Osteoarthritis and Cartilage 22 (2014) 1453e1460

Changes to the articular cartilage thickness profile of the tibiafollowing anterior cruciate ligament injury

E.C. Argentieri y, D.R. Sturnick y, M.J. DeSarno z, M.G. Gardner-Morse y, J.R. Slauterbeck y,R.J. Johnson y, B.D. Beynnon y *

y Department of Orthopaedics and Rehabilitation, University of Vermont, Burlington, VT, USAz Department of Medical Biostatistics, University of Vermont, Burlington, VT, USA

a r t i c l e i n f o

Article history:Received 4 February 2014Accepted 22 June 2014

Keywords:Cartilage thicknessCartilage morphologyJoint traumaACL injuryPTOATibia

* Address correspondence and reprint requests to:fessor, McClure Musculoskeletal Research Center, DepRehabilitation, University of Vermont, College of Medi1-(802)-656-2250; Fax: 1-(802)-656-4247.

E-mail addresses: Erin.Argentieri@uvm.edu (E.C.uvm.edu (D.R. Sturnick), Michael.DeSarno@uvm.eGardner-Morse@uvm.edu (M.G. Gardner-Morse),(J.R. Slauterbeck), Robert.Johnson@uvm.edu (R.J. Johedu (B.D. Beynnon).

http://dx.doi.org/10.1016/j.joca.2014.06.0251063-4584/Published by Elsevier Ltd on behalf of Ost

s u m m a r y

Objectives: We sought to determine if anterior cruciate ligament (ACL)-injured subjects demonstratedside-to-side differences in tibial cartilage thickness soon after injury, and if uninjured-control subjectsdisplayed side-to-side symmetry in cartilage thickness. Second, we aimed to investigate associationsbetween body mass index (BMI), cross-sectional area (CSA) of the proximal tibia, and articular cartilagethickness differences.Methods: Bilateral Magnetic Resonance Images (MRIs) were obtained on 88 ACL-injured subjects (27male; 61 female) a mean 27 days post-injury, and 88 matched uninjured control subjects. Within ACL-injured and uninjured control subjects, side-to-side differences in medial and lateral tibial articularcartilage thickness were analyzed with adjustment for tibial position relative to the femur during MRIacquisition. Associations between tibial CSA and cartilage thickness differences were tested within highand low BMI groups.Results: Within the medial tibial compartment, ACL-injured females displayed significant increases:mean (confidence interval (CI)) ¼ þ0.18 mm (0.17, 0.19) and decreases: mean (CI) ¼ �0.14 mm(�0.13, �0.15) in tibial cartilage thickness within the central and posterior cartilage regions respectively.Adjustment for tibial position revealed a decreased area of significant cartilage thickness differences,though 46% of points maintained significance.In the lateral compartment anterior region, there was a significantly different relationship betweencartilage thickness differences and CSA, within high and low BMI groups (BMI group*CSA interaction,P ¼ 0.007). Within the low BMI group, a significant negative correlation between cartilage thickness andCSA was identified (P ¼ 0.03).Conclusions: ACL-injured females displayed cartilage thickness differences in the central, and posteriormedial tibial cartilage regions. Tibial position effected thickness differences, but did not account for allsignificant differences.

Published by Elsevier Ltd on behalf of Osteoarthritis Research Society International.

B.D. Beynnon, McClure Pro-artment of Orthopedics andcine, Burlington, VT, USA. Tel:

Argentieri), Daniel.Sturnick@du (M.J. DeSarno), Mack.James.Slauterbeck@uvm.edunson), Bruce.Beynnon@uvm.

eoarthritis Research Society Intern

Introduction

Severe joint traumas, including injury to the anterior cruciateligament (ACL), have lasting effects on joint heath, and often resultin the early onset of post traumatic osteoarthritis (PTOA)1e10.Though surgical reconstruction of the ACL has become the con-ventional way to improve patient function, it does not prevent theultimate development of PTOA1,2,5,11. Reports estimate the highestincidence of ACL-injury occurs in individuals between 15 and 26years of age5,12. As a consequence, many young and otherwisehealthy individuals inevitably present with PTOA within 10e15years of the index injury2,3,13,14. Presently, little is known about theonset of cartilage degeneration after injury, though several studies

ational.

Table IICC values for measurement reliability between two separate time points presentedfor medial and lateral compartments, anterior, central and posterior regions

Compartment Regions ICC

Lateral Anterior 0.88Central 0.97Posterior 0.96

Medial Anterior 0.91Central 0.95Posterior 0.85

E.C. Argentieri et al. / Osteoarthritis and Cartilage 22 (2014) 1453e14601454

suggest that the initial joint trauma2,5e10,15, and corresponding al-terations in joint contact mechanics2,3,16e18 are paramount factorsin the development of PTOA.

Magnetic Resonance Imaging (MRI) studies of ACL-injured pa-tients often reveal concomitant injuries within the tibiofemoraljoint19,20, that are indicative of the large shear and compressiveloads developed across the joint at the time of injury2,3,5,19. Thoughthere might be an absence of gross defects within the articularcartilage surface, bone marrow lesions below the articular cartilageare common19, and serve as evidence of the substantial forcestransmitted through the cartilage at the time of ACL-injury2,19.Transmitted forces of this magnitude are capable of disrupting theorganization of the collagen matrix, and there is evidence to sug-gest that this initial impact could initiate a cascade of biologicalevents culminating in chondrocyte necrosis, alterations in cellmetabolism, loss of proteoglycan, and the ultimate development ofPTOA2,4e10. Previous research focused on cartilage status after ACL-injury has revealed important links between the presence ofconcomitant injuries and patient outcomes2,5,9,19,21,22. ACL-injuredpatients who sustain concomitant injuries develop PTOA athigher rates than patients who sustain isolated-ACL injuries5,22,23.

