|Home | About | Journals | Submit | Contact Us | Français|
Modular, metal-backed tibial (MBT) components are associated with locking mechanism dysfunction, breakage, backside wear, and osteolysis, which compromise survivorship. All-polyethylene tibial (APT) components eliminate problems associated with MBTs, but, historically, APT utilization has generally been limited to older, less active patients. However, it is unclear whether APT utilization can be expanded to a nonselected patient population.
We therefore determined the survivorship of APT components compared with MBT components in a non-age- or activity-selected population who underwent TKA.
Using a longitudinal database, we identified 775 patients with primary TKAs utilizing a single implant design between 1999 and 2007. Of these, 558 (72%) patients had APT components (APT2), while 217 (28%) patients with tibial bone loss or defects, contralateral MBT components, or a BMI of greater than 37.5 received MBT components. We determined the survivorship in the two groups. The minimum followup was 2 years for both groups (mean ± SD: MBT, 80 ± 29 months; APT, 63 ± 27 months). The APT group was older (average age: APT2, 70 years; MBT, 64.7 years) and had a lower BMI than the MBT group (APT2, 30.8; MBT, 33.8).
Survivorship, as defined by revision for any reason, was 99% for the APT group and 97% for the MBT group. There were four (2%) tibial failures in the MBT group in patients with a BMI of greater than 40. There were no revisions for loosening or osteolysis in the APT group.
APT implants perform as well as MBT implants in a non-age- or activity-selected TKA population with a BMI of less than 37.5.
Level III, therapeutic study. See Guidelines for Authors for a complete description of levels of evidence.
Early TKA designs used cemented all-polyethylene tibial (APT) components [6, 7, 9, 12, 18, 21, 28, 33, 38, 39]. Both posterior-cruciate-preserving and posterior-cruciate-substituting designs utilizing polymethylmethacrylate bone cement for fixation showed greater than 90% survivorship at followup of 10 or more years [11, 31, 32, 36, 37]. APT utilization decreased markedly in the late 1980s and 1990s due to several concerning reports about APT components [6, 7, 9, 10, 12]. Metal-backed tibial (MBT) components with modularity were introduced to improve intraoperative flexibility [11, 21, 36], provide options for revision surgery, and improve the survivorship of the tibial implant [2, 37, 38] by improving load transfer to the tibial plateau and improving fixation [33, 38, 39].
However, in the late 1990s, the benefits of modular MBT components were questioned by reports of backside wear of the polyethylene modular insert, poorly functioning tibial base plate locking mechanisms, and osteolysis associated with these problems [4, 8, 24, 27]. Alternatives to MBT components were therefore developed. Mobile-bearing tibial components, nonmodular MBT components, and improved versions of MBT components with better locking mechanisms and improved polyethylene sterilization techniques became available. Modern MBT components continue to be the most commonly used tibial implant [11, 13, 17, 35].
Although some early cemented APT components provided survivorship ranging from 96% to 98%, with revision as an end point at 5 to 14 years of followup [18, 28], survivorship rates of 72% to 90% at 5 to 10 years’ followup were reported in some TKA designs when indications were expanded to a higher-demand patient population [7, 10, 11, 31, 32, 37]. Consequently, APT components became less frequently utilized, most often being implanted in older or lower-demand patients [1, 22, 26]. In the late 1990s, the survivorship of APT components in low-demand patients was greater than 95% at 10 or more years of followup, and utilization in non-age- or activity-selected patients began [3, 5, 14, 25, 29, 32, 34]. Surgeons implanted APT components fabricated with polyethylene sterilized with gamma irradiation and sealed in a vacuum, designed with a stiff tibial keel, and manufactured in a variety of sizes to cover the proximal tibial cortical surface in patients throughout the demand spectrum. Previous studies have limited use of APT components in elderly or low-demand patients [1, 9, 22, 26, 28, 29]. In this study, we expanded the use of APT components to a non-age- or activity-selected population.
