The press-fit condylar modular total knee prosthesis (PFC; Johnson and Johnson, Raynham, Massachusetts) was designed to address deficiencies that had been noted with earlier designs of prosthetic knees. Although the total condylar prosthesis has demonstrated excellent durability, function, and relief of pain at long-term follow-up evaluations, the mean range of motion of the knee has been only between 90 and 95 degrees11,13,14,19,20,22. The early designs of posterior stabilized prostheses improved the range of motion, but issues of patellofemoral complications became more apparent10. In addition, large condylar erosions or osseous loss that was not obvious radiographically required the surgeon to switch to a different prosthesis in order to augment fixation with a stem or to address the loss of bone.
The condylar geometry of the total condylar prosthesis was adapted by adjusting the length and the radius of the curvature of the posterior condyles. In addition, a tibial post-and-cam mechanism that substituted for the posterior cruciate ligament was added to provide greater clearance for the posterior aspect of the tibial component with the knee in full flexion and to optimize the contact area for improved contact stresses on the polyethylene. The trochlear groove was deepened, and its congruency with the patellar component was optimized to allow a broad surface of contact, which extended the range of motion to 110 degrees, and thereby reduced the compressive and shear forces of the patellofemoral articulation. Modular components, such as optional stems, wedges, and augments as well as more constrained polyethylene inserts, were included in the design to allow greater intraoperative versatility.
The posterior stabilized version of the press-fit condylar modular knee has been used by the senior one of us (C. S. R.) since 1988 and was released for general use in 1989. The purpose of the present study was to assess whether the design changes reflected in the press-fit condylar modular knee system have improved the functional results of total knee replacement compared with those in historical controls.
*One or more of the authors has received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
†Center for Total Joint Replacement, Lenox Hill Hospital, William Black Hall, 130 East 77th Street, 11th Floor, New York, N.Y. 10021.
‡1406 N.E. Oakridge Drive, Clairemont, Oklahoma 74017.
Between February 1988 and February 1990, 118 patients who had had 150 consecutive primary total knee arthroplasties with cement performed by the senior one of us were enrolled in a prospective study. The underlying diagnosis was osteoarthrosis in ninety-eight patients, rheumatoid arthritis in sixteen, avascular necrosis in three, and psoriatic arthritis in one. The mean age at the time of the index procedure was seventy years (range, twenty-nine to eighty-five years), and the mean weight was 73.9 kilograms (range, 45.4 to 109.1 kilograms). Sixteen patients (thirty-two knees) had the procedure bilaterally, with the two procedures performed simultaneously in fourteen (twenty-eight knees). A lateral retinacular release was performed in forty-four knees (29 per cent; thirty-five patients) to allow for proper tracking of the patella.
Of the 118 patients who had a total knee replacement, ninety-six patients (125 knees) were followed for an adequate interval (mean, 4.8 years; range, 3.8 to 6.2 years). Thirteen patients (fifteen knees) died, and nine patients (ten knees) were lost to follow-up. Of the ninety-six patients, sixty-six (69 per cent) completed the questionnaire. The other thirty patients were asked the same questions by telephone interview.
Clinical Results
Preoperatively, the mean functional score was 40 points (range, 0 to 70 points) and the mean knee score was 44 points (range, 0 to 60 points). At the time of the most recent follow-up, the mean functional score was 78 points (range, 0 to 100 points) and the mean knee score was 93 points (range, 57 to 100 points). The clinical result was excellent for 103 knees, good for thirteen, fair for three, and poor for six. At the most recent follow-up examination, the mean functional score was 81 points for the patients who had osteoarthrosis and 62 points for those who had rheumatoid arthritis (p < 0.05). The mean knee score was 94 points for the patients who had osteoarthrosis and 90 points for those who had rheumatoid arthritis. Twenty-four patients (25 per cent) were classified as a Charnley functional-class C (multiple joints involved by arthritis or a medical infirmity).
