Aim

To evaluate outcome and complications of knee arthrodesis with a modular prosthetic system (MUTARS(r) Implantcast), as primary and revision implants in musculoskeletal oncology.

We retrospectively studied 14 patients with proximal and diaphyseal tumours and disappearing bone (Gorham’s) disease of the humerus treated with wide resection and reconstruction using an allograft-resurfacing composite (ARC). There were ten women and four men, with a mean age of 35 years (8 to 69). At a mean follow-up of 25 months (10 to 89), two patients had a fracture of the allograft. In one of these it was revised with a similar ARC and in the other with an intercalary prosthesis. A further patient had an infection and a fracture of the allograft that was revised with a megaprosthesis. In all patients with an ARC, healing of the ARC-host bone interface was observed. One patient had failure of the locking mechanism of the total elbow replacement. The mean post-operative Musculoskeletal Tumor Society score for the upper extremity was 77% (46.7% to 86.7%), which represents good and excellent results; one patient had a poor result (46.7%).

In the short term ARC effectively relieves pain and restores shoulder function in patients with wide resection of the proximal humerus. Fracture and infection remain significant complications.

Objective of this study was to analyse results of two stage revisions in infected megaprostheses in lower limb.

Material and Methods: Between April 1983 and December 2007, 1036 modular uncemented megaprostheses were implanted in 605 males and 431 females with mean age 33.5 yrs: 160 KMFTR®, 633 HMRS® prostheses, 68 HMRS® Rotating Hinge and 175 GMRS®. Sites: distal femur 659, proximal tibia 198, proximal femur 145, total femur 25, distal femur and proximal tibia 9. Histology showed 612 osteosarcomas, 113 chondrosarcomas, 72 Ewing’s sarcoma, 31 metastatic carcinomas, 89 GCT, 36 MFH, 68 other diagnoses.

Infection occurred in 80 cases (7.7%) at mean time of 4 yrs (min 1 month, max 19 yrs) in 18 KMFTR®, 47 HMRS®, 5 HMRS® Rotating Hinge, 10 GMRS®. Sites: 51 distal femurs, 21 proximal tibias, 6 proximal femurs, 1 total femur and 1 extrarticular knee resection. Most frequent bacteria causing infection were: Staphilococcus Epidermidis (39 cases), Staphilococcus Aureus (17) and Pseudomonas Aeruginosa (5). Infection occurred postoperatively within 4 weeks in 9 cases, early (within 6 months) in 12 cases, late (after 6 months) in 59 cases.

Usual surgical treatment was “two stage” (removal of implant, one or more cement spacers with antibiotics, new implant), with antibiotics according with coltures. One stage treatment was used for immediate postoperative infections, only since 1998.

Functional results after treatment of infection were assessed using the MSTS system.

Results: A two stage revision was attempted in 73 pts (91.2%): in 58 cases a new prostheses was implanted (with negative laboratory tests for infection) at mean time of 5 months (min 2, max 16 months), but in 3 pts infection recurred and they were amputated; 4 pts died before implanting a new prosthesis; 11 pts were amputated after several spacers since infection did not heal. One stage revision was performed in 4 of the 9 immediate postoperative infections, with successful results.

In 3 cases an amputation was primarily performed, to proceed with chemotherapy.

Revisions for infection were successful in 63 pts (79%), while 17 pts were amputated (21%).

Functional results evaluated in 53 revised cases were good or excellent in 43 (81.1%).

Conclusions: Two stage treatment of infected megaprostheses is successful in most cases. One stage has selected indications, mainly in postoperative immediate infections.

Introduction: Tumours of sacrum are rare. Treatment depends on malignancy or local aggressiveness: resection is indicated for malignant lesions, intralesional surgery for benign. Purpose of this study was to analyse risk of infection and its treatment after surgery for the two most common primary sacral tumours.

Material and Methods: Between 1976 and 2005, 82 patients with sacral chordoma or giant cell tumour were treated in our Institution. Demographic data, surgery and adjuvant treatments were analysed in the two histotypes. All patients were periodically checked with imaging studies. Special attention was given to the assessment of deep infections, their treatment and outcome.

Patients included 44 females and 38 males, ranging in age from 14 to 74 years. Mean follow-up was 9.5 years (min. 3, max. 27). Histopathological findings included chordomas in 55 cases and giant cell tumor (GCT) in 27. Most pts. had iv antibiotic therapy with amikacin and teicoplanin. Surgery of chordoma was resection, surgery of GCT was intralesional excision. In 6 sacral resections a miocutaneous transabdominal flap of rectus abdominis was used for posterior closure.

