Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Unicompartmental Knee Arthroplasty
In theory, unicompartmental knee arthroplasty (UKA) is an attractive alternative to osteotomy and total knee arthroplasty (TKA) in selected osteoarthritic patients. Advantages of UKA over osteotomy include higher initial success, fewer early complications, greater longevity, better cosmetic limb alignment, easier conversion to TKA, and the potential to perform bilateral procedures on the same day. Later conversion of osteotomy to TKA is potentially complicated by many factors (see Chapter 9 ).
Advantages of UKA over TKA include the preservation of both cruciate ligaments, resulting in more normal knee kinematics and potential for a higher level of performance. Bone stock is preserved in the opposite and patellofemoral compartments, allowing easier conversion to TKA should this be necessary. Initial reports from my institution did not confirm that UKA conversion was necessarily an easy procedure. A later report, however, showed that if the UKA was done in a conservative fashion, conversion was easy and results were the same as those for primary TKA. Potential revision problems involved in osteotomy, TKA, and UKA conversion are listed in Table 16.1 . In UKA, the only potential revision problem should be medial tibial plateau deficiency treated with bone graft or metal wedge augmentation (see Chapter 11 ).
TABLE 16.1
Revision Problems Seen in Osteotomy, Total Knee Arthroplasty, and Unicompartmental Knee Arthroplasty Conversion
| Problems | Tibial Osteotomy | TKA | UKA |
|---|---|---|---|
| Unusable prior incision | + | − | − |
| Poorly accessible prior hardware | + | − | − |
| Joint line angle distortion | +/− | +/− | − |
| Malunion | +/− | − | − |
| Nonunion | +/− | − | − |
| Patella baja | +/− | +/− | − |
| Offset tibial shaft | +/− | − | − |
| Deficient femoral bone | − | + | − |
| Deficient patellar bone | − | + | − |
| Deficient lateral tibial bone | +/− | + | − |
| Deficient medial tibial bone | +/− | + | +/− |
Despite these arguments, UKA has been a controversial procedure since its introduction in the early 1970s. Initial reports were discouraging regarding UKA in medial compartment osteoarthritis. A few years later, R. Santore and I published more favorable results using the unicondylar prosthesis. We studied 100 knees with 2- to 6-year follow-up. Three repeat operations had been performed, for an incidence of revision of roughly 1% per year at this short follow-up. The average flexion was 114 degrees, significantly better than in any contemporary reports using bicompartmental arthroplasty. When the same series was followed at 5 to 9 years (mean 7 years), seven revisions had been performed for a revision rate continuing at 1% per year. The total condylar experience at that time for the same follow-up also showed a revision rate of 1% per year. Armed with this information, our enthusiasm for unicompartmental arthroplasty began to grow. We felt that when the knee was exposed for arthroplasty and the patient fulfilled the criteria for unicompartmental replacement, it was reasonable to perform this procedure on these patients. By the early 1980s, approximately 10% of my patients with osteoarthritis underwent unicompartmental replacement.
Reasons for failure of UKA included patient selection, prosthetic design, and surgical technique. We learned the lesson that patients with lateral compartment arthritis and a lax medial collateral ligament (MCL) could not be stabilized by a unicompartmental procedure ( Fig. 16.1 ). We also began to see failures in obese patients. These were due to either tibial or femoral loosening. We saw some femoral component loosening by subsidence into the subchondral bone ( Fig. 16.2 ). The force across the knee in level gait is approximately three times body weight. This force is ideally distributed evenly to the medial and lateral compartment over the surface area provided by each compartment. The pounds per square inch increase if more pounds and fewer square inches are covered by the prosthetic components. The relatively small size of early UKA components therefore made them vulnerable to loosening in heavy patients.
Unicompartmental knee arthroplasty failed to restore stability to this severe valgus knee with medial collateral ligament laxity.
Subsidence of the femoral component in a heavy patient.
To address this issue, P. Walker and I redesigned the unicondylar prosthesis to the Brigham unicompartmental knee in 1981 ( Fig. 16.3 ). The femoral component was made 5 mm wider to better cap the subchondral bone and resist subsidence. The tibial component was a nonmodular, metal-backed component with a composite thickness starting at 6 mm ( Fig. 16.4 ). The articulation was flat-on-flat to diminish stress on the polyethylene and increase surface contact. We used this prosthesis exclusively for the next 8 or 9 years. The specifics of its design and surgical technique taught us more important lessons about unicompartmental arthroplasty.
A, The relatively narrow runner of the unicondylar femoral component made it vulnerable to subsidence. B, The Brigham knee was 5 mm wider to better cap the subchondral bone and resist subsidence.
The 6-mm metal-backed tibial component of the Brigham knee had only 2 mm of polyethylene in parts of its articulating surface.
Because the articulation was flat-on-flat, it became apparent that imprecise surgical technique was unforgiving. If the component articulating surfaces were not parallel during weight-bearing, edge-loading would occur, leading to accelerated wear of the polyethylene ( Fig. 16.5 ). We also learned that mediolateral and rotational congruity between the components must be assessed in extension rather than flexion. Through the classic median parapatellar approach with patellar eversion, the tibia is artificially externally rotated in flexion owing to the force of the everted quadriceps. Normally, of course, the tibia tends to internally rotate in flexion. If component congruency is assessed in flexion, it will give an inaccurate view of component congruency when the quadriceps is relocated and the patient bears weight in extension.