Range of Motion
The mean preoperative range of flexion of the knee was 107 degrees (range, 60 to 135 degrees). The mean postoperative range of flexion was 111 degrees (range, 75 to 135 degrees). The mean range of flexion of the 105 knees that were affected by osteoarthrosis was 112 degrees (range, 60 to 135 degrees), and that of the eighteen knees that were affected by rheumatoid arthritis was 104 degrees (range, 85 to 125 degrees). Of the knees that were affected by osteoarthrosis, two had less than 80 degrees of flexion; four, between 80 and 89 degrees; ten, between 90 and 99 degrees; fifteen, between 100 and 109 degrees; twenty-six, between 110 and 119 degrees; forty-six, between 120 and 129 degrees; and two, more than 130 degrees.
Preoperatively, thirty-one patients (forty-one knees; 33 per cent) had a flexion contracture (mean, 17 degrees; range, 5 to 40 degrees). The contracture measured 5 degrees in ten knees, 10 degrees in ten knees, 15 degrees in four knees, 20 degrees in six knees, 25 degrees in three knees, 30 degrees in four knees, and 40 degrees in four knees. A biplanar deformity, defined as a flexion contracture of more than 15 degrees combined with varus angulation of more than 10 degrees or valgus angulation of more than 15 degrees, was present in twenty knees. At the most recent follow-up examination, seven knees (five patients) had a flexion contracture of 5 degrees and one knee had a flexion contracture of 10 degrees.
Pain
Seventy-three patients (ninety-nine knees; 79 per cent) had no pain in the knee at the most recent follow-up evaluation. Twenty-three patients (twenty-six knees; 21 per cent) reported discomfort in the knee. Of this group, fourteen patients (sixteen knees) reported pain that was mild and infrequent and nine patients (ten knees) reported moderate pain. Seven of the nine needed medication to control the symptoms, and two of the seven took medication with a narcotic derivative. In eight patients (nine knees; 7 per cent), the pain was in the anterior aspect of the knee and therefore was presumed to be referable to the patellofemoral articulation.
Ability to Walk
Fifty-one patients (sixty-seven knees; 54 per cent) were able to walk an essentially unlimited distance without pain, fifteen were able to walk five to ten blocks, twenty-four were able to walk five blocks, four were able to walk in the house only, and two were unable to walk. Five of the six patients who could not walk at least six blocks had polyarticular rheumatoid arthritis that limited the ability to walk, and the other patient had severe spinal stenosis with back pain and neurogenic claudication that limited the ability to walk.
Before the operation, none of the patients were able to ascend or descend stairs in a reciprocal fashion. After the operation, sixty-one patients (64 per cent) were able to do so. Sixty-six patients (ninety-two knees) did not need support for walking. Twenty-two patients (twenty-five knees) used a cane, three patients used two crutches, and five patients used a walker. Fifteen patients (nineteen knees; 15 per cent) had a limp: seven reported that they limped all of the time, three reported that they limped intermittently, and five reported an infrequent limp.
Satisfaction and Activity
Sixty-two patients were fully satisfied with the outcome of the operation, twenty-one patients were very satisfied, nine patients were satisfied, and four patients were dissatisfied. Of the four patients who were dissatisfied, one had had a revision operation because of tibiofemoral instability, two had constant moderate pain in the knee, and one had not regained an appreciable ability to walk after the operation because of ankylosing polyarthritis. The mean score for satisfaction was 8.8 of a possible 10 points for the patients who had osteoarthrosis and 8.6 points for the patients who had rheumatoid arthritis.
Survivorship Analysis
Survivorship analysis was performed with a revision operation for any reason or a recommendation for revision as the end point. The projected clinical and radiographic rate of survival of the knee prosthesis was 97 per cent at six years, at which point thirty-five knees remained at risk. The standard error of the mean for this calculation was 1.6 per cent.
Revision Operations
Three revision total knee replacements were performed. Two of the revisions were done because of infection of the involved joint. Both patients were managed with a standard two-stage revision consisting of débridement and subsequent reimplantation after a six-week course of intravenous antibiotic therapy. The third revision was performed because of recurrent posterior dislocation of the prosthesis. The instability was treated successfully with revision of the tibial component four months after the initial procedure.
