Results of Unicompartmental Knee Arthroplasty at a Minimum of Ten Years of Follow-up

Berger, Richard A. MD; Meneghini, R. Michael MD; Jacobs, Joshua J. MD; Sheinkop, Mitchell B. MD; Della Valle, Craig J. MD; Rosenberg, Aaron G. MD; Galante, Jorge O. MD

Journal of Bone & Joint Surgery - American Volume: May 2005 - Volume 87 - Issue 5 - p 999–1006
doi: 10.2106/JBJS.C.00568
Scientific Articles

Background: There is a renewed interest in unicompartmental knee arthroplasty. The present report describes the minimum ten-year results associated with a unicompartmental knee arthroplasty design that is in current use.

Methods: Sixty-two consecutive unicompartmental knee arthroplasties that were performed with cemented modular Miller-Galante implants in fifty-one patients were studied prospectively both clinically and radiographically. All patients had isolated unicompartmental disease without patellofemoral symptoms. No patient was lost to follow-up. Thirteen patients (thirteen knees) died after less than ten years of follow-up, leaving thirty-eight patients (forty-nine knees) with a minimum of ten years of follow-up. The average duration of follow-up was twelve years.

Results: The mean Hospital for Special Surgery knee score improved from 55 points preoperatively to 92 points at the time of the final follow-up. Thirty-nine knees (80%) had an excellent result, six (12%) had a good result, and four (8%) had a fair result. At the time of the final follow-up, thirty-nine knees (80%) had flexion to at least 120°. Two patients (two knees) with well-fixed components underwent revision to total knee arthroplasty, at seven and eleven years, because of progression of patellofemoral arthritis. At the time of the final follow-up, no component was loose radiographically and there was no evidence of periprosthetic osteolysis. Radiographic evidence of progressive loss of joint space was observed in the opposite compartment of nine knees (18%) and in the patellofemoral space of seven knees (14%). Kaplan-Meier analysis revealed a survival rate of 98.0% ± 2.0% at ten years and of 95.7% ± 4.3% at thirteen years, with revision or radiographic loosening as the end point. The survival rate was 100% at thirteen years with aseptic loosening as the end point.

Conclusions: After a minimum duration of follow-up of ten years, this cemented modular unicompartmental knee design was associated with excellent clinical and radiographic results. Although the ten-year survival rate was excellent, radiographic signs of progression of osteoarthritis in the other compartments continued at a slow rate. With appropriate indications and technique, this unicompartmental knee design can yield excellent results into the beginning of the second decade of use.

Level of Evidence: Therapeutic Level IV. See Instructions to Authors for a complete description of levels of evidence.

1 Department of Orthopaedic Surgery, Rush Medical College, Rush-Presbyterian-St. Luke's Medical Center, 1725 West Harrison Street, Suite 1063, Chicago, IL 60612. E-mail address for R.A. Berger: r.a.berger@sbcglobal.net

Article Outline

Early reports on the results of unicompartmental knee arthroplasty were discouraging1,2, but more recent studies have demonstrated ten-year survival rates of >90% in association with improvements in surgical technique, strict preoperative selection criteria, and improved instrumentation3-8.

Unicompartmental knee arthroplasty has certain advantages over total knee arthroplasty in the properly selected patient, including shorter hospital stays, fewer serious complications, improved walking ability, and lower cost9,10. In one study, patients who received a unicondylar replacement in one knee and a total knee replacement in the other knee during the same period of hospitalization preferred the knee with the unicondylar replacement and had a better range of motion in that knee11. Additionally, patients who have been managed with unicompartmental arthroplasty have a more normal gait, better quadriceps function, and better knee flexion than those who have been managed with total knee arthroplasty12,13. When a failed unicompartmental knee arthroplasty is converted to a total knee arthroplasty, the results compare favorably with those of primary total knee arthroplasty14,15.

In 1996 and 1997, only 2500 unicompartmental knee arthroplasties were performed in the United States, comprising approximately 1% of all knee arthroplasties performed16. In contrast, in 2000 and 2001, 33,900 unicompartmental knee arthroplasties were performed, comprising 6% of all knee arthroplasties performed17.

