The etiology of congenital dislocation of the patella remains unknown. In adulthood, osteoarthritis of the knee may develop, and it usually affects the lateral compartment. Valgus malalignment of the lower extremity and a laterally dislocated patella are the typical physical findings.
Various techniques to reconstruct a dislocated extensor mechanism have been described1-4. In the presence of severe osteoarthritis, total knee arthroplasty is a surgical option, which may be combined with realignment of the extensor mechanism. We report a new technique of total knee arthroplasty with proximal realignment of the extensor mechanism with use of a V-W quadricepsplasty (quadriceps turndown and medial advancement) in two patients. The patients were informed that data concerning the cases would be submitted for publication, and they consented.
Case 1. A forty-nine-year-old man was evaluated for bilateral knee pain and reduced walking distance secondary to knee arthritis. At the age of six years, he was diagnosed with bilateral congenital patellar dislocation and was managed nonoperatively. On examination, standing lower-extremity alignment measured 15° of valgus bilaterally. The passive range of motion of both knees measured 0° to 135°, with painful crepitus noted in the lateral compartments. The patient had a 30° extensor lag. Lateral dislocation of the patella was present bilaterally. Radiographic examination revealed bilateral valgus deformity with complete loss of the lateral compartment joint space. The patellae were laterally dislocated.
A lateral unicompartmental hemiarthroplasty of the right knee was performed when the patient was fifty years of age, and the dislocated patella was not addressed. Because of recurrent disabling pain, the hemiarthroplasty was revised to a posterior stabilized total knee arthroplasty with cement four years later, when the patient was fifty-four years of age. The extensor mechanism was reconstructed with use of a V-W quadricepsplasty technique, as described below. Ten years later, when the patient was sixty-four years of age, he had the same procedure performed on the left knee.
At twelve years after it was operated on, the right knee was pain-free and had a passive range of motion of 0° to 125°, with a 15° extensor lag. The patella tracked centrally. At two years after it was operated on, the left knee was also pain-free, with a passive range of motion of 0° to 125° and a 35° extensor lag. The patella tracked centrally.
Case 2. A fifty-three-year-old woman with a history of pseudoachondroplasia presented with bilateral knee pain secondary to degenerative arthritis. As a child, she had undergone several soft-tissue procedures in an attempt to reduce the congenitally dislocated patellae, but these operations had failed. On examination, the standing lower-limb alignment was neutral with lateral dislocation of both patellae. The passive range of motion of each knee was 15° to 110°, with painful crepitus localized to the lateral compartments. The patient had an extensor lag of 45° bilaterally. Radiographic examination revealed bicompartmental arthritis. The patellae were dislocated laterally with dysplastic trochlear grooves (Figs. 1-A and 1-B).
At the age of fifty-four, the patient underwent a right total knee arthroplasty with use of cemented posterior stabilized components. The patella was resurfaced with a cemented all-polyethylene component. The extensor mechanism was reconstructed and realigned with use of a V-W quadricepsplasty, and this resulted in central tracking of the patella. Three months later, the same procedure was performed on the left knee.
Two years following the second knee replacement, the patient was walking with one cane and reported that she was pain-free. The right knee had a passive range of motion of 15° to 105° and active extension of 15° (no extensor lag). The left knee had a passive range of motion of 10° to 115° and active extension of 15° (a 5° extensor lag). Both patellae tracked centrally (Figs. 1-C and 1-D).
A midline longitudinal skin incision was made under tourniquet control. As a result of lateral dislocation of the extensor mechanism, the vastus medialis obliquus muscle was encountered in an anterior position, covering the femoral trochlea. The patella and the quadriceps and patellar tendons were identified in the lateral gutter of the knee, and a medial parapatellar arthrotomy was performed (Fig. 2-A). An inverted "V" incision was then initiated from the top of the "medial" arthrotomy along the insertion of the vastus lateralis, continuing distally approximately 2 cm posterior to the lateral margin of the patella. The lateral superior geniculate vessels were cauterized in order to gain adequate mobility of the patellar pedicle to reestablish central patellar tracking.
A total knee arthroplasty was then performed with posterior stabilized components and a cemented all-polyethylene patellar button. The distal part of the femur was osteotomized in 3° of valgus in order to facilitate patellar tracking. Femoral component rotation was based on establishing flexion gap symmetry with the tibial cut being perpendicular to the mechanical axis. All four knees in this series had relative lateral laxity in flexion, resulting in as much as 8° of external rotation of the femoral component when referenced in relationship to the posterior condyles. The senior author (R.D.S.) believes that increasing external rotation of the femoral component facilitates patellar tracking. The rotational alignment of the tibial component was based on the middle of the tibial tubercle to optimize the Q angle.
An extensive medial advancement was required to establish and maintain central patellar tracking (Fig. 2-B). The medial aspect of the capsule and the vastus medialis obliquus were brought over the top of the inverted V-shaped pedicle, which includes the quadriceps tendon, the patella, and the patellar tendon. The lateral side of the arthrotomy was not closed. The medial aspect of the capsule was advanced at least 2 cm. The closure was performed with two rows of sutures. The first row consisted of vertical mattress sutures beginning in the medial aspect of the capsule, 1 cm medial to its edge. The suture was then passed through the retinaculum of the patellar pedicle, just lateral to the patellar edge, and then back to a few millimeters from the initial entry point in the medial aspect of the capsule (Fig. 2-B). In between these sutures, figure-of-eight sutures were placed at the edge of the point of advancement of the medial aspect of the capsule to add security to the repair and flatten the capsule's leading edge.
After the repair was complete, the hip was flexed 90° and passive knee flexion against gravity was recorded. This was done to assess the excursion of the quadriceps and the integrity of the repair. Postoperative flexion was restricted to this amount for three weeks to protect the repair.
There is a paucity of information in the literature on the surgical management of osteoarthritis of the knee associated with congenital patellar dislocation. Marmor5 and Pradhan et al.6 reported total knee arthroplasty, without an attempt to relocate the extensor mechanism, as a surgical option for the treatment of knee arthritis in association with congenital dislocation of the patella. They theorized that, if the patient compensated well for his or her disability, the surgeon could avoid addressing the difficult issues associated with relocation of the extensor mechanism. These reports, however, had short follow-up, and the effects of neglecting patellar stability on component survival are unknown. Bullek et al.7 and Bergquist et al.8 successfully relocated and stabilized a chronically dislocated patella during total knee arthroplasty by employing a modified lateral release and medial vastus medialis obliquus imbrication. In the two cases described in our report, these maneuvers alone were not enough to relocate the patella in the trochlear groove. An additional proximal or distal realignment was required. A distal realignment by means of a tibial tubercle transfer alone might not have been successful and is associated with a high risk of nonunion. Proximal realignment eliminates this risk and allows early knee motion.
One concern about our technique is the potential for loss of knee flexion because the extensor mechanism is anatomically repositioned anterior to the plane of knee motion. In our two patients, the final flexion arc in the four knees was essentially unchanged from the preoperative arc. Another concern is the potential for osteonecrosis of the patella due to the extensive release of the extensor mechanism. We have yet to see evidence of this in these two patients, whose knees have been followed for two to twelve years. A third concern is whether active extension can actually be improved with this technique. Three of the four knees showed marked improvement in the amount of active extension, as compared with the preoperative status, whereas one knee showed no change. We must stress, therefore, that although the technique centralized patellar tracking in all four knees, it did not completely resolve a preoperative extensor lag.