Finite Element Analysis of Patella Alta: A Patellofemoral Instability Model

Abstract

Background:

This study aims to provide biomechanical data on the effect of patella height in the setting of medial patellofemoral ligament (MPFL) reconstruction using finite element analysis. The study will also examine patellofemoral joint biomechanics using variable femoral insertion sites for MPFL reconstruction.

Methods:

A previously validated finite element knee model was modified to study patella alta and baja by translating the patella a given distance to achieve each patella height ratio. Additionally, the models were modified to study various femoral insertion sites of the MPFL (anatomic, anterior, proximal, and distal) for each patella height model, resulting in 32 unique scenarios available for investigation.

Results:

In the setting of patella alta, the patellofemoral contact area decreased, resulting in a subsequent increase in maximum patellofemoral contact pressures as compared to the scenarios with normal patellar height. Additionally, patella alta resulted in decreased lateral restraining forces in the native knee scenario as well as following MPFL reconstruction. Changing femoral insertion sites had a variable effect on patellofemoral contact pressures; however, distal and anterior femoral tunnel malpositioning in the setting of patella alta resulted in grossly elevated maximum patellofemoral contact pressures as compared to other scenarios.

Conclusions:

Patella alta after MPFL reconstruction results in decreased lateral restraining forces and patellofemoral contact area and increased maximum patellofemoral contact pressures. When the femoral MPFL tunnel is malpositioned anteriorly or distally on the femur, the maximum patellofemoral contact pressures increase with severity of patella alta.

Clinical Relevance:

When evaluating patients with patellofemoral instability, it is important to recognize patella alta as a potential aggravating factor. Failure to address patella alta in the setting of MPFL femoral tunnel malposition may result in even further increases in patellofemoral contact pressures, making it essential to optimize intraoperative techniques to confirm anatomic MPFL femoral tunnel positioning.

Introduction

The stability of the patellofemoral joint is maintained by a complex interaction between static soft tissue restraints, the dynamic action of the quadriceps, and bony anatomy about the knee. Patella alta, or high riding patella, is frequently associated with patellar instability^1,2 as the patella only engages the bony constraints of the trochlear groove in higher degrees of knee flexion³. Several studies^2,4-7 have provided radiographic methods to quantify patellar height with associated values defining patella alta and patella baja, or low riding patella. While several studies^8-10 have described increased contact pressures in patients with patella alta, the biomechanical effects of treatments aimed at patellar instabilty, including medial patellofemoral ligament (MPFL) reconstruction, have yet to be determined.

While patella alta frequently plays a role in patients with patellar instability, it often does so in combination with other pathology. Soft tissue restraints play a critical role in patellar stability. The medial patellofemoral ligament (MPFL) is the primary soft tissue restraint to lateral translation of the patella^11-13, acting as a check-rein during the first 30° of knee flexion prior to the patella engaging the trochlear groove^14,15. Following acute lateral patellar dislocation, the MPFL is the most consistently injured ligamentous structure^16-18. With nonoperative management of acute lateral patellar dislocations, recurrent dislocation is not uncommon, with resultant episodic dislocation more common in patients with multiple instability events^19,20. MPFL reconstruction aims to restore the form and function of the native MPFL. However, experimental studies have shown that even anatomic MPFL reconstruction may alter contact pressures within the patellofemoral joint²¹, while non-anatomic reconstruction and excessive graft tension may lead to increased medial patellofemoral contact pressures, resultant medial instability, or even catastrophic failure^15,22,23.

More frequently than not, several factors play a role in patellar instability. Patella alta and MPFL insufficiency are often concomitantly identified in this patient population1, and failure to address both pathologies can lead to treatment failure²⁴. To our knowledge, no studies have examined patellofemoral biomechanics after MPFL reconstruction in the setting of patella alta or patella baja. Therefore, the purpose of this study was to use finite element (FE) analysis in order to better define patellofemoral biomechanics and patellofemoral contact pressures before and after MPFL reconstruction in the setting of patella alta and baja while also analyzing the effects of the MPFL insertion site. We hypothesize that patella alta will increase the magnitude of contact pressures within the patellofemoral joint. Additionally, we hypothesize that changes in the femoral MPFL insertion site will further alter patellofemoral mechanics and contact pressures in knees with concomitant pathology.

