- Email: [email protected]
Conversion of failed modern unicompartmental arthroplasty to total knee arthroplasty
The Journal of Arthroplasty Vol. 11 No. 7 1996
Conversion of Failed M o d e r n U n i c o m p a r t m e n t a l Arthroplasty to Total K n e e Arthroplasty William N. Levine, MD,* t Richard M. Ozuna, MD,$§ Richard D. Scott, MD,t§ and Thomas S. Thornhill, MD~-§
Abstract: Between January I983 and January 1991, 29 patients (31 knees) with a failed Robert Brigham metal-backed knee arthroplasty (Johnson & Johnson, Raynham, MA) underwent revision to a total knee arthroplasty (TKA). TWenty-five patients had osteoarthritis, three avascular necrosis, and one rheumatoid arthritis. The average patient age was 72.3 years (range, 49-88 years), and the average weight was 179 lb. (range, 112-242 lb.). The interval between the primary and secondary index procedures averaged 62 months (range, 7-i06 months), and mean postrevision follow-up period was 45 months (range, 24-104 months). The primary mechanism of failure of the UKA was tibial polyethylene wear in 21 knees and opposite compartment progression of arthritis in 10 knees. Sixteen knees had particulate synovitis with dense metallic staining of the synovium. At revision, the posterior cruciate ligament was spared in 30 knees and substituted in 1 knee. Restoration of bony deficiency at revision required cancellous bone-graft for contained defects in seven knees, tibial wedges in four knees, and femoral wedges in two knees. No defects received structural allografts. The data suggest that failed, modern unicompartmental knee arthroplasty can successfully be converted to TKA. In most cases, the posterior cruciate ligament can be spared and bone defects corrected with simple wedges or cancellous grafts. Moreover, the results of revision of failed unicompartmental knee arthroplasty are superior to those of failed TKA and failed high tibial osteotomy and comparable to the authors' results of primary TKA with similar-length follow-up periods. Although these results are encouraging, longer-term follow-up evaluation is required to determine survivorship of these revision arthroplasties. K e y words: unicompartmental, arthroplasty, revision, total, knee.
w h e n used in properly selected patients. Resection of excessive tibia and failure to cap the femoral condyle (resurface the entire mediolateral width) were two major flaws in early implant designs. These led to earlier failures and m o r e difficult revision surgery [10,11]. Moreover, stricter indications have decreased the incidence of UKA in patients with osteoarthritis [12]. These criteria have evolved over the last decade and were n o t specifically used in the patients presented in this study. A theoretical advantage of UKA over total knee arthroplasty (TKA) is that the UKA is a m o r e conservative technique that sacrifices less bone. Two reports in the literature reviewed the revision of
Unicompartmental knee arthroplasty (UKA) is a controversial surgical treatment option for unicompartmental osteoarthritis. Some surgeons believe that it is a viable surgical option in older patients with unicompartmental degenerative disease {1-7], whereas others have criticized the procedure [8-10]. We have performed 750 UKAs since 1974 and continue to be satisfied with the results
From *Tufts University, J-New England Baptist Hospital, SHarvard UniversiO~,and fBrigham and Women's Hospital, Boston, Massachusetts.
Reprint requests: William N. Levine, MD, San Diego Orthopaedic Associates, 4060 Fourth Avenue, Suite 700, San Diego, CA 92130.
797
798
The Journal of Arthroplasty Vol. 11 No. 7 October 1996
UKA to TKA and did not support this p r e s u m e d advantage [10,11]. Both of these studies, however, reported results of earlier prostheses with the aforementioned design flaws. Our previous report showed that revision of failed early-design UKA to TI(A required bone-grafts, augmentation with screws and cement, and long-stem components in more than half of the patients [10]. The purpose of this study is to report the results of conversion of failed m o d e r n UKA to TKA and to determine if the alterations in design from the earlier Robert Brigham prosthesis (Johnson & Johnson, Raynham, MA) allowed for an easier revision to TKA if necessary.
Materials and Methods Between J a n u a r y i983 and J a n u a r y 1991, t w e n t y - n i n e patients (31 knees) u n d e r w e n t conversion of a failed metal-backed Robert Brigham UKA to TKA. All of the patients were available for review. Each was evaluated by the Knee Society score, the Knee Society function score, and the Knee Society roentgenographic evaluation [ 13,14J. There were 18 w o m e n and 11 m e n in this study. The average age at the time of the UKA was 68.2 years (range, 40-82 years). The average weight was 179 lb. (range, 112-242 lb.). Twenty-five patients had osteoarthritis, three patients had osteonecrosis, and one patient had unrecognized inflammatory arthritis. The medial compartment was involved in 29 knees and the lateral compartment in 2. In all patients, the indications for the index UKA was pain localized to one c o m p a r t m e n t with corresponding radiographic evidence of unicompartmental disease. The primary m o d e of failure was tibial polyethylene wear in 2i knees. Sixteen of these 21 knees had particulate synovitis with dense metallic staining of the synovium, as well as evidence of tibial polyethylene wear. These patients all presented with pain, swelling, and a grinding sensation, but did not have evidence of gross loosening. Progression of osteoarthritis in the remaining compartments led to revision in the remaining 10 knees. The average age at revision surgery was 72.3 years (range, 49-88 years). The interval b e t w e e n the index and secondary procedures averaged 62 m o n t h s (range, 7-106 months). The average follow-up period after revision to TKA was 45 m o n t h s (range, 24-106 months). Thirty-one of the 32 knees were revised to posterior cruciate ligament-sparing devices, whereas 1 was revised with posterior cruciate-sacrificing components. PFC components (Johnson & Johnson) were
used in I3 knees, Kinemax components (Howmedica, Rutherford, NJ) in 12, and Osteonics components (Allendale, NJ) in the remaining 6 revisions. Six femoral components and one tibial component were uncemented; the rest were cemented. Two long-stem femoral components and one long-stem tibial component were used. All but three patellas were resurfaced. Indications for patellar resurfadng were significant preoperative patellofemoral complaints and i n t r a o p e r a t i v e d o c u m e n t a t i o n of loss of patello-femoral articular cartilage. The tibial component thickness ranged from 8 to 15 mm. Twentysix of the 31 knees used thicknesses of 12.5 m m or smaller. The remaining five knees were 15.0 m m thick. Cancellous bone-graft was used to fill contained defects in seven kt~ees. No structural bone-graft was used. Four tibias and two femurs required the use of a metal wedge. Seven knees required lateral retinacular release for proper patellofemoral tracldng.
