Introduction

The gold standard in the treatment of knee osteoarthritis remains the total knee arthroplasty. Each year over 750,000 total knee arthroplasties are performed in the USA alone [1]. Unfortunately, patient satisfaction has varied between 82% in younger patients and 91% in older patients [2••, 3]. Stiffness and poor range of motion are key driving forces behind poor satisfaction rates [3]. As a result, several types of tibial polyethylene inserts have been developed with an emphasis on recreating native knee biomechanics [4, 5•, 6, 7••, 8, 9]. The goal of this review is to summarize the studies that have been done on the biomechanics of the classic posterior stabilized and cruciate retaining inserts as well as the more modern inserts. A brief review of outcomes will also be discussed.

Biomechanics of the Native Knee

The knee is a modified hinge joint which has six total degrees of freedom, three translational and three rotational. Translational degrees of freedom are; anterior-posterior, medial-lateral, and inferior-superior [5, 10]. Rotational degrees of freedom are; flexion-extension, internal-external rotation, and adduction-abduction [11•].

Through range of motion of the knee, the medial and lateral condyles experience different amounts of translation in various activities. Fluoroscopic studies have shown the medial femoral condyle moves −4.7 mm while the lateral condyle moves −11.8 mm during normal gait [12••]. However, during deep bending, the medial condyle moves −4.5 mm while the lateral condyle moves −16.3 mm [12••]. This results in internal rotation of the tibia relative to the distal femur as well as posterior translation of the femur relative to the tibia. The knee typically has the ability to achieve 140–160° of flexion [13] (Table 2). The difference in movement of the lateral condyle versus the medial condyle while the knee is bending is termed “femoral roll-back.” During the first 15 to 20°, we see primarily rotation, while beyond 20°, we get primarily posterior translation of the femur. The opposite of this is termed the “screw home mechanism,” where the tibia externally rotates and the femur translates anteriorly [11].

Debate remains on whether or not to sacrifice the posterior cruciate ligament (PCL) in total knee arthroplasty (TKA) [4, 5, 14, 15•, 16]. The primary role of the PCL in the knee is to prevent excessive posterior translation of the tibia [17•, 18]. Another function of the PCL is to prevent excessive tibial internal rotation between 90° and 120° of flexion [19]. Through a cadaveric analysis, it was found that an intact and functional PCL allows for appropriate levels of femoral rollback and prevents excessive posterior translation of the tibia relative to the femur [17]. This supports the belief that retention of the PCL allows for more physiologic knee kinematics after TKA specifically allowing the screw home mechanism to occur in the final 20 degrees of extension [20].

Posterior Stabilized Implant Biomechanics

Posterior stabilized (PS) TKA implants utilize a tibial post to function as the PCL, which is sacrificed during joint preparation (Table 1). Compared with the native knee, conventional PS-TKA demonstrates less overall motion in both normal gait and with deep flexion. However, the most notable difference is that, on average, the posterior stabilized cam-post design demonstrates paradoxical anterior translation of the lateral condyle relative to the tibia during normal gait. This is opposite that of the native knee which demonstrates posterior translation of the lateral condyle with femoral rollback during gait [13, 21•].

Table 1 Summary of polyethylene inserts
Full size table

However, during deep flexion, PS-TKA demonstrated normal axial rotation of 10.4° with the tibia rotating internally with progressive knee flexion and posterior femoral rollback of −7.1 mm [21•] (Table 2).

Table 2 Summary of amount of medial and lateral condylar movement for normal gait (heel strike to toe off 0–30° of knee flexion) and deep bending of the knee (0–90°)
Full size table

Within the PS umbrella, there are a multitude of customized cam-post designs which are all aimed at best re-creating normal knee kinematics [22••]. All conventional and customized cam-post designs demonstrate less rollback and internal rotation when compared to a native knee [22••].

The rationale for the paradoxical movement in normal gait with return of normal rollback during deep flexion is that the post does not engage the cam during normal gait. However, during deep flexion, the post uses the cam as a lever to induce the normal femoral rollback seen in the native knee [21•].

One variation of the PS design seeks to replace both the ACL and PCL with posts, bicruciate stabilized (BCS) TKA designs utilize an anterior cam-post to substitute for the ACL, similar to the mechanism used for replication of the PCL. These implants theoretically provide the technical implantation ease of traditional CR components with increased stability and more natural biomechanics provided by the anterior post. However, the ability of the anterior post-cam design to mimic the biomechanics of a native knee may be limited. Grieco et al. [12••] used fluoroscopy to compare in vivo kinematics of BCS-TKA and the normal knee. They conclude that the dual cam-post mechanism does not adequately substitute the cruciate ligaments in mid-flexion, a point at which ACL tension is decreasing and PCL tension is increasing.