The present report describes the ten to thirteen-year results associated with a unicompartmental knee system that is in current use.

Back to Top | Article Outline

Materials and Methods

Between June 1987 and June 1993, the senior authors (J.O.G., M.B.S., and A.G.R.) performed 517 knee arthroplasties, of which sixty-two (12%) were unicompartmental knee arthroplasties. All procedures were performed through a standard medial parapatellar approach that included patellar eversion.

Patients were deemed to be candidates for unicompartmental knee arthroplasty rather than total knee replacement on the basis of the criteria of Kozinn and Scott18: a diagnosis of unicompartmental osteoarthritis or osteonecrosis, radiographic evidence of preservation of the opposite compartment, and only mild radiographic signs of deterioration of the patellofemoral joint. The inclusion criteria were a range of motion of at least 90° with a flexion contracture of <15°, minimal pain at rest, a relatively sedentary lifestyle, a weight of <275 lb (124.7 kg), and an age of more than fifty years. The exclusion criteria were inflammatory arthritis, hemochromatosis, chondrocalcinosis, hemophilia, patellofemoral joint symptoms, a positive patellar grind test, and symptomatic knee instability.

The patients included thirty-four women and seventeen men with an average age of sixty-eight years (range, fifty-one to eighty-four years). The average weight was 176 lb (79.8 kg) (range, 115 to 270 lb [52.2 to 122.5 kg]). Eleven patients weighed 200 lb (90.7 kg) or more. The diagnosis was osteoarthritis in fifty-three knees (85%) and osteonecrosis in nine knees (15%). Fifty-nine arthroplasties (95%) involved the medial compartment, and three (5%) involved the lateral compartment. Equal numbers of right and left knee arthroplasties were performed. Fifty knees had undergone previous arthroscopy, and two had had a previous proximal tibial osteotomy.

Thirty-seven patients underwent a single unicompartmental knee arthroplasty, ten underwent a sequential bilateral unicompartmental knee arthroplasty, and one had a staged bilateral procedure. The three remaining patients had sequential unicompartmental knee arthroplasty and contralateral tricompartmental knee arthroplasty.

No patient was lost to follow-up. Thirteen patients (thirteen knees) died after less than ten years of follow-up as a result of causes that were unrelated to the arthroplasty. Thus, thirty-eight patients (forty-nine knees) had a minimum of ten years of follow-up. The preoperative diagnosis was osteoarthritis for forty-six of these forty-nine knees and osteonecrosis for three. The average duration of follow-up for these forty-nine knees was twelve years (range, ten to thirteen years).

The Miller-Galante unicompartmental knee system (Zimmer, Warsaw, Indiana) was used in all patients (Fig. 1). The system consisted of a cobalt-chromium alloy femoral component, a titanium alloy tibial tray, and an ultra-high molecular weight polyethylene insert. The femoral components were available in five sizes, and the polyethylene inserts were available in four thicknesses. The thinnest tibial component that was used had a total thickness of 8 mm, of which 5.7 mm comprised the polyethylene liner and 2.3 mm comprised the tibial baseplate. The system was designed to be used with cement fixation. To provide enhanced fixation, the fixation surface of the femoral component had two pegs and the fixation surface of the tibial component had two pegs and a fin. In addition, the fixation surface of the tibial component was precoated with polymethylmethacrylate. The shape of the tibial articulating surface was essentially flat to allow for unconstrained motion of the femur on the tibia. The surgical instrumentation included an extramedullary cutting jig for the femur and an extramedullary cutting attachment for the tibial component.

The intramedullary femoral guide was set to remove 6 mm of bone both distally and posteriorly. The femoral component replaced the resected bone millimeter for millimeter. The femoral component was aligned in the coronal plane so that its long axis was perpendicular to the tibial component and therefore parallel to the long axis of the tibia (Fig. 2). The component was sized to be congruent with the curvature of the anterior part of the femur (Fig. 3).

The tibia was cut perpendicular to its long axis. In varus knees, the medial collateral ligament was released from the tibial plateau approximately 1 cm from the joint line from the patellar tendon to the posteromedial corner of the tibia. This 1-cm release of the medial collateral ligament was not done to correct deformity but instead was done for exposure. An appropriate polyethylene liner thickness allowed 2 mm of joint laxity in both full extension and flexion. This distance was tested with use of a calibrated spacer.