Results At the time of latest follow-up examination, 14 patients had no pain, 15 patients had only mild or occasional pain, and n o n e of the 29 patients had moderate or severe pain as defined by the Knee Society score. The n u m b e r of patients is too small, but no differences were n o t e d in the patients with lateral UKAs nor in those w h o did not undergo patellar resurfacing. The flexion arc averaged 115 ° (range, 90°-140°). Flexion contractures occurred in two patients (2 °, 8°). None of the patients demonstrated ligamentous instability. All patients were able to walk without difficulty, although one patient was severely limited by her contralateral knee. Four patients used a cane outside the house for longer ambulation. None of the patients had difficulty ascending stairs, but most used the handrail for balance w h e n descending stairs. The average Knee Society score was 91 (range, 59-100), and the average Knee Society function score was 81 (range, 75-100). Revision roentgenograms were evaluated as outlined by the Knee Society Roentgenographic Evaluation and Scoring System [13]. The position of the tibial and femoral c o m p o n e n t s was assessed in the anteroposterior and lateral views. The average position of the femoral c o m p o n e n t in the anteroposterior view was 97 ° or 7 ° of anatomic valgus (range, 93 °10I°). The position of the tibial c o m p o n e n t was 90 ° or 0 ° of varus tilt (range, 86°-94°). The average angle of the femoral c o m p o n e n t in the lateral view
Conversion of Failed AnthroplastytoTKA
was 3 ° of flexion, and no femoral c o m p o n e n t was extended. The average tibial angle was 87 ° or 3 ° of posterior tilt in the lateral view. There was no evidence of osteolysis. Each zone was assessed for the presence of radiolucent lines, their position, and evidence of progression on successive radiographs. According to this system, a score of 4 or less indicates no progression and is not significant; 5-9 should be closely followed for progression; and 10 or greater signifies i m p e n d i n g failure. In the follow-up lateral view of the femur, 78% of knees h a d no radiolucent lines in a n y zones. Six knees h a d scores of 1 and one k n e e h a d a score of 2. In the follow-up anteroposterior v i e w of the tibial c o m p o n e n t , 81% of the knees h a d no radiolucent lines in a n y zones. Three knees had scores of 1 a n d three knees h a d scores of 2. In the follow-up lateral v i e w of the tibial c o m p o nent, 97% of the knees had no radiolucent lines in a n y zones. One k n e e h a d a score of 1. In the foll o w - u p sunrise view of the patella, 93% of the knees h a d no radiolucent lines in a n y zones. One k n e e h a d a score of 1 and a n o t h e r k n e e had a score of 2. Three patellas w e r e not resurface& One r h e u m a t o i d patient developed a deep w o u n d infection 8 m o n t h s after her revision TICk. This was initially treated with arthroscopic debridement, but eventually required a second revision TICk. Therefore, the rerevision and reoperation rates were 1 of 31 knees (3%). There were no other significant complications in this study group.
Discussion U n i c o m p a r t m e n t a l arthroplasty continues to be a controversial procedure. Insall and Walker [8] and Laskin [9] h a v e b o t h criticized their l o n g - t e r m results of UKA. Insall and Walker f o u n d that 63% of patients r e v i e w e d at an average 5 years h a d fair or p o o r results. Nearly one third of the knees had b e e n revised to TKAs [8]. P r o p o n e n t s of UKA believe that a m a j o r factor contributing to inferior l o n g - t e r m results c o m p a r e d with TKA is i m p r o p e r patient selection. Chestnut designed a u n i c o m p a r t m e n t a l osteoarthrosis protocol to help predict w h i c h patients w o u l d be appropriate candidates for UKA. The u n i c o m p a r t m e n t a l osteoarthrosis protocol was correct in predicting u n i c o m p a r t m e n t a ] osteoarthrosis in 207 of 208 arthroplasties [1]. Kozinn a n d Scott also advocated strict inclusion criteria: (1) age greater t h a n 60, (2) weight less t h a n 82 kg, (3) no h e a v y laborers, (4) m i n i m a l k n e e pain at rest, (5) p r e o p e r a t i v e flexion arc of 90 ° or more, (6) no m o r e t h a n 5 ° of flexion
•
Levine et al.
