Introduction

Knee osteoarthritis (OA), one of the most common orthopedic diseases, is a leading cause of pain and a huge burden on the healthcare system [1,2,3]. Unicompartmental knee arthroplasty (UKA) is a successful and reliable treatment option for end-stage OA of the knee (especially anteromedial osteoarthritis (AMOA)), which benefits patients' quality of life [4, 5]. So far, numerous studies have reported that patients with mobile-bearing UKA have good clinical outcomes and long-term survival [6,7,8,9,10].

Appropriate patient selection is critical to the success of UKA surgery. Medial UKA should be performed in patients with AMOA, correctable varus deformity, intact ligaments, and less than 15 degrees of flexion and varus deformity [11]. However, the indications and contraindications of UKA were still evolving. The classic UKA contraindications proposed by Kozinn and Scott [12] were constantly being expanded, and many studies had shown that good clinical outcomes were observed in obese patients, younger patients, extremely physically active patients, and patients with the patellofemoral joint disease [13,14,15,16,17].

Coronal tibiofemoral subluxation (CTFS), defined as the distance between the tangent line to the outermost joint edge of the lateral condyle of the femur and the tangent line of the lateral tibial plateau, is considered a controversial and potential contraindication to UKA but is less discussed [18, 19]. A few studies had reported that excessive CTFS could decrease postoperative patient report outcome measurements (PROMs) and increase the risk of lateral compartment OA in UKA patients [20, 21]. At the same time, some scholars believed that postoperative CTFS was related to the impingement of the medial part of the lateral femoral condyle and the tibial spine and was a potential cause of unexplained knee pain after UKA [22,23,24]. Therefore, CTFS is an important indicator to pay more attention to UKA surgery.

CTFS can be corrected to some extent by UKA. However, not all CTFS can be satisfactorily corrected by UKA, and excessive postoperative CTFS may have adverse effects on patient prognosis [20, 24, 25]. Therefore, predicting postoperative CTFS through preoperative indicators is particularly important for surgeons to select suitable patients. A previous study suggested that preoperative CTFS under valgus stress may be a predictor of postoperative CTFS in patients with fixed-bearing UKA [20]. However, no study explored the preoperative factors which were associated with postoperative CTFS in patients with mobile-bearing UKA.

We therefore asked: (1) Can preoperative CTFS be significantly improved after mobile-bearing UKA? (2) Do preoperative variables (preoperative CTFS and preoperative CTFS under valgus stress) correlate with postoperative CTFS after mobile-bearing UKA?

Methods

This retrospective study was conducted using patients undergoing UKA from September 1 2019 to December 31 2021. The indications for UKA were as follows: (1) patients with anteromedial osteoarthritis (AMOA); (2) patients with intact knee ligaments (anterior cruciate ligament (ACL) and medial collateral ligament (MCL)); (3) patients with flexion contracture < 15° and preserved knee range of motion; (4) patients with varus deformity < 15° and this kind of deformity is correctable [11]. The inclusion criteria were as follows: (1) patients undergoing UKA for varus knee with OA; (2) patients with anterior–posterior (AP) standing knee radiographs, lateral radiographs, hip-to-ankle AP standing radiographs, and AP valgus stress force radiographs before knee surgery; (3) patients with AP standing knee radiographs, lateral radiographs and hip-to-ankle AP standing radiographs after knee surgery. The exclusion criteria were as follows: (1) patients with poor radiographs which cannot be used for research, poor valgus stress force radiographs (excessive internal or external rotation, and the radiograph beam not parallel to the tibial plateau) especially; (2) patients with lateral CTFS; (3) patients with secondary OA; (4) patients with fractures around the knee. A total number of 181 patients (224 knees, treated as 224 patients in this study) were enrolled in this study. The flowchart is shown in Fig. 1. The study was approved by the institutional review board (approval number 2020–50-k28).

Fig. 1
figure 1

Flowchart of the study

Full size image

Radiographic assessments

Before surgery, all patients had AP standing knee radiographs, lateral radiographs, hip-to-ankle AP standing radiographs, and AP valgus stress force radiographs. With the patients in the supine position and the knee flexed at 20 degrees, valgus stress force radiographs were obtained by the junior physician manually applying a firm valgus force to the affected knee joint without anesthesia. The junior physician was experienced and applied valgus stress as homogeneously as possible each time. Neutral rotation and parallelism of the tibial plateau to the radiograph beam were required. Three days after surgery, the same set of radiographs (except the valgus stress force radiograph) was repeated for all patients.

