Intercalary defects reconstruction of the femur and tibia after primary malignant bone tumour resection. A series of 13 cases

Orthopaedics & Traumatology: Surgery & Research (2011) 97, 512 519 ORIGINAL ARTICLE Intercalary defects reconstruction of the femur and tibia after primary malignant bone tumour resection. A series of 13 cases O. Brunet, P. Anract, S. Bouabid, A. Babinet, V. Dumaine, B. Toméno, D. Biau Paris Descartes University, Cochin Hospital, Orthopaedic Surgery Department, 26, rue du faubourg Saint-Jacques, 75014 Paris, France Accepted: 22 March 2011 KEYWORDS Malignant Bone tumour; Autograft; Allograft; Bone tumour resection; Bone reconstruction Summary Introduction: Performing intercalary segment reconstruction after malignant bone tumour resection results in both mechanical and biological challenges. Fixation must be solid enough to avoid short-term or mid-term mechanical failure. The use of an allograft or autograft must ensure long-term survival of the reconstruction. The goal of this study was to analyse the clinical and radiological outcomes of these reconstructions. Patients and methods: Thirteen patients were operated on eight femurs and five tibias. The median age was 20 years old (range 14 50). The most common diagnosis was osteosarcoma. The median resection length was 15 cm (Q1 Q3: 6 26). A plate was used for fixation in nine cases and an intramedullary locked nail in four cases. An isolated bone autograft was used in two cases, an isolated bone allograft in one case, a dual autograft-allograft composite in six cases, and vascularised fibula and allograft combination in four cases. Results: The cumulative probability of union was 46% (95% CI: 0 99%) at 1 year; at the final follow-up, union was achieved in 12 patients (92%). Because of non-unions, 13 iterative procedures were needed to obtain these results. A non-displaced fracture of a cuboid-shaped tibial graft occurred in one patient, which was treated conservatively. Three infections occurred. Discussion: The results of intercalary segmental defects reconstruction after bone tumour resection were good, both from an oncologic and radiological point-of-view. One or more iterative procedures are sometimes needed to finally obtain bone union. We prefer to use a free rectangular cuboidal tibial graft since reconstruction with a vascularised autograft is technically more difficult. The choice of fixation methods is still controversial and no approach was found to be superior. Level of evidence: Level IV. Retrospective study. 2011 Published by Elsevier Masson SAS. Corresponding author. E-mail address: brunet.oceane@free.fr (O. Brunet). 1877-0568/$ see front matter 2011 Published by Elsevier Masson SAS. doi:10.1016/j.otsr.2011.03.021

Intercalary defects reconstruction of the femur and tibia 513 Table 1 Preoperative data. Patient Age (years) Gender Pathology Location Neoadjuvant chemotherapy Adjuvant chemotherapy Radiotherapy 1 23 F OO Femur Yes Yes No 2 23 F OO Femur Yes Yes No 3 18 F CO Femur Yes Yes No 4 22 M Ewing Femur Yes Yes Yes 5 31 M OO Femur Yes Yes No 6 17 M Ewing Femur Yes Yes Yes 7 20 M CO Femur No No No 8 50 F JCO Femur No No No 9 16 M OO Tibia Yes Yes No 10 18 F A Tibia No No no 11 20 M Ewing Tibia Yes Yes No 12 14 F OO Tibia Yes Yes No 13 18 F CO Tibia Yes Yes No OO: osteoblastic osteosarcoma; CO: chondroblastic osteosarcoma; JCO: juxtacortical osteosarcoma; A: adamantinoma. Introduction Primary, malignant bone tumours are rare. They represent less than 1% of all cancers and have an estimated incidence of six per million per year [1]. The life expectancy of patients operated for a malignant bone tumour has increased, particularly due to advances in chemotherapy. This has resulted in changes to the initial surgical management, with fewer amputations and more conservative treatments [2 4]. A prosthesis is often used to reconstruct the metaphysis or metaphysis-epiphysis junction, the most common regions for a tumour[5 7]. Conversely, tumours distal to the epiphysis are rarer. The possibility of joint preservation exists but post-resection reconstruction methods are not standardised. Malignant bone tumours located at the metaphysisdiaphysis junction of long bones in the lower limb are amenable to a resection and intercalary reconstruction procedure that preserves the proximal joint, however the long-term outcome of this reconstruction is not well defined. One aspect of this procedure requires that the fixation be solid enough to avoid mechanical failure of the reconstruction in the short and medium term. The location and extent of resection determine the choice of fixation methods [8,9]. The other aspect is the need for the bone allograft (cryopreserved or irradiated) or autograft to integrate into the host bone where the resection was performed and ensure the long-term survival of the reconstruction. Resection and intercalary reconstruction often results in mechanical problems: non-union rates range from 17 to 57% in various series [10 18]. These can be attributed to either mechanical (fixation stability) or biological (graft integration) failures. Secondary