Tibiofemoral Cartilage Contact Biomechanics in Patients afterReconstruction of a Ruptured Anterior Cruciate Ligament

Ali Hosseini, Samuel Van de Velde, Thomas J. Gill, Guoan Li

Bioengineering Laboratory, Massachusetts General Hospital/Harvard Medical School, 55 Fruit Street, GRJ 1215, Boston, Massachusetts 02114

Received 3 August 2011; accepted 19 March 2012

Published online 23 April 2012 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jor.22122

ABSTRACT: We investigated the in vivo cartilage contact biomechanics of the tibiofemoral joint in patients after reconstruction of aruptured anterior cruciate ligament (ACL). A dual fluoroscopic and MR imaging technique was used to investigate the cartilage contactbiomechanics of the tibiofemoral joint during in vivo weight-bearing flexion of the knee in eight patients 6 months following clinicallysuccessful reconstruction of an acute isolated ACL rupture. The location of tibiofemoral cartilage contact, size of the contact area,cartilage thickness at the contact area, and magnitude of the cartilage contact deformation of the ACL-reconstructed knees were com-pared with those previously measured in intact (contralateral) knees and ACL-deficient knees of the same subjects. Contact biomechan-ics of the tibiofemoral cartilage after ACL reconstruction were similar to those measured in intact knees. However, at lower flexion, theabnormal posterior and lateral shift of cartilage contact location to smaller regions of thinner tibial cartilage that has been described inACL-deficient knees persisted in ACL-reconstructed knees, resulting in an increase of the magnitude of cartilage contact deformationat those flexion angles. Reconstruction of the ACL restored some of the in vivo cartilage contact biomechanics of the tibiofemoral jointto normal. Clinically, recovering anterior knee stability might be insufficient to prevent post-operative cartilage degeneration due tolack of restoration of in vivo cartilage contact biomechanics. � 2012 Orthopaedic Research Society. Published by Wiley Periodicals, Inc.J Orthop Res 30:1781–1788, 2012

Keywords: anterior cruciate ligament (ACL); cartilage deformation; fluoroscopy; ACL reconstruction; postoperative osteoarthritis (OA)

Rupture of the ACL is a common acute injury that isassociated with an increased risk for developing osteo-arthritis (OA) in the affected knee.1,2 Each year�200,000 patients opt for ACL reconstruction in theUnited States,3 drawn by the excellent postoperativestability, health-related quality of life, and the abilityto return to sports.4,5 However, no long-term differencein OA prevalence has been detected between patientsthat opted for conservative treatment and those thatopted for surgery.6,7

Since current ACL reconstruction techniques canrestore the main mechanical function of the ACL (i.e.,control of AP translation of the tibiofemoral joint) onclinical evaluation, other pathogenic processes such asthe release of inflammatory cytokines in the synovialfluid8 or occult cartilage abnormalities9,10 that arealready present following ACL injury have been impli-cated in post-reconstruction cartilage degeneration.However, the contribution of mechanics should not beruled out in the long-term postoperative OA pathogen-esis. Studies that examined knee motion under func-tional loading conditions with advanced imagingmodalities all showed that abnormal translations androtations of the joint persist following ACL reconstruc-tion in spite of clinically satisfactory anterior stabili-ty.11–13 These studies have greatly enhanced ourinsight into the efficacy of contemporary reconstruc-tion techniques to reproduce normal in vivo knee kine-matics. More importantly, they triggered the quest forsurgical techniques that more closely restore normalkinematics,11–13 since even minimal alterations are

believed to affect cartilage contact,14 which in turn hasbeen hypothesized to contribute to the initiation ofOA.15 However, no methodology is available with thenecessary accuracy to directly quantify the extent ofthe biomechanical cartilage alteration in response tosuch persistent minimal alterations in knee kinemat-ics after ACL reconstruction.

