Joint laxity varies in response to partial and complete anterior cruciate ligament injuries throughout skeletal growth

Abstract

Anterior cruciate ligament (ACL) injuries are increasingly common in the skeletally immature population. As such there is a need to increase our understanding of the biomechanical function of the joint following partial and complete ACL injury during skeletal growth. In this work, we aimed to assess changes in knee kinematics and loading of the remaining soft tissues following both partial and complete ACL injury in a porcine model. To do so, we applied anterior-posterior tibial loads and varus-valgus moments to stifle joints of female pigs ranging from early juvenile to late adolescent ages and assessed both kinematics and in-situ loads carried in the bundles of the ACL and other soft tissues including the collateral ligaments and the menisci. Partial ACL injury led to increased anterior tibial translation only in late adolescence and small increases in varus-valgus rotation at all ages. Complete ACL injury led to substantial increases in translation and rotation at all ages. At all ages, the medial collateral ligament and the medial meniscus combined to resist the majority of applied anterior tibial load following complete ACL transection. Across all ages and flexion angles, the contribution of the MCL ranged from 45 to 90% of the anterior load and the contribution of the medial meniscus ranged from 14 to 35% of the anterior load. These findings add to our current understanding of age-specific functional properties of both healthy and injured knees during skeletal growth.

1. Introduction

The incidence of reported anterior cruciate ligament (ACL) injury in pediatric and adolescent patients has been steadily increasing over the past few decades (Dodwell et al., 2014). A recent study found that the most common age of ACL injury has dropped to 17 years, while the most rapid increases in injury rates occurred for 10–14 year olds (Collins et al., 2014). Skeletally immature patients experience more partial ACL injuries to either the anteromedial (AM) or posterolateral (PL) ACL bundles in comparison to adults (Kocher et al., 2002). Treatment of partial injury can vary based on the extent and location of the tear (Frank and Gambacorta, 2013). Treatments are partially informed by functional outcomes including joint stability, muscle strength, or accrual of secondary injuries.

Complete ACL injury is frequently associated with secondary injuries of tissues such as the medial collateral ligament (MCL) and the medial meniscus (Millett et al., 2002). In pediatric age groups, chances of subsequent injury to the meniscus or the reconstructed graft were higher in children compared to adults (Webster et al., 2018, 2014). Studies of skeletally mature joints have shown increased loads in tissues including the MCL and medial meniscus in the ACL-deficient joint, leading to increased functional demands (Allen et al., 2000; Daniel et al., 1994; Fisher et al., 2011; Haimes et al., 1994; Millett et al., 2002; Rasmussen et al., 2016; Sakane et al., 1999). Recent work in the porcine model has shown that the ACL is the primary stabilizer against anterior tibial loads throughout skeletal growth (Cone et al., 2019), so it follows that secondary stabilizers, such as the MCL and medial meniscus, would experience increased functional demands following ACL injuries at younger ages. However, little has been done to assess age-dependent changes in the function of these tissues following partial or complete ACL injury during growth.

The porcine model has been established as a robust surrogate for the human ACL in morphometric and functional studies in skeletally mature specimens (Cone et al., 2017b; Proffen et al., 2012; Xerogeanes et al., 1998b). Previous work in our lab has found that the porcine ACL undergoes steady increases in angular orientation over a period equivalent to newborn to late adolescence with timing and magnitude similar to orientation changes in the human ACL (Cone et al., 2017a). We have also identified bundle- and age-dependent changes in the porcine ACL biomechanical function during skeletal growth in ages corresponding with early juvenile through late adolescent groups in humans (Cone et al., 2019). Under anterior and varus loading in the intact joint, the in-situ force carried by the PL bundle declines relative to the AM bundle during adolescence. A recent study of pediatric human cadaveric specimens suggested that pediatric ACL tissue is weaker than adult tissue (Schmidt et al., 2019). However, in both humans and animal models, age-dependent changes in the impact of ACL injury on joint kinetics and kinematics during growth remain unclear.

Here, we assessed joint kinematics and soft tissue function under applied loads after partial and complete ACL transection throughout skeletal growth in the porcine model. Given that the ACL is a primary stabilizer throughout growth (Cone et al., 2019), we hypothesized that joint kinematics would increase under applied loads following partial and complete ACL injury across ages. Additionally, we hypothesized that the distribution of forces carried by secondary stabilizers would vary with age given reports of age-specific secondary medial meniscus injury rates (Millett et al., 2002; Webster et al., 2018). To test these hypotheses, we used a robotic testing system to apply loads to porcine joints ranging from early youth to late adolescence to measure changes in anterior-posterior tibial translation and varus-valgus rotation, and analyzed changes following partial and complete ACL injury in the forces carried by secondary tissues.

2. Methods

All animals were obtained from a university-owned herd (Swine Educational Unit, North Carolina State University) and were healthy and of normal size (Bouchard et al., 2017). Animals were cared for according to management practices outlined in the Guide for the Care and Use of Agricultural Animals in Teaching and Research, and experimental protocols were approved by the North Carolina State University Institutional Animal Care and Use Committee (Federation of Animal Science Societies, 2010). Hind limbs were collected from 30 female Yorkshire cross-breed pigs throughout growth: early youth (1.5 months), juvenile (3 months), early adolescence (4.5 months), adolescence (6 months), and late adolescence (18 months) (n = 6/group). Human age equivalencies were based on a combination of skeletal and sexual maturity (Cone et al., 2017b; Reiland, 1978). Specimens were wrapped in saline soaked gauze and stored at −20° C.

Prior to testing, specimens were thawed gradually overnight (~15 h) at room temperature. The femur, tibia, and fibula were cut in the central diaphysis and all soft tissue was removed from the bones. The femur and tibia were set within custom molds using an epoxy compound (Everglass, Evercoat, Cincinatti, OH). Joints were wrapped in saline soaked gauze with supplemental saline applied throughout testing to ensure the tissues remained moist.

Joint and tissue function were assessed using a 6 degree-of-freedom (DOF) robotic system (KR300 R3500, KUKA) operated by a separate controller (KRC4, KUKA) along with a 6-DOF force sensor (Omega160 IP65, ATI). These systems were integrated and controlled through a software package (SimVitro, Cleveland Clinic, Cleveland, Ohio), and data were processed using custom codes (Matlab2018b, Mathworks, Natick, MA). This system is capable of testing under kinematic and kinetic control with repeatability 0.1 mm and 0.25 N. We assessed repeatability in preliminary tests and found the system could target forces within 1.4 ± 0.7 N and moments within 0.2 ± 0.1 N*m with kinematic repeatability of <0.1 mm and <0.1°.

Custom clamps were used to attach each specimen to the system, and a point digitizer (G2X, Microscribe) was used to define the anatomic coordinate system of the joint relative to the robotic coordinate system (Livesay et al., 1997; Noble et al., 2010; Sakane et al., 1997). A passive path was established by varying joint flexion under kinematic control by 1° increments from full extension (40°) to flexion (90°) while minimizing forces and moments in the remaining 5 DOF.

A series of loading conditions were applied to the joint (Table 1). Specifically, an anterior-posterior tibial load was applied at 40°, 60°, and 90° of flexion. A varus-valgus moment was applied at 60° of flexion . This loading condition was not applied at 40° or 90° due to instability of the joint in younger specimens resulting in excessive internal-external rotation or anterior tibial translation leading to soft tissue injury at these flexion angles. We focused on loading conditions that would allow comparisons across all ages. Testing under force control was performed with 4 DOF, as flexion angle was under kinematic control and internal-external rotation was under kinematic control due to high rotational laxity in young joints. Age-specific loads and moments (Table 2) were determined based on tibial plateau size, measured via MRI. Load-displacement curves from anterior-posterior load testing revealed that loading levels were age-appropriate, as all specimens exceeded the toe region and no specimens experienced yield behavior or mechanical failure.

Table 1.

Loading Protocol for Robotic Testing. Superscripts represent the in-situ force of a specific tissue, subscripts represent the kinematic state.