APT components are less expensive than other tibial implants, and the demand for TKA is projected to increase over the next 20 years . In 2008, total joint arthroplasty accounted for the single largest outlay of Center for Medicare & Medicaid Services funds for a single-diagnosis-related group [16, 20, 23]. In the face of rising costs, declining revenues, and increasing demand, there was a renewed interest in assessing APT components in TKA [3, 13–16, 19, 25, 26]. While MBT components offered increased options in terms of stems and augments for complex primary or revision TKAs, the majority of patients undergoing TKA did not require this degree of intraoperative versatility [13, 14].
We therefore determined whether (1) a non-age- and activity-selected patient population would have equivalent survivorship and complication rates (reoperation, osteolysis, tibial implant loosening) using APT and MBT tibial implants, (2) a BMI of greater than 37.5 would affect survivorship in the MBT cohort, and (3) a historical control population with APT components used in an elderly, low-demand population would maintain previously reported survivorship at long-term followup.
Using an institutional database, we identified 823 knees in 664 patients who underwent TKA by the operating surgeon (RI) between 1999 and 2007. From 1992 to 1998, the Lahey Clinic only utilized APT components (PFC®; DePuy Orthopaedics Inc, a Johnson and Johnson company, Warsaw, IN, USA) only in low-demand (Lahey Clinic Demand Category 4)  or elderly (older than 75 years) patients (APT1). The implant was a posterior-cruciate-substituting design, and the femoral component did not have lugs on the distal femur. All other patients had MBT components from the same manufacturer implanted. In late 1998, one surgeon (RI) elected to use APT components (PFC® Sigma®) for primary TKA in almost all patients, without regard for age or activity, due to concerns about the performance of MBT components. The femoral component had lugs in all cases after January 1, 1999. The APT component design did not change. After January 1, 1999, patients undergoing TKA received APT components if their BMI was less than 37.5, there were no proximal tibial bone defects, and they did not have a MBT component implanted on a contralateral knee before index surgery. Patients with a BMI of greater than 37.5, proximal tibial bone defects, and a preexisting MBT component on the contralateral side received a MBT component (MBT routine and high BMI/demand). Eighteen of the 823 patients (20 TKAs) did not have adequate 2-year clinical or radiographic followup for inclusion in this study. Twenty-six patients (28 TKA) died before the 2-year followup interval. This left 775 knees in 620 patients for study.
We included 76 APT components implanted in 59 patients from 1994 to 1999 (APT1) in this study for historical, longitudinal survivorship analysis. These patients were typically older and low-demand and were a small portion of the cohort of patients undergoing TKA before 1999 (Lahey Clinic Demand Category 3 or 4) . For comparative purposes, we only included the APT2 (TKA after 1999) (n = 558) and MBT (n = 217) components implanted from 1999 to 2007 in the main analysis (Table 1). We subdivided the MBT group further into MBT routine BMI/demand (MBT-3B) and high BMI/demand (MBT-3A) groups to standardize comparison with the APT2 cohort. There were 62 patients in the MBT-3B group who had a MBT component on the contralateral side and 155 patients in the MBT-3A group who had a BMI of greater than 37.5 or proximal tibial bone defects (three of 155, 1.9%) (Table 2).
The average BMI values differed between the two groups: the APT2 group averaged 30.8 and the MBT group averaged 33.8 (95% CI, 1.96–4.08) (Table 1). Twenty-two of 217 TKAs (9.1%) were simultaneously bilateral in the MBT group and 40 of 558 (7.3%) in the APT2 group. The primary preoperative diagnosis for all groups was osteoarthritis. Average radiographic and clinical followup was longer for the MBT group (Table 1). Comparison between the younger and heavier MBT-3A and MBT-3B groups demonstrated differences in terms of BMI (95% CI, 12.47–15.11) and age at time of surgery (95% CI, −9.44 to −4.67).