In a recent study with a dual fluoroscopic andMRI technique, we indirectly found that—during aquasi-static lunge activity—ACL deficiency shifted thearticular contact location to smaller regions of thinnercartilage, and increased the cartilage contact deforma-tion, providing insight in the possible underlyingbiomechanical dimension of cartilage degeneration.14

Our current objective was to re-analyze the cartilagecontact biomechanics in our study sample of eightpatients with an acute isolated ACL rupture,14

now 6 months following clinically successful ACLreconstruction. We hypothesized that reconstruction isunable to correct the abnormal contact location andsize, the cartilage thickness at the contact area, andmagnitude of cartilage contact deformation caused byrupture of the ACL during in vivo weight-bearingflexion from 08 to 908. The healthy contralateral kneeswere considered as a normal group.

METHODSEight patients (5 men, 3 women, from 19 to 38 years old)that were recruited for our previous study were included inthe present study.14 Injury to other ligaments, noticeable car-tilage lesions or meniscal damage on MRI at 4.5 � 3 monthsafter injury and during arthroscopic examination, and injuryto the underlying bone were reasons for exclusion. Thepatients were active on a minimal to moderate athletic levelbefore injury based on the IKDC demographic form, and hada primary complaint of persistent instability that limited

Correspondence to: Guoan Li (T: 617-726-6472; F: 617-724-4392;E-mail: gli1@partners.org)

� 2012 Orthopaedic Research Society. Published by Wiley Periodicals, Inc.

JOURNAL OF ORTHOPAEDIC RESEARCH NOVEMBER 2012 1781

their athletic participation or work and prompted thedecision to undergo ACL reconstruction. Each patient signed aconsent form that had been approved by our Institutional Re-view Board. All the patients were scheduled for ACL recon-struction surgery within 1 week after their preoperative study.

ACL Reconstruction TechniqueThe patients underwent arthroscopic ACL reconstruction at4.5 � 3 months after injury. All surgeries were done by onesurgeon. A diagnostic arthroscopy was performed before graftplacement. Reconstruction was performed with a central 10-mm bone-patellar tendon-bone (BPTB) autograft. A 10-mmtibial tunnel was drilled using a 558 guide (Linvatec-Conmed,Largo, FL) centered 7 mm anterior to the PCL on the down-slope of the medial tibial spine. A 10-mm femoral tunnel wasdrilled using a 6-mm femoral offset guide (Arthrex, Naples,FL) centered at the 10:30 position for right knees (1:30 forleft). The graft was passed in retrograde fashion, and thefemoral and tibial bone blocks were secured with titaniuminterference screws (Guardsman, Linvatec-Conmed, Largo,FL). The femoral screw length was 25 mm and was placedwith the knee in maximal flexion. The tibial screw lengthwas 30 mm. The graft was fully tensioned with the knee infull extension. Screw diameter was determined based ongraft-tunnel fit. Examination confirmed that there was nonotch impingement, and cycling of the knee revealed <2 mmof graft motion. The anterior laxity of the reconstructed kneeas measured with the KT-1000 arthrometer was similar tothat of the intact contralateral knee.

Imaging Procedure14,16

Six months post-operatively, the ACL-reconstructed knee ofeach patient was simultaneously imaged using two orthogo-nally placed fluoroscopes (OEC 9800; GE Healthcare, SaltLake City, UT) as the patient performed a single-leg quasi-static lunge at 08, 158, 308, 608, and 908 of flexion—similar tothe lunge activity performed by the patient prior to the re-construction.14 At each angle, the patient was asked to pausefor 5 s while simultaneous fluoroscopic images were taken.Throughout the experiment, the leg being tested supportedthe patient’s body weight, while the other leg provided stabil-ity. Next, the images were imported into solid modeling soft-ware (Rhinoceros; Robert McNeel and Assoc, Seattle, WA)and placed in the orthogonal planes based on the position ofthe fluoroscopes during imaging of the patient.

