Total Knee Arthroplasty—Which Implant Do I Prefer for my Patients? International Roundtable Discussion

Chapter 114 Total Knee Arthroplasty—Which Implant Do I Prefer for my Patients? International Roundtable Discussion

Steve Haas, John Callaghan, Bruno Violante, Christopher Hulet, James B. Stiehl, Aaron G. Rosenberg, Hong Zhang, Thomas Thornhill

Dr. Rosenberg: Although multiple differing design philosophies have come and gone over the past several decades, in many cases no consensus has been reached as to which features are preferable. There are multiple confounding factors in the comparative evaluation of implants (e.g., function, patient satisfaction, implant longevity, complication rates), as well as the influence of tradition in the implant choices of most surgeons that make comparisons difficult. Why do debates continue on such matters as whether to preserve the posterior cruciate ligament (PCL) in total knee arthroplasty (TKA), so-called cruciate-retaining (CR), or to substitute for it with a posterior-stabilized (PS) knee, to use mobile versus fixed bearings, and to use cemented or cementless fixation?

I think it’s important to keep in mind the changes in implant design as well as surgical technique made as a result of evaluating early outcomes, and that for the most part these have led to systematic improvements in the quality and consistency of TKA. The implants we use now are different in important ways from those reported on in most long term follow-up studies. These longer term studies also reflect a significantly different patient population than that currently presenting for knee replacement.¹¹ Historically, concerns about implant longevity, as well as relatively higher complication rates, along with the fear of early failure, led many surgeons to delay arthroplasty until more advanced disease was present. These advances have given surgeons the confidence to offer knee replacements to younger and less severely affected individuals who were not previously considered sufficiently debilitated or aged to warrant the risk. In the modern era, younger age, larger size, and what were considered high physical activity demands are much more common in all of our practices. Thus, the surgeon must be aware of the fact that data in the literature may reflect both implants and patient types that are not identical to those seen in today’s practice. In addition, we are getting better at mobilizing our patients by reducing postoperative pain.⁷

It is my pleasure to have these experienced TKA surgeons gathered from around the world to share their thoughts on contemporary total knee issues with us: Steve Haas, Bruno Violante, Hong Zhang, Christopher Hulet, James Stiehl, and John Callaghan.

Let’s start by discussing the evolution of our implant choices. Dr. Callaghan, what is the current TKA of choice in your practice, how did you come to use this particular knee, and how has your thinking evolved?

Dr. Callaghan: I have been in practice for 26 years and, in the first 5, I used posterior CR designs in patients with smaller deformities and PS designs in larger deformities. As time went along, I decided that I could not really make definite cut-offs on when to go from CR to PS, so for standardization and predictability, I went to using posterior-stabilized designs in all patients. It was also during this time that there was some question about rollback kinematics in CR designs. In addition, with the difficult patient population that I operate on at a state university, I believe it is essential to perform one operation and see how that works in the long term.

We now have 15-year survivorship of the PFC knee (DePuy, Warsaw, Ind) that I have been using and, other than a few infections, most of which were hematogenous, the only failures were related to polyethylene wear. This particular implant had an issue with the polyethylene used in the early to mid-1900s and we and others have reported this. Hence, in summary, I use a posterior-stabilized design because it allows me to feel comfortable that I have provided the best operation for patients with little as well as maximum deformity. I think the survivorship is similar to cruciate-retaining designs of the same brand. I do, however, think that the literature is clouded because of the variance in polyethylenes used in the late 1980s and 1990s. The implant I use allows for fixed- or mobile-bearing options and I have gone to using the mobile-bearing design for 60% of knees.

Dr. Rosenberg: Dr. Stiehl, I know your experience is quite similar.

Dr. Stiehl: My initial experience for over 15 years was with posterior cruciate–retaining implants. Back in the 1980s, I saw numerous implant failures caused by excessive wear, patellar failure, and loss of fixation. I was then heavily influenced by fluoroscopic kinematic studies that I performed beginning in 1992 and subsequently published extensively. My early experience was with a flat-on-flat condylar design that used radiation-sterilized polyethylene.

