In vivo Compositional Changes in the Articular Cartilage of the Patellofemoral Joint following Anterior Cruciate Ligament Reconstruction

Abstract

Objective:

To compare T1ρ relaxation times of the medial and lateral regions of the patella and femoral trochlea at 6 and 12 months post-anterior cruciate ligament reconstruction (ACLR) on the ACLR and contralateral limb. Greater T1ρ relaxation times are associated with a lesser proteoglycan density of articular cartilage.

Methods:

Twenty individuals (11 males, 9 females; age=22±3.9yrs; mass=76.11±13.48kg; height=178.32±12.32) who underwent a previous unilateral ACLR using a patellar tendon autograft. Magnetic resonance images from both limbs were acquired at 6 and 12 months post-ACLR. Voxel by voxel T1ρ relaxation times were calculated using a five-image sequence. The medial and lateral regions of the femoral trochlea and patellar articular cartilage were manually segmented on both limbs. Separate limb (ACLR and contralateral limb) by time (6-months and 12-months) ANOVAs were performed for each region (P<0.05).

Results:

For the medial patella and lateral trochlea, T1ρ relaxation times increased in both limbs between 6 and 12-months post-ACLR (medial patella: P=0.012; lateral trochlea: P=0.043). For the lateral patella, T1ρ relaxation times were significantly greater on the contralateral limb compared to the ACLR limb (P=0.001). The T1ρ relaxation times of the medial trochlea on the ACLR limb were significantly greater at 6 (P=0.005) and 12-months (P<0.001) compared to the contralateral limb. T1ρ relaxation times of the medial trochlea significantly increased from 6 to 12-months on the ACLR limb (P=0.003).

Conclusion:

Changes in T1ρ relaxation times occur within the first 12 months following ACLR in specific regions of the patellofemoral joint on the ACLR and contralateral limb.

The incidence of anterior cruciate ligament (ACL) injury is highest between the ages of 14 and 25 years (1) when individuals are more likely to engage in dynamic physical activity. Unfortunately, those who sustain an ACL injury are at increased risk of developing posttraumatic osteoarthritis (OA) regardless of ACL reconstruction (ACLR) and rehabilitation (2, 3). Approximately 50% of individuals are reported to have signs of radiographic OA in the tibiofemoral (TF) and/or patellofemoral (PF) joints within 12 years following ACLR (4, 5). Younger individuals who desire to be engaged in a high level of physical activity and develop posttraumatic OA early in life may experience a disproportionately high level of disability compared to individuals who develop idiopathic OA later in life (6). Therefore, a better understanding of the deleterious joint tissue changes that occur following ACL injury may improve the capacity to detect and manage the early development of posttraumatic OA in young physically active individuals.

Compositional changes to articular cartilage often predate morphological alterations in the early phases of OA development (7, 8). Proteoglycans are macromolecules embedded in the extracellular matrix of articular cartilage and play a critical role in force attenuation in the tissue (9). Proteoglycan depletion within TF articular cartilage is an early marker associated with OA development (10, 11). T1ρ magnetic resonance imaging (MRI) relaxation times are associated with the proteoglycan density of articular cartilage and have been used to evaluate early in vivo compositional changes in articular cartilage (12, 13). Increased T1ρ MRI relaxation times, interpreted as decreased proteoglycan density, have been reported in the TF cartilage of the ACLR limb compared to uninjured limbs within the first 2 years following ACL injury (14–16). Although the prevalence of PF joint OA is reported to be similar to that of TF joint OA following ACLR(17), only a few studies have investigated longitudinal changes in T1ρ relaxation times at the PF joint following ACLR (18, 19). The results from these studies demonstrate significant increases in T1ρ relaxation times from baseline to 6 months post ACLR occurring in the femoral trochlea articular cartilage but no changes in the entire patellar articular cartilage on the injured limb (18, 19). These previous studies assessed T1ρ relaxation times of the articular cartilage averaged across the entire PF cartilage and did not evaluate changes to different regions of the patella or femoral trochlea (18, 19), which could allow for a more sensitive analysis.

