Similar kinematic patterns in posterior-stabilized and condylar constrained knee prostheses in revision knee arthroplasty: a prospective cohort study

Lenka Stroobant^a,, Hannes Vermue^a, Ewoud Jacobs^e, Nele Arnout^a, Stefaan Van Onsem^a,b, Scott A. Banks^c, Jan Victor^a, Amélie Chevalier^d

a Department Orthopaedic Surgery, University Hospital Ghent, Ghent, Belgium

b Department Orthopaedic Surgery, AZ Alma, Eeklo, Belgium

c Department of Mechanical & Aerospace Engineering, University of Florida, Gainesville, FL, USA

d Department of Electromechanics, University of Antwerp, Antwerp, Belgium

e Department of Rehabilitation Sciences, Ghent University Faculty of Medicine and Health Sciences, Ghent, Belgium

⁎Corresponding author: Lenka Stroobant. lenka.stroobant@ugent.be

Received: 2025-3-17, Accepted: 2025-4-20, Published: 2025-4-22

Professor P K Surendran Memorial Education Foundation

Disclaimer:
This article was originally published by Reed Elsevier India Pvt. Ltd. and was migrated to Scientific Scholar after the change of Publisher.

Abstract

The rising incidence of primary total knee arthroplasty (pTKA) is expected to lead to more revision TKAs (rTKA), which pose greater challenges and poorer outcomes, burdening both patients and healthcare systems. Knee kinematics play a key role in pTKA outcomes, but less is known about rTKA, where high-constraint implants and joint line elevation (JLE) may affect knee kinematics. The study aimed to: (1) Compare kinematic patterns between posterior-stabilized (PS) and condylar constrained (CCK) implants during open- and closed-chain exercises in rTKA; (2) Assess the impact of JLE on anteroposterior translation and post-cam engagement.

Thirty patients (19 with PS implants and 11 with CCK implants) who underwent rTKA between 2022 and 2024 were tested at a minimum six-month follow-up. Tibiofemoral kinematics during open-chain flexion-extension (FE) and closed-chain exercises (sit-to-stand (STS) and squatting (SQ)) were analyzed using fluoroscopy. (1) Kinematic patterns were compared between PS and CCK implants, focusing on AP translation, internal-external (IE) rotation, varus-valgus (VV) rotation, range of motion (ROM) and post-cam engagement. (2) Joint line elevation was defined as a ≥4 mm increase compared to the native knee on a weight-bearing radiograph, and comparisons were made between patients with and without JLE.

(1) No significant differences were observed in kinematic patterns between PS and CCK implants in rTKA. (2) A JLE ≥4 mm led to instability during squatting, particularly in early- and mid-flexion, with a significantly more anterior position in the medial compartment (0–30°: p = 0.037; 30–60°: p = 0.021). Although post-cam engagement was delayed in patients with JLE, the difference was not statistically significant (p = 0.173)

(1) CCK implants have a kinematic pattern similar to PS implants in revision setting, supporting their use when appropriate, (2) JLE is associated with instability during squatting in early- and mid-flexion.

Keywords

Total knee arthroplasty

Revision

Kinematics

Constrained

Fluoroscopy

Show Related Articles from PubMed

1 Introduction

By 2050, the rate of primary Total Knee Arthroplasty (pTKA) is expected to increase by 143 %.¹ This trend is primarily driven by an aging population and favorable clinical outcomes, which have expanded TKA indications to younger patients.^2–4 Additionally, the rate of revision Total Knee Arthroplasty (rTKA) is escalating more rapidly than pTKA, with projections suggesting a 149 % increase by 2040 and a 520 % rise by 2060 in the United States.³ In contrast to pTKA, rTKA presents greater technical challenges and often yields less favorable outcomes, imposing a substantial burden on both patients and healthcare systems.²^,⁵^,⁶ As knee kinematics are one of the main drivers of patient satisfaction, a comprehensive understanding of the kinematics in the context of revision knee arthroplasty is crucial for improving rTKA outcomes.⁴^,⁷^,⁸

In revision knee arthroplasty, implants with higher constraints might be imperative to deal with bone defects and ligamentous insufficiency. A commonly used implant is the posterior-stabilized (PS) prosthesis, which incorporates a post-cam mechanism to control posterior translation later in the flexion arc.⁴^,⁹^,¹⁰ When PS implants fail to provide sufficient stability, a condylar constrained knee (CCK) is considered, distinguished by a larger and broader post.⁴ Although the CCK is designed to enhance stability, concerns have been raised regarding potential kinematic conflicts once the post-cam engages.⁴^,¹¹ While some studies have suggested that CCK implants perform comparably to PS implants in terms of ROM and clinical outcomes,¹¹ these findings are primarily based on pTKA populations and cannot be directly extrapolated to rTKA populations.

Another technical challenge in revision surgery is the restoration of the joint line height.¹² Studies have demonstrated that an average joint line elevation (JLE) of 5–8 mm is found to occur in revision arthroplasty.¹³ Beyond a reported negative correlation with the outcome, such elevation may result in reduced ROM and instability, potentially affecting knee kinematics.¹³^,¹⁴ Despite extensive research on knee kinematics in primary knee arthroplasty populations, this is the first study to specifically address the revision population, where higher-constraint implants and joint line elevation can substantially affect knee kinematics.

The primary study aim was (1) to compare the kinematic patterns between PS and CCK implants following revision knee arthroplasty during both open- and closed-chain exercises, specifically focusing on (1a) anteroposterior (AP) translation, (1b) internal-external (IE) rotation, (1c) varus-valgus (VV) rotation, (1d) ROM, and (1e) the moment of post-cam engagement. The secondary aim was (2) to assess the impact of JLE on AP translation and post-cam engagement after revision arthroplasty. The hypothesis was that both implants had similar kinematic patterns and that JLE would lead to AP instability with a delayed post-cam engagement.

2 Methodology

2.1

2.1 Study design and setting

After receiving institutional review board approval (BC-11194-AM01), a prospective comparative single-center study was conducted at Ghent University Hospital, Belgium. All patients participating in this study provided informed consent.