In addition to the consequences of the initial joint trauma, ACL-injury results in alterations to the biomechanics of the tibiofemoraljoint including: altered location and orientation of the tibia relativeto the femur17,24, as well as changes in the location and distributionof contact stresses17,25e27. These alterations are important toconsider, and add to an understanding of the relationship betweenACL-injury and subsequent PTOA development7,16. Though quan-titative MRI measurements of cartilage thickness have been shownto accurately depict changes within cartilage morphology5,28e34,current research has yet to include consideration for the effect ofaltered joint position on the measurement of cartilage thickness. Toour knowledge no previous studies have measured significant side-to-side differences in tibial cartilage thickness soon after ACL-injury, though previous research has described both increasedand decreased cartilage thickness on the medial femoral condyleand trochlear sulcus, respectively, at one- and two-year follow-upsafter ACL-injury1,35 Additionally, prior work has used radiographicbased techniques to measured acute changes in joint space width,and has indicated both thinning and thickening of tibiofemoralarticular cartilage within months after ACL-injury36.

Presently, a paradigm exists where interventions for both pri-maryosteoarthritis (OA) and PTOAhavebeen targeted at individualswho already exhibit clinical signs and symptoms of an irreparabledisease process. Quantification of early changes within the articularcartilage soon after ACL-injurymay help to describe early structuraldamage, and give insight to onset of PTOA; at a point in time wherestructural deterioration may be reversible4,37,38.

Building on prior research the objectives of this study weretwofold. First, we sought to determine if ACL-injured subjectsexhibited significant side-to-side differences in tibial plateauarticular cartilage thicknesses within a short time interval afterinjury and, if uninjured control subjects displayed side-to-sidesymmetry in tibial cartilage thickness. Additionally, we aimed todetermine if cartilage thickness differences were affected by theposition of the tibia relative to the femur during MRI acquisition.Our second objective was designed to build on our first, by definingeffects of overall subject size (body mass index (BMI)) and the sizeof a subject's proximal tibia (defined by cross-sectional area (CSA)),on the measurement of cartilage thickness.

Methods

Our Institutional Review Board gave approval for this study andall subjects provided written consent prior to their participation.

MRI data in this study were originally collected as part of a largerlongitudinal cohort study with a nested matched case-controldesign39. The aforementioned study was designed to identify therisk factors for noncontact ACL-injury, and resulted in a multivar-iate model of risk39. Additional details of the MRI cohort, includingthe recruitment protocol and subject demographics, have beenpreviously described40.

Varsity athletes from 28 high schools and 8 colleges weremonitored prospectively over a 4-year time interval for the occur-rence of non-contact ACL-injures. Non-contact ACL-injury wasdefined as: an ACL-injury that had occurred in the absence of adirect blow to the knee. 88 ACL-injured subjects (27 male, 61 fe-male) who had sustained a first time, grade III, non-contact ACL-injury, during participation in organized high school or collegesport were included in this study. Uninjured control subjects (27males, 61 females), were recruited teammates of the ACL-injuredsubjects, who were matched on age, and sex. Case-control match-ing of subjects from the same sports teamwas performed to controlfor the type, amount, and level of activity of each case subject.

Using the Phillips Achivia 3.0T Research MRI (Fletcher AllenHealthcare, Burlington, VT) bilateral MRI scans were obtained onboth ACL-injured, and uninjured-control subjects. Subjects werepositioned supine,with their knees in extension inside an 8-channelSENSE coil. Sagittal plane, T1 weighted, Fast-Field Echo (FFE) scanswith a slice thickness of 1.2 mm, and a within-plane resolution of0.3 mm by 0.3 mm, were obtained on all subjects. MRIs were ac-quired on ACL-injured subjects after injury, but preceding any sur-gical reconstruction (days post-injury: median 15, average 27, range1e110). Using OsiriX Software (Pixmeo, Geneva, Switzerland,version 3.6.1, open source) and a Cintiq 21 UK Digitizing tablet(Wacom Tech Corp, Vancouver, WA, USA), digital imaging andcommunications in medicine (DICOM) images were viewed andmanually segmented. Medial and lateral compartments of the tibialplateau were segmented and considered separately. Articularcartilage surfaces were segmented in both the sagittal and coronalplanes, from the edge of each compartment to the last slice on thetibial spinewhere the cartilage surface was discernable. Medial andlateral tibial plateau subchondral bone surfaces were segmented inthe sagittal plane (See Fig. 1 of Supplemental text). Intraclass cor-relation coefficient (ICC) values were obtained from an analysis ofreliability between two separate time points, indicated a high levelof cartilage thicknessmeasurement reliability (Table I). BecauseMRIdatawere acquired in the coordinate system of the scanner, and theposition of the tibia relative to the femur varied both within andbetween subjects, MRI data were transformed into three-dimensional bone based femoral and tibial coordinate systems[Fig. 1(A)]. Tibial and femoral coordinate systems were used toobtainmeasurements thatwere referenced to the anatomical planesof each bone, allowing for comparison of data both between, andwithin subjects.Methods used to establish and locate the coordinatesystems have been described41. Respective locations of the tibial andfemoral coordinate system origins allowed for a quantitativedescription of the position of the tibia relative to the femur duringMRI acquisition. With this approach medial-lateral (ML) and

A B

Fig. 1. A: Tibial coordinate system in which digitized articular cartilage (light gray) and subchondral bone (dark gray) surface data were repositioned, including AP length and MLwidth of the proximal tibia used to scale data to a mean size. B: Cartilage thicknesses were measured where a normal vector to the underlying subchondral bone intersected withthe overlying articular cartilage at each point on a standard 1 mm by 1 mm grid.

