Retropatellar resurfacing in primary total knee arthroplasty does not improve clinical outcomes but increases revision rates

Abstract

Purpose

Controversy still remains regarding the necessity of primary retropatellar resurfacing (RPR) in total knee arthroplasty (TKA). While some studies suggest retropatellar resurfacing reduces anterior knee pain (AKP) and revision rates, others report no significant differences compared to non-resurfaced TKA. This study aims to evaluate postoperative complications, pain levels, and functional outcomes in patients undergoing TKA with and without RPR.

Methods

A retrospective, single-center observational study was conducted including 121 patients who underwent primary TKA. Fifty-five patients (45.5%) had RPR, while 66 patients (54.5%) retained their native patella. Clinical outcomes were assessed using the knee society score (KSS), while radiological parameters such as femoral and patellar offset were analyzed. Statistical comparisons were performed using independent t-tests and chi-square tests, with significance set at p < 0.05.

Results

Radiological analysis revealed a significant reduction in femoral offset postoperatively (p < 0.001), with greater changes observed in the non-RPR group (p = 0.009). Patellar offset was also significantly reduced in non-RPR patients compared to RPR patients (p < 0.001). The revision rate was higher in the RPR group (10 cases) compared to the non-RPR group (1 case). Functional assessment using the KSS showed no significant differences in knee pain (p = 0.418) or knee scores (p = 0.461) between groups. However, patients in the RPR group were more likely to require walking aids postoperatively (p = 0.012).

Conclusions

RPR in primary TKA did not result in superior pain relief or functional outcomes compared to non-resurfacing but was associated with a higher revision rate and increased postoperative use of walking aids. These findings suggest that the decision to resurface should be individualized based on clinical and anatomical considerations.

Level of evidence

Level III (Retrospective Cohort).

Introduction

Total knee arthroplasty (TKA) is a well-established surgical procedure for patients with end-stage knee osteoarthritis, commonly resulting in substantial pain relief and improved joint function [1]. Despite ongoing advancement in implant design and surgical techniques, the decision to routinely resurface the patella remains controversial [2]. Although modern prosthetic designs have contributed to a reduction in the incidence of anterior knee pain (AKP), between 2% and 12% of patients continue to report persistent symptoms [3–7]. This postoperative AKP often presents without a clearly identifiable cause, thereby reducing patients` satisfaction and quality of life and complicating effective management strategies [8, 9]. A number of contributing factors have been reported being responsible for the development of AKP, including patellar maltracking due to component malalignment [10], lower limb malalignment, soft tissue imbalance, arthrofibrosis and quadriceps weakness [11]. Given the multifactorial etiology of AKP, the accurate assessment of patellofemoral alignment and biomechanics prior to surgery is critical in identifying patients at elevated risk and in guiding strategies to enhance postoperative outcomes.

Among these considerations, the role of retropatellar resurfacing (RPR) remains controversial. The decision to resurface the patella is multifactorial, influenced by intraoperative findings, implant design, surgeon`s preference, and patient-reported outcomes [12]. Altered patellofemoral mechanics following TKA without patellar resurfacing may increase retropatellar pressure, potentially contributing to persistent AKP [13]. Hence, resurfacing of the patella has been discussed as a potential method to prevent AKP [14, 15]. While several meta-analyses report lower rates of AKP and reoperations in resurfaced cohorts, others fail to show significant differences, underscoring the ongoing debate [12, 16–24].

The discrepancy in clinical outcomes may be attributable to heterogeneity in study designs, patient populations, implant geometries, and surgical techniques [16]. Despite numerous randomised controlled trials, consensus guidelines on patellar resurfacing remain lacking. Registry data reflect this uncertainty: the American Joint Replacement Registry (AJRR) indicates a gradual decline in patellar resurfacing rates, from 95.9% in 2012 to 87% in 2023 [25]. Conversely, the Australian registry reports a marked increase in resurfacing, reaching 78.1% by 2022 [26], while in Sweden, resurfacing is virtually abandoned [27].

This global variation reflects the absence of definitive evidence guiding best practice. Furthermore, many studies suffer from limitations such as short follow-up durations, inconsistent reporting of complications, and lack of standardised radiological assessments [12]. Importantly, resurfacing decisions often rely on surgeon experience rather than objective criteria [28]. In this context, improving our understanding of patient- and procedure-specific factors contributing to AKP is vital.