On the hip-to-ankle AP standing radiographs and valgus stress force radiographs, preoperative CTFS, preoperative CTFS under valgus stress, and postoperative CTFS were measured using the hospital's imaging system (picture archiving and communication system, PACS) (Fig. 2). CTFS was defined as the distance between the tangent line to the outermost joint edge of the lateral condyle of the femur and the tangent line of the lateral tibial plateau [18, 19]. Positive values represent medial coronal tibiofemoral subluxation, negative values represent lateral coronal tibiofemoral subluxation, and zero means no coronal tibiofemoral subluxation [24]. As lateral CTFS (CTFS < 0) was rare in UKA patients and such patients were not representative, patients with lateral subluxation on preoperative AP standing knee radiographs were excluded from the study. The subjects of this study were patients with medial CTFS (CTFS > 0) or no CTFS (CTFS = 0).

Fig. 2
figure 2

Measurement of coronal tibiofemoral subluxation (CTFS) (a–g). a The preoperative hip-to-ankle anterior–posterior (AP) standing radiograph; b partial magnification of the preoperative hip-to-ankle AP standing radiograph, and the preoperative CTFS (recorded as pre-CTFS in the figure) was measured; c the preoperative AP valgus stress force radiograph; d partial magnification of the preoperative AP valgus stress force radiograph, and the preoperative CTFS under valgus stress (recorded as pre-CTFS under valgus stress in the figure) was measured; e the postoperative hip-to-ankle AP standing radiograph; f partial magnification of the postoperative hip-to-ankle AP standing radiograph, and the postoperative CTFS (recorded as post-CTFS in the figure) was measured; g charts showing preoperative, preoperative (under valgus stress), and postoperative CTFS. The means ± standard deviations (SD) of preoperative CTFS, preoperative CTFS under valgus stress, and postoperative CTFS were 4.96 ± 1.82 mm, 3.06 ± 1.37 mm, and 3.19 ± 1.27 mm. The difference between preoperative CTFS and preoperative CTFS under valgus stress was statistically significant (p < 0.001), as was the difference between preoperative CTFS and postoperative CTFS (p < 0.001). However, no significant difference was found between preoperative CTFS under valgus stress and postoperative CTFS

Full size image

Many previous studies used the CTFS value of 5 mm as the cutoff value [20, 26,27,28]. It has been shown that when the femoral prosthesis was placed at a horizontal distance of 5 mm from the center of the tibial prosthesis, there would be negative biomechanical effects [29]. Therefore, 5 mm was used as the cutoff value for this study. Patients were divided into two groups based on postoperative CTFS, group A (postoperative CTFS ≤ 5 mm) and group B (postoperative CTFS > 5 mm). In addition, patients could be divided into three groups based on preoperative and postoperative CTFS, group C (both preoperative and postoperative CTFS ≤ 5 mm), group D (preoperative CTFS > 5 mm and postoperative CTFS ≤ 5 mm), and group E (both preoperative and postoperative CTFS > 5 mm).

On the hip-to-ankle AP standing radiographs, preoperative and postoperative hip–knee–ankle angle (HKA), preoperative mechanical proximal tibial angle (mPTA), and preoperative mechanical lateral distal femoral angle (mLDFA) were measured using the hospital's imaging system (PACS) (Fig. 3). HKA was the angle between the tibial mechanical axis and femoral mechanical axis, mPTA was the angle between the tangent of the medial and lateral tibial plateau and the mechanical axis of the tibia, and mLDFA was the angle between the tangent of the medial and lateral femoral condyle and the mechanical axis of the femur [30,31,32].

Fig. 3
figure 3

Measurement of different alignment parameters on hip-to-ankle anterior–posterior (AP) standing radiographs (a–c). a The preoperative hip-to-ankle AP standing radiograph, and the preoperative hip–knee–ankle angle (HKA) (recorded as pre-HKA in the figure), mechanical lateral distal femoral angle (mLDFA), and mechanical proximal tibial angle (mPTA) were measured; b the postoperative hip-to-ankle AP standing radiograph, and the postoperative HKA (recorded as post-HKA in the figure) was measured; c charts showing preoperative and postoperative HKA. The mean ± standard deviation (SD) of preoperative HKA was 172.12 ± 3.85°, while the mean ± SD of postoperative HKA was 176.80 ± 2.30°, a difference between them was statistically significant (p < 0.001)

Full size image

Two orthopedic surgeons made all measurements using the hospital's imaging system (PACS). Intra-class correlation coefficient (ICC) was used for continuous variables (preoperative HKA, preoperative mPTA, preoperative mLDFA, preoperative CTFS, preoperative CTFS under valgus stress, postoperative HKA, and postoperative CTFS) to test the interobserver reliability.

UKA procedures

All UKA operations were performed by the same surgeon using the mobile-bearing Oxford medial UKA (Zimmer-Biomet, Warsaw, Indiana) concerning the standard Oxford UKA protocol [33]. The surgeon exposed the knee joint through a small incision (an anteromedial incision along the medial border of the patella from the superior pole of the patella to the medial side of tibial tuberosity, approximately 8–10 cm) during the operation, preserving the quadriceps femoris without patella eversion. All medial osteophytes were completely removed using an osteotome. A femoral spherical mill was used to mill the distal femur condyle to balance the flexion and extension gap. After the balance of the extension and flexion gap, the tibia keel was prepared, and the prosthesis was fixed with bone cement. The ligaments were balanced in line with the polyethylene insert’s thickness. No medial release was performed during the operation.

Statistical analysis

The continuous variables (age, body mass index (BMI), preoperative HKA, postoperative HKA, preoperative mLDFA, preoperative mPTA, preoperative CTFS, preoperative CTFS under valgus stress, and postoperative CTFS) were presented as means and standard deviations (SD), while the categorical variables (sex, side, size of femoral component, size of tibial component, and size of polyethylene bearing) were given as frequencies and percentages (%).

The Student’s t-test was used to compare the differences between preoperative HKA and postoperative HKA, preoperative CTFS and preoperative CTFS under valgus stress, preoperative CTFS and postoperative CTFS, and preoperative CTFS under valgus stress and postoperative CTFS. The Student’s t-test or one-way ANOVA together with Tukey’s post hoc test (for continuous variables) and the chi-square test or Fisher’s exact test (for categorical variables) were used to compare the clinical variables between group A (postoperative CTFS ≤ 5 mm) and group B (postoperative CTFS > 5 mm) or among group C (both preoperative and postoperative CTFS ≤ 5 mm), group D (preoperative CTFS > 5 mm and postoperative CTFS ≤ 5 mm), and group E (both preoperative and postoperative CTFS > 5 mm). Pearson correlation analysis was performed among these variables, including preoperative CTFS, preoperative CTFS under valgus stress, postoperative CTFS, preoperative HKA, and postoperative HKA. Simple linear regression was carried out between the clinical variables (independent variables) and the postoperative CTFS (dependent variable). The variables (p < 0.10 in simple linear regression) were further analyzed in multiple linear regression. To assess the predictive variables of postoperative CTFS (> 5 mm), univariate and multiple logistic regression analyses (variables with p < 0.10 in univariate logistic regression) were carried out.

All analyses were performed using SPSS24.0 (IBM, New York, USA), and a p < 0.05 (two-sided) was considered statistically significant.

Results

Subject characteristics

A total of 181 patients (224 knees, treated as 224 patients in this study) were enrolled in the study. The mean ± SD (range) of age and BMI were 67.58 ± 7.51 years and 27.59 ± 3.44 kg/m2, respectively. Of all patients, 43 patients were male, and 181 patients were female; 112 patients had surgery on the left knee, while the others (112 patients) were on the right knee (Table 1). The mean ± SD of preoperative and postoperative radiographic measurements and prosthesis parameters are presented in Table 1. Excellent inter-observer reliabilities were shown in preoperative and postoperative radiographic measurements (Table 2).

