Preservation of limb function after resection of malignant tumors in skeletally immature children is challenging [1, 9, 11, 12, 21, 25, 30]. Surgical options for this specific population have evolved from amputation and rotationplasty to limb salvage with endoprosthesis reconstruction and biologic (allograft and vascularized fibula) or expandable prostheses in the last decade [12, 13, 16, 20, 25]. Although allograft reconstructions are a valid option for reconstruction in skeletally immature patients, the complications are reported to be higher than in the adult population . In addition, the complication of limb length discrepancy at skeletal maturity related to the loss of a physis often cannot be avoided .
Some short- and midterm results of allograft reconstructions have been reported [3, 6, 7, 10], but those analyzing only reconstruction in children younger than 10 years old are scarce [11, 30]. Although we previously reported on complications resulting in second operations that are to be expected in children , we showed that this type of reconstruction resulted in acceptable function in four of 18 surviving patients with a minimum followup of 2 years; however, evaluation of these patients with longer followup has not been reported. The longer term complications such as the final limb length discrepancy at skeletal maturity were not included in our previous study.
We therefore sought to determine whether this type of reconstruction allows acceptable limb function in patients with malignant tumors followed for ≥ 10 years. To address this issue, we looked at the longer term followup of these patients and asked the following questions: what is (1) the limb length discrepancy at skeletal maturity and how was it managed; (2) the risk of amputation; (3) the risk of allograft removal; and (4) the risk of second surgery resulting from complications?
Patients and Methods
Between 1994 and 2006, we performed 22 bone allografts after bone sarcoma resections in children younger than 10 years of age (Table 1). Of those, none were lost to followup before the minimum followup of 10 years was reached, and an additional six had died of disease (of whom three died since our last report on this group of patients), leaving 16 patients whom we studied here. Followup on these patients was at a mean of 13.5 years (range, 10-22 years). During the period in question, no other treatments (such as extendible prostheses, amputations, etc) were used. The mean age at the time of the original surgery was 7 years (range, 2–10 years), and the mean age of the 16 surviving patients at last followup was 20 years (range, 15–28 years). This series included 10 boys and six girls who had a diagnosis of osteosarcomas in 14 and Ewing sarcoma in two. Ten reconstructions were performed with an intercalary allograft and six with an osteoarticular allograft. The growth plate was unaffected by the procedure in three of the patients with intercalary reconstruction, whereas in 13, the growth plate was compromised after resection and reconstruction.
Surgical treatment consisted of wide resection of primary bone sarcoma and the resulting bony defect was reconstructed with a fresh-frozen structural allograft. The type of reconstruction included intercalary and osteoarticular allografts. All allografts were harvested and packaged under sterile conditions and stored frozen at -80° C in the bone bank that is established at the authors’ institution according to a technique that has been previously described . No attempt was made to preserve the viability of the articular cartilage, and we performed bacteriologic and viral studies in accordance with the recommendations of the American Association of Tissue Banks and the tests available at the time. After tumor resection, the grafts were taken out of their packaging and placed directly in a warm normal saline solution. After being thawed, we cut the donor bone to the proper length and soft tissue structures were prepared for implantation. If the epiphysis was removed with the tumor resection, the joint was reconstructed by attaching the ligaments to the corresponding allograft tissues to improve stability. Reattachment of the allograft ligamentous tissue to the host tissue was performed using a direct continuous suture. A transverse osteotomy was performed in every patient and rigid fixation was used for stabilization. In patients in whom a growth plate would be sacrificed with resection, we attempted to overlengthen by placing an allograft that was 1 to 2 cm longer than the resected bone segment to potentially compensate for future limb length discrepancy.
Survivorship of the allograft free from revision or amputation was calculated at 10 years using a competing-risks survivorship estimator procedure . Data were drawn from a longitudinally maintained institutional registry and supplemented by in-office and telephone interviews to ascertain fractures, local recurrence, infection, and nonunion. Orthopaedic surgeons (LAA-T, GLF) who were involved in the clinical care interviewed patients by telephone or at their latest followup. At the latest followup, none of the surviving 16 patients was lost to followup and all patients reached skeletal maturity (Table 1). Limb length discrepancy at skeletal maturity was measured with full-length standing radiographs (Fig. 1), and data were collected by retrospective study of a longitudinally maintained institutional database. Lengthening was performed when the limb length discrepancy was > 3 cm and at least 4 years free of disease. Because all the patients whose limbs were elongated were patients with allografts of the distal femur (Fig. 1A), the bone was lengthened in the proximal part of the femur with a monolateral frame and to control the alignment during distraction, an intramedullary nail was used to avoid compromising the allograft. When the final planned length was obtained, the nail was locked and the external fixator was removed (Fig. 1B).