Cruciate Retaining Biomechanics

The theoretical benefit of maintaining the PCL in a total knee arthroplasty is the maintenance of normal knee kinematics (Table 2). This is accomplished by decreasing sheer stress over the femoro-tibial interface by decreasing posterior translation [17•]. Cruciate retaining total knee arthroplasties (CR-TKA) were designed as an improvement over PS-TKA to improve kinematics and function. However, in a biomechanical study of CR-TKAs, Dennis et al. [21•] demonstrated paradoxical anterior translation of the femur relative to the tibia during regular gait [13, 21•].

In the cruciate retaining knee during normal gait, the medial femoral condyle moves anteriorly 0.9 mm and the lateral condyle moves anteriorly 0.1 mm which results in a paradoxical external rotation of the tibia in addition to paradoxical anterior translation [21•] (Table 1).

During deep bends, there was noted to be −1.7 mm of medial condyle motion and −0.7 mm of lateral condyle motion [21•]. Though there was no paradoxical anterior motion as seen in normal gait, the tibia continues to paradoxically rotate externally during deep flexion in the CR-TKA design. This shows that, though the PCL is retained in an effort to improve kinematics, femoral rollback was not restored in these designs and paradoxical anterior translation is seen during normal gait [13, 21•] (Table 1).

One possible explanation for the paradoxical motion has been the amount of slope built into the tibial component. Several studies have shown that in CR-TKA when the posterior tibial slope is greater than 3°, the femoral condyles demonstrate normal femoral rollback from 20 to 90°. However, when this slope is decreased to less than 3°, the motion becomes paradoxical [23,24,25].

Outcomes and Shortcomings

The overall goal in developing the various designs of TKA is to improve outcomes and survivorship in an effort to improve patient care. Though PS-TKA total knees have demonstrated improved motion in kinematic studies, survivorship has been better with CR-TKA designs. Fifteen year outcome data puts survivorship of CR-TKA around 90% and 77% for PS-TKA [4]. This trend holds true across multiple long-term studies [5, 15•, 16].

CR-TKA and PS-TKA are also not without fault. Though CR-TKA has been shown to have improved survivorship in patients with fixed varus deformities, overall outcomes have been mixed with numerous studies showing no difference at short, middle, and long-term follow-up, while patients with rheumatoid are predisposed to late PCL rupture in CR-TKA [5, 26, 27, 28••, 29••, 30, 31].

Eliminating the PCL during surgery raises a separate set of concerns. The polyethylene post has been reported to fracture and the post can be jumped by the femoral component in an unstable knee [32, 33, 34••]. Additionally, the removal of bone from the inter-condylar notch during surgery possesses a greater risk for possible condyle fracture in PS-TKA when compared to CR-TKA [35, 36].

With both designs showing significant shortcoming and overall range of motion being lower for CR-TKA at long-term follow-up, room for improved designs remains [5•, 10, 37].

“Congruent Articular Surface”

With the long-standing debate over the merits of cruciate retaining versus posterior stabilized designs, most authors agree if the PCL is sacrificed, then it should be substituted. One alternative to the PS design is the congruent articular surface (CAS) tibial insert in which there is no cam-post mechanism and the PCL can be substituted with a higher anterior and posterior lip (Table 1).

This design has the theoretical benefits of bone preservation as no box cut is required, and it distributes the loads over a larger surface area disturbing the forces more evenly at the bone-implant interface [38•]. This design also avoids the risks of the cam-post mechanism such as fracture, post wear, and jumping the post.

A potential drawback of the CAS insert is having increased contact surface area potentially leading to increased wear characteristics; however, the radiographic data has not been shown to have increased rates of osteolysis [38•]. Another potential drawback is decreased ROM on deep knee flexion due to the posterior lip preventing effective femoral rollback. In vivo fluoroscopic studies have shown that congruent designs have less AP-translation when compared to flat, or mobile bearing, polyethylene cruciate-retaining designs but had non-physiologic rollback when compared to the native knee [39, 40•] (Table 2). With regards to AP stability and ROM, the CAS insert has been shown to be similar when compared to CR inserts [41].

In a retrospective study with short-term results, the ultra-congruent, or anterior stabilized, polyethylene inserts have demonstrated no difference in Knee Society Scores and rates of complications when compared to cruciate-retaining designs, and they demonstrated a lower rate of revision surgery (1.5% vs 4.6%) [42].