Table 1 Basic characteristics
Full size table
Table 2 Interobserver reliability
Full size table

The mean ± SD of preoperative CTFS, preoperative CTFS under valgus stress, and postoperative CTFS were 4.96 ± 1.82 mm, 3.06 ± 1.37 mm, and 3.19 ± 1.27 mm. The difference between preoperative CTFS and preoperative CTFS under valgus stress was statistically significant (p < 0.001), as was the difference between preoperative CTFS and postoperative CTFS (p < 0.001). However, there was no statistically significant difference between preoperative CTFS under valgus stress and postoperative CTFS (Fig. 2). The mean ± SD of preoperative HKA was 172.12 ± 3.85°, while the mean ± SD of postoperative HKA was 176.80 ± 2.30°, the difference between them was statistically significant (p < 0.001) (Fig. 3).

Comparison analysis

Of all patients, 202 patients were classified as Group A (postoperative CTFS ≤ 5 mm), while 22 patients were classified as Group B (postoperative CTFS > 5 mm). The preoperative CTFS in Group B was significantly higher than that in Group A (6.35 ± 1.34 mm vs 4.81 ± 1.82 mm, p < 0.001), so was the variable-preoperative CTFS under valgus stress (5.41 ± 1.00 mm vs 2.80 ± 1.14 mm, p < 0.001). No significant difference was found in other variables between Group A and Group B (Table 3).

Table 3 Comparisons between Group A (postoperative coronal tibiofemoral subluxation (CTFS) ≤ 5 mm) and Group B (postoperative CTFS > 5 mm)
Full size table

Of all patients, 122 patients were classified as Group C (both preoperative and postoperative CTFS ≤ 5 mm), 80 patients were classified as Group D (preoperative CTFS > 5 mm and postoperative CTFS ≤ 5 mm), and 22 patients were classified as Group E (both preoperative and postoperative CTFS > 5 mm). The preoperative CTFS (p < 0.001) and preoperative CTFS under valgus stress (p < 0.001) differed significantly across the three groups, while no significant difference was found in other variables (Table 4).

Table 4 Comparisons among Group C (both preoperative and postoperative coronal tibiofemoral subluxation (CTFS) ≤ 5 mm), Group D (preoperative CTFS > 5 mm and postoperative CTFS ≤ 5 mm), and Group E (both preoperative and postoperative CTFS > 5 mm)
Full size table

Correlation analysis

In Pearson correlation analysis, preoperative HKA was positively related to postoperative HKA (r = 0.690, p < 0.001). There was also a correlation between preoperative CTFS and postoperative CTFS (r = 0.493, p < 0.001), and the correlation between preoperative CTFS under valgus stress and postoperative CTFS was stronger (r = 0.842, p < 0.001) (Table 5). No significant correlation was found between preoperative HKA and preoperative CTFS (r = 0.048) or preoperative CTFS under valgus stress (r = 0.033), nor between postoperative HKA and postoperative CTFS (r = – 0.033).

Table 5 Correlation analysis
Full size table

Simple and multiple linear regression analysis

In simple linear regression analysis, only preoperative CTFS (β = 0.344, 95%CI = 0.264–0.424, and p < 0.001) and preoperative CTFS under valgus stress (β = 0.783, 95%CI = 0.717–0.850, and p < 0.001) were significantly correlated with postoperative CTFS. In addition, the p values were 0.063 (p < 0.1) and 0.092 (p < 0.1) for preoperative mPTA and size of tibial component. For other variables, no significant correlations were found with postoperative CTFS (Table 6).

Table 6 Simple and multiple linear regression analysis of postoperative coronal tibiofemoral subluxation (CTFS)
Full size table

These variables (preoperative mPTA, preoperative CTFS, preoperative CTFS under valgus stress, and size of tibial component) with p < 0.1 in the simple linear regression analysis were further analyzed using multiple linear regression analysis, and only preoperative CTFS under valgus stress (β = 0.798, 95%CI = 0.714–0.883, p < 0.001) was significantly correlated with postoperative CTFS (Table 6).

Univariate and multivariate logistic regression analysis

In univariate analysis, significant correlations were found between preoperative CTFS (odds ratio (OR) = 1.642, 95%CI = 1.256–2.148, p < 0.001), preoperative CTFS under valgus stress (OR = 11.589, 95%CI = 4.773–28.138, p < 0.001) and postoperative CTFS (> 5 mm). In addition, p < 0.1 was shown in age (p = 0.055) and size of tibial component (p = 0.052). However, no significant correlations were found between other variables and postoperative CTFS (> 5 mm) (Table 7).