Chondral damage of the articular surfaces of the patellofemoral and opposite compartments was graded at the time of surgery according to the method of Outerbridge19. Patients with Outerbridge Grade-3 or 4 lesions were considered to be unsuitable for unicompartmental knee arthroplasty and were managed with tricompartmental knee replacement. (This was the case for two patients who were not included in this series.) At the time of surgery, Outerbridge Grade-1 or 2 changes were seen in the patellofemoral compartment in forty-eight knees (77%) and in the opposite compartment in twenty knees (32%).

All patients were evaluated prospectively. Postoperative knee function was evaluated by independent observers (clinical nurses and fellows) with use of the Hospital for Special Surgery knee score before surgery and at yearly intervals after surgery20. Radiographic analysis included measurement of the mechanical axis, measurement of the femorotibial axis, and assessment of the degree of correction. On standing anteroposterior radiographs, we measured the varus or valgus tilt of the prosthesis with respect to the longitudinal tibial axis. On lateral radiographs, we measured the posterior slope of the tibial component with respect to the longitudinal tibial axis and the flexion or extension of the femoral component with respect to the longitudinal femoral axis. The cement interfaces were evaluated for the presence and extent of radiolucent lines in each of ten zones3. A radiolucent line was considered to be progressive if it increased in size or if it progressed from one zone to an adjacent zone over time. Sequential radiographs were reviewed for evidence of component subsidence, or change in position. Definite loosening was defined as a change in position (subsidence) of >2 mm or an angular change of >3° relative to the surrounding bone as seen on sequential radiographs, with use of the early radiographs as a baseline3.

In addition to evaluating the components, we evaluated arthritic progression in the opposite compartment and the patellofemoral joint on standing radiographs3. Radiographic changes were defined as Grade 1 (evidence of radiographic changes such as osteophytes, but with no measurable loss of joint space), Grade 2 (≤25% loss of joint space), Grade 3 (≤50% loss of joint space), or Grade 4 (>50% loss of joint space).

Kaplan-Meier survivorship analysis of all sixty-two knees was performed with revision for any reason and with revision or radiographic signs of loosening as the end points21.

Back to Top | Article Outline

Results

Clinical Results

The average Hospital for Special Surgery knee score improved from 55 points (range, 30 to 79 points) preoperatively to 92 points (range, 60 to 100 points) at the time of the final follow-up. Thirty-nine knees (80%) had an excellent result (85 to 100 points), six (12%) had a good result (70 to 84 points), and four (8%) had a fair result (60 to 69 points). Two of the four patients who had a fair result (with scores of 63 and 60 points) underwent conversion to total knee replacement. The other two patients who had a fair result (with scores of 65 and 67 points) had limited walking ability secondary to severe cardiopulmonary disease at the time of the latest follow-up and subsequently died of causes that were unrelated to the arthroplasty. All living patients who had the unicompartmental replacement in place at the time of the final follow-up had an excellent result (thirty-nine knees, 87%) or a good result (six knees, 13%).

Patients reported no pain in twenty-nine knees (59%), slight pain in seventeen knees (35%), moderate pain in two knees (4%), and severe pain prior to revision in one knee (2%). Twenty-nine patients (76%) did not limp, six patients (16%) had a slight limp, two patients (5%) had a moderate limp, and one patient was unable to walk. At the time of the latest follow-up, thirty patients (79%) used no assistive devices, six patients (16%) used a cane for walking long distances, one patient used a cane full time, and one patient was unable to walk because of severe cardiopulmonary disease.

The average arc of knee flexion was 118° (range, 85° to 135°) preoperatively and 121° (range, 100° to 140°) at the time of the final follow-up. Thirty-nine knees (80%) had at least 120° of flexion at the time of the latest follow-up.

At the time of surgery, the anterior cruciate ligament was found to be intact in all patients but one. This patient had a Hospital for Special Surgery knee score of 100 points at the time of the latest follow-up, 153 months postoperatively. Two patients who had had a previous proximal tibial osteotomy had Hospital for Special Surgery knee scores of 95 and 81 points at the time of the final follow-up.