Joint State	Loading Conditions	Data Acquired
Intact Joint	Anterior-posterior tibial load (40°, 60°, 90°) and varus-valgus (60°) moment	Kinematics from intact joint (Kintact) In-situ force of joint (${\mathrm{F}}_{\text{intact}}^{\text{intact}}$)
AM Bundle Transected	Anterior-posterior tibial load (40°, 60°, 90°) and varus-valgus (60°) moment Repeat (Kintact)	Kinematics from partial injury (KAMt) In-situ force of AM bundle (${\mathrm{F}}_{\text{intact}}^{\text{AM}}$, ${\mathrm{F}}_{\text{AMt}}^{\text{AM}}$)
ACL Transected	Anterior-posterior tibial load (40°, 60°, 90°) and varus-valgus (60°) moment Repeat (Kintact, KAMt	Kinematics from ACL deficient knee (KACLt) In-situ force of PL bundle and ACL (${\mathrm{F}}_{\text{intact}}^{\text{PL}}$, ${\mathrm{F}}_{\text{AMt}}^{\text{PL}}$, ${\mathrm{F}}_{\text{intact}}^{\text{ACL}}$, ${\mathrm{F}}_{\text{AMt}}^{\text{ACL}}$, ${\mathrm{F}}_{\text{ACLt}}^{\text{ACL}}$)
MCL Transected	Repeat (Kintact, KAMt, KACLt)	In-situ force of MCL (${\mathrm{F}}_{\text{intact}}^{\text{MCL}}$, ${\mathrm{F}}_{\text{AMt}}^{\text{MCL}}$, ${\mathrm{F}}_{\text{ACLt}}^{\text{MCL}}$)
LCL Transected	Repeat (Kintact, KAMt, KACLt)	In-situ force of LCL (${\mathrm{F}}_{\text{intact}}^{\text{LCL}}$, ${\mathrm{F}}_{\text{AMt}}^{\text{LCL}}$, ${\mathrm{F}}_{\text{ACLt}}^{\text{LCL}}$)
Medial Meniscus Removed	Repeat (Kintact, KAMt, KACLt)	In-situ force of medial meniscus (${\mathrm{F}}_{\text{intact}}^{\text{MMEN}}$, ${\mathrm{F}}_{\text{AMt}}^{\text{MMEN}}$, ${\mathrm{F}}_{\text{ACLt}}^{\text{MMEN}}$)
Lateral Meniscus Removed	Repeat (Kintact, KAMt, KACLt)	In-situ force of lateral meniscus (${\mathrm{F}}_{\text{intact}}^{\text{LMEN}}$, ${\mathrm{F}}_{\text{AMt}}^{\text{LMEN}}$, ${\mathrm{F}}_{\text{ACLt}}^{\text{LMEN}}$)

Table 2.

Age-Specific Applied Loads and Moments for Robotic Testing.

Age	Anterior-Posterior Load	Varus-Valgus Moment
1.5 Months	20 N	1 N*m
3 Months	40 N	2 N*m
4.5 Months	80 N	4 N*m
6 Months	100 N	5 N*m
18 Months	140 N	7 N*m

Kinematics resulting from the applied loads and moments were measured from the robotic manipulator position and recorded. Intact joint kinematics were repeated under kinematic control on the intact joint, and resulting forces and moments were recorded. The AM bundle, identified as anterior to the insertion of the lateral meniscus in the pig, was transected in the distal 1/4 of the tissue. Partial transections were only performed on the AM bundle to control for the effect of injury location. Independent behavior of the AM and PL bundles was assessed by examining the mechanical contribution of the interaction forces (forces transferred) between the bundles due to physical attachment. These forces did not exceed 3% of the total applied load across age groups, suggesting that the contribution was negligible. The loading conditions were applied to the joint under force control, and resulting kinematics were recorded as AM-deficient kinematics. Both intact and AM-deficient kinematic paths were repeated on the specimen, and resulting forces and moments were recorded.

The remainder of the ACL (PL bundle) was transected, and loading conditions were again applied. Resulting kinematics were recorded as the ACL-transected path. The kinematics from the intact, AM-deficient, and ACL-transected paths were repeated under kinematic control on the ACL-transected joint, and all forces and moments were recorded. To determine the contributions of the remaining soft tissues (MCL, LCL, PCL, medial meniscus, and lateral meniscus), each tissue was transected/removed, and kinematics from the intact, AM-deficient, and ACL-transected states were repeated while recording forces and moments.

All data was analyzed using a custom Matlab code. Anterior-posterior tibial translation (APTT) with respect to the femur was calculated as the change in position between maximum applied anterior and posterior loads. Varus-valgus rotation was calculated as rotation of the tibia between maximum varus and valgus moments. Normalization of APTT was performed relative to the sagittal width of the tibial plateau as measured from magnetic resonance imaging (MRI) scans (Simpleware 7.0, Synopsys, Chantilly, Virginia) (Cone et al., 2017a).

Kinetic data was processed using custom Matlab codes to determine the contribution of each tissue to joint function under anterior translation, varus torque, and valgus torque (Supplemental Table 1). This was performed by applying the principle of superposition described previously (Rudy et al., 1996). Resultant forces were calculated as the mathematical resultant of forces in the anterior-posterior, medial-lateral, and proximal-distal directions. Force data were normalized to applied forces to facilitate comparisons across ages. These values may exceed 100% (when one tissue is dominant) or be negative due to experimental noise and the presence of individual tissue loads at the neutral position. Thus, even if the net joint load is near zero at the neutral position, opposing tissues may carry a small amount of load, typically under 5% of the target load within an individual tissue (i.e. <1 N under anterior tibial loads for the 1.5 month group, or <7 N for the 18 month group). For example, in the passive path position at 60° of flexion, one 6 month old specimen carried the following loads in each tissue in the anterior-posterior direction: ACL 6.6 N anterior, MCL 0.5 N posterior, LCL 0.3 N posterior, PCL 0.6 N anterior, medial meniscus 1.4 N posterior, lateral meniscus 0.2 N anterior, and cartilage contact 6.1 N posterior. For this same joint, the anterior force in the joint at maximum anterior translation was 97 N for the intact joint and −8 N following ACL transection, leading to a total anterior contribution of 105 N by the ACL (108% of the applied anterior load). Percent contributions of the AM and PL bundles were calculated as the bundle contribution divided by the combined contribution of both bundles.

Statistical analyses were performed using commercial software (JMP Pro 13.0, SAS Institute, Cary, North Carolina). Normality was confirmed for each data set. For APTT and tissue contributions to anterior tibial translation, a multi-way ANOVA test was performed with flexion angle as a repeated measure and age and state as main effects. Varus-valgus rotation was analyzed using a two-way ANOVA with age and state as main effects. Tukey’s post-hoc analysis was performed (α = 0.05). Complete summary data and statistical results are reported in the Supplemental Materials.

3. Results

Loss of ACL function impacted anterior-posterior tibial translation (APTT) normalized to joint size in response to applied loads (Fig. 1, Supplemental Table 2). Statistical analyses revealed significant interactions between injury state and age (p < 0.05) as well as injury state and flexion angle (p < 0.05). Average values for the intact joint ranged from 0.17 to 0.34 across flexion angles. AM bundle transection led to normalized APTT values ranging from 0.21 to 0.40 on average. Statistically significant increases relative to the intact joint were only found for the late adolescent (18 month old) age group (41–52% increase relative to the intact joint across all flexion angles tested, (p < 0.05)). Across all ages, values for the AM-deficient group were statistically significant from the intact condition at 60° of flexion (p < 0.05) but not 40° or 90° of flexion (p > 0.05). Complete ACL transection further increased average normalized APTT values to 0.42–0.74. These values were approximately 2-fold higher than intact and AM-deficient conditions across all ages and flexion angles (p < 0.05). In addition to normalized APTT, absolute values of APTT (not normalized to joint size, Supplemental Table 10) revealed similar differences due to partial and complete ACL transection (Supplemental Fig. 1, Supplemental Table 3).

Fig. 1.

Normalized anterior-posterior tibial translation (APTT) in response to an applied anterior-posterior tibial load increased following partial and complete ACL transection. Two-fold increases occurred following complete ACL transection. Values were normalized to the anterior-posterior length of the tibial plateau. Data represented as points, with bars showing mean and 95% confidence interval. * denotes p < 0.05 from intact state, + denotes p < 0.05 from intact and AM deficient states.

ACL transection also resulted in increased varus-valgus rotation (Fig. 2, Supplemental Table 4). AM bundle transection resulted in 0.5–1.5° increases from the intact state (p < 0.05, Supplemental Table 4), while increases due to complete ACL transection ranged from approximately 3°-5° compared to intact values across ages (p < 0.05). Additionally, varus-valgus rotation decreased with increasing age (p < 0.05). For the intact state, varus-valgus rotation decreased from an average of 25.4° in youth to 6.3° in late adolescence. For the ACL-transected state, values decreased from 30.1° in youth to 10.6° in late adolescence.

Fig. 2.

Varus-valgus rotation increased under applied moments following both partial and complete ACL transection across ages. Data represented as points, with bars showing mean and 95% confidence interval. * denotes p < 0.05 from intact state, + denotes p < 0.05 from intact and AM deficient states.

In addition to changes in kinematic parameters, we studied the force distribution across tissues. In the intact state, the ACL served as the primary soft tissue stabilizer to anterior tibial translation across ages and flexion angles, carrying 75–111% of the applied anterior load (Supplemental Table 5). Division of this load across bundles varied with age, with substantial contributions from both the AM and PL bundles in younger age groups and the AM bundle resisting the majority of the anterior load in adolescence (average 80–91% across flexion angles, Fig. 3, Supplemental Figs. 2 and 3, Supplemental Tables 6 and 7). Following AM bundle transection, the PL bundle carried the majority of applied anterior load under anterior tibial translation across ages at 60° of flexion (Fig. 3). Specifically, at 60° and 90° of flexion, the average PL bundle contribution dropped no lower than 88% following AM bundle transection (Supplemental Table 7). While the contributions of the other soft tissues were minimal in both intact and partial injury states, they were substantial following ACL transection.

Fig. 3.

In-situ force contributions of the AM and PL bundles of the ACL relative to all other soft tissues in response to an applied anterior-posterior tibial load are shown for all ages for the intact, AM-deficient, and ACL-transected states at 60° of flexion normalized to the total force carried in the joint. The PL bundle resists the majority of the load across ages in the partial transection state, and other tissues carry all of the functional contributions in the ACL-transected state. Data represented as points, with bars showing mean and 95% confidence interval. Percent contribution is normalized to the total anterior force resisted by the joint under an applied load.

As such, we assessed the forces carried by specific secondary stabilizers, namely the medial meniscus and the MCL which consistently carried the majority (75–100% of the applied load), after ACL transection. Although the demand under anterior translation on secondary soft tissues did not increase following AM-bundle transection, contributions of the MCL and medial meniscus increased significantly following complete ACL transection (Fig. 4). In the ACL transected state, the MCL carried the greatest proportion of the applied anterior tibial load across ages, an average of 52–90% of the total load across flexion angles and ages (Fig. 4, Supplemental Table 8). The medial meniscus also carried up to 35% of the anterior load on average in the ACL transected state. Of note, the contribution of the MCL was greater at 90° of flexion than either 40° (p < 0.05) or 60° (p < 0.05) of flexion in the ACL transected state. The contribution of the medial meniscus was greater at 40° (p = 0.01) and 60° (p < 0.01) compared to 90° of flexion (Supplemental Table 9).

Fig. 4.

The MCL and medial meniscus carry the majority of in-situ force within the ACL-transected joint under an applied anterior tibial load across age groups. Data represented as points, with bars showing mean and 95% confidence interval. Percent contribution is normalized to the total force resisted by the joint under an applied load. * denotes p < 0.05 from both 40° and 60° of flexion.