E.C. Argentieri et al. / Osteoarthritis and Cartilage 22 (2014) 1453e1460 1455

anterior-posterior (AP) positions of the tibia relative to the femurwere characterized bilaterally within each subject.

Using data obtained from the tibial coordinate system, cartilageareas within each subject were scaled based on the AP, and MLdimensions of the proximal tibia. Using a 1 mm by 1mm grid, threetibial plateau cartilage regions: anterior, central, and posterior,were defined within the medial and lateral compartments[Fig. 1(B)]. Cartilage thickness measurements were made at eachpoint on the 1 mm by 1 mm grid, for each region.

Cartilage thickness was measured at each point on the afore-mentioned grid, where a normal vector from the cartilage surfaceintersected with the underlying subchondral bone [Fig. 1(B)]. Eachsubject's right knee was mirrored to match their left, so that side-to-side differences in cartilage thickness could be measured.Within ACL-injured subjects, differences in cartilage thickness weremeasured between the injured and uninjured knees. Differenceswithin uninjured-control subjects were measured using the sameinjured and uninjured-knee convention as their matched ACL-injured subjects. In order to account for the size of the tibialplateau, cross sectional area (CSA) of the proximal tibia wasmeasured at a standardized location inferior to the articular carti-lage surface41.

Statistical analyses

Previous studies have reported differences in tibiofemoral jointmorphology42,43 and risk factors for ACL-injury40,41 between malesand females. These findings suggest differences between the sexesin the biomechanics and mechanism of ACL-injury, as well as po-tential differences in the subsequent affects of injury. As such, itwas important to analyze male and female data separately.

Because ACL-injury alters the position of the tibia relative to thefemur, there was concern that the position of the articular surfacesof the joint would have an affect on the cartilage thickness distri-bution. Therefore within each subject, the side-to-side differencesin AP and ML tibial position (relative to the femur) were used ascovariates in analyses. Evaluation of the first objective includedmeasurement of tibial plateau articular cartilage thickness withinACL-injured and uninjured-control subjects. Side-to-side differ-ences in cartilage thickness were compared using a mixed modelrepeatedmeasures of variance at each point on the aforementioned1 mm by 1 mm grid. Comparisons were made in a region (anterior,central, posterior) and compartment (medial, lateral) specificmanner, and the determined P-values were adjusted for a FalseDiscovery Rate, using the Benjamini-Hochberg method. Our second

objective aimed to evaluate if a relationship existed betweenmeasured cartilage thickness profile differences and subject size.Categorical BMI groups were defined based on the respectiveranges of male and female subject BMI data. Median BMI valueswere used as the cut off for high and low BMI groups. Femalesubjects were considered to have low BMI if it was less than 22.5, orhigh BMI if greater than 22.5. Male subjects were considered tohave a low BMI if it was less than 23.2 or high BMI if it was greaterthan 23.2. Analyses of variance (ANOVA) were used to test the as-sociations between measurements of CSA, and magnitude ofcartilage thickness change for each BMI group, and to test for dif-ferences in these associations between BMI groups. All statisticalanalyses were conducted using SAS, version 9.2 (SAS Institute, Cary,NC).

Results

Summary statistics for measured articular cartilage thicknesswithin each region of both the medial and lateral tibial compart-ments, including mean and 95% confidence intervals (CIs), areprovided in Table II.

Analyses of within subject comparisons of cartilage thickness

Within ACL-injured females (n ¼ 61), significant side-to-sidedifferences in cartilage thickness were found in the central andposterior cartilage regions of the medial tibial compartment(Table III and Fig. 2). Specifically, significant cartilage thickness in-creases: mean (CI) ¼ þ0.18 mm (0.17, 0.19), were found within15.9% of points in the central cartilage region, while significantcartilage thickness decreases: mean (CI) ¼ �0.14 mm(�0.13, �0.15), were found within 28.5% of points in the posteriorcartilage region of the medial compartment. Within ACL-injuredfemales, no significant differences in lateral tibial cartilage thick-ness were seen in the anterior, central, or posterior cartilageregions.

Further analyses of ACL-injured females that were adjusted forthe AP and ML position of the tibia relative to the femur revealed adecrease in the area where cartilage where significant differenceswere found, though 46% of points in the central and posterior re-gions of the medial compartment maintained significance.