799
contracture, (7) angular deformity of the k n e e less t h a n 15 °, (8) intact anterior cruciate ligament, (9), and no significant cartilaginous erosions involving weight-bearing areas of the opposite c o m p a r t m e n t [3]. Insall and Walker predicted that only 5% of osteoarthritic patients w o u l d m e e t these criteria for UKA [81. Two studies h a v e r e v i e w e d conversion of failed UKA to TKA. Padgett et al. reported the results in 19 patients (21 knees) w h o u n d e r w e n t revision f r o m a UKA to a TKA b e t w e e n 1973 and 1983. They f o u n d a m a j o r osseous defect in 76% of the knees and concluded that the UKA is not a conservative procedure. These unicondylar c o m p o n e n t s were early designs, sacrificed m o r e bone stock, and had femoral c o m p o n e n t s that w e r e n a r r o w in the medial-to-lateral dimension, allowing for femoral c o m p o n e n t subsidence. The tibial tray in the original Hospital for Special Surgery unicondylar implant had no tibial lugs, which also led to earlier failure. They concluded that results of a TICA following a failed UKA could be expected to be similar to those for revision TKA and inferior to those for primary TKA [ 10]. Barrett and Scott reported the results of 29 patients w h o had conversion of a failed UICA to a TKA b e t w e e n 1974 and 1984 [11]. Six patients included in this study had r h e u m a t o i d arthritis and would not m e e t the inclusion criteria today. In addition, three patients w h o weighed m o r e t h a n 90 kg at the time of the index procedure would also fail inclusion criteria. Nearly half of the patients in this series required additional procedures such as bone-grafting, long-stem components, or custom components. This was due to the early prosthetic design of the u n i c o m p a r t m e n t a l c o m p o n e n t s that were inserted prior to 1981. Surgical techniques at that time compromised femoral bone stock by using implants with long femoral spikes and failing to adequately cap the prepared surface of the femur. This led to loosening and subsidence of components. In an a t t e m p t to address these design concerns, a n e w e r - d e s i g n u n i c o m p a r t m e n t a l device was introduced in 1981. The femoral c o m p o n e n t was wider so it could cap the resurfaced condyle and distribute weight-bearing forces m o r e evenly. The fixation devices for the femoral c o m p o n e n t w e r e decreased in size and two smaller fixation lugs appear to control rotation sufficiently. The tibial c o m p o n e n t in this design is m e t a l - b a c k e d and continues to be controversial. The purpose of this study was to review the results of conversion of a failed m o d e r n UKA to a TICk to determine if design changes in the implant facilitated revision to TKA w h e n necessary. Twenty-nine
800
The Journal of Arthroplasty Vol. 11 No. 7 October 1996
patients (31 knees) were evaluated clinically and radiographically. The average Knee Society score was 91 (range, 59-100) and the average Knee Society function score was 81 (range, 60-100). These results compare favorably to our results using primary hybrid TKA, w h e r e the average postoperative Knee Society score at follow-up periods of 2 to 4 years was 93, and the average postoperative Knee Society functional score was 77 [15]. The results of this study also compare favorably to our results of primary TKA with follow-up periods of 5 to 9 years [16]. One knee in this series failed because of deep sepsis. Interestingly, this patient had an unrecognized diagnosis of r h e u m a t o i d arthritis at the time of UKA and would not have been a candidate today. The posterior cruciate ligament was spared in all but one knee, and the bone defects encountered were less substantial t h a n in those found in the earlier prostheses. These defects were easily treated with cement, wedges, or long-stem prostheses. Structural grafts were not used to fill any of these defects. We prefer nonstructural allograft for contained, small defects (< 1.5 cm). Metal wedges are used, however, if the defect is n o n c o n t a i n e d and less than 1.5 cm. Structural allograft is used for noncontained defects greater than 1.5 cm. If structural allograft is required to provide stability of the tibial base plate ( n o n s t e m m e d tibial trial), a long-stem c o m p o n e n t is used to stress-shield the tibia. Likewise, a stemmed prosthesis is preferred if structural allograft is required to provide stability of the femoral trial c o m p o n e n t . High tibial osteotomy (HTO) remains a viable surgical option for patients with u n i c o m p a r t m e n t a l arthritis. Although this provides properly selected patients with early pain relief, m a n y of these patients experience r e t u r n of a painful joint. Conversion to a TKA has several disadvantages. Patella infera is a c o m m o n finding in knees following tibial o s t e o t o m y and can alter the patellofemoral mechanics. In addition, the results following conversion of failed HTO to TKA have b e e n disappointing in several studies [17-20]. Windsor et al. f o u n d their results similar to those obtained in revision TKA and had a higher infection rate compared with primary TKA [18]. Katz et al. found that range of motion was decreased in patients following conversion of HTO to TKA as compared with patients after primary TKA (average arc of motion, 95 ° vs 103°). In addition, three patients had an average of 6 ° of fixed flexion contracture in the p o s t o s t e o t o m y group c o m p a r e d with no flexion contracture in the primary TKA group [ 19].
