Introduction

Postoperative range of motion (ROM) is an important outcome parameter following knee arthroplasty [22, 28, 42]. This is especially so for Asian populations, that tend to require deep knee flexion in their daily activities, such as kneeling during prayers [22, 29].

Posterior tibial slope (PTS), posterior condylar offset (PCO) and femoral sagittal angle (FSA) are surgically modifiable determinants of postoperative ROM after unicompartmental knee arthroplasty (UKA) [8, 9, 11, 32]. A steeper PTS can facilitate greater flexion [16] but could result in tighter extension gaps [40]. Greater flexion in FSA can also facilitate greater knee flexion [8], but increases contact stress in the unreplaced compartment [32]. Component positioning in the sagittal plane is vital to prevent early failure in UKA [2, 32].

However, the optimal sagittal positioning remains controversial, partly because previous studies have evaluated PTS, PCO and FSA in isolation [8, 16, 32, 40]. This study aimed to propose the optimal alignment for sagittal plane parameters in UKA by accounting for PTS, PCO and FSA alignments concurrently. We hypothesize that there is an optimal range for sagittal plane parameters that can maximize postoperative ROM.

Methods and materials

Patients

This study was approved by a Centralized Institutional Review Board (CIRB 2020/2157) and performed in accordance to the amended ethical standards laid out in the World Medical Association Declaration of Helsinki [47]. Prospectively collected data of patients who underwent primary fixed-bearing UKA at a single institution from 2004 to 2007 was reviewed. A total of 392 UKAs were performed during this period. After excluding those whose radiographs went missing when patient records went digital and PROMs of less than 2-year follow-up, 265 UKAs with complete follow up from pre operation to 10-year were available for analysis (68%). The 265 UKAs belonged to 213 patients. 52 patients had bilateral UKAs, while 161 patients had unilateral UKAs. The mean age of patients was 61.0 ± 7.6 years old, mean body mass index was 27.2 ± 4.4 kg/m2, and 75% were females (Table 1). On latest review at mean 15-years, 19 knees have been revised. Reasons for revision included: aseptic loosening of tibial component (n = 1), progression of OA (n = 13), polyethylene failure (n = 3), subsidence of tibial tray (n = 2). Two types of implants were used in this study, the Preservation (DePuy Synthes, Leeds, UK) (n = 59) and the Miller-Galante (Zimmer Biomet, Warsaw, IN) (n = 206) UKAs. Both prostheses had a similar round-on-flat geometry. Failure rates by implants were at 8% for the Preservation (n = 5/59) and 7% for the Miller-Galante (n = 14/206). Only a minority of revisions occurred before the 10-year mark (n = 4), giving an overall 10-year survivorship of 98%. The final survivorship of mean 15-year (range 13–17 years) was 93%.

Table 1 Summary of patient demographics (n = 265 knees)
Full size table

Surgical technique

All UKAs were consecutively done by the two senior authors who are fellowship-trained arthroplasty surgeons. The indication for surgery was based on the Kozinn and Scott criteria [20] and all surgeries were indicated for unicompartmental noninflammatory primary OA of the knee with intact cruciate ligaments. The surgical technique was performed in accordance with the respective manufacturer’s surgical instrumentation guide using conventional instruments. An abbreviated medial approach without patellar eversion was performed.

Clinical evaluation

The patients included in this study had few medical comorbidities at index operation, which was translated into the Charlson Comorbidity Index (CCI) scores [34] (Table 1). A variety of patient-reported outcome measures (PROMs) were used to assess all patients routinely in the clinic at 6-month, 2-year and 10-year follow-up. Knee function was assessed using the: adapted Oxford Knee Score (OKS) scoring system [30]; Knee Society knee (KSKS) and function (KSFS) scores [15], Short Form-36 (SF-36) [17, 23] health survey which was transformed into physical (PCS) and mental (MCS) component scores [43] (Table 2). Fulfilment of patients’ expectations and satisfaction after surgery was done with a 6-point scale using similar questions to Bourne et al. [6]. ROM was assessed using a two-arm goniometer with the patient in a supine position. All clinical assessment was performed by physiotherapists who were not involved in this study. The electronic records of all patients were reviewed in July 2020 to determine if and when they had undergone revision, giving a mean duration of follow-up of 15 years.

