Current Concepts

Current Trends in Anterior CruciateLigament ReconstructionPart 1: Biology and Biomechanics of Reconstruction

Freddie H. Fu,* MD, Craig H. Bennett, MD, Christian Lattermann, MD, andC. Benjamin Ma, MD

From the Department of Orthopaedics, University of Pittsburgh, Pittsburgh, Pennsylvania

ABSTRACT

With today’s increasing emphasis on sporting activi-ties, the incidence of anterior cruciate ligament injurieshas also increased. Epidemiologic studies estimatethat the prevalence of anterior cruciate ligament inju-ries is about 1 per 3000 Americans. Management ofthese injuries has evolved from nonoperative treatmentto extracapsular augmentation and primary ligamentrepair to anterior cruciate ligament reconstruction.Treatment of these injuries has significantly improvedover the last few decades with the application of knowl-edge gained from both basic science and clinical re-search. This article is composed of two parts. The firstpart reviews the biology and biomechanics of the in-jured anterior cruciate ligament and the basic scienceof reconstruction. In the second part, to be publishedlater, current operative concepts of reconstruction, aswell as clinical correlations, are reviewed. Summariz-ing the latest information on basic scientific as well asclinical studies regarding the anterior cruciate liga-ment, this article intends to demonstrate the correlationbetween the application of basic science knowledgeand improvement of clinical outcomes.

Biomechanical analysis is recognized as a firm entity inknee ligament research. The biology of ligament healinghas received a tremendous amount of interest over thepast few decades with advances in knowledge of molecular

biology. In this first part of the article, the biology of theinjured ACL as well as pertinent and recent biomechani-cal and biologic studies involving the reconstructed ACLare reviewed. A goal of this section is to demonstrate therelationship between the biomechanics and the biology ofACL reconstructions.

BIOLOGY OF THE INJURED ACL

Clinical results of nonoperative treatment of the rupturedACL have shown that the majority of these injuries lead tofunctionally unacceptable outcomes.25,60,61,79 Patients of-ten complain of persistent joint instability and knee painafter such treatment. Moreover, other structures withinthe joint are at risk of injury in the ACL-deficient knee.74

Both biomechanical and clinical studies have found thatpatients with chronic ACL deficiency have an increasedincidence of meniscal injuries.51,101,106 Unlike the intra-articular ACL, extraarticular ligaments such as the me-dial collateral ligament (MCL) have good healing potentialand can have functional recovery with conservative treat-ment.10 The difference in healing potential has led clini-cians and scientists to believe that intraarticular and ex-traarticular ligaments have distinctly different biologicproperties.

Injuries to ligaments usually lead to the formation oflocal hematoma. In extraarticular ligaments, the hema-toma is organized into a fibrinogen mesh where invadingcells such as macrophages, monocytes, and other inflam-matory cells can settle. These inflammatory cells secrete amultitude of cytokines and growth factors that can medi-ate inflammation and attract fibroblasts and stem cells tothe injured ligament. The inflammatory response fadesafter a short period of time and granulation tissue isformed. The fibroblastic cells then gradually reorganizethis granulation tissue to form a scar or fibrous tissue.

* Address correspondence and reprint requests to Freddie H. Fu, MD,Department of Orthopaedic Surgery, 1010 Kaufmann Building, 3471 FifthAvenue, Pittsburgh, PA 15213.

No author or related institution has received financial benefit from researchin this study.

0363-5465/99/2727-0821$02.00/0THE AMERICAN JOURNAL OF SPORTS MEDICINE, Vol. 27, No. 6© 1999 American Orthopaedic Society for Sports Medicine

821

Unlike the MCL, the ACL is not embedded in a strong, softtissue envelope. The ACL has a thin, vascularized enve-lope formed by the synovial lining. The synovial lining ofthe knee joint is a highly organized tissue with the capac-ity to expand inside the joint. When this synovial sheath istorn during injury, blood dissipates within the joint anddoes not form a local fibrinous mesh at the site of injury.