4. Discussion

In this study, we analyzed changes in joint function in the porcine model during growth following partial and complete ACL injury. Under applied anterior-posterior tibial loads, complete ACL transection resulted in greater anterior-posterior tibial translation throughout skeletal growth, and partial ACL transection resulted in increased translation only in adolescence. The increased impact of partial ACL injuries in older age groups suggests that responses to partial injury found in studies of skeletally mature joints should not be assumed in younger joints. These findings may indicate the need for age-specific treatment of partial injuries if properly diagnosed through physical examination and MRI. If a partial injury does not greatly alter joint stability in younger patients, there may be merit in nonsurgical treatment and physical therapy versus invasive surgical reconstruction. Under varus-valgus moments, both complete and partial ACL transection led to increased varus-valgus rotation at all ages, although rotational changes of 0.5–1.5° following partial ACL transection may be lower than a clinically relevant threshold (Bates et al., 2017). Together, these findings partially affirmed our first hypothesis that joint kinematics would increase following partial and complete ACL transection for all age groups. The PL bundle was the primary restraint to anterior tibial loads following AM bundle transection. After complete ACL transection, the MCL and medial meniscus provided the majority of functional restraint across flexion angles and age groups compared to the LCL, the lateral meniscus, and cartilage contact. Aside from the MCL and the medial meniscus, these other tissues contributed ~5–10% individually, and not more than 25% as a group. Overall, these findings were contrary to our second hypothesis that the role of the secondary stabilizers following ACL transection would vary with age.

This study reported the immediate ex-vivo changes in kinematics resulting from loss of ACL function, and these changes are consistent with prior work. We reported a two-fold increase in anterior-posterior tibial translation following partial ACL injury only in the late adolescent group and 2- to 3-fold increases due to complete ACL injury across all ages. A human cadaveric study using biplanar radiography and manual joint manipulation found that complete, but not partial, ACL transection resulted in a significant 1.5- to 2-fold increase in anterior tibial translation under anterior tibial loads (Lintner et al., 1995). Previous studies testing human specimens on robotic testing systems have also found that anterior translations doubled following ACL transection (Woo et al., 1998) while a porcine study reported that anterior tibial translations tripled following ACL transection (Ishibashi et al., 1997). Additionally, we found an increase in varus-valgus rotation following both partial (5–20% increases) and complete (20–70% increases) ACL injuries at all ages. This is consistent with 2-fold increases in varus-valgus rotational laxity found for ACL-sectioned knees subjected to passive flexion-extension paths (Li et al., 2007). Although our results may suggest difficulty in detecting partial ACL injuries in skeletally immature joints through clinical exams (due to smaller changes relative to the intact joint), prior studies in human patients reported greater sensitivity to partial ACL injuries in skeletally immature patients through clinical examination (76.5%) compared to MRI analysis (52.9%) (Kocher et al., 2001). The sensitivity of clinical exams to partial ACL injuries is similar in mature patients, with a meta-analysis of 8 studies reporting an average sensitivity of 68% (Leblanc et al., 2015). In addition, more sophisticated imaging in combination with clinical examinations may allow better identification of partial ACL injuries (Cavinatto et al., 2018). A study employing 3 T MRI scanners to diagnose partial ACL injuries found that sensitivity for partial ACL tears was 77%, representing a significant improvement from the earlier value in MRI sensitivity, matching the reported sensitivity of clinical examination by Van Dyck et al. (2011).

With recently increasing numbers of pediatric and adolescent ACL injuries, an improved understanding of the functional properties of the soft tissues of the immature knee is increasingly necessary. A need for subsequent reconstruction has been noted in cases of injuries affecting >50% of the midsubstance of the ACL and also in injury mainly to the PL bundle (Kocher et al., 2002). In order to investigate the biomechanical behavior of the ACL in both healthy and injured states, many groups have performed studies in large animal models, establishing important findings regarding the function of the ACL under both isolated loading conditions and complex conditions simulating activities such as walking and landing (Boguszewski et al., 2011; Nesbitt et al., 2014; Rosvold et al., 2016; Woo et al., 1998). Previous studies have considered the importance of the ACL in stabilizing the knee against anterior tibial translation, reporting that the ACL provides the majority (80–125%) of the overall restraint in both humans and animals (Boguszewski et al., 2011; Butler et al., 1980; Cone et al., 2019; Xerogeanes et al., 1998). Furthermore, our findings on the function of the AM and PL bundles under anterior tibial translation in the intact adolescent joint were in line with previous reports showing that that 60–70% of the force carried in the ACL was carried through the AM bundle in adolescent pigs (Xerogeanes et al., 1998a). The increased role of the PL bundle following AM bundle injury, and that of the MCL and medial meniscus following complete ACL transection may provide insight into secondary injuries in young patients with partial ACL injuries, although more work is needed. Along these lines, a recent study in an in-vivo sheep model reported that AM bundle transection had varied results in joint kinematics and cartilage health between animals, with noticeable effects on some animals and minimal changes in others (Barton et al., 2019).