In the group of knees that were treated with medial compartment arthroplasty, the average preoperative deformity was 8° of varus (range, 3° of varus to 14° of varus) from the mechanical axis. The average postoperative alignment was 2° of varus (range, 2° of valgus to 10° of varus), for an average correction of 6°. Ten (17%) of the fifty-nine knees that were treated with medial unicompartmental arthroplasty were not corrected to within 5° of the neutral mechanical axis (range, 5° of varus to 10° of varus).

Back to Top | Article Outline

Conversion to Total Knee Arthroplasty

Two knees, with Hospital for Special Surgery scores of 63 and 60 points, underwent conversion to total knee replacement because of progressive patellofemoral arthritis. One of these knees was in a fifty-eight-year-old woman who had had a medial unicompartmental knee arthroplasty for the treatment of osteoarthritis. At the time of the index arthroplasty, the patellofemoral articulation demonstrated Outerbridge Grade-2 chondrosis19. Heterotopic ossification in the quadriceps tendon developed postoperatively, and anterior knee pain subsequently developed five years postoperatively. By seven years postoperatively, the patient had increasing pain anteriorly as well as evidence of osteophytes around the patellofemoral joint. At eighty-seven months, she underwent total knee arthroplasty.

The second patient who required total knee replacement was a fifty-one-year-old woman who had undergone sequential bilateral medial unicompartmental knee arthroplasty for the treatment of osteoarthritis. At the time of the index arthroplasty, there was no evidence of degenerative changes in the lateral compartment of the right knee and the patellofemoral joint exhibited a large area of Outerbridge Grade-2 involvement19. Approximately 120 months later, degeneration of the patellofemoral compartment was evident radiographically and the patient had anterior knee pain that was refractory to conservative measures. She underwent total knee arthroplasty at 127 months.

In both cases, the unicompartmental knee arthroplasty was converted to a straightforward tricompartmental knee replacement without the need for blocks, wedges, or augmentation. At the time of the revision operation, the unicompartmental components in both knees were found to be well fixed intraoperatively.

Back to Top | Article Outline

Technical Complications

Technical complications included three intraoperative tibial plateau fractures and one avulsion of the medial collateral ligament. One of the fractures was unrecognized at the time of surgery but was noted as a displaced marginal medial tibial plateau fracture on the first postoperative radiograph. The patient was managed nonoperatively with restricted weight-bearing until the fracture healed. At the time of the latest evaluation, ten years postoperatively, the Hospital for Special Surgery score was 88 points and radiographic evaluation revealed a stable tibial component. The knee had a mechanical axis of 1° of valgus and an anatomic axis of 7° of valgus. Radiographs revealed a 2-mm radiolucent line where the fracture intersected the cement mantle of the tibial component at its outer edge. The radiolucent line had not progressed since forty-four months postoperatively.

The second intraoperative fracture involved the medial tibial plateau and was fixed with a 6.5-mm screw. The fracture healed without adverse sequelae. At the time of the last followup, the patient was pain-free and the Hospital for Special Surgery knee score was 82 points. The relatively low overall score was attributed to severe peripheral neuropathy that was unrelated to the arthroplasty.

The third tibial plateau fracture was observed on the six-week postoperative radiographs of a patient who had undergone bilateral unicompartmental knee arthroplasty. On the lateral radiograph, the well-fixed tibial component had changed position as a result of rotation and/or subsidence of the fracture fragment. The tibial component, which had had 9° of posterior slope on the initial postoperative radiograph, was now noted to have 0° of posterior slope. The tibial component remained well fixed, and the fracture healed uneventfully. At 129 months postoperatively, the knee score was 96 points.

A fourth patient required stapling of the medial collateral ligament after it was avulsed during exposure of the medial part of the tibia. This knee remained clinically stable, and the knee score was 91 points at 161 months postoperatively.

One patient had only 80° of flexion at three weeks postoperatively and underwent manipulation under anesthesia to 145°. At the time of the most recent follow-up, 120 months postoperatively, the range of motion was 0° to 140°. Another patient had pain in the popliteal region and a 10° flexion contracture that persisted until eight months postoperatively. Radiographs revealed a piece of retained cement in the posterior aspect of the knee; the cement particle was removed arthroscopically. The patient reported relief of pain, and the Hospital for Special Surgery knee score was 100 points at 148 months postoperatively.