The dependence of ACL healing on an intact synoviallining has been demonstrated by the poor healing re-sponse in ACL transection studies performed in ani-mals.41 When the synovial lining is intact, the hematomacan be kept in place, provide an organized fibrin mesh-work, and promote an early inflammatory response withthe associated release of cytokines and growth factors.Besides containing the blood clot, the synovium can alsoinitiate scar formation in partial ACL ruptures. Recentresearch demonstrates that fibroblasts within the syno-vial lining can migrate toward the site of ligament inju-ry.41,71 For complete ruptures, however, there is no localhealing response detectable at the injury site.41 Recentclinical work suggests that surgical formation of a bloodclot at the femoral insertion of the ACL after proximal,partial ACL ruptures can lead to reattachment of the ACLat its origin (J. R. Steadman, unpublished data, 1998).Once the blood clot has been able to form at the site ofinjury, there seems to be a potent reparative responseinitiated by cells within the synovial lining. Therefore, thehealing capacity of complete ACL ruptures may depend onthe extent of injury to the synovial sheath.

Besides loss of synovial sheath integrity, there are alsochanges in the intraarticular cytokine profile after ACLinjury. In chronic ACL-deficient knees, the levels of proin-flammatory cytokines such as interleukin-1 and tumornecrosis factor-a are markedly elevated; whereas protec-tive, antiinflammatory proteins such as the interleukinreceptor antagonist protein are significantly decreased.This change in cytokine profile can lead to a potentiallyaggressive environment, which may interfere with theregular healing response.21

The differences in healing potential between intraartic-ular and extraarticular ligaments are not completely un-derstood. Recent in vitro studies have shown that thereare distinct differences between the intrinsic properties ofACL and MCL fibroblasts. For example, the production ofextracellular matrix and collagenous proteins is signifi-cantly higher for fibroblasts within the ACL.77,105 Prolif-eration and migration of cells in the ACL, however, isslower than that in the MCL.91 Furthermore, under in-flammatory conditions fibroblasts in the ACL have lowermobility than those in the MCL.111 Further in vivo and invitro studies are needed to fully characterize the compli-cated healing process of these ligaments.

BIOMECHANICS AND BIOLOGY OF ACLRECONSTRUCTION

The poor healing capacity of the ACL observed clinicallyand confirmed in multiple in vitro and in vivo experimentshas led orthopaedic surgeons to perform ACL reconstruc-tions rather than repairs. In this section, we will address

a few important biomechanical and biologic issues thatare important for ACL reconstructions.

ACL Anatomy and Strength

The ACL is a two-bundle ligament consisting of antero-medial and posterolateral bundles that originate from thelateral femoral condyle within the intercondylar notch.The ACL inserts on the tibial plateau, medial to the in-sertion of the anterior horn of the lateral menis-cus.6,7,31,35 Recent studies performed on young humancadaveric knees have shown the ultimate tensile load andstiffness of the human femur-ACL-tibia complex to be2160 6 157 N and 242 6 28 N/mm, respectively.114 Thisintraarticular ligament is the primary restraint to ante-rior tibial translation.19,29,74 An understanding of thestructural properties of the intact ACL is important, asreplacement grafts should have similar tensile and dimen-sional properties as those of the intact ACL to best repro-duce its in vivo function.

Graft Material

Because of the unfavorable results of prosthetic ligamentreplacements,23,57,64,69,87 the most popular and success-ful surgical replacements for the ACL have been biologictissue grafts because of their potential for graft remodel-ing and integration into the joint.4 Biologic tissue graftsare available either as autografts or allografts. The advan-tages of autograft include low risk of adverse inflamma-tory reaction and virtually no risk of disease transmission.Allograft use, however, avoids donor site morbidity, de-creases surgical time, and diminishes postoperative pain.

Autografts. Throughout the history of ACL reconstruc-tion, various autograft choices have been used. The mostcommon current graft choices are the bone-patellar ten-don-bone graft8,47,80,94,96 and the quadruple semitendino-sus/gracilis tendon graft.2,49,58,70,73 The bone-patellartendon-bone graft is usually an 8- to 11-mm wide grafttaken from the central third of the patellar tendon with itsadjacent patellar and tibial bone blocks. The bone-patellartendon-bone graft has been a popular ACL replacementgraft because of its high ultimate tensile load (approxi-mately 2300 N), its stiffness (approximately 620 N/mm),88

and the possibility for rigid fixation with its attached bonyends.