fractures are a common complication (between 9 and 19% in various series [10 15,19,20]) because the grafted and intercalated bone segments are not repopulated. These mechanical and biological failures arise from insufficient stability of the initial fixation [14,16,21]. Infection is also a common complication [15] of this lengthy and damaging surgery, which is often performed in immunosuppressed patients. We performed a retrospective study to evaluate the clinical and radiological outcomes of patients who underwent a resection and intercalary reconstruction of a primary bone tumour in the lower limb. The goal was to assess the time needed for bone union to occur and to critically analyse the mechanical, infectious and biological challenges of the different surgical methods used. Patients and methods This was a single centre, retrospective study on a series of 13 patients operated on between 1996 and 2005 in the orthopaedic surgery department of the Cochin Hospital in Paris, a group specialised in musculoskeletal tumour surgery. All of the procedures were performed by surgeons who were experienced in tumour surgery. The inclusion criteria were: patients with a primary malignant bone tumour of the lower limb; tumour located at the diaphysis, metaphysis or metaphysis-diaphysis junction; resection and intercalary reconstruction surgery. Patients with a secondary or benign bone tumour, tumour located at the epiphysis, patients where the bone resection left some bone continuity, patients with a prosthesis reconstruction or amputation were excluded. During this defined period, 13 of the 159 patients operated for a primary bone tumour in the lower limb corresponded to the inclusion criteria. The median age was 20 years old [range: 14 50]; there were six men and seven women. Osteosarcoma was the most common diagnosis (Table 1). Eight tumours were in the femur and five in the tibia. In eight cases, the tumour was located at the diaphysis (four in the femur and four in the tibia), in three cases the location was the metaphysis-diaphysis junction (two in the femur and one in the tibia) and in two case the location was the metaphysis (both in the femur). One patient (patient 6) had a metastatic disease with bone and lung involvement. Ten patients had adjuvant chemotherapy. Two patients had postoperative radiotherapy (patients 4, 6). Resection was performed according to the surgical principles for primary malignant bone tumours [22]. The tumour

514 O. Brunet et al. Figure 1 Length of the tibial and femoral resection. was removed as a block, leaving a layer of healthy tissue in contact with the tumour. The height of the bone cuts were based on preoperative MRI, performed before the chemotherapy if applicable, so that 2 cm of healthy bone remained. The scar and biopsy track were removed in block with the resected piece. The median length of resection (Fig. 1) was 15 cm (first and third quartile [Q1 Q3]: 6 26). The resected edges showed no signs of contamination in 12 patients and microscopic contamination in one patient (patient 5). Different surgical techniques were used to reconstruct the 13 resected bone segments. Three cortical-cancellous autografts were harvested from the tibia in the shape of a rectangular cuboid (patients 7, 8, 9). In two cases, two cortical rectangular cuboid pieces of about 10 cm in length were embedded into the proximal and distal shafts on both sides of the resection; in one case, only one piece was used. Two massive allografts were performed, an isolated one in the femur (patient 6) and one in the tibia in combination with an iliac crest cancellous bone autograft (patient 10). In four cases, a vascularised fibula bone graft was encased into a massive allograft (patients 1, 2, 11, 13). The fibula was included in the diaphysis of massive femur or tibia allografts and jutted out about 1 cm above the allograft, such that it was embedded into the two host bone segments on both Figure 3 Diagram of an intercalary reconstruction of a femur diaphysis performed with a vascularised fibula graft and plate/nails. 1: allograft sleeve opened over the entire length; 2: vascularised fibula; 3: fixation with plate and nails. Anract P, Vastel L, Tomeno B. Techniques and indications for bone and cartilage grafting and transplantation. EMC, Techniques Chirugicales Orthopédie Traumatologie. 44-030-A, 1999. sides of the resection (Fig. 2 and 3). Each time, the graft was performed as a second surgical procedure a few months after the resection. A cement spacer was used to stabilise the fixation between the two surgical procedures. Finally, in four cases (patients 3, 4, 5, 12), a massive allograft was used in combination with a cortical-cancellous rectangular cuboidal tibia graft harvested from the contralateral side. The 10 massive allografts (five femurs and Figure 2 Intercalary resection of an osteoblastic osteoscarcoma at the metaphysis-diaphysis junction of the femur and reconstruction with a plate, massive allograft and vascularised fibula (case 1). A: preoperative; B: first intervention: cement spacer; C: second intervention; D: 48-month follow-up.