Back to Top | Article Outline

Other Complications

Two patients had a symptomatic deep venous thrombosis in the calf. The thromboses were confirmed with Doppler ultrasound at one and two weeks. Both patients were managed with Coumadin (warfarin) for three months.

Back to Top | Article Outline

Radiographic Results

At the time of the final radiographic evaluation, no component showed evidence of definite loosening and no knee had evidence of osteolysis. Three knees had a complete tibial radiolucent line. All complete and partial radiolucent lines were <2 mm in thickness. The three knees that had a complete tibial radiolucent line were followed for 11.2, 12.4, and 12.4 years. At the time of the final follow-up, the Hospital for Special Surgery knee scores for these three knees were 83, 83, and 100 points. There were no complete femoral radiolucent lines.

Partial tibial radiolucent lines were observed in at least one zone in nineteen of the forty-nine knees. In the majority (sixteen) of these knees, the tibial radiolucent lines were at the cement-bone interface. In four knees, the lines appeared to be progressive initially but did not progress after three years of follow-up. Seven knees showed partial femoral radiolucent lines. In contrast to the tibial radiolucent lines, which occurred at the cement-bone interface, the majority of the partial femoral radiolucent lines occurred at the cement-prosthesis interface. None of the femoral radiolucent lines was progressive.

Overall, twenty-three (47%) of the forty-nine knees had at least one partial radiolucent line around either the tibial or the femoral component. In the subgroup of eleven knees in patients who weighed >200 lb (90.7 kg), seven (64%) had at least one partial radiolucent line around either the tibial or the femoral component. None of the lines was complete or progressive.

Of the forty-nine knees that had at least ten years of follow-up, forty knees (82%) had some radiographic evidence of deterioration of the opposite compartment or the patellofemoral compartment at the time of the final evaluation and nine knees (18%) did not. However, only fourteen (29%) of the forty-nine knees had measurable loss of joint space: five had a loss of patellofemoral joint space only, seven had a loss of tibiofemoral joint space only, and two had a loss of both patellofemoral and tibiofemoral joint space.

Nineteen knees (39%) had Grade-1 radiographic changes (no loss of joint space) in the opposite compartment, six knees (12%) had Grade-2 changes (≤25% loss of joint space), and three knees had Grade-3 changes (≤50% loss of joint space). The three knees with Grade-3 changes in the opposite compartment had Hospital for Special Surgery knee scores of 92, 93, and 100 points at the time of the latest follow-up (at 156, 144, and 152 months, respectively).

Thirteen knees (27%) had Grade-1 radiographic changes in the patellofemoral joint, one knee had Grade-2 changes, three knees had Grade-3 changes, and four knees had Grade-4 changes. Two of the three knees with Grade-3 radiographic changes underwent revision to a tricompartmental knee arthroplasty, as described earlier. The four knees with Grade-4 radiographic changes (three of which had had a medial arthroplasty and one of which had had a lateral arthroplasty) demonstrated severe patellofemoral joint-space loss secondary to impingement with the femoral component. In all four knees, the subchondral erosion of the patella was not observed radiographically until the time of the latest follow-up evaluation and in all cases it was preceded by progressive patellofemoral joint-space loss on serial radiographs. The average duration of follow-up for these four knees was 151 months, and the average Hospital for Special Surgery knee score was 93 points (range, 89 to 96 points).

Back to Top | Article Outline

Survivorship

Kaplan-Meier analysis of all sixty-two knees revealed a ten-year survival rate of 98.0% (95% confidence interval, 96.0% to 100%) with revision for any reason or radiographic loosening as the end point (Fig. 4), a thirteen-year survival rate of 95.7% (95% confidence interval, 91.4% to 100%) with revision or radiographic loosening as the end point, and a thirteen-year survival rate of 100% with definite aseptic loosening of either component as the end point.

Back to Top | Article Outline

Discussion

The present study of sixty-two cemented modular unicompartmental knee replacements demonstrated excellent clinical and radiographic results after ten to thirteen years of follow-up. The results of the present study compare favorably with the reported results associated with total knee replacement22-25 and with contemporary designs of unicompartmental knee implants4,6,8,26,27.