A recent report by Jackson et al. compared the results of revision TKA following either UKA or HTO. Thirty-eight percent of the postosteotomy patients suffered complications. Major complications including deep w o u n d infections, patellar t e n d o n rupture, and peroneal nerve palsy occurred in 28% of the patients in the study. In addition, there was a 19% reoperation rate following the revision surgery. The UKA-TKA group had no major complications, although there were two superficial w o u n d infections successfully treated with oral antibiotics. One of 23 patients in this group required a rerevision at 4 years for tibial loosening (4%) {20]. Conversion of failed TKA to a second tricompartmental arthroplasty has also yielded less favorable results. In a recent study of 40 revision arthroplasties, M u r r a y et al. found that the Knee Society score was 83 (range, 33-100) and the Knee Society function score was 64 (range, 0-100). Range of m o t i o n averaged 101 ° (range, 90°-135°), and 28% of the knees had flexion contractures greater than 5 ° (11/40). The incidence of tibial radiolucencies was 32%. A deep w o u n d infection developed in one knee, which eventually required a resection arthroplasty. The reoperation and rerevision rate, therefore, was 3% (1/40) [21]. Stuart et al. reported a 52% clinical failure rate in 46 knees requiring revision TKA following a failed primary TKA. Nearly 50% percent of these failures were due to problems with the extensor mechanism or patellar c o m p o n e n t [22]. Ritter et al. found that after 4 to 6 years, revision TKA survived in only 53.9% compared with 84.1% of primary TKAs. This difference was statistically significant {23]. We believe that modified implants have made revision surgery easier. A UKA is still a viable option in the treatment of u n i c o m p a r t m e n t a l osteoarthritis in a sedentary, elderly patient. Earlier discouraging results of conversion of failed UKA were based on obsolete implants. At a m i n i m u m follow-up period of 2 years from conversion of UKA to TKA, patients in this study had clinical, functional, and radiographic results comparable to our results for primary TKA with similar-length follow-up periods and superior to revision following tricompartmental arthroplasty or HTO. Longerterm follow-up evaluation is required to determine the survivorship of these revision arthroplasties.
References 1. Chesnut WJ: Preoperative diagnostic protocol to predict candidates for unicompartmental arthroplasty. Clin Orthop 273:146, 1991
Conversion of Failed AnthroplastytoTKA 2. Christensen NO: Unicompartmental prosthesis for gonarthrosis. Clin Orthop 273:165, 1991 3. Kozinn SC, Scott RD: Unicondylar knee arthroplasty. J Bone Joint Surg 71A:145, 1989 4. M a r m o r L: Unicompartmental arthroplasty of the knee with a m i n i m u m ten-year follow-up period. Clin Orthop 228:171, 1988 5. Shurley TIC, O'Donoghue DH, Smith WD et ah Unicompartmental arthroplasty of the knee: a review of three-five year follow-up. Clin Orthop 164:236, 1982 6. Sisto D J, Blazina ME, Heskiaoff D, Hirsh LC: Unicompartmental arthroplasty for osteoarthrosis of the knee. Clin Orthop 286:149, 1993 7. Swank M, Stulberg SD, Jiganti J, Machairas S: The natural history of unicompartmental arthroplasty: an eight-year follow-up with survivorship analysis. Clin Orthop 286:130, i993 8. Insall JN, Walker PS: Unicondylar knee replacement. Clin Orthop 120:83, 1976 9. Laskin RS: Unicompartmental tibiofemoral resurfacing arthroplasty. J Bone Joint Surg 60A:182, 1978 10. Padgett DE, Stern SH, Insall JN: Revision total knee arthroplasty for failed unicompartmental replacement. J Bone Joint Surg 73A:186, 1991 11. Barrett WP, Scott RD: Revision of failed unicondylar unicompartmental knee arthroplasty. J Bone Joint Surg 69A:1328, 1987 I2. Thornhill TS, Scott RD: UnicompartmentaI total knee arthroplasty. Orthop Clin North Am 20:245, 1989 13. Ewald FC: The Knee Society total knee arthroplasty roentgenographic evaluation and scoring system. Clin Orthop 248:9, 1989
•
Levine et al.
801
14. Insall JN, Dorr LD, Scott RD, Scott WN: Rationale of the Knee Society clinical rating system. Clin Orthop 248:13, 1989 15. Wright RJ, Lima J, Scott RD, Thornhill TS: Two-to four-year results of posterior cruciate-sparing condylar total knee arthroplasty with an uncemented femoral component. Clin Orthop 260:80, 1990 16. Martin SD, McManus JL, Scott RD, Thornhil] TS: Press-fit condylar (PFC) total knee replacement: five to nine year follow-up. Presented at the 61st Annual Meeting of the American Academy of Orthopaedic Surgeons, New Orleans, LA, February 1994 17. Mont MA, Antonaides S, Krakow KA, Hungerford DS: Total knee arthroplasty after failed high tibial osteotomy. Clin Orthop 299:125, 1994 18. Windsor RE, Insall JN, Vince KG: Technical considerations of total knee arthroplasty after proximal tibial osteotomy. J Bone Joint Surg 70A:547, 1988 19. Katz MM, Hungerford DS, Krackow KA, Lennox DW: Results of total knee arthroplasty after failed proximal tibial osteotomy for osteoarthritis. J Bone Joint Surg 69A:225, 1987 20. Jackson M, Sarangi PP, N e w m a n JH: Revision total knee arthroplasty: comparison of outcome following primary proximal tibial osteotomy or unicompartmental arthroplasty. J Arthroplasty 9:539, 1994 21. Murray PB, Rand JA, Hanssen AD: Cemented longstern revision total knee arthroplasty. Clin Orthop 309:116, 1994 22. Stuart M J, Larson JE, Morrey BF: Reoperation after condylar revision total knee arthroplasty. Clin Orthop 286:168, 1993
Conversion of Failed M o d e r n U n i c o m p a r t m e n t a l Arthroplasty to Total K n e e Arthroplasty William N. Levine, MD,* t Richard M. Ozuna, MD,$§ Richard D. Scott, MD,t§ and Thomas S. Thornhill, MD~-§
Abstract: Between January I983 and January 1991, 29 patients (31 knees) with a failed Robert Brigham metal-backed knee arthroplasty (Johnson & Johnson, Raynham, MA) underwent revision to a total knee arthroplasty (TKA). TWenty-five patients had osteoarthritis, three avascular necrosis, and one rheumatoid arthritis. The average patient age was 72.3 years (range, 49-88 years), and the average weight was 179 lb. (range, 112-242 lb.). The interval between the primary and secondary index procedures averaged 62 months (range, 7-i06 months), and mean postrevision follow-up period was 45 months (range, 24-104 months). The primary mechanism of failure of the UKA was tibial polyethylene wear in 21 knees and opposite compartment progression of arthritis in 10 knees. Sixteen knees had particulate synovitis with dense metallic staining of the synovium. At revision, the posterior cruciate ligament was spared in 30 knees and substituted in 1 knee. Restoration of bony deficiency at revision required cancellous bone-graft for contained defects in seven knees, tibial wedges in four knees, and femoral wedges in two knees. No defects received structural allografts. The data suggest that failed, modern unicompartmental knee arthroplasty can successfully be converted to TKA. In most cases, the posterior cruciate ligament can be spared and bone defects corrected with simple wedges or cancellous grafts. Moreover, the results of revision of failed unicompartmental knee arthroplasty are superior to those of failed TKA and failed high tibial osteotomy and comparable to the authors' results of primary TKA with similar-length follow-up periods. Although these results are encouraging, longer-term follow-up evaluation is required to determine survivorship of these revision arthroplasties. K e y words: unicompartmental, arthroplasty, revision, total, knee.
w h e n used in properly selected patients. Resection of excessive tibia and failure to cap the femoral condyle (resurface the entire mediolateral width) were two major flaws in early implant designs. These led to earlier failures and m o r e difficult revision surgery [10,11]. Moreover, stricter indications have decreased the incidence of UKA in patients with osteoarthritis [12]. These criteria have evolved over the last decade and were n o t specifically used in the patients presented in this study. A theoretical advantage of UKA over total knee arthroplasty (TKA) is that the UKA is a m o r e conservative technique that sacrifices less bone. Two reports in the literature reviewed the revision of
Unicompartmental knee arthroplasty (UKA) is a controversial surgical treatment option for unicompartmental osteoarthritis. Some surgeons believe that it is a viable surgical option in older patients with unicompartmental degenerative disease {1-7], whereas others have criticized the procedure [8-10]. We have performed 750 UKAs since 1974 and continue to be satisfied with the results
From *Tufts University, J-New England Baptist Hospital, SHarvard UniversiO~,and fBrigham and Women's Hospital, Boston, Massachusetts.
Reprint requests: William N. Levine, MD, San Diego Orthopaedic Associates, 4060 Fourth Avenue, Suite 700, San Diego, CA 92130.
797
798
The Journal of Arthroplasty Vol. 11 No. 7 October 1996
UKA to TKA and did not support this p r e s u m e d advantage [10,11]. Both of these studies, however, reported results of earlier prostheses with the aforementioned design flaws. Our previous report showed that revision of failed early-design UKA to TI(A required bone-grafts, augmentation with screws and cement, and long-stem components in more than half of the patients [10]. The purpose of this study is to report the results of conversion of failed m o d e r n UKA to TKA and to determine if the alterations in design from the earlier Robert Brigham prosthesis (Johnson & Johnson, Raynham, MA) allowed for an easier revision to TKA if necessary.
Materials and Methods Between J a n u a r y i983 and J a n u a r y 1991, t w e n t y - n i n e patients (31 knees) u n d e r w e n t conversion of a failed metal-backed Robert Brigham UKA to TKA. All of the patients were available for review. Each was evaluated by the Knee Society score, the Knee Society function score, and the Knee Society roentgenographic evaluation [ 13,14J. There were 18 w o m e n and 11 m e n in this study. The average age at the time of the UKA was 68.2 years (range, 40-82 years). The average weight was 179 lb. (range, 112-242 lb.). Twenty-five patients had osteoarthritis, three patients had osteonecrosis, and one patient had unrecognized inflammatory arthritis. The medial compartment was involved in 29 knees and the lateral compartment in 2. In all patients, the indications for the index UKA was pain localized to one c o m p a r t m e n t with corresponding radiographic evidence of unicompartmental disease. The primary m o d e of failure was tibial polyethylene wear in 2i knees. Sixteen of these 21 knees had particulate synovitis with dense metallic staining of the synovium, as well as evidence of tibial polyethylene wear. These patients all presented with pain, swelling, and a grinding sensation, but did not have evidence of gross loosening. Progression of osteoarthritis in the remaining compartments led to revision in the remaining 10 knees. The average age at revision surgery was 72.3 years (range, 49-88 years). The interval b e t w e e n the index and secondary procedures averaged 62 m o n t h s (range, 7-106 months). The average follow-up period after revision to TKA was 45 m o n t h s (range, 24-106 months). Thirty-one of the 32 knees were revised to posterior cruciate ligament-sparing devices, whereas 1 was revised with posterior cruciate-sacrificing components. PFC components (Johnson & Johnson) were
used in I3 knees, Kinemax components (Howmedica, Rutherford, NJ) in 12, and Osteonics components (Allendale, NJ) in the remaining 6 revisions. Six femoral components and one tibial component were uncemented; the rest were cemented. Two long-stem femoral components and one long-stem tibial component were used. All but three patellas were resurfaced. Indications for patellar resurfadng were significant preoperative patellofemoral complaints and i n t r a o p e r a t i v e d o c u m e n t a t i o n of loss of patello-femoral articular cartilage. The tibial component thickness ranged from 8 to 15 mm. Twentysix of the 31 knees used thicknesses of 12.5 m m or smaller. The remaining five knees were 15.0 m m thick. Cancellous bone-graft was used to fill contained defects in seven kt~ees. No structural bone-graft was used. Four tibias and two femurs required the use of a metal wedge. Seven knees required lateral retinacular release for proper patellofemoral tracldng.
