Cruciate-Retaining Total Knee Arthroplasty: Current Concepts Review

Posterior cruciate-retaining (CR) total knee arthroplasty for osteoarthritis of the knee is a popular implant choice. At present, there is no consensus on whether sacrifice or retention of the posterior cruciate ligament (PCL) offers superior outcomes. This review explores the current literature available on CR total knee arthroplasty (TKA). PubMed was searched by keyword to find relevant articles for inclusion. Additional sources came from article references and joint registry reports. CR design knees have distinct kinematic gait patterns from posterior-stabilizing (PS) knees and exhibit paradoxical anterior femoral movement with less femoral rollback. While CR implants offer less flexion than PS designs, the difference is not clinically detectable as clinical scores are similar in the short and long term. CR implants have better long-term survival compared to PS knees, likely due to lower risk of aseptic loosening. CR total knee arthroplasties also have shorter operating times and lower risk of peri-prosthetic fractures. Because the CR implant is unconstrained, there may be an increased risk of instability compared to PS designs, but the literature is mixed. Overall, the current literature supports the continued use of CR TKAs due to their lower risk of complications, durability, and demonstrated equivalence in function to posterior-substituting models.


Introduction And Background
Osteoarthritis (OA) is one of the leading causes of disability worldwide and its true burden on population health is likely underestimated. Approximately 3.8% of the global population is estimated to be affected by knee OA [1], and the knee is the most commonly replaced joint [2]. Primary total knee arthroplasty (TKA) offers definitive surgical management for refractory pain or functional limitation due to OA of the knee, amongst other painful knee conditions such as rheumatoid arthritis. The number of primary TKAs is projected to increase in the next 20 years due to age-and obesity-associated increase in both prevalence and incidence of OA [2].
Various joint replacement systems are available for TKA. Cruciate-retaining (CR) and posterior-stabilizing (PS) implants are the most widely utilized. In the United States. Surgeons prefer PS implants over CR, although CR implant designs are growing in popularity [3]. In European countries, the CR model is more widely used [4][5][6][7][8]. Both CR and PS models offer unique advantages and drawbacks, but there remains no consensus on which offers superior outcomes. This study aims to review the scope of current literature on CR implants for TKA, specifically focusing on kinematics, functional outcomes, survival, and complications.

Methods
For each section of the article, a PubMed search utilizing relevant keywords was conducted. Keywords used were "native knee flexion kinematics" (n = 308), "CR PS total knee kinematics" (n = 80), "CR PS knee outcomes" (n = 107), "CR knee survival" (n = 51), "PS knee survival" (n = 77), "CR PS knee flexion" (n = 97), "CR TKA revision" (n = 61), "CR TKA instability" (n = 89). Six additional references were obtained from 2022 national registry reports. Articles that compared CR and PS TKAs in terms of any metric were of special interest. Two independent reviewers screened articles based on titles. Those that were deemed suitable for abstract review were compiled into a list that was independently screened by two reviewers for retention or exclusion based on relevance. A total of 51 articles were included in the review (Figure 1). Articles were excluded if they were not about CR TKAs, were about TKAs for treatment of conditions other than knee OA, TKAs on patients who had undergone previous knee surgeries, focus on posterior tibial slope, focus on soft tissue deformities, and focus on component-specific design such as polyethylene design.

Review
The native knee

Structures of the Knee
Articular surfaces of the native knee involve medial and lateral femoral condyles, which articulate with the respective condyles of the tibial plateau and are supported by the surrounding menisci. The ligamentous structures, which include the anterior cruciate ligament, medial collateral ligament, lateral collateral ligament, and posterior cruciate ligament (PCL) serve to provide tension and stability to the knee joint during movement.
The PCL is an intra-articular structure of the knee joint with its own synovial sheath. It is composed of the anterolateral bundle (ALB) and the posteromedial bundle (PMB). The ALB originates anteriorly compared to the PCL on the medial epicondyle of the femur and inserts anteriorly and medially to the PMB on the posterior tibia. Compared to the PMB which runs an oblique course relative to axial load on the knee joint, the ALB is situated upright. Both bundles resist posterior translation of the tibia in a codominant fashion during extension and flexion of the knee joint. During maximal extension of the knee joint, the PMB is stretched taut and acts as the dominant bundle to resist posterior tibial translation. In contrast, the ALB elongates during mid-flexion and resists posterior tibial translation while the PMB is lax. During extreme flexion, the PMB is again dominant and serves to resist posterior translation and internal rotation of the tibia.