Intercalary defects reconstruction of the femur and tibia 515 Table 2 Peri- and postoperative data. Patient Resection Fixation Graft Complications 1 Diaphysis Metaphysis Plate Allo + VF Non-union ALLI Compartment syndrome 2 Diaphysis Plate Allo + VF No Metaphysis 3 Diaphysis Intramedullary locked nail Allo + auto cortical-cancellous No 4 Diaphysis Intramedullary locked nail Allo + auto cortical-cancellous Non-union Fracture 5 Diaphysis Plate Allo + auto cortical-cancellous No 6 Diaphysis Intramedullary locked nail Allo Non-union 7 Metaphysis Plate Auto cortical-cancellous Sepsis Hematoma 8 Metaphysis Plate Allo + auto cortical-cancellous No 9 Diaphysis Plate Auto cortical-cancellous Fracture 10 Diaphysis Intramedullary locked nail Allo + cancellous auto (iliac crest) No 11 Diaphysis Plate Allo + VF Non-union Sepsis LPN paralysis 12 Diaphysis Plate Allo + auto cortical-cancellous Sepsis Metaphysis 13 Diaphysis Plate Allo + VF No VF: vascularised fibula; ALLI: acute lower limb ischemia; LPN: lateral popliteal nerve. five tibias) were taken from the bone bank at Cochin hospital and had been previously frozen and irradiated. Cancellous bone was added to improve the strength of the fixation in eight cases. In six patients, the bone came from the contralateral tibia (patients 3, 4, 5, 7, 9, 12), the ipsilateral iliac crest in one patient (patient 10) and a banked femoral head in one patient (patient 8). This bone was packed in at the junction between the host bone and cortical graft. In cases of tibial resection, a graft between the tibia and fibula was systematically performed with this cancellous bone. A plate was used for fixation in nine patients (five femoral and four tibial reconstructions) and an intramedullary locked nail in four patients (three femoral and one tibial reconstruction) (Table 2). When an allograft was used, the plates were set up as a bridge with typically one or two screws at the end of each allograft and at least three screws on both sides of the resection (Fig. 4). The nails acted as a Figure 4 Intercalary resection of an osteoblastic osteoscarcoma at the metaphysis-diaphysis junction of the tibia and reconstruction with a plate, massive allograft and cortical-cancellous autograft (case 12). A: preoperative; B: immediately postoperative; C: 39-month follow-up.