2.2

2.2 Participants

This study included 30 patients who underwent rTKA between January 2022 and January 2024, comprising 19 patients with a PS implant and 11 with a CCK implant. Revision arthroplasty was defined as the removal and replacing of the femoral and/or tibial component. Patients with periprosthetic joint infections (PJI) or periprosthetic fractures as indications for revision were excluded. To be eligible for participation, patients had to be six months postoperative (mean 9.4 months (SD, 4.1)), with a minimum knee flexion of 90° and retained the ability to walk. The cohort consisted of 14 women and 16 men, with a mean age at time of revision of 64 years (range, 49 to 82) and a mean body mass index (BMI) of 29.8 kg/m² (range, 22 to 38). A primary revision was performed in 25 patients, while five patients underwent a re-revision, with four of these patients in the CCK group (p = 0.047). Detailed patient characteristics are presented in Table 1, there were no other significant differences between the PS and CCK group.

Table 1 Patient characteristics.

Variable	All patients (n = 30)	PS implant (n = 19)	CCK implant (n = 11)	P-value
Sex				0.919
Men	16 (53.3)	10 (52.6)	6 (54.5)
Women	14 (46.7)	9 (47.4)	5 (45.5)
Age (years)
Primary arthroplasty	57.4 ± 6.9	57.6 ± 7.7	57.0 ± 5.7	0.814^a
Revision arthroplasty	63.6 ± 6.9	63.5 ± 7.0	63.8 ± 6.9	0.897^a
Re-revision: first revision	54 [52.5–66]	53.0	57.0 [52.5–69.0]	0.800†
BMI (kg/m²)	29.8 ± 3.9	29.1 ± 3.8	30.9 ± 4.0	0.246^a
ASA classification				0.507
1	6 (20)	4 (21.1)	2 (18.2)
2	22 (73.3)	13 (68.4)	9 (81.8)
3	2 (6.7)	2 (10.5)	0 (0)
4	0 (0)	0 (0)	0 (0)
Type of primary procedure				0.488
TKA	24 (80)	14 (73.7)	10 (90.9)
UKA	5 (16.7)	4 (21.1)	1 (9.1)
UKA + PF	1 (3.3)	1 (5.3)	0 (0)
Type of revision procedure				0.047
Primary revision	25 (83.3)	18 (94.7)	7 (63.6)
Re-revision	5 (16.7)	1 (5.3)	4 (36.4)
Δ Time primary – revision arthroplasty (months)
Primary revision patients (n=25)	41 [20–95.5]	43 [24.5–119.3]	21 [20–53]	0.423^b
Re-revision patients (n=5)
Primary – first revision	14 [8.5–74]	14	13.5 [6.8–103.5]	1.000^b
First – current revision	85 [28.5–137.5]	94	60.5 [24.8–157]	1.000^b
Indication revision arthroplasty ^c				0.655
Aseptic loosening	10 (33.3)	6 (31.6)	4 (36.4)
Malalignment	12 (40)	8 (42.1)	4 (36.4)
Instability	6 (20)	4 (21.1)	2 (18.2)
Progressive disease	1 (3.3)	1 (5.3)	0 (0)
Polyethylene wear	1 (3.3)	0 (0)	1 (9.1)
Side				1.000
Left	12 (40)	8 (42.1)	4 (36.4)
Right	18 (60)	11 (57.9)	7 (63.6)

a Independent Student T-Test.

b Mann-Whitney U Test.

c Based on definitions from a previously published study (Stroobant L, Ackerman T, Vermue H et al. Revision Knee Arthroplasty in a Tertiary Center: Infection Remains the Leading Cause for Primary Revisions, While Aseptic Loosening Led Causes for Rerevisions. J Arthroplasty. Published online February 19, 2025.https://doi.org/10.1016/j.arth.2025.02.045).

2.3

2.3 Procedure

The revision surgeries were performed by two senior surgeons (NA, JV) using a similar technique.¹⁵ All patients received the Legion Revision implant (Smith & Nephew, Memphis, TN, USA), which is compatible with both the PS and CCK variant of the Genesis II polyethylene insert. The CCK insert differs by having a larger, higher and reinforced central post on the constrained insert. The CCK insert allows approximately 3° of varus/valgus laxity and 4° of internal/external rotation. The indication for selecting the CCK insert was insufficient stability after trialing with a less constrained PS insert.

All patients underwent the same postoperative rehabilitation protocol for outpatient care. Standard follow-up assessments were scheduled at six weeks, three months, six months and 12 months postoperatively.

2.4

2.4 Kinematic analysis

Fluoroscopic videos were obtained in the sagittal plane using a flat-panel fluoroscopy system (Siemens Axiom Luminos dRF, Erlangen, Germany) at 15 frames per second. The patients performed three exercises, including one open-chain and two closed-chain exercises. The open kinetic chain exercise involved flexion-extension (FE) from a seated position, with patients instructed to extend and flex the knee maximally. As the knee passed the same flexion angle twice, the average of the two measurements was calculated. The two closed kinetic chain exercises included the sit-to-stand (STS) test and the squatting (SQ) exercise. For the STS, the patients began with the knee in 90 degrees of flexion while seated on a chair. Lastly, the squatting exercise was performed to maximal knee flexion (Fig. 1). During the exercises the patients were allowed to use limited support from a handrail. The knee under investigation was positioned closest to the radiographic table during each exercise. Patients performed each exercise three times, with the best sequence selected. The best sequence was defined as the exercise with the most consistent execution, clearly capturing both the beginning and the end of the movement as observed by the fluoroscope.

Patient performing a squatting exercise while under fluoroscopic surveillance.

The fluoroscopic videos were cropped to isolate the relevant frames of interest. Kinematic analysis was performed using the fluoroscopic images with a standard 2-D to 3-D image registration technique, employing the open-source software package JointTrack Auto (University of Florida, Gainesville, FL, USA).¹⁶ In this technique, computer-aided design models – based on the implant component size – were superimposed onto the fluoroscopic images based on their silhouette, allowing for the generation of transformation matrices for the components in the image space. This method was recently validated, demonstrating errors ranging from 0.37° to 1.2° for in-plane rotations, with an average of 0.73°, and from 0.72 mm to 0.92 mm for translations, with an average of 0.81 mm.¹⁶ Given a polyethylene insert is nearly invisible on fluoroscopic images, it was assumed to be rigidly attached to the tibial components, consistent with the fixed bearing design. For both the femur and tibia a joint coordinate system was defined based on the mathematical framework of Grood and Suntay.¹⁷ Kinematics were assessed during three flexion ranges: early flexion (0°–30°), mid-flexion (30°–60°) and deep flexion (60°–90°). Additionally, the difference between the endpoint and the starting point of each exercise was calculated, providing information of the entire ROM.1a.AP translation

Landmarks were defined on the femoral and tibial components. The femoral origins for medial and lateral condyle translation were defined by the center of spheres, fitting the bearing surfaces of the femoral condyles. These two points were then plotted relative to a tibial reference frame, with 0 % and 100 % corresponding to the most posterior and anterior points of the tibia respectively (Fig. 2). This normalization ensured that the values were independent of tibial size.1b.IE rotation

The AP dimension of the tibial insert is shown, with 0 % and 100 % corresponding to the most posterior and anterior points of the tibia, respectively. On the right, an overview of the different tibial sizes and their corresponding AP lengths is provided. AP, Anteroposterior.

IE rotation was defined as the angle created between the line joining the medial and lateral femoral origins (=mediolateral femoral axis) and the line connecting the medial and lateral tibial plateau (=mediolateral tibial axis), projected onto the transverse plane of the tibia implant. Rotation is considered negative when the femur is externally rotated and positive when internally rotated relative to the tibia.1c.VV rotation

VV rotation was defined as the angle between the mediolateral femoral axis and the mediolateral tibial axis in the frontal plane. A negative degree was defined as varus, and a positive degree as valgus.1d.ROM

The flexion angle was calculated as the angular difference between the mechanical axis of the femur (= line between hip center and intercondylar notch) and the mechanical axis of the tibia (= line between ankle center and tibial spine) in the sagittal plane. The ROM was assessed during the SQ exercise by measuring the difference between maximum extension and flexion, with an accuracy of 5°.1e.Post-cam engagement

Post-cam engagement was defined as a distance of less than one mm between the post and cam, evaluated at five degrees increments in flexion.2.Effect of joint line elevation on knee kinematics

Postoperative joint line height was measured using a scaled weight-bearing radiograph, preferably a full-leg (FL) view (n = 27), or an AP view (n = 3), based on availability. A line was drawn from the superior aspect of the fibular head, perpendicular to the tibial axis, and the perpendicular distance to the inferior femoral component was recorded.¹³^,¹⁸^,¹⁹ Given the similarity of the contralateral knee's joint line (mean difference 0.0, range −1.1 to 1.1), it was evaluated on the same radiograph (for FL views) to estimate the native joint line of the revision knee.²⁰^,²¹ If the contralateral knee had a prosthesis (n = 12), an alternative native radiograph was used. The joint line change was calculated as difference between postoperative and native joint line height, with positive values indicating proximalization and negative values indicating distalization. Grouping was categorized based on literature indicating that joint line elevation (JLE) of 4 mm or more is associated with poorer outcomes.²¹ Patients with a distalization of more than 4 mm were excluded from the analysis (n = 3).

2.5

2.5 Data analyses

A power analysis was not conducted due to absence of similar studies in revision patients. Based on kinematic studies involving pTKA patients, a sample size of 30 was proposed, with a minimum of 10 patients in each group.⁴^,⁷^,¹¹^,²²

Statistical analyses were performed using SPSS (SPSS version 29, Chicago, IL, United States). Descriptive data are reported as mean and standard deviation (SD) or median and interquartile range (IQR) for continuous variables, based on normality tests (Shapiro-Wilk), and count (percentage) for categorical data. Fluoroscopic data are all reported as means and SDs for uniformity. Differences between groups were analyzed using t-test or Mann-Whitney U tests for continuous variables and Chi-square or Fisher's exact tests for categorical variables. The Chi-square test was chosen when less than 20 % of the cells in the contingency table had expected frequencies of less than five, and all expected frequencies were above one. The level of significance was set at 0.05.

3 Results

(1)Kinematic analysis: PS versus CCK1a)AP translation

From full extension to maximum flexion in the medial compartment, the average posterior femoral rollback (PFR) was 20.64 % (SD, 8.40) relative to the AP distance of the tibial component for the PS group and 22.17 % (SD, 7.94) for the CCK group during the squatting exercise (p = 0.653). During open-chain FE, these values were and 13.96 % (SD, 9.49) and 16.24 % (SD, 6.29) for PS and CCK patients, respectively (p = 0.485). During the STS exercise, an anterior femoral displacement of 15.87 % (SD, 7.11) was observed in the PS group and 12.26 % (SD, 7.82) in the CCK group (p = 0.220). No statistically significant differences were observed between the AP positions of CCK and PS implants across the different flexion ranges, regardless of the exercise performed (Table 2, Fig. 3).

Table 2 Comparison of AP translation (expressed as percentage based on the tibial reference frame) between revision TKA patients with a PS implant and a CCK implant.