One surgeon (RI) performed all operations at one high-volume academic medical center utilizing a standardized clinical care pathway during the entire study period [15, 19]. All TKA operations were three-compartment resurfacing cemented arthroplasties in which the surgeon inserted a condylar prosthesis through a medial parapatellar arthrotomy. All implants for primary TKA were posterior-cruciate-substituting design, with either an APT or MBT tibial component. Preferred femoral alignment was 5° of valgus. Preferred tibial alignment was parallel to the ankle and/or perpendicular to the mechanical axis with 3° of posterior slope. The surgeon used Specialist One instruments (DePuy Orthopaedics Inc) as alignment and cutting guides. He placed tibial components in identical fashion: he pulse-lavaged the cut tibial surface, dried it, and manually impacted it with cement into the tibia. The surgeon also placed cement on the bone-facing surface of either component type before implantation and impaction. The details of the operative technique and postoperative management followed an identical protocol between the two groups [15, 19]. Patients met for radiographic followup at 6 weeks, 1 year, and biannually thereafter. If for any reason patients did not return for radiographic followup, we contacted them to return for radiographic followup and queried them about the TKA concerning reoperations. Sixty-one patients had confirmation of implant survival without radiographic evaluation of the knee implant within 2 years of closing this study.
One author (JT) reviewed all postoperative radiographs. Then, one of the senior authors (RI) also reviewed positive radiographic findings for confirmation. We defined osteolysis as an area of focal bone resorption identified by the absence of radiodense bony trabeculae and/or an area of reactive bone formation around a metaphyseal or epiphyseal osseous defect. We compared any such findings on followup films to the 6-week postoperative radiographs and, if the lucent area represented a change, recorded it as osteolysis, following it closely for the manifestation of symptoms or change in component position. We also evaluated radiolucencies at each followup interval and noted progressive radiolucencies of greater than 1 mm as evidence of possible loosening. We did not consider lucencies that were greater than 1 mm, but not progressive, as an indication of loosening [14, 24]. We recorded major complications, those occurrences that would impact survivorship or cause a reoperation, and reoperations for all patients. We considered instability, periprosthetic fracture, infection, loosening, synovitis, crepitus, and patella-femoral clunk as possible indications for reoperation and recorded them, as well as reoperations for instability, open reduction and internal fixation of a periprosthetic fracture, infection, loosening, synovitis, or patellar-femoral clunk.
The main outcomes we examined were failure of the tibial component and survivorship of the implant, defined by need for revision of the TKA for any reason. We recorded clinical and radiographic assessment for 217 MBT, 76 APT1, and 558 APT2 operations. We determined differences between revision, reoperation, and complication rates between the APT2 and MBT groups with a chi square analysis. We performed additional analysis between the APT2 and the MBT-3A and MBT-3B cohorts. We analyzed survivorship of the implant using Kaplan-Meier survival analysis with two end points: (1) revision of the TKA operation for any reason and (2) tibial revision; we then determined differences in rates of survivorship between groups We performed all statistical analysis using SPSS® software (v11.5; SPSS Inc, Chicago, IL, USA).
Failure of the implant for any reason, such that revision operation was necessary, occurred in six (1.1%) APT2 knees and six (2.8%) MBT knees. Survivorship analysis that compared the APT2 and MBT groups using revision operation for any reason as an end point was similar at all followup intervals (Fig. 1). Survivorship analysis that compared the APT2 and MBT groups using tibial failure as an end point was similar at all followup intervals (Fig. 2). Failure of the tibial implant was observed in four (1.8%) of the MBT patients and none of the APT patients in either cohort.
The APT1 TKAs had no isolated tibial component failures and no revisions for loosening of any component.
Survivorship analysis revealed, for revision for any reason (Fig. 2) or tibial failure (Fig. 3) the MBT-3A cohort had decreased survivorship when compared to the APT2 group or the MBT-3B group. Both of the MBT groups were heavier than the APT2 group. The MBT-3A group was younger than the APT2 cohort (Table 2). Additionally, there were four knees in the MBT-3A group with aseptic tibial loosening for which we performed revision, but none in the MBT-3B group.
Complications requiring a return to the operating room included revisions for any reason (APT2: two for deep infections, two for instability, two for stiffness; MBT: two for deep infections, four for tibial failures [all in the MBT-3A]). Supracondylar fractures (MBT: one), extensor mechanism interruptions (APT2: four; MBT: two), refractory knee stiffness for which we performed manipulation (APT2: one; MBT: one), and synovitis or clunk for which we performed arthroscopic débridement (APT2: 10; MBT: three) comprised the remainder of the complications requiring return to the operating room. These complications lead to reoperation in 21 (3.8%) patients in the APT2 group and 13 (6%) in the MBT group. There were no cases of osteolysis, progressive radiolucency, femoral or tibial implant loosening, or breakage in either APT cohort. Reoperation for tibial component failure (p < 0.001), occurring exclusively in the MBT-3A group (four of 135, 3%), was the only reoperation demonstrating a statistical difference in intervention rate between groups.
The aging demographic of the US population, a decreasing age of intervention, and the success of TKA are expected to drive annual demand for TKA from the current level of approximately 600,000 to between 3 and 4 million procedures per year by 2030 . MBT components are the most commonly used tibial implants [13, 14]. The largest expense for TKA is the cost of the implant [15, 23]. One method to reduce individual case expenditures for TKA is the use of APT components, which can provide savings of $300 to $750 per case when compared with MBT components [13–15, 23]. We therefore determined whether (1) a non-age- and activity-selected patient population had equivalent survivorship and complication rates (reoperation, osteolysis, tibial implant loosening) using APT and MBT tibial implants, (2) BMI would have an effect on tibial implant survivorship, and (3) a historical control population with APT components used in an elderly, low-demand population would maintain previously reported survivorship at long-term followup.
There are several limitations in this study. First, the retrospective nature of the study reflects selection bias based on indications at the time of surgery. Therefore, the choice of tibial implant was not randomized between groups. We made no attempt to match patients in the APT and MBT groups in terms of demographics or demand profile, which may be a source of bias or confounding variables. The younger age, larger BMI, and longer followup for the MBT group may bias the survivorship statistics since the APT group is older, smaller, and followed for shorter intervals. However, we designed this longitudinal, observational study to capture a broad cross section of primary TKA patients without special implants or surgical techniques. Second, we only included patients available for 2-year followup. However, 90% of the knees had at least 2-year followup, so we do not believe this would jeopardize the conclusions. We evaluated patients without adequate radiographic followup via telephone and checked their medical records to confirm survivorship of the implant and any reoperations. Third, we used components from a single implant manufacturer, with a single implant geometry, and reported only the experience of a single surgeon, with a standardized perioperative protocol at a high-volume, academic, joint arthroplasty setting. While this reduced variability, it may also have made the conclusions less generalizable.
Several retrospective, nonrandomized, or matched-pair analyses reinforce the survivorship equivalence between APT and MBT components in cemented TKA (Table 3) [1, 3, 5, 11, 14, 18, 26–29, 32, 34]. More recently, there have been several double-blind, randomized, controlled trials comparing TKA with APT and MBT components. Gioe et al.  randomized 316 TKAs to APT or MBT tibial implants, following them for 8 to 12 years, and noted no differences at 10 years regarding subjective knee function, clinical examination, or radiographic findings. The 10-year survivorship of TKA with APT components was 91.6% versus 88.9% for TKA with MBT components. Although the overall rate of TKA revision for any reason was similar between the groups, three MBT components underwent revision for aseptic loosening of the tibial component, while no APT component had a revision. Muller et al.  reported equivalent negligible rates of tibial component migration (< 1 mm) between cemented APT and MBT components at 24 months.