Results: No deep infections were observed in the GCT series. Three patients with sacral chordoma died for postoperative complications and were excluded from this analysis. Of the remaining 52 patients with chordoma, 23/52 had deep wound infection (44%), that required one or more additional operative procedures. In 16 pts. (70%) infection occurred within 4 weeks postoperatively, in 7 within 6 months. Most frequent bacteria causing infection were Enterococcus (23%), Escherichia Coli (20%), Pseudomonas Aeruginosa (18%). In 74% of cases a multiagent infection was detected. Surgical treatment consisted in 1 (52%) or more (48%) surgical debridements, combined with antibiotics therapy according to coltural results.

Mean surgical time was 14 hours for resections and 6 hours for excisions.

No significant difference was found comparing deep wound infections with levels of resection (15/33 resections proximal to S3-45% and 8/19 resections below or at S3-42%), previous intralesional surgery elsewhere (4/9 patients previous treated elsewhere-44% and 19/46 primarily treated patients-41%) and age at surgery.

Conclusions: Type of surgery was the prominent factor related with a major risk of infection. Operating procedure time correlated as well. Resection of sacral chordomas with wide margins improves survival although extensive soft-tissue resection in proximity to the rectum favours deep infections. Intralesional excision is the recommended surgical treatment for GCT of the sacrum and does not imply a significant risk of infection.

Objective: Aim of this study was to analyse the incidence of infections in primary prosthetic reconstructions of the humerus after resection for bone tumours and their treatment and results.

Material and Methods: Between 1974 and 2006 at Rizzoli 344 reconstructions of the humerus using prosthetic devices (alone or in association with allografts) were performed. Sites of reconstruction were: proximal humerus 311, distal humerus 19, diaphysis 5, total humerus 9. Histological diagnoses included 24 benign tumors, 253 malignant tumors and 67 metastatic carcinomas. Patients were followed periodically in the clinic. Informations were obtained from clinical charts and imaging studies with special attention to major complications requiring revision surgery. Univariate analysis through Kaplan-Meier actuarial curves was used in evaluating implant survival to major complications. Infections developing in the first 4 weeks were considered postoperative infections, those diagnosed in the first 6 months were judged early infections, while late infections those diagnosed after 6 months from surgery.

Results: In 20 patients (5.8%) a revision for deep infection was required. In 19 of these cases tumor was localized in the proximal humerus and in 1 in the distal humerus. There were 18 infections in prostheses and 2 in allograft prosthesis composites. Two infections were postoperatively diagnosed, seven were early infections and eleven late infections. Revision was required in 18 cemented prosthesis, 1 uncemented prostheses and 1 Coonrad-Morrey.

S. Epidermidis and S. Aureo were the most frequent bacteria causing infection (45%). Two stage treatment of infection was chosen: removal of the implant and temporary substitution with cement spacer with antibiotics (usually vancomycin) until infection healed. But a new prostheses was actually implanted in 3 cases only (at mean time of 5.7 mos), while in 17 the spacer was never removed by patients choice due to the acceptable result with the spacer. Systemic antibiotics were associated according to cultural results. Infection healed in all patients.

Conclusions: Infection is the most severe complication in prosthetic reconstructions for tumours of the humerus. Its incidence (5.8%) is lower than in lower limb. Treatment requires a team work: surgeon, microbiologist and infectious disease physician. One stage is indicated in postoperative infections, two stage is recommended in both early and late infections. Two stage surgery offered good results, although in most cases a new prosthesis was not implanted, since actually humeral megaprostheses act as a spacer and don’t provide a much better function.

Purpose: Aim of this study was to analyse our Institution experience with osteosarcoma in Paget’s disease (PD).

Methods: Twenty-six patients treated between 1961 and 2006 were retrospectively analysed. Information focusing on treatment, imaging and oncologic outcome were obtained from clinical charts.

Results Fifteen patients had previous diagnosis (mean time 9 yrs) of monostotic (80%) or polyostotic (20%) PD; in 11 sarcoma and PD were contemporarily diagnosed. The most frequents osteosarcoma histotypes observed were osteoblastic in 18 cases (69%), fibroblastic in 5 cases (19%), teleangectasic in 2 patients and chondroblastic in one case. In 6 patients surgery only was performed: 3 amputations and 3 resections. In 3 pts surgery (amputation), adjuvant chemotherapy and radio-therapy were given. In 1 pt surgery (amputation) and radiotherapy. In 12 patients surgery and chemotherapy: adjuvant in 10 cases (8 amputations and 2 resections) and neoadjuvant in 2 (both amputations). Two patients had only radiotherapy and 2 had only chemotherapy. Oncologic outcome showed 4 pts with no evidence sisease at a mean follow up of 139 months (min.42.6, max.257.4) and 22 died with disease at a mean time of 20.15 months (min.1, max. 84). One patient only of 6 (11%) treated with surgery only is NED at 10 years, while the other 5 died from disease at a mean time of 30 mos. Three of 12 patients (25%) treated with surgery and chemotherapy are NED at a mean follow up of 12 years, while 9 died of disease at a mean time of 24 months. All patients treated without surgery died at of 7.5 mos on average (min.1 – max.13.7).