Building upon previous findings suggesting that preoperative patellofemoral alignment influences postoperative outcomes, the present study aims to evaluate postoperative complications, pain levels, and functional results in TKA with and without patellar resurfacing. By identifying clinical and radiographic predictors of AKP, this study seeks to inform a more evidence-based approach to patellar management in primary TKA. It was hypothesized that the outcome after primary TKA differs significantly between patients undergoing patella resurfacing and those remaining with the native patella. Specifically, it was expected that RPR would not be associated with a lower incidence of anterior knee pain, but rather with an increased risk of implant-related complications.

Materials and methods

A total of 121 patients with primary TKA were enrolled in this study. The mean age of patients at the time of surgery was 68.4 ± 9.7 years (45–88 years), with 54 (44.6%) males and 67 (55.4%) females. RPR was performed in 55 patients (45.5%), while 66 patients (54.5%) had undergone TKA without.

This was a single center, retrospective observational study at a knee arthroplasty center. The records consisted of patients who attended the clinic due to AKP after TKA performed between 2012 and 2022. In our cohort, femoral component constraint comprised cruciate-retaining (CR) designs in 80 cases (66.1%), posterior-stabilized (PS) designs in 39 cases (32.2%), in two cases (1.7%) the femoral constraint was not explicitly documented. Tibial bearing design was documented as rotating-platform (RP) in 116 cases (95.9%) and fixed-bearing (FB) in 5 cases (4.1%). Preoperative and postoperative radiographs were reviewed by a research fellow and a medical student after a training of 20 x-rays performed by an experienced knee surgeon. The two reviewers investigated postoperative complications in both cohorts - patellar resurfacing and non-patellar resurfacing. All included patients had a minimum follow-up period of 12 months. Radiological, clinical, and functional parameters were assessed using standardized measures. Both groups—resurfaced and non-resurfaced—were matched based on preoperative knee pathology and baseline functional status to ensure comparable clinical conditions across cohorts.

All knee replacements were performed by the same team of specialized knee surgeons using a standard parapatellar approach with patellar displacement. The posterior cruciate ligament was preserved in 66.1% of knees CR designs and sacrificed in 32.2% PS designs; in 1.7% the constraint type was not explicitly documented. Bone resections were started from the proximal tibial, and the femoral distal cut was performed perpendicular to the mechanical axis of the tibia, approximately 9 mm in thickness from the lateral femoral condyle. The external rotation angle was determined as the relative angle between the lines of the posterior condylar and surgical transepicondylar axes. Following the completion of bone resections, preliminary gap assessments in the flexion-extension position were conducted using a gap spacer block. The flexion and extension gaps were considered balanced if the medio-lateral gap difference was less than 2 mm. The medial release was performed according to the principles of selective release of tight structures in primary TKA. After confirming patellar tracking and the balance of the flexion and extension gaps, the tibial and femoral components were fixed with cement (120 cases; 99.2%). One case (0.8%) was classified as missing data, yielding a total of 121 cases.

Exclusion criteria included patients who did not give general consent for the use of their data, those that had undergone revision surgery at other hospitals between the time of their primary TKA and assessment, and those with periprosthetic joint infections or neuropathic pain. 167 patients were screened in total. 46 patients were excluded from analysis because of the following reasons: 29 for lack of general consent foe data use, 4 for lack of Knee Society Score (KSS), and 13 for poor or inadequate radiographs. After the initial exclusions, 121 patients were included in the final analysis.

The study was designed in line with the Declaration of Helsinki (1975, as revised in 2008) and was approved by a local ethics committee. Approval Number: 2022–02293. Informed consent was collected from all participants prior to their inclusion in the study.

Radiological measurements

Radiological images were obtained from the PACS (patient archiving and communication system). Two independent raters measured the anterior patellar and anterior femoral offset twice with a 4-week interval between interpretations in a random order. An interval of 2 weeks was also kept between the interpretations of the two imaging modalities. The two raters were blinded to both their own previous measurements and the measurements of the other rater. The intra- and inter-rater reliability was then calculated. Rater one (R1) was a medical student (T.S.) and rater two (R2) was a fully trained musculoskeletal research fellow (A.L.). Calibration was performed based on prosthesis size using the manufacturer’s specifications.