516 O. Brunet et al. Figure 5 Intercalary resection of a chondroblastic osteoscarcoma at the femur diaphysis and reconstruction with an intramedullary locked nail, massive allograft and cortical-cancellous autograft (case 3). A: preoperative; B: immediately postoperative; C: 39-month follow-up. bridge over the bone graft and were systematically locked at the distal and proximal ends (Fig. 5). The primary outcome was the time needed for union in the reconstruction. A union was defined as the presence of fusion between the host bone and graft at both ends and full weight-bearing without pain. The time was calculated from the resection date, including for procedures done in multiple steps such as the vascularised fibula graft. Secondary outcome measures were the cumulative probability of surgical revision, independent of the reason, and the cumulative probability of surgical revision for mechanical reasons only. The cumulative probability of events under consideration were estimated by taking into account competitive risks, notably death, and depending on the event of interest, revisions due to infection, tumour or other reasons [23]. The cumulative probability estimates are given with their 95% confidence intervals (95% CI). Quantitative data are reported with the median and first and third quartile (Q1 Q3), except if specified otherwise. The analyses were performed with the R software [24]. Results of surgical revision, independent of the reason, was 38% at 1 year (95% CI: 0% 97%) and 88% at 5 years (1% 100%). The cumulative probability of surgical revision for mechanical reasons was 23% (95% CI: 0% 99%) at 1 year and 73% at 5 year (95% CI: 0% 100%). Ten patients were re-operated for a total of 22 procedures. Thirteen procedures in eight patients were performed for mechanical reasons, six procedures were performed for infection reasons in three patients and three procedures were performed in one patient for other reasons. Seven procedures for mechanical reasons and two for infection reasons were performed in four patients where union was not initially achieved. Three patients (patients 1, 4, 8) required decortication with a cortical-cancellous bone autograft from the ipsilateral iliac crest; two of these patients needed the fixation material changed and one did not. The fourth patient (patient 11) required six procedures, four for mechanical reasons and two for infection reasons, before union was achieved. The first procedure consisted of the second surgical intervention for the initial reconstruction, with placement of a vascularised fibula inside an allograft sleeve. Three years and four months after tumour The median duration of hospitalisation was 12 days (Q1 Q3: 9 21). The median follow-up was 48 months (Q1 Q3: 24 131). At the last follow-up, 12 patients were still alive with no evidence of disease and one patient (patient 6) was alive with on-going disease. There was no morbidity at the autograft donor sites. There was no local tumour recurrence. The cumulative probability of union was 46% (95% CI: 0% 99%) at 1 year and 90% (95% CI: 0% 100%) at 5 years (Fig. 6). At the last follow-up, 12 patients (92%) had graft union. The union occurred without further procedures in nine patients and after one additional procedure in four other patients. One patient (patient 6) did not have union at the last follow-up. However, no procedure was planned because of this patient s poor general status (Ewing sarcoma with lung and bone metastases). The cumulative probability Figure 6 Cumulative incidence of the recorded events.

Intercalary defects reconstruction of the femur and tibia 517 Table 3 Published results. Series Case Average follow-up (months) Type of graft Non-union (%) Sepsis (%) Fracture (%) Local recurrence (%) Kreig [9] 16 49 Irradiated massive auto + VF 16 0 6 0 Gebhardt [14] 53 25 Frozen massive allo 22 30 11 9 Muscolo [18] 59 60 Frozen massive allo 15 5 7 10 Donati [20] 113 12 Frozen massive allo 57 14 15.5 0 Enneking [25] 40 42 Cortical autograft + VF 33 5 20 0 Sugiura [26] 19 57 Sterilised massive auto 32 5 10 0 Ahmed [27] 31 69 Sterilised massive auto 6 16 10 0 Manabe [28] 23 52 Sterilised massive auto 8 20 12 0 Khattak [29] 12 49 Sterilised massive auto 8 41 0 16 Sugiura [30] 15 48 Sterilised massive auto 14 13 13 0 Uyttendaele [31] 15 60 Irradiated massive auto 6 13? 