Conclusions: Sarcomas in PD have a poor prognosis in pts treated without surgery or with surgery only. Surgery with chemotherapy – when feasible – improves prognosis.

Introduction: Extraskeletal osteosarcoma is a rare malignant tumor of the soft tissues. Overall, this malignancy has been associated with worse local control and overall survival rates than its skeletal counterpart despite multimodal approach. Purpose of this study was to review a single Institution experience and analyse results of management to identify factors affecting the outcome.

Methods: Retrospective study of 48 patients observed between 1966 and 2007 was undertaken. Of the total, 36 patients were admitted and managed at our Institution while 12 patients were sent for consultation and therefore not included in this study. Clinico-pathologic features and details of treatment of all 36 patients were reviewed and correlated with outcome. Updated follow-up was available in all patients.

Results: There were 21 males and 15 females, mean age was 53.6+/−19.3 years (range 14–84 yrs); 23 patients (63.9%) presented with localised disease while distant metastases were present in 13 patients (36.1%). Surgery consisted of a limb-salvage procedure in 25 patients (69.4%), amputation in 9 patients (25%) and 2 patients were considered inoperable (5.6%). Postoperative radiation therapy was given to 6 patients (16.7%) and multiagent chemotherapy administered to 19 patients (52.8%). At mean follow-up of 5.8 years, 23 patients had died of disease, expected 5 and 10 year overall survival rates were 41% and 31%. Tumor size and age at presentation were the most important predictors of survival while chemotherapy showed a trend towards improved survival in patients with localised disease.

Discussion and Conclusion: Extraskeletal osteosarcoma was associated with substantially worse prognosis than skeletal osteosarcoma despite multimodal management.

Purpose was to evaluate the incidence of complication in lower limb reconstructions with modular prostheses comparing cemented versus uncemented stems in two different orthopedic tumor Centers.

Methods: retrospective analysis of implant survival, complications and functional results assessed according to MSTS system of the Rizzoli and Moffitt series of modular megaprosthesis in lower limb.

From 2002 and 2007, 238 modular prostheses of same design with a rotating hinge knee, were implanted in the lower limb as primary reconstruction in these two Centers. In 130 cases the prosthesis was implanted with cemented stems and in 108 cases with uncemented stems. Sites included: 120 distal femur, 61 proximal femur, 46 proximal tibia, 10 total femur and in 1 case both distal femur and proximal tibia. Histologic diagnoses included: 89 osteosarcoma, 13 Ewing sarcoma, 21 chondrosarcoma, 19 sarcoma, 55 metastasis, 17 TGC, 7 MFH, 11 other diagnoses and 6 non oncologic cases.

Major prostheses-related complications were analysed and functional results according to Muscolo Skeletal Tumor Society system, at a mean follow up of 2.03 yrs. Also a statistical evaluation with Kaplan Meier curves, a comparative statistical analysis with Wilcoxon test and multivariate Cox regression analysis were performed.

Results: Outcome in 223 evaluated oncologic pts, showed: 121 pts continuously disease free, 26 NED after treatment of relapse, 54 AWD, 16 DWD. Margins were wide in 94.3% (214/227 pts) of evaluated pts. Complications causing implants failure were: infections in 20/238 (8.41%), aseptic loosening in 7/238 (2.95%). No breakages of prosthetic components were observed.

Infections occurred at mean time of 1.3 yrs, sites included: 9 distal femurs, 4 proximal femurs, 5 proximal tibias, 2 total femurs. Aseptic loosening occurred at mean time of 2.3 yrs, sites included: 3 proximal tibias, 2 distal femurs, 2 proximal femurs. Rate of aseptic loosening was higher for cemented stems 5/130 (3.85%) vs uncemented 2/108 (1.85%). This difference is not statistically significant. Infection rate was 10% (13/130) for cemented stems and 6.5% (7/108) for uncemented stems. This difference is statistically significant at comparison of survival curves logrank test. Average MSTS function score was 83.5%. Multivariate Cox regression analysis showed that the significant factor favourable reluted with lower incidence of complications was the use of uncemented stems.