The PF joint is complex, as different regions of the patella articular cartilage contact multiple regions of the femoral trochlea during normal knee movements required for activities of daily living. Changes to joint tissue metabolism (20) and PF joint biomechanics (21, 22) following ACLR may lead to alterations in cartilage composition in different regions of the PF joint. Furthermore, there is evidence to support a higher prevalence of medial PF cartilage damage as compared to lateral PF cartilage damage among individuals with PF OA (23). Due to the high incidence (47%) of PF joint degenerative changes between 5 and 9 years following ACLR(4), more research is needed to understand the factors contributing to the long-term articular cartilage changes within specific regions of the PF joint. Sub-sectioning the patella and femoral trochlea into medial and lateral regions of interest is important for understanding the nature of compositional changes following ACLR and how these changes may be related to long term damage at the PF joint. No studies to date have assessed the T1ρ relaxation times for the medial or lateral regions of the PF joint following ACLR. Therefore, the objective of this investigation was to determine if changes in T1ρ relaxation times occur in the medial and lateral regions of the patella and femoral trochlea on the ACLR limb (ACLR) and uninvolved contralateral limb from 6 to 12 months post-ACLR. We hypothesized that significant longitudinal increases in T1ρ relaxation times would be observed over time in specific regions (medial and lateral) of the PF joint on the ACLR limb but not on the uninvolved contralateral limb. A secondary purpose of this study was to determine if T1ρ relaxation times are greater in specific regions of the PF joint on the ACLR limb compared to the uninvolved contralateral limb at both 6 and 12 months post-ACLR. We hypothesized the T1ρ relaxation times for all regions of the PF joint on the ACLR limb would be significantly greater than the uninvolved contralateral limb at both 6 and 12 months post-ACLR.

PATIENTS & METHODS

Study Design

We conducted a longitudinal cohort study from a subset of individuals who underwent MRI analysis as part of a larger prospective longitudinal cohort study. The subset included all individuals with a primary ACLR who had completed the T1rho MRI collections in both limbs for the 6 and 12 month follow-up exams at the time of this study. No statistically significant differences were found between the sample used in the current study and the overall cohort for participant age (P=0.21), height (P=0.09), or weight (P=0.90). All participants were initially identified upon presentation in the orthopaedic clinic within 14 days of ACL injury and prior to ACLR. Participants who attended both the 6 (198.5±23.0 days) and 12-month (369.2±18.6 days) follow-up examination after their ACLR were included in the current study. While formal outpatient rehabilitation services were prescribed following ACLR by each surgeon, the rehabilitation protocol was not standardized across the cohort (24–26). During the rehabilitation process, participants were prescribed therapeutic exercise by their physician to be supervised by an Athletic Trainer or Physical Therapist. Standardized evidence-based rehabilitation guidelines were provided to the participants and clinicians to help progress the rehabilitation process (27). The Knee injury and Osteoarthritis Outcomes Score (KOOS) (28) was collected from each patient at the 6 and 12-month follow-up exams in order to describe the self-reported function of the cohort (Table 1). All participants provided informed consent that was approved by Institutional Review Board at the University of North Carolina at Chapel Hill (13–2385) prior to participating in any research related procedures.

Table 1.

Participant demographics^*

Characteristic	Value
No. participants men/women	11/9
Age	22±3.9yrs
Height	178.32±12.32cm
Mass	76.11±13.48kg
Concomitant Medial Meniscus Injury	n=6 (30%)
Concomitant Lateral Meniscus Injury	n=15 (75%)
Concomitant Chondral Injury	n=8 (40%)
KOOS-Symptoms @ 6 mos	80.15±11.91
KOOS-Pain @ 6 mos	86.50±8.26
KOOS-ADL @ 6 mos	96.85±3.66
KOOS-Sport @ 6 mos	69.50±14.95
KOOS-QOL @ 6 mos	54.75±18.36
KOOS-Symptoms @ 12 mos	84.85±9.46
KOOS-Pain @ 12 mos	91.70±7.41
KOOS-ADL @ 12 mos	97.35±4.18
KOOS-Sport @ 12 mos	84.00±14.47
KOOS-QOL @ 12 mos	74.50±17.96

^*Values are presented as mean±SD unless otherwise indicated. KOOS = Knee injury and Osteoarthritis Outcome Score; mos= months; ADL = activities of daily living; QOL=quality of life.

Participants

Individuals between 18 and 35 years old with a history of a unilateral primary ACL injury were included. We excluded those with a previous history of ACL injury on either limb, as well as those who sustained a second ACL injury at any point during the observation period. We did not exclude individuals with a concomitant meniscal or chondral injury. Those who were pregnant at the time of consent or planned to become pregnant during the 12-month observation period, had been previously diagnosed with any form of arthritis, needed a multi-ligament reconstruction, or were not planning to undergo ACLR were excluded.