Patellofemoral Articulation
Two patients (two knees) had a patellar tilt (a medially opening angle) of 20 degrees that was seen on skyline radiographs, but both patients were asymptomatic with reference to the patellofemoral joint. One patient had a vertical patellar fracture that resulted in a complete radiolucent line around the cement-bone interface of the patella 2.5 years after the operation; the fracture was treated non-operatively. At the most recent follow-up examination, the patient had maintained full strength of the quadriceps muscles without extensor lag or pain but had recurrent effusions in the joint related to activity. As noted previously, eight patients (nine knees) reported pain in the anterior aspect of the knee, according to the questionnaire. On examination, two of the nine knees had patellofemoral crepitation. In addition, eight knees (eight patients) were noted to have patellofemoral crepitation through the arc of motion. However, because the crepitation was not associated with pain and did not limit the ability of the patient to climb stairs, the knees were not considered to be symptomatic. Therefore, the over-all rate of symptomatic patellofemoral articulations (pain or effusion, or both) was 8 per cent (ten knees). None of the knees had loosening of the patellar component, clunk syndrome, or patellofemoral instability.
Instability
In addition to the patient who had a revision operation because of instability, two patients sustained posterior dislocations of the prosthesis: one was diagnosed clinically at two months and one, at one year after the operation. Both of these patients had a closed reduction. In all three patients, the posterior dislocations occurred with the knee in a flexed position. One patient was managed successfully with a hinged knee brace for four months and had no recurrence during forty-eight months of follow-up. The other patient who had posterior instability was advised to have a revision operation to correct the problem, and the knee was recorded as a failure in the survivorship analysis. The patient subsequently died of cancer. All three index operations had been performed within one week of each other in 1988 with use of the same early design of the tibial cam, which was eight millimeters high. Subsequent to those procedures, the height of the tibial cam was elevated to 14.3 millimeters, and no other instances of instability occurred.
Radiographic Results
Preoperative radiographs revealed that seventy-three knees had a valgus angulation and seventy-seven knees had a varus angulation. The valgus angulation was 21 to 30 degrees in nine knees, 11 to 20 degrees in twenty-six, and 1 to 10 degrees in thirty-eight. The varus angulation was 1 to 10 degrees in sixty-two knees and 11 to 20 degrees in fifteen.
Ninety-nine knees (seventy-six patients) had complete radiographic follow-up. The position of the femoral component in the anteroposterior plane was a mean of 96 degrees (range, 91 to 103 degrees), or 6 degrees of valgus inclination. The position of the tibial component in this plane was a mean of 89 degrees (range, 82 to 95 degrees), or 1 degree of varus inclination. In the sagittal plane, the femoral component was positioned in a mean of 5 degrees (range, 0 to 16 degrees) of flexion and the angle of the tibial component was a mean of 85 degrees (range, 78 to 97 degrees), or 5 degrees of posterior tilt. Sixty knees had no evidence of radiolucent lines around any of the components. Therefore, the prevalence of radiolucent lines around at least one component was 39 per cent (thirty-nine knees).
The anteroposterior radiographs of the knees, made at the latest follow-up examination, demonstrated seventeen tibial radiolucent lines in zone 1, three in zone 2, six in zone 3, and seven in zone 4. One patient had tibial radiolucent lines in zones 1 and 2. Lateral radiographs showed four tibial radiolucent lines in zone 1 and one in zone 2. On the femoral side, eight radiolucent lines were found in zone 2 and two were found in zone 4. Evaluation of the skyline patellar radiographs revealed two radiolucent lines in zone 1 and nine in zone 2. All of the radiolucent lines were grade 2 with the exception of the complete line around the cement-bone interface associated with the patellar fracture in one knee and the one grade-3 radiolucent line in zone 1 on the lateral radiograph of the tibial component. None of the grade-2 radiolucent lines evident on the immediate postoperative radiograph had progressed. None of the knees had a radiolucent line in more than two contiguous zones, and no radiolucent lines were observed around the tibial keel.
The patellar tilt, which was measured on the radiographs of ninety knees, consisted of a mean medial opening of 3 degrees (range, 11 degrees of lateral opening to 20 degrees of medial opening).