Conclusions: lower limb modular prostheses with both cemented and uncemented stems gave good results and a low complication rate. Better results were observed with uncemented stems and statistically confirmed, but this needs to be further investigate in a future study at long term.

Purpose: Aim of this study was to analyse the incidence of infection in orthopaedic oncology after major surgical procedures for bone tumors.

Materials and Methods: We included patients with primary sacral tumors treated by major surgical procedure and patients with bone tumors of the upper and lower limb treated by resection and prosthetic reconstruction. Demographic data, surgery, adjuvant treatments, type of reconstruction were analyzed. Special attention was given to the infection: incidence, classification, microbic agents, treatment and outcome. Infections in the first 4 weeks were considered “postoperative”, those in the first 6 months were judged “early”, while “late” those diagnosed after 6 months. Overall 1462 patients treated in one institution from 1076 to 2007. Were considered 1036 patients with tumors of the lower limb, 344 patients with tumors of the upper limb and 82 sacral tumors. Univariate analysis with Kaplan-Meier actuarial curves was used in evaluating risk factors and implant survival to infections.

Results: In the lower limb, infection occurred in 80 cases (7.7%). Most frequent bacteria were gram positive. Infection was postoperative in 9 cases, early in 12, late in 59 cases and generally monomicrobial. Surgical treatment was “two stage” in 73 patients, “one stage” in 4 and primary amputation in 3 cases. Revisions for infection were successful in 63 pts (79%), while 17 pts were amputated (21%).

In the upper limb, in 20 patients (5.8%) a revision for deep infection was required. Two infections were postoperative, 7 early and 11 late. S. Epidermidis and S. Aureo were the most frequent bacteria causing infection (45%). “Two stage” treatment of infection was performed, but a new prostheses was implanted in 3 cases. In 17 the spacer was never removed.

In the sacrum, no deep infections were observed after intralesional excision for giant cell tumors. In 23/52 resections (44%) for chordoma (3 pts. died postoperatively and were excluded), infection occurred: in 16 patients postoperatively, in 7 within 6 months. Bacteria causing infection were mostly gram negative: in 74% of cases infection was multiagent. Surgical treatment consisted in one or more surgical debridements with antibiotics therapy according to coltures: infection healed in all cases.

Conclusion: Infection is a severe complication in prosthetic reconstructions for tumors of the upper and lower limb. Its incidence in the extremities (7.7% and 5.8%) is lower than after sacral surgery (44%). Infections are mostly late, monomicrobial and caused by gram positive in extremities, while early, multimicrobial and caused by gram negative in the sacrum.

Purpose: GSD, also known as massive osteolysis or disappearing bone disease, is rare, characterized by proliferation of vascular channels of hematic and lymphatic origin resulting in progressive distruction of bone. This study about Gorham-Stout disease is a retrospective review of the Rizzoli files with special attention given to treatment and outcome.

Materials and Methods: This study is based on a retrospective analysis of a single institution experience. In the Rizzoli files we found 15 cases of GSD from 1968 to 2008. Two were excluded for insufficient documentation. For 13 cases clinical data, imaging and histology were analysed. Histopatologically benign vascular proliferation of thin-walled endothelial capillaries surrounded by a fibrous stroma is present. Adipose involution of the bone marrow and extreme thinning of bony trabeculae represent other histopatologic features. A final diagnosis was established based on clinical, radiological and histopathologic features, as recommended in the literature. Imaging included X-rays in 11 cases and CT or MRI in 5. All lesions were lytic, with an associated sclerosis in two cases. There was one lesion in four cases, several lesions in the same bone in one, and multiple bones involved in six patients. Primary sites were proximal femur in 7 cases, pelvis in 2, hip and knee, calcaneus, humerus and cervical spine in 1 case each.

Results: Two patients had no treatment, 2 conservative treatment (cast or brace), 5 surgery, 6 medical treatment (byphosphonates, calcitonin, zoledronic acid, interferon, steroids), 1 radiotherapy, 2 selective arterial embolization. Surgery consisted of internal fixation of 4 pathologic fractures and reconstruction of the entire humerus with a double composite allograf in 1. Overall, surgery only in 2 patients, medical treatment only in 4 (1 also embolization), surgery and medical treatment in 2 (1 also embolization), radiotherapy only in 1, conservative treatment in 2. Four patients were lost at follow up. In the remaining 9 patients mean follow up was 17 ys.(min 2, max 30). These 9 patients had the following results: 2 dead, 3 healed, 3 with stable disease, 1 alive with asymptomatic disease at 24 ys.