The risk of amputation, allograft removal, and secondary surgery resulting from a complication was calculated by a competing-risks analysis method  with limb salvage reconstruction censored at the time of failure or last followup (Fig. 2). The statistical analysis was performed using R programming language (R Foundation for Statistical Computing, Vienna, Austria) .
We observed no limb length discrepancy in the three patients who had their growth plates preserved (two femurs and one tibia); however, one patient developed malalignment that was treated with corrective osteotomy of the tibia. The remaining 13 patients developed limb length discrepancy as a result of loss of one growth plate. We did not perform epiphyseodesis in any of these patients. Seven patients (four femurs, two tibias, and one humerus) developed shortening of ≤ 3 cm (mean, 2.4 cm; range, 1–3 cm) and no lengthening was performed. Those patients were managed with a shoe lift. Six patients developed > 3 cm of limb discrepancy (all distal femoral reconstructions) (Fig. 1A). In four patients femoral lengthening was done (Fig. 1B), whereas two declined this procedure (each with 6 cm of shortening). In the four patients who underwent a lengthening procedure, one patient had a persistent discrepancy of 4 cm, whereas the other three had equal limb lengths at last followup.
Survivorship free from amputation was 96% (95% confidence interval [CI], 85%-100%) with a risk of amputation of 4% (95% CI, 0-15) (one amputation out of 22 patients). The cause for this amputation was local recurrence and it was described in the previous report .
Survivorship free from allograft removal was 85% (95% CI, 71-99). In these four patients, the allografts were removed and were described in the previous report (one infection, one local recurrence, and two fractures) .
The risk of a second operation resulting from a complication was 38% (95% CI, 19-57) (11 second operations out of 22 patients). Eleven patients underwent a second procedure as a result of a complication. Four allografts were removed and were previously reported; no further allograft removals occurred in this group of patients. Seven patients had revision of the allograft but the allograft was preserved. The procedures included two soft tissue resections for local recurrence, three fractures (two articular fractures and one diaphyseal), one nonunion, and one tibial deformity. The two local recurrences were in soft tissue with no contact with the reconstructions and were resected with wide margins maintaining the original allograft. In the three patients with fractures, one articular fracture was treated with a conventional hip prosthesis, whereas the other was treated with an ankle arthrodesis; a patient with a diaphyseal fracture was treated with revision of the internal fixation and two allograft struts were used to reinforce the original allograft. In the patient with a diaphyseal nonunion, a single operation was performed in which the internal fixation was replaced and an autogenous graft was added to the site, resulting in union at the host-donor junction. The tibial deformity was treated with an open-wedge osteotomy.
One of the 16 patients (Patient 5, Table 1) developed a second sarcoma in the scapula (Ewing sarcoma) 11 years after an osteosarcoma of the distal femur. After neoadjuvant chemotherapy, the scapula was resected with no further reconstruction. One year after ending chemotherapy, pulmonary metastases were detected, and the patient is actually undergoing chemotherapy with evidence of disease.
Preservation of limb function after resection of malignant tumors in skeletally immature children is challenging. Previously, we reported that resection of bone sarcomas and reconstruction with an allograft in patients younger than 10 years old was an alternative to endoprosthesis, rotationplasty, or amputation . The procedure had acceptable results regarding allograft survival and functional scores. However, long-term studies analyzing late complications and final limb length discrepancy in those patients in whom growth potential was altered as a result of the loss of one physis during tumor resection are scarce [11, 29]. Therefore, we studied our long-term results in these patients that were previously reported analyzing late complications and limb length discrepancy at skeletal maturity.
The limitations of this study include the small number of patients and the lack of a comparison with other types of reconstructions. However, because the incidence of bone sarcomas is low, it is difficult for a single center to accumulate a large number of patients to do a study such as this. We do, however, have substantial followup of these patients and the ability to carefully analyze issues of limb length inequality, which have been lacking in other published studies. Another limitation is that we included patients with tumors in different bones including the femur, tibia, and humerus so that the impact of limb length discrepancy differed depending on the site of resection. Obviously limb length is more of an issue in the lower extremity than the upper extremity so we only needed to surgically correct discrepancies in the femur. Another limitation is that surgeons involved in treating the patients were involved in reviewing their own work; however, we believe that the endpoints in question were pretty clear–limb length discrepancy, amputation–and so the risk of bias seems low. Finally, no patient-reported or validated outcomes tools are reported here, and without those, it is impossible to know the quality of life of these patients or how well they function. Despite that, the long-term results of objective data such as amputation, survival of reconstruction, and shortening make these results possible to compare with other types of reconstruction such as endoprosthesis and rotationplasty in future studies.