All participants underwent a unilateral arthroscopically assisted single incision ACLR (31±16 days following ACL injury) using a patellar tendon autograft performed by one of three participating orthopaedic surgeons as previously reported (29). In brief, the middle third of the patellar tendon was harvested via an anterior longitudinal incision. Next, a target was determined on the lateral wall of the intercondylar notch of the femur and a femoral tunnel was drilled through the infra-medial arthroscopic portal with the knee in 120° of flexion. A pin was drilled and over-reamed into the ACL footprint from the infra-medial tibia to create a tibial tunnel. The proximal bone-plug of the patellar tendon graft was affixed to the femur with a metal interference screw. Finally, a metal interference screw was used to affix the distal bone-plug of the patellar tendon graft to the tibia. The attending orthopaedic surgeon recorded data regarding meniscal and articular cartilage injury at the time of surgery.

Previous work has demonstrated interlimb effects which range from moderate to strong for articular cartilage T1ρ relaxation times at 12 months following ACLR in different TF regions of interest (30). Based on these data we estimated (G*Power v3.1.9.2) 20 individuals would be needed to demonstrate a statistically significance difference between limbs and over time if a moderate effect was found (d=0.65; 1-β =0.8; α= 0.05).

Magnetic Resonance Image Acquisition

Magnetic resonance images from both limbs were acquired using either a Siemens Magnetom TIM Trio 3 Tesla scanner with a 4-channel Siemens large flex coil (516 mm x 224 mm, Siemens, Munich, Germany) or a Siemens Magnetom Prisma 3T PowerPack scanner with a XR 80/200 gradient coil (60 cm x 213 cm, Siemens, Munich, Germany). Strong inter-scanner reliability for absolute agreement has been previously determined for T1ρ relaxation times in the entire medial (ICC _2,1=0.99) and lateral (ICC _2,1=0.96) weight bearing regions of the femoral condyle in a separate cohort of 6 knees assessed in both scanners approximately 45 days apart.(25) Upon arrival to the imaging center, participants remained seated for 30 minutes to unload the knee cartilage (31). We used a T1ρ prepared three-dimensional Fast Low Angle Shot (FLASH) with a spin-lock power at 500Hz, five different spin-lock durations (40, 30, 20, 10, 0 ms) and a voxel size of 0.8mm x 0.4mm x 3mm (field of view= 288mm, slice thickness=3.0mm, TR= 9.2ms, 160 × 320 matrix, gap= 0mm, flip angle=10°, echo-train duration time= 443ms, phase encode direction of anterior/posterior) (32, 33).

T1ρ Relaxation Time Quantification, Registration and Segmentation

Prior to segmentation, an affine technique was performed to register the ACLR limb image to the uninjured limb image using the 0 ms spin lock image with 3-D Slicer software (http://www.slicer.org) (34). Following the affine registration, a non-rigid deformable, voxel by voxel intensity-based registration technique was applied to accurately align the ACLR femur and tibia to that of the uninvolved contralateral limb at each time point. The articular cartilage of the femoral trochlea and the patella acquired during the 0 ms spin-lock duration were manually segmented using ITK-SNAP software (version 3.6; http://www.itksnap.org) (35) for both the ACLR and uninvolved contralateral limbs. Following the initial segmentation, the articular cartilage of the femoral trochlea and patella were evenly divided into medial and lateral regions of interest (ROI). We separately determined the total number of image slices that included the patella and femur and divided each bone in half to derive the medial and lateral ROI for the patella and femoral trochlea. The medial and lateral ROI of the femoral trochlea and the patella were included in the data analysis. Voxel by voxel T1ρ relaxation times were calculated using a five-image sequence created with a MatLab program (MatLab R2014b [8.4.0] MathWorks, Natick, MA, USA) with the following equation: S(TSL) = S₀ exp(−TSL/T1ρ),(14) where TSL is the duration of the spin-lock time, S₀ is signal intensity when TSL equals zero, S corresponds to signal intensity, and T1ρ is the T1 relaxation time in the rotating frame, as previously reported (32, 33).

