With the population aging and an increasing desire for a high quality of life, millions of patients who suffered knee arthropathy have undergone total knee arthroplasty (TKA). Although the success rate for primary TKA has been quite high, the number of revision surgeries can not be ignored. Bone loss is one of the challenges confronting surgeons who perform revision TKAs as it can impair alignment accuracy and the long-term stability of the implant. Although several options are available for bone loss after TKA—cement, bone grafting, standard augments, and hinged implants—there is no single ideal option available for all patients with a severe bone defect because of the variety and severity of the defects.¹The burgeoning rapid prototyping (RP) technique, which has been used to fabricate components with complex and unique structures, may offer a novel option in these cases.

A 56-year-old woman underwent simultaneous bilateral TKA. She presented to another hospital 2 years later with pain and swelling of the right knee. An infection in the prosthesis was identified. The area was debrided, all implants were removed, and a non-articulating spacer made of antibiotic cement was implanted.

When the patient came to our department, her right knee had a constrained range of motion (ROM 20°–70°) with flexion and varus deformities. Although the infection had been well controlled for more than 6 months, severe bone loss was found on plain radiography (Figure 1A). Computed tomography (CT) confirmed the presence of bone defects on both femoral and tibial sides. According to the Anderson Orthopaedic Research Institute classification,²the defects were classified as F2b and T2b, respectively. Traditional options were not satisfactory for this case because of the large size and irregular shape of the bone defect. Therefore, we decided to work with the Beijing AKEC Medical Co., Ltd. (Beijing, China), using an RP technique to design and produce a customized implant for this patient.

First, CT scans were obtained with 1.3 mm thick slices. The data were saved in digital imaging and communication in medicine (DICOM) format and processed using the Mimics 10.01 software (Materialise®, Leuven, Belgium) to build a precise three-dimensional (3D) model of the defect in the computer. Second, the 3D model was proposed to the UG 6.0 software (Siemens PLM Software®, Plano TX, USA). A virtual revision procedure was performed to estimate the proper size of the component, and we then previewed the alignment of the component. A computer-aided design (CAD) model of the device was designed precisely to fit the bone defect and the alignment of the knee component, as shown in Figure 1B. The CAD file was then saved and transferred toARCAM EBM A1 (ARCAM AB, Krokslätts, Sweden). The implant was fabricated by melting metal powder (Ti6Al4V) layer-by-layer through the Arcam electron beam melting (EBM) process. Finally, the patient-specific implant, with its unique shape and high porosity (Figure 1C), was sterilized and packed.

Revision surgery was carried out in our department with the approval of the ethics committee and informed consent from the patient. The cement was cautiously removed, and the device was implanted as planned preoperatively. Any gaps in the implant were filled with impacted morselized bone. A condylar constrained knee was then fixed in place (Figure 1D). Antibiotics and anticoagulant drugs were prescribed for 2 weeks until the incision healed. At the 6-month follow-up visit, the Hospital for Special Surgery (HSS) knee score had increased to 83, and the ROM was 0°–100°. There were no signs of loosening on follow-up radiography (Figure 1E).