The results of our series of 851 TKAs were consistent with recent literature confirming excellent survivorship of APT implants in a broad cross section of TKA patients [1, 3, 5, 14, 18, 22, 25, 26, 29, 32, 34]. The operative surgeon had previous experience with APT components, illustrated by the clinical results of the APT1 group in this study and a previous study . His experience eliminated any learning curve issues with wide spread use of APT components.
We implanted APT components in all patients except those with tibial bone defects, BMI of greater than 37.5, or a preexisting contralateral MBT component. When we stratified the MBT group according to BMI and demand profile, we observed a relatively similar rate of TKA revision for any reason (3.8% in MBT-3A and 3.2% in MBT-3B). The MBT-3A group exhibited a 2.6% (four of 155 TKAs) revision rate for aseptic loosening of the tibial implant, different when compared to the APT2 group, which had no tibial revisions for loosening. However, patients in the MBT-3A group were heavier (BMI range, 40.4–30.8) and younger (age range, 63–70 years) than the APT2 group. The MBT-3B group was slightly heavier (BMI range, 33.5–30.8) but similarly aged when compared with the APT2 group. In the MBT-3B, we noted no tibial failures secondary to loosening.
All aseptic tibial revisions in this study occurred in the largest patients (BMI > 40) with MBT components. The largest patients remain problematic for modern TKA. No patient with a BMI of greater than 37.5 received an APT component in this study. It remains to be seen whether APT components may be more enduring in the super obese than MBT components. Two studies from the same institution [31, 34] reported only one failure and good overall results after 2 to 11 years’ followup of 242 APT components in younger, more active TKA patients. The results of this study include younger patients, and the 98.9% survivorship is similar to previous studies of APT components demonstrating survivorships of greater than 97% at more than 10 years of followup in older patients [3, 5, 14, 22, 25, 28, 29, 34]. The results of the APT1 group, with no revisions noted at an average 152 months of followup, validate the enduring performance of APT components in an age- and activity-selected patient population. It remains to be seen whether the APT2 group will match this performance in a population not selected by age or activity but limited to patients with a BMI of less than 37.5.
The cost savings of APT components (at least $300 per case) was noted as early as 1991 . Subsequently, Gioe et al.  found a savings of $675 per case, and in a recent study, the same group  noted APT components averaged $729 less than MBT components. However, the profit margin realized by hospitals for TKA continues to decline. In some hospitals, TKA is performed at a net loss [15, 16, 30, 35]. Since the early 1990s, diagnosis-related group payments for total joint arthroplasty have failed to keep pace with inflation, and the costs associated with TKA patient care have increased [15, 16, 30]. The cost of the implant is the largest expense associated with TKA [15, 16, 23, 30]. APT implants offer an opportunity to control hospital costs for TKA while maintaining quality. APT implants in this study provided at least equivalent clinical outcomes over MBT implants in patients with a BMI of less than 37.5.
APT implants are equivalent to MBT implants when surgeons consider tibial component survivorship for cemented TKA in patients with a BMI of less than 37.5. This study demonstrated APT components, when utilized without regard for activity and age, were equivalent to MBT components in patients with a BMI of less than 37.5. In this study, patients with BMI of greater than 37.5, poor proximal tibial bone, or proximal tibial bone defects did not receive APT components. It remains to be seen whether APT components may be used successfully in the larger patient group. APT utilization also can be associated with a substantial reduction in primary TKA hospital cost in this age of healthcare reform and cost control [15, 26, 30].
The authors thank John Garfi for his editorial and statistical contribution.
One author (WLH) has a product development agreement with DePuy Orthopaedics Inc, a Johnson and Johnson company (Warsaw, IN, USA). No institutional support was received for this study.
Clinical Orthopaedics and Related Research neither advocates nor endorses the use of any treatment, drug, or device. Readers are encouraged to always seek additional information, including FDA approval status, of any drug or device before clinical use.
Each author certifies that his or her institution approved the human protocol for this investigation, that all investigations were conducted in conformity with ethical principles of research, and that informed consent for participation in the study was obtained.