High Rates of Aseptic Loosening in Modern Posterior-stabilized Femoral Components From a Single Manufacturer

Abstract

Background

Total knee arthroplasty (TKA) is one of the most commonly performed surgical procedures in the United States. There are concerns and reports of early aseptic loosening due to debonding at the cement-implant interface in a specific posterior-stabilized femoral design. This study describes failure at the cement-implant interface with 2 femoral implant designs produced by a single manufacturer.

Methods

This is a review of failed primary total knee arthroplasties performed with Optetrak Logic and Truliant posterior-stabilized femoral components between 2010 and 2020. Cases with revision surgery performed for femoral component loosening were reviewed.

Results

Seventeen of 896 knees were revised for femoral component loosening. The mean time from index arthroplasty to revision was 73 months. Patients presented with knee pain and recurrent aseptic effusions. More than half of patients had no evidence of component loosening on plain radiographs. Bone scans were positive in 77% of patients with loosening. Femoral components were easily disimpacted from the cement mantle at the time of revision surgery. Patients were treated with femoral only revision or femoral and tibial revision. Two complications occurred in this series.

Conclusions

Providers should maintain vigilance in the surveillance of patients with these implants. Clinicians should have a low threshold for additional diagnostic testing if these patients develop pain and/or recurrent effusions. Isolated femoral or complete revision both seem to be acceptable treatment strategies for this failure mechanism.

Introduction

Total knee arthroplasty (TKA) is one of the most commonly performed surgical procedures in the United States [1], with 1.28 million TKAs estimated to be performed annually by 2030 [2]. Although infection remains the most common reason for revision TKA, mechanical loosening remains a close second [3]. There have been concerns and reports of early aseptic loosening and implant failure among a specific posterior-stabilized (PS) femoral design from a single manufacturer (Fig. 1) [4]. This specific femoral implant, based on previous designs, was introduced in 2009 and featured a bone preserving intercondylar box geometry for PS femurs as well as changes to allow a higher degree of flexion [5]. Recent data is concerning for early aseptic loosening due to debonding at the cement-implant interface of this specific implant [4]. In this report, we describe the clinical presentation and treatment of 17 patients who underwent revision TKA for femoral aseptic loosening with implants from a single manufacturer. We hypothesize that there is a substantial failure risk of debonding at the cement-implant interface with 2 femoral implant designs produced by a single manufacturer.

Figure 1.

AP and lateral radiographs of the knee with Exactech components.

Material and methods

This is a single-institution review of failed primary TKAs performed with Optetrak Logic and Truliant PS femoral components (Exactech, Inc., Gainesville, FL). Institutional review board approval was obtained prior to proceeding with this study. Patients who underwent primary TKA with these implants between 2010 and 2020 and had a subsequent revision for aseptic loosening were identified from a maintained institutional database. Revisions for diagnoses other than aseptic loosening were excluded. Demographic and clinical data were recorded, and operative records were obtained and reviewed for those who underwent revision.

Primary and revision TKA was performed by the senior author (J. C. M.) on all patients included in this study. The technique for primary TKA did not significantly change during the study period. Primary TKA was performed with a tourniquet through a midline incision with a standard medial parapatellar arthrotomy utilizing a gap-balancing technique. High viscosity cement (HVC) was used in all cases after July 2011, representing >90% of the cement used in this study. Cement was pressurized into the bony cut surface and applied to the backside of the implant at the time of implantation. Cement was allowed to harden in full extension with a trial insert in place prior to the placement of the final polythene insert.

Infection was ruled out in all patients prior to revision surgery. Revision TKA was performed through the prior midline incision with a medial parapatellar arthrotomy. No extensile approaches were required in this series. After thorough synovectomy, the polyethylene insert was removed, and the femur was disimpacted by impacting the anterior flange of the component. The tibia was checked and revised if found to be loose or if there was a size mismatch after femoral revision. Isolated femoral revisions were performed with the Optetrak revision knee system (Exactech Inc., Gainesville, FL), and full revisions were performed either with the Optetrak revision knee system or the Attune Revision Knee System (DePuy Orthopaedics, Warsaw, IN). Metaphyseal bone loss was addressed with sleeves or cones as needed.