For the measurement of the anterior femoral offset all patients underwent a standard lateral x-ray of the knee, flexed at 30 degrees. For standardized measurements a reference line was drawn along the anterior cortex of the femur. In the preoperative image, the distance from this reference line to the lateral femoral condyle was measured (Fig. 1a). In the post-operative image, the distance from the reference line to the most anterior surface of the femoral shield of the prosthesis was measured (Fig. 1b).

Fig. 1.

Lateral knee x-ray (30° flexion) show anterior femoral offset measurements: preoperatively (a), from the anterior cortex to the lateral femoral condyle; postoperatively (b), to the prosthesis’ anterior surface

For the anterior patellar offset measurement, a skyline view x-ray was used. In both preoperative and postoperative images, the distance from femoral trochlea to the patellar base was measured (Fig. 2ab).

Fig. 2.

Skyline x-ray measure anterior patellar offset: preoperatively (a) and postoperatively (b), measured the distance from the femoral trochlea to the patellar base

Clinical data

Clinical data were collected postoperatively using the Knee Society Score (KSS) [29]. Assessments were conducted during routine outpatient follow-up visits by the orthopedic fellows not involved in surgery. The collected data were documented in the patient’s clinical records and subsequently entered into the hospital’s electronic data management system for analysis.

Data management and analysis

Data was collected and managed using REDCap. For statistical analysis, SPSS statistics for windows, version 26.0 (Armonk, NY: IBM Corp, USA), was used by an independent professional statistician (F.A.). A power analysis to define the needed sample size was calculated on KSS variability reported in a randomized trial (KSS knee SD ≈ 12; KSS function SD ≈ 21) [30] Assuming a one-sided hypothesis favoring resurfacing, α = 0.05, and a moderate standardized effect (Cohen’s d = 0.50), a sample of approximately 2 × 51 patients is required to achieve 80% power. With 55 resurfaced and 66 non-resurfaced patients, our cohort provides ~ 0.86 power for d = 0.50 but is underpowered for small effects, consistent with meta-analytic evidence that between-group KSS differences are very small [23].

The intra-rater reliability was calculated using ICCs with a 95% confidence interval and was calculated separately for each rater. The mean value of each rater’s measurements was calculated, and the interrater reliability was reported. Comparisons between groups were done by independent and chi-square t-tests, considering p-values < 0.05 significant.

Results

Mean values and findings for intra- and interrater reliability for pre- and postoperative femoral and patellar offset measurements expressed as ICCs are outlined in Table 1. The intra- and interrater reliability for all measurements was excellent with ICC values close to 1.0 (Table 2).

Table 1.

M1: measure 1, M2 measure 2 femoral and patellar measurements displayed excellent intra- and inter-rater reliability (ICC 0.97–0.998)

Intra-Rater-Reliability	Inter-Rater-Reliability						Mean Rater 1	Mean Rater 2	ICC (95% CI)
	Rater 1			Rater 2
	M1	M2	ICC (95% CI)	M1	M2	ICC (95% CI)
	Mean ± SD	Mean ± SD		Mean ± SD	Mean ± SD		Mean ± SD	Mean ± SD
Pre-OP Femoral	6.1 ± 2.4	6.1 ± 2.4	0.993 (0.989; 0.995)	6.2 ± 2.3	6.1 ± 2.3	0.994 (0.991; 0.996)	6.1 ± 2.4	6.1 ± 2.3	0.985 (0.978; 0.989)
Post-OP Femoral	4.8 ± 2.1	4.9 ± 2.1	0.991 (0.987; 0.994)	4.9 ± 2.0	4.9 ± 2.0	0.989 (0.985; 0.993)	4.8 ± 2.1	4.9 ± 2.0	0.986 (0.981; 0.991)
Pre-OP Patellar	24.9 ± 3.5	25.0 ± 3.5	0.990 (0.986; 0.993)	24.9 ± 3.4	24.9 ± 3.4	0.994 (0.991; 0.995)	24.9 ± 3.5	24.9 ± 3.4	0.991 (0.987; 0.993)
Post-OP Patellar	22.6 ± 3.0	22.6 ± 2.9	0.969 (0.956; 0.979)	22.5 ± 2.9	22.5 ± 2.9	0.998 (0.997; 0.999)	22.6 ± 3.0	22.6 ± 2.9	0.987 (0.981; 0.991)

Table 2.