Poor surgical technique can lead to edge-loading of a flat-on-flat articulation.
Modes of failure during this era were most frequently secondary to wear, loosening, and degeneration of the opposite compartment. The flat-on-flat articulation highlighted the problem of edge-loading as a cause for wear. The specific design of the Brigham nonmodular metal-backed tibial component also showed the importance of having an adequate thickness of polyethylene in the articulating area. The “6-mm” Brigham tibial component consisted of a 2-mm-thick titanium metal backing with 4 mm of polyethylene. The coupling of the polyethylene to the tray was designed in a manner that provided this full 4 mm in the middle 60% of the articulation, but only 2 mm of polyethylene in the anterior and posterior 20% of the components. Although most of these 6-mm components survived for at least 6 or 7 years after implantation, a significant number began to wear through by the end of the first decade ( Fig. 16.6 ). This complication occurred when the articulating pattern of the arthroplasty replicated the preoperative articulating pattern of the arthritic knee. As described by White and colleagues, this is most frequently an anteromedial wear pattern. This pattern would bring the femoral component onto an area of the tibial component with only 2 mm of polyethylene thickness and often at the right-angle metallic junction between the 2- and 4-mm depths.
Wear-through of a composite 6-mm Brigham unicondylar femoral component.
We now realize that all fixed-bearing UKA designs are subject to this wear pattern. Attempts to address this issue by providing fixed bearings with more conforming articulations failed because the increased constraint transmitted too much force to the fixation interface. Failures were seen on both the femoral and tibial sides, and the incidence of tibial radiolucent lines with more conforming articulations was increased. If a conforming insert is used to address top-side wear, the system must be a mobile-bearing design to eliminate the adverse effects of constraint on the fixation interface.
Degeneration of the opposite compartment is a second mechanism of failure in UKA designs implanted today. This complication is usually late, occurring after the first decade, unless the UKA was overcorrected ( Fig. 16.7 ). The ideal corrected alignment after medial unicompartmental replacement is probably somewhere between 2 and 5 degrees of valgus. Varus malalignment of the limb secondary to medial unicompartmental arthritis is best corrected by removal of peripheral osteophytes from the femur and tibia that tent-up the MCL and medial capsule ( Fig. 16.8 ). Adequate passive correction of the deformity is usually achievable after osteophyte débridement (see Chapter 17 ).
Secondary degeneration of the lateral compartment 24 years after the initial surgery.
A, Peripheral osteophytes (arrow) tent-up the medial collateral ligament and medial capsule. B, Removal of peripheral osteophytes allows passive correction of the deformity.
Classic Selection Criteria
In 1989, S. Kozinn and I reported on the ideal candidate for unicompartmental arthroplasty. Criteria included an elderly patient, noninflammatory osteoarthritis, a mechanical axis deformity of less than 10 degrees of varus or 5 degrees of valgus, an intact anterior cruciate ligament (ACL) without mediolateral subluxation, a flexion contracture less than 15 degrees, a body weight less than 80 to 90 kg, and patellofemoral changes not greater than grade II or III.
Several years later, Stern and colleagues studied their patient population with osteoarthritis and found that 6% of patients fulfilled all these selection criteria. Coincidentally, approximately 6% of the knee arthroplasty market received unicompartmental replacement. This percentage remained relatively stable until the advent of minimally invasive surgery at the beginning of the twenty-first century. Enthusiasm for the procedure began to grow quickly, perhaps to a point at which some perspective was lost regarding indications for the procedure.
My own perspective changed somewhat when we examined our second decade survivorship in unicompartmental devices implanted in the mid-1970s. Although the reoperation rate progressed for the first decade at roughly 1% per year of follow-up, the need for revision surgery appeared to grow faster in the second decade than that seen for bicompartmental replacement. This made me question the advisability of performing UKA in patients with approximately 15 to 20 years of life expectancy. These patients would have a greater statistical chance of requiring no further surgery throughout the remainder of their lifetime with TKA rather than UKA. Using this rationale, I began to stratify my UKA candidates into two categories. The first would be a middle-aged patient, especially female, and the second would be an octogenarian. I began to think about the procedure as the first arthroplasty in middle-aged patients that would buy them 10 or more years of longevity with an easy conversion to TKA when that was inevitably necessary in later life. They would benefit from a high initial success, few early complications relative to high tibial osteotomy (HTO), an acceptable cosmetic appearance relative to HTO, preservation of both cruciate ligaments relative to TKA, and an easier revision in comparison with both HTO and TKA.
Advantages for the octogenarian would include faster recovery, less blood loss, less medical morbidity, and a less expensive procedure to be borne by the health care system. Given the life expectancy and activity level of octogenarians, the UKA would be unlikely to require revision in their lifetime.
We have reported a small experience with UKA in patients under the age of 60 years. Only 28 knees were included in the study, with a 2- to 6-year follow-up. Ninety percent had good-to-excellent results, but two revisions had been done in heavy, active males as a result of femoral component loosening. This series was generated at a time when we were using conforming femorotibial geometry, which we now know can promote too much stress at the fixation interface. This prosthetic design issue diminishes the significance of this series.
You're Reading a Preview
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here