In this study, we did not find age-related differences in the immediate loading of the MCL and medial meniscus. Age-specific differences in secondary injuries following ACL injury have been reported. Specifically, patients over the age of 15 suffered from medial meniscus tears at a higher rate (Dumont et al., 2012). However, additional factors related to different species, patient age and weight, or time to surgery may contribute to this discrepancy (Dumont et al., 2012; Millett et al., 2002). Our findings regarding increased loads in the medial meniscus following ACL transection are in agreement with a previous study which found significant increases in the loads in the medial meniscus in ACL deficient knees compared to intact knees (Allen et al., 2000). Interestingly, recent work studying the strain in skeletally mature human cadaveric MCL tissues during simulated ACL failure events did not identify an increase in MCL strain concomitant with the ACL injury event (Schilaty et al., 2019). This may suggest that changes in the MCL behavior are more directly identified under isolated anterior tibial drawer tests compared to complex kinematic simulations, although the impact is skeletal maturity is not known. Future in-silico and in-vivo studies will aim to investigate the impact of tissue adaptation to ACL injury throughout growth.

This work used a large animal model, the Yorkshire cross-breed pig. The porcine ACL has closer structural and functional properties to human ACL properties in comparison to other large animals, and our mature age group corresponded well to previously published works on porcine ACL function (Cone et al., 2017b; Proffen et al., 2012; Xerogeanes et al., 1998). However, a possible limitation to this work is related to the lack of available information on the similarities or differences between the porcine stifle and the human knee joint during skeletal growth. Current literature suggests that changes in the orientation and geometry of the ACL are similar between porcine and human growth. Limited data on mechanical properties of the skeletally immature human ACL indicates that ACL stiffness increases with maturity (Schmidt et al., 2019) as in the pig model (Cone et al., 2019), although establishing a similar time course in humans is difficult due to a lack of specimens. Additionally, we lack the information to claim direct similarities in the in-vivo mechanics of the ACL during pediatric growth. As with all studies involving large animal models, differences exist relative to humans in the growth timeline, locomotion modality, and body size. Of specific relevance to this study, the porcine joint reaches full extension at approximately 30–40° of flexion between the tibia and the femur. As such, care should be taken in extending these findings to a human population where full extension occurs near 0°. Another limitation is the inclusion of only specimens from female animals, and in the future, a similar study will be completed in a male population to assess the effect of sex on parameters measured in this study. Additionally, this study involved only passive soft tissue restraints, whereas the behavior of the joint in-vivo is influenced by the activity of the muscles crossing the tibiofemoral joint. Further analysis of the differences between normalized and raw force data is limited by the current lack of a normalization of rotations. In the future, studies could incorporate measurement of a moment arm between the center of joint rotation and the application point of the moments. Finally, we only studied partial ACL injury to the AM bundle, although comparison to injuries to the PL bundle would be an interesting direction for future studies.

In conclusion, in response to applied loads, partial ACL injury led to increased anterior tibial translation only in late adolescence and increased varus-valgus rotation at all ages. Complete ACL injury led to increased translation and rotation at all ages. The PL bundle, not the MCL or medial meniscus, carried the majority of load in case of a partial ACL injury to the AM bundle. However, with the additional loss of PL bundle function, the MCL and medial meniscus provided functional resistance against excessive anterior tibial translation. These findings add to our understanding of knee function during growth in normal and injured skeletally immature joints.