Medial Pivot

Normal knee kinematics based upon multiple in vivo fluoroscopic studies have demonstrated that during flexion, the medial femoral condyle pivots while there is a posterior translation or “rollback” of the lateral femoral condyle [43••].

To address the paradoxical movement of the femoral condyles of traditional knee arthroplasty designs, the medially congruent or medial pivot (MP) design was developed (Table 2). This unique design involves a concave shape medially with an anterior lip, or “ball and socket articulation” and a less congruent shape to the lateral compartment which allows anterior to posterior translation meant to replicate normal knee kinematics 39 (Table 1, Fig. 1).

Fig 1
figure 1

Schematic diagram of the medial pivot polyethylene biomechanics. As the knee goes into flexion, the medial compartment’s rotation moment is converted to translation moment in the lateral plateau allowing for femoral rollback

Full size image

Fluoroscopic kinematic studies have demonstrated that MP designs produce similar motion to the native knee with posterior translation of the lateral femoral condyle without the paradoxical anterior motion during weight bearing activities including deep knee bending [44•, 45, 46••] (Table 2).

KSS and OKS scores have been shown to be similar comparing MP and PS fixed bearing designs in patients, and have been shown to have better FJS-12 scores, especially with deep knee flexion and stability, at 1 year [46••, 47••]. Patients reportedly were less aware that they had an artificial joint when they had an MP design compared to PS cam post designs [47••]. 15-year outcome data reported by Karachalios et al. in a retrospective study showed that the MP knee design has a 96.4% survivability rate for any cause, and significant improvement in knee outcome scores [48]. Similar rates of survivability and improvements were found in the study performed by Macheras et al., and they noticed no difference in functional outcome scores between patients in which the PCL was retained versus resected at 15 years [49].

Lateral and Dual Pivot

Despite the development of medial and lateral pivot designs, various kinematic studies have demonstrated a complex pattern of pivot motions at various degrees of flexion [50,51,52,53,54,55]. Early flexion angles, such as those that occur with activities such as walking and running, are characterized by a lateral pivot movement. In contrast, deep flexion angles are characterized by a medial pivot pattern. Lateral pivot implants have been developed to improve patient outcomes by providing more natural kinematics during activities such as walking. However, lateral pivot designs are far less common than medial pivoting implants.

Dual-pivot implants are designed to better mimic natural knee kinematics by providing a lateral and medial pivot point at lower and higher flexion angles, respectively. Meneghini et al. [50] used sensor trials to record intra-operative knee kinematics in one hundred twenty TKAs. The authors report higher Knee Society Function Scores, greater one-year functional scores, more normal feelings knees, and higher overall satisfaction in patients with early lateral/later medial pivot kinematics. The authors conclude that replicating the dual-pivot kinematic pattern may improve function and satisfaction after TKA.

Sandberg et al. [56••] compared outcomes of two similar cohorts who underwent TKA with either traditional (n = 183) or dual-pivot (n = 183) TKA designs and minimum 1-year follow-up. Patients with a dual-pivot TKA reported less walking pain (p = 0.022), greater participation in very active activities or impact sports (p = 0.067), and were more likely to report that their knee feels normal (p = 0.091). Importantly, 89% of patients in both groups reported that they were satisfied or very satisfied with their outcome. The authors conclude that the dual-pivot design may provide potential benefit, but not similar overall satisfaction between the two groups.

Bicruciate Retaining

Bicruciate retaining (BCR) or bi-unicompartmental surgeries are desirable given the retention of native knee kinematics and proprioception through the preservation of the native ACL and PCL (Table 1). Preservation of both cruciate ligaments is supported by studies demonstrating improved patient satisfaction in BCR TKA designs. Patients surveyed after TKA have reported abnormal feeling knees after replacement, which is possibly explained by a failure or inability to recreate normal kinematics and loss of intra-ligamentous sensory function [57, 58]. Implants that preserve or substitute the anatomy of both cruciate ligaments provide the theoretical benefit of closely mimicking the natural biomechanics of the native knee and preserving proprioceptive feedback. The contribution of the ACL to the “screw-home” mechanism and the PCL’s function to control femoral rollback both play a major role in knee stability [59,60,61,62, 63•]. Furthermore, various studies have demonstrated more native-like kinematics when the ACL is preserved during TKA [63•, 64, 65, 66•, 67, 68•]. Despite potential benefits of preserving the ACL, it is commonly sacrificed. Reasons for ACL sacrifice include technical difficulties, implant availability, and absence of a properly functioning ACL [63•]. In addition, BCR-TKA has been shown not to restore native tibiofemoral articular contact kinematics [69].