In our series, six of 13 patients who lost one physis developed > 3 cm of limb discrepancy (all distal femoral reconstructions); in four patients, femoral lengthening was done, whereas two refused this procedure (each with 6 cm of shortening). In one of the four patients who had lengthening, 4 cm of discrepancy persisted, whereas in the other three, the discrepancy was solved. Final limb length discrepancy at skeletal maturity is usually described as a major concern in young children in whom limb salvage procedures are performed in the lower limb [1, 12, 21, 25, 30]. Although epiphyseodesis is an alternative to lengthening [25, 29], we did not perform this on any of our patients. Patients with limb discrepancies in the upper extremity did not undergo any surgical treatment, because this is more of a cosmetic rather than a functional problem in the upper extremity . Although rotationplasty  is an option is this specific population, Benedetti et al.  described that excessive residual thigh-shank length in adult patients might lead to consideration of contralateral lengthening to improve functional results in some patients undergoing rotationplasty. Expandable prostheses have been developed to solve the problem of limb length discrepancy [2, 17, 24, 27]; however, recent publications showed high revision rates [8, 28] and subsequent bone loss . Also, around the knee, the adjacent growth plate must be crossed with a metal stem potentially altering growth in two epiphyses. Ruggieri et al.  analyzed 32 children with a mean age of 9 years with bone sarcomas of the femur treated with limb salvage using expandable prostheses. Three of the nine children who reached skeletal maturity had limb length equality and six had discrepancies of 15 to 30 mm. However, the rate of implant-related complications was 51% at a mean followup of 49 months. Cipriano et al.  analyzed 10 patients (mean age, 10 years) with distal femoral osteosarcoma using an Repiphysis® (Wright Medical Technology, Inc, Memphis, TN, USA) extendible prosthesis and found that bone loss around the stem of this prosthesis limits subsequent revision options, often resulting in a total femoral prosthesis. Neel et al.  described that implantation of a smooth, press-fit stem through the central portion of the uninvolved adjacent physis does not result in growth retardation or arrest; however, Arteau et al.  found that 15 of 23 (65%) of the immature patients reconstructed with a prosthesis of the distal femur experienced less proximal tibial growth in the operative compared with the contralateral limb. Biologic reconstruction has been described in children [3, 6, 7, 10, 14, 16, 18, 19], but there are only a few reports in children younger 10 years old that analyzed limb length discrepancy [11, 29]. Futani et al.  analyzed 33 patients in a multicentric retrospective study (22 who had endoprosthetic reconstruction [expandable and conventional endoprostheses] and 11 who had biologic reconstruction [allograft, vascularized fibula, and distraction osteogenesis]) with a mean age of 10 ± 1 years at the time of diagnosis and 22 ± 4 years at the time of latest followup. Nineteen (58%) patients had a limb lengthening procedure (12 endoprosthetic reconstruction and seven biologic reconstruction). Of the remaining 14 patients, eight preferred not to undergo limb lengthening, five needed an amputation, and one had a limb length discrepancy of < 2 cm. At the time of final followup, a limb length discrepancy of ≥ 2 cm was present in 15 of the patients who had lengthening and in eight others who had not had lengthening. San-Julian et al.  described 40 children younger than 10 years old with a minimum followup of 5 years; 31 of these patients had their limbs reconstructed after resection with allografts in 24 and with an endoprosthesis in seven. Lengthening was performed when limb length discrepancy was > 4 cm and patients were at least 3 years free of disease. They found that the most problematic location regarding ultimate growth was the distal femur.
In our series, the risk of amputation was 4% (95% CI, 0%-15%) and it was related to local recurrence. The only infection in our series was managed with resection of the reconstruction, a cemented spacer with antibiotics, and a second reconstruction. The main causes of amputation in limb salvage surgery are local recurrence and infection. A multiinstitutional report  described a 15% amputation rate in children younger than 10 years old with a total success rate for limb salvage of 85%, a success rate of 91% for the endoprosthetic reconstructions, and 73% for the biologic reconstructions. San-Julian et al.  reported four amputations in 40 patients younger than 10 years old with bone sarcomas as a result of three local recurrences and one infection.
