Introduction

Patellofemoral joint complications are a major and frequent cause for revision surgery after total knee arthroplasty (TKA) [9, 20, 21, 23]. In TKA with or without primary patellar resurfacing, the number of patients with patellofemoral problems appears to be comparably high [4, 22, 25]. The patients typically complain about persistent or new onset of anterior knee pain [19]. In patients with unresurfaced patella, patellar maltracking, patellofemoral overloading and progression of patellofemoral osteoarthritis (OA) are common pain generators [23].

Radionucleotide imaging such as bone scintigraphy (bone scan) has been used to obtain more functional loading information of the patellofemoral joint and assess overloading due to patellar maltracking [11, 20, 32]. The hot patella sign has been controversially discussed [6]. Recent studies demonstrated a clear correlation between increased bone tracer uptake (BTU) and anterior knee pain after TKA. Possible explanations of increased patellar BTU are an increased in vivo loading at the patellofemoral joint and subsequent bone remodelling. Moreover, patients with a hot patella who underwent secondary patellar resurfacing reported a significant symptomatic relief [1]. Some authors even recommended the use of bone scans as a screening tool for diagnosis of patellar-related problems [1]. However, the main drawback of bone scans is their poor ability to localize the increased BTU [28].

Hybrid SPECT/CT, which combines the strengths of 3D bone scan (SPECT) and CT, has been highlighted as an important tool for evaluating patients after TKA [2, 7, 10, 12,13,14,15,16]. In these studies, SPECT/CT was particularly helpful in identifying patellofemoral problems such as maltracking, overloading or progression of osteoarthritis [15]. It is the combined assessment of mechanical, structural and functional information, which offers a richer source for establishment of the correct diagnosis [6, 8, 15]. As patellar tracking and loading of the patella are influenced by TT–TG distance, patellar height, thickness and tilt as well as TKA component position, it was our hypothesis that these parameters significantly correlate with patellar BTU intensity and localization in SPECT/CT.

The purpose of the study was to investigate whether aforementioned patellar criteria, mechanical alignment and TKA component position influence the intensity and the BTU distribution pattern of the unresurfaced patella after TKA. This is the first study assessing in vivo loading of the patella using SPECT/CT.

Materials and methods

A total of 64 knees of 62 consecutive patients (mean age ± standard deviation 66 ± 11, range 23–87 years, male-to-female ratio 31:31) who underwent primary TKA without patellar resurfacing between 2011 and 2013 were prospectively included. All patients underwent clinical and radiological examination in a specialized knee clinic, including standardized radiographs (anterior–posterior and lateral weight bearing, patellar skyline view) and Tc-99m-HDP-SPECT/CT before, 12 and 24 months after TKA. Demographic data such as age, gender, side and type of primary TKA were noted.

Tc-99m-HDP-SPECT/CT was performed using a hybrid system (Symbia T16, Siemens, Erlangen, Germany) which consists of a pair of low-energy, high-resolution collimators and a dual-head gamma camera and an integrated 16 × 0.75 mm slice thickness CT. All patients received a commercial 500–700 MBq Tc-99m-HDP injection (CIS Bio International Sur Yvette, France). Planar scintigraphic images were taken in the perfusion phase (immediately after injection), the soft tissue phase (1–5 min after injection) and the delayed metabolic phase (2 h after injection). SPECT/CT was performed with a matrix size of 128 × 128, an angle step of 32 and a time per frame of 25 s 2 h after injection.

SPECT/CT images were analysed on 3D reconstructed images. Rotational (internal–external rotation), sagittal (flexion–extension, anterior–posterior slope) and coronal (varus–valgus) position of the tibial and femoral TKA components was assessed using a customized analysis software (OrthoExpert©, London, UK) [26].