Methods

A previously validated patellofemoral finite element (FE) model²⁵ was modified to meet the aims of this study. Briefly, the patellofemoral model was generated using surfaces obtained from a magnetic resonance (MR) image of a cadaveric knee specimen. Bones were modeled using three-dimensional rigid elements since bone is significantly stiffer than the soft tissues, which were the structures of interest for the purposes of this study. Similar to previous computational models of patellofemoral biomechanics^26-29, the model did not include the fibula. The cartilage, patellar tendon (PT), and quadriceps tendon (QT) were modeled using 8-noded hexahedral elements. The MPFL and the medial patellotibial ligament (MPTL) were also modeled using hexahedral elements. The meniscus was not included since the study focused on the patellofemoral interaction with the knee fixed in 30° of flexion. The model included 39,515 elements.

The viscoelastic nature of the cartilage was simplified to linear elastic material properties (E=12 MPa, =0.45) based on previous literature^28-30. The tendon and ligaments were modeled as hyperelastic with the material properties adapted from stress-strain and force-displacement curves reported previously^31-33. The reconstructed MPFL assumed material property characteristics of the split anterior tibialis tendon³⁴. The anterior tibialis tendon was used to mirror the allograft used in previous corresponding experimental study³⁵. The anatomic reconstruction insertion site and dimensions were considered to be the same as the native MPFL ligament model. The model was validated against a corresponding experimental study³⁵.

Patella Height

A validated finite element patellofemoral knee model was modified to account for four different patella heights. The validated model depicts a ‘normal’ knee with a Caton-Deschamps Index ratio of 1.036. The patella height was adjusted to study patella alta and baja ratios of 1.4, 1.2, and 0.8 (Figure 1). To create various patellar heights, the patella in the ‘normal’ model was translated a given distance to achieve each patella height ratio. The new nodal coordinates for the entire model were output and used as the starting coordinates for the patella alta and patealla baja models.

Figure 1.

The patella baja (Caton-Deschamps 0.8), “normal” (Caton-Deschamps 1.0), and patella alta (Caton-Deschamps 1.2 and 1.4) finite element models. Each model includes the quadriceps tendon (purple), patellar tendon (red), cartilage (yellow), medial patellofemoral ligament (blue), and medial patellotibial ligament (green).

MPFL Insertion Site

Each patella height model was also modified to study multiple MPFL reconstruction insertion sites. Specifically, the insertion sites were repositioned in increments of 5 mm from the anatomic position on the femur^13,37. The femoral insertion was modeled at 5 mm and 10 mm anterior, proximal, and distal to the anatomic position with the patellar insertion remaining at the anatomic location, thus creating seven unique insertion scenarios (Figure 2). The different insertion sites were modeled for each patellar height, resulting in 28 unique MPFL reconstruction scenarios and four native MPFL model variants.

Figure 2.

Various femoral insetion sites for the reconstructed medial patellofemoral ligament. The red dot represents the anatomical insertion site. The black dots show locations 5mm and 10mm anterior, proximal, and distal to the anatomic insertion site.

Boundary conditions

For each patellar height and MPFL insertion site, the knee was positioned at 30° of flexion with the femur fixed in all directions. The tibia was free to translate and rotate about the anterior-posterior axis (z-axis), allowing anterior-posterior translation and varus-valgus rotation. The patella was free to rotate and displace in all directions. The quadriceps was physiologically loaded to 178 N, with separate forces through each individual component of the quadriceps as previously described^38-40. With the quadriceps loaded, the tibia and femur were fixed in all directions and the patella was displaced laterally 10 mm. The resultant patellar restraining force, contact pressure, and contact area were compared for the different scenarios described above. Analyses were completed using Abaqus/Standard (Version 6.12-1; Dassault Systèmes Simulia, Providence, RI).