The survivorship free from allograft removal was 85% (95% CI, 71-99) in the current review of our patients. These were the result of infection in one patient, local recurrence in another, and two patients had fractures of their allograft. Futani et al.  found a limb survival reconstruction rate of 77% (95% CI, 60%-95%) at 5 years and 51% (95% CI, 29%-73%) at 10 years for endoprosthetic reconstructions and 46% (95% CI, 16%-75%) at both 5 and 10 years for the biologic reconstructions. Interestingly, they found similar survival rates for both types of reconstructions at 10 years.
The survivorship free of a second surgery resulting from a complication was 62% (95% CI, 43%-81%) in our series. We observed that the new complications such as fractures and malalignment observed with longer followup could be solved without resection of the original allograft. San-Julian et al.  described 14 complications (four infections, four fractures, four nonunions, and two local recurrences) in 24 biologic reconstructions in children younger than 10 years old, although their minimum followup was 5 years. Futani et al.  found in their series that at 10 years, the rate of complications increased to 69% (95% CI, 47%-91%) for the endoprosthetic reconstruction group and to 86% (95% CI, 63%-109%) for the biologic reconstruction group. Cipriano et al.  described in 10 patients reconstructed with expandable prostheses 37 implant-related complications resulting in a total of 15 reoperations. Picardo et al.  described 16 of 55 patients (29%) reconstructed with expandable prostheses with complications; however, only 10 children (18.2%) underwent 11 revisions for complications, three for gearbox failure, four for subluxation of the femoral head, two for periprosthetic fracture, and two for infection.
In conclusion, limb length inequalities and subsequent procedures to correct them were common in this small series of very young patients as were complications resulting in revision operations, but overall most allografts remained intact at long-term followup. We consider allograft reconstruction after sarcoma resection an appropriate reconstructive procedure in selected skeletally immature children and it may be a reasonable option to consider as an alternative to rotationplasty or endoprosthesis in some patients. Future studies with long followup are needed to compare the available options with one another.
1. Abed R, Grimer R. Surgical modalities in the treatment of bone sarcoma in children. Cancer Treat Rev. 2010;36:342–347.
2. Abudu A, Grimer R, Tillman R, Carter S. The use of prostheses in skeletally immature patients. Orthop Clin North Am. 2006;37:75–84.
3. Alman BA, de Bari A, Krajbich JI. Massive allografts in the treatment of osteosarcoma and Ewing sarcoma in children and adolescents. J Bone Joint Surg Am. 1995;77:54–64.
4. Arteau A, Lewis VO, Moon BS, Satcher RL, Bird JE, Lin PP. Tibial growth disturbance following distal femoral resection and expandable endoprosthetic reconstruction. J Bone Joint Surg Am. 2015;97:e72(1–9).
5. Benedetti MG, Okita Y, Recubini E, Mariani E, Leardini A, Manfrini M. How much clinical and functional impairment do children treated with knee rotationplasty experience in adulthood? Clin Orthop Relat Res. 2016;474:995–1004.
6. Campanacci DA, Dursky S, Totti F, Frenos F, Scoccianti G, Beltrami G, Capanna R. Osteoarticular allografts in paediatric bone tumor reconstruction of the knee. J Biol Regul Homeost Agents. 2015;29(Suppl):111–119.
7. Campanacci L, Manfrini M, Colangeli M, Alí N, Mercuri M. Long-term results in children with massive bone osteoarticular allografts of the knee for high-grade osteosarcoma. J Pediatr Orthop. 2010;30:919–927.
8. Cipriano CA, Gruzinova IS, Frank RM, Gitelis S, Virkus WW. Frequent complications and severe bone loss associated with the Repiphysis expandable distal femoral prosthesis. Clin Orthop Relat Res. 2015;473:831–838.
9. Eleutério SJ, Senerchia AA, Almeida MT, Da Costa CM, Lustosa D, Calheiros LM, Barreto JH, Brunetto AL, Macedo CR, Petrilli AS. Osteosarcoma in patients younger than 12 years old without metastases have similar prognosis as adolescent and young adults. Pediatr Blood Cancer. 2015;62:1209–1213.
10. Erol B, Basci O, Topkar MO, Caypinar B, Basar H, Tetik C. Mid-term radiological and functional results of biological reconstructions of extremity-located bone sarcomas in children and young adults. J Pediatr Orthop B. 2015;24:469–478.
11. Futani H, Minamizaki T, Nishimoto Y, Abe S, Yabe H, Ueda T. Long-term follow-up after limb salvage in skeletally immature children with a primary malignant tumor of the distal end of the femur. J Bone Joint Surg Am. 2006;88:595–603.