Kinematics
During knee flexion, the asymmetric articular surfaces of the knee along with the PCL enable femoral rollback and external rotation of the femur with respect to the tibia. The medial tibial plateau is larger and concave, while the lateral tibial plateau is convex. As a result, medial condylar contacts are more static while lateral contact points are more variable, causing a pivot about the medial condyles. From extension to 20º of flexion, condylar contact points rotate externally and posteriorly as the PMB exerts anterior and external rotary force on the tibia. During mid-flexion up to about 80º of flexion, the medial and lateral condylar contact points both translate posteriorly owing to the centrally located ALB providing anterior forces on the tibia. During flexion >90º, the lateral contact points translate posteriorly while the medial contact points remain static due to the PMB [9].
The net effect of knee flexion is a posterior translation of the femoral head with respect to the tibia known as "femoral rollback" (Figure 2). This is opposed by the "screw home" mechanism in extension, which is an external tibial rotation with relative internal femoral rotation centered on the medial condyle that provides stability during full knee extension. Coordination between the ALB and PMB of the PCL creates constraint and stability throughout all ranges of extension and flexion.

Kinematics after arthroplasty
Due to preservation of the native PCL, CR arthroplasty designs allow for stability throughout the flexionextension arc, detailed above. Kinematic studies of CR designs, often in comparison to PS designs, provide understanding of the effects of PCL preservation. In addition to passive flexion, mobility patterns critical for activities of daily living such as walking gait and stair climbing are of particular interest when studying TKA kinematics.

Gait Kinematics
CR implant designs generally allow posterior translation of femorotibial condylar contact points in early flexion <30º, followed by paradoxical anterior femoral sliding from 30-60º during normal gait kinematic analysis [10][11][12]. In comparison to PS designs, which do not retain the native PCL, condylar contact points in CR designs are overall relatively anterior from 30-60º. The difference was most significantly exaggerated in deeper knee flexion >60º, which is not engaged during normal gait [13]. Studies comparing kinematics of CR and PS designs during normal gait report CR knees demonstrated an increased "pivot" moment about the medial compartment, as evidenced by higher lateral-to-medial rollback ratios [12][13][14].

Stair-Climbing Kinematics
During stair climbing, Hamai et al. found that both CR and PS designs demonstrate paradoxical posterior femoral translation during early flexion due to absence of the ACL [15]. In CR models, the PCL provided midflexion stability, while PS designs demonstrated risk of paradoxical anterior translation in mid-flexion. Flexion during stair-climbing was not sufficient to engage the cam-post mechanism of PS implants [15]. Fantozzi et al. found the medial condylar contact point for CR implants consistently translated posteriorly during stair climbing to a degree similar to PS designs until about 60º of flexion, upon which PS medial condylar contacts shifted significantly more posteriorly [14].

Passive Flexion Kinematics
Intra-operative studies investigating the kinematics of passive flexion in CR implants against PS implants in the same knee demonstrate different patterns of condylar contact at varying degrees of flexion. From 0-30º of flexion, studies agree both CR and PS designs demonstrate a posterior shift of the condylar contact points [10,11,13]. From 30-60º of flexion, CR knee contact points were variable, and some knees demonstrated paradoxical anterior femoral motion [10][11][12]. Translation of contact points during 30-60º of flexion was not observed in PS knees, in which condylar contact points remained stable [10,12]. In midflexion from 60-90º, CR knee contact points were significantly anterior compared to the PS knee contact points, although both knees demonstrated posterior translation [11,13]. In deep passive flexion >90º, CR knees had significantly higher rates of paradoxical anterior femoral translation than PS knees; however, not all CR knees demonstrated paradoxical anterior femur translation [10,13].

Femoral Rollback
A recent meta-analysis by Li et al. suggests that CR knees retain less native femoral rollback than PS knees (80% vs 90%) [16], consistent with the kinematics literature previously discussed. Yoshiya et al. report that CR implants also demonstrate more variable results in terms of kinematics, which may reflect variability of native PCL condition [10]. Exclusive lateral condylar liftoff has been noted in CR TKAs, as opposed to PS TKAs in which both medial and lateral condylar liftoff has been noted [12].

Short-Term Functional Outcomes
Most studies on CR TKA functional outcomes are limited to short-term follow-up, defined here as two to five years after primary TKA. Overall, in this period, studies suggest CR and PS TKAs offer equivalent clinical scores as assessed by the Oxford Knee Score (OKS), Knee Society Pain Score and Knee Society Functional Score (KSS) [17][18][19][20][21][22]. However, certain studies have demonstrated better clinical scores associated with PS implant designs [23][24][25][26]. The Forgotten Joint Score (FJS) is a newer, validated, and reliable metric with lower ceiling effect than the OKS and KSS and allows for differentiation among patients with positive TKA outcomes [27][28][29]. FJS score showed no significant difference between CR and PS TKAs overall and in all subsets of age, gender, or laterality in a study by Thippanna et al. [17], but was significantly higher for CR knees in a study by Thuysbayert et al. [28]. Based on the current literature, it appears that neither CR nor PS implants have a distinct advantage over the other on short-term follow-up.

Long-Term Functional Outcomes
Long-term functional outcomes, defined as more than five years after primary TKA, appear similar between CR and PS designs as well.  [31]. Although the current data is mixed, further study on long-term functional outcomes is warranted.