As I learned more about the deleterious effects of sliding and ploughing wear on polyethylene surfaces, kinematic issues such as poor femoral condylar rollback in many patients, and femoral condylar liftoff with activity, I began to use a posterior cruciate–sacrificing mobile-bearing implant. These implants seemed immune to these problems from long-term studies, and typically offered higher ranges of motion with activity.

Following my early kinematic studies, I sought methods, such as flexion ligament balancing, that appeared to limit the problems of instability and wear that were often seen with posterior cruciate–retaining, fixed-bearing knees implanted with measured resection methods. I remember a brief conversation with one of the prominent originators of posterior cruciate–retaining implants when I asked if he had seen significant laxity in the lateral compartment in deep flexion. He explained that he witnessed that phenomenon commonly but never recognized a clinical problem with it. He may be correct in his answer to this day, but the kinematics that I could demonstrate on weight-bearing fluoroscopy were truly scary (Fig. 114-1).

Figure 114-1 Lateral fluoroscopic image of a total knee in maximum flexion. Note the degree to which one of the femoral condyles has lifted off from the tibia.

Dr. Rosenberg: Dr. Zhang, isn’t your experience affected by implant availability?

Dr. Zhang: Because there are almost no CR implants available in China, almost all implants I have used are PS fixed or mobile bearing. However, I have had the chance to use different knee implants from different companies. I have used Innex (Zimmer, Winterthur, Switzerland), Insall-Burstein II (IB II, Zimmer, Warsaw, Ind), NexGen LPS and LPS-flex II (Zimmer), Scorpio and NRG (Stryker, Mahwah, NJ), LCS, PFC, PFC-RP, and PFC-RPF (DePuy, Warsaw Ind), and Genesis-II (Smith & Nephew, Memphis, Tenn). In China knee arthroplasty, surgery has only become popular in the past 5 years; in 2001, our department did 55 total knees, but last year we did almost 1000. Although I can’t report on longevity and survivorship, in my experience different implants have shown different function and outcomes.

Dr. Rosenberg: Dr. Violante, I remember going to Italy 20 years ago and was struck by the state of the art in TKA, clearly somewhat behind the United States. But, what a difference 2 decades have made!

Dr. Violante: The story of the TKA in Europe is quite different than that in North America. Despite the popularity of total hip arthroplasty in the 1980s and 1990s, there was a lack of acceptance of total knee arthroplasty. During that period, proximal tibial osteotomy was the standard treatment for knee osteoarthritis. The most common indications for a total knee were failed osteotomies or real disasters in terms of deformity and function.

In the early 1990s, I started with a CR design (Genesis, Smith & Nephew), which required managing the PCL, which might be loose, tight, or appropriately balanced. Loose was more desirable in the presence of an optimal flexion space, but a tight PCL led to bad outcomes with lack of flexion, pain, and early tibial loosening.

Because of this early experience and looking more to the anatomy of the knee, I shifted toward a PS design (IB II, Zimmer) to eliminate the PCL management problem and to work on a more easily reproduced surgical technique and more reliable outcomes. So, from the early 1990s I used the IB II and in 1997, switched to the LPS NexGen Legacy (Zimmer), also designed by John Insall.

Dr. Rosenberg: Dr. Violante, like Dr. Callaghan and Dr. Stiehl, made the move from CR to PS. Dr. Hulet, what has been your experience?

Dr. Hulet: In the early 1980s and 1990s, we used the IB II knee system and the Kinemax knee. But, in 1994, we developed a new prosthesis in France, which was called the Jade TR prosthesis (Howmedica [Stryker Orthopaedics], Mahwah, NJ). It was a complete system with the PCL either sacrificed or retained. There was also a revision system with stems and wedges. We do not resurface the patella in our practice. The trochlear design accommodates the shape of the patella and the needed posterior translation during knee flexion.

Dr. Rosenberg: Dr. Hulet, I think this is an important point that most knee systems have both CR and PS implants but, Dr. Haas, you have some experience with bicruciate-substituting knee design.

Dr. Haas: For many years, I used the Genesis II knee system (Smith & Nephew). This system provided excellent results. I have used primarily the high-flexion version of the Genesis II posterior-stabilized knee. This provided excellent clinical results. However, as more of my patients were doing higher levels of activity, such as tennis, skiing, and competitive golf, I wanted to have a knee arthroplasty that would provide normal rotational motions of the knee in addition to high flexion. I currently choose to use the Journey Knee (Smith & Nephew) as my main implant of choice. I use this system because my practice involves a large number of younger active patients. Many of the features of the Genesis II system were incorporated into the Journey Knee and this system restores near-normal kinematics and allows for a high level of function.

Dr. Rosenberg: My own experience has been primarily with cruciate-retaining arthroplasty. Early on I used the Miller-Galante I (Zimmer) with both cemented and cementless fixation but found that although my cemented implants almost never loosened, my cementless tibias failed to obtain ingrowth at about a 3% rate.¹⁷ I subsequently began cementing everyone by the late 1980s and, at longer term follow-up, the cementless cohort had loosenings of the tibial component that the cemented components just did not have.⁴ The implant had a unique tibial component that featured four small pegs and had no conventional keel. Although our initial experience demonstrated a high rate of metal-backed patellar failures,¹⁶ wear, lysis, and component loosening have ever been an issue, even in our long-term follow-up of that implant.⁵ The lessons learned from that experience led to the MG II (Zimmer), which featured a deepened trochlear sulcus, all-polythylene patella of improved design, and minor modifications of the articular surface configuration of the polyethylene in the sagittal plane.

We have not yet talked much about fixed versus mobile bearings. Dr. Stiehl, I know that you have done a lot of research and given a lot of thought to this.

Dr. Stiehl: I would state categorically that I now use mobile bearings in all primary patients and in virtually all settings. This choice is then divided into two groups. For patients who I consider at high risk for long-term wear problems and loss of fixation, I will perform the classic LCS (DePuy), posterior cruciate–sacrificing or rotating platform mobile-bearing, usually without resurfacing of the patella. Specifically, those patients are under 65, desiring a normal functional outcome, with the ability to perform many sports including tennis, golf, snow skiing, horseback riding, power walking, and moderate lifting in occupational settings. The second group consists of patients who are at lower risk for long-term failure including most patients over 65 or those with severe bone loss or preoperative deformities with loss of motion. These patients generally desire to decrease pain, improve motion, and kneel without restriction. This group receives the Nexgen LPS High Flex (Zimmer), a mobile-bearing posterior cruciate–stabilized design, often without resurfacing of the patella. These patients are told that they may accomplish very high range of motion and have a higher likelihood of feeling normal on stairs than with other choices.

Dr. Rosenberg: I think the whole issue of mobile bearings is an interesting one, but my impression is that there are no published data, at least at intermediate follow-up, that show lower revision rates, better function, or longevity. So, where are the advantages?

Dr. Stiehl: I believe that mobile bearings offer two very notable features that enhance the potential outcome. First, I am convinced that the bearing surfaces have better wear performance with higher surface contact. The high contact forces seen with limited surface contact were indigenous to flat-on-flat designs. This idea is supported by numerous wear studies and observations made from large numbers of implant retrievals.

Second, the problem of tibial tray malrotation on implantation is eliminated with mobile bearings.²² The knee has the normal function of tibial internal rotation that occurs with flexion. This can be distorted or eliminated in the diseased state. Combined with anatomic variation, there are few reliable landmarks that predict the position of the tibial base plate position in all comers. Add the challenge of placing a tibial tray correctly through smaller surgical incisions that have become the fashion of the day, and you have a problem that cannot be eliminated, even with the experiences of several thousand total joint procedures.

Mobile-bearing polyethylene adds an important safety factor, in my view. I recently published my retrospective experience evaluating preoperative and postoperative tibial rotation measured with computer navigation, and found that fixed trays tended to be more internally placed and had less internal rotation with flexion compared with the mobile-bearing knees.²²

Dr. Rosenberg: Dr. Zhang, is the mobile bearing harder to implant?

Dr. Zhang: Whether or not an implant is easy to put in depends on the surgeon’s surgical experience and the implantation equipment. Most PS prostheses are easy to put in if the manufacturer’s instructions and surgical steps are followed. However, it is a little more complicated to achieve a good balance of the soft tissue and prosthesis stability for mobile- bearing implantation, such as LCS or Innex. The learning curve is longer for mobile-bearing TKA.