The measurements show very high inter- and intra-measurement realibility

Radiological outcomes

Anterior femoral offset showed significant reduction from 6.06 mm (± 2.42 mm) preoperatively to 4.82 mm (± 2.03 mm) postoperatively (p < 0.001). Non-RPR patients showed significantly more relative reduction anterior in femoral offset with an average change of 6.55 mm (± 2.44 mm) to 4.75 mm (± 2.18 mm), as compared to RPR group having anterior femoral offset of 5.47 mm (± 2.28 mm) to 4.90 mm (± 1.86 mm) (not significant, p = 0.080).

With regard to the patellar offset, the data indicate that the mean patellar offset changed significantly from 24.88 mm (± 3.43 mm) preoperatively to 22.58 mm (± 2.93 mm) postoperatively (p < 0.001). Patients without RPR presented a higher reduction in the patellar offset from 25.36 mm (± 3.75 mm) to 21.58 mm (± 2.45 mm) (p < 0.001), while the values changed only from 24.3 mm (± 2.92 mm) to 23.79 mm (± 3.03 mm) in the RPR group, which was not statistically significant (p = 0.197) (Table 3).

Table 3.

Preoperative and postoperative femoral and patellar offset measurements, stratified by retropatellar resurfacing (RPR). Both femoral and patellar offsets significantly decreased. The reduction was significant only in the group without RPR. Cohen’s d values indicate the effect size, with d = 0.5 representing a half standard deviation difference

Femoral Offset	Pre-OP Mean +/- SD	Post-OP Mean +/- SD	N	Pearson r	p	Mean Diff. +/- SD	t	Cohen’s d	p
Total	6.1 +/-2.4	4.8 +/-2.0	121	0.3	0.000	1.2 +/-2.6	5.3	0.6	0.000
Without RPR	6.6 +/-2.4	4.8 +/-2.2	66	0.4	0.004	1.8 +/-2.6	5.5	0.8	0.000
With RPR	5.5 +/-2.3	4.9 +/-1.9	55	0.4	0.006	0.6 +/-2.4	1.8	0.3	0.080
Patellar Offset
Total	24.9 +/-3.4	22.6 +/-2.9	121	0.5	0.000	2.3 +/-3.1	8.2	0.7	0.000
Without RPR	25.4 +/-3.8	21.6 +/-2.5	66	0.8	0.000	3.8 +/-2.4	13.0	1.2	0.000
With RPR	24.3 +/-2.9	23.8 +/-3.0	55	0.5	0.000	0.5 +/-2.9	1.3	0.2	0.197

In the non-RPR group, the preoperative femoral offset was significantly larger than that of the RPR group, with a mean of 6.6 mm (± 2.4 mm) compared to 5.5 mm (± 2.3 mm) (p = 0.014). The postoperative patellar offset was larger in RPR group, with a mean of 23.79 mm (± 3.03 mm) as compared with the non-RPR group, at a mean of 21.58 mm (± 2.45 mm) (p < 0.001). Out of the parameters compared radiologically, changes in femoral and patellar offsets were more visible in the non-RPR group and significantly differed from femoral offset change (p = 0.009) and patellar offset change (p < 0.001) (Table 4). Out of the 121 patients, 11 (9.1%) underwent revision surgery. 10 occurred in the RPR group, and only one was found in the non-RPR group.

Table 4.

Preoperative femoral and postoperative patellar offsets show significant differences, while other differences are in the same direction but not significant. In patients without RPR, the reduction in offsets is greater than in those with RPR

RPR	Total (N = 121)	without (N = 66)	with (N = 55)
	Mean + SD	Mean + SD	Mean + SD	t	R2	P
Pre-OP Femoral Offset	6.1 (+/-2.4)	6.6 (+/-2.4)	5.5 (+/-2.3)	2.5	0.05	0.014
Pre-OP Patellar Offset	24.9 (+/-3.4)	25.4 (+/-3.8)	24.3 (+/-2.9)	1.7	0.02	0.090
Post-OP Femoral Offset	4.8 (+/-2.0)	4.8 (+/-2.2)	4.9 (+/-1.9)	0.4	0.00	0.699
Post-OP Patellar Offset	22.6 (+/-2.9)	21.6 (+/-2.5)	23.8 (+/-3.0)	4.4	0.14	0.000
Femoral Offset Change	-1.2 (+/-2.6)	-1.8 (+/-2.6)	-0.6 (+/-2.4)	2.7	0.06	0.009
Patellar Offset Change	-2.3 (+/-3.1)	-3.8 (+/-2.4)	-0.5 (+/-2.9)	6.8	0.28	0.000