Conclusions: No clear treatment recommendations were desumed. Surgery is indicated in pathologic fractures or reconstruction of massively destroyed bones, medical treatment and selective embolization are helpful. In the literature prostheses are mostly recommended for reconstructions due to the risk of allografts resorption.

Short intense electrical pulses transiently increase the permeability of the cell membrane, an effect known as electroporation. This can be combined with antiblastic drugs for ablation of tumours of the skin and subcutaneous tissue. The aim of this study was to test the efficacy of electroporation when applied to bone and to understand whether the presence of mineralised trabeculae would affect the capability of the electric field to porate the membrane of bone cells.

Different levels of electrical field were applied to the femoral bone of rabbits. The field distribution and modelling were simulated by computer. Specimens of bone from treated and control rabbits were obtained for histology, histomorphometry and biomechanical testing.

After seven days, the area of ablation had increased in line with the number of pulses and/or with the amplitude of the electrical field applied. The osteogenic activity in the ablated area had recovered by 30 days. Biomechanical testing showed structural integrity of the bone at both times.

Electroporation using the appropriate combination of voltage and pulses induced ablation of bone cells without affecting the recovery of osteogenic activity. It can be an effective treatment in bone and when used in combination with drugs, an option for the treatment of metastases.

Purpose of this study was to analyse the different techniques of prosthetic reconstruction of the humerus (also in association with bone grafts) after resection of primary tumors, discussing indications and evaluating implant survival.

Material and Methods: Between 1974 and 2006, 277 patients had prosthetic reconstruction of the humerus after tumor resection. These included 253 reconstructions of the proximal humerus: 225 cemented modular prostheses (MRS), 18 uncemented modular prostheses, 9 allograft/prostheses composites, 1 custom-made prosthesis; 2 diaphyseal reconstructions:1 intercalary prosthesis and 1 MRS; 13 reconstructions of the distal humerus: 11 uncemented modular prostheses, 1 allograft/prosthesis, 1 Coonrad-Morrey prosthesis; 9 total humerus reconstructions: 6 uncemented modular prostheses, 2 allograft/prostheses composites, 1 custom made prosthesis. The uncemented modular prosthesis used was the HMRS^® and the cemented modular was the MRS^®.

Histologically 24 were benign tumors and 253 primary malignant tumors.

All patients were periodically followed in the clinic, imaging studies and histology were reviewed and special attention given to prostheses-related complications and implant survival.

Univariate analysis through actuarial Kaplan Meier curves was used in evaluating implant survival to major complications. Functional results were assessed using the MSTS system.

Results: at a medium follow-up of 9 years 121 patients were NED, 14 NED1pm, 7 NED1lr, 1 NED1bm, 3 NED2pm, 2 NED2lr, 1 NED3pm, 102 died of disease, 19 died of other disease, 7 were lost to follow-up.

Major complications of the implants included 19 cases of deep infection (6.8%), 8 aseptic loosenings (2.9%), 4 breakages (1.4%) causing failure of the implants requiring revisions. Further complications were observed in revised cases. Actuarial curve of implant survival to major complications showed over 80% at 10 years and over 70% at 20 years.

Functional results according to the MSTS system were good or excellent (over 50%) in more than 90% of the patients, with an average score of 79%.

Conclusion: different techniques are available in reconstructions of the humerus after tumor resection, the indications depending on the type of resection and removal of soft tissues required.

Introduction: Based on their experience of over 25 years in musculoskeletal oncology the Authors review indications and problems of the different types of biopsies.

Methods: From the Rizzoli files and the literature most critical procedural problems and mistakes in performing biopsies are examined, with special attention to the consequences of mistakes and impact on treatment. Data of 749 consecutive cases of biopsies over a 12 year period were analysed: cases included were bone lesions with clinical and pathological features of malignancy requiring biopsy. Of these 198 had already had a biopsy elsewhere. Moreover the Rizzoli experience was reviewed in comparison to what reported in major studies in the literature.

Results: Of the 551 cases primarily biopsied at the Rizzoli 28 (5%) required a repeated biopsy. Of 198 cases biopsied elsewhere in 35 cases there was a major diagnostic error and in 18 a minor error (same grade of malignancy but different histotype).

Most common mistakes adversely affecting treatment were wrong skin incisions and/or surgical approach, amount and quality of the biopsy sample, infection.