Tibial baseplate design and insertion have played a significant role in the creation of bicruciate retaining implants. With sacrifice of the ACL, the joint space can be opened widely and the tibia can be subluxated anteriorly, allowing for insertion of a stemmed tibial baseplate. Two approaches to baseplate design have predominated to allow ACL preservation. The first is similar to a standard PCL-retaining baseplate and polyethylene; however, the recess is extended anteriorly to accommodate the ACL. One drawback to this design is the narrow connecting bridge, which can fail secondary to torsional loading [70]. The second approach to ACL preserving implants is the use of two separate baseplates, one for each tibial condyle [70, 71]. The main challenge to this approach is component malalignment both intraoperatively and throughout the implants lifespan. Intraoperative error or different patterns of subsidence between the two over time can lead to accelerated wear and need for early revision [70]. Nowakowski et al. [70, 72] have designed the transversal support tibial plateau (TSTP) concept to address subsidence rates between the medial and lateral baseplate [68•, 70]. Their design incorporates a transverse post, which connects the two baseplates beneath the joint line; however, long-term clinical data is not yet available.

In addition to baseplate design, the quality of the intact ACL, its attachments, and the effects of ligament balancing play important roles in the outcome and implant survival. Despite an intact ACL, the ligament may be attenuated from prior trauma or chronic wear. Furthermore, balancing during TKA has potential to increase tension on the ACL and tibial bone cuts, which spare the ACL, leave a bone island that can fracture [9, 63•, 73, 74]. Therefore, BCR-TKA is technically challenging and typically longer to perform, which may help explain poor acceptance among orthopedic surgeons.

Despite potential advantages of BCR-TKA, it may carry a higher reoperation rate compared to other implant designs. Christensen et al. [75•] evaluated revision rates in 475 primary TKAs, which included 78 BCR-TKAs and 294 CR-TKAs. They demonstrate a higher all-cause revision rate in the BCR group (5%) compared to other groups (1.3%). In addition, the BCR group had a higher frequency of irrigation and debridement with component retention. They conclude that BCR implants may have inferior survivorship compared with conventional CR implants. In contrast, Pritchett [9] has reported an 89% survivorship at 23 years with a 5.6% revision rate. Revisions were most commonly performed for polyethylene wear.

Varus-Valgus Constrained

Varus-Valgus Constrained Total Knee Arthroplasty (VV-TKA) uses a polyethylene with a taller and wider post that is more constrained by the femoral box to prevent coronal instability (Table 1). The wider post, through contact with the intercondylar notch, restricts varus-valgus and rotation motion more than a traditional PS [76] (Fig. 2).

Fig 2
figure 2

a–c Standard PS insert demonstrating that with even significant amounts of varus or valgus displacement, the post does not engage the condyles thereby relying on the collateral ligaments and bony anatomy for stability. Post’s main function is to provide anterior-posterior stability. d–f Varus-Valgus constrained total knee polyethylene insert demonstrating how it engages the medial/lateral condyle earlier compared to the standard PS insert thereby allowing for increased varus-valgus stability

Full size image

This is in addition to the primary function of the post, which is to resist posterior translation and serve as a lever for the CAM mechanism. Additional constraint without the need for extra bone cuts or component revision is useful when coronal balancing cannot be achieved through soft tissue techniques alone. Another advantage of VV-TKA includes less resection of bone than a Rotating Hinge [77]. Indications for a VV-TKA include a severe valgus deformity, incompetent medial collateral ligament (MCL), or severe flexion contracture in which the knee cannot be balanced [78, 79].

One drawback to increasing the constraint of the implant is that this leads to additional stress on the femoral, tibial, and polyethylene components. The increased constraint provided by the VV-TKA is a risk factor for accelerated wear rates of the polyethylene and component loosening [80]. In addition, knee biomechanics are altered and the polyethylene post may undergo permanent deformation due to bending loads and torques applied by the femoral box [81]. Excessive forces have caused post breakage and failure of the locking mechanism that can secure the polyethylene into the tibial base-plate [82,83,84,85]. Pitta et al. [86••] found that VV-TKA implants had a higher rate of failure and were twice as likely to undergo revision compared to PS implants in a prospective study of 18,065 modern total knee arthroplasties; however, early outcomes of VV-TKA are similar to standard PS designs at 2 years [81, 86,•• 87••].