Statistical Analysis

Means and standard deviations were calculated for all continuous demographic variables and T1ρ relaxation times for all ROI, while frequencies were counted for all non-continuous demographic variables. Data distributions were assessed using the Shapiro–Wilk test for normality and stem and leaf plots were visually inspected for potential outliers. Separate limb (ACLR limb and uninvolved contralateral limb) by time (6 months and 12 months) ANOVAs were performed for each ROI (medial patella, lateral patella, medial trochlea, lateral trochlea). There were no covariates included in this analysis. An a priori level of significance for all analyses was set at p<0.05 and all analyses were performed using the Statistical Package for the Social Sciences software (SPSS, Version 21.0, IBM Corp., Somers, NY).

RESULTS

Demographics

Table 1 provides participant demographics. If an individual sustained a concomitant meniscal injury or chondral injury, this was addressed during the ACLR procedure (medial meniscal tear: 6 repairs; lateral meniscal tear: 4 repairs, 7 meniscectomy, 3 repairs and meniscectomy, 1 tear did not require surgical intervention; 8 chondral injuries: 1 chondroplasty, 1 microfracture, 6 did not require surgical intervention). Table 2 provides the average T1ρ relaxation times in all ROI on the ACLR and uninvolved contralateral limb at 6 and 12 months post-ACLR. All outcome measures were normally distributed.

Table 2.

T1ρ relaxation times (ms) at 6 months and 12 months post-ACLR for each region of interest.^*

	6 months	12 months
	Mean±SD	Mean±SD	MD (95% CI)
Medial Patella †
ACLR	53.67±3.74	54.62±4.37	−0.95(−2.90, 1.01)
Uninvolved	52.94±2.77	55.51±2.08	−2.57(−4.01, -1.12)
MD (95% CI)	0.73(−0.92, 2.38)	−0.89(−2.6, 0.89)
Lateral Patella ‡
ACLR	55.24±4.53	54.52±5.62	0.71(−1.82, 3.25)
Uninvolved	56.96±3.26	57.73±3.59	−0.77(−2.51, 0.97)
MD (95% CI)	−1.73(−3.42, -0.03)	−3.21(−4.97, -1.45)
Medial Trochlea
ACLR	57.69±4.50	61.10±4.22	−3.41(−5.48, −1.34)†
Uninvolved	55.07±3.01	56.20±3.88	−1.13(−2.77, 0.50)
MD (95% CI)	2.62(0.87, 4.37)‡	4.90(3.52, 6.28)‡
Lateral Trochlea †, ‡
ACLR	56.10±3.23	57.80±4.69	−1.70(−3.62, 0.22)
Uninvolved	53.65±3.22	55.30±3.03	−1.66(−3.28, -0.03)
MD (95% CI)	2.45(1.04, 3.86)	2.50(0.54, 4.46)

^*ACLR= anterior cruciate ligament reconstruction; SD=standard deviation; MD=Mean Difference; CI=confidence interval.

^†Significant increase from 6 to 12 months for main effect or interaction effect (P<0.05)

^‡Significant difference between limbs for main effect or interaction effect (P<0.05)

Patella Articular Cartilage

For the medial patella, T1ρ relaxation times increased in both limbs from 6 to 12 months post-ACLR [F(1,19)=7.79, P=0.012; η_ρ²=0.29; mean difference=1.76ms, 95% CI: −0.04, 3.55]. For the lateral patella, T1ρ relaxation times were significantly greater on the uninvolved contralateral limb compared to the ACLR limb [F(1,19)=14.156, P=0.001; η_ρ²=0.43; mean difference=2.47ms, 95% CI: 0.13, 4.82].

Trochlear Articular Cartilage

For the lateral trochlea, T1ρ relaxation times increased in both limbs from 6 to 12 months post-ACLR [F(1,19)=4.698, P=0.043; η_ρ²=0.20; mean difference=1.68ms, 95% CI: −0.32, 3.67] and the T1ρ relaxation times were significantly greater on the ACLR limb compared to the uninvolved contralateral limb [F(1,19)=11.311, P=0.003; η_ρ²=0.37; mean difference=2.47ms, 95% CI: 0.51, 4.43]. A significant limb by time interaction was found for T1ρ relaxation times in the medial trochlea [F(1,19)=6.136, P=0.023; η_ρ²=0.24]. T1ρ relaxation times in the medial trochlea on the ACLR limb were significantly greater at 6 (P=0.005; mean difference=2.62ms, 95% CI: 0.87, 4.37) and 12 months (P<0.001; mean difference=4.90ms, 95% CI: 3.52, 6.28) post-ACLR compared to the uninvolved contralateral limb. T1ρ relaxation times in the medial trochlea on the ACLR limb significantly increased from 6 to 12 months on the ACLR limb (P=0.003; mean difference=−3.41ms, 95% CI: −5.48, −1.34). There were no significant changes in T1ρ relaxation times of the medial trochlea from 6 to 12 months on the uninvolved contralateral limb (P=0.163; mean difference=−1.13ms, 95% CI: −2.77, 0.50).