13 Araki [32] 20 45 Irradiated massive auto 20 15 5 0 Auto: autograft; Allo: allograft; VF: vascularised fibula graft. resection, a varus seesaw motion of the leg axis to 18 required treatment for non-union with decortication and cortical-cancellous autograft and changing of the fixation hardware. An Enterobacter cloacae and Propionibacterium acnes infection occurred 4 months after this procedure. This presented as a skin fistula, which first required lavage and removal of the fixation hardware (plate) and implantation of a gentalline loaded cement spacer, which was covered by a posterolateral sural skin-fascia flap. Repeated lavage was required because CRP levels did not return to normal, along with placement of a gentalline cement spacer and a skin graft. Piperacillin and peflacin antibiotic treatment was initiated for 2 months. Finally, the two-step Masquelet technique resulted in bone union. At the last follow-up, 35 months after the last surgical procedure, there were no clinical or biological signs suggesting that the infection had recurred. Six procedures because of an infection were performed in three patients (patients 7, 11, 12). One patient (patient 7), had an early coagulase-negative staphylococcus infection on the fourth postoperative day. Surgical debridement was performed, followed by triple antibiotic treatment (vancomycin, pefloxacin, clindamycin) for 10 days. At the last follow-up, 62 months after the last surgical procedure, the infection had not reoccurred. One patient (patient 11) developed a multi-microbial infection with E. cloacae and P. acnes on the tibia reconstruction, which occurred 3 years and 8 months after the tumour was resected. The two procedures, because of infection in this patient, were described above. Finally, one patient (patient 12) had a wound dehiscence on the 6th postoperative week that was initially treated by excision and lavage. A medial saphenous skinfascia flap was performed 5 weeks later when the dehiscence recurred. An infection (coagulase-negative Staphylococcus lugdunensis) occurred during the 7th month postoperative that required surgical debridement and antibiotic therapy with vancomycin during 7 days, then ofloxacin and lincosanide for 4 months. At the last follow-up, 38 months after the surgical procedure, the infection had not recurred. Six other procedures were performed for mechanical failures: three procedures consisted of the second intervention of the initial reconstruction with placement of a vascularised fibula graft with allograft sleeve; two hardware removal procedures; one total knee arthroplasty with prosthesis for osteoarthritis. Three procedures were performed for other reasons in a single patient (patient 1): a femoro-popliteal bypass using a saphenous vein was needed as a second intervention on the day of the reconstruction procedure because of acute ischemia of the operated lower limb due to superficial femoral artery thrombosis during the anastomosis of the vascularised fibular graft; an aponeurosis release had to be performed the next day because of compartment syndrome secondary to revascularisation of this bypass; another procedure for a keloid scar was performed later on. A nondisplaced fracture in the cuboid tibial autograft occurred in one patient (patient 9); unloading of the limb for one and half months was sufficient to obtain union. Discussion In this study, clinical and radiological results in 13 patients were evaluated after resection of a primary bone tumour in the lower limb followed by intercalary reconstruction with a bone graft. This work has limitations that can affect the internal and external validity of the results. First, the small number of patient did not allow us to analyse the effect of the different co-variables, such as the type of graft or fixation hardware, on the time needed for bone union. Second, the reconstructions were heterogeneous and the estimated results have a certain amount of uncertainty because of the small number of subjects analysed. However, these limitations are directly related to the low incidence of malignant bone tumours and they are prevalent in other published series. Primary malignant bone tumours are truly rare and their initial surgical management has become more conservative [2 4]. These tumours are amenable to a resection and intercalary reconstruction procedure that spares the neighbouring joint, but reconstruction lead to particular problems and the long-term outcome is not well defined. Non-unions, infections and fractures are the most common complications

518 O. Brunet et al. that require one or more additional procedures (Table 3). In retrospective series of 53 to 112 patients, where results of the resection and intercalary reconstruction of malignant bone tumours are reported with an average follow-up of 25 to 63 months, the non-union rate of massive allografts at the final follow-up was from 1 to 17% [14 16,20]. When isolated autografts are used in the reconstruction, retrospective series with eight to 40 patients with an average follow-up of between 20 and 69 months reported a non-union rate between 0 and 10% at the last follow-up [9,25 33]. The following factors were listed as being related to delayed union: type of graft, fixation, length of resection, associated chemotherapy [17]. Union occurs more quickly when a vascularised fibula graft is used [26,34,35], but instead of this long and difficult technique, we prefer to add a rectangular cuboid-shaped piece of tibia to the allograft, as this can be carried out easily, induces little pain, and provides a sufficient amount of cancellous and cortical bone. Different reconstruction techniques have been used with similar results [8,9,13,15,18] (Table 3). There were three infection cases that resolved after one or more surgical revisions and appropriate antibiotic treatment. In a retrospective series with 53 patients after resection and intercalary reconstruction of malignant bone tumours by allograft, Gebhardt [14] found a 30% infection rate, despite a shorter follow-up than ours (25 versus 55 months). In cases of reconstruction with an isolated autograft, retrospective series with 12 to 40 patients found an infection rate between 0 and 42% [9,25 32]. The infection rate seems related to the length of the resection and recourse to adjuvant chemotherapy [14]. It is less when a vascularised fibula graft is used [35]. Our results are comparable to studies reporting on results of reconstruction using isolated autografts (Table 3). One case of fracture was reported, which was treated conservatively. The fracture rate ranged from 9 to 19% in retrospective series with 26 to 120 patients where reconstruction was carried out with massive allografts after intercalary resection of malignant bone tumours [10 14,19,20,36] (Table 3). When isolated autografts are used, retrospective series with 12 to 40 patients have found a fracture rate ranging from 0 to 45% [9,25 30,32]. The length of the resection, type of graft and type of fixation were reported as being related to the risk of secondary fracture [21,25,37 42]. It seems that fixation with plates or intramedullary nails are less likely to result in fractures than fixation with screws alone [9,32]. We prefer to use intramedullary locked nails when resection is performed at the diaphysis, and plates or reconstruction with a vascularised fibula in metaphysis-diaphysis junction cases. We no longer recommend a two-step procedure because of the difficulty encountered during secondary dissection and as a consequence, increased risk of acute ischemia in the lower limb following the surgery. Conclusion Mechanical results of intercalary segment reconstruction after resection of a bone tumour were satisfactory, with bone union occurring in more than 90% of patients. However, union takes a long time and secondary procedures are often needed because of non-union or infection. Patients must be informed of these risks. We recommend adding a rectangular cuboid tibial autograft to the intercalary allograft during a single surgical procedure, as this is simpler to perform than a vascularised fibula graft. The intercalary allograft and fixation ensure immediate mechanical stability until the autograft can integrate sufficiently to ensure the long-term survival of the reconstruction. The