Measurements of BTU including intensity and anatomical distribution pattern were also performed from 3D data [26]. BTU intensity was measured in 3D for each anatomical area, as reflected by a previously validated localization scheme [14]. The localization scheme equally divides the patella into eight different areas to accurately map the BTU activity (Fig. 1). Maximum intensity values were recorded, and ratios between the respective value in the measured area and the background BTU (proximal midshaft of the femur) were calculated.

Fig. 1
figure 1

3D SPECT/CT tracer uptake analysis in the coronal, sagittal and axial planes using a customized software (OrthoExpert©, London, UK). The quantification of the maximum, minimum and mean uptake values was done in eight different anatomical areas of the patella. The patellar grid in those images of a 73-year-old woman 14 months after CR unresurfaced TKA shows significantly higher uptake in the superior–lateral patellar area

Full size image

The patellar height, according to the modified Insall–Salvati index and the Caton–Deschamps index, and the patella thickness were measured using lateral knee radiographs in 30° flexion [8, 24]. The patellar tilt was measured in a “skyline view” using the Sasaki index [27]. The distance between the tibial tuberosity and the trochlear groove (TT–TG) was measured using axial images from the CT scan [5].

Ethical approval was obtained from the Ethics Committee of Northwestern and Central Switzerland (EKNZ 2016-01890). All procedures performed were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.

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

Statistical analysis

All data were analysed by an independent professional statistician using IBM SPSS Statistics for Windows, version 24.0 (Armonk, NY: IBM Corp, USA). Continuous variables were described using means with standard deviations and medians with ranges. All statistical analysis was carried out non-parametrically: To compare KSS before and after operation, Wilcoxon signed-rank tests were performed and Spearman rho was calculated to identify any correlations between BTU and TKA component position and patellar measurements (p < 0.05). A post hoc analysis using G*Power, version 3.1.9 (University of Kiel, Germany) showed that with the given N = 64 and a two-tailed alpha <0.05, correlations of r > 0.35 could be found with a power of 80%.

Results

The mean TKA alignment measurements and the patellar thickness, height and the TT–TG distance in the investigated population are shown in Table 1. The mean KSS score significantly improved from 125 ± 27 preoperatively to 184 ± 19 post-operatively (Table 2). The highest mean BTU was measured in the superior and joint adjacent parts of the patella. The second highest BTU was found in the inferior joint adjacent patellar parts. The anterior patellar areas were lower than the joint adjacent areas. The lowest BTU was found in the inferior anterior patellar area (Table 3). A significant correlation was found between valgus alignment of the femoral TKA and increased BTU in the lateral patellar areas (p < 0.05). In addition, a valgus-aligned femoral TKA component correlated with increased medial BTU, although this observation did not reach statistical significance (Table 4). External rotation of the tibial TKA component correlated with increased BTU in the articular lateral superior patellar parts. No correlation was found between the position of the tibial TKA component in the coronal or sagittal plane. There was also no significant correlation between patellar thickness and TT–TG distance to increased BTU at the patella.

Table 1 TKA component measurements using 3D reconstructed CT images and patellar measurements including patellar thickness, patellar height and TT–TG distance
Full size table
Table 2 Pre- and post-operative outcome scoring using KSS
Full size table
Table 3 Mean BTU measurements of patellar areas in SPECT/CT
Full size table
Table 4 Correlation between TKA component measurements and BTU measurements in different patellar areas
Full size table

Discussion

The most important findings of the present study were as follows: Firstly, valgus alignment of the femoral TKA component was significantly correlated with increased BTU in the patella, in particular the lateral patellar areas. This finding clearly indicates that TKA position in the coronal plane influences in vivo loading of the patella in patients after TKA, a fact which is to date underreported. It appears that knee surgeons focus more on the influence of rotational and sagittal TKA alignment on patellar tracking. It is well established that internal rotation of the femoral TKA leads to increased stress of the lateral patellar facet, which is due to an increased Q-angle and lateral patellofemoral tracking [3, 21]. In the present study, it was found that with more valgus in the femoral component BTU increased in all patellar areas, but particularly in the lateral patellar areas. One explanation could be that femoral TKA valgus alignment results in increased laterally directed forces, which in its maximum extent tend to sublux the patella laterally [21]. Another explanation might be the fact that the contact zone of patella and trochlea changes in valgus-aligned femoral TKA. The trochlear geometry might also play a role here.

