Popliteus impingement after TKA may occur with well-sized prostheses

Residual pain and poor outcomes after total knee arthroplasty (TKA) can be attributed to soft-tissue impingements, which could arise due to prosthetic overhang at the femur [10, 30] or the tibia [9]. Impingements may involve various anatomic structures such as the medial collateral ligament (MCL), the iliotibial band, the popliteus tendon, the patellar tendon, and the medial and lateral patellar retinaculum [3, 8, 10, 30]. The popliteus tendon is of special interest due to its intra-articular location and its close contact with the posterolateral tibial plateau and the lateral condylar margin [46].

In a normal knee, the popliteus remains in close contact with the convex posterolateral area of the lateral tibial plateau, up to the popliteus hiatus, where it is stabilized by the popliteomeniscal fascicles [11, 16, 38, 43, 46, 47]. It then crosses the margin of the lateral condyle and inserts anterior and distal to the lateral epicondyle [24, 29, 32, 44]. In full extension, the popliteus is engaged in a distal indentation of the lateral condyle, called the sulcus statorius [43]. During flexion, it glides over the bumpy margin of the lateral condyle, and beyond 100° of flexion, it lies entirely within the groove of the sulcus popliteus (Fig. 1) [28, 44].

In a TKA, the thickness of the tibial component is selected to restore the joint line and to match the contours of the resected surfaces [13, 14, 31, 36]. Therefore, a superstructure of polyethylene is generally built above the posterolateral area of the tibial plateau, leading to a potential risk of popliteus impingement (Fig. 2). At the femur, any shape difference between the prosthetic and the native lateral condylar margin, such as induced by the design, the sizing or the positioning of the femoral component, potentially affects the tracking of the popliteus [4, 45]. Indeed, impingements have been reported secondary to friction against femoral osteophytes or overhanging prosthetic condyles [1, 4, 26] and have been successfully treated by arthroscopic popliteus release [1].

It has been demonstrated that mediolateral overhang of the femoral component could cause residual pain after TKA, and that slightly undersizing the femoral component may improve pain scores [10, 30]. However, undersizing may lead to implant subsidence or tibiofemoral instability [5] and could compromise bone–implant fit [7]. Many recent femoral components are available in ‘standard’ or ‘narrow’ versions and allow greater tibia–femur size mismatch that enables surgeons to fine-tune mediolateral sizing. Nevertheless, the optimal sizing, exact fit or slight under-coverage, remains controversial.

The purpose of this study was to determine the mechanisms and extents of popliteus impingements by examining the bony and prosthetic contours of knees before and after TKA and to investigate the influence of implant sizing. The study hypotheses were that (1) popliteus impingements after TKA may occur at both the tibia and the femur, and (2) even with an apparently well-sized prosthesis, the position or tracking of the popliteus during knee flexion is modified compared to the preoperative situation. A validation (or invalidation) of these hypotheses might have clinical consequences concerning the optimal sizing and positioning of the prosthetic components and the design evolution of the components towards a more ‘popliteus friendly’ design.

The location of the popliteus tendon was studied on three fresh frozen cadaver knees throughout the flexion–extension range, using computed tomography (CT), before and after implantation of TKA. The cadavers had been donated for research by testament, and none of the cadaver knees had history of previous surgery.

The main finding of this study is that a well-sized tibial component modifies popliteal tracking, while an undersized tibial component maintains more physiologic patterns. The data also demonstrate that oversizing the tibial component by one size increment shifts the popliteus considerably throughout the full arc of motion. These results confirm previous clinical investigations that reported better pain scores in patients with ‘undersized’ implants [10], and poorer outcomes in patients with posterior tibial overhang [9]. Why an ‘anterior translation’ of the popliteus was observed in deep flexion with a normosized prosthesis remains unclear. An explanation could be that sacrificing the popliteo–meniscal fascicles during implantation, secondary to lateral meniscectomy, destabilizes the tendon in deep flexion [38, 47].

Another finding was that the popliteus–condyle contact is modified after TKA throughout the flexion range, whatever the sizing option, which resulted in a ‘reversed pattern’. From full extension to mid-flexion, the popliteus was medialized because the margin of the prosthetic lateral condyle does not reproduce the smooth ridge of the native lateral condyle, which is often removed during surgery. From mid to deep flexion, the tendon was lateralized due to prosthetic overhang at the lateral condyle.