The survival rate in the present study was 96% after thirteen years of follow-up. The revision rate of 4% is comparable with the rates in other studies4,5,8,28,29. Notably, there were no radiographic signs of definite loosening of any component. The 47% prevalence of nonprogressive partial radiolucent lines in the present series is lower than the prevalences reported in other studies2,4,20,22,23,26,30-32. Also, the 6% prevalence of complete tibial radiolucent lines in this series is lower than that reported in other studies4,26,30. Thus far, these radiolucent lines have not led to failure or poor clinical scores.

At the time of the latest follow-up, 18% of the knees had progressive loss of joint space in the opposite compartment and 14% of the knees had progressive loss of joint space in the patellofemoral compartment. Degeneration in the opposite compartment is a common cause of failure after unicompartmental arthroplasty2,28,29,33-35. Some authors have suggested that overcorrection of joint deformity results in the transfer of increased forces to the uninvolved compartment and accelerates degeneration2,29,36. We believe that the overall undercorrection reported in the present series may be responsible for the low rate of degeneration of the opposite compartment. We do not recommend formal medial collateral ligament release because we believe that it will result in overcorrection of the varus deformity. Many authors have suggested that the presence of extensive disease, such as an area of eburnated bone in the patellofemoral joint or contralateral compartment at the time of the index procedure, is the primary factor that predisposes these compartments to further degeneration2,20,28,34. We believe that the stringent selection criteria used in the present study (i.e., not more than Grade-2 chondrosis19 in the other compartments) is probably responsible for our excellent results.

Recently, Hernigou and Deschamps37 documented patellar impingement radiographically in twenty-eight (28%) of ninety-nine knees at an average of fourteen years after unicompartmental knee arthroplasty. In that study, impingement of the patella on the femoral component was associated with anterior placement of the femoral component and resulted in patellofemoral symptoms in the most severe cases. In the present series, impingement was documented radiographically in four knees; all four patients had an excellent clinical score without patellofemoral symptoms after an average duration of follow-up of thirteen years. The impingement occurred secondary to progressive joint-space narrowing as observed on sequential radiographs and, in all four knees, it was not observed until the time of the latest radiographic evaluation.

Alignment of the femoral component in the sagittal plane is important in order to avoid patellar impingement. As shown in Figure 3, the femoral component should be congruent with the curvature of the anterior part of the femur. Any prominence of the femoral component anteriorly may result in patellar impingement. Patellar impingement can occur in association with medial or lateral unicompartmental arthroplasty; however, because of lateral tracking of the patella in many patients, it is more commonly associated with lateral unicompartmental arthroplasty.

The rate of perioperative fracture in this series was high (three of forty-nine). These three fractures were technique-related and occurred in association with the first thirteen procedures performed in this series. At the time of those procedures, we did not predrill the fixation pins for the tibial alignment guide. Since that time, we have predrilled the holes and tibial fracture has not occurred.

The mean flexion at the time of the final follow-up was 121°, with 80% of the knees having at least 120° of flexion. This excellent range of motion slightly exceeds that reported in most studies of total knee replacement22-25,38.

Tibial aseptic loosening and accelerated polyethylene wear are two of the most common causes of failure of unicompartmental knee arthroplasty3,7,15,28,39-41. However, these complications were not seen in the present series despite the use of a thin (5.7-mm) polyethylene liner in more than half the knees, the use of a tibial component with a relatively flat articular surface, and intentional undercorrection of the preoperative angular deformity. We believe that undercorrection of the preoperative angular deformity precludes overstuffing the compartment and thus minimizes polyethylene wear2,28,34.

The metal backing of the tibial component was intended to provide more uniform stresses in the proximal part of the tibia and to protect the cancellous bone. Its use required minimal tibial bone resection and did not adversely affect conversion to total knee arthroplasty15,18,26,31. Total knee replacement can be performed readily after a failed unicompartmental knee arthroplasty if an appreciably small amount of bone was resected during the original implantation and if the holes for fixation with cement do not deeply invade the condylar bone stock26. In the two knees in the present study that underwent conversion to total knee arthroplasty, a standard posterior-stabilized and a standard posterior cruciate-retaining total knee replacement were used without blocks, wedges, or bone grafts. Furthermore, excessive tibial resection was not needed, as demonstrated by our use of 10 and 12-mm tibial components in these two knees.