It is also a well established fact that flexion of the femoral TKA component increases the patellofemoral pressure and leads to a “pseudo” patella baja. Here an increase in the stress at the superior patellar areas was found [23]. Furthermore, a recent study in native knees has shown that a patella baja correlates with increased BTU in all patellar areas. In addition, a higher patellofemoral tilt angle led to increased BTU in the inferior and superior patellar areas [29]. Another effect of the sagittal component alignment on the patellar kinematics was recently reported in the cadaveric study by Keshmiri et al. [17], as femoral component aligned in 5° of flexion significantly altered patellar mediolateral shift. However, in the present study no correlation between sagittal TKA alignment and BTU in any patellar area was found. There was also no evidence that a patella baja led to more loading of any patellar areas. This is in clear contrast to a biomechanical study by Singerman et al. [30]. Secondly, a specific BTU pattern in patients after TKA with unresurfaced patellae was identified and described in detail. There were significant BTU differences between the patella areas in SPECT/CT. As previously shown, BTU represents the in vivo loading of the tibiofemoral and patellofemoral joint [29]. To date, the normal and pathological BTU pattern after TKA is unknown. In the last 20 years, the discussion with regard to this topic was limited to the question whether a hot patella is a clinically relevant finding in patients after TKA. Some authors stated that the hot patella sign is an important abnormal finding indicating overloading of the patellofemoral joint. Others did not see any clinical value in reporting a hot patella sign as it includes both the normal remodelling of an unresurfaced patella and abnormal loading of the patella.

In a recent retrospective study, Ahmad et al. [1] found a significant correlation between the “hot patella” sign and post-operative patellofemoral problems in 55 symptomatic patients after TKA. In this study, patients underwent bone scans. The authors concluded that patients with a “hot patella” sign are more likely to benefit from a secondary patellar resurfacing. However, this study was done on 2D and the present study used 3D data. Our group has shown in many previous publications that 3D data should be used to open a more realistic and detailed window into the bone homoeostasis [11, 12, 16, 26, 31]. The “hot patella” sign, which is defined as greater BTU in the patella than in the ipsilateral distal femur or the proximal tibia [18], is a rather rough and broad assessment of patella BTU. Based on our data, the “hot patella” sign is not useful for assessment of patellar loading. However, when the patellar bone is divided into several different areas, BTU evaluation in 3D might be helpful for guidance of treatment. In a similar way to the “hot patella” concept, Awengen et al. [2] found significantly higher BTU intensity in symptomatic TKA in all the different anatomical regions in the patella using SPECT/CT. The SPECT/CT algorithm presented here localizes and quantifies patellar BTU in eight different areas of interest. In future, analysis of patellar BTU after TKA might be beneficial in establishing more differential diagnoses and possible reasons for anterior knee pain following TKA. This algorithm might then also guide further treatment in unhappy patients after TKA [16].

The present study bears a considerable number of limitations. This was a prospective series of TKA in a teaching hospital setting. Hence, a team of different surgeons operated on a consecutive series of patients. In addition, the patients investigated were not specifically evaluated for anterior knee pain, which would allow to differentiate between normal and abnormal BTU pattern in symptomatic and asymptomatic patients after TKA. In this study, this was only done based on KSS, which is not a patellofemoral-specific outcome instrument. Furthermore, increased patellar BTU might not only be due to the factors investigated, but also due to other reasons such as residual patellar osteoarthritis.

Conclusion

The strongest correlation of increased patellar BTU was found with a femoral TKA in valgus orientation. These findings highlight the clinical importance of femoral TKA position in the coronal plane with regard to post-operative patellar tracking. In unhappy TKA with anterior knee pain, these facts should be carefully considered.