Converting unicompartmental knee arthroplasty (UKA) to total knee arthroplasty can be difficult, and specialized techniques are needed. Issues include bone loss, joint-line, sizing, and rotation. Determining the complexity of conversion preoperatively helps predict the need for augmentation, grafting, stems, or constraint. We examined insert thickness, augmentation, stem use, and effect of failure mode on complexity of UKA conversion. Fifty cases (1997-2007) were reviewed: 9 implants (18%) were modular fixed-bearing, 4 (8%) were metal-backed nonmodular fixed-bearing, 8 (16%) were resurfacing onlay, 10 (20%) were all-polyethylene step-cut, and 19 (38%) were mobile bearing designs; 5 knees (10%) failed due to infection, 5 (10%) due to wear and/or instability, 10 (20%) for pain or progression of arthritis, 8 (16%) for tibial fracture or severe subsidence, and 22 (44%) due to loosening of either one or both components. Complexity was evaluated using analysis of variance and chi-squared 2-by-k test (80% power; 95% confidence interval). Insert thickness was no different between implants (P=.23) or failure modes (P=.27). Stemmed component use was most frequent with nonmodular components (50%), all-polyethylene step-cut implants (44%), and modular fixed-bearing implants (33%; P=.40). Stem use was highest in tibial fracture (86%; P=.002). Augment use was highest among all-polyethylene step-cut implants (all-polyethylene, 56%; metal-backed, 50%; modular fixed-bearing, 33%; P=.01). Augmentation use was highest in fracture (86%) and infection (67%), with a significant difference noted between failure modes (P=.003). Failure of nonmodular all-polyethylene step-cut devices was more complex than resurfacing or mobile bearing. Failure mode was predictive of complexity. Reestablishing the joint-line, ligamentous balance, and durable fixation are critical to assuring a primary outcome.
Unicompartmental knee arthroplasty (UKA) or partial knee replacements are an important element in the armamentarium of orthopedic surgeons. Unicompartmental knee arthroplasty has been well documented in the literature, and the survivorship is reported to be comparable to total knee arthroplasty (TKA).1 While UKA is considered by most to be acceptable, any surgical procedure has an underlying failure rate. The current recommendation for management of UKA failure is conversion to TKA.2 Conversion to TKA can be a technically demanding procedure. Often, bony landmarks are either lost due to implant failure or sacrificed during the index procedure, and this complicates subsequent reconstructive efforts. Stems and/or metallic augmentation are sometimes necessary for adequate conversion from a failed UKA to a stable TKA.
Recent reports have argued that with modern implant designs, conversion from UKA to TKA may not be as complex as once believed.3-5 The majority of conversion procedures can be successfully completed using primary TKA components with outcomes that rival primary TKA.4,5 The surgeon must, however, be prepared for the technical difficulties that may arise. The ultimate goal in converting UKA to TKA is to assure a primary outcome. Knowledge of these technical difficulties and predicting them requires an understanding of the variables that correlate with increased challenges and the need for more complex reconstructive options.
This article reports the experiences of a single institution in converting failed UKA to TKA. Our objective was to correlate UKA failure mode and UKA implant type with increasing technical difficulty and the need for more complex revision components.
Materials and Methods
Between 1997 and 2007, 59 consecutive TKAs converted from failed UKAs were identified. Nine cases were excluded due to incomplete perioperative data, leaving a study group of 50 knees in 50 patients.
Unicompartmental knee arthroplasty implant design was identified and stratified according to 5 basic categories. Nine implants (18%) were modular fixed-bearing, 4 (8%) were metal-backed fixed-bearing nonmodular, 8 (16%) were resurfacing onlay, 10 (20%) were all-polyethylene step-cut, and 19 (38%) were mobile bearing designs.
Mode of UKA failure was identified preoperatively for all revised knees. Five knees (10%) failed due to infection, 5 (10%) due to wear and/or instability, 10 (20%) for pain or progression of arthritis, 8 (16%) for tibial fracture or severe subsidence, and 22 (44%) due to loosening of either one or both components.