Table 2 Summary of functional outcomes
Full size table

Radiological evaluation

Preoperative and postoperative lateral radiograph of all 265 knee joints were used for radiological evaluation. Radiographs measurements were performed using the Picture Archiving and Communication Systems (PACS: Carestream Health, Rochester, New York, USA). Measurements were recorded to a precision of 0.1 mm and 1°. Measurements were performed using a similar method to previous authors [12, 18, 40], where PTS is the acute angle between the tibial plateau in the sagittal view, and the line perpendicular to the tibial mechanical axis (Figs. 1and 2); PCO is the perpendicular distance of the posterior-most aspect of the femoral condyle from a projection of the posterior femoral line (Figs. 1and 2); FSA is the acute angle between the cut surface of the posterior femoral condyle and the anatomical axis of the femur (Fig. 3a, b).

Fig. 1
figure 1

Preoperative lateral radiograph showing measurement of posterior condylar offset (PCO) and posterior tibial slope (PTS)

Full size image
Fig. 2
figure 2

Postoperative lateral radiograph showing measurements of posterior condylar offset (PCO) and posterior tibial slope (PTS)

Full size image
Fig. 3
figure 3

Postoperative lateral radiograph showing the measurement of the femoral sagittal angle (FSA) on a Preservation and b Miller-Galante implants

Full size image

Radiographs were evaluated consecutively by two blinded independent assessors who were not involved in the surgical procedure. Evaluation was performed by both assessors independently, in December 2019. Subsequently, the evaluation was repeated by both assessors after 6 months. The test–retest reliability was assessed using intraclass correlation coefficient (ICC). Using a two-way random-effects model, the ICC value was 0.98 (95% CI 0.97–0.99, p < 0.05). An ICC > 0.9 represents excellent reliability, which suggests that measurement of sagittal plane parameters meticulously performed using this method, can achieve excellent reliability and reproducibility.

Statistical analysis

All statistical analysis was done using RStudio (RStudio: Integrated Development for R. Rstudio, Inc., Boston, MA). A p value of < 0.05 was used as the threshold for statistical significance. Sample size calculation was performed with power at 0.8 and a significance level at 0.05. To detect 1º of change, with a standard deviation of 3º, the minimum sample size was 142. Descriptive statistics included count, mean, standard deviation and proportions. Kendall’s ranked correlation test was used to evaluate the correlation between sagittal plane parameters with various clinical outcomes. Multivariable regression analysis was performed to elucidate the significant predictors of postoperative ROM. Multivariable regression models also explored potential interaction effects between PTS, PCO and FSA. Regression models adjusted for potential confounders of outcomes, such as age [21], body mass index (BMI) [5, 13, 25, 35, 48], CCI, operated side, different implant designs, coronal and limb alignments, and preoperative ROM [16]. Subsequently, adjusted outputs from the multivariable regression models were channelled into constructing effect plots and interaction plots, to isolate the individual effect of variables on postoperative ROM and propose recommendations accordingly. Loess lines, which are locally weighted smooth polynomial regression lines set to a default span of 0.5, were utilised to better capture non-linear trends in the data and reduce any skewing effect from outliers.

Results

Correlations

PTS, PCO and FSA are significantly correlated with each other (Table 3). A steeper PTS and larger PCO was correlated with greater flexion. Greater flexion of the FSA was correlated with greater FFD. A larger PCO and greater flexion of the FSA was correlated with better PROMs (Table 3).

Table 3 Kendall’s ranked correlation coefficients showing relationship of posterior tibial slope (PTS), femoral posterior condylar offset (PCO), and femoral sagittal angle (FSA) against different clinical parameters
Full size table

Predictors

There were significant two-way interaction effects between PTS and PCO (Table 4). Younger age, lower BMI and greater preoperative flexion are significant predictors of better postoperative flexion (Tables 4, 5and 6). The only significant predictor of FFD was preoperative FFD (Tables 4, 5and 6).

Table 4 Multivariable linear regression on knee range of motion at 6-month
Full size table
Table 5 Multivariable linear regression on knee range of motion at 2-year
Full size table
Table 6 Multivariable linear regression on knee range of motion at 10-year
Full size table

Sagittal plane alignment

A PTS between 2° to 8° is optimal for better postoperative flexion (Fig. 4). Restoration of PCO within 1.5 mm from preoperative values is optimal for better postoperative flexion (Fig. 5). At shallow PTS of 2°, it is preferable to reduce PCO by 1.0 mm, whereas at steeper PTS of 8°, it is preferable to restore PCO at 0 mm (Fig. 6).