Unadjusted analyses of ACL-injured females revealed no sig-nificant differences in cartilage thickness for any of the lateralcompartment cartilage regions, though differences within the

Table IIMean (95% CI) cartilage thicknesses (mm) by sex, compartment and region for cases and controls in injured (or matched injured knees in controls) and contralateral uninjured(or matched uninjured knees in controls) knees

Region Subjects Knee Mean (95% CI) cartilage thicknesses (mm)

Females Males

Lateral Medial Lateral Medial

Anterior Cases Injured 1.30 (1.21e1.39) 1.62 (1.55e1.69) 1.42 (1.26e1.57) 1.79 (1.72e1.87)Uninjured 1.21 (1.13e1.29) 1.57 (1.50e1.64) 1.38 (1.24e1.51) 1.71 (1.62e1.81)

Controls Injured* 1.25 (1.16e1.34) 1.54 (1.48e1.61) 1.46 (1.32e1.61) 1.87 (1.75e1.99)Uninjured* 1.25 (1.17e1.32 1.56 (1.50e1.63) 1.41 (1.26e1.55) 1.81 (1.71e1.91)

Center Cases Injured 2.75 (2.63e2.86) 2.08 (1.99e2.16) 2.99 (2.87e3.11) 2.30 (2.20e2.41)Uninjured 2.69 (2.58e2.80) 2.08 (2.00e2.16) 2.97 (2.82e3.11) 2.24 (2.14e2.35)

Controls Injured* 2.61 (2.51e2.71) 1.99 (1.92e2.05) 3.08 (2.92e3.24) 2.32 (2.21e2.43)Uninjured* 2.58 (2.48e2.68) 2.02 (1.95e2.10) 3.02 (2.83e3.20) 2.32 (2.21e2.43)

Posterior Cases Injured 2.81 (2.69e2.93) 1.68 (1.61e1.75) 2.77 (2.64e2.89) 1.62 (1.52e1.73)Uninjured 2.84 (2.72e2.96) 1.76 (1.69e1.84) 2.81 (2.68e2.94) 1.70 (1.59e1.81)

Controls Injured* 2.71 (2.62e2.81) 1.72 (1.65e1.79) 2.86 (2.73e2.99) 1.88 (1.75e2.00)Uninjured* 2.68 (2.58e2.77) 1.70 (1.62e1.78) 2.86 (2.71e3.01) 1.84 (1.70e1.98)

* Uninjured control subjects with same injured and uninjured side convention as matched case subject.

Table IIISignificant side-to-side (injured-to-uninjured knee) differences in medial tibialarticular cartilage thickness within ACL-injured females (n ¼ 61)

Cartilage region Mean (mm) 95% CI (mm) Range (mm)

Anterior e e e

Central þ0.18 (þ0.17, þ0.19) 0.10 to 0.29Posterior �0.14 (�0.13, �0.15) �0.27 to 0.07

E.C. Argentieri et al. / Osteoarthritis and Cartilage 22 (2014) 1453e14601456

central region of the lateral compartment displayed a trend towardsignificance.

Analyses of male ACL-injured subjects revealed no significantside-to-side differences in cartilage thickness for any of the threecartilage regions (anterior, central, posterior), within the medial

Fig. 2. Significant injured-to-uninjured side differences in articular cartilage thicknessfound in the medial compartment within ACL-injured females. Articular cartilage wasthicker in the central region and thinner in the posterior region of ACL-injured knees.

and lateral compartments after adjustment for tibial positionrelative to the femur (Supplemental text Fig. 2). Uninjuredmale andfemale control subjects exhibited no significant side-to-side dif-ferences in tibial cartilage thickness for any of the anterior, central,or posterior cartilage regions within the medial or lateralcompartments.

ANOVA results for cartilage thickness difference vs categorical BMI,and mean CSA

Within the categorized high and low BMI groups, relationshipswithin ACL-injured subjects between mean injured-uninjured sidedifferences in cartilage thickness, and injured-uninjured side meanCSA were tested. A significant relationship was only found withinfemale ACL-injured subjects, for the anterior cartilage region of thelateral compartment. Within the low BMI group, there was a sig-nificant negative correlation between cartilage thickness mean andCSA mean (P-value ¼ 0.03; r ¼ �0.39). There was also a nearlysignificant positively correlated relationship between mean carti-lage thickness mean and CSA mean for the High BMI group (P-value ¼ 0.08; r ¼ 0.32). In addition, the BMI group*CSA meaninteraction was significant (P-value ¼ 0.007), indicating a signifi-cant difference in the thickness vs CSA relationships between BMIgroups.

Discussion

ACL-injured females displayed significant side-to-side differ-ences in articular cartilage thickness within the medial compart-ment. Specifically, significant increases in tibial cartilage thicknesswere found within the central region of the medial tibial cartilage,while significant decreases in tibial cartilage thickness were foundwithin the posterior cartilage region. Notably, within the currentstudy ACL-injured males did not display significant side-to-sidedifferences in cartilage thickness. The cartilage thickness data formale ACL-injured subjects was more variable than ACL-injured fe-males. Greater variability within ACL-injured male data, inconjunction with a comparably smaller male sample size may havedecreased the level of statistical power and obscured significantfindings.

Measured cartilage thickness differences within ACL-injuredfemales are biomechanically significant as they represent amagnitude of change that is more that 10% of the mean cartilagethickness within the medial compartment, and were not fully

E.C. Argentieri et al. / Osteoarthritis and Cartilage 22 (2014) 1453e1460 1457

explained by position of the tibia relative to the femur. Blunt impactforces to the articular cartilage, such as those produced at the timeof ACL-injury, have been shown to alter the organization of thecollagen matrix even in the absence of gross cartilage defects2,6e10.Disruption of the cartilage matrix is a potential cause for themeasured increase in cartilage thickness; increased thickness maybe reflective of a shift in water content, created by an acute loss ofproteoglycan from the cartilage. This notion is further supported byresearch that has employed the use of specialized MRI techniquesto observe changes in both cell metabolism, and glycosaminoglycanindices of the cartilage within a short time interval after ACL-injury4,6,7. Previous studies have revealed significant side-to-sidedifferences in the delayed Gadolinium-Enhanced MR Imaging ofCartilage (dGEMRIC) indices within the medial tibiofemoralcompartment of ACL-injured subjects44,45. Additionally, at one-yearfollow up after ACL-injury Li et al. described the prolongation of T1rho, indicative of a loss of proteoglycan within a similar region ofthe tibial cartilage in the medial compartment4.

Biomechanically, ACL-injury causes changes within the loadingenvironment of the tibiofemoral joint2,3,16e18,25,27. In the currentstudy alterations in the position of the tibia relative to the femur atthe time of MRI acquisition were found to be an important meth-odological factor to consider when making side-to-side compari-sons of cartilage thickness. Statistical analysis with adjustment forthe position of the tibia relative to the femur duringMRI acquisitionhad a significant influence on the magnitude of side-to-side carti-lage thickness differences. If the 3D position of the tibia relative tothe femur were not considered in the statistical analysis, themagnitude of measured side-to-side differences would have beenamplified. ACL-injury causes a shift in tibiofemoral contact me-chanics to a more lateral and posterior location on the tibialplateau17,27. Within the current study, the decreases in tibial carti-lage thickness seen within the posterior cartilage region may be aresult of increased load being transmitted through the posteriormedial meniscus. Decreased cartilage thickness in this region couldbe further propagated by regional variations in the material andstructural properties of the tibial cartilage; therefore loadingcartilage unsuited to bear such loads3,16,27,46,47. Conversely, thecartilage within the central cartilage region that is conditioned forfunctional weight bearing, but might be subjected to a decrease inload do to the more anterior position of the tibia relative to thefemur.

Abnormal biomechanics and subsequent changes in the tibio-femoral joint loading environment have been proposed as acontributing factors for the progression of cartilage degeneration(PTOA)3,16,27,46e48. Altered orientations include a two-degree in-ternal rotational orientation, a three-millimeter anterior shift, and aone-millimeter medial translation of the tibia relative to the femurwithin ACL-injured subjects performing a static lunge24. Theloading environment of tibial plateau articular cartilage is influ-enced by altered kinematics after an ACL-injury25,27,49. In an ACL-injured knee, locations of highest contact stress transmissionthroughout gate can shift to thinner cartilage regions that are notconditioned to withstand such stress3,16,25e27,47.

Advancements of the study include its rigorous study designand highly standardized measurement methodology. Only subjectswho had suffered their first ACL-injury (to either knee) wererecruited for participation in this study. Uninjured controls werematched to the ACL-injured subjects on age, sex and participationon the same sports team. As such, therewas control for the amount,level, and type of activity that subjects participated in prior toinjury. A consistent definition of non-contact injury mechanismswas applied to all subjects, which avoided confounding results thatmay have been introduced by differences in joint biomechanicsbetween contact and non-contact injury mechanisms. For each

subject, bilateral MRI data were acquired in a consistent manner,with knees placed in extension, and patella aligned vertically, usingthe same 8-channel SENSE coil and MRI unit. Data analysesincluded the use of tibial and femoral coordinate systems thatallowed us to establish the complete 6 degree-of-freedom positionof the tibia relative to the femur during MRI acquisition and makereliable measures of cartilage thickness both within and betweensubjects.

Some limitations exist with the application of this study design.MRI data were acquired at a single time point after ACL-injury butprior to surgical reconstruction; a time interval which varied be-tween ACL-injured subjects. However, themajority of subjects wererecruited and evaluated within a short time interval (median 15days) after ACL-injury. Consequently, this study should be consid-ered to have focused on the response of articular cartilage to thetrauma produced at the time of ACL disruption and its initialhealing response. We were interested in determining if the timeinterval between ACL disruption and MRI data acquisition had aneffect on cartilage thickness and consequently we performed post-hoc analysis. This revealed the time interval between injury andMRI acquisition was not significantly correlated with differences inarticular cartilage thickness. Data collected from this unique patientdemographic allowed for the investigation of the early traumaticeffects of ACL-injury on cartilage thickness. However, withoutlongitudinal data it is difficult to determine how the early changesin cartilage thickness are associated with the future onset andprogression of PTOA. Future work is needed to characterize theprogressive response of articular cartilage and bone. With suchinformation, the time interval during which an intervention couldbe effective may be identified.

Additionally, leg axis and muscle forces could have an influenceon articular cartilage thickness through alteration of contactstresses developed across the joint. Such measurements requiredata acquired with standing radiographs, which were not collectedin this study. Recent work by Marsh et al. reported correlationsbetween tibiofemoral joint space width measured off of weight-bearing X-ray, weight bearing MRI, and non-weight-bearing MRI50.In addition, frontal plane (varus-valgus) alignment of the tibia wascharacterized in the current study through the orientation of thetibial coordinate system relative to the femoral coordinate systemduring MRI acquisition. No side-to-side differences in varus-valgusalignment within ACL-injured and uninjured control subjects werefound. Consequently, limb alignment and muscle forces werethought to have minimal influence on both cartilage thickness orthe within ACL-injured subject differences reported in this study.Other limitations of this study include the incomplete under-standing of the effect of concomitant articular cartilage andmeniscal injuries on the differences in cartilage thickness profileseen within ACL-injured females. Most ACL-injured subjects suffersome degree of bone marrow lesions and a large number sufferarticular cartilage and meniscus lesions at the time of injury5,19.Data on concomitant injury were not collected on all subjects. As aresult, conclusions on their influence cannot be made. However,medical records including intra-operative arthroscopic documen-tation of meniscus and articular cartilage pathology were collectedon a subset of 25 female subjects, 13 and 15 of which had ameniscus or cartilage injury diagnoses, respectively. Post-hocanalysis was completed to investigate potential effects. No signifi-cant differences in articular cartilage thickness were identifiedbetween groups based on meniscus or cartilage injury status.Consequently, the influence of such concomitant pathology may besmall. Lastly, this study focused on characterizing changes in thetibial plateau articular cartilage. Changes in articular cartilagemorphology throughout the femoral condyles, trochlea and patellamay be equally, if not more, important. Future research should

E.C. Argentieri et al. / Osteoarthritis and Cartilage 22 (2014) 1453e14601458

investigate these anatomic structures with similar methodologiesin order to obtain a more comprehensive understanding of theeffects of an ACL-injury.

In conclusion, tibial plateau articular cartilage thickness differ-ences were identified between the injured and uninjured legs ofACL-injured females. These differences were located within themedial compartment only, and were significant after adjustmentfor tibial position relative to the femur during MRI data acquisition.No side-to-side differences in tibial cartilage thickness were foundwithin ACL-injured male subjects; and both male and femaleuninjured-control subjects displayed no side-to-side differences incartilage profile. Consequently, differences seenwithin ACL-injuredfemales represent a significant change in tibial plateau articularcartilage morphology as a result of ACL-injury. Quantification ofthese acute changes could act as a predictor for the progression ofcartilage damage, and yield insight to the onset of PTOA, at a pointin time where structural deterioration may be reversible.

Contributions

ECA: I am the primary author of this paper. I was involved in theconception and design of the study, as well as data collection.Additionally, I was involved in the post-processing, analyses andinterpretation of these data as well as the drafting, and finalapproval of the article.

DRS: I am a coauthor of this manuscript and was involved in thedesign of the experiment, Matlab coding, calculation of thicknesses,interpretation of the data, drafting, revision, and final approval ofthe article.

MJD: I am the head biostatistician for this project. I was involvedin the design of this study, the analysis and interpretation of data aswell as the critical revision of the article for important intellectualcontent.

MGG: Design of the experiment, Matlab coding for data surfacegeneration, calculation of thicknesses and correlation analyzes,interpretation of the data, drafting, revision, and final approval ofthe article.

JRS: I was involved in the designing the study, observing theacquisition of data, interpreting the data, revising the manuscriptcritically for important intellectual content and providing approvalof the version of the submitted.

RJJ: I was involved in the conception and design of this study asdemonstrated by attendance at many meetings concerning themethodology and analysis and interpretation of the data. Thearticle was circulated several times and I contributed by editing thearticle for its intellectual content, clarity and accuracy. I partici-pated in final approval of the version of the article that will besubmitted for publication.

BDB: I am the PI of this study and have been involved in theconception and design of this study, interpretation of these data, aswell as revision, and final approval of the article.

Role of funding sourcesResearch support provided by NIH NIAMS R01 AR050421.

MRI equipment funded by DOE SC 000017.

Competing interestsAuthors of this manuscript have no competing interests to disclose.

Acknowledgments

We would like to thank Timothy W. Tourville, Ph.D. for hissupport and technical insight, as well as Pamela M. Vacek, Ph.D. forher contributions and oversight of statistical methods.

Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.joca.2014.06.025.

References

1. Frobell RB, Le Graverand MP, Buck R, Roos EM, Roos HP,Tamez-Pena J, et al. The acutely ACL injured knee assessed byMRI: changes in joint fluid, bone marrow lesions, and cartilageduring the first year. Osteoarthritis Cartilage 2009;17:161e7.

2. Lohmander LS, Englund PM, Dahl LL, Roos EM. The long-termconsequence of anterior cruciate ligament and meniscus in-juries: osteoarthritis. Am J Sports Med 2007;35:1756e69.

3. Chaudhari AM, Briant PL, Bevill SL, Koo S, Andriacchi TP. Kneekinematics, cartilage morphology, and osteoarthritis after ACLinjury. Med Sci Sports Exerc 2008;40:215e22.

4. Li X, Kuo D, Theologis A, Carballido-Gamio J, Stehling C,Link TM, et al. Cartilage in anterior cruciate ligament-reconstructed knees: MR imaging T1{rho} and T2 e initialexperience with 1-year follow-up. Radiology 2011;258:505e14.

5. Potter HG, Jain SK, Ma Y, Black BR, Fung S, Lyman S. Cartilageinjury after acute, isolated anterior cruciate ligament tear:immediate and longitudinal effect with clinical/MRI follow-up.Am J Sports Med 2012;40:276e85.

6. Buckwalter JA. Articular cartilage injuries. Clin Orthop RelatRes 2002:21e37.

7. Martin JA, Brown T, Heiner A, Buckwalter JA. Post-traumaticosteoarthritis: the role of accelerated chondrocyte senescence.Biorheology 2004;41:479e91.

8. Kramer WC, Hendricks KJ, Wang J. Pathogenetic mechanismsof posttraumatic osteoarthritis: opportunities for early inter-vention. Int J Clin Exp Med 2011;4:285e98.

9. van Eck CF, Vyas SA, Fu FH. Ben would approve. Am J SportsMed 2010;38:NP2e3.