Results

Eight hundred and ninety-six TKAs were performed with logic (n = 432, 48%) and Truliant (n = 464, 52%) femoral components in 671 patients from 2010-2020. In total, 17 knees (1.9%) were revised for femoral aseptic loosening (Table 1). Logic components had a higher rate of revision when compared to Truliant components (3% vs 0.9%). The mean time from index arthroplasty to revision was 73 months.

Table 1.

Summary of cases.

Case	Sex	Age (y)	BMI (kg/m2)	Time to revision (mo)	Primary implant	Revision details
1	Male	60	32	68	Logic PS	Femur only
2	Male	70	28	103	Logic PS	Femur and tibia
3	Female	56	25	34	Truliant PS	Femur, tibia, patella
4	Male	66	32	64	Logic PS	Femur and tibia
5	Male	84	35	77	Logic PS	Femur, tibia, patella
6	Male	77	35	30	Truliant PS	Femur only
7	Male	72	31	16	Truliant PS	Femur only
8	Male	65	32	27	Truliant PS	Femur only
9	Female	73	40	103	Logic PS	Femur only
10	Female	69	33	90	Logic PS	Femur only
11	Male	71	39	68	Logic PS	Femur only
12	Female	81	28	96	Logic PS	Femur and tibia
13	Male	70	32	84	Logic PS	Femur and patella
14	Female	69	29	139	Logic PS	Femur only
15	Male	53	35	73	Logic PS	Femur only
16	Male	57	41	79	Logic PS	Femur and tibia
17	Female	55	48	92	Logic PS	Femur only

BMI, body mass index.

Patients generally presented with mild knee pain and persistent, large aseptic effusions. One patient presented with a large recurrent hemarthrosis. Radiographs were obtained in all patients during the workup. Nine patients (56%) had no radiolucent lines or evidence of implant migration on plain radiographs. One patient had subtle radiolucent lines under the femoral component, and 1 patient had subtle flexion of the femoral component over sequential radiographs. Three patients had osteolysis behind the femoral component, and 2 patients had obvious femoral component loosening. Triple-phase bone scans were obtained in 13 patients; 10 were considered positive, 2 negative, and 1 inconclusive.

At the time of revision surgery, most femoral components were not grossly loose, but were easily disimpacted from the cement mantle with impaction against the edge of the anterior flange, leaving an intact cement mantle (Fig. 2). Two patients had significant osteolysis beneath the cement mantle. Two patients had concurrent aseptic loosening of the patella, and 1 had delamination of the patellar polyethylene. One patient (patient 16) underwent polyethylene liner exchange, synovectomy, and bone grafting of femoral osteolysis for suspected polyethylene wear 26 months prior to revision for loosening.

Figure 2.

Intraoperative photograph showing intact cement mantle after femoral component removal.

Patients in this series were treated with femoral only revision (n = 11) or femoral and tibial revision (n = 6). Three patients also required patellar component revision. All patients had relief of prerevision weight-bearing pain after revision TKA. One patient required hematoma evacuation 1-week post-revision (patient 2), and 1 patient required tibial component revision for aseptic loosening 30 months after femoral component revision (patient 17). No other known complications occurred.

Discussion

This study supports previously reported findings of increased failure rates of a specific implant from a single manufacturer due to aseptic loosening. Additionally, we highlight the potential for a similar failure mechanism in a second implant from the same manufacturer. Malahias et al. previously reported a 0.4% revision rate due to femoral sided aseptic loosening among Logic PS TKAs. While this was not statistically different from the revision rate among nonlogic PS TKAs included in their study, the time to revision as well as the presence and severity of backside burnishing were statistically higher for logic femoral components when compared to other PS TKA designs [4]. The revision rate of logic components in our cohort is 3%, which is much higher than the reported rate by Malahias et al. Additionally, we report a 0.9% revision rate of Truliant implants for aseptic loosening.

The mechanism of debonding at the implant-cement interface likely begins with the design of these femoral implants. A biomechanical study would be required to completely elucidate the mechanism behind debonding. However, we hypothesize that the backside roughness of the femoral component plays a role in the high rates of aseptic loosening. Running changes were made to the backside roughness of the Truliant femoral component around June 2021 (Figure 3, Figure 4). Prior to this, the backside of the femoral component was smoother, which may have contributed to debonding.

Figure 3.

Photograph of the femoral component backside after extraction.

Figure 4.