Statistical evaluation was performed on all partial knees that were converted to TKA. Complexity of revision was assessed using analysis of variance (ANOVA) and chi-squared 2-by-k testing (80% power; 95% confidence interval).
Conversion/Revision Complexity Stratified by UKA Implant Design
Revision details were correlated to unicondylar knee implant design used in the index procedure (Table 1). There were 10 all-polyethylene step-cut implants, which failed at an average of 61.2 months from the time of the index procedure. All 10 were converted to primary, unconstrained TKA, with an average polyethylene size of 11.9 mm. Of these, 4 (40%) required the use of stems, and 5 (50%) required metallic augmentation.
Resurfacing onlay implants were used in 8 knees, which failed at an average of 40.3 months from the index procedure. Of these failures, 1 was converted to another UKA, and 7 were converted to primary, unconstrained TKAs. The average polyethylene size in TKAs measured 11.6 mm. One conversion to TKA (12.5%) required a stem, and 1 (12.5%) required metallic augmentation.
Metal-backed fixed-bearing nonmodular implants failed at an average of 13.7 months from the index procedure. Three of these were converted to primary unconstrained TKAs, and 1 was converted to a constrained implant design. The average polyethylene size measured 12.5 mm. Of these revisions, 2 (50%) required stems, and 2 (50%) required metallic augmentation.
Metal-backed fixed-bearing modular implants accounted for 9 failures in this cohort, and they failed at an average of 100.7 months from the time of implantation. All 9 were converted to primary unconstrained TKA, with an average polyethylene thickness of 13.4 mm. Two of these revisions (22.2%) required stems, and 2 (22.2%) required the use of metallic augmentation for reconstruction.
Finally, mobile-bearing UKA accounted for 19 failures in this study. These knees failed at an average of 12.3 months from the index procedure, and 18 were converted to primary unconstrained TKAs with an average polyethylene thickness of 13.1 mm. Four of these revisions (21.6%) required stems, and 5 (26.3%) required metallic augmentation. One was revised to another UKA.
Conversion/Revision Complexity Stratified by Failure Mode
The complexity of revision was correlated with the mode of UKA failure (Table 2). Eight cases failed secondary to fracture at an average of 16.8 months from the time of UKA. All 8 were converted to primary unconstrained TKAs, with an average polyethylene size of 12.6 mm. Six cases (75%) required the use of stems, and 6 (75%) required supplementation with metallic augmentation.
Five UKAs failed secondary to infection at an average of 9.7 months. After debridement, placement of an antibiotic spacer, and appropriate antibiotic treatment, 4 were converted to primary unconstrained implants, and 1 was converted to a constrained TKA. The average polyethylene thickness was 14.0 mm. Two cases (40%) required stems, and 3 (60%) required augmentation.
Eight UKAs failed secondary to wear and/or instability at an average of 70.6 months. Four were converted to primary unconstrained TKAs, and 1 was revised to another UKA. Average polyethylene thickness for the TKAs measured 11.6 mm. Two cases (25%) required stems, and 1 case (12.5%) required augmentation.
Pain and/or progression of osteoarthritis in the other compartments accounted for 10 revisions in this series. All were converted to primary unconstrained TKA at an average of 50.6 months, and the average polyethylene thickness measured 13.0 mm. No cases required either stems or metallic augmentation.
Aseptic loosening accounted for 22 failures in this series. The average time to failure was 49.4 months. One was exchanged for another UKA, and the remaining 22 were converted to primary, unconstrained TKAs. Average polyethylene thickness was 12.4 mm. Two cases (22.2%) required stems and 4 (18.1%) required augmentation.
Analysis of variance testing revealed the fixed-bearing implant designs requiring a tibial step cut to be statistically associated with increased complexity of conversion (P<.01; figure="" 1).="" unicompartmental="" knee="" arthroplasty="" failure="" modes="" of="" tibial="" fracture="" and="" infection="" were="" found="" to="" be="" statistically="" associated="" with="" increased="" complexity="" of="" conversion="">P<.01). the="" most="" significant="" factors="" were="" a="" metal-backed="" fixed-bearing="" nonmodular="" tibial="" design="" and="" tibia="" fracture="">P<.01).>
| || |
Figure 1: Statistical evaluation. The failure etiology of tibial plateau fracture or implant designs requiring a step-cut tibial resection and a nonmodular component was associated with significantly increased surgical complexity (P=.01). Figure 2: Tibial preparation. The tibial resection is templated preoperatively. Resection should not go below the failed UKA device, as this defect may be addressed with cement, screws and cement, bone grafting, or augmentation.