Results At the time of latest follow-up examination, 14 patients had no pain, 15 patients had only mild or occasional pain, and n o n e of the 29 patients had moderate or severe pain as defined by the Knee Society score. The n u m b e r of patients is too small, but no differences were n o t e d in the patients with lateral UKAs nor in those w h o did not undergo patellar resurfacing. The flexion arc averaged 115 ° (range, 90°-140°). Flexion contractures occurred in two patients (2 °, 8°). None of the patients demonstrated ligamentous instability. All patients were able to walk without difficulty, although one patient was severely limited by her contralateral knee. Four patients used a cane outside the house for longer ambulation. None of the patients had difficulty ascending stairs, but most used the handrail for balance w h e n descending stairs. The average Knee Society score was 91 (range, 59-100), and the average Knee Society function score was 81 (range, 75-100). Revision roentgenograms were evaluated as outlined by the Knee Society Roentgenographic Evaluation and Scoring System [13]. The position of the tibial and femoral c o m p o n e n t s was assessed in the anteroposterior and lateral views. The average position of the femoral c o m p o n e n t in the anteroposterior view was 97 ° or 7 ° of anatomic valgus (range, 93 °10I°). The position of the tibial c o m p o n e n t was 90 ° or 0 ° of varus tilt (range, 86°-94°). The average angle of the femoral c o m p o n e n t in the lateral view
Conversion of Failed AnthroplastytoTKA
was 3 ° of flexion, and no femoral c o m p o n e n t was extended. The average tibial angle was 87 ° or 3 ° of posterior tilt in the lateral view. There was no evidence of osteolysis. Each zone was assessed for the presence of radiolucent lines, their position, and evidence of progression on successive radiographs. According to this system, a score of 4 or less indicates no progression and is not significant; 5-9 should be closely followed for progression; and 10 or greater signifies i m p e n d i n g failure. In the follow-up lateral view of the femur, 78% of knees h a d no radiolucent lines in a n y zones. Six knees h a d scores of 1 and one k n e e h a d a score of 2. In the follow-up anteroposterior v i e w of the tibial c o m p o n e n t , 81% of the knees h a d no radiolucent lines in a n y zones. Three knees had scores of 1 a n d three knees h a d scores of 2. In the follow-up lateral v i e w of the tibial c o m p o nent, 97% of the knees had no radiolucent lines in a n y zones. One k n e e h a d a score of 1. In the foll o w - u p sunrise view of the patella, 93% of the knees h a d no radiolucent lines in a n y zones. One k n e e h a d a score of 1 and a n o t h e r k n e e had a score of 2. Three patellas w e r e not resurface& One r h e u m a t o i d patient developed a deep w o u n d infection 8 m o n t h s after her revision TICk. This was initially treated with arthroscopic debridement, but eventually required a second revision TICk. Therefore, the rerevision and reoperation rates were 1 of 31 knees (3%). There were no other significant complications in this study group.
Discussion U n i c o m p a r t m e n t a l arthroplasty continues to be a controversial procedure. Insall and Walker [8] and Laskin [9] h a v e b o t h criticized their l o n g - t e r m results of UKA. Insall and Walker f o u n d that 63% of patients r e v i e w e d at an average 5 years h a d fair or p o o r results. Nearly one third of the knees had b e e n revised to TKAs [8]. P r o p o n e n t s of UKA believe that a m a j o r factor contributing to inferior l o n g - t e r m results c o m p a r e d with TKA is i m p r o p e r patient selection. Chestnut designed a u n i c o m p a r t m e n t a l osteoarthrosis protocol to help predict w h i c h patients w o u l d be appropriate candidates for UKA. The u n i c o m p a r t m e n t a l osteoarthrosis protocol was correct in predicting u n i c o m p a r t m e n t a ] osteoarthrosis in 207 of 208 arthroplasties [1]. Kozinn a n d Scott also advocated strict inclusion criteria: (1) age greater t h a n 60, (2) weight less t h a n 82 kg, (3) no h e a v y laborers, (4) m i n i m a l k n e e pain at rest, (5) p r e o p e r a t i v e flexion arc of 90 ° or more, (6) no m o r e t h a n 5 ° of flexion
•
Levine et al.