Dr. Stiehl: Mobile-bearing implants do have the notorious clinical experience of bearing spin-out or dislocation if the prosthetic ligament balancing is not adequate (Fig. 114-2). Both implants that I use are particularly stable if a reasonable surgical technique is performed and I can confess not one single bearing dislocation with 15 years of clinical experience. Early on, I avoided mobile bearings in valgus deformities over 20 degrees and, in some severe varus deformities, deferred to a fixed-bearing, posterior-stabilized device, as would be done in total knee revision arthroplasties. Today, I would use the LCS (DePuy) mobile rotating platform or the LPS High Flex mobile (Zimmer) with impunity in any primary total knee setting. The LPS actually has a spin-out stop to limit rotation over 20 degrees in one plane and might be safer for the community surgeon.

Figure 114-2 Lateral radiograph of a mobile-bearing total knee with a dislocated bearing. Note the lateral view of the femoral component abutting the dislocated polyethylene just posterior to it.

I would stongly discourage surgeons from using a measured resection method with a mobile-bearing device. Although measured resection may work in straight forward cases with little deformity, it will fail if there is poor ligament balancing after correction of marked deformity or bone defect and the gaps are not appropriately matched. Mobile-bearing implants are not for the occasional knee surgeon, but more appropriate for the engaged surgeon who is willing to learn from experienced technicians, as I did with my first few.

Dr. Rosenberg: Dr. Callaghan, are you wedded to fixed or mobile bearings in the majority of your primary knees?

Dr. Callaghan: One of the specific advantages of the current implant I use is the fact that I can use a fixed-bearing or mobile-bearing option. In addition, the fixed-bearing option has a polished tray, which I believe has the potential benefit of decreasing backside wear. However, this has never been proven in the literature. It also allows for using a more constrained component. The mobile-bearing design allows for more conformity of the polyethylene to the femoral component, with potential for decreasing wear. This is why I use it in the younger patients instead of the fixed-bearing design, which is not as conforming.

Being as it allows for both fixed bearing as well as mobile bearing the implant I use allows for much versatility. In patients with large valgus deformities, I am a little more apt to use the fixed-bearing design as well as in cases that have tremendous range of motion—that is, more than 140 degrees as I would imagine the patient’s femoral runners run past the tibial inserts which are only designed to obtain 120 degrees of flexion.

Dr. Rosenberg: What about bearing spin-out? Dr. Callaghan, have you had any experience with this phenomenon?

Dr. Callaghan: Although I have not had a bearing spin-out with a PS component, it still is a concern with mobile-bearing knees. The capability of the system that I use allows the surgeon to use constraint, whether it be a fixed or mobile bearing. This helps me in the consideration of which of these two implants I use. In a younger patient, if I am going to use constraint, I will tend to use a mobile-bearing design, which I hope will lead to less issues with loosening of the implants.

I have personally reviewed minimum 20-year follow-ups of mobile-bearing designs, performed with polyethylene gamma-irradiated in air, and have seen no loosening over that period. Although the average age of the patients at the time of surgery was 70 years old, I have gained more confidence in believing that these devices have potential for greater long-term durability in the most kinematically challenging patients.

Dr. Rosenberg: Dr. Violante, please tell us what you believe to be the specific advantages of the current implant you use and how its design features compare with previous implants you used.

Dr. Violante: With the exception of the post-cam mechanism, the difference between the IB II (Zimmer) and the LPS NexGen Legacy (Zimmer) is substantial. With the IB II, patellar clunk, caused by the soft tissue impingement between the infrapatellar scar and the femoral trochlea–box junction during flexion, was seen occasionally.¹⁵ With the LPS NexGen, asymmetrical femoral components were introduced and, along with this, changes in the patellofemoral geometry with a deeper and prolonged trochlear groove, obliquely oriented by 7 degrees, and smoother transition of the distal femoral condylar geometry to achieve better patellar tracking, patellofemoral kinematics, and reduced patellar clunk.¹⁰ More femoral component sizes and more accurate shaping of the native femur also helped reduce patellofemoral overstuffing.

The polyethylene tibial fixation was subject to micromovements, with subsequent backside polyethylene wear as a cause of tibia lysis in the IB II. In the NexGen Legacy, the polyethylene was compression-molded and gamma-sterilized in inert gas and the polyethylene locking mechanism was reinforced with a posterior dovetail mechanism and peripheral anterior compression.