In recent years, attention has shifted to the increaseduse of the hamstring tendon graft with its relatively lowdonor site morbidity.102 Hamstring tendon graft use hasevolved from a single-strand semitendinosus tendon graftto a quadruple-strand semitendinosus/gracilis tendongraft, where both the semitendinosus and gracilis tendonsare folded in half and combined.2,70,108 The dimension ofa round, 10-mm quadruple semitendinosus/gracilis ten-don graft is more comparable with that of the intact ACL,and its ultimate tensile load has been reported to be ashigh as 4108 N.16 The quadruple semitendinosus/gracilistendon graft also provides a multiple-bundle replacementgraft that may better approximate the function of thetwo-bundle ACL.113 Disadvantages of this soft tissue graft

822 Fu et al. American Journal of Sports Medicine

include the concern over tendon healing within the osse-ous tunnels and the lack of rigid bony fixation.

The quadriceps tendon graft for ACL reconstruction hasalso gained recent attention.30,47,62,88,110 Biomechanicalstudies have shown the ultimate tensile load of this graftto be as high as 2352 N.88 This replacement graft has beenfound to have adequate tensile properties, as well as size,for ACL reconstruction. The quadriceps tendon graft hasbecome an alternative replacement graft, especially forrevision ACL surgeries110 and for knees with multipleligament injuries. Table 1 shows a comparison of the ul-timate tensile loads of the various autografts used in ACLreconstruction.

Allografts. Allografts are soft tissue grafts harvestedfrom human donors. The morbidity associated with au-tograft harvest and the supply of allografts have increasedthe interest in allografts. However, the decrease in tensileproperties with sterilization and preservation as well asrisk of inflammatory reaction has been a concern. Allo-grafts are now commonly used for multiple ligament re-constructions, revision ligament reconstructions, and forpatients who are not high-performance athletes.

Commonly available allografts for ACL reconstructioninclude the patellar tendon and Achilles tendon. Withtheir immunogenic potential and risk of disease transmis-sion, allografts need adequate preparation before implan-tation. The effects of preservation and sterilization on thebiomechanical properties of the allograft have been thor-oughly studied. Allografts are usually preserved by deepfreezing or freeze-drying. Deep freezing without dryinghas little or no effect on the mechanical properties ofligaments,53,115 with no significant differences in the stiff-ness, ultimate load, or modulus noted between treated andcontrol ligaments. However, freeze-thaw treatments candamage cells and matrices, resulting in inferior biome-chanical properties.63

Sterile procurement of allograft tissues is optimal toavoid contamination; however, this process can be bothexpensive and time-consuming. Secondary sterilizationcan be achieved by ethylene oxide or gamma irradiation,but each has detrimental effects on the allograft. Ethyleneoxide derivatives remain on the tissue after ethylene oxidesterilization, despite extensive aeration techniques. Theseremains, namely ethylene glycol residues, can be retrievedat surgery in reconstructed knees.56 Currently, thismethod is not recommended for allograft sterilization.Gamma irradiation can reduce the risk of disease trans-

mission; however, it also has adverse effects on the tensileproperties of ligaments. A trend toward decreased struc-tural and mechanical properties of the tissue has beenobserved with irradiation, with a significant decrease seenat the 3-Mrad level of dosage.34,36,103 Irradiation alsoalters tissue morphology, during which collagen fasciclesbecome separated and the ligaments become visiblycrimped. To date, allografts are usually harvested in asterile environment and kept sterile by deep freezing, orlow-dose irradiation is used to achieve sterilization.

Besides differences in structural properties, the biologicproperties of allografts are also different from those ofautografts. In an animal study comparing goat patellartendon allograft and autograft, the knees that were recon-structed with autografts had smaller increases in antero-posterior laxity, higher ultimate tensile load, and moreadvanced biologic incorporation when compared with al-lograft-reconstructed knees.54 This study indicated thatallografts have slower incorporation than autografts at 6months after surgery. Both animