Orientation changes in the cruciate ligaments of the knee during skeletal growth: a porcine model

Abstract

Musculoskeletal injuries in pediatric patients are on the rise, including significant increases in anterior cruciate ligament (ACL) injuries. Previous studies have found major anatomical changes during skeletal growth in the soft tissues of the knee. Specifically, the ACL and the posterior cruciate ligament (PCL) change in their relative orientation to the tibial plateau throughout growth. In order to develop age-specific treatments for ACL injuries, the purpose of this study was to characterize orientation changes in the cruciate ligaments of the Yorkshire pig, a common pre-clinical model, during skeletal growth in order to verify the applicability of this model for pediatric musculoskeletal studies. Hind limbs were isolated from female Yorkshire pigs ranging in age from newborn to late adolescence, and were then imaged using high field strength magnetic resonance imaging. Orientation changes were quantified from the magnetic resonance images using image segmentation software. Statistically significant increases were found in the coronal and sagittal angles of the ACL relative to the tibial plateau during pre-adolescent growth. Additional changes were observed in the PCL angle, Blumensaat angle, intercondylar roof angle, and the aspect ratio of the intercondylar notch. Only the sagittal angle of the ACL relative to the tibial plateau experienced statistically significant changes through late adolescence. The age-dependent properties of the ACL and PCL in the female pig mirrored results found in female human patients, suggesting that the porcine model may provide a pre-clinical platform to study the cruciate ligaments during skeletal growth.

Graphical abstract

High-field magnetic resonance imaging is used to analyze changes in the orientation of the cruciate ligaments of the knee in the porcine model throughout skeletal growth. Age-dependent changes occur in parameters including the coronal and sagittal angles of the anterior cruciate ligament relative to the tibial plateau. These changes are similar to those previously found in human knees throughout growth.

INTRODUCTION

Anterior cruciate ligament (ACL) injuries of the knee are increasingly common in the pediatric population.¹ Incidence rates of ACL injury in patients under 18 years of age have been increasing rapidly, with the overall rate of ACL injury in this population nearly tripling over the past twenty years from 17.6 to 50.9 per 100,000 people aged 13–20.² Concurrently, the rate of ACL reconstruction in patients under 14 years old has increased by 11% every year since 2006.³

Clinical approaches to pediatric ACL injuries can be divided into conservative and surgical treatments. Conservative treatments avoid the risk of introducing additional damage to musculoskeletal structures (including open physes) and include functional bracing and activity modification to address the instability associated with ACL deficiency.⁴ Surgical reconstruction of the disrupted ACL in the pediatric patient can require modification of the traditional adult techniques to avoid direct interruption of the growth plates, and can be defined as all-epiphyseal, partial-transphyseal, transphyseal, or extraphyseal techniques depending on the graft placement.^{5, 6} Additionally, treatment of pediatric ACL injury can be complicated by altered anatomic features relative to adult knees, requiring specialized procedures which are compatible with both the current anatomy and future anatomic changes throughout growth. Despite the modifications employed in current treatment of pediatric ACL injuries, complications include limb length discrepancy, high rates of graft failure requiring surgical revision, and early-onset osteoarthritis (OA).⁵ Early onset osteoarthritis is a major concern with young patients, as the incidence of OA is approximately 50% at 10 year follow-up after ACL injury regardless of patient age.⁷ Between the increased pediatric injury incidence rates, challenges to surgery in growing patients, and long term risk factors, there is an increased need for improved understanding of the age-dependent structure and function of the ACL in both healthy and injured states.

One potential complication in developing age-appropriate treatments is the changing structure and function of the skeletally immature knee. Decreases in both cellularity and vascularity in the ACL with increasing age have previously been reported in skeletally immature animal models.⁸ Moreover, changes in matrix organization of the ACL have been found during growth, with increasing collagen alignment in older specimens.⁹ In humans, the angular orientation of the ACL measured relative to the tibial plateau increases substantially from birth through adolescence.¹⁰ Specifically, the orientation in both the sagittal and coronal planes changes over time with increases in the coronal plane of approximately 20° and increases in the sagittal plane of 15°, creating a more “vertical” orientation over time. The posterior cruciate ligament (PCL) also experiences an approximately 15° increase in the relative angle of its horizontal and vertical components, and an overall decrease in the horizontal-to-vertical aspect ratio with age.¹⁰ By furthering the general knowledge of these changes in healthy patients, age-specific treatments can be designed to work within the growing joint and facilitate future changes due to growth and maturation.

A unique challenge in studying pediatric ACL structure and function is the limited availability of human cadaveric specimens. As such, validated large animal models provide a mechanism to study injuries and treatments during skeletal growth.¹¹ Relative to other animal models, the adult porcine ACL is closest to human ACL in terms of dimensional parameters such as relative ACL width and length.¹² Additionally, the adult porcine ACL proved more similar to the human ACL in terms of the magnitude and direction of its in situ force under anterior tibial loading than either the sheep or the goat.¹³

A complicating factor in performing translational studies in a large animal model is establishing the relative ages of the models. Age can be described using several different scales, including chronological age, skeletal age, and sexual age.¹⁴ Pigs and humans experience growth on different chronological scales, with pigs experiencing far more rapid growth and shorter lifespans. Human skeletal growth is often indexed based on a left-hand radiograph,¹⁵ a system which does not easily translate to the porcine model. As such, the most commonly applied age equivalency between pigs and humans is based on a combination of skeletal age and sexual age, which defines age relative to pubertal changes. The age groups in this study are representative of the spectrum of growth, namely, newborn (0 month old), juvenile (1.5 and 3 month old), early adolescent (4.5 month old), adolescent (6 month old), and late adolescent (18 month old) age groups.¹⁶

In order to establish a pre-clinical large animal model for skeletally immature knee injuries, the objective of this study was to characterize the changes in the orientation of the porcine ACL and PCL during post-natal skeletal growth, and to compare these changes to corresponding human values available in the literature. Given the large changes in cruciate ligament orientation observed in humans during skeletal growth and the similarities in the ACL and PCL in skeletally mature pigs and humans, we hypothesized that the porcine model would exhibit significant changes in ACL and PCL orientation throughout skeletal growth.

METHODS

Study Design

A total of 36 stifle (knee equivalent) joints were collected from female Yorkshire pigs (one joint per animal) at 0, 1.5, 3, 4.5, 6, and 18 months of age (n=6 per time point, Swine Educational Unit at North Carolina State University). The animals used in this study were obtained from a university owned herd, and all animals were healthy and of normal size. Swine were cared for according to the management practices outlined in the Guide for the Care and Use of Agricultural Animals in Teaching and Research, and their use in the current experimental protocols was approved by the North Carolina State University Institutional Animal Use and Care Committee.¹⁷ Animals were euthanized by one of two IACUC approved methods, intravenous injection of sodium pentobarbital, or penetrating captive bolt euthanasia followed by jugular exsanguination. The limbs were isolated from the pigs immediately following euthanasia and were stored at −20 degrees Celsius until testing.

Magnetic Resonance Imaging

Limbs were removed from the freezer and allowed to thaw at room temperature prior to imaging. All limbs were imaged at full extension, which in the porcine model is approximately 35 degrees of flexion. Tissues were wrapped with saline-soaked gauze for imaging. Due to the small feature size in 0 month old specimens, imaging of the 0 month old limbs was performed in a separate scanner with improved resolution relative to the scanner employed for the other age groups. Magnetic resonance (MR) imaging for 0 month old limbs was performed in a 9.4-Tesla Bruker BioSpec 94/30USR machine (Bruker BioSpin Corp, Billerica, MA) using a 3D fast low angle shot scan sequence (3D-FLASH, flip angle: 10°, TR: 38 ms, TE: 4.42 ms, acquisition time: 13 hours 18 minutes, FOV: 30x30x30 mm) and a 35 mm volume coil with a voxel size of 0.1x0.1x0.1 mm and no gap between slices. MR imaging for 1.5, 3, 4.5, 6, and 18 month old limbs was performed on a 7.0-Tesla Siemens Magnetom machine (Siemens Healthineers, Erlangen, Germany) with a 28 channel knee coil (Siemens Healthineers, Erlangen, Germany) using a double echo steady state scan sequence (DESS, flip angle: 25°, TR: 17 ms, TE: 6 ms, acquisition time: 24 minutes, FOV: 123x187x102 mm) with a voxel size of 0.42x0.42x0.4mm and no gap between slices. These sequences were selected as they allow for visualization of the boundaries between all musculoskeletal soft tissues within the knee joint at an adequately high resolution for post-processing and analysis.

Image Post-Processing

Image post-processing was performed using commercial tissue segmentation software (Simpleware 7.0, Synopsys, Chantilly, VA) using the “Distance”, “Angle”, and “Disconnected Angle” tools. All image processing and measurements were performed by a single author (SC) and analysis techniques had high intrareader repeatability (Intraclass Correlation Coefficient (ICC)=0.7113–0.9942) and interreader repeatability (ICC=0.7091–0.9702) (Table S-5). The orientation and horizontal-to-vertical aspect ratios of the cruciate ligaments were collected using definitions established in the literature.¹⁰ As shown in Figure 2A, the sagittal angle of the ACL was collected by measuring the angle between the anterior edge of the ACL relative to the anterior-posterior line of the tibial plateau. This measurement was taken from the first full slice of the ACL approached from the medial aspect of the joint. The coronal angle of the ACL was collected from the medial edge of the ACL relative to the tibial plateau, as shown in Figure 2C. This measurement was performed on the first scan slice with a full image of the ACL moving from the anterior aspect of the knee.

Figure 2.

The angle of orientation of the ACL in both the sagittal and coronal planes is measured as shown in panels A and C. The sagittal angle (B) increases significantly between 3 and 18 months (late adolescence equivalent) while the changes in coronal angle (D) becomes insignificant after 4.5 months (early adolescence) in the porcine model. Data are presented as mean ± standard deviation, and age groups with different letters are statistically significant from one another. For example, for the sagittal ACL angle (panel B below), the 0 and 1.5 month age groups are not statistically different since they share a letter (p>0.05). The 3 and 4.5 month age groups are statistically different because they do not share a letter (p<0.05).

Analysis of the posterior cruciate ligament (PCL) included measurement of the PCL angle and the horizontal-to-vertical ratio, as seen in Figure 3A. The angle of inclination of the intercondylar roof was the angle between the intercondylar roof and the long axis of the femur in the sagittal plane. Notch width was calculated in the coronal plane following methods established in previous literature, in which the notch width index (NWI) was defined as the ratio of the intercondylar notch width (horizontal length) and the width of the distal femur. These widths were measured from a coronal image in the medial-lateral direction at the superior-inferior midpoint of the notch (one-half of the vertical notch length).

Figure 3.

The PCL angle, measured between the horizontal and vertical components shown in panel A, increases with increasing age (B). Significant increases occur between consecutive age groups between 1.5 and 4.5 months of age, equivalent to an early adolescent age in humans. The horizontal-to-vertical ratio (C), calculated from these components, nearly doubles throughout skeletal growth, with significant changes occurring through late adolescence (18 months). Data are presented as mean ± standard deviation, and age groups with different letters are statistically significant from one another. For example, for PCL angle (panel B), the 0 and 1.5 month age groups are not statistically different since they share a letter (p>0.05). The 1.5 and 3 month age groups are statistically different because they do not share a letter (p<0.05).

The Blumensaat line-to-ACL angle was described as the angle between the Blumensaat line and the anterior surface of the ACL in the intercondylar notch in the sagittal plane. The angle of inclination of the intercondylar roof was the sagittal plane angle of the intercondylar roof relative to the long axis of the femur. Notch width was calculated in the coronal plane following methods established in previous literature, in which the notch width index (NWI) was defined as the ratio of the intercondylar notch width and the width of the distal femur.^18–21 These widths were measured from a coronal image along the medial-lateral line which intersected the notch at one-half the total notch height.

Statistical Analysis

Statistics were performed using SPSS (v21.0, IBM, Armonk, NY). All specimens (36/36) were used in data analyses and statistical testing. Normality was verified for each data set with the Kolmogorov-Smirnoff test. Analysis for all data sets consisted of a one-way analysis of variance (ANOVA) with specimen age as the independent variable. Tukey’s and Games-Howell post hoc tests were used for further analysis depending on if the groups had equal or unequal variances, respectively. Overall significance was set at p<0.05. Data are presented as mean ± standard deviation, with different letters representing statistically significant differences between age groups (Figures 2–4). Alternatively, figures with symbolic representation of statistically significant differences can be found in the Supplemental Information (Supplemental Figures 1–3).

Figure 4.

Notch width aspect ratio is calculated as a ratio of the horizontal (“H”) and vertical (“V”) notch measurements (A). Values decrease with increasing age, and statistically significant changes occur prior to early adolescence. Data are presented as mean ± standard deviation, and age groups with different letters are statistically significant from one another. For example, for notch width aspect ratio (panel B below), the 0 and 1.5 month age groups are not statistically different since they share a letter (p>0.05). The 1.5 and 3 month age groups are statistically different because they do not share a letter (p<0.05).

RESULTS

Magnetic resonance images of porcine knees ranging from newborn to 18 month age groups revealed overall size increases, gross morphology changes, and differences in tissue orientation (Figure 1). In young specimens, the femur and tibia had a higher proportion of epiphyseal cartilage relative to the epiphyseal bone. By the adolescent age groups, only a thin layer of articular cartilage remained. Open growth plates were observed throughout the early adolescent stages, reaching a state of near or complete fusion by the late adolescent (18 month) age.

Figure 1.

Magnetic resonance images of porcine stifle joints in the sagittal and coronal planes at 0, 1.5, 3, 4.5, 6, and 18 months of age. Scale bars are 10mm.

Analysis of the sagittal angle of the ACL relative to the tibial plateau revealed a statistically significant effect due to age (Figure 2B, Table S1), as mean values increased from 30° in the 0-month old group to 60° in the 18 month old group which resulted in an overall change of 30° throughout growth. The most rapid change in sagittal angle occurred between 3 and 4.5 months (11° change between the group means over a 1.5 month timespan, p<0.05). Interestingly, the mean sagittal angle continued to increase by 11° between the 6 and 18 month time points (p<0.05). Analysis of the coronal angle revealed statistically significant effects due to age through the 4.5 month (early adolescent) age group (Figure 2D, Table S1), with insignificant change occurring afterwards, which represented an overall increase of 41° throughout skeletal growth. Similar to the sagittal angle, the most rapid changes occurred between 3 and 4.5 months (15° change on average over 1.5 months, p<0.05); however, unlike the sagittal angle where major changes continued throughout adolescence, no statistically significant differences (3° and 5°) were found between 4.5 and 6, and 6 and 18 months, respectively (p>0.05).

The angle of the PCL increased from 112° to 142° from 0 to 18 months of age, and these changes were significant between consecutive age groups up to the 4.5 month (early adolescent) time point (Figure 3B, Table S2). The largest changes in PCL angle occurred between 1.5 and 3, and 3 and 4.5 months of age with differences of 12° and 9° on average, respectively (p<0.05). No statistically significant changes were found following the onset of adolescence, with a mean difference between 6 and 18 month groups of only 2° over the course of 12 months (p>0.05). Likewise, the PCL horizontal-to-vertical aspect ratio increased by nearly two-fold from 0.50 to 0.93 during skeletal growth (Fig. 3C, Table S2, p<0.05).

The angle between the Blumensaat line and the ACL decreased from 10° to 5.6° from birth to skeletal maturity in the porcine model (Table 1). These changes occurred gradually and only reached statistical significance across multiple time points (e.g., 1.5 month to 6 months), opposed to other orientation changes which were significant between sequential groups. Additionally, the angle of incidence of the intercondylar roof experienced a 5-fold decrease with increasing age (mean values of 32.9° and 5.7° at birth and 18 months, respectively) (Table 1); however, these changes were only significant between sequential age groups in youth (1.5 to 3 months) and early adolescence (4.5 to 6 months) (p<0.05).

Table 1.