Results

A total of 32 scenarios were available for finite element analysis using variable femoral insertion sites for MPFL reconstruction in the setting of normal patellar height, patella baja, and patella alta. Lateral restraining force was smallest for the native MPFL scenario regardless of patellar height. MPFL reconstruction increased the lateral restraining force in all scenarios as compared to the native MPFL, with anterior translation of the femoral MPFL insertion site resulting in the largest lateral restraining forces. Patella alta, with separate scenarios created for Caton-Deschamps ratios of 1.2 and 1.4, consistently decreased the lateral restraining force (Table 1).

Table I.

Restraining Force (N)

	Ratio 0.8	Ratio 1.0	Ratio 1.2	Ratio 1.4
Anatomical – Native	78.6	80.0	90.3	70.0
Anatomical – Reconstruction	136.2	148.9	147.6	103.5
Anterior 5mm	145.4	163.1	160.1	113.6
Anterior 10mm	157.6	176.9	166.5	119.7
Distal 5mm	142.8	159.4	154.0	109.7
Distal 10mm	138.3	154.5	156.2	112.2
Proximal 5mm	115.1	125.1	127.5	90.7
Proximal 10mm	102.9	114.1	113.1	79.8

The lateral restraining force (N) of the patella after 10mm lateral displacement for each patella height and corresponding femoral insertion site.

The patellofemoral joint contact force increased following MPFL reconstruction regardless of femoral insertion site as compared to the native MPFL scenario. Contact forces were similar in the setting of normal patella height (Caton-Deschamps 1.0), patella baja (Caton-Deschamps 0.8), and mild patella alta (Caton-Deschamps 1.2); there was a noted decrease in contact forces in the setting of severe patella alta (Caton-Deschamps 1.4) (Table 2).

Table II.

Contact Force (N)

	Ratio 0.8	Ratio 1.0	Ratio 1.2	Ratio 1.4
Anatomical – Native	183.9	190.6	184.3	118.2
Anatomical – Reconstruction	220.4	226.9	231.2	141.3
Anterior 5mm	224.3	232.4	233.3	147.7
Anterior 10mm	229.5	236.5	226.7	152.4
Distal 5mm	222.8	230.0	232.7	146.1
Distal 10mm	217.1	222.1	226.0	148.9
Proximal 5mm	208.6	215.4	215.3	131.5
Proximal 10mm	201.3	211.7	204.2	124.3

The patellofemoral contact force (N) after 10mm lateral patella displacement for each patella height and corresponding femur insertion site.

Patellofemoral contact area was greatest in the setting of patella baja and gradually decreased with interval increases in patellar height (Table 3). The differences in contact area based on patellar height drove subsequent changes in maximum patellofemoral contact pressures, which were decreased in patella baja and increased for patella alta after anatomic MPFL reconstruction. Particularly large maximum patellofemoral contact pressures were noted with anterior and distal femoral insertion sites in the setting of severe patella alta (Caton-Deschamps 1.4) (Figure 3).

Table III.

Contact Area (mm²)

	Ratio 0.8	Ratio 1.0	Ratio 1.2	Ratio 1.4
Anatomical – Native	117.0	67.6	59.9	30.8
Anatomical – Reconstruction	139.3	71.8	54.4	48.0
Anterior 5mm	143.7	86.7	55.5	49.7
Anterior 10mm	143.2	95.2	59.3	49.3
Distal 5mm	137.8	83.8	54.4	48.8
Distal 10mm	140.2	89.1	65.2	45.4
Proximal 5mm	132.6	67.2	63.0	42.9
Proximal 10mm	131.3	60.1	59.4	35.4

The patellofemoral contact area (mm2) after 10mm lateral patella displacement for each patella height and corresponding femur insertion site.

Figure 3.

The contact pressure and area for each loading scenario after 10mm lateral patella displacement. The value below each image is the maximum contact pressure (MPa).