The cam and post mechanism was also optimized to reduce the risk of posterior dislocation and was moved posteriorly so that the cam engages the post at 75 degrees. High-flexion, specific design geometry variations were made so the patient could safely bend up to 155 degrees.^19,²⁰

Dr. Rosenberg: I was involved with the development of the same system (NexGen) but focused on the cruciate-retaining side. We were building on the improved results we had experienced going from the Miller-Galante I to the Miller-Galante II (Zimmer) and after having developed a lot of confidence in PCL retention at that point in time. The results with PCL retention in the MG II appeared to be much better than those reported for multiple other devices. Thus, the NexGen CR retained the kinematic advantages we thought were inherent in the shape of the femoral and tibial articulating surfaces but added a lot of attractive design features related to component sizing, matching, and deepening of the trochlea to improve patellofemoral kinematics, and to other features.

Over time, we made modifications to the implant by creating the Flex design, which allowed for more precise tensioning of the PCL by introducing a larger number of anteroposterior sized implants, and then the Gender design, which allowed for better overall femoral component size matching, medially and laterally, to the individual patient. However, it was particularly gratifying when the kinematic studies that Dr. Stiehl had helped develop demonstrated that our implant design actually produced normal-appearing femoral rollback and minimized liftoff, in comparison to all the other CR designs that Dennis and colleagues ¹² had tested.

Dr. Stiehl, what are the distinguishing features of your current implant choice?

Dr. Stiehl: My choice of the DePuy LCS Mobile and the NexGen LPS High Flex Mobile relies on two specific advantages that I perceive to have guided my overall technical approach. I believe that the mobile feature allows for dramatically increased articulation surface areas over older flat-on-flat designs that arose in my early years and were developed with posterior cruciate–retaining methods. The original predicted advantage of mobile-bearing devices was decreased wear, which was based on diminished contact surface forces that may be exaggerated by sliding and ploughing point contact. What I find ironic is that my current choices were conceived and perfected well before these interim devices, and are now emerging as contemporary replacements of those improved options.

The second important advantage is the ability to correct errors with tibial tray placement in terms of rotational position because the device self-corrects with final ligament tension. I have studied this problem extensively with computer navigation and believe that the tibial tray position for a given patient is highly variable and can be affected by anatomic variation, degree of arthritic deformity, and surgical technique.

Dr. Rosenberg: Dr. Zhang, you have used a lot of different implants. What are your particular findings?

Dr. Zhang: I am very interested in bone conservation. The Scorpio, NRG, LCS, and Innex remove very little bone when shaping the femoral intracondylar box whereas the PFC and PFC-RP require removal of a bigger piece of bone. The difference is significant.

Dr. Hulet: The Jade system TKA can be used with PS constraint or PCL retention. It features a molded polyethylene (PE) insert with curved geometry for the femoral component. The tibial component theoretically allows for normal femoral rollback. The geometry and dimensions of the femoral component reproduce the normal femur based on a medial pivot design with a single radius curve for the femoral condyle. The geometry of the femoral component should restore normal kinematics and also reproduce normal ligament tension in both extension and flexion. The tibial tray is flat and the PE insert needs to be compatible with femoral rollback (curved-on-curved). The PE insert in both designs is symmetrical, such as the design of the femoral condyles for both the PCL-substituting or PCL-retaining version. We used cement fixation for both components.

Dr. Haas: The Journey system concept is fundamentally a resurfacing-type arthroplasty with a completely left- and right-sided anatomically shaped femur, with medial distal and posterior condyles 2 mm thicker than the lateral distal and posterior condyles. The anterior flange is also thinner medially and thicker laterally because the implant is placed in external rotation. The tibia is similarly asymmetrical to mimic the thickness and shape of the cut bone and reproduce the natural shape of the tibia. These features allow more consistent joint line restoration and result in less tension in the patellofemoral articulation.

The tibial base plate is also asymmetrical. It allows increased tibial coverage and leads to more accurate rotational alignment. The anatomic shape of the base plate also helps avoid posterolateral overhang, which makes minimally invasive insertion easier. Although there is high coronal conformity medially and laterally, the sagittal conformity is greater on the medial tibial plateau compared to the lateral tibial plateau. This differential geometry helps promote more natural rotational motions with a screw-home mechanism in extension and lateral rollback in flexion.