Compartment	Exercise	Flexion range (°)	PS implant (%)	CCK implant (%)	P-value
Medial	SQ	0–30	50.76 ± 4.82 (n=14)	49.63 ± 5.69 (n=8)	0.627^a
		30–60	49.84 ± 5.17 (n=17)	49.09 ± 4.26 (n=9)	0.833^b
		60–90	39.52 ± 3.01 (n=19)	38.98 ± 5.73 (n=9)	0.744^a
		fROM (Extension→ Flexion)	−20.64 ± 8.40 (n = 18)	−22.17 ± 7.94 (n = 9)	0.653^a
	FE	0–30	49.27 ± 5.31 (n=17)	47.89 ± 2.49 (n=11)	0.427^a
		30–60	51.62 ± 4.61 (n=19)	50.08 ± 2.64 (n=11)	0.321^a
		60–90	41.44 ± 3.99 (n=18)	39.06 ± 3.36 (n=11)	0.111^a
		fROM (Flexion→ Extension)	+13.96 ± 9.49 (n = 19)	+16.24 ± 6.29 (n = 1 1 )	0.485^a
	STS	0–30	49.83 ± 6.69 (n=19)	51.18 ± 6.77 (n=10)	0.610^a
		30–60	50.28 ± 4.11 (n=19)	49.36 ± 3.85 (n=10)	0.565^a
		60–90	40.81 ± 3.77 (n=19)	39.96 ± 5.03 (n=10)	0.610^a
		fROM (Flexion→ Extension)	+15.87 ± 7.11 (n = 19)	+12.26 ± 7.82 (n = 10)	0.220^a
Lateral	SQ	0–30	47.98 ± 7.12 (n=14)	45.01 ± 9.28 (n=8)	0.409^a
		30–60	46.26 ± 5.67 (n=17)	44.26 ± 8.41 (n=9)	0.475^a
		60–90	34.48 ± 4.81 (n=19)	34.87 ± 5.19 (n=9)	0.849^a
		fROM (Extension→ Flexion)	−24.17 ± 10.02 (n = 19)	−21.80 ± 9.56 (n = 9)	0.558^a
	FE	0–30	46.81 ± 3.71 (n=17)	45.99 ± 4.30 (n=11)	0.598^a
		30–60	47.87 ± 4.25 (n=19)	45.07 ± 3.99 (n=11)	0.086^a
		60–90	35.41 ± 4.30 (n=18)	32.42 ± 4.91 (n=11)	0.096^a
		fROM (Flexion→ Extension)	+20.32 ± 12.49 (n = 19)	+24.09 ± 11.92 (n = 11)	0.471^b
	STS	0–30	49.41 ± 5.91 (n=19)	49.25 ± 5.19 (n=10)	0.943^a
		30–60	46.98 ± 4.38 (n=19)	46.31 ± 4.30 (n=10)	0.698^a
		60–90	35.62 ± 4.96 (n=19)	36.14 ± 3.93 (n=10)	0.776^a
		fROM (Flexion→ Extension)	+22.49 ± 9.37 (n = 19)	+17.64 ± 8.38 (n = 10)	0.211

a Independent Student T-Test.

b Mann-Whitney U Test.

AP translation of the medial (superior row) and lateral condyles (inferior row) for CCK (black) and PS (red) implants, during three exercises, with respect to the flexion angle. The line represents the mean value, and the shaded area denotes the 95 % confidence interval. AP translation is expressed as a percentage based on the tibial reference frame. The starting point of the x-axis is 10°. AP, Anteroposterior. CCK, Condylar Constrained Knee. PS, Posterior-Stabilized. %, Percentage. °, Degrees.

Regarding the lateral condyle, a PFR of 24.17 % (SD, 10.02) was observed in the PS group and 21.80 % (SD, 9.56) in the CCK group during squatting (p = 0.558). A similar posterior displacement was observed during FE from extension to flexion (PS: 20.32 (SD, 12.49), CCK: 24.09 (SD, 11.92); p = 0.471). Greater anterior displacement was noted during the STS movement in PS patients (22.49 %, SD 9.37) versus CCK patients (17.64 %, SD, 8.38), although the difference was not statistically significant (p = 0.211). Again, there were no statistically significant differences in AP positions of CCK and PS implants across the various flexion ranges, irrespective of the exercise performed (Table 2, Fig. 3).

Since no differences were observed between CCK and PS patients, a comparison was made between the medial and lateral compartments for the entire cohort. The results revealed a greater AP displacement in the lateral compartment compared to the medial compartment. This difference was statistically significant in both the FE and STS exercises (p < 0.001). Specifically, during FE, the lateral compartment showed a displacement of 21.7 % (SD, 12.22), while the medial compartment had a displacement of 14.79 % (SD, 8.41 (p < 0.001). Similarly, in the STS exercise, the lateral compartment exhibited a displacement of 20.81 % (SD, 9.19) compared to 14.63 % (SD, 7.43) in the medial compartment (p < 0.001). The lateral displacement of 23.41 % (SD, 9.95) was not statistically higher than the medial displacement of 21.15 % (SD, 8.13) during the squatting (p = 0.125).1b)IE rotation

During the squatting movement, the femoral component exhibited an average external rotation of 2.71° (SD, 6.48) from maximum extension to flexion in the PS group, while the CCK group showed a slight internal rotation of 0.10° (SD, 4.12) (p = 0.248). Both groups demonstrated comparable external rotation from extension to flexion during the FE exercise (p = 0.588). Additionally, during the STS transition, the femoral component underwent approximately 5° of internal rotation from seated position to extension in both groups (p = 0.728) (Table 3, Fig. 4).1c)VV rotation

Table 3 Comparison of internal-external rotation between revision TKA patients with a PS implant and a CCK implant.

Exercise	Flexion range (°)	PS implant (°)	CCK implant (°)	P-value
SQ	0–30	−2.15 ± 7.10 (n=15)	−3.8 ± 7.90 (n=8)	0.616^a
	30–60	−3.07 ± 5.93 (n=17)	−4.07 ± 8.79 (n=9)	0.732^a
	60–90	−4.30 ± 5.14 (n=19)	−3.57 ± 8.24 (n=9)	0.777^a
	fROM (Extension→ Flexion)	−2.71 ± 6.48 (n = 19)	0.10 ± 4.12 (n = 9)	0.248^a
FE	0–30	−2.02 ± 6.06 (n=17)	−1.57 ± 4.12 (n=11)	0.832^a
	30–60	−3.19 ± 5.30 (n=19)	−4.20 ± 3.99 (n=11)	0.586^a
	60–90	−5.10 ± 4.77 (n=18)	−5.62 ± 5.38 (n=11)	0.792^a
	fROM (Flexion→ Extension)	+5.33 ± 5.10 (n = 19)	+6.63 ± 7.93 (n = 11)	0.588^a
STS	0–30	−0.42 ± 3.56 (n=19)	−1.61 ± 3.00 (n=10)	0.377^a
	30–60	−2.84 ± 3.78 (n=19)	−2.64 ± 4.01 (n=10)	0.895^a
	60–90	−4.44 ± 6.10 (n=19)	−3.34 ± 5.28 (n=10)	0.377^b
	fROM (Flexion→ Extension)	+5.59 ± 8.21 (n = 19)	+4.59 ± 5.04 (n = 10)	0.728^a

a Independent Student T-Test.

b Mann-Whitney U Test.

IE rotation for CCK (black) and PS (red) implants, during three exercises, with respect to the flexion angle. The line represents the mean value, and the shaded area denotes the 95 % confidence interval. The starting point of the x-axis is 10°. IE, Internal-external. CCK, Condylar Constrained Knee. PS, Posterior-Stabilized. %, Percentage. °, Degrees.