Although this unicompartmental design was associated with excellent clinical and radiographic results after a minimum duration of follow-up of ten years, we urge its use only in properly selected patients. ▪

NOTE: The authors thank Regina Barden, RN, for all of her help with our total joint program. Without her knowledge and help, we could not have completed this paper.

In support of their research or preparation of this manuscript, one or more of the authors received grants or outside funding from Zimmer. In addition, one or more of the authors received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity (Zimmer). Also, a commercial entity (Zimmer) paid or directed, or agreed to pay or direct, benefits to a research fund, foundation, educational institution, or other charitable or nonprofit organization with which the authors are affiliated or associated.

Investigation performed at the Department of Orthopaedic Surgery, Rush Medical College, Rush-Presbyterian-St. Luke's Medical Center, Chicago, Illinois

1. , Walker P. Unicondylar knee replacement. Clin Orthop. 1976;120: 83-5.
2. . Unicompartmental tibiofemoral resurfacing arthroplasty. J Bone Joint Surg Am. 1978;60: 182-5.
3. , Nedeff DD, Barden RM, Sheinkop MM, Jacobs JJ, Rosenberg AG, Galante JO. Unicompartmental knee arthroplasty. Clinical experience at 6- to 10-year follow-up. Clin Orthop. 1999;367: 50-60.
4. , Sanouiller JL, Grelsamer RP. Unicompartmental knee arthroplasty surgery. 10-year minimum follow-up period. J Arthroplasty. 1996;11: 782-8.
5. , Marmor L, Gibson A, Rougraff BT. Unicompartmental knee arthroplasty. A multicenter investigation with long-term follow-up evaluation. Clin Orthop. 1993;286: 154-9.
6. , Goodfellow JW, O'Connor JJ. The Oxford medial unicompartmental arthroplasty: a ten-year survival study. J Bone Joint Surg Br. 1998;80: 983-9.
7. , Callaghan JJ, Goetz DD, Sullivan PM, Johnston RC. Unicompartmental knee replacement. A minimum 15 year follow-up study. Clin Orthop. 1999;367: 61-72.
8. , Price AJ. Oxford medial unicompartmental knee arthroplasty. A survival analysis of an independent series. J Bone Joint Surg Br. 2001;83: 191-4.
9. , Borgquist L, Knutson K, Lewold S, Lidgren L. Use of unicompartmental instead of tricompartmental prostheses for unicompartmental arthrosis in the knee is a cost-effective alternative. 15,437 primary tricompartmental prostheses were compared with 10,624 primary medial or lateral unicompartmental prostheses. Acta Orthop Scand. 1999;70: 170-5.
10. , Heck DA, Gibson AE. A comparison of tricompartmental and unicompartmental arthroplasty for the treatment of gonarthrosis. Clin Orthop. 1991;273: 157-64.
11. , Zelicof SB, Scott RD, Ewald FC. Unicompartmental versus total knee arthroplasty in the same patient. A comparative study. Clin Orthop. 1991;273: 151-6.
12. , Galante JO, Fermier RW. The influence of total knee-replacement design on walking and stair-climbing. J Bone Joint Surg Am. 1982;64: 1328-35.
13. , Mikosz RP, Andriacchi TP, Rosenberg AG. Functional analysis of cemented medial unicompartmental knee arthroplasty. J Arthroplasty. 1996;11: 553-9.
14. , Landsiedl F. Revision surgery after failed unicompartmental knee arthroplasty: a study of 35 cases. J Arthroplasty. 2000;15: 982-9.
15. , Engh GA, Ammeen DJ. Revision of failed unicompartmental knee arthroplasty. Clin Orthop. 2001;392: 279-82.
16. US markets for reconstructive devices 2001. Toronto: Millennium Research Group; 2002.
17. US markets for reconstructive devices 2002. Toronto: Millennium Research Group; 2003.