Photograph of the backside of the Truliant femoral implant after changes to the backside to increase roughness.

The overall use of HVC in primary elective TKA is increasing [6], and there is concern that its increased use may lead to higher rates of aseptic loosening. Greater than 90% of knees in this study were cemented with HVC, which could be an alternative reason for the increased rate of failures seen in this study. Multiple studies have demonstrated higher rates of aseptic loosening with the use of HVC. Buller et al. demonstrated a higher rate of aseptic loosening among a single center database [7]. Kopinski et al. reported a series of cases of tibial component loosening in knees cemented with HVC [8]. Another single center database study demonstrated higher rates of overall revision with the use of HVC; however, there was no difference in rates of aseptic loosening in knees where HVC was utilized [9].

We do not believe the use of HVC to be a factor in the higher rates of loosening seen in this study. Additional studies in the literature demonstrate no increased rates of aseptic loosening when HVC is used [10,11] and there was only 1 case of tibial component loosening in this study.

The workup for a painful TKA is a complex process. The workup for patients in this series included a thorough history, physical exam, and radiographs of the affected joint. All patients in this series reported pain. Some patients presented with classical start-up pain, which has been associated with component loosening [12], while others reported vague persistent pain. Palpable effusions were noted in all patients. As previously noted, most patients had no overt signs of loosening or implant changes on plain radiographs, which is not surprising, as radiographs for aseptic loosening are not sensitive or specific [13]. Infection is always a concern in any patient with a painful TKA. In this series, infection was ruled out with routine laboratory analysis and joint aspiration. Advanced imaging techniques, such as nuclear studies, are reported to be sensitive but not specific. In this series, 77% of bone scans were considered positive [12]. However, it is important to note that bone scans can detect normal inflammatory reactions up to 2 years after TKA, which can produce false-positive results during this time period [12]. Although not utilized in this series of patients, a diagnostic anesthetic-only injection could be considered. These injections can be helpful with the work-up of a painful TKA when the diagnosis remains unclear [14].

It is important to note that the diagnosis of aseptic loosening is complex, as previously described. However, it should be noted that other coexisting pathologies may be present, and the surgeon must be aware of all potential failure mechanisms associated with the implant under investigation. One patient in this series underwent revision surgery for polyethylene wear with subsequent revision for aseptic loosening. Physicians should be aware that this company has recalled polyethylene inserts [15] associated with the femoral implants described in this report. Surgeons planning an isolated polyethylene revision should closely evaluate the distal femoral component at the time of the revision.

One limitation is the retrospective design of this study. Because of this study design, we are unable to provide any findings related to the implants themselves. Additionally, revision surgery was used as the end point for this study. Because of this, the rate of loosening reported may underestimate the true prevalence of loosening, as patients who have not returned for follow-up or are awaiting revision surgery are not captured in the data. We do, however, have a database of patients awaiting revision surgery, patients being monitored with suspected loosening, and patients undergoing evaluation for suspected loosening. To better define the problem, this would add more than 30 additional patients to this study. These patients were not included in the analysis because debonding can only be confirmed at revision surgery.

Another perceived limitation is that all primary TKAs were performed by a single surgeon, which may indicate loosening due to poor surgical technique or surgeon-related factors. However, the debonding noted in this study is likely not associated with surgeon-related factors. Femoral debonding has not been identified with any substantial frequency among other implant designs used by the senior author with the same surgical technique.

Conclusions

We report 17 cases requiring revision TKA for aseptic loosening among femoral implants from a single manufacturer. Most patients presented with pain and recurrent effusions. Most plain radiographs did not have any signs of loosening, and bone scans were positive in 77% of patients. Additional studies are needed to further evaluate these implant designs and the associated femoral debonding. We recommend vigilance in the surveillance of patients with these implants. Clinicians should have a low threshold for additional diagnostic testing if these patients develop pain and/or recurrent effusions. Isolated femoral or complete revision both seem to be acceptable treatment strategies for this failure mechanism.

Funding

This research was supported (in whole or in part) by HCA Healthcare and/or an HCA Healthcare affiliated entity. The views expressed in this publication represent those of the authors and do not necessarily represent the official views of HCA Healthcare or any of its affiliated entities.

Conflicts of interest

The authors declare there are no conflicts of interest.

For full disclosure statements refer to https://doi.org/10.1016/j.artd.2023.101132.