Replacement of the patellofemoral and medial tibiofemoral joints has been performed since the 1980s. Bicompartmental replacement was modified. Two different designs were developed: one custom implant and one with multiple predetermined sizes. The surgical technique and instruments are unique and training is helpful. There are no clinical reports for the custom design as of yet. The standard implant has several reports in the literature with only fair to good results and has subsequently been withdrawn from the market. Bicompartmental arthroplasty remains a questionable area of knee surgery. At present, the two separate implant technique is the best choice.
Key points
- •
Bicompartmental arthroplasty is a bone-preserving procedure that also preserves both of the cruciate ligaments.
- •
The two separate implant technique allows the surgeon to concentrate on one area and then the other without compromising either arthroplasty.
- •
Presently, bicompartmental arthroplasty does not have acceptable long-term results. However, there may still be a place for the procedure with custom-made implants that are positioned more accurately either with navigation control or patient-specific cutting blocks.
Introduction
Partial knee arthroplasty was developed in the 1950s with such devices as the McKeever and MacIntosh implants. Unicondylar (UKA) and patellofemoral (PFA) prostheses were used in the late 1970s and publications showed acceptable results at midterm follow-up. Some surgeons combined the UKA and PFA when the pathology presented itself at the time of the surgical procedure. The results were once again acceptable at midterm follow-up and had the advantage of ligament preservation and improved proprioception. However, long-term follow-up showed a high revision rate. As the total knee arthroplasty (TKA) designs improved, there was less interest in partial knee arthroplasty until Repicci and coworkers offered a smaller incision for UKA. Limited incisions for knee arthroplasty became more popular and partial knee arthroplasty became more common. The bicompartmental replacements from the early 1970s had some recognized advantages over TKA with preservation of the cruciate ligaments, increased motion, and improved proprioception; however, the two separate implants removed a considerable amount of bone and the operative procedure was complex. A monoblock femoral component was designed to simplify the surgery and limit the amount of resected bone (Journey-Deuce; Smith and Nephew, Memphis, TN). The prosthesis removed less bone and spared all of the ligaments of the knee. A similar prosthesis is now available that is custom designed from a computed tomography of each individual knee (iDuo; Conformis, Burlington, MA).
The separate prostheses for the patellofemoral and the medial tibiofemoral joints have also been improved over the past few years and resect less bone with improved accuracy. The question is whether either of the approaches can produce clinical results that compare with standard UKA or TKA.
Introduction
Partial knee arthroplasty was developed in the 1950s with such devices as the McKeever and MacIntosh implants. Unicondylar (UKA) and patellofemoral (PFA) prostheses were used in the late 1970s and publications showed acceptable results at midterm follow-up. Some surgeons combined the UKA and PFA when the pathology presented itself at the time of the surgical procedure. The results were once again acceptable at midterm follow-up and had the advantage of ligament preservation and improved proprioception. However, long-term follow-up showed a high revision rate. As the total knee arthroplasty (TKA) designs improved, there was less interest in partial knee arthroplasty until Repicci and coworkers offered a smaller incision for UKA. Limited incisions for knee arthroplasty became more popular and partial knee arthroplasty became more common. The bicompartmental replacements from the early 1970s had some recognized advantages over TKA with preservation of the cruciate ligaments, increased motion, and improved proprioception; however, the two separate implants removed a considerable amount of bone and the operative procedure was complex. A monoblock femoral component was designed to simplify the surgery and limit the amount of resected bone (Journey-Deuce; Smith and Nephew, Memphis, TN). The prosthesis removed less bone and spared all of the ligaments of the knee. A similar prosthesis is now available that is custom designed from a computed tomography of each individual knee (iDuo; Conformis, Burlington, MA).
The separate prostheses for the patellofemoral and the medial tibiofemoral joints have also been improved over the past few years and resect less bone with improved accuracy. The question is whether either of the approaches can produce clinical results that compare with standard UKA or TKA.
Clinical reports
Bicompartmental Arthroplasty Using a Single Piece Femoral Component
The author performed 100 of the monoblock designed components and followed the first 40 patients for 5 years. Two operations were bilateral procedures. The patients were chosen for the operation based on the preoperative office interview, physical examination, and radiographic evaluation.
The patients were asked to indicate the location of their pain and the prevalence. If the pain was medial tibiofemoral with associated medial patellofemoral symptoms, the patient was considered a good candidate. The indications were very similar to those used for UKA but allowed more symptoms relating to the patellofemoral joint. Global knee pain that was equally distributed in all areas of the knee was a definite contraindication despite any of the physical examination and radiographic findings to the contrary.
The physical examination included medial tibiofemoral and patellofemoral tenderness. The clinical deformity did not exceed 10 degrees of varus or flexion contracture. When the varus deformity corrected to neutral with valgus stress, the knee was a more ideal one for the replacement. However, it was not absolutely necessary for the knee to correct completely. The cruciate ligaments were clinically intact. Some degree of anterior laxity was accepted but grade four instability was not included. Inflammatory arthritis and knees with previous ligament reconstructions or osteotomies were excluded.
The standing anteroposterior radiograph showed an anatomic varus deformity that was less than 10 degrees with minimal translocation of the tibia beneath the femur. Patellofemoral arthritic changes of any extent were acceptable. Mild lateral osteoarthritic changes were considered acceptable. If there were changes in the lateral compartment, there should be no significant symptoms of pain or tenderness on physical examination.
The operative procedure was performed using a curvilinear medial incision with an associated similar medial arthrotomy. No attempt was made to specifically limit the incision; however, the smallest possible exposure was chosen that permitted adequate joint visualization and implant positioning. Most of the procedures included a 1- to 2-cm incision into the quadriceps tendon as part of the medial arthrotomy.
The tibial plateau was addressed first and an extramedullary guide was used to cut the surface 2 to 4 mm below the lowest point ( Fig. 1 ). The space was then evaluated using a spacer block with the 8-mm insert ( Fig. 2 ). The gap at 90 degrees of flexion was compared with the gap in extension to plan the distal femoral resection. An intramedullary instrumentation system was used to make the femoral cuts. The anteroposterior axis was used for referencing proper external rotation. The first femoral guide sized the femur from anterior to posterior and set the anterior femoral resection depth and rotation ( Fig. 3 ). The distal femoral resection was performed by setting the depth on the medial side of the femur and referencing a point on the lateral side of the previously anterior cut surface where the edge of the prosthesis would meet the lateral aspect of the femoral cortex ( Fig. 4 ). After the distal medial resection was completed, the space in full extension was compared with the space in flexion and minor adjustments made before proceeding. The finishing block for the femur was set on the cut surface and adjusted with reference to the medial femoral condyle and the lateral femoral cortex ( Fig. 5 ). It was often difficult to match the femur with a femoral component size that would allow the medial femoral runner to sit exactly in the center of the tibial tray and also extend across the anterior surface to the lateral cortex.