The 3D surface mesh models of the tibia, fibula, femur,and articulating cartilage of the ACL-deficient knees con-structed based on 3T MR images pre-operatively were usedfor the present study: an MR scanner (Siemens, Malvern,PA) equipped with a surface coil and 3D double-echo waterexcitation sequence (field of view 16 cm � 16 cm � 12 cm)acquired sagittal plane images that were then imported intosolid modeling software to construct the bone and cartilagesurfaces of the knee.14 The 3D MRI-based model wasimported into the same software that contained the postoper-ative fluoroscopic images, viewed from the two orthogonaldirections corresponding to the orthogonal fluoroscopic setup,and independently manipulated in 6DOF inside the softwareuntil the projections of the model matched the outlines of thefluoroscopic images, so that the model reproduced the in vivoposition of the ACL-reconstructed knee. This analysis tech-nique has an error of <0.1 mm and a repeatability of <0.38in measuring the position and orientation of matched bones,respectively.17

Cartilage thickness was calculated by finding the smallestEuclidian distance connecting a vertex of the articular sur-face to the cartilage-bone interface of the 3D surface models.The size of the contact area was determined by computingthe area of tibial cartilage that intersected the femoral carti-lage.14 The accuracy and repeatability of cartilage thicknessmeasurement using MRI-based models of the knee joint hasbeen validated to be 0.04 � 0.01 mm.14 Cartilage contact de-formation was then defined for each vertex of the articularsurface mesh as the amount of cartilage surface intersection(Fig. 1) divided by the sum of the tibial and femoral cartilagesurface thicknesses.14 Previous validation studies showed anaccuracy of 4% and 14% when this technique was used tomeasure the cartilage contact deformation18 and cartilagecontact area,16 respectively. In this study, cartilage contactdeformation and its location were defined as the magnitudeand location of peak cartilage deformation, referenced toCartesian coordinate systems on the tibial plateaus.14 Theorigin of each coordinate system was located at the center ofa circle that was fit to the posterior edge of each tibialcompartment. The AP and mediolateral axes split eachplateau into quadrants. In the AP direction, a location ante-rior to the mediolateral axis was considered positive. In themediolateral direction, a location lateral to the AP axis wasconsidered positive.

Statistical AnalysisA two-way repeated-measures ANOVA and Newman–Keulspost-hoc test were used to determine if differences betweengroup-means of cartilage contact biomechanics existedmeasured in the eight subjects at the five flexion angles. Foreach knee compartment (medial and lateral), the dependentvariables were location, contact area, thickness, and cartilagecontact deformation. The independent variables were thestate of the knee (healthy contralateral before ACL recon-struction, ACL-deficient,14 and ACL-reconstructed) and theflexion angle. p-values < 0.05 were considered significant.Analyses were performed with Statistica software (Statis-tica�, StatSoft Inc., Tulsa, OK).

RESULTS

Location of Cartilage ContactACL reconstruction did not restore the combined later-al and posterior shift of the location of peak cartilagecontact deformation on the tibial plateaus that wasmeasured in ACL-deficient knees to normal levels atlower flexion angles. At 08 in the medial (lateral)compartment, contact in ACL-reconstructed kneesremained 4.8 � 2.7 mm (2.8 � 2.7 mm) posterior and4.5 � 3.5 mm (3.8 � 2.7 mm) lateral (p < 0.0001) fromthe location of contact in intact knees (Fig. 2). Between308 and 908, no significant differences were foundbetween the cartilage contact location in eithermediolateral or AP direction between the intact andACL-reconstructed knees (p > 0.05).

Size of Contact AreaThe cartilage contact area after ACL reconstructionremained significantly smaller than the normalcartilage contact area at 08 [314 � 114 mm2 in intactknees, 220 � 69 mm2 in ACL-deficient knees, and237 � 98 mm2 in ACL-reconstructed knees in the

1782 HOSSEINI ET AL.

JOURNAL OF ORTHOPAEDIC RESEARCH NOVEMBER 2012

medial compartment, p ¼ 0.008; 193 � 75 mm2 inintact knees, 137 � 64 mm2 in ACL-deficient knees,and 155 � 77 mm2 in ACL-reconstructed knees in thelateral compartment, p ¼ 0.005]. The cartilage contactareas between 158 and 908 were not significantly dif-ferent (although smaller) compared to those of normalgroup at these flexion angles (p > 0.05; Fig. 3).