Regarding VV rotation, no statistically significant differences were observed between the groups across the different flexion ranges during both open- and closed-chain exercises. Throughout the full range of motion, VV rotation changed minimal (Table 4, Fig. 5).1d)Closed-chain ROM

Table 4 Comparison of varus-valgus rotation between revision TKA patients with a PS implant and a CCK implant.

Exercise	Flexion range (°)	PS implant (°)	CCK implant (°)	P-value
SQ	0–30	0.28 ± 1.13 (n=14)	0.66 ± 2.37 (n=8)	0.607^a
	30–60	1.15 ± 1.00 (n=17)	0.81 ± 1.31 (n=9)	0.200^b
	60–90	1.39 ± 1.22 (n=19)	1.01 ± 1.09 (n=9)	0.410^b
	fROM (Extension→ Flexion)	+0.93 ± 2.5 5 (n=18)	+0.66 ± 3.25 (n=9)	0.813^a
FE	0–30	0.50 ± 0.89 (n=17)	0.63 ± 0.92 (n=11)	0.721^a
	30–60	0.98 ± 1.36 (n=19)	1.48 ± 1.10 (n=11)	0.145^b
	60–90	1.16 ± 1.42 (n=18)	1.70 ± 1.81 (n=11)	0.611b
	fROM (Flexion→ Extension)	−0.2 0± 2.81(n=19)	−1.33 ± 2.29 (n=11)	0.269^a
STS	0–30	0.20 ± 1.12 (n=19)	0.55 ± 0.79 (n=10)	0.382^a
	30–60	0.79 ± 1.21 (n=19)	1.10 ± 0.87 (n=10)	0.485^a
	60–90	1.03 ± 1.24 (n=19)	1.16 ± 0.79 (n=10)	0.762^a
	fROM (Flexion→ Extension)	0.89 ± 2.65 (n=19)	0.78 ± 1.82 (n=10)	0.908^a

a Independent Student T-Test.

b Mann-Whitney U Test.

VV rotation for CCK (black) and PS (red) implants, during three exercises, with respect to the flexion angle. The line represents the mean value, and the shaded area denotes the 95 % confidence interval. The starting point of the x-axis is 10°. VV, Varus-valgus. CCK, Condylar Constrained Knee. PS, Posterior-Stabilized. %, Percentage. °, Degrees.

The average maximum ROM during the squatting exercise in this study was 77.11° (SD, 16.86) for the PS patients, and 76.11° (SD, 20.73) for the CCK patients (p = 0.893).1e)Post-cam engagement

The flexion angle at which post-cam engagement occurred was comparable between PS and CCK patients during both open- (p = 0.654) and closed-chain (p = 0.901) exercises (Table 5). Across the entire cohort, the post-cam engagement occurred at a significantly lower flexion angle during open-chain exercises compared to closed-chain exercises (p = 0.032).(2)Effect of joint line elevation on knee kinematics

Table 5 Comparison of post-cam engagement between revision TKA patients with a PS implant and a CCK implant.

Exercise	PS implant (°)	CCK implant (°)	P-value
Closed-Chain	47.29 ± 10.67 (n=19)	49.05 ± 9.32 (n=11)	0.654
Open-Chain	44.44 ± 10.95 (n=19)	44.08 ± 4.86 (n=11)	0.901

A significantly more anterior position in the medial compartment during squatting between 0 and 60° was observed in patients with proximalization of more than 4 mm (n = 8, 26.7 %), compared to those without (n = 19, 63.3 %) (0–30°: p = 0.037; 30–60°: p = 0.021). No other significant differences were found (Table 6, Fig. 6). During closed-chain exercises, the post-cam engagement occurred at 52.31° (SD, 11.77) in the JLE group compared to 46.32° (SD, 9.42) in the non-JLE group, although without statistical significance (p = 0.173). During FE, post-cam engagement was similar in both groups (non-JLE: 44.56° (SD, 10.65); JLE: 42.96° (SD, 4.71); p = 0.594).

Table 6 Comparison of AP translation (expressed as percentage based on the tibial reference frame) between revision TKA patients with joint line proximalisation ≥4 mm and without a joint line elevation (−4 mm < JLE <4 mm).

Compartment	Exercise	Flexion range (°)	−4 mm < JLE <4 mm (n = 19) (%)	JLE ≥4 mm (n = 8)(%)	P-value
Medial	SQ	0–30	48.29 ± 4.42 (n=13)	53.60 ± 5.47 (n=6)	0.037 ^a
		30–60	48.04 ± 2.97 (n=17)	53.42 ± 7.68 (n=6)	0.021 ^a
		60–90	38.75 ± 4.23 (n=19)	41.02 ± 3.61 (n=6)	0.248^a
		fROM (Extension→ Flexion)	−19.20 ± 7.78 (n=19)	−25.95 ± 8.22 (n=6)	0.080^a
	FE	0–30	48.77 ± 4.76 (n=19)	47.45 ± 3.36 (n=7)	0.509^a
		30–60	51.52 ± 4.31 (n=19)	49.65 ± 3.66 (n=8)	0.292^a
		60–90	41.23 ± 3.90 (n=18)	38.86 ± 3.70 (n=8)	0.160^a
		fROM (Flexion→ Extension)	+12.14 ± 8.98 (n=19)	+19.04 ± 5.66 (n=8)	0.057^a
	STS	0–30	49.58 ± 5.55 (n=18)	50.24 ± 9.06 (n=8)	0.821^a
		30–60	49.76 ± 3.57 (n=18)	48.95 ± 3.83 (n=8)	0.609^a
		60–90	41.09 ± 3.84 (n=18)	39.63 ± 4.87 (n=8)	0.416^a
		fROM (Flexion→ Extension)	+12.78 ± 6.29 (n=18)	+14.83 ± 7.47 (n=8)	0.429^b
Lateral	SQ	0–30	46.95 ± 9.06 (n=13)	46.40 ± 7.04 (n=6)	0.964^a
		30–60	45.40 ± 7.71 (n=17)	45.15 ± 4.51 (n=6)	0.943^a
		60–90	34.49 ± 5.59 (n=19)	33.75 ± 2.35 (n=6)	0.756^a
		fROM (Extension→ flexion)	−21.80 ± 9.15 (n=19)	−26.86 ± 13.24 (n=6)	0.299^a
	FE	0–30	46.34 ± 4.51 (n=19)	47.16 ± 2.50 (n=7)	0.655^a
		30–60	46.99 ± 4.73 (n=19)	46.65 ± 4.21 (n=8)	0.862^a
		60–90	34.68 ± 5.25 (n=18)	33.79 ± 4.27 (n=8)	0.678^a
		fROM (Flexion→ Extension)	+18.05 ± 12.95 (n=19)	+28.49 ± 9.20 (n=8)	0.050^a
	STS	0–30	48.65 ± 4.39 (n=18)	49.87 ± 8.36 (n=8)	0.629^a
		30–60	45.99 ± 3.57(n=18)	46.94 ± 5.62 (n=8)	0.978^b
		60–90	34.74 ± 4.71 (n=18)	37.02 ± 4.45 (n=8)	0.765^b
		fROM (Flexion→ Extension)	+19.95 ± 10.78 (n=18)	+20.49 ± 5.83 (n=8)	0.461^b

a Independent Student T-Test.

b Mann-Whitney U Test.

AP translation of the medial condyle for patients with proximalization of the joint line (black) and no proximalization of the joint line (red), during three exercises, with respect to the flexion angle. The line represents the mean value, and the shaded area denotes the 95 % confidence interval. AP translation is expressed as a percentage based on the tibial reference frame. The starting point of the x-axis is 10°. AP, Anteroposterior. %, Percentage. °, Degrees.

4 Discussion

Although kinematics in pTKA have been extensively studied, these findings cannot be generalized to revision patients. These patients are typically characterized by ligament alterations, instability, scarring, bone loss and muscular insufficiency, requiring higher constraints and complicating joint line restoration. To the authors’ knowledge, this is the first study to investigate kinematics in a revision population, comparing two different constraints. The principal findings are: (1) No significant differences were observed in kinematic patterns, including post-cam engagement, between PS and CCK implants in revision knee arthroplasty patients. (2) A JLE ≥4 mm leads to significant instability during early and mid-flexion during squatting. A delayed post-cam engagement is observed compared to patients who had less than 4 mm change in joint line height, although this was not statistically significant.(1)Kinematic analysis: PS versus CCK

The literature comparing PS and CCK implants reports mixed findings, with some groups reporting no differences in clinical outcomes,²³ while other highlighting increased complications and motion restrictions.¹¹ Overconstraining may lead to complications such as loosening and wear.¹¹ Therefore, selecting the least constrained implant is recommended to maintain soft tissue influence on kinematics and minimize stress at the bone-implant surface.²⁴ Building on this rationale, our study investigated whether the additional stability of the CCK inserts allowed for unrestricted motion in rTKA patients. In our study, the CCK implant revealed its efficiency in the frontal plane without impacting other movements, thereby supporting previous studies that reported comparable results between PS and CCK implants.⁴ Both implants showed similar normal-like kinematic patterns, pertaining to PFR, axial rotation patterns, and weight-bearing flexion.¹¹ Although comparing our results with the literature is challenging due to variations in surgical technique and implant design, our study focused on comparing the two types of constraints rather than absolute values.²⁵1a)AP translation

The kinematic pattern during squatting was similar between both groups. Medially, the mean posterior shift during squatting was 20.64 % for PS and 22.17 % for CCK implants, corresponding to an absolute shift of 8.4–13 mm, depending on the tibial component size. These values are similar to the 14 mm mean PFR reported by Victor et al. for the Journey Cruciate substituting (BCS) implant (Smith and Nephew, Memphis, TN, USA).²⁶^,²⁷ More recently, LaCour et al. observed shifts of 7.7 mm for the traditional and 10.8 mm for the constrained insert of the Journey II BCS (Smith and Nephew, Memphis, TN, USA).¹¹ However, these translations are higher than those observed in the native knee, raising concerns about the potential overloading the medial compartment.²⁶^,²⁸^,²⁹ Laterally, our results align with those of LaCour et al.,¹¹ showing a smaller shift compared to the native knee, consistent with current literature.¹⁰^,¹¹^,²⁹ As the PFR prevents posterior impingement, this might explain why knee flexion is reduced following TKA compared to the native knee.²⁹

Similarly, during both FE and STS movements, the kinematic patterns were comparable between both constraints, suggesting that our findings are applicable to both open- and closed-chain exercises. During these exercises the PFR was significant greater in the lateral compartment compared to the medial compartment resulting in a medial pivot kinematic pattern with femoral external rotation as knee flexion progresses.¹⁰^,²⁶^,²⁹^,³⁰1b)IE rotation

No significant difference was found between PS and CCK implants in terms of IE rotation. Both showed increasing femoral external rotation with flexion, following the same pattern as the native knee, although smaller in total rotation.¹⁰^,³⁰1c)VV rotation

No significant differences were observed between CCK and PS implants, suggesting that PS patients may compensate for constraints through intrinsic ligamentous structures, while the CCK design inherently limits this.³¹1d)Closed-chain ROM

Both the CCK and PS implants exhibited a ROM of approximately 77° during squatting, compared to the average range of 105–110° reported in the literature.¹¹^,³⁰ This discrepancy can be attributed to potential absence of full extension images, as a minimum flexion of 90° was an inclusion criterion. Nonetheless, revision patients are at higher risk for reduced ROM, and insufficient quadriceps strength may still be present after nine months (mean follow-up). However, both implants exhibited similar ROM, suggesting no inferiority of the CCK compared to the PS implant.³²1e)Post-cam engagement

The moment of post-cam engagement was similar between the PS and CCK implants. After this engagement, the pronounced and controlled femoral rollback was initiated. In the native knee, the posterior cruciate ligament becomes functional beyond 50 degrees of flexion.⁷ Since the post serves as a substitute for the PCL, it is suggested that the optimal initial contact angle should be around 50° of flexion.⁷^,³³ In the literature, values are mainly reported between 50 and 90 degrees of flexion.³³^,³⁴(2)Effect of joint line elevation on knee kinematics

During closed-chain exercises, patients with joint line proximalization demonstrated a significantly more anterior femoral position in early and mid-flexion, followed by a delayed post-cam engagement. This anterior position reduces the knee extension lever arm, thereby possibly affecting patient satisfaction.⁸ Although post-cam engagement occurred later, both groups ultimately reached the same endpoint (Fig. 6). This is suggestive for a greater posterior translation in mid-flexion, which could contribute to mid-flexion instability.⁷ As noted by Van Onsem et al., anterior femoral positioning in early and mid-flexion is linked to lower patient-reported outcome measures (PROMs), and literature indicates that a JLE of more than 4 mm may also result in lower PROMs.⁷^,²¹ However, further research with a larger patient population is needed to confirm this assumption and establish a direct correlation between JLE, kinematics and PROMs.