Discussion: the analysis as well as major series reported in literature confirm that chosing the technique of biopsy and performing is not so simple. Critical task is first of all to properly chose the best technique:fine needle, trocar, incisional, frozen and excisional biopsies have proper indications, as well radioguidance or CT guidance or ultrasound guidance. Main needs are to avoid contamination, to provide an adequate sample of viable tissue and to place the biopsy tract so that it can be removed at definitive surgery. Today CT or MRI guided trocar biopsies are preferable for most bone lesions while ultrasound guided tru-cut biopsies in most soft tissue lesions. Mistakes concern the surgical approach, the site of biopsy, the quality of sample and tissue preservation. Most common mistakes of the unexperienced surgeon are to remove a lesion without a previous histology or to inadequately excise a soft tissue lesion.

Conclusions: Biopsy is the last step of staging before treatment and it is a compromise between the need of having significant tissue and the need to avoid contamination, yet this is by definition an intralesional procedure. Prof. Mario Campanacci used to say that biopsy is an important surgical procedure in the treatment of musculoskeletal tumors and it should be planned and performed by an experienced surgeon or radiologist.

Purpose of this study is to analyze the results of a modular reconstructive tumor prosthesis for the lower limb (GMRS^®) with a comparative statistical analysis of primary and secondary implants.

Material and methods: From October 2003 to September 2007 at Rizzoli 161 GMRS^® prostheses were implanted, most after resection of osteosarcoma (94 cases, 58%). It is a modular system with a rotating hinge mechanism for the knee, cemented and uncemented stems, in titanium and chromium-cobalt-molybdenum, curved and straight-fluted, with or without hydroxyapatite coating. Moreover adaptors are available to revise HMRS^® implants. This series includes 88 males and 73 females ranging in age from 9 to 80 years. Sites of reconstruction were 109 distal femurs, 19 proximal femurs, 1 total femur and 32 proximal tibias. There were 149 oncologic and 12 non oncologic diagnoses, including 96 primary reconstructions and 65 revisions after failure of previous implant. A retrospective analysis of imaging and complications was performed and functional results assessed according to MSTS system. Statistical analysis with Kaplan-Meier curves was used to study implant survival.

Results: At a mean follow up of 2 yrs. 106 patients are continuously NED, 31 are NED after treatment of one or more local recurrence or metastasis, 7 AWD, 5 DWD. There were 10 major complications: 8 infections (4,7%) (5 in primary and 3 in secondary implants, 2 previously infected) and 2 aseptic loosening (1,2%) (1 each). There were 9 minor complications (4 wound sloughs, 1 stiff joint, 3 disrupted extensor apparatus,1 patellar instability) requiring revision. Comparative statistical analysis of primary and secondary implants survival at major complications shows no statistically significant difference. Functional results were good or excellent in 95% of the evaluated patients, without any poor.

Conclusions: Middle term results are promising: good function, very low incidence of major complications, no breakage of implant components. This prosthetic reconstruction is indicated in oncological cases as well as in selected in some non oncological settings, such as challenging revisions of prosthetic failures with massive bone loss or post-radiation non unions or allografts failures. Although a higher incidence of complications was expected in secondary implants, statistical analysis shows similar survival.

Gorham-Stout disease (GSD) is rare, characterized by proliferation of vascular channels resulting in progressive distruction of bone. In the Rizzoli files we found 15 cases of GSD from 1968 to 2008. Two were excluded for insufficient documentation. For 13 cases clinical data, imaging and histology were analysed. Histopathologic features included benign vascular proliferation, vascular pattern of osteolytic angioma, fibro-connective tissue component and bony destruction. A final diagnosis was established based on clinical, radiological and histopathologic features.

Imaging included X-rays in 11 cases and CT or MRI in 5. All lesions were lytic, with associated sclerosis in two cases. There was one lesion only in 4 cases, multiple lesions in the same bone in 1 and multiple bones involved in 6. Primary sites were proximal femur in 7 cases, pelvis in 2, hip and knee, calcaneus, humerus and cervical spine in 1 case each. Two patients had no treatment, 2 conservative treatment (cast or brace), 5 surgery, 6 medical treatment (byphosphonates, calcitonin, zoledronic acid, interferon, steroids), 1 radiotherapy, 2 selective arterial embolization. Surgery consisted of internal fixation of pathologic fractures in 4 patients and reconstruction of the entire humerus with a double composite allograft in 1. Treatment was surgery only in 2 patients, medical treatment in 4 (1 also embolization), surgery and medical treatment in 2 (1 also embolization), radiotherapy only in 1, conservative treatment in 2. Four patients were lost at follow up. Mean follow up was 17 ys.(min 2, max 30) in 9 patients: 2 dead, 3 healed, 3 with stable disease, 1 alive with disease at 24 ys.