10. Van Ginckel A, Verdonk P, Witvrouw E. Cartilage adaptationafter anterior cruciate ligament injury and reconstruction:implications for clinical management and research? A sys-tematic review of longitudinal MRI studies. OsteoarthritisCartilage 2013;21:1009e24.

11. Spindler KP, Wright RW. Clinical practice. Anterior cruciateligament tear. N Engl J Med 2008;359:2135e42.

12. Griffin LY, Agel J, Albohm MJ, Arendt EA, Dick RW, Garrett WE,et al. Noncontact anterior cruciate ligament injuries: risk fac-tors and prevention strategies. J Am Acad Orthop Surg 2000;8:141e50.

13. Daniel DM, Stone ML, Dobson BE, Fithian DC, Rossman DJ,Kaufman KR. Fate of the ACL-injured patient. A prospectiveoutcome study. Am J Sports Med 1994;22:632e44.

14. von Porat A, Roos EM, Roos H. High prevalence of osteoar-thritis 14 years after an anterior cruciate ligament tear in malesoccer players: a study of radiographic and patient relevantoutcomes. Ann Rheum Dis 2004;63:269e73.

15. Karsdal MA, Christiansen C, Ladel C, Henriksen K, Kraus VB,Bay-Jensen AC. Osteoarthritis e a case for personalized healthcare? Osteoarthritis Cartilage 2014;22:7e16.

16. Andriacchi TP, Mundermann A, Smith RL, Alexander EJ,Dyrby CO, Koo S. A framework for the in vivo pathomechanicsof osteoarthritis at the knee. Ann Biomed Eng 2004;32:447e57.

17. Li G, Moses JM, Papannagari R, Pathare NP, DeFrate LE, Gill TJ.Anterior cruciate ligament deficiency alters the in vivo motionof the tibiofemoral cartilage contact points in both the

E.C. Argentieri et al. / Osteoarthritis and Cartilage 22 (2014) 1453e1460 1459

anteroposterior and mediolateral directions. J Bone Joint SurgAm 2006;88:1826e34.

18. Bevill SL, Briant PL, Levenston ME, Andriacchi TP. Central andperipheral region tibial plateau chondrocytes respond differ-ently to in vitro dynamic compression. Osteoarthritis Cartilage2009;17:980e7.

19. Frobell RB, Roos HP, Roos EM, Hellio Le Graverand MP, Buck R,Tamez-Pena J, et al. The acutely ACL injured knee assessed byMRI: are large volume traumatic bone marrow lesions a sign ofsevere compression injury? Osteoarthritis Cartilage 2008;16:829e36.

20. Slauterbeck JR, Kousa P, Clifton BC, Naud S, Tourville TW,Johnson RJ, et al. Geographic mapping of meniscus and carti-lage lesions associated with anterior cruciate ligament injuries.J Bone Joint Surg Am 2009;91:2094e103.

21. Lohmander LS, Ostenberg A, Englund M, Roos H. High preva-lence of knee osteoarthritis, pain, and functional limitations infemale soccer players twelve years after anterior cruciate lig-ament injury. Arthritis Rheum 2004;50:3145e52.

22. Oiestad BE, Holm I, Aune AK, Gunderson R, Myklebust G,Engebretsen L, et al. Knee function and prevalence of kneeosteoarthritis after anterior cruciate ligament reconstruction:a prospective study with 10 to 15 years of follow-up. Am JSports Med 2010;38:2201e10.

23. Li RT, Lorenz S, Xu Y, Harner CD, Fu FH, Irrgang JJ. Predictors ofradiographic knee osteoarthritis after anterior cruciate liga-ment reconstruction. Am J Sports Med 2011;39:2595e603.

24. Defrate LE, Papannagari R, Gill TJ, Moses JM, Pathare NP, Li G.The 6 degrees of freedom kinematics of the knee after anteriorcruciate ligament deficiency: an in vivo imaging analysis. Am JSports Med 2006;34:1240e6.

25. Hosseini A, Van de Velde S, Gill TJ, Li G. Tibiofemoral cartilagecontact biomechanics in patients after reconstruction of aruptured anterior cruciate ligament. J Orthop Res 2012;30:1781e8.

26. Scanlan SF, Favre J, Andriacchi TP. The relationship betweenpeak knee extension at heel-strike of walking and the locationof thickest femoral cartilage in ACL reconstructed and healthycontralateral knees. J Biomech 2013;46:849e54.

27. Van de Velde SK, Bingham JT, Hosseini A, Kozanek M,DeFrate LE, Gill TJ, et al. Increased tibiofemoral cartilage con-tact deformation in patients with anterior cruciate ligamentdeficiency. Arthritis Rheum 2009;60:3693e702.

28. Bowers ME, Trinh N, Tung GA, Crisco JJ, Kimia BB, Fleming BC.Quantitative MR imaging using “LiveWire” to measure tibio-femoral articular cartilage thickness. Osteoarthritis Cartilage2008;16:1167e73.

29. Eckstein F, Heudorfer L, Faber SC, Burgkart R, Englmeier KH,Reiser M. Long-term and resegmentation precision of quanti-tative cartilage MR imaging (qMRI). Osteoarthritis Cartilage2002;10:922e8.

30. Graichen H, von Eisenhart-Rothe R, Vogl T, Englmeier KH,Eckstein F. Quantitative assessment of cartilage status inosteoarthritis by quantitative magnetic resonance imaging:technical validation for use in analysis of cartilage volume andfurther morphologic parameters. Arthritis Rheum 2004;50:811e6.

31. Eckstein F, Charles HC, Buck RJ, Kraus VB, Remmers AE,Hudelmaier M, et al. Accuracy and precision of quantitativeassessment of cartilage morphology by magnetic resonanceimaging at 3.0T. Arthritis Rheum 2005;52:3132e6.

32. Eckstein F, Cicuttini F, Raynauld JP, Waterton JC, Peterfy C.Magnetic resonance imaging (MRI) of articular cartilage inknee osteoarthritis (OA): morphological assessment. Osteoar-thritis Cartilage 2006;14(Suppl A):A46e75.

33. Su F, Hilton JF, Nardo L, Wu S, Liang F, Link TM, et al. Cartilagemorphology and T1rho and T2 quantification in ACL-reconstructed knees: a 2-year follow-up. Osteoarthritis Carti-lage 2013;21:1058e67.

34. Koo S, Giori NJ, Gold GE, Dyrby CO, Andriacchi TP. Accuracy of3D cartilage models generated from MR images is dependenton cartilage thickness: laser scanner based validation of in vivocartilage. J Biomech Eng 2009;131:121004.

35. Frobell RB. Change in cartilage thickness, posttraumatic bonemarrow lesions, and joint fluid volumes after acute ACLdisruption: a two-year prospective MRI study of sixty-onesubjects. J Bone Joint Surg Am 2011;93:1096e103.

36. Tourville TW, Johnson RJ, Slauterbeck JR, Naud S, Beynnon BD.Assessment of early tibiofemoral joint space width changesafter anterior cruciate ligament injury and reconstruction: amatched case-control study. Am J Sports Med 2013;41:769e78.

37. Hunter DJ. Lower extremity osteoarthritis management needsa paradigm shift. Br J Sports Med 2011;45:283e8.

38. Li G, Thomson M, Dicarlo E, Yang X, Nestor B, Bostrom MP,et al. A chemometric analysis for evaluation of early-stagecartilage degradation by infrared fiber-optic probe spectros-copy. Appl Spectrosc 2005;59:1527e33.

39. Smith HC, Johnson RJ, Shultz SJ, Tourville T, Holterman LA,Slauterbeck JR, et al. A prospective evaluation of the LandingError Scoring System (LESS) as a screening tool for anteriorcruciate ligament injury risk. Am J Sports Med 2012;40:521e6.

40. Beynnon BD, Hall JS, Sturnick DR, Desarno MJ, Gardner-Morse M, Tourville TW, et al. Increased slope of the lateraltibial plateau subchondral bone is associated with greater riskof noncontact ACL injury in females but not in males: a pro-spective cohort study with a nested, matched case-controlanalysis. Am J Sports Med 2014;42:1039e48.

41. Beynnon BD, Vacek PM, Sturnick DR, Holterman LA, Gardner-Morse M, Tourville TW, et al. Geometric profile of the tibialplateau cartilage surface is associated with the risk of non-contact anterior cruciate ligament injury. J Orthop Res2014;32:61e8.

42. Hashemi J, Chandrashekar N, Gill B, Beynnon BD,Slauterbeck JR, Schutt Jr RC, et al. The geometry of the tibialplateau and its influence on the biomechanics of the tibiofe-moral joint. J Bone Joint Surg Am 2008;90:2724e34.

43. van Diek FM, Wolf MR, Murawski CD, van Eck CF, Fu FH. Kneemorphology and risk factors for developing an anterior cru-ciate ligament rupture: an MRI comparison between ACL-ruptured and non-injured knees. Knee Surg Sports TraumatolArthrosc 2013;22:987e94.

44. Fleming BC, Oksendahl HL, Mehan WA, Portnoy R, Fadale PD,Hulstyn MJ, et al. Delayed Gadolinium-Enhanced MR Imagingof Cartilage (dGEMRIC) following ACL injury. OsteoarthritisCartilage 2010;18:662e7.

45. Tiderius CJ, Olsson LE, Nyquist F, Dahlberg L. Cartilageglycosaminoglycan loss in the acute phase after an anteriorcruciate ligament injury: delayed gadolinium-enhanced mag-netic resonance imaging of cartilage and synovial fluid anal-ysis. Arthritis Rheum 2005;52:120e7.

46. Scanlan SF, Chaudhari AM, Dyrby CO, Andriacchi TP. Differ-ences in tibial rotation during walking in ACL reconstructedand healthy contralateral knees. J Biomech 2010;43:1817e22.

47. Quinn TM, Hunziker EB, Hauselmann HJ. Variation of cell andmatrix morphologies in articular cartilage among locations inthe adult human knee. Osteoarthritis Cartilage 2005;13:672e8.

48. Li G, DeFrate LE, Park SE, Gill TJ, Rubash HE. In vivo articularcartilage contact kinematics of the knee: an investigation

E.C. Argentieri et al. / Osteoarthritis and Cartilage 22 (2014) 1453e14601460

using dual-orthogonal fluoroscopy and magnetic resonanceimage-based computer models. Am J Sports Med 2005;33:102e7.

49. Zabala ME, Favre J, Scanlan SF, Donahue J, Andriacchi TP.Three-dimensional knee moments of ACL reconstructed andcontrol subjects during gait, stair ascent, and stair descent.J Biomech 2013;46:515e20.

50. Marsh M, Souza RB, Wyman BT, Hellio Le Graverand MP,Subburaj K, Link TM, et al. Differences between X-ray and MRI-determined knee cartilage thickness in weight-bearing andnon-weight-bearing conditions. Osteoarthritis Cartilage2013;21:1876e85.