Range of Motion
In general, CR TKA designs have smaller flexion arcs than PS designs, reported to be between 2.24-4.05º less (

NA: data not available
Survival Implant survivability is of paramount importance to both the patient and the treating surgeon. Short-term TKA failure is commonly due to infection, while aseptic loosening is the most common cause of long-term TKA failure [35][36][37]. A meta-analysis by Migliorini et al. found no significant difference in revision rate over a mean follow-up period of 3.37 years [26]. Rates of revision arthroplasty after primary TKA are equivalent for CR and PS knees at mid-term follow-up [19,20].  Aseptic loosening has been found to occur significantly less frequently in CR TKAs compared to PS TKAs, while infection rates are not significantly different between the two arthroplasty types [37,38]. At 13-year follow-up, 93% of revisions for PS designs were due to loosening, and males were at significantly higher risk [37]. Greater frequency of aseptic loosening in PS knees may explain better long-term survival of CR TKAs compared to PS TKAs. While certain studies do not support CR design superiority regarding longevity [30], more suggest at least equivalent long-term survival outcomes (Table 3) [21,22,30,31,38,39].

Fractures
Complications for TKA include fractures, pain, and unplanned conversion to a PS implant design. The incidence of intraoperative fracture risk during TKA varies between 0.2-4.4% [42,43]. Intraoperative fracture risk is significantly higher in PS than in CR TKAs, with a relative risk of 4.74 [42,44]. Periprosthetic fractures are devastating complications following TKA. Elkabbani et al. conducted a retrospective review in which CR TKAs demonstrated a lower risk of periprosthetic fracture compared to PS TKAs, with a risk ratio of 0.10 [45]. CR implant designs carry the advantage of avoiding additional femoral resection and iatrogenic creation of bony stress risers that risk intraoperative fracture and later periprosthetic fracture.

Pain
Persistent pain after TKA is one of the major causes of TKA revision [37]. Pain can be idiopathic or associated with mechanical symptoms. Knee Society Pain Scores are mostly equivalent between CR and PS designs [19,20,23,24]. A systematic review and meta-analysis by Lewis et al. found that persistent postoperative knee pain was more closely associated with greater severity of pre-operative pain, psychiatric illness, and presence of other medical comorbidities rather than choice of TKA prosthetic model [46].

Instability
Due to preservation of the PCL, CR implants are less constrained than PS implants and theoretically less stable. Instability following TKA is clinically important with an incidence of approximately 7.5% [35] and is a significant cause of revision [36]. A retrospective review by Hannon et al. found CR design to be a risk factor for surgical revision due to instability [47]. Savov et al. reviewed patients with valgus OA who underwent either CR or PS TKA, and results showed 8% of patients in the CR group underwent revision due to instability compared to no patients in the PS group [48].
Contrarily, a systematic review by Rouquette et al. on tibiofemoral dislocation after TKA revealed dislocations were 30% CR designs and 62% PS designs [49]. Additionally, a meta-analysis of over 4,000 TKA procedures by Migliorini et al. found no differences between CR and PS designs in terms of joint instability [26]. Thus, while some literature suggests CR TKA may increase risk of instability compared to PS TKA, uncertainty remains as other studies support equivalent risks.

Intraoperative Conversion to PS
Unexpected intraoperative conversion from a CR implant design to a PS implant design is another intraoperative consideration during TKA. Iatrogenic PCL injury or ligamentous incompetency may require the additional constraint provided by PS designs. One study reports a 9% PS conversion rate of attempted CR TKAs [50]. Wang et al. report PS conversion is associated with a thicker polyethylene insert, a result of additional bony resection [51]. Severe varus deformity, flexion contracture, and narrow femurs are additional reported risk factors for conversion from CR to PS implants [50,51].

Conclusions
The cruciate-retaining TKA design is widely used, both in the United States and worldwide. Currently, the benefits and drawbacks of PCL retention in TKA for primary OA is a topic of debate with no clear advantage between CR or PS designs. Studies were also highly heterogeneous in terms of how data were presented. This review offers a summary of the recent literature available on CR TKAs.
Kinematics after TKA for OA show distinct patterns between CR and PS knees, though both models demonstrate different kinematics from the native knee. CR designs encourage less femoral rollback, resulting in clinically insignificant decreased flexion. More external rotation about the medial aspect of the knee is also seen in comparison to PS designs.
Short-term functional outcomes are relatively similar between CR and PS designs, aside from ROM. Overall, the current literature and registry data suggest improved long-term survival of CR TKA designs compared to PS designs, which becomes more pronounced greater than 10 years post-operatively. As more data becomes available, continued investigation into the long-term survival of CR and PS prostheses is indicated to corroborate or reject these early findings.
CR TKAs have the advantage of decreased incidence of intra-operative and post-operative periprosthetic fractures compared to PS designs. Post-operative pain scores are similar between CR and PS designs. A possible increased risk of instability may be associated with CR models, but the literature is conflicting. A large database study of knee dislocation rates between patients with CR and PS implants may help clarify this hypothesis. As TKA designs and techniques continue to advance, the current body of literature supports the continued use of cruciate-retaining implants.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.