Clinical outcomes

The average KSS pain score was 44.01 (± 9.99), with no statistically significant differences in scores between the RPR (44.82 ± 8.71) and the non-RPR group (43.33 ± 10.97) (p = 0.418).

The average KSS knee score was 90.44 (± 11.8), with no significant differences seen in RPR and non-RPR groups, which were 91.31 (± 10.84) and 89.71 (± 12.58), respectively (p = 0.461).

Functional scores (KSS B) were higher in the non-RPR group, which scored 91.67 (± 15.38), than in the RPR group, which scored 85.91 (± 19.41), with p value at 0.071. The use of walking aids was significantly higher in the RPR group (p = 0.012) (Table 5).

Table 5.

The only significant difference is the more frequent use of assistive devices in the RPR group clinical outcomes:

RPR		Total (N = 121)		Without (N = 66)		With (N = 55)
	N	Mean + SD	N	Mean + SD	N	Mean + SD	t	R2	P
KSS Pain Score	121	44.0 (+/-10.0)	66	43.3 (+/-11.0)	55	44.8 (+/-8.7)	0.8	0.01	0.418
KSS A Knee Score	121	90.4 (+/-11.8)	66	89.7 (+/-12.6)	55	91.3 (+/-10.8)	0.7	0.00	0.461
KSS Walking	121	45.5 (+/-8.0)	66	46.5 (+/-6.7)	55	44.2 (+/-9.2)	1.6	0.02	0.109
KSS Stair Climbing	121	44.2 (+/-9.7)	66	45.2 (+/-9.9)	55	43 (+/-9.5)	1.3	0.01	0.210
KSS walking aid (minus)	121	-0.6 (+/-2.6)	66	-0.1 (+/-0.6)	55	-1.3 (+/-3.8)	2.6	0.05	0.012
KSS B Function Score	121	89.1 (+/-17.5)	66	91.7 (+/-15.4)	55	85.9 (+/-19.4)	1.8	0.03	0.071
KSS Total Score	121	179.5 (+/-25.4)	66	181.4 (+/-24.7)	55	177.2 (+/-26.3)	0.9	0.01	0.373
OKS Score	36	21,0 (+/-7.4)	18	21.5 (+/-7.5)	18	20.4 (+/-7.5)	0.4	0.01	0.676

Discussion

The most important findings of the present study were that the non-RPR group exhibited a more substantial reduction in both femoral offset and patellar offsets compared to the RPR group. Notably, the non-RPR cohort demonstrated a greater decrease in anterior femoral and patellar offset between pre- and postoperative values compared to the RPR group. Despite these radiological changes, there were no statistically significant differences in KSS pain or knee scores between the two groups. Revision rates and use of walking aids differed between groups, as discussed below. AKP remains a prevalent concern following TKA, often impacting patient satisfaction despite advancements in surgical techniques and implant designs [3–7]. The relationship between patellar resurfacing and AKP is complex and has been the subject of extensive research. Some studies suggest that resurfacing may reduce the incidence of AKP and the need for revision surgeries [20, 31, 32]. This study found no statistically significant difference in KSS pain or knee scores between the RPR and non-RPR groups, in agreement with other high-quality studies [33, 34]. Burnett et al. [35] and Li et al. [36] similarly reported that patellar resurfacing had no significant impact on long-term pain or functional outcomes. Likewise, in a meta-analysis of randomized controlled trials, Chen et al. [23] found that resurfacing may reduce reoperation rates, however, it did not significantly reduce AKP. These findings question the rationale of patellar resurfacing solely for the purpose of minimizing postoperative anterior knee pain.