799
contracture, (7) angular deformity of the k n e e less t h a n 15 °, (8) intact anterior cruciate ligament, (9), and no significant cartilaginous erosions involving weight-bearing areas of the opposite c o m p a r t m e n t [3]. Insall and Walker predicted that only 5% of osteoarthritic patients w o u l d m e e t these criteria for UKA [81. Two studies h a v e r e v i e w e d conversion of failed UKA to TKA. Padgett et al. reported the results in 19 patients (21 knees) w h o u n d e r w e n t revision f r o m a UKA to a TKA b e t w e e n 1973 and 1983. They f o u n d a m a j o r osseous defect in 76% of the knees and concluded that the UKA is not a conservative procedure. These unicondylar c o m p o n e n t s were early designs, sacrificed m o r e bone stock, and had femoral c o m p o n e n t s that w e r e n a r r o w in the medial-to-lateral dimension, allowing for femoral c o m p o n e n t subsidence. The tibial tray in the original Hospital for Special Surgery unicondylar implant had no tibial lugs, which also led to earlier failure. They concluded that results of a TICA following a failed UKA could be expected to be similar to those for revision TKA and inferior to those for primary TKA [ 10]. Barrett and Scott reported the results of 29 patients w h o had conversion of a failed UICA to a TKA b e t w e e n 1974 and 1984 [11]. Six patients included in this study had r h e u m a t o i d arthritis and would not m e e t the inclusion criteria today. In addition, three patients w h o weighed m o r e t h a n 90 kg at the time of the index procedure would also fail inclusion criteria. Nearly half of the patients in this series required additional procedures such as bone-grafting, long-stem components, or custom components. This was due to the early prosthetic design of the u n i c o m p a r t m e n t a l c o m p o n e n t s that were inserted prior to 1981. Surgical techniques at that time compromised femoral bone stock by using implants with long femoral spikes and failing to adequately cap the prepared surface of the femur. This led to loosening and subsidence of components. In an a t t e m p t to address these design concerns, a n e w e r - d e s i g n u n i c o m p a r t m e n t a l device was introduced in 1981. The femoral c o m p o n e n t was wider so it could cap the resurfaced condyle and distribute weight-bearing forces m o r e evenly. The fixation devices for the femoral c o m p o n e n t w e r e decreased in size and two smaller fixation lugs appear to control rotation sufficiently. The tibial c o m p o n e n t in this design is m e t a l - b a c k e d and continues to be controversial. The purpose of this study was to review the results of conversion of a failed m o d e r n UKA to a TICk to determine if design changes in the implant facilitated revision to TKA w h e n necessary. Twenty-nine
800
The Journal of Arthroplasty Vol. 11 No. 7 October 1996
patients (31 knees) were evaluated clinically and radiographically. The average Knee Society score was 91 (range, 59-100) and the average Knee Society function score was 81 (range, 60-100). These results compare favorably to our results using primary hybrid TKA, w h e r e the average postoperative Knee Society score at follow-up periods of 2 to 4 years was 93, and the average postoperative Knee Society functional score was 77 [15]. The results of this study also compare favorably to our results of primary TKA with follow-up periods of 5 to 9 years [16]. One knee in this series failed because of deep sepsis. Interestingly, this patient had an unrecognized diagnosis of r h e u m a t o i d arthritis at the time of UKA and would not have been a candidate today. The posterior cruciate ligament was spared in all but one knee, and the bone defects encountered were less substantial t h a n in those found in the earlier prostheses. These defects were easily treated with cement, wedges, or long-stem prostheses. Structural grafts were not used to fill any of these defects. We prefer nonstructural allograft for contained, small defects (< 1.5 cm). Metal wedges are used, however, if the defect is n o n c o n t a i n e d and less than 1.5 cm. Structural allograft is used for noncontained defects greater than 1.5 cm. If structural allograft is required to provide stability of the tibial base plate ( n o n s t e m m e d tibial trial), a long-stem c o m p o n e n t is used to stress-shield the tibia. Likewise, a stemmed prosthesis is preferred if structural allograft is required to provide stability of the femoral trial c o m p o n e n t . High tibial osteotomy (HTO) remains a viable surgical option for patients with u n i c o m p a r t m e n t a l arthritis. Although this provides properly selected patients with early pain relief, m a n y of these patients experience r e t u r n of a painful joint. Conversion to a TKA has several disadvantages. Patella infera is a c o m m o n finding in knees following tibial o s t e o t o m y and can alter the patellofemoral mechanics. In addition, the results following conversion of failed HTO to TKA have b e e n disappointing in several studies [17-20]. Windsor et al. f o u n d their results similar to those obtained in revision TKA and had a higher infection rate compared with primary TKA [18]. Katz et al. found that range of motion was decreased in patients following conversion of HTO to TKA as compared with patients after primary TKA (average arc of motion, 95 ° vs 103°). In addition, three patients had an average of 6 ° of fixed flexion contracture in the p o s t o s t e o t o m y group c o m p a r e d with no flexion contracture in the primary TKA group [ 19].