Studies are conflicting on the actual impact of higher constraint on implant survivability [80,81,82, 87••]. This may be due to several factors, including the need for more constraint with severe deformities and when soft tissues laxity requires the implant to bear large biomechanical stresses. Cholewinski et al. [88] reported on long-term outcomes of primary VV-TKA. Their retrospective study comprised 43 knees in 41 patients with a mean age of 66 years and minimum 10-year follow-up. Complications other than venous thrombosis occurred in 16% of patients, which included 2 cases of septic loosening and 1 case of major instability in a patient with an ipsilateral hip arthrodesis. No cases of osteolysis or constraint-mechanism failure were identified. The 11-year prosthesis survival rate was 88.5% after excluding cases of infection. They conclude that long-term functional gains were similar to PS-TKA with a higher complication rate and decreased survival. However, they note that pre-operative deformities were severe and two-thirds of knees had one or more prior surgical procedures.

Siqueira et al. [89] also reported on survivorship and modes of failure in 685 cases: primary (n = 247), aseptic (n = 315), and septic (n = 123) revision TKAs with constrained implants. The authors report 10-year survival as 89%, 75%, and 55%, respectively. The most common mode of failure was infection. Mechanical failure was typically due to peri-prosthetic fracture (45%) and soft tissue instability (19%).

Despite the known complications that arise from increased constraint, several studies support its use when appropriately indicated. Rai et al. [90••] found 95% prostheses survival at an average of 6.5 years in a retrospective review of 38 primary VV-TKAs. Patients had improved Knee Society Scores (KSS) scores, improved Hospital for Special Surgery (HSS) scores, and improved Range of Motion (ROM). Radiolucency was noted in 13 knees, primarily affecting the tibial implant. Sabatini et al. show improved functional KSS, no evidence of loosening, or peri-prosthetic fractures, and full recovery of extension in a series of 28 primary VV-TKAs with an average of 31 (6–48) months of follow-up. They conclude that the implants are safe and practical for primary TKA in cases of severe deformity that cannot be managed with soft tissue releases alone.

Some authors recommend using stemmed tibial and femoral component to address increased stresses of constrained implants [91, 92]. Moussa et al. [92] report a significantly higher revision rate (2.20% vs 0.98%) for constrained prostheses (n = 817) compared to PS knees (n = 817) due to mechanical failure. In contrast, Anderson et al. [93] demonstrated a 2.5% failure rate in 192 knees after VV-TKA without stem extensions at a mean of 47 months. Failures were comprised of 2 infections, 1 aseptic loosening, 1 supracondylar femoral fracture, and 1 tibial post fracture. They concluded the use of a non-modular constrained condylar knee for difficult primary TKA demonstrated reliable short- to mid-term results and question the use of routine use of intramedullary stem extensions.

Conclusion

Overall outcomes comparing the various types of modern polyethylene inserts have all yielded results suggesting that though there are benefits to certain inserts, there is no one insert that is superior to the others.

Congruent articular surface inserts provide a true cruciate retaining type implant which preserves femoral bone stock. Additionally, its highly congruent articular surface allows a more even distribution of forces [38•]. However, due to the increased congruence and lipped design, deep flexion is limited and greater amounts of polyethylene wear is possible with the increased congruence [38•, 39, 94].

Similar to the congruent articular surface design, medial pivot inserts have a highly congruent concave medial side which prevents deep flexion, but this design was able to solve the conundrum of paradoxical anterior translation that is seen with the other traditional CR designs [13, 39, 41, 44•, 45, 94]. In spite of the improved kinematics, there is still no significant difference in outcomes at 15 years when compared to posterior stabilized total knees [46••, 47••, 48, 49].

Dual pivot inserts seem to have improved patient-reported outcomes with their more natural feeling knee that has both a medial and lateral pivot. Though overall outcome medium and long-term data are still needed, early data suggests that patients may notice improved outcomes with this design [50, 56]. Additionally, while patients are reporting a more natural feeling knee, we are also lacking kinematic studies showing how close to a native knee the articulation truly is.

Bicruciate retaining total knees have demonstrated the best kinematics and range of motion, but the technically challenging procedure and high revision rate seem to be the limiting factor for this design [63•, 64, 65, 66•, 67, 68, 75•].

Though the VV-TKA serves a different purpose than the other CR inserts discussed in this paper, it allows the surgeon the ability to increase stability without modifying the femoral cut [76, 77]. Though short-term/smaller studies have demonstrated no significant increase in revision rate, larger studies with longer follow-up have shown a significant increase in the need for revision in VV-TKA [86••, 88, 89, 90••].

The optimal polyethylene inserts still remain to be determined. Long-term studies on newer designs need to be completed and older designs such as the CR and PS may have a good track record of survivorship lack the ability to truly recreate the biomechanics of the native knee [4, 5•, 15•, 16, 21•, 37]. Ultimately, we are not able to recommend for or against any one design. The optimal insert is one the surgeon feels most comfortable using and the one that serves that specific patient’s needs.