Data on the angle between the Blumensaat line and the ACL and the angle of incidence of the intercondylar roof. Both angles decrease with increasing age. Data are presented as mean ± standard deviation, and age groups with different letters are statistically significant from one another. For example, for Blumensaat angle, the 0 and 1.5 month age groups are not statistically different since they share a letter (p>0.05). The 0 and 18 month age groups are statistically different because they do not share a letter (p<0.05).

Age [months]	Blumensaat-ACL Angle [deg]	Intercondylar Roof Angle [deg]
0	10.0±1.3a,b	32.9±4.3a
1.5	12.4±2.1a	35.7±5.5a
3	10.8±1.5a	18.9±6.8b
4.5	9.6±2.6a,b	14.4±2.9b
6	7.4±1.2b,c	6.1±1.0c
18	5.6±2.3c	5.7±3.1c

The notch width aspect ratio (Figure 4, Table S3), calculated as the ratio of the horizontal length to the vertical length of the intercondylar notch, and notch width index (Table S3) decreased with increasing age during the juvenile age groups (0–3 months), with minimal changes throughout the remainder of skeletal growth (p>0.05). Specifically, 84% of the total change in aspect ratio occurred prior to 3 months of age. This data reflected the changing bony morphology near the cruciate ligaments, and coincided with the change in epiphysis composition from a highly cartilaginous tissue to more defined bony structure.

DISCUSSION

In order to study age-specific treatments in childhood musculoskeletal injuries, including ACL tears, a pre-clinical model exhibiting age-dependent changes similar to those seen in humans must be validated. Additionally, the angle of the PCL changes during skeletal growth. All of the changes occur in an age-dependent manner, in accordance with the stated hypothesis.

Humans experience significant orientation changes in the angular orientation of the ACL relative to the tibial plateau, with both the sagittal and coronal angles increasing by approximately 20° relative to the tibial plateau from early childhood through adulthood.¹⁰ The current study found that the sagittal and coronal ACL angles also increase in the porcine model during skeletal growth. Interestingly, the human data vary between male and female patients, with the sagittal angle of the ACL increasing through late adolescence in females and reaching a plateau during adolescence in male patients. These changes are concurrent with other sex-dependent differences, as significantly smaller intercondylar notches were found in the distal femur of adolescent female pigs relative to male pigs.¹⁹ Female specimens in this study matched sex-specific findings in humans. The sagittal angle of the porcine ACL increases in the female pig throughout late adolescence (18 months) and human data from female subjects also change through late adolescence (17–20 years).¹⁰ However, the coronal ACL angle increases only through an earlier stage of adolescence (6 months) in the pig. Likewise, the coronal angle of human female patients reaches a plateau at approximately 65° shortly after the onset of puberty.¹⁰ This suggests that future work in the porcine model needs to include similar characterization of soft tissue changes in a male population in order to determine the validity of the pre-clinical model for determining potential sex-based differences. Interestingly, the specific timing of these changes are similar in the human and porcine models, with major increases in the sagittal angle through late adolescence, while the coronal angle increases up to the onset of adolescence, with little change thereafter (Table 2, Table S4).

Table 2.

Data showing the specific timing of orientation changes in the cruciate ligaments between humans and the porcine model. Both species show continued changes in the sagittal angle throughout adolescent growth, unlike the changes in the coronal ACL angle and the PCL angle which occur primarily during the early stages of growth. Data are presented as a percentage of the total orientation change from early youth through late adolescence in each species.

	Human Early Youtha to Pre-Adolescenceb	Porcine Early Youthc to Pre-Adolescenced	Human Pre-b to Late-Adolescencee	Porcine Pre-d to Late-Adolescencef
Sagittal ACL Angle	~50%	~60%	~50%	~40%
Coronal ACL Angle	~71%	~75%	~29%	~25%
PCL Angle	~75%	~77%	~25%	~23%

Age groups are defined as follows:

^a0–3 years,

^b10–13 years,

^c1.5–3 months,

^d4.5–6 months,

^e18–20 years,

^f18 months. Human data estimated from Kim et al.¹⁰

While the PCL tends to be less frequently studied due to lower injury incidence rates (around 8–10% of the rate of ACL injury),^22–25 it serves as an additional comparison between the porcine model and humans. The angle between the horizontal and vertical aspects of the PCL experiences a significant increase during skeletal growth (~15° increase) in human subjects.¹⁰ Similar changes were seen in the porcine model (~30° increase). However, a comparison of horizontal-to-vertical component ratios of the PCL between human and porcine images differs as the ratio decreases with increasing age in humans, and the ratio increases with increasing age in pigs. This disparity could be caused by several factors including different anatomy of the femur, altered loading patterns, and a major variation in the flexion angle of the joint at full extension (0° in humans, ~35° in pigs). Previous studies of pre-clinical knee models have found that the human PCL is relatively wider than the porcine PCL; however, the porcine model exhibits a similar length between the femoral and tibial insertion sites to the human knee.¹²

In both human and porcine studies, the angle of incidence of the intercondylar roof decreased with increasing age, in a manner potentially related to the altered angle of the ACL.¹⁰ Intercondylar roof angle has been correlated with ACL injury and tibial spine fractures at the distal insertion of the ACL in previous literature, with significant differences in the average intercondylar roof angle between the two injury mechanisms.²⁶ When considering the angle between the Blumensaat line and the ACL, an interesting relationship between the models appears. In human subjects, the relationship between age and angle is only statistically significant when considering values from patients 2 years old and under.¹⁰ In the porcine model, no statistically significant changes in this parameter are found in consecutive time points, and the changes across all ages are quite small.

Anatomical differences between the human knee joint and the porcine stifle joint make a comparison of notch widths difficult, as human notch width index is traditionally measured at the level of the popliteal groove, yet a direct comparison of this anatomic landmark is not available in pigs across all age groups. However, the notch measurements in this work are taken using similar metrics to previous studies.¹⁹ As such, the notch width aspect ratio may aid in understanding trends in the growth of the porcine joint, as the bony anatomy can affect the structure and function of the soft tissues within the joint.

A limitation of this study is the inclusion of only female porcine specimens. Given the promising results of this initial work, future plans include similar studies on male pigs throughout skeletal growth in order to study the impact of both age and sex. However, using a combination of sexual and skeletal maturity as a comparison scale, many similarities between pigs and humans were found in terms of ACL and PCL orientation. However, using sexual maturity as a comparison scale, many similarities between pigs and humans were found in terms of ACL and PCL orientation. Additionally, the specific accuracy and repeatability measures established in this work are only applicable to in vitro work, and would need to be re-established for any in vivo tests as limb positioning would be more challenging. Finally, the study is limited by comparison of specific porcine data to approximated human data collected from the literature. Without a wide range of age-specific human scans, this limits the analysis performed in this work to a subjective comparison instead of a statistical one.

The anterior and posterior cruciate ligaments are two of the primary soft tissue stabilizers within the knee. Other properties, including geometric measures, biomechanical properties, and intrinsic force distributions must be studied in order to develop a more complete understanding of the changes during post-natal growth and their impact on total knee behavior. Other properties, including geometric measures, material properties, and intrinsic force distributions must be studied in order to develop a more complete understanding of the changes during post-natal growth and their impact on total knee behavior. Further work is needed to expand data that can be studied in both human and porcine models through non-invasive methods, including imaging based analysis of tissue size and geometry (e.g. tissue volume, CSA, and length). These parameters can be evaluated at a higher resolution through high-strength MR imaging in both humans and the porcine model, which may be particularly important for smaller (i.e. younger) specimens.^27–29 This may provide further verification that the porcine model has the potential to mimic complex processes involved in the growth of pediatric patients. Following further investigation of non-invasive parameters, characteristics including the biomechanical properties of musculoskeletal soft tissues will be investigated in the skeletally immature porcine model and/or through correlation analysis of imaging parameters and tissue properties.^{30, 31}

In summary, this work demonstrates that the female porcine model experiences age-dependent changes in the orientation of the cruciate ligaments that mirror prior findings in skeletally immature humans during post-natal growth. This suggests that the porcine model may be appropriate for studying the ACL during normal growth, ACL injury, and response to clinical interventions in future studies. Given the growing prevalence of pediatric ACL injuries, an appropriate pre-clinical model will be instrumental for studying the long-term effects of ACL reconstructions, including graft remodeling and return of joint function.