A recent report by Jackson et al. compared the results of revision TKA following either UKA or HTO. Thirty-eight percent of the postosteotomy patients suffered complications. Major complications including deep w o u n d infections, patellar t e n d o n rupture, and peroneal nerve palsy occurred in 28% of the patients in the study. In addition, there was a 19% reoperation rate following the revision surgery. The UKA-TKA group had no major complications, although there were two superficial w o u n d infections successfully treated with oral antibiotics. One of 23 patients in this group required a rerevision at 4 years for tibial loosening (4%) {20]. Conversion of failed TKA to a second tricompartmental arthroplasty has also yielded less favorable results. In a recent study of 40 revision arthroplasties, M u r r a y et al. found that the Knee Society score was 83 (range, 33-100) and the Knee Society function score was 64 (range, 0-100). Range of m o t i o n averaged 101 ° (range, 90°-135°), and 28% of the knees had flexion contractures greater than 5 ° (11/40). The incidence of tibial radiolucencies was 32%. A deep w o u n d infection developed in one knee, which eventually required a resection arthroplasty. The reoperation and rerevision rate, therefore, was 3% (1/40) [21]. Stuart et al. reported a 52% clinical failure rate in 46 knees requiring revision TKA following a failed primary TKA. Nearly 50% percent of these failures were due to problems with the extensor mechanism or patellar c o m p o n e n t [22]. Ritter et al. found that after 4 to 6 years, revision TKA survived in only 53.9% compared with 84.1% of primary TKAs. This difference was statistically significant {23]. We believe that modified implants have made revision surgery easier. A UKA is still a viable option in the treatment of u n i c o m p a r t m e n t a l osteoarthritis in a sedentary, elderly patient. Earlier discouraging results of conversion of failed UKA were based on obsolete implants. At a m i n i m u m follow-up period of 2 years from conversion of UKA to TKA, patients in this study had clinical, functional, and radiographic results comparable to our results for primary TKA with similar-length follow-up periods and superior to revision following tricompartmental arthroplasty or HTO. Longerterm follow-up evaluation is required to determine the survivorship of these revision arthroplasties.
References 1. Chesnut WJ: Preoperative diagnostic protocol to predict candidates for unicompartmental arthroplasty. Clin Orthop 273:146, 1991
Conversion of Failed AnthroplastytoTKA 2. Christensen NO: Unicompartmental prosthesis for gonarthrosis. Clin Orthop 273:165, 1991 3. Kozinn SC, Scott RD: Unicondylar knee arthroplasty. J Bone Joint Surg 71A:145, 1989 4. M a r m o r L: Unicompartmental arthroplasty of the knee with a m i n i m u m ten-year follow-up period. Clin Orthop 228:171, 1988 5. Shurley TIC, O'Donoghue DH, Smith WD et ah Unicompartmental arthroplasty of the knee: a review of three-five year follow-up. Clin Orthop 164:236, 1982 6. Sisto D J, Blazina ME, Heskiaoff D, Hirsh LC: Unicompartmental arthroplasty for osteoarthrosis of the knee. Clin Orthop 286:149, 1993 7. Swank M, Stulberg SD, Jiganti J, Machairas S: The natural history of unicompartmental arthroplasty: an eight-year follow-up with survivorship analysis. Clin Orthop 286:130, i993 8. Insall JN, Walker PS: Unicondylar knee replacement. Clin Orthop 120:83, 1976 9. Laskin RS: Unicompartmental tibiofemoral resurfacing arthroplasty. J Bone Joint Surg 60A:182, 1978 10. Padgett DE, Stern SH, Insall JN: Revision total knee arthroplasty for failed unicompartmental replacement. J Bone Joint Surg 73A:186, 1991 11. Barrett WP, Scott RD: Revision of failed unicondylar unicompartmental knee arthroplasty. J Bone Joint Surg 69A:1328, 1987 I2. Thornhill TS, Scott RD: UnicompartmentaI total knee arthroplasty. Orthop Clin North Am 20:245, 1989 13. Ewald FC: The Knee Society total knee arthroplasty roentgenographic evaluation and scoring system. Clin Orthop 248:9, 1989
•
Levine et al.
801
14. Insall JN, Dorr LD, Scott RD, Scott WN: Rationale of the Knee Society clinical rating system. Clin Orthop 248:13, 1989 15. Wright RJ, Lima J, Scott RD, Thornhill TS: Two-to four-year results of posterior cruciate-sparing condylar total knee arthroplasty with an uncemented femoral component. Clin Orthop 260:80, 1990 16. Martin SD, McManus JL, Scott RD, Thornhil] TS: Press-fit condylar (PFC) total knee replacement: five to nine year follow-up. Presented at the 61st Annual Meeting of the American Academy of Orthopaedic Surgeons, New Orleans, LA, February 1994 17. Mont MA, Antonaides S, Krakow KA, Hungerford DS: Total knee arthroplasty after failed high tibial osteotomy. Clin Orthop 299:125, 1994 18. Windsor RE, Insall JN, Vince KG: Technical considerations of total knee arthroplasty after proximal tibial osteotomy. J Bone Joint Surg 70A:547, 1988 19. Katz MM, Hungerford DS, Krackow KA, Lennox DW: Results of total knee arthroplasty after failed proximal tibial osteotomy for osteoarthritis. J Bone Joint Surg 69A:225, 1987 20. Jackson M, Sarangi PP, N e w m a n JH: Revision total knee arthroplasty: comparison of outcome following primary proximal tibial osteotomy or unicompartmental arthroplasty. J Arthroplasty 9:539, 1994 21. Murray PB, Rand JA, Hanssen AD: Cemented longstern revision total knee arthroplasty. Clin Orthop 309:116, 1994 22. Stuart M J, Larson JE, Morrey BF: Reoperation after condylar revision total knee arthroplasty. Clin Orthop 286:168, 1993