use of an intramedullary locked nail is favoured after resection at the diaphysis, while the use of plates should be limited to reconstruction after resection at the metaphysis-diaphysis. Disclosure of interest The authors declare they have no conflicts of interest concerning this article. References [1] Anract P, Golwasser F. Epidémiologie des cancers osseux. In: Cancers osseux, p. 2. Édité par John Libbey, Paris, 2007. [2] Harris IE, Leff AR, Gitelis S, Simon MA. Function after amputation, arthrodesis, or arthroplasty for tumours about the knee. J Bone Joint Surg Am 1990;72:1477 85. [3] Rougraff BT, Simon MA, Kneisl JS, Greenberg DB, Mankin HJ. Limb-salvage compared with amputation for osteosarcoma of the distal end of the femur. A long-term oncological, functional, and quality-of-life study. J Bone Joint Surg Am 1994;76:649 56. [4] Simon MA, Aschliman MA, Thomas N, Mankin HJ. Limb-salvage treatment versus amputation for osteosarcoma of the distal end of the femur. J Bone Joint Surg Am 1986;68:1331 7. [5] Biau DJ, Faure F, Katsahian S, Jeanrot C, Tomeno B, Anract P. Survival of total knee replacement with a megaprosthesis after bone tumour resection. J Bone Joint Surg Am 2006;88:1285 93. [6] Kassab M, Dumaine V, Babinet A, Ouaknine M, Tomeno B, Anract P. Twenty nine shoulder reconstructions after resection of the proximal humerus for neoplasm with mean 7-year follow-up. Rev Chir Orthop Reparatrice Appar Mot 2005;91(1):15 23. [7] Anract P, Coste J, Vastel L, Jeanrot C, Mascard E, Tomeno B. Proximal femoral reconstruction with megaprosthesis versus allograft prosthesis composite. A comparative study of functional results, complications and longevity in 41 cases. Rev Chir Orthop Reparatrice Appar Mot 2000;86:278 88. [8] Toméno B, Gerber C. Les résections-reconstructions diaphysaires des grands os des membres en pathologie tumorale. À propos de 23 cas. Rev Chir Orthop Reparatrice Appar Mot 1987;73:131 6. [9] Kreig AH, Davidson AW, Stalley PD. Intercalary femoral reconstruction with extracorporeal irradiated autogenous bone graft in limb-salvage surgery. J Bone Joint Surg Br 2007;89:366 71. [10] Cara JA, Laclériga A, Cañadell J. Intercalary bone allografts. Twenty-three tumour cases followed for 3 years. Acta Orthop Scand 1994;65:42 6. [11] Mnaymneh W, Malinin TI, Lackman RD, Hornicek FJ, Ghandur- Mnaymneh L. Massive distal femoral osteoarticular allografts after resection of bone tumours. Clin Orthop Relat Res 1994;303:103 15. [12] San-Julian M, Cañadell J. Fractures of allografts used in limb preserving operations. Int Orthop 1998;22:32 6. [13] Vander Griend RA. The effect of internal fixation on the healing of large allografts. J Bone Joint Surg Am 1994;76:657 63. [14] Gebhardt MC, Flugstad DI, Springfield DS, Mankin HJ. The use of bone allografts for limb-salvage in high-grade extremity osteosarcoma. Clin Orthop Relat Res 1991;270:181 96.

Intercalary defects reconstruction of the femur and tibia 519 [15] Ortiz-Cruz E, Gebhardt MC, Jennings LC, Springfield DS, Mankin HJ. The results of transplantation of intercalary allografts after resection of tumours. A long-term follow-up study. J Bone Joint Surg Am 1997;79:97 106. [16] Donati D, Di Liddo M, Zavatta M, Manfrini M, Bacci G, Picci P, et al. Massive bone allograft reconstruction in high-grade osteosarcoma. Clin Orthop Relat Res 2000;377:186 94. [17] Hornicek FJ, Gebhardt MC, Tomford WW, Sorger JI, Zavatta M, Menzner JP, et al. Factors affecting non-union of the allografthost junction. Clin Orthop Relat Res 2001;382:87 98. [18] Muscolo DL, Ayerza MA, Aponte-Tinao L, Ranalletta M, Abalo E. Intercalary femur and tibia segmental allografts provide an acceptable alternative in reconstructing tumour resections. Clin Orthop Relat Res 2004;426:97 102. [19] Mankin HJ, Doppelt SH, Sullivan TR, Tomford WW. Osteoarticular and intercalary allograft transplantation in the management of malignant tumours of bone. Cancer 1982;15(50):613 30. [20] Donati D, Capanna R, Campanacci D, Del Ben M, Ercolani C, Masetti C, et al. The use of massive bone allografts for intercalary reconstruction and arthrodeses after tumour resection. A multicentric European study. Chir Organi Mov 1993;78:81 94. [21] Ozaki T, Hillmann A, Bettin D, Wuisman P, Winkelmann W. Intramedullary, antibiotic-loaded cemented, massive allografts for skeletal reconstruction. Twenty-six cases compared with 19 uncemented allografts. Acta Orthop Scand 1997;68:387 91. [22] Vichard P, Cagneux E. Traitement chirurgical des tumeurs des os. In: Encyclopédie Médico-Chirurgicale Appareil locomoteur 2001; volume 5, p. 14 701. Édité par Elsevier SAS. [23] Biau DJ, Latouche A, Porcher R. Competing events influence estimated survival probability: when Kaplan-Meier analysis appropriate? Clin Orthop Relat Res 2007;462:229 33. [24] R Development Core Team (2006). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.r-project.org. [25] Enneking WF, Eady JL, Burchardt H. Autogenous cortical bone grafts in the reconstruction of segmental skeletal defects. J Bone Joint Surg Am 1980;62:1039 58. [26] Sugiura H, Yamamura S, Sato K, Katagiri H, Nishida Y, Nakashima H, et al. Remodelling and healing process of moderately heat-treated bone grafts after wide resection of bone and soft-tissue tumours. Arch Orthop Trauma Surg 2003;123:514 20 [Epub 2003 Jul 3]. [27] Ahmed AR, Manabe J, Kawaguchi N, Matsumoto S, Matsushita Y. Radiographic analysis of pasteurised autologous bone graft. Skeletal Radiol 2003;32:454 61. [28] Manabe J, Ahmed AR, Kawaguchi N, Matsumoto S, Kuroda H. Pasteurised autologous bone graft in surgery for bone and softtissue sarcoma. Clin Orthop Relat Res 2004;419:258 66. [29] Khattak MJ, Umer M, Haroon-ur-Rasheed, Umar M. Autoclaved tumour bone for reconstruction: an alternative in developing countries. Clin Orthop Relat Res 2006;447:138 44. [30] Sugiura H, Takahashi M, Nakanishi K, Nishida Y, Kamei Y. Pasteurised intercalary autogenous bone graft combined with vascularised fibula. Clin Orthop Relat Res 2007;456:196 202. [31] Uyttendaele D, De Schryver A, Claessens H, Roels H, Berkvens P, Mondelaers W. Limb conservation in primary bone tumours by resection, extracorporeal irradiation and re-implantation. J Bone Joint Surg Br 1988;70:348 53. [32] Araki N, Myoui A, Kuratsu S, Hashimoto N, Inoue T, Kudawara I, et al. Intraoperative extracorporeal autogenous irradiated bone grafts in tumour surgery. Clin Orthop Relat Res 1999;368:196 206. [33] Böhm P, Fritz J, Thiede S, Budach W. Re-implantation of extracorporeal irradiated bone segments in musculoskeletal tumour surgery: clinical experience in eight patients and review of the literature. Langenbecks Arch Surg 2003;387:355 65. [34] Chang DW, Weber KL. Use of a vascularised fibula bone flap and intercalary allograft for diaphyseal reconstruction after resection of primary extremity bone sarcomas. Plast ReconstrSurg 2005;116:1918 25. [35] Chen CM, Disa JJ, Lee HY, Mehrara BJ, Hu QY, Nathan S, et al. Reconstruction of extremity long bone defects after sarcoma resection with vascularised fibula flaps: a 10-year review. Plast Reconstr Surg 2007;119:915 24. [36] Berrey Jr BH, Lord CF, Gebhardt MC, Mankin HJ. Fractures of allografts. Frequency, treatment, and end-results. J Bone Joint Surg Am 1990;72:825 33. [37] Bernd L, Sabo D, Zahlten-Hinguranage A, Niemeyer P, Daecke W, Simank HG. Experiences with vascular pedicled fibula in reconstruction of osseous defects in primary malignant bone tumours. Orthopade 2003;32:983 93. [38] Pollock R, Stalley P, Lee K, Pennington D. Free vascularised fibula grafts in limb-salvage surgery. J Reconstr Microsurg 2005;21:79 84. [39] O Brien BM, Gumley GJ, Dooley BJ, Pribaz JJ. Folded free vascularised fibula transfer. Plast Reconstr Surg 1988;82:311 8. [40] Ceruso M, Falcone C, Innocenti M, Delcroix L, Capanna R, Manfrini M. Skeletal reconstruction with a free vascularised fibula graft associated to bone allograft after resection of malignant bone tumour of limbs. Handchir Mikrochir Plast Chir 2001;33:277 82. [41] Abe S, Tateishi A, Tokizaki T, Takeyama S, Nakano H, Matsushita T. Mechanical failure of unsupported Huckstep intramedullary nail in malignant bone tumour reconstruction. Clin Orthop Relat Res 2001;393:272 8. [42] Wunder JS, Davis AM, Hummel JS, Mandelcorn J, Griffin AM, Bell RS. The effect of intramedullary cement on intercalary all ograft reconstruction of bone defects after tumour resection: a pilot study. Can J Surg 1995;38:521 7.