Cartilage Thickness at the Contact AreaMaximum cartilage contact deformation remained atareas where cartilage thickness was an average of0.5 � 0.2 mm thinner than the normal conditionbetween 08 (p ¼ 0.004) and 308 (p ¼ 0.04) on themedial compartment and an average of 0.6 � 0.1 mmthinner than the normal condition at 08 (p ¼ 0.041)and 158 (p ¼ 0.036) on the lateral compartment(Fig. 4).

Magnitude of Cartilage Contact DeformationIn the medial compartment, significant differences inthe magnitude of cartilage contact deformationpersisted between healthy contralateral knees andACL-reconstructed knees at 08 (p < 0.0001) and 158(p ¼ 0.038) (Fig. 5). The maximum increase in

cartilage deformation occurred at 08, where a deforma-tion of 19 � 4%% was found in intact knees, 29 � 9%in ACL-deficient knees, and 27 � 3% in ACL-recon-structed knees (p < 0.0001) (a 42% increase from thedeformation in the intact knee). Similarly, at 08 carti-lage contact deformation in the lateral compartment ofACL-reconstructed knees was increased 29% comparedto the value measured in intact knees (24 � 9% intactknee, 33 � 6% ACL-deficient knee, 31 � 3% ACL-reconstructed knee, p ¼ 0.006) (Fig. 5). In Figure 6,the cartilage contact areas of intact, ACL-deficient andACL-reconstructed knees of a typical patient areshown at full extension (Fig. 6A) and 908 of flexion(Fig. 6B).

DISCUSSIONWe used a dual fluoroscopic and MR imaging tech-nique to investigate the cartilage contact biomechanicsindirectly in eight patients after ACL reconstruction.We found that the in vivo tibiofemoral cartilage con-tact biomechanics after clinically successful ACL re-construction were not restored to those measured inintact contralateral knees at lower flexion angles ina lunge activity (when the ACL is primarily

Figure 1. A 3D MRI-based knee model to illustrate the intersection of tibiofemoral cartilage surfaces in the medial (A) and lateral(B) compartments in response to weight bearing. Reproduced, with permission, from Hosseini A, et al. Osteoarthritis Cartilage. 2010;18:909–916.

CARTILAGE BIOMECHANICS AFTER ACL SURGERY 1783

JOURNAL OF ORTHOPAEDIC RESEARCH NOVEMBER 2012

functioning). In general, around 08 and 158 of flexion,the abnormal posterior and lateral shift of cartilagecontact location to smaller regions of thinner cartilageseen in ACL-deficient knees persisted in the ACL-reconstructed knees, resulting in a persistent increaseof cartilage contact deformation at those flexionangles.

Many clinical studies have demonstrated that APstability can be restored after ACL reconstruction.19,20

However, when the forces in the graft were measuredin cadaver knees, the graft forces were dramaticallylarger compared to those of the intact ACL.21,22 Suchabnormal force vectors could explain the altered kneekinematics that have been observed both in vitro23,24

Figure 3. Cartilage contact area on the medial (A) and lateral (B) tibial plateaus in intact, ACL-deficient knees,14 and ACL-reconstructed knees. Reprinted, with permission of the original publisher, Elsevier, from Hosseini A, et al.42

Figure 2. Location of cartilage contact on the medial tibial plateau in the anteroposterior (AP) (A) and mediolateral (ML) (B) direc-tions, and on the lateral tibial plateau in the AP (C), and ML (D) directions in intact, ACL-deficient knees,14 and ACL-reconstructedknees.

1784 HOSSEINI ET AL.

JOURNAL OF ORTHOPAEDIC RESEARCH NOVEMBER 2012

and in vivo11,12,25–27 after ACL reconstruction. Al-though the absolute changes in tibiofemoral kinemat-ics vary widely from study to study, a general trend ofpersistent increased anterior tibial translation com-bined with rotational changes in knee kinematics wereobserved in ACL-reconstructed knees during weight-bearing.28–30 These changes in tibiofemoral kinematicsafter ACL reconstruction are hypothesized to lead tochanges in the cartilage contact characteristics ulti-mately wearing down the articular cartilage.31 In thepresent study, we found that cartilage contact was lo-cated posteriorly and laterally on the tibial plateaus at08 and 158 of flexion, despite a clinically successful nor-malization of anterior knee stability. Analogous to ourobservations in ACL deficiency,14 a shift in cartilagecontact resulted in a considerable change in cartilage

loading distribution within the knee joint despite sur-gical reconstruction of the ACL.

In a recent study, two different ACL reconstructionmethods (anteroproximal vs. anatomic graft place-ment) were compared to quantify the effect of femoralgraft placement on the ability of ACL reconstructionin restoring normal knee kinematics.32 During a qua-si-static lunge activity, the knees with anatomic graftplacement more closely matched the contralateralknee kinematics (AP and mediolateral tibial transla-tions and internal tibial rotation). In the currentstudy, only a transtibial reconstruction technique wasused. Therefore, no solid conclusion could be madeabout the effect of surgical techniques and femoralgraft placement on restoration of tibiofemoral cartilagecontact biomechanics.

Figure 5. Peak cartilage contact deformation on the medial (A) and lateral (B) tibial plateaus in intact, ACL-deficient knees,14 andACL-reconstructed knees.

Figure 4. Cartilage thickness in regions of contact on the medial (A) and lateral (B) tibial plateaus in intact, ACL-deficient knees,14

and ACL-reconstructed knees. Broken horizontal line indicates the total average cartilage thickness.

CARTILAGE BIOMECHANICS AFTER ACL SURGERY 1785

JOURNAL OF ORTHOPAEDIC RESEARCH NOVEMBER 2012

We acknowledge several limitations to this study.Similar to the study of cartilage contact biomechanicsof ACL-deficient knees,14 no ground reaction forceswere measured to document whether global knee jointloading for the ACL-reconstructed knees were replicat-ed with the intact and the ACL-deficient knees. Weacquired data from only one functional activity, asingle leg lunge, using a goniometer to measure theflexion angle. However, knee joint kinematics andconsequently cartilage contact deformation are activi-ty-dependent33,34 so the present findings cannot beextrapolated to other knee movements. For example,with biplane radiography, Deneweth et al. found thatsingle-legged hopping—similar to the demandingjumping and cutting movements of high-level competi-tive sports—elicited significant differences in all exam-ined degrees of freedom except adduction of the tibiarelative to the femur,35 whereas downhill running onlycaused differences in tibial adduction and external ro-tation.30 If feasible, the incorporation of cartilage mor-phology into such high-speed motion analysis mightfurther quantify the impact of persistent kinematicchanges following ACL reconstruction during move-ments that are relevant to the this patient population.We did not include meniscus–cartilage contact. Howev-er, the fluoroscopic images reflected the in vivo struc-ture, including menisci, so they influenced thekinematics and the relative position of the cartilagesurfaces.

We cannot discern whether the enduring deforma-tion seen in the ACL-reconstructed knees will eventu-ally cause healthy cartilage to breakdown or whetherthe increased deformation was the result of cartilage

damage that was already present at the time of injuryand surgery, remaining more compliant throughoutthe postoperative follow-up. In addition, longitudinaldegenerative changes in response to altered joint load-ing have been well-demonstrated in animal modelswhere a transection of the ACL reliably triggers carti-lage degeneration.36,37 However, an analogous studythat tracks the longitudinal effect of instability-in-duced cartilage deformation on the integrity of undam-aged cartilage tissue at baseline has not beenperformed in patients.

The same 3D surface mesh models previously con-structed of the ACL-deficient knees were used for thecorresponding ACL-reconstructed knees at the 6-months follow-up visit to reduce variability in modelconstruction. Although in general significant changesare not expected in cartilage within such a time frame,there might be changes from the acute stage rightafter injury to this sub-acute stage.38 Another predica-ment has been the absence of valid markers that corre-late with and precede structural and biomechanicalcartilage changes. Patients with discernible cartilageand meniscal lesions on 3T MRI at 4.5 � 3 months ofinjury were excluded from the study. However, con-ventional MRI sequences such as the 3D double-echowater excitation that was used in this study are notparticularly sensitive to cartilage and meniscal lesions:cartilage damage is often only visible multiple yearsafter the initiation of OA on conventional MRimages.9,10,39 Such delay in the identification of under-lying bone marrow or cartilage abnormalities in theknees of some ACL deficient patients9 has historicallyimpeded recruitment to only patients with truly

Figure 6. Comparison of the cartilage contact areas of intact, ACL-deficient, and ACL-reconstructed knees of a typical patient at fullextension (A) and 908 of knee flexion (B).

1786 HOSSEINI ET AL.

JOURNAL OF ORTHOPAEDIC RESEARCH NOVEMBER 2012

isolated ruptures of the ACL. Therefore, rather thanrelying solely on clinical MR imaging, the final evalua-tion of cartilage and meniscal damage in our studyoccurred during arthroscopic examination at the timeof ACL reconstruction. We measured the kinematics ofhealthy contralateral knees pre-operatively. However,kinematics might have changed after ACL reconstruc-tion. It would be interesting to study and compare thekinematics of contralateral knees pre-and post-operatively.

Progress toward a more comprehensive insight inthe development of OA in ACL-reconstructed kneeshas been made by OA biomarker imaging assessment,such as T1r, delayed gadolinium-enhanced magneticresonance imaging of cartilage (dGEMRIC), and sodi-um MR.9,40,41 In a recent study, Li et al.39 evaluatedcartilage matrix changes with T1r and T2 relaxationtime quantification in 12 patients with acute ACL inju-ries and 1 year following reconstruction. The elevatedT1r values found at baseline in the posterolateralregion of the tibial cartilage of ACL-deficient knees,which were related to underlying bone marrow lesionsat the time of injury, decreased closer to normal valuesat 1 year, whereas the initially normal T1r values inthe medial tibiofemoral compartment significantlyincreased in the weightbearing subcompartments.Interestingly, in the present study, a relatively greaterpersistent increase in cartilage deformation wasmeasured in the medial tibiofemoral compartment ascompared with the lateral compartment both at thetime of injury and at 6-months follow-up. Thosechanges in T1r values would indicate changes in thecartilage matrix, which would agree with the changesin cartilage deformation seen in this study.

We believe our results provide insight into thechanges in in vivo tibiofemoral cartilage contactdeformation following ACL reconstruction and identifyimportant directions for future research. Clinicallysuccessful ACL reconstruction restored some of the invivo cartilage contact biomechanics of the tibiofemoraljoint to normal, but at lower flexion angles anincreased magnitude of cartilage contact deformationpersisted.

ACKNOWLEDGMENTSThe authors gratefully acknowledge the support of NIH (R01AR055612, F32 AR056451) and thank Bijoy Thomas, LouisDeFrate, and Jeffrey Bingham for their technical assistance.

REFERENCES1. Roos H, Adalberth T, Dahlberg L, et al. 1995. Osteoartliritis

of the knee after injury to the anterior cruciate ligament ormeniscus: the influence of time and age. OsteoarthritisCartilage 3:261–267.

2. Oiestad BE, Engebretsen L, Storheim K, et al. 2009. Kneeosteoarthritis after anterior cruciate ligament injury: asystematic review. Am J Sports Med 37:1434–1443.

3. Brophy RH, Wright RW, Matava MJ. 2009. Cost analysis ofconverting from single-bundle to double-bundle anteriorcruciate ligament reconstruction. Am J Sports Med 37:683–687.

4. Lewis PB, Parameswaran AD, Rue JP, et al. 2008. Systemat-ic review of single-bundle anterior cruciate ligament recon-struction outcomes: a baseline assessment for considerationof double-bundle techniques. Am J Sports Med 36:2028–2036.

5. Barenius B, Nordlander M, Ponzer S, et al. 2010. Quality oflife and clinical outcome after anterior cruciate ligamentreconstruction using patellar tendon graft or quadrupledsemitendinosus graft: an 8-year follow-up of a randomizedcontrolled trial. Am J Sports Med 38:1533–1541.

6. Daniel DM, Stone ML, Dobson BE, et al. 1994. Fate of theACL-injured patient. A prospective outcome study. Am JSports Med 22:632–644.

7. Lohmander LS, Englund PM, Dahl LL, et al. 2007. Thelong-term consequence of anterior cruciate ligament andmeniscus injuries: osteoarthritis. Am J Sports Med 35:1756–1769.

8. Cameron M, Buchgraber A, Passler H, et al. 1997. Thenatural history of the anterior cruciate ligament-deficientknee. Changes in synovial fluid cytokine and keratan sulfateconcentrations. Am J Sports Med 25:751–754.

9. Bolbos RI, Ma CB, Link TM, et al. 2008. In vivo Tlrho quan-titative assessment of knee cartilage after anterior cruciateligament injury using 3 Tesla magnetic resonance imaging.Invest Radiol 43:782–788.

10. Hernandez-Molina G, Guermazi A, Niu J, et al. 2008.Central bone marrow lesions in symptomatic knee osteoar-thritis and their relationship to anterior cruciate ligamenttears and cartilage loss. Arthritis Rheum 58:130–136.

11. Tashman S, Collon D, Anderson K, et al. 2004. Abnormalrotational knee motion during running after anterior cruci-ate ligament-reconstruction. Am J Sports Med 32:975–983.

12. Papannagari R, Gill TJ, Defrate LE, et al. 2006. In vivokinematics of the knee after anterior cruciate ligamentreconstruction: a clinical and functional evaluation. Am JSports Med 34:2006–2012.

13. Ristanis S, Stergiou N, Patras K, et al. 2005. Excessive tibialrotation during high-demand activities is not restored byanterior cruciate ligament reconstruction. Arthroscopy 21:1323–1329.

14. Van de Velde SK, Bingham JT, Hosseini A, et al. 2009.Increased tibiofemoral cartilage contact deformation inpatients with anterior cruciate ligament deficiency. ArthritisRheum 60:3693–3702.

15. Andriacchi TP, Mundermann A, Smith RL, et al. 2004. Aframework for the in vivo pathomechanics of osteo arthritisat the knee. Ann Biomed Eng 32:447–457.

16. Bingham JT, Papannagari R, Van de Velde SK, et al. 2008.In vivo cartilage contact deformation in the healthy humantibiofemoral joint. Rheumatology (Oxford) 47:1622–1627.

17. Defrate LE, Papannagari R, Gill TJ, et al. 2006. The6 degrees of freedom kinematics of the knee after anteriorcruciate ligament deficiency: an in vivo imaging analysis.Am J Sports Med 34:1240–1246.

18. Wan L, de Asla RJ, Rubash HE, et al. 2008. In vivo cartilagecontact deformation of human ankle joints under full bodyweight. J Orthop Res 26:1081–1089.

19. Fox JA, Nedeff DD, Bach BR Jr, et al. 2002. Anterior cruci-ate ligament reconstruction with patellar autograft tendon.Clin Orthop Relat Res 402:53–63.

20. Howe JG, Johnson RJ, Kaplan MJ, et al. 1991. Anteriorcruciate ligament reconstruction using quadriceps patellartendon graft. Part I. Long-term followup. Am. J Sports Med19:447–457.

21. Hoher J, Kanamori A, Zeminski J, et al. 2001. The positionof the tibia during graft fixation affects knee kinematics andgraft forces for anterior cruciate ligament reconstruction.Am J Sports Med 29:771–776.

CARTILAGE BIOMECHANICS AFTER ACL SURGERY 1787

JOURNAL OF ORTHOPAEDIC RESEARCH NOVEMBER 2012

22. Li G, Papannagari R, DeFrate LE, et al. 2006. Comparisonof the ACL and ACL graft forces before and after ACL recon-struction: an in-vitro robotic investigation. Acta Orthop 77:267–274.

23. Shoemaker SC, Adams D, Daniel DM, et al. 1993. Quadri-ceps/anterior cruciate graft interaction. An in vitro study ofjoint kinematics and anterior cruciate ligament grafttension. Clin Orthop 294:379–390.

24. Yoo J, Papannagari R, Park SE, et al. 2005. The effect ofACL reconstruction on knee joint kinematics under simulat-ed muscle loads. Am J Sports Med 33:240–246.

25. Georgoulis AD, Papadonikolakis A, Papageorgiou CD, et al.2003. Three-dimensional tibiofemoral kinematics of theanterior cruciate ligament-deficient and reconstructed kneeduring walking. Am J Sports Med 31:75–79.

26. Nordt WE, 3rd, Lotfi P, Plotkin E, et al. 1999. The in vivoassessment of tibial motion in the transverse plane in anteri-or cruciate ligament-reconstructed knees. Am J Sports Med27:611–616.

27. Brandsson S, Karlsson J, Sward L, et al. 2002. Kinematicsand laxity of the knee joint after anterior cruciate ligamentreconstruction: pre- and postoperative radiostereometricstudies. Am J Sports Med 30:361–367.

28. Georgoulis AD, Ristahis S, Chouliaras V, et al. 2007. Tibialrotation is not restored after ACL reconstruction with ahamstring graft. Clin Orthop Relat Res 454:89–94.

29. Ferretti A, Monaco E, Labianca L, et al. 2009. Double-bundleanterior cruciate ligament reconstruction: a comprehensivekinematic study using navigation. Am J Sports Med 37:1548–1553.

30. Tashman S, Kolowich P, Collon D, et al. 2007. Dynamicfunction of the ACL-reconstructed knee during running.Clin Orthop Relat Res 454:66–73.

31. Stergiou N, Ristanis S, Moraiti C, et al. 2007. Tibial rotationin anterior cruciate ligament (ACL)-deficient and ACL-recon-structed knees: a theoretical proposition for the developmentof osteoarthritis. Sports Med 37:601–613.

32. Abebe ES, Utturkar GM, Taylor DC, et al. 2011. The effectsof femoral graft placement on in vivo knee kinematics after

anterior cruciate ligament reconstruction. J Biomech 44:924–929.

33. Kozanek M, Hosseini A, Liu F, et al. 2009. Tibiofemoral ki-nematics and condylar motion during the stance phase ofgait. J Biomech 42:1877–1884.

34. Liu F, Kozanek M, Hosseini A, et al. 2009. In vivo tibiofe-moral cartilage deformation during the stance phase of gait.J Biomech 43:658–665.

35. Deneweth JM, Bey MJ, McLean SG, et al. 2010. Tibiofe-moral joint kinematics of the anterior cruciate ligament-reconstructed knee during a single-legged hop landing. Am JSports 38:1820–1828.

36. Brandt KD. 1991. Transection of the anterior cruciate liga-ment in the dog: a model of osteoarthritis. Sernin ArthritisRheum 21:22–32.

37. Liu W, Burton-Wurster N, Giant TT, et al. 2003. Spontane-ous and experimental osteoarthritis in dog: similarities anddifferences in proteoglycan levels. J Orthop Res 21:730–737.

38. Frobell RB, Le Graverand MP, Buck R, et al. 2009. Theacutely ACL injured knee assessed by MRI: changes in jointfluid, bone marrow lesions, and cartilage during the firstyear. Osteoarthritis Cartilage 17:161–167.

39. Li X, Kuo D, Theologis A, et al. 2010. Cartilage in anteriorcruciate ligament-reconstructed knees: MR imaging Tl{rho}and T2—initial experience with 1-year follow-up. Radiology258:505–514.

40. Samosky JT, Burstein D, Eric Grimson W, et al. 2005.Spatially-localized correlation of dGEMRIC-measured GAGdistribution and mechanical stiffness in the human tibialplateau. J Orthop Res 23:93–101.

41. Taylor C, Carballido-Gamio J, Majumdar S, et al. 2009.Comparison of quantitative imaging of cartilage for osteoar-thritis: T2, Tlrho, dGEMRIC and contrast-enhanced comput-ed tomography. Magn Reson Imaging 27:779–784.

42. Hosseini A, Van de Velde SK, Kozanek M, et al. 2010.In-vivo time-dependent articular cartilage contact behaviorof the tibiofemoral joint. Osteoarthritis Cartilage 18:909–916.

1788 HOSSEINI ET AL.

JOURNAL OF ORTHOPAEDIC RESEARCH NOVEMBER 2012