Table 7 Univariate and multivariate logistic regression analysis of predictive factors for postoperative coronal tibiofemoral subluxation (CTFS) residual (> 5 mm)
Full size table

Variables (p < 0.1 in univariate analysis), including age, preoperative CTFS, preoperative CTFS under valgus stress, and size of the tibial component, were further analyzed using multiple logistic regression analysis. The variable-preoperative CTFS under valgus stress (OR = 12.412, 95%CI = 4.757–32.384, and p < 0.001) was expressed as the risk factor of postoperative CTFS (> 5 mm) in the multiple logistic regression analysis (Table 7).

Discussion

The main findings of this study were that preoperative CTFS could be improved significantly after mobile-bearing UKA, postoperative CTFS was correlated with preoperative CTFS under valgus stress, and a higher preoperative CTFS under valgus stress would increase the risk of higher postoperative CTFS (> 5 mm). It was worth mentioning that to our knowledge, this was the first study to quantitatively assess the effect of preoperative CTFS under valgus stress on postoperative CTFS in the mobile-bearing UKA (Oxford UKA).

CTFS has a variety of measurement methods [19, 34,35,36,37]. In this study, we measured CTFS as the distance between the tangent line to the outermost joint edge of the lateral condyle of the femur and the tangent line of the lateral tibial plateau concerning the method by Springer et al. [18]. The method had ICC values of 0.919 (preoperative CTFS), 0.908 (preoperative CTFS under valgus stress), and 0.847 (postoperative CTFS) in this study, and had an ICC value of 1.0 in the study of Springer et al. [18], indicating the high reproducibility of this measurement method. The value of preoperative CTFS in this study was 4.96 ± 1.82 mm (all patients had an intact ACL), similar to the results by Springer et al. [18], in which the preoperative CTFS value was 4.7 ± 1.4 mm (ACL functionally intact group, n = 68). Moreover, by using the measurement method, this study found that the difference between preoperative CTFS (4.96 ± 1.82 mm) and postoperative CTFS (3.19 ± 1.27 mm) was statistically significant (p < 0.001) (Table 1 and Fig. 2). Such a finding suggested that preoperative CTFS could be improved significantly with mobile-bearing UKA, which was similar to the results by Nam et al. [24] and Xi et al. [28].

CTFS is considered a controversial and potential contraindication to UKA, but it is less discussed [18, 19]. Some scholars believed that excessive postoperative CTFS may result in a mismatch in relative positioning between femoral and tibial components, which could lead to intercondylar notch impingement, component edge loading, polyethylene wear, lateral compartment OA, and poor PROMs after surgery [20,21,22,23,24,25, 38]. Kang et al. [29] reported that the femoral prosthesis placed at a horizontal distance of 5 mm from the center of the tibial prosthesis would increase the contact stress on the lateral compartment of the knee in UKA surgery. In addition, Kamenaga et al. found that UKA patients with postoperative CTFS > 5 mm (Group U: 6.5 ± 1.5 mm) had poorer PROMs 2 years after surgery than patients with postoperative CTFS < 5 mm (Group A: 3.0 ± 1.2 mm, and Group C: 3.2 ± 1.2 mm) [20] and that UKA patients with lateral osteoarthritis progression (mean follow-up: 80.8 ± 10.5 months) had higher postoperative CTFS than patients with no progression (6.6 ± 2.3 mm > 5 mm vs 4.6 ± 2.9 mm < 5 mm) [21]. Therefore, 5 mm was used as the cutoff value of postoperative CTFS, and patients were divided into two groups, group A (postoperative CTFS ≤ 5 mm) (n = 202) and group B (postoperative CTFS > 5 mm) (n = 22) in the study. The preoperative CTFS under valgus stress in Group B was significantly higher than that in Group A (5.41 ± 1.00 mm vs 2.80 ± 1.14 mm, p < 0.001) (Table 3), which was similar to the results by Kamenaga et al. [20] that the mean ± SD of postoperative CTFS was 6.5 ± 1.5 mm in Group U (patients with preoperative CTFS > 5 mm and preoperative CTFS under valgus stress > 5 mm), higher than that in Group A (patients with preoperative CTFS < 5 mm and preoperative CTFS under valgus stress < 5 mm, postoperative CTFS: 3.0 ± 1.2 mm) and Group C (patients with preoperative CTFS > 5 mm and preoperative CTFS under valgus stress < 5 mm, postoperative CTFS: 3.2 ± 1.2 mm). This might explain why we still have cases with postoperative CTFS > 5 mm after surgery and indicate that patients with high preoperative CTFS under valgus stress should be performed UKA with caution.