Nonetheless, contrasting evidence exists. Tang et al. performed a trial-sequential meta-analysis of 50 randomized controlled trials including over 5500 knees comparing patellar resurfacing versus non- resurfacing. They found significantly lower rates of AKP with resurfacing (13% versus 18% non-resurfacing). They also reported a lower rate of patellar clunk syndrome with patellar resurfacing [31]. Similarly, Migliorini et al. [12] and Parvizi et al. [15] also found decreased rates of AKP and revision surgeries in resurfaced groups. This discrepancy is likely due to the multifactorial nature of AKP, which can result from component malalignment, poor patellar tracking, soft-tissue imbalances, or other biomechanical factors [8–10]. Thus, the presence or absence of patellar resurfacing may only be one of many contributing variables.

A further novel aspect of this study is the analysis of anterior femoral and patellar offset changes following TKA. Patients with resurfacing showed less patellar offset reduction compared with non-resurfaced patellae. This would originally suggest that resurfacing will restore or at least maintain the anterior offset and, thus, the patellofemoral joint line intact. This, however, is likely a demonstration of baseline anatomic variation and not so much an immediate mechanical advantage of resurfacing. The resurfaced group also had less preoperative anterior femoral offset, meaning less projection of the anterior condyle and maybe more substantial trochlear wear, with proportionally smaller postoperative decreases. Thus, resurfacing appears to compensate for underlying cartilage loss but not actively restrain offset loss [37, 38].

However, our findings did not demonstrate superior function or pain outcomes for the resurfaced cohort. This discrepancy demonstrates that even native offset restoration alone does not lead to superior biomechanics or symptomatic relief if other factors such as trochlear orientation, sagittal alignment, or soft-tissue balance are not also optimized [39, 40]. This approach reflects the concept of “inverse kinematic alignment,” which combines patellofemoral balancing and anterior offset restoration using conventional instrumentation rather than exclusively mechanical targets [41]. In addition, Koutserimpas et al. placed much emphasis on dynamic, intraoperative assessment of the patellofemoral compartment to illustrate how static radiographs are powerless in the ability to detect changes in anterior offset and tracking across flexion angles [42].

Recent robotic and dynamic research suggests that minor anterior offset variations (< 5 mm) are biomechanically well tolerated and that functional alignment principles may achieve equivalent results without resurfacing in the presence of a balanced anterior compartment [43–45] Together, these findings favor the move towards multi-compartment, phenotype-driven alignment constructs, as reported by Hess et al., involving coronal, sagittal, and rotational balance to achieve truly individualized TKA [46].

From the surgical perspective, these findings mean that anterior offset modifications are due to both intraoperative decision and patient-specific anatomy rather than resurfacing status. Restoration of native anterior compartment anatomy thus requires individualized assessment of preoperative trochlear morphology and selective resurfacing in the case of severe cartilage or osseous loss [39, 40].

Interestingly, while clinical knee scores did not significantly differ between RPR and non-RPR groups, the use of walking aids was significantly higher in the RPR group. It is plausible that subtle alterations in the extensor mechanism due to offset changes could lead to downstream functional effects [47]. The increased use of walking aids in the RPR group may reflect neuromuscular or proprioceptive adaptations, though this requires further investigation. Moreover, although not reaching statistical significance, the non-RPR group trended toward better functional scores. These findings may suggest a potential functional advantage in preserving the native patella in selected patients. However, the stark contrast in revision rates − 10 revisions in the RPR group compared to just 1 in the non-RPR group - raises concerns about the long-term implications of patellar resurfacing. The higher revision rate in the RPR group aligns with earlier findings that patellar resurfacing may not always mitigate the risk of patellofemoral complications and may even contribute to them under certain anatomical circumstances.

Although patellar offset is not traditionally emphasized in assessing postoperative AKP risk, the findings in this study highlight its potential role, particularly when considered alongside other structural changes. The postoperative patellar offset was significantly higher in the RPR group, which may relate to overstuffing or altered tracking dynamics. Abnormal patellar morphologies, such as Wiberg type III, and dysplastic trochlear grooves have been associated with an increased risk of patellar maltracking and anterior knee pain (AKP) following total knee arthroplasty (TKA), even when standard alignment parameters appear normalized postoperatively. Studies have demonstrated that these anatomical variations can lead to compromised patellofemoral kinematics and may necessitate additional surgical interventions to address maltracking issues [48–50].