No conclusive treatment recommendations are possible; surgery is indicated in pathologic fractures or reconstruction of massively destroyed bones, medical treatment and selective embolization are helpful. In literature prosthetic reconstruction is preferred due to the risk of allografts resorption.

Solitary fibrous tumour (SFT) is a relatively uncommon mesenchymal neoplasm that most frequently arises in the pleura, but is also known to affect extrathoracic sites. About 15 % of SFT’s behave in an aggressive way, giving rise to local recurrence and/or distant metastasis. However, the behaviour of SFT remains unpredictable and due to the rarity of this tumour, it is difficult to define prognostic factors. The purpose of this study was to describe our experience with SFT, trying to define the pathologic features of this rare entity and better understand its clinical behaviour.

We performed a clinicopathologic review of all cases treated for a SFT at the Istituto Ortopedico Rizzoli in Bologna, between 1996 and 2008. We included 24 patients, nine males and fifteen females, ranging in age from 22 to 82 years (median 43.5 years). The anatomical sites involved were: the thigh (12 cases), shoulder region (four cases), gluteus (three cases), foot (two cases), extrapleural thoracic wall (two cases), and the lower leg (one case).

The tumour was > 5 cm in 15 cases, ranging in diameter from 2.5 cm to 18 cm (median 7.5 cm). Pain and swelling were the most frequently reported symptoms at presentation, with a mean duration of symptoms of 10 months. All patients were treated by excisional surgery (wide margins in 11, marginal margins in 13). Three patients had undergone pre-operative radiotherapy (44Gy) and one of these had also adjuvant radiotherapy after marginal excision of the tumour. Six tumours showed at least one atypical histologic feature (moderate to marked cytological atypia, extensive tumor necrosis, ≥ four mitoses per ten high-power fields, or infiltrative margins). On immunohistochemistry, 21 cases were positive for CD-34, 10 for CD-99, 17 for vimentin, three for CD-31, four for actin and one for S-100. Subsequent follow-up (average 33 months, range 5 to 112 months) revealed tumour relapse in only one case: a bone metastasis after 36 months of follow-up. The initial lesion was considered a large, deep, malignant SFT of the thigh, treated with wide surgical excision.

In the current review, including 24 extrathoracic solitary fibrous tumours, all lesions but one had a benign course. Nevertheless, this entity has a potential to recur or metastasize, and therefore careful long-term follow-up is necessary for all patients, even after wide excisional surgery. Although specific prognostic factors are yet to be defined, a high degree of suspicion for malignant behaviour is warranted for those cases in which atypical histologic features are present, particularly in the context of a deep tumor > 5cm in diameter.

Between 1983 and 2006 at Rizzoli 669 knee modular uncemented prostheses were implanted after resection of the distal femur, total femur or proximal tibia. These prostheses include 126 KMFTR prostheses and 543 second generation HMRS prostheses. Patients were followed periodically in the clinic. Data was obtained from clinical charts and imaging studies with special attention to major complications requiring revision surgery. Revision for polyethylene wear was considered a minor complication, since it did not imply failure of the implant. Functional results were assessed according to the MSTS system. Since data could be misleading due to deaths in an oncologic population (although 2/3 of patients were cured or long survivors), to censore the implant unrelated events Kaplan-Meyer curves of implant survival were studied.

In 126 KMFTR group infection rated 13.5%, stem breakage 13%, aseptic loosening 9.5%; change of polyethylene rated 44%. In 543 HMRS prostheses infection rated 8.6%, stem breakage 3%, aseptic loosening 4.8%; revision for polyethylene wear rated 9.6%. Techniques of revisions and their outcome analysed showed about 2/3 of good results, but increased risk of further complications in revised implants. Functional results (MSTS system) were good or excellent in 80% of KMFTR prostheses and in 90% of HMRS.

Decrease of major complications in newer prosthetic design was statistically significant and possibly due to newer materials and modified stem design. Polyethylene wear also decreased significantly. Function was satisfactory in most patients without complications for both groups. Revision surgery is technically demanding and appropriate timing of revision is crucial, since early treatment can improve final outcome.

From October 2003 to September 2007 at Rizzoli 161 GMRS^® prostheses were implanted after resections of the lower limb. This is a modular system with a knee rotating hinge mechanism, cemented and uncemented stems, in titanium and chromium-cobalt-molybdenum, curved and straight-fluted, with or without hydroxyapatite coating. Adaptors are available to revise HMRS^® prostheses with hybrid implants. This study includes 88 males and 73 females ranging in age from 9 to 80 years. Sites of reconstruction were 109 distal femurs, 19 proximal femurs, 1 total femur and 32 proximal tibias. There were 149 oncologic and 12 non oncologic diagnoses, including 96 primary reconstructions and 65 revisions for failures of previous reconstructions. Analysis of imaging and complications was performed and function assessed according to MSTS system. Kaplan-Meier curves were used to statistically evaluate implant survival.