4.1

4.1 Limitations

Several limitations of this study should be acknowledged. The absence of randomization introduced the potential for selection bias during recruitment. However, cohort demographics were largely similar. In addition, given the nature of this study, the sample size was small, and larger patient groups would enable more robust comparisons. Furthermore, this is a short-term study without long-term information, though the primary aim was to assess early kinematics following rTKA. Additionally, implant alignment, coronal stability, and perioperative information of the soft tissue conditions were not included in the analysis. Further studies are needed to explore the impact of these factors. It is also important to note that the results are specific to the prosthetic design used in this study. Other designs, such as those with a rotating platform, may exhibit entirely different kinematic patterns.

Despite these limitations, this is the first study to provide valuable insights into kinematic patterns of revision arthroplasty patients. It offers a thorough analysis of kinematics during both open- and closed-chain exercises, including post-cam engagement, and compares two types of constraints. Furthermore, the impact of joint line position on kinematics was assessed.

4.2

4.2 Future perspectives

Looking forward, linking kinematic patterns to PROMs could offer further understanding of how knee kinematics impact functional outcomes and overall patient satisfaction, specifically in revision TKA.³⁰

5 Conclusion

In conclusion, this study offers valuable insights into the kinematic patterns of rTKA, comparing PS and CCK implants during both open- and closed-chain exercises. Both implants exhibited a similar kinematic pattern in terms of AP translation, IE rotation, VV rotation, ROM, and post-cam engagement. The CCK implant demonstrated effective performance in the frontal plane without restricting other movements. These findings are encouraging and reinforce the rationale for selecting semi-constrained implants in TKA revision when indicated. Additionally, the study found a significantly greater anterior translation in early and mid-flexion, along with a delayed post-cam engagement in patients with a joint line proximalization ≥4 mm.

CRediT authorship contribution statement

Lenka Stroobant: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Methodology, Visualization, Writing – original draft. Hannes Vermue: Investigation, Methodology, Writing – review & editing. Ewoud Jacobs: Writing – review & editing. Nele Arnout: Writing – review & editing. Stefaan Van Onsem: Methodology, Supervision, Writing – review & editing. Scott A. Banks: Conceptualization, Formal analysis, Resources, Writing – review & editing. Jan Victor: Funding acquisition, Supervision, Writing – review & editing. Amélie Chevalier: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization.

Data availability

The data that support the findings of this study are available on reasonable request from the corresponding author.

Ethical approval and patient consent

After receiving institutional review board approval (BC-11194-AM01), a prospective comparative single-center study was conducted at Ghent University Hospital, Belgium. The study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki (2013). Informed consent was obtained from all patients participating in the study.

Funding

This project was supported by grant number T001620N from ‘Research Foundation Flanders’, awarded to Jan Victor.

Author contribution statement

J Victor received Royalties from Smith + Nephew, was Speakers bureau/paid presentations for a Medacta, Research support from Moximed, Board member/committee appointments for BVOT, BKS, EKS. Nele Arnout was Board member/committee appointments for Belgian Knee Society board. Scott A. Banks received Royalties from DJO Surgical, Stryker, Speakers bureau/paid presentations for DJO Surgical, Stock or stock options in Scientific Motion Technologies, Orthopaedic Driven Imaging, Board member/committee appointments for Research Committee, The Knee Society. Stefaan Van Onsem was Paid consultant for Stryker, Medacta, Stock or stock options in Astriqs, Surgical sensors, Board member/committee appointments for Belgian Knee Society. The rest of the authors declare no competing interests.

References

Klug A., Gramlich Y., Rudert M., et al . The projected volume of primary and revision total knee arthroplasty will place an immense burden on future health care systems over the next 30 years. Knee Surg Sports Traumatol Arthrosc. 2021;29:3287-3298.
[Google Scholar]
Brown L.R., Clement N.D., MacDonald D.J., Breusch S.J., . The survivorship of the link endo-rotational hinge total knee arthroplasty: 5–12-year follow-up of 100 patients. Arch Orthop Trauma Surg. 2019;139:107-112.
[Google Scholar]
Shichman I., Askew N., Habibi A., et al . Projections and epidemiology of revision hip and knee arthroplasty in the United States to 2040-2060. Arthroplast Today. 2023;21
[Google Scholar]
Batailler C., Foissey C., Fary C., Naaim A., Servien E., Lustig S., . Similar kinematic patterns between revision total stabilized (TS) and primary posterior stabilized (PS) knee prostheses: a prospective case–controlled study with gait assessment. Knee Surg Sports Traumatol Arthrosc. 2022;30:2714-2722.
[Google Scholar]
Saleh K.J., Dykes D.C., Tweedie R.L., et al . Functional outcome after total knee arthroplasty revision: a meta-analysis. J Arthroplast. 2002;17:967-977.
[Google Scholar]
Baker P., Cowling P., Kurtz S., Jameson S., Gregg P., Deehan D., . Reason for revision influences early patient outcomes after aseptic knee revision knee. Clin Orthop Relat Res. 2012;470:2244-2252.
[Google Scholar]
Van Onsem S., Verstraete M., Van Eenoo W., Van Der Straeten C., Victor J., . Are TKA kinematics during closed kinetic chain exercises associated with patient-reported outcomes? A preliminary analysis. Clin Orthop Relat Res. 2020;478:255-263.
[Google Scholar]
Banks S.A., Catani F., Deckard E.R., et al . Total knee arthroplasty kinematics predict patient-reported outcome measures: implications for clinical kinematic examinations. J Arthroplast. 2024;39:S224-S229.
[Google Scholar]
Sekeitto A.R., McGale J.G., Montgomery L.A., Vasarhelyi E.M., Willing R., Lanting B.A., . Posterior-stabilized total knee arthroplasty kinematics and joint laxity: a hybrid biomechanical study. Arthroplasty. 2022;4:53.
[Google Scholar]
Angerame M.R., Holst D.C., Jennings J.M., Komistek R.D., Dennis D.A., . Total knee arthroplasty kinematics. J Arthroplast. 2019;34:2502-2510.
[Google Scholar]
LaCour M.T., Dessinger G.M., Haas S.B., Komistek R.D., . In vivo weight-bearing kinematics for constrained versus traditional bicruciate stabilized total knee arthroplasty cohorts compared to the normal knee. J Arthroplast. 2024;39:1589-1594.
[Google Scholar]
Bellemans J., . Restoring the joint line in revision TKA: does it matter? Knee. 2004;11:3-5.
[Google Scholar]
Watts D.T., Kendoff D., Sussmann P., Klauser W., Heinert G., . A comparison of joint line height after revision total knee replacement using a rotating hinge versus a condylar revision knee system. J Clin Orthop Trauma. 2023;47
[Google Scholar]
van Lieshout W.A.M., Valkering K.P., Koenraadt K.L.M., van Etten-Jamaludin F.S., Kerkhoffs G.M.M.J., van Geenen R.C.I., . The negative effect of joint line elevation after total knee arthroplasty on outcome. Knee Surg Sports Traumatol Arthrosc. 2019;27:1477-1486.
[Google Scholar]
Hardeman F., Londers J., Favril A., Witvrouw E., Bellemans J., Victor J., . Predisposing factors which are relevant for the clinical outcome after revision total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc. 2012;20:1049-1056.
[Google Scholar]
Jensen A.J., Flood P.D.L., Palm-Vlasak L.S., et al . Joint track machine learning: an autonomous method of measuring total knee arthroplasty kinematics from single-plane X-ray images. J Arthroplast. 2023;38:2068-2074.
[Google Scholar]
Grood E.S., Suntay W.J., . A joint coordinate system for the clinical description of three-dimensional motions: application to the knee. J Biomech Eng. 1983;105:136-144.
[Google Scholar]
Snider M.G., MacDonald S.J., . The influence of the posterior cruciate ligament and component design on joint line position after primary total knee arthroplasty. J Arthroplast. 2009;24:1093-1098.
[Google Scholar]
Babazadeh S., Dowsey M.M., Swan J.D., Stoney J.D., Choong P.F.M., . Joint line position correlates with function after primary total knee replacement. J Bone Joint Surg Br. 2011;93-B:1223-1231.
[Google Scholar]
Maderbacher G., Keshmiri A., Zeman F., Grifka J., Baier C., . Assessing joint line positions by means of the contralateral knee: a new approach for planning knee revision surgery? Knee Surg Sports Traumatol Arthrosc. 2015;23:3244-3250.
[Google Scholar]
Clavé A., Le Henaff G., Roger T., Maisongrosse P., Mabit C., Dubrana F., . Joint line level in revision total knee replacement: assessment and functional results with an average of seven years follow-up. Int Orthop. 2016;40:1655-1662.
[Google Scholar]
Kono K., Inui H., Tomita T., et al . Bicruciate-stabilised total knee arthroplasty provides good functional stability during high-flexion weight-bearing activities. Knee Surg Sports Traumatol Arthrosc. 2019;27:2096-2103.
[Google Scholar]
Sumino T., Tomita T., Sugamoto K., Yamazaki T., Okazaki K., . Semi-constrained posterior stabilized total knee arthroplasty reproduces natural deep knee bending kinematics. BMC Muscoskelet Disord. 2020;21
[Google Scholar]
Stroobant L., de Taeye T., Byttebier P., et al . Condylar constrained and rotating hinged implants in revision knee arthroplasty show similar survivorship and clinical outcome: a systematic review and meta-analysis. Knee Surg Sports Traumatol Arthrosc 2023
[Google Scholar]
Digennaro V., Zambianchi F., Marcovigi A., Mugnai R., Fiacchi F., Catani F., . Design and kinematics in total knee arthroplasty. Int Orthop. 2014;38:227-233.
[Google Scholar]
Victor J., Mueller J.K.P., Komistek R.D., Sharma A., Nadaud M.C., Bellemans J., . In vivo kinematics after a cruciate-substituting TKA. Clin Orthop Relat Res. 2010;468:807-814.
[Google Scholar]
Victor J., Banks S., Bellemans J., Victor J., Banks S., Bellemans J., . Kinematics of posterior cruciate ligament-retaining and-substituting total knee arthroplasty A PROSPECTIVE RANDOMISED OUTCOME STUDY. J Bone Joint Surg 2005
[Google Scholar]
Komistek R.D., Dennis D.A., Mahfouz M., . In vivo fluoroscopic analysis of the normal human knee. Clin Orthop Relat Res. 2003;410:69-81.
[Google Scholar]
Dennis D.A., Komistek R.D., Mahfouz M.R., Haas B.D., Stiehl J.B., . Multicenter Determination of in Vivo Kinematics after Total Knee Arthroplasty. Clin Orthop Relat Res. 2003;vol. 416:37-57.
[Google Scholar]
Khasian M., LaCour M.T., Coomer S.C., Bolognesi M.P., Komistek R.D., . In vivo knee kinematics for a cruciate sacrificing total knee arthroplasty having both a symmetrical femoral and tibial component. J Arthroplast. 2020;35:1712-1719.
[Google Scholar]
Deshmukh A.J., Rathod P.A., Moses M.J., Snir N., Marwin S.E., Dayan A.J., . Does a non-stemmed constrained condylar prosthesis predispose to early failure of primary total knee arthroplasty? Knee Surg Sports Traumatol Arthrosc. 2016;24:3194-3199.
[Google Scholar]
King B.R., Gladnick B.P., Lee Y yu, Lyman S., Della Valle A.G., . Range of motion and function are not affected by increased post constraint in patients undergoing posterior stabilized total knee arthroplasty. Knee. 2014;21:194-198.
[Google Scholar]
Arnout N., Vanlommel L., Vanlommel J., et al . Post-cam mechanics and tibiofemoral kinematics: a dynamic in vitro analysis of eight posterior-stabilized total knee designs. Knee Surg Sports Traumatol Arthrosc. 2015;23:3343-3353.
[Google Scholar]
Mihalko W.M., Lowell J., Higgs G., Kurtz S., . Total knee post-cam design variations and their effects on kinematics and wear patterns. Orthopedics. 2016;39
[Google Scholar]