DISCUSSION

In agreement of our hypotheses, T1ρ relaxation times in the medial femoral trochlea were greater in the ACLR limb compared to the uninvolved contralateral limb at both time points and increased in the ACLR limb between 6 and 12 months post-ACLR. In partial agreement of our hypotheses, we found that T1ρ relaxation times in the medial patella and lateral trochlea increased bilaterally between 6 and 12 months following unilateral ACLR; yet T1ρ relaxation times were always higher in the lateral trochlea on the ACLR limb compared to the uninvolved contralateral limb. Contrary to our hypotheses we found that T1ρ relaxation times were greater in the lateral patellar cartilage on the uninvolved contralateral limb compared to the ACLR limb. Overall, these findings provide novel evidence that specific regions of the PF joint may be more susceptible to deleterious tissue changes associated with posttraumatic OA and that these compositional changes in specific portions of the PF joint occur bilaterally following unilateral ACLR.

Previous studies have reported longitudinal T1ρ relaxation time changes in the femoral trochlea of the ACLR limb and patella on the uninvolved contralateral limb (18, 19). Both Amano et al. (19) and Pedoia et al. (18) reported a significant increase in T1ρ relaxation times of the articular cartilage in the femoral trochlea between ACL injury (prior to ACLR) and 6 months post-ACLR. In addition, Pedoia et al. (18) reported increased T1ρ relaxation times at the patella on the uninvolved contralateral limb from ACL injury (prior to ACLR) to 6 months. The findings from our study build on previous work, by defining more specific regions of the PF joint that demonstrate changes in T1ρ relaxation times.

The underlying mechanisms leading to these compositional changes in PF joint articular cartilage composition remain unclear; yet, both biochemical and biomechanical changes following ACLR may be important factors related to these degenerative joint tissue changes. Increased proinflammatory cytokines and degenerative enzymes associated with cartilage breakdown have been reported within the first 12 months following ACL injury and ACLR (20). In addition to biochemical changes, both overloading and underloading of the lower extremity during dynamic tasks have been associated with deleterious cartilage compositional changes following ACLR.(25, 36, 37)

Aberrant gait biomechanics, which impact knee joint loading, are common following ACLR (38). Individuals with an ACLR demonstrate altered joint loading [decreased vertical ground reaction force (vGRF) in early stance and increased vGRF in midstance] of the ACLR limb and uninvolved contralateral limb during the stance phase of gait in the first 12 months following ACLR compared to uninjured controls (39). During more dynamic tasks, such as squatting and jumping, individuals with an ACLR display decreased vGRF on the ACLR limb compared to the uninvolved contralateral limb one to two years post-ACLR.(40–43) The changes in vGRF likely lead to alterations in loads placed across the PF joint. In a recent study investigating PF joint contact forces during running 12–24 months post-ACLR, peak PF joint contact forces were decreased on the ACLR limb as compared to the uninvolved contralateral limb (22). Decreased loading on the ACLR limb has been associated with altered TF cartilage composition (25) but it is not clear if decreased loading leads to changes in PJ joint cartilage composition. Future research needs to investigate how these changes in joint loading contribute to articular cartilage changes at the PF joint.

There is also evidence to suggest specific kinematic changes occur at the PF joint following ACLR. Lin et al. (44) demonstrated a significant increase in patellar external rotation, lateral patellar tilt, and lateral translation following ACLR. It would be expected that these changes in PF kinematics would lead to increased loads placed across the lateral patella and femoral trochlea and decreased loads across the medial patella and femoral trochlea. Our study demonstrated that the T1ρ relaxation times in the articular cartilage of medial femoral trochlea significantly increased between 6 and 12 months post-ACLR in the injured limb, which could be a sign of altered loading across this joint. Further investigation into changes in PF kinematics on the ACLR limb following ACLR and their effect on T1ρ relaxation times is needed to better understand how these factors influence articular cartilage compositional changes at the PF joint.