At a mean follow up of 2.5 yrs. 106 patients are continuously NED, 31 NED after treatment of relapse, 7 AWD, 5 DWD. There were 10 major complications: 8 infections (4.7%) (5 in primary and 3 in secondary implants) and 2 aseptic loosenings (1.2%) (1 each). There were 9 minor complications requiring minor revisions. Comparative statistical analysis of implant survival showed no statistically significant difference between primary and secondary implants. Functional results were good or excellent in 95% of patients, without any poor.

Middle term results are promising with no breakages of implant components. Besides oncological cases, there are selected indications in non oncological settings, such as revisions of prosthetic or allografts failures. While a higher complication rate was expected in secondary implants, statistical analysis shows similar survival.

Forty-six hemangioendotheliomas (HE) of bone treated at Rizzoli from 1985 to 2004 were studied with minimum follow up of 4 years: 19 females and 27 males, mean age 37 years, mean follow-up 9 years, 35 cases unifocal at diagnosis (10 spine – 1 with lung metastasis also- 11 lower limb, 8 upper limb, 6 pelvis) and 11 with multifocal involvement. In 10 patients intralesional surgery was previously performed elsewhere. In 27 patients primarily treated at Rizzoli with unifocal localization, surgery was used in 15 cases, surgery and radiotherapy in 7, surgery with radio/chemotherapy in 1 and no surgery in 4 (2 radiotherapy, 1 radio/chemotherapy and 1 embolization). Eight unifocal patients already treated elsewhere had surgery in 3 cases, surgery and radiotherapy in 3, surgery with radio/chemotherapy in 1 and surgery plus chemotherapy in 1. Three of the unifocal cases had further bone involvement subsequently. Nine multifocal patients primarily treated at Rizzoli had surgery in 4 cases, surgery and radiotherapy in 4, surgery with radio/chemotherapy in 1. The 2 previously treated multifocal HE had 1 surgery and 1 radiotherapy.

Six patients died: 3 of disease, 1 of radio-induced osteosarcoma, 2 of different disease. Two patients are AWD. Of remaining 40 patients, 26 are NED (mean follow up 9 years), 11 NED after treatment of recurrence, 1 NED after treatment of radio-induced sarcoma. No lung metastases were diagnosed after treatment. All 10 cases previously treated intralesionally had recurrence. Two of 15 unifocal cases treated with surgery recurred (13%). None of 9 resected unifocal cases previously untreated recurred. Two of 21 pts. with radiotherapy (9.5%) had radio-induced sarcoma.

Surgery is recommended, resection when feasible. Radiotherapy, implying risk of induced sarcoma, should be reserved to multifocal or unresectable cases. Adverse prognostic factor was previous intralesional surgery.

Osteosarcoma is the most common second malignancy seen in retinoblastoma survivors. Risk of developing osteosarcoma in this group is estimated approximately 500 times higher than the general population. Prognosis in this setting has been reported significantly worse than conventional osteosarcoma despite multimodal management. Purpose of this study was to evaluate clinical features, molecular aspects and outcome of treatment in this subgroup of osteosarcoma patients.

Between 1985 and 2004, from a total of about 1100 osteosarcomas, 7 survivors of retinoblastoma developing high-grade osteosarcoma as second malignancy presented at the authors’ Institution. Retrospective study was undertaken to analyze presentation, tissue expression of RB1, P53, PGP and DHFR, treatment and outcome of both retinoblastoma and osteosarcoma.

Retinoblastoma was bilateral in 5 cases and unilateral in two. All the patients had been treated with a combination of surgery +/− chemotherapy +/− radiation.

None of them had evidence of retinoblastoma at the time of second malignancy diagnosis. Average age at diagnosis of osteosarcoma was 14 years (9–17 years), mean interval between the two malignancies was 155 months. All the osteosarcomas were in the appendicular skeleton, all but one around the knee. Molecular analysis showed defective RB1 gene in all cases All the seven patients received contemporary multimodal management for osteosarcoma. All but one patient died of osteosarcoma within 30 months from diagnosis. The living patient had local recurrence 9 years after limb salvage and is currently disease free following amputation.

Prognosis of osteosarcoma in retinoblastoma patients remains poor as compared to conventional high grade osteosarcoma despite multimodal management. No obvious correlation was found between poor prognosis and P53, PGP and DHFR expression.