12. Groundland JS, Binitie O. Reconstruction after tumor resection in the growing child. Orthop Clin North Am. 2016;47:265–281.
13. Haidar R, Sagghieh S, Muwakitt S, Mumtaz G, Akel S, Kamzoul R, Farah R, Noun P, Khalifeh H, Farhat F, Razzouk B, Rodriguez Galindo C, Rao B, Abboud MR. Limb salvage surgery for children and adolescents with malignant bone tumors in a developing country. Pediatr Blood Cancer. 2008;51:787–791.
14. Han CS, Chung DW, Lee JH, Jeong BO. Lengthening of intercalary allograft combined with free vascularized fibular graft after reconstruction in pediatric osteosarcoma of femur. J Pediatr Orthop B. 2010;19:61–65.
15. Hillmann A, Hoffman C, Gosheger G, Krakau H, Winkelmann W. Malignant tumor of the distal part of the femur or the proximal part of the tibia: endoprosthetic replacement or rotationplasty. Functional outcome and quality-of-life measurements. J Bone Joint Surg Am. 1999;81:462–468.
16. Jager T, Journeau P, Dautel G, Barbary S, Haumont T, Lascombes P. Is combining massive bone allograft with free vascularized fibular flap the children's reconstruction answer to lower limb defects following bone tumour resection? Orthop Traumatol Surg Res. 2010;96:340–347.
17. Kenan S, Bloom M, Lewis MM. Limb-sparing surgery in skeletally immature patients with osteosarcoma. The use of an expandable prosthesis. Clin Orthop Relat Res. 1991;270:223–230.
18. Kohler R, Lorge F, Brunat-Mentigny M, Noyer D, Patricot L. Massive bone allografts in children. Int Orthop. 1990;14:249–253.
19. Manfrini M, Gasbarrini A, Malaguti C, Ceruso M, Innocenti M, Bini S, Capanna R, Campanacci M. Intraepiphyseal resection of the proximal tibia and its impact on lower limb growth. Clin Orthop Relat Res. 1999;358:111–119.
20. Manfrini M, Tiwari A, Ham J, Colangeli M, Mercuri M. Evolution of surgical treatment for sarcomas of proximal humerus in children: retrospective review at a single institute over 30 years. J Pediatr Orthop. 2011;31:56–64.
21. Mercuri M, Capanna R, Manfrini M, Bacci G, Picci P, Ruggieri P, Ferruzzi A, Ferraro A, Donati D, Biagini R, De Maio M, Cazzola A, Campanacci M. The management of malignant bone tumors in children and adolescents. Clin Orthop Relat Res. 1991;264:156–168.
22. Muscolo DL, Ayerza MA, Aponte-Tinao L, Farfalli G. Allograft reconstruction after sarcoma resection in children younger than 10 years old. Clin Orthop Relat Res. 2008;466:1856–1862.
23. Neel MD, Heck R, Britton L, Daw N, Rao BN. Use of a smooth press-fit stem preserves physeal growth after tumor resection. Clin Orthop Relat Res. 2004;426:125–128.
24. Ness KK, Neel MD, Kaste SC, Billups CA, Marchese VG, Rao BN, Daw NC. A comparison of function after limb salvage with non-invasive expandable or modular prostheses in children. Eur J Cancer. 2014;50:3212–3220.
25. Ozger H, Bulbul M, Eralp L. Complications of limb salvage surgery in childhood tumors and recommended solutions. Strategies Trauma Limb Reconstr. 2010;5:11–15.
26. Picardo NE, Blunn GW, Shekkeris AS, Meswania J, Aston WJ, Pollock RC, Skinner JA, Cannon SR, Briggs TW. The medium-term results of the Stanmore non-invasive extendible endoprosthesis in the treatment of paediatric bone tumours. J Bone Joint Surg Br. 2012;94:425–430.
27. R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria; 2014. Available at: http://www.Rproject.org
. Accessed July 17, 2016.
28. Ruggieri P, Mavrogenis AF, Pala E, Romantini M, Manfrini M, Mercuri M. Outcome of expandable prostheses in children. J Pediatr Orthop. 2013;33:244–253.
29. San-Julian M, Dolz R, Garcia-Barrecheguren E, Noain E, Sierrasesumaga L, Canadell J. Limb salvage in bone sarcomas in patients younger than age 10: a 20-year experience. J Pediatr Orthop. 2003;23:753–762.
30. Weisstein JS, Goldsby RE, O’Donnell RJ. Oncologic approaches to pediatric limb preservation. J Am Acad Orthop Surg. 2005;13:544–554.