Residual pain is a very frustrating situation after TKA, and identifying the cause can be challenging [40]. Soft-tissue impingements have been described as an aetiology but mostly in patients with overhanging components [9, 10, 30], and little is known about the behaviour of the soft tissues surrounding a well-sized TKA. There is little published literature on the diagnosis and treatment of popliteus impingement after TKA. While some authors reported successful pain relief after tendon release [1, 4, 26], none investigated the pathophysiology in detail. Allardyce et al. [1] reported results of arthroscopic release in two patients presenting ‘popliteus tendon dysfunction’ but did not describe the nature and location of the impingement. Likewise, Kazakin et al. [26] observed snapping popliteus tendon during TKA. The consequences of such popliteus tendon release in prosthetic knees are unclear. While De Simone et al. [15] reported lower function score, neither Kesman et al. [27] nor Ghosh et al. [20] observed adverse effects after popliteus transection, in vivo or in vitro. Recently, Cottino et al. [12] reported an increased TKA laxity after popliteus section, both with cruciate-retaining and postero-stabilized prostheses.

To avoid popliteus impingements after TKA, slightly undersizing the tibial component could be an option, in order to preserve a peripheral bony margin at its posterolateral corner. It must be noted, however, that excessive undersizing could lead to implant subsidence and failure [5]. The use of anatomic base plates, which replicate the tibial asymmetry, may help the surgeon to both undersize laterally and preserve a good medial coverage. Dai et al. [13] and Martin et al. [31] recently demonstrated that asymmetric tibial base plates provide better conformity to resected surfaces. Theoretically an ideal TKA should closely reproduce the shape of the resected articular surfaces, but this is difficult to achieve at the tibia because the polyethylene must respect a degree of congruency with the prosthetic condyles [18, 21, 25] and the lateral compartment presents high geometric variability [8, 18, 37, 41]. Consequently, none of the current TKA designs reproduce the convex shape of the posterolateral tibial plateau, though some lateral UKA models feature ‘dome-shaped’ tibial base plates with bi-concave polyethylene inserts, to better reproduce natural anatomy and kinematics [2, 48].

This study reports that a posterolateral condyle overhang may occur even with normosized implants. The reality of this phenomenon may be criticized as it has been infrequently reported in the literature. Hirakawa et al. [22] in a series of 40 TKAs in a Japanese population reported a overhang of the posterolateral condyle greater than 3 mm in 25 patients and suggested to reduce the dimensions of the posterolateral condyle in TKA. Shah et al. [42] in an Indian population also reported overhang of the posterolateral condyle when implanting a standard TKA. Mahoney et al. [30] measured intra-operatively overhang with the Scorpio prosthesis in several zones of the femur but did not detail the incidence of posterior condyle overhang. However, it must be noted that in Mahoney’s series, patients were operated via subvastus approach, which limits visualization in the posterolateral area.

Prosthetic posterior condyles implanted in this experiment were symmetric, as most TKAs available on the market. The shape of the posterior condyles has been investigated mostly in the sagittal plane [23, 33, 39, 51], and little is known about their morphometry (symmetric or asymmetric) and their mediolateral dimensions. Recently, Monk et al. [35] analysed a series of MRI in healthy volunteers and demonstrated that the posterolateral condyle (mean width, 24 ± 3.5 mm) was narrower than the posteromedial condyle (mean width, 26 ± 3.0 mm), but further morphometric investigations are required to improve our understanding of this specific topic.

This study had some limitations: First, even if this study was based on pre- and post-operative comparison, the limited sample size remains a limitation of this study and anatomic variations may modify the bone–popliteus or implant–popliteus relationships that we observed. Second, only one implant design was used, and it is unclear whether our conclusions can be extended to other implants. It would be valuable to do this investigation with other designs such as medial-pivot or asymmetric TKAs. The posterior translation of the tendon may also be influenced by prosthetic kinematics, and it could be useful to compare postero-stabilized, cruciate-retaining and deep-dished TKAs [34, 50]. Third, this was a non-weight-bearing investigation, which could limit our conclusions, though by virtue of conformity of articular surfaces in postero-stabilized TKA, it can be assumed that the contact of the popliteus with bone and implant surfaces is similar in weight-bearing [6, 52, 53]. Fourth, it is a cadaveric work, and even with fresh specimens, one may argue that the behaviour of the soft tissues was altered. Fifth, the use of CT scans required injecting the popliteus via a lateral incision, which can modify its elasticity. In a preliminary investigation, the authors tested MRI, which proved inaccurate due to technical problems such as mirror images in deep flexion, poor-quality images due to temperature variations within specimens, small diameter of the MRI tube, which does not accommodate the knee in flexion. Finally, the study did not analyse the strain in the popliteus because inserting strain gauges [19] would compromise the image quality by scattering and even the use of chromium steel spheres precludes good soft-tissue analysis [17].

This work demonstrates that sizing in TKA is challenging due to the non-anatomic design of current implants. It suggests that tibio-popliteus impingement might play a role in residual pain after TKA. The practical teaching is that surgeons should choose the size of the tibial implant considering not only the cortical contours of the resected tibia but also the volume of the implant. Hence, tibial components should be sized and positioned in a way to undercover by few mm the cortical contours of the lateral plateau. Design of tibial plateau should also be revisited in the future in order to avoid tibio-popliteus impingements.