While there appears to be a general tendency to underload the ACLR limb following ACLR, it is important to be aware of the increased loads placed on the uninvolved contralateral limb during dynamic tasks. The alterations in T1ρ relaxation times of the medial and lateral compartments of the patella and lateral trochlea on the uninvolved contralateral limb in this study could be attributed to increased loading on the contralateral limb during dynamic activities. Evidence supports increased vGRF during squatting and jumping on the uninvolved contralateral limb compared to the ACLR limb (40–43). The higher vGRF on the uninvolved contralateral limb could lead to increased loads placed on the PF joint and compositional alterations from overloading the articular cartilage (45).

The findings of compositional cartilage changes at the PF joint on both the ACLR and uninvolved contralateral limb highlight the need for clinicians to focus on both limbs following ACLR to ensure symmetrical and sufficient loading is restored upon return to sport or activity. Gaining an understanding of optimal loading of articular cartilage and the ability to counteract these deleterious articular cartilage changes through interventions should help to inform the development of effective interventions post-ACLR. Increases in T1ρ relaxation times of the knee articular cartilage following a period of non-weightbearing have been shown to be transient and return to baseline levels upon return to normal loading of the joint (46). Although it remains unknown how adjusting loading in those with knee injury may impact proteoglycan density over time. More research is needed to understand if there are intervention strategies that can be implemented to improve proteoglycan concentrations in articular cartilage at the PF joint post-ACLR.

When interpreting the results of this study, it is important to consider that all participants in this investigation underwent a bone patellar tendon bone autograft. Increased degenerative changes have been reported at the PF joint as compared to the TF joint 7 years post-ACLR in individuals undergoing a bone patellar tendon bone autograft (47). It is not known if these same changes in T1ρ relaxation times within specific regions of the PF joint would also occur following other ACL graft procedures. Furthermore, it is not clear that changes in T1ρ relaxation times in the PF joint will predispose individuals to the development of posttraumatic OA. The articular cartilage in the femoral trochlea has been reported to thin 24 months post-ACLR (48). Yet, additional research is needed to determine if an increase in T1ρ relaxation times between 6 and 12 months post-ACLR will result in the eventual thinning of articular cartilage in the femoral trochlea or patella.

While this is the first study to evaluate T1ρ relaxation times within specific regions of the PF joint post-ACLR, there are some limitations which can inform future investigations. We did not collect baseline T1ρ relaxation times in this cohort; therefore, we cannot determine preinjury T1ρ relaxation times or how cartilage composition may have changed within the first 6 months following ACLR on the ACLR limb and uninvolved contralateral limb. Furthermore, proteoglycan density changes in the articular cartilage may be reversable and we do not know if the changes in T1ρ relaxation times are transient or if these changes will continue. Longer follow-up times are needed to determine if the changes in T1ρ relaxation times at the PF joint following ACLR lead to chronic symptoms and radiographic PF osteoarthritis on the ACLR or uninvolved contralateral limb. We also did not assess the influence of patient function, rehabilitation progression or their physical activity levels at the time of the 6 and 12-month MRI acquisitions. An assessment of the progression of the participant through their rehabilitation program and their physical activity levels at each time point could help to inform the interpretation of the biological changes occurring within the PF joint post-ACLR. We recommend the inclusion of an assessment of physical activity levels and rehabilitation progress in future investigations on articular cartilage changes at the PF joint following ACLR. While all participants were prescribed therapeutic exercise and provided standardized evidence-based rehabilitation guidelines, we did not standardize the rehabilitation protocol in this study. Finally, this is a relatively small subset of individuals from a larger investigation. Due to the sample size, we were unable to determine how other covariates may have impacted the change in the T1ρ relaxation times (i.e. sex, age, concomitant meniscal/chondral injury).

In conclusion, compositional changes in articular cartilage occur within the first 12 months following ACLR in specific regions of the PF joint on the ACLR and uninvolved contralateral limb. These compositional changes may increase the risk for the development of posttraumatic OA early in the PF joint. Continued research is needed to understand if these compositional changes at the PF joint are associated with biochemical and/or biomechanical changes that occur following ACLR and if these changes lead to long term damage at the PF joint. Additional research on compositional changes in the articular cartilage of the PF joint following ACLR could help to inform the development of effective treatment strategies aimed at preventing the development of posttraumatic OA at the PF joint.

SIGNIFICANCE AND INNOVATION.

• Following unilateral ACLR, the involved and uninvolved limbs display compositional changes of the articular cartilage of the PF joint.

• The findings from this study provide novel evidence that deleterious tissue changes associated with posttraumatic OA are occurring within specific regions of the PF joint post-ACLR.