 75% and the five-year disease-free survival rate is 60%.

According to the Society for Musculoskeletal Oncology criteria, orthopaedic results were excellent in 7 patients, good in 30, fair in 6, and bad in 6. The mean functional score of 46 patients who still have their prostheses is 83% with usually no pain, excellent acceptance, length discrepancy of less than 1 cm, average flexion of 100 degrees, and unlimited walking without support.

We conclude that the rapidity and flexibility of this procedure are the positive aspects of this reconstructive technique. However, perfect positioning of the prosthesis remains difficult in a very large periacetabular resection. A computed guide is of great help to specify safe margins and prosthesis positioning. Longer follow-up is needed to ensure that the rate of late loosening will not be too high.

Purpose: Between January 1975 and December 2000, 498 cases of sarcoma of limb bones were treated by our multi-disciplinary team. Mean patient age was 27.1 years. Mean tumour size was 13.1 cm. Histology revealed osteosarcoma (n= 231), chondrosarcoma (n=118), Ewing sarcoma (n=104), MFH (n= 25), fibrosarcoma (n=12) and diverse tumours (n=8). The most frequent localisations were: femur (n=203), pelvis (n=98), tibia (n=86), humerus (n=60). Metastasis was present in 64 cases when first seen by our team. The histological resection was wide in 295 cases, marginal in 185 and contaminated in 18. Adjuvant treatment was adapted to patient age, histology and tumour localisation. Postoperative radiotherapy (34 to 50 Gy) was given for certain adults with osteosarcoma or Ewing tumours who had little histological response to preoperative chemotherapy or who had a marginal or contaminated resection.

Results: Median follow-up was 12 years. Two hundred fifty-two patients were living and disease free, eight were still under treatment, and 238 had died of their disease or treatment complications. There were 35 cases of local relapse, most of them (n=26) in referred patients, particularly after insufficiently effective chemotherapy. Complications were mainly deep infections (n=42). Secondary amputation was required for 24 patients (5%). Functional outcome at last follow-up was excellent in 52% of the patients, good in 35%, fair in 7% and unsatisfactory in 6%. Outcome depended basically on tumour size and localisation and deteriorated with infectious complications and radiotherapy.

Conclusion: 1. In our series, conservative surgery was performed in 95% of the cases, even for large tumours with fractures or for young children. 2. Functional outcome was better after conservative surgery: more than 85% excellent or good function. 3. Risk of local relapse was 2% for patients seen for initial diagnosis of high-grade malignant sarcoma. For patients with low-grade malignant tumours, or those who could not be given effective chemotherapy, the risk of local relapse was higher. 4. Radiotherapy improved local control for Ewing sarcomas and mesenchymatous chondro-sarcomas but its effect could not be assessed for the other tumours. As most of the secondary amputations and most of the poor functional results were observed in patients given complementary radiotherapy, this therapeutic modality, should, in our opinion, be avoided. 5. Preoperative chemotherapy being potentially dangerous for poor responders when the preoperative phase is continued too long, we advocate one month of preoperative chemotherapy for osteosarcoma and six weeks for Ewing sarcoma. This should be sufficient to allow conservative surgery (reduced tumour size) and chemotherapy (precise dose and protocols).

Purpose: Benign giant-cell tumours of the lower radius constitute a therapeutic challenge. Curettage with bone graft is generally used, but in this localisation recurrence varies from 25% to 80% requiring repeated surgery with a high risk of losing function.

Material and methods: Eight patients (five men and three women), aged 27 to 59 years were treated by our team between 1972 and 1994. Primary care was given in our unit for six of them and two others were referred for secondary surgery after recurrence. Radiographically, five of the tumours were aggressive (two had already led to fracture) and three were progressing. Three patients were treated first by curettage and bone graft to fill the gap. For the five other patients treatment included enucleation, curretage, and cement filling followed by osteosynthesis and immediate mobilisation. All patients were followed regularly at visits every three months for two years then every six months for two more years and every year thereafter. Median follow-up was 15 years (six–25 years).

Results: There were 12 recurrences (including four in soft tissue) in five patients (three patients initially treated with bone filling and two others among the five treated with cement filling). Recurrence was noted six to 30 months after surgery. Two patients initially treated with bone filling later had an arthrodesis that was filled with cement. Patients whose gap had been filled with cement and who had recurrence were treated again with cement filling. At last follow-up, all patients were in remission but two of them had lost wrist mobility. According to the function criteria established by the European Society for Bone Tumours, the final functional result was excellent in five, good in two and fair in one. Patients treated with cement filling had wrist mobility comparable to the healthy side and did not exhibit any radiographic alteration of the joint line.

Conclusion: The risk of relapse is high after treatment of benign giant-cell tumours of the lower radius. Filling the gap with cement does not avoid the risk of relapse but can be repeated without major inconvenience as long as autologous bone does not have to be harvested and immediate mobilisation is possible. The long-term functional outcome is the best argument favouring cement filling for benign giant-cell tumours of the lower radius, even in case of voluminous, aggressive tumours leading to fracture or relapse.

Purpose: Several studies have been conducted to search for factors affecting the prognosis of osteosarcoma. In this work, we attempted to assess the prognostic value of the biopsy technique and initial management on long-term prognosis of localised osteosarcoma of the limbs.

Material and methods: The series included 139 patients (88 males and 51 females, aged 4 to 58 years) with high-grade malignant osteosarcoma of the lower limbs treated or followed by our team between 1984 and 1998. Seventy-eight patients were referred to our unit for biopsy performed by a team surgeon after careful search for local extension and conception of the future extratumour en bloc resection. The 84 other patients were referred to our team after biopsy or after induction chemotherapy. There was no statistical significant difference between the two groups for known prognostic factors (localisation, tumour size, gender). All patients were given pre- and postoperative chemotherapy using the protocols generally applied at the time of their treatment. Three referred patients had already undergone amputation. All others were treated with conservative surgery even in case of fracture, very large tumour or young age. Patients were followed by their surgeon and chemotherapist independently with regular visits every three months for two years then every six months for two years and every year thereafter. Median follow-up was ten years (2.5–16.5 years).

Results: There were 12 local relapses (9%). Two were complications observed among the 75 patients followed from the start in our unit and ten were among the 84 secondary referral patients. Relapse-free survival reached 54% (46/84) in the referral patients compared with 73% (40/55) for the patients initially treated in our unit. For patients treated initially in our unit after 1986, the relapse-free survival rate reached 93%. Multivariate analysis demonstrated that the difference was significant (p < 0.02).

Conclusion: Initial management by an experimented team is a major prognosis factor for long-term survival and for risk of local recurrence in patients with high-grade malignant osteosarcoma of the limbs first seen without metastasis. When the diagnosis of osteosarcoma cannot be ruled out, these patients should be referred to a team specialised in malignant bone tumours before biopsy.