Discussion

Patellar stability is maintained by a complex interplay between static soft tissue restraints, dynamic muscle action, and osseous anatomy about the knee. MPFL reconstruction has become a popular and effective method for treatment of MPFL pathology, either from acute injury or chronic insufficiency^18,41,42. The present study provides biomechanical data using an established finite element knee model following MPFL reconstruction while taking into consideration the effect of patellar heightas well as the effect of the femoral insertion point during MPFL reconstruction. We found that MPFL reconstruction increased the lateral restraining force of the patella as compared to the native knee, with decreasing lateral restraining force noted with increasing patellar height. Additionally, we noted increased maximum patellofemoral contact pressures following MPFL reconstruction in the setting of patella alta, with notably large increases in maximum contact pressures when the femoral insertion site was placed anterior or distal to the anatomic insertion. Several of these findings warrant further discussion.

The femoral insertion of the MPFL and its subsequent effect on patellofemoral biomechanics has been studied extensively throughout the literature^12,18,22. Stephen et al.²³ found that MPFL femoral tunnel placement that was proximal or distal to the anatomic insertion point significantly increased contact pressures within the patellofemoral joint. The present study found that the femoral insertion site studied in isolation had variable effects on maximum patellofemoral contact pressures, but in the setting of patella alta (Caton-Deschamps 1.4), misplaced femoral tunnels dramatically increased maximum contact pressures within the patellofemoral joint. We find this particularly important given the general difficulty in consistently identifying the anatomic femoral insertion of the MPFL^18,43, as well as the high frequency of patella alta in this patient population1. Clinically, both femoral tunnel malpositioning and untreated patella alta have been identified as potential causes of failure following MPFL reconstruction^15,24. Our findings echo these clinical concerns, as femoral tunnel malpositioning in the setting of patella alta increases maximum contact pressures within the patellofemoral joint.

Even prior to concerns regarding MPFL femoral tunnel positioning, Dejour and colleagues identified patella alta as one of the contributing factors to patellar instability, suggesting tibial tubercle distalization in patients with patellar instability that exhibited this radiographic finding1. Since that time, several studies have investigated the biomechanical effects of patellar height^8-10. Luyckx et al.⁸ investigated the influence of patellar height on patellofemoral joint biomechanics and found that patella alta decreased contact area. Similarly, we found that patella alta caused a decrease in contact area. Although this study does not compare directly to the previous studies that looked at the patellofemoral biomechanics throughout flexion^8-10, the current study allowed us to determine the role patella height had on lateral restraining force. We found that patella alta decreased lateral restraining force even after MPFL reconstruction, suggesting that unaddressed patella alta may biomechanically serve as risk factor for recurrent episodic patellar dislocation. Our findings provide a biomechanical basis for addressing patella alta at the time of surgery in this patient population.

Inherent limitations of finite element models, including boundary conditions and simplified material properties, exist for the present study. In this model, the MPFL and medial patellotibial ligament were modeled based on anatomic data reported in the literature^44-47 and were not specimen-specific. Although defining these ligaments from medical images would be ideal, it presents a challenge due to the thin anatomy and complex nature of the attachment site. With advances in medical imaging, future models may be able to define all soft tissues on a specimen-/subject- specific basis. Additionally, this model did not incorporate the meniscus since static loading options were considered. To study various loading conditions and angles of flexion, the meniscus should be included. Also, the model boundary conditions do not capture in vivo scenarios; however, the biomechanics do mimic in vitro loading conditions. The model boundary constraints should be considered when applying these predicted trends to clinical situations. Lastly, this study was limited to one angle of flexion, 30 degrees, and no graft pretension. Future work should continue to investigate various tibiofemoral flexion angles and different MPFL graft pretensions to gain a better understanding of patellar height and MPFL insertion location on the patellofemoral biomechanics.

In conclusion, treatment of patellar instability requires recognition and understanding of the multiple factors that affect patellar stability, often leading to an algorithmic-based treatment approach. The present study provides the first data on patellofemoral biomechanics following MPFL reconstruction in the setting of variable patellar height and femoral insertion sites. We found that persistent patella alta after MPFL reconstruction decreases lateral restraining force, potentially providing a biomechanical explanation for clinical failures of isolated MPFL reconstruction in the setting of patella alta. Additionally, we found that patella alta increased maximum patellofemoral contact pressures, particularly with anterior and distal femoral tunnel malpositioning. We feel these findings are important, as patellar instability is often caused by a constellation of pathologies, and failure to address all of these factors at the time of surgery may result in treatment failure.