This study also confirms that changes in femoral and patellar offsets were statistically more significant in the non-RPR group, underscoring the mechanical impact of omitting patellar resurfacing. Yet, this mechanical alteration did not translate into worse pain or function in the non-RPR group, rather, outcomes were comparable or superior. These observations support the growing evidence questioning the routine use of patellar resurfacing in all primary TKA cases.

Interestingly, our findings demonstrate a higher revision rate in the RPR group (10 out of 55 patients) compared to the non-RPR group (1 out of 66 patients), which stands in contrast to earlier literature [15, 18, 19]. Multiple large-scale studies have shown that patellar resurfacing is associated with lower revision risks. For example, the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) observed significantly reduced revision rates in resurfaced knees, especially in minimally stabilized designs [51]. Bourne and Burnett [18, 35] similarly reported that failure to resurface the patella increased the likelihood of secondary surgeries. The higher revision rate in our RPR group may be attributed to a number of factors not fully captured in our study, such as prosthetic design, surgical technique variability, or patient-specific characteristics. Importantly, our study’s retrospective nature limits the ability to determine causality. Additionally, potential confounders such as implant malalignment or patient activity levels were not stratified, further cautioning against overinterpretation.

Study limitations

Several limitations must be acknowledged. First, this was a single-center, retrospective study, thus the findings may not be generalizable to broader populations. Second, factors such as surgeon experience, implant brand, and variations in postoperative rehabilitation may have influenced both radiological and clinical outcomes. Nevertheless, the excellent intra- and inter-observer reliability for radiological measurements supports the robustness of the measured parameters. Third, this study focused on radiologic offsets, other preoperative factors such as trochlear dysplasia and TT-TG distance also warrant consideration for a comprehensive risk model of AKP. Fourth, although the greater reduction in anterior femoral and patellar offset in the non-RPR group reduces the quadriceps lever arm and possibly diminishes extensor efficiency, this was not reflected in the clinical scores. This may mean that the KSS alone is not sufficiently sensitive to pick up on subtle biomechanical differences in extensor function. Finally, the follow-up period of at least 12 months, while adequate for short-term analyses, may not fully capture late-onset complications associated with TKA.

Conclusion

In conclusion, this study adds to the evolving narrative around patellar resurfacing in TKA. This study does not support the routine use of patellar resurfacing in primary TKA, particularly given the absence of improved pain or functional outcomes and the observation of higher revision rates in the resurfaced cohort. These findings highlight the importance of preoperative anatomical evaluation and support selective, rather than routine, patellar resurfacing in TKA.

Abbreviations

TKA

Total knee arthroplasty

AKP

Anterior knee pain

RPR

Retropatellar resurfacing

Non-RPR

Non-retropatellar resurfacing

KSS

Knee Society Score

Author contributions

T.S., MT.H., and DT.M. conceived the study and wrote the manuscript. T.S. and A.L. contributed to data acquisition, analysis, and interpretation. F.A. and T.S. performed the statistical analyses. R.H. provided radiological expertise and supported imaging interpretation. A.M.N. contributed to study design and clinical interpretation. All authors reviewed and approved the final manuscript.

Funding

The project is funded by Kantonsspital Baselland and the research group “DKF Knee” of the University of Basel (Departement Klinische Forschung). No other sources of funding will be used for this study.

Data availability

The generated and analyzed datasets used in this study are not publicly available due to institutional restrictions but are available from the corresponding author upon reasonable request.

Declarations

The authors have no financial or proprietary interests in any material discussed in this article.

Conflict of interest

There is no conflict of interests for any of the researchers involved in this study.

Ethics approval

This project complies with the regulatory requirements of the HFG and HFV. The project has been approved by the Ethics Committee.of Central Switzerland (Ethikkommission Nordwest- und Zentralschweiz (EKNZ)). Project-ID: 2022–02293.

Patient consent statement

Only patients who have signed a declaration of consent for further use of their biological material and health related data for biomedical research will be included in this study. In this written document, the patient was also informed about the basic content of further use. With the signature, the patient confirms that he has read the information, has been informed about the further use and can withdraw his consent at any time.

Permission to reproduce material from other sources

No figures, tables, or other content in this manuscript have been reproduced from external sources. All materials presented are original and created by the authors.

Clinical trial registration

This study is retrospective and was not registered as a clinical trial.

Consent to participate

Informed consent was obtained from all individual participants included in the study.