Reconstruction after distal femoral resection for patients in early childhood is challenging. Rotationplasty is a viable reconstruction option, offering generally acceptable long-term functional outcomes (9), but this surgical procedure is not always optimal as it is not always accepted emotionally by patients and their guardians. Alternatively, an osteo-articular allograft can be used for reconstruction after distal femoral removal, but this procedure is susceptible to complications. In a retrospective study of allograft reconstructions for sarcoma patients younger than 10 years by Aponte-Tinao et al., four of the seven patients undergoing osteo-articular allograft reconstruction with long-term follow-up experienced major complications: two fractures, one infection, and one non-union (2). In our case, mega-prosthetic reconstruction was not feasible due to the intramedullary diameter of the affected femur being too narrow, and rotationplasty was refused. As allograft supply is rather limited in Japan, we had to seek an alternative option to act as a bridge until mega-prosthetic reconstruction became feasible.
The use of a temporary spacer after the removal of the distal femoral tumor was first reported in 1982 by Volz et al., but their concept differed from ours. They used a temporary spacer for a 31-year-old woman with a recurrent malignant bone tumor as a stabilizing filler for only 13 weeks until an appropriate customized implant was prepared (10). In 2015, Chung et al. published a case series of temporary hemiarthroplasty utilizing synthetic mesh for children with osteosarcoma around the knee. The authors performed hemiarthroplasty with or without an allograft in 12 children aged 8–12 years with distal femoral osteosarcoma, four of which eventually underwent non-expandable tumor prosthetic reconstruction 23–47 months after hemiarthroplasty. Synthetic mesh was used to reconstruct the soft tissue and stabilize the joint. They reported good clinical outcomes, with MSTS scores of 24–28 points and the most recent LLD of 1.0–4.6 cm at 29–76 months after the first surgery (5).
In our case, our first hand-made spacer was dislocated at 7 months post-operatively. Due to the misfitting of the first spacer to the proximal tibia, we chose a commercially-available custom-made spacer (Figure 4) for the second surgery. It was characterized by its fitting the anatomical shape of the distal femur, its smooth surface suitable for articulation, and an anti-rotation fin at the proximal end of its body. Due to the absence of anti-rotation screw(s) and looseness between the anti-rotation fin and the residual bone, we decided to partially fix the spacer using PMMA to prevent rotation at the bone-spacer junction. In addition, we sought to achieve the formation of an osseous bridge over the bone-spacer junction by covering the junction with the periosteum and induced membrane. Instability of the knee joint can be an issue without ligamentous reconstruction, but we did not use synthetic mesh as a fibrous pseudo-capsule had already formed at the time of the second hemiarthroplasty. After the second hemi-arthroplasty, the patient could begin ambulation with weight bearing on the affected leg through use of hinged knee brace but without any other support, and this functioned well for 18 months, allowing periosteal bone formation over the junction and the successful mega-prosthetic reconstruction with insertion of a durable stem.
The use of PMMA may have played an essential part to achieve durable bone-prosthesis junction. In 2003, Masquelet reported 35 cases of long bone defect successfully treated with 2-step reconstruction: cement spacer insertion followed by autologous cancellous bone grafting. The concept is that spacer is responsible for the formation of synovium-like membrane, which can prevent grafted bone from resorption and favor its revascularization and corticalization (11). Since then, “Masquelet technique” has become a standard treatment for long bone defect. In our case, we also preserved induced membrane around PMMA and grafted cancellous bone tips under them at the third surgery. After surgery, the cortex of the residual bone became satisfactorily thickened and peri-osteal/peri-membranous bone formation overlapped the bone-prosthesis junction. It is difficult to prove that such successful bone formation is attributed to cementation and following membrane formation, but we believe that temporal spacer accompanied by partial cementation is better than immediate prosthetic reconstruction in terms of bone formation around the bone-prosthesis junction.
The functional results of our case were slightly worse than those reported by Chung et al., but such inferior outcomes can mostly be attributed to the younger age of our case, along with the instability of the first spacer. The surgical procedure and post-operative management are obviously more difficult in younger patients due to their smaller bone size and expected longer LLD. Our case, however, indicates that hemiarthroplasty can be applied to younger children with malignant distal femoral tumor by utilizing the aforementioned surgical technique.
In conclusion, this case report provides an example of limb salvage surgery after distal femoral resection in a small child. Along with allograft reconstruction, the use of a temporary spacer utilizing partial cementation and preservation of the periosteum and induced membrane after PMMA removal appears to afford a viable limb-salvage option after distal femoral resection for younger children refusing femoral rotationplasty.