Most studies report that conversion of UKA to TKA is usually possible with a primary unconstrained implant.3-11 Osseous defects, bone loss, and instability have not been found to complicate revision efforts in most modern series.3-5 We report that conversion to TKA has the highest complexity when revision is performed for tibial fracture and with failed modular step-cut UKA designs. The fixed-bearing step-cut designs produce more bone loss, similar to the situation encountered in fractures.
Springer et al7 reported on a cohort of 22 patients who required conversion from UKA to TKA. All patients were converted to primary, cruciate-retaining implants, although 6 patients (27%) required autograft supplementation of contained femoral defects. None of these patients required metallic augmentation or stems on the femoral side. Ten patients (45%) required autografting of contained tibial defects, as these defects were more problematic than femoral defects. Five patients (23%) required metallic tibial augmentation and 2 (9%) required tibial stems. Thus, it appears that contained femoral defects do not impact revision efforts as much as contained tibial defects.7 Our data support the contention that, in failed UKA, tibial-sided defects require augmentation and stems more frequently than femoral reconstruction. As noted, the need for stems and augmentation also correlates with failure modes of fracture or infection, and fixed-bearing UKA implant types that require a tibial step-cut for implantation.
Similar to the findings of the present study, Aleto et al6 correlated failure mode with implant type. They classified UKA failure into 5 different modes: medial tibial collapse, tricompartmental progression, component failure, aseptic loosening, and polyethylene wear. As in our study, failure from progression, loosening, and wear were managed with primary TKA, without supplemental stabilization or constraint. In the study by Aleto et al,6 UKAs revised due to medial tibial collapse or component failure required constrained TKA more frequently. This is further supported by Saldanha et al3 who correlated intraoperative findings with the type of TKA components used in revision. In short, as we also found, the majority of revisions were adequately performed with primary nonconstrained implants, but the surgeon must be ready to stem, augment, bone graft, and/or use constraint when appropriate. Failure mode and implant type are strong predictors of these needs and can be used by surgeons to better assure a primary outcome.
Technical aspects of ensuring a primary outcome can be divided into dealing with the implant, dealing with the failure mode, and adequate reconstruction. First, it is strongly recommended that the UKA implants be left in-situ for as much of the conversion procedure as possible. This allows more accurate approximation of the joint-line and enables the surgeon to better reference accurate landmarks. It is not necessary to resect down to the level of the intact medial tibial bone (Figure 2). Referencing the tibial resection off the intact lateral bone allows for a more conservative resection, and screws and cement or occasionally metallic augmentation can be used to raise the level of the medial compartment to the level of the lateral compartment resection (Figure 3).
Figure 3: Use of medial tibial augments and stems. Tibial plateau fracture, as shown here, was associated with the highest need for stems and augmentation. This is an example of a patient who sustained a traumatic plateau fracture 4 months after UKA. At the time of treament, the implant was found to be loose from the bone. Final construct using a primary femoral component, medial tibial augment, and cementless stem fixation is shown.
The femoral component should be left in place during distal femoral resection and only removed to complete the cut. This better estimates the joint line in extension. Another important consideration is ensuring proper femoral rotation (Figure 4). When converting a partial medial knee to a total knee, bone loss from the posterior portion of the medial femoral condyle is not uncommon. This can be due to bone preparation in the index surgery, femoral loosening, or bone loss incurred during the conversion procedure. Posterior femoral condyle referencing for rotation and size is rendered unreliable. Great care must be taken to align the femoral component with the AP axis, epicondylar axis, or both to allow for balanced flexion and extension gaps. If the femoral implant can be left in place or temporarily reinserted during sizing and establishing rotation, a more accurate flexion gap can be created. Alternatively, the surgeon may insert a spacer measuring approximately 3 to 5 mm onto the posterior femoral condyle during posterior condylar referencing (Figure 5). The back end of an osteotome works well for this, and in so doing, the appropriate rotation can be obtained for the femoral component.
Conversion from UKA to TKA has been found to have nearly equivalent survivorship to primary TKA.4,5,8 This becomes yet another important piece of evidence that points to UKA as being a truly conservative and minimally invasive procedure. Ensuring a primary outcome demands that the surgeon consider implant type and failure mode preoperatively to predict the complexity of the conversion procedure.
| || |
Figure 4: Femoral rotation. As the posterior condyle is distorted by bone loss from the UKA and implant removal, posterior referencing will not accurately demonstrate femoral rotation and sizing. Instead, the AP axis and epicondylar axis should be used to rotate the TKA femoral component. Figure 5: Femoral rotation. Standardized posterior referencing cannot be used for UKA conversion. The medial jointline will be elevated in flexion and the component severely externally rotated. One method to address this defect is to place the back end of an osteotome between the posterior condyle and the referencing guide.
- Newman J, Pydisetty RV, Ackroyd C. Unicompartmental or total knee replacement; the 15-year results of a prospective randomized controlled trial. J Bone Joint Surg Br. 2009; 91(1):52-57.
- Lewold S, Pobertssn O, Knutson K, Lidgren L. Revision of unicompartmental knee arthroplasty; outcome in 1135 cases from the Swedish Knee Arthroplasty study. Acta Orthop Scand. 1998; 69(5):469-474.
- Saldanha KA, Keys GW, Svard UC, White SH, Rao C. Revision of Oxford medial unicompartmental knee arthroplasty to total knee arthroplasty-results of a multicentre study. Knee. 2007; 14(4):275-279.
- Saragaglia D, Estour G, Nemer C, Colle PE. Revision of 33 unicompartmental knee prostheses using total knee arthroplasty: strategy and results. Int Orthop. 2009; 33(4):969-974. Epub 2008 Jun 18.
- Johnson S, Jones P, Newman JH. The survivorship and results of total knee replacements converted from unicompartmental knee replacements. Knee. 2007; 14(2):154-157.
- Aleto TJ, Berend ME, Ritter MA, Faris PM, Meneghini RM. Early failure of unicompartmental knee arthroplasty leading to revision. J Arthroplasty. 2008; 23(2):159-163.
- Springer BD, Scott RD, Thornhill TS. Conversion of failed unicompartmental knee arthroplasty to TKA. Clin Orthop Relat Res. 2006; (446):214-220.
- McAuley JP, Engh GA, Ammeen DJ. Revision of failed unicompartmental knee arthroplasty. Clin Orthop Relat Res. 2001; (392):279-282.
- Bohm I, Landsiedl F. Revision surgery after failed unicompartmetal knee arthroplasty: a study of 35 cases. J Arthroplasty. 2000; 15(8):982-989.
- Knight JL, Atwater RD, Guo J. Early failure of the porous coated anatomic cemented unicompartmental knee arthroplasty. Aids to diagnosis and revison. J Arthroplasty. 1997; 12(1):11-20.
- Levine WN, Ozuna Rm, Scott RD, Thornhill TS. Conversion of failed modern unicompartmental arthroplasty to total knee arthroplasty. J Arthroplasty. 1996; 11(7):797-801.
Drs Berend and Lombardi are from Joint Implant Surgeons, Inc, New Albany and The Ohio State University Department of Orthopedic Surgery, Columbus, and Dr George is from the Cleveland Clinic Foundation, Cleveland, Ohio.
Drs Berend and Lombardi receive royalties, consulting income, and research funding from Biomet, Inc. Dr George has no relevant financial relationships to disclose.
Presented at Current Concepts in Joint Replacement 2008 Winter Meeting; December 10-13, 2008; Orlando, Florida.
Correspondence should be addressed to: Keith R. Berend, MD, Joint Implant Surgeons, Inc, 7277 Smith’s Mill Rd, New Albany, OH 43054.