Fig. 4
figure 4

Effect plot showing the adjusted influence of posterior tibial slope (PTS) on knee flexion at 6-month

Full size image
Fig. 5
figure 5

Effect plot showing the adjusted influence of change in posterior condylar offset (∆PCO) on knee flexion at 6-month

Full size image
Fig. 6
figure 6

Interaction plot showing the effect of posterior tibial slope (PTS) on knee flexion at 6-month, for different values of change in posterior condylar offset (∆PCO)

Full size image

Discussion

The first important finding of this study was the interaction effect between PTS and PCO. This interaction means that there are differential effects of PTS depending on different sizes of PCO, and vice versa. This finding was consistent with a previous study by Han et al. which found an interaction effect between PTS and PCO in total knee arthroplasty (TKA) [12]. Even though knee kinematics between UKA and TKA differ significantly, this finding seems to be similar for both. This is crucial because surgeons need to balance the component gaps yet be mindful that the resultant size of the posterior condyle on different tibial plateau inclination may influence the femoral roll-back mechanism [46]. A mismatch between the PTS and PCO may lead to excessive translation during weight-bearing activities [14, 46], and exacerbate stress on the native knee structures [37, 38].

The second important finding of this study was that PTS between 2° to 8° is optimal in maximising knee flexion. This finding was consistent with previous studies that have made similar recommendations [1, 37, 39]. In addition, the study by Small et al. demonstrated that contact strain in the posterior compartment increased by 80% (p = 0.00) when PTS was increased from 5° to 10°, suggesting that 10° PTS is not advisable [38]. The study by Weber et al. in a cohort of fixed-bearing UKA, also found that there was a significant reduction in wear rate from 10.4 mg/million cycles to 3.22 mg/million cycles when PTS was increased from 0° to 8° [45]. These findings are consistent with the recommendation regarding optimal PTS angles within 2°–8°.

The third important finding of this study was that restoration of PCO to within 1.5 mm of its preoperative value was ideal in maximising knee flexion. Restoration of PCO has been described to be an important predictor of better functional outcomes after TKA [27], but this has not yet been described in UKA. Restoration of PCO is likely to be crucial due to two reasons. Firstly, UKA may be primarily thought of as a resurfacing procedure [3, 44], restoration of PCO is more likely to maintain the complex kinematics of the native knee. Secondly, given the interactive effects between PTS and PCO detailed above, preserving the patient’s native PCO will likely avoid unnecessary counter-productive effects when coupled with PTS. Surgeons just have to be mindful not to cut PCO > 1.5 mm when balancing the component gaps.

In terms of survivorship, PTS, PCO and FSA were not associated with failures in this study. This finding was consistent with previous studies [7, 10], although some authors have reported PTS to be a crucial factor in UKA survivorship [26, 33]. The mechanism for this is because extremes of PTS has been shown to be a risk factor for bearing dislocation [36], although this is still contested [19, 24]. However, this study is unable to corroborate if a steeper PTS might predispose patients to a higher risk of bearing dislocation as it did not include any mobile-bearing UKA. Park et al. also described that excessive flexion of FSA may contribute to the progression of lateral compartment OA due to higher contact stresses [32]. However, none of the patients with revision for the progression of OA in this study had extreme FSA positioning, which suggests that it may not be a clinically significant mechanism for failure.

Given that there has been significant improvement in implant design since 2007, the effects of these parameters may be different with newer generations of UKA implants that have a different geometry and hence biomechanics. A limitation of this study was the use of short radiographs for radiological evaluation, which has slightly lesser precision compared to long film radiographs. Although short radiographs have been described to be an acceptable alternative to long film radiographs in assessing component positioning [41]. Another limitation was the use of 2-dimensional films, therefore tilt, rotation or non-strict lateral views could influence radiological measurements. It would be useful for future studies with access to 3-dimensional imaging to corroborate the findings from this study. Despite this limitation, this study has various strengths. Firstly, it has a long duration of follow-up. Secondly, this study utilized a myriad of different yardsticks and survivorship in assessing patients. Thirdly, this study used a string of rigorous statistical approaches that adjusted for various covariates and interactions, to accurately capture the effect of each factor free from the influence of other confounders.

Clinically, UKA surgeons should aim to cut PTS between 2° to 8° and restore PCO to within 1.5 mm of its preoperative value. Since PTS is often augmented depending on the condition of the cruciate ligaments, surgeons should prioritize increasing the posterior condylar cut when PTS is shallow, and decrease the posterior condylar cut when PTS is steep. Moreover, these targets would be useful to guide planning in robotic surgery given the extremely precise cuts involved [4, 31].

Conclusion

UKA surgeons and future studies should be mindful of the relationship between PTS, PCO and FSA, and avoid considering them in isolation. When deciding on the method of balancing component gaps in UKA, surgeons should rely on the PTS. Decrease the posterior condylar cut when PTS is steep, and increase the posterior condylar cut when PTS is shallow. The acceptable range for PTS is between 2° to 8° and PCO should be restored to 1.5 mm of native values.