Dr. Rosenberg: Dr. Violante, what sort of clinical results are available on the Nex Gen LPS?

Dr. Violante: There are as yet no long-term published follow-up data on function, outcomes, longevity, and radiographs but at up to 6-year follow the LPS NexGen showed better results than the IB II. In an early series of 279 knees from August 1997 to December 1998, all made by senior surgeons (Drs. Insall, Scott, and Scuderi), with a mean follow-up of 48 months, the mean Knee Society score for 238 knees was 96 points at the latest follow-up, with no case of patellar clunk, maltracking, or posterior dislocation. There was no radiographic evidence of loosening or osteolysis and no revisions performed or recommended for loosening, osteolysis, instability, or PE wear. Some radiographic evidence of radiolucent lines no larger than 1 mm located at the tibial component in zone 1 was observed in about 4% of cases not associated with any clinical symptoms.¹⁰

Dr. Rosenberg: That certainly seems acceptable! Our own results with the NexGen CR at 10-year minimum follow-up ¹⁸ have been published on a cohort of 161 patients with 179 cemented TKAs in whom the patella was resurfaced; we had originally reported on them at 5 years.⁶ Of these, 40 patients had died and 8 were lost, leaving 113 patients with 126 knees. Survivorship at a minimum of 10 years, using revision for any reason, was 98% and, with revision for aseptic loosening, it was 100%. Three knees were revised—one each for infection, periprosthetic fracture, and arthrofibrosis. There were no r-operations for patellar component problems or wear and neither lysis nor aseptic loosening were seen. Interestingly, both the Australian and Swedish arthroplasty registry seem to confirm these findings, noting the lowest cumulative revision rates for this implant in both countries.

Dr. Hulet, what about the Jade system?

Dr. Hulet: We began to use this design in 1996 and recently at the last Effort (European Federation) Meeting in Nice, in 2008; we have published the results of our first 122 cases with 10-year follow-up.¹ These were cemented PCL-sparing, with the patella resurfaced in only eight cases. No reoperation was performed for patellar pain. There were no early complications. The mean Knee Society score was 90 points, with mean motion of 109 degrees. The mean Knee Society function score was 85 points, with survival of 95% at 9.5 years (Kaplan-Meier method), considering all reasons for removal, and 99.4% for removal for aseptic loosening.

Dr. Rosenberg: Dr. Haas, what sort of results have been reported with the Journey knee?

Dr. Haas: Victor and Ries have recently reported excellent clinical results in the Journey knee, with a 30-degree improvement in average range of motion. In addition, in vivo kinematic evaluations have shown consistent and reproducible flexion and rotational motions. Komistek has performed several in vivo two- and three-dimensional fluoroscopic studies and has shown that patients with Journey BCS TKA “consistently experience kinematic patterns similar to that of the normal knee.” Studies by Komestek and Catani have shown a normal pattern of knee rotation without significant paradoxical anterior sliding commonly seen in TKA. Greenwald has performed a kinematic analysis of several contemporary TKA designs and found that “ the Journey most closely replicated the healthy un-operated knee kinematics.”

Dr. Rosenberg: I am consistently intrigued by reports of TKA without patellar resurfacing; to me this seems more like a hemiarthroplasty at the patellofemoral joint. We have a great deal of experience with hemiarthroplasty at other joints and the results are consistently inferior to what Charnley has demonstrated at the hip and has also been found in most other joints. That is, replacing both sides of the articulation generally results in better pain relief and function than replacing only one side of the joint and letting it articulate with native cartilage or what remains of it.

Dr. Stiehl: I believe that mobile-bearing devices optimize the patellofemoral articulation. This is a controversial problem and there are numerous conflicting ideas about how to manage the patella in total knees. In favor of the mobile-bearing method is that flexion gap balance must be very precise and anatomic, close to the normal 2 to 3 mm. I believe that this will optimize the patellar tracking as the knee goes into deep flexion.²¹ A number of authors, including myself, have minimized the problem of abnormal femoral rotation that could occur by using flexion-based tensor instruments (Figs. 114-3 and 114-4).