Preoperative valgus stress radiography has been widely used in predicting coronal alignment after UKA [39,40,41]. However, few studies have explored the relationship between preoperative valgus stress radiographs and postoperative CTFS [20, 21]. Some scholars believed that patients were suitable for UKA as long as CTFS was correctable on preoperative stress force radiographs [18, 42] and a previous study by Kamenaga et al. [20] suggested that preoperative CTFS under valgus stress might be a predictor of postoperative CTFS in patients with fixed-bearing UKA. To our knowledge, no study has explored the relationship between the preoperative CTFS under valgus stress and postoperative CTFS in patients with mobile-bearing UKA. This research first reported that preoperative CTFS under valgus stress was significantly correlated with postoperative CTFS (Table 5 and Table 6) and was shown as an independent risk factor for postoperative CTFS (> 5 mm) (Table 7) in patients undergoing mobile-bearing UKA. Such results provided a reference for clinicians in selecting suitable UKA patients, which could avoid excessive postoperative CTFS that affected the prognosis of patients undergoing mobile-bearing UKA.

In this study, there was no significant correlation between preoperative HKA and preoperative CTFS (r = 0.048) or preoperative CTFS under valgus stress (r = 0.033), nor between postoperative HKA and postoperative CTFS (r = – 0.033). Such results suggested that CTFS was a variable independent of overall mechanical alignment, which was similar to previous findings by Springer et al. [18] and Nam et al. [24]. This phenomenon might be related to the quality of the bone in the affected knee compartment. For example, a patient with medial knee OA is more likely to develop bone deformity with varus angulation than CTFS if the medial compartment has soft bone. Correspondingly, if the bone in the medial compartment is sclerotic, the patient is more likely to have CTFS with varus angulation [24, 43].

Limitations

This study had its limitations. Firstly, although the same protocol was used for each patient’s radiograph, some imaging measurements, including CTFS, were still affected by lower extremity rotation and osteophyte. Therefore, computer tomography (CT) evaluation may be a better choice for CTFS measurement. However, this study aimed to measure CTFS on readily available radiographs and the interobserver agreement for this measure was excellent, so CT was not used in the study. Secondly, the study was a retrospective analysis involving only one prosthesis (mobile-bearing medial UKA). Therefore, prospective studies using multiple types of prostheses (including fixed-bearing medial UKA) should be considered in the future. Thirdly, clinical follow-up of patients has not been performed in this study, and follow-up in the future is needed to confirm the clinical significance of CTSF (postoperative CTSF, > 5 mm vs ≤ 5 mm). However, some studies reported that excessive CTFS after UKA would lead to a decrease in PROMs and increase the risk of lateral compartment OA in patients following medial UKA [20, 21]. Fourthly, the magnitude of valgus stress of the valgus stress force radiographs in this study was not quantified using a machine such as a Telos device. However, the surgeon who performed the valgus stress force radiographs had a lot of experience and applied valgus stress as homogeneously as possible, and patients with poor quality of the valgus stress force radiographs (excessive internal or external rotation, and the radiograph beam not parallel to the tibial plateau) were excluded in the study. Finally, the cutoff value of the preoperative CTFS under valgus stress had not been calculated in this study. However, our results in the study have suggested that postoperative CTFS was correlated with preoperative CTFS under valgus stress and a higher preoperative CTFS under valgus stress would increase the risk of higher postoperative CTFS (> 5 mm), which had served the purpose of this study and was clinically meaningful.

Conclusion

Preoperative CTFS can be improved significantly after mobile-bearing UKA. In addition, postoperative CTFS is correlated with preoperative CTFS under valgus stress and a higher preoperative CTFS under valgus stress will increase the risk of higher postoperative CTFS (> 5 mm).