18. , Scott R. Unicondylar knee arthroplasty. J Bone Joint Surg Am. 1989;71: 145-50.
19. . The etiology of chondromalacia patellae. J Bone Joint Surg Br. 1961;43: 752-7.
20. , Aglietti P. A five to seven-year follow-up of unicondylar arthroplasty. J Bone Joint Surg Am. 1980;62: 1329-37.
21. , Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53: 457-81.
22. , Padgett DE, Ohashi Y. Total knee arthroplasty for patients younger than 55 years. Clin Orthop. 1989;248: 27-33.
23. , Shine JJ. Duo-condylar total knee arthroplasty. Clin Orthop. 1973;94: 185-95.
24. , Herbst SA, Keating EM, Faris PM, Meding JB. Long-term survival analysis of a posterior cruciate-retaining total condylar total knee arthroplasty. Clin Orthop. 1994;309: 136-45.
25. , Barden RM, Galante JO. Cemented and ingrowth fixation of the Miller-Galante prosthesis. Clinical and roentgenographic comparison after three- to six-year follow-up studies. Clin Orthop. 1990;260: 71-9.
26. , Santore RF. Unicondylar unicompartmental replacement for osteoarthritis of the knee. J Bone Joint Surg Am. 1981;63: 536-44.
27. , Lindstrand A, Lewold S. Unicompartmental knee arthroplasty with special reference to the Swedish Knee Arthroplasty Register. In: Cartier P, Deschamps G, Epinette JA, Hernigou P, editors. Unicompartmental knee arthroplasty. Paris: Expansion Scientifique Francaise; 1997. p 159-62.
28. . Unicompartmental knee arthroplasty. Ten- to 13-year follow-up study. Clin Orthop. 1988;226: 14-20.
29. , Cobb AG, McQueary FG, Thornhill TS. Unicompartmental knee arthroplasty. Eight- to 12-year follow-up evaluation with survivorship analysis. Clin Orthop. 1991;271: 96-100.
30. , Kershaw CJ, Benson MK, O'Connor JJ. The Oxford Knee for unicompartmental osteoarthritis. The first 103 cases. J Bone Joint Surg Br. 1988;70: 692-701.
31. . Results of single compartment arthroplasty with acrylic cement fixation. A minimum follow-up of two years. Clin Orthop. 1977;122: 181-8.
32. , Becker MW, Insall JN. Unicondylar knee arthroplasty. An evaluation of selection criteria. Clin Orthop. 1993;286: 143-8.
33. , Newman JH, Ackroyd CE. Revision of unicompartmental arthroplasty of the knee. Clinical and technical considerations. J Arthroplasty. 1998;13: 191-6.
34. , Sarangi PP, Newman JH. Revision total knee arthroplasty. Comparison of outcome following primary proximal tibial osteotomy or unicompartmental arthroplasty. J Arthroplasty. 1994;9: 539-42.
35. , McAuley JP, Ammeen DJ, Engh GA. The effect of alignment of the knee on the outcome of unicompartmental knee replacement. J Bone Joint Surg Br. 2002;84: 351-5. Erratum in: J Bone Joint Surg Br. 2002;84:1091.
36. . Marmor modular knee in unicompartmental disease. Minimum four-year follow-up. J Bone Joint Surg Am. 1979;61: 347-53.
37. , Deschamps G. Patellar impingement following unicompartmental arthroplasty. J Bone Joint Surg Am. 2002;84: 1132-7.
38. , Rosenberg AG, Borden RM, Sheinkop MB, Jacobs JJ, Galante JO. Long term followup of the Miller Galante total knee replacement. Clin Orthop. 2001;388: 58-67.
39. , Scott RD. Revision of failed unicondylar unicompartmental knee arthroplasty. J Bone Joint Surg Am. 1987;69: 1328-35.
40. , McAuley JP. Unicondylar arthroplasty: an option for high-demand patients with gonarthrosis. Instr Course Lect. 1999;48: 143-8.
41. , Ozuna RM, Scott RD, Thornhill TS. Conversion of failed modern unicompartmental arthroplasty to total knee arthroplasty. J Arthroplasty. 1996;11: 797-801.
Copyright 2005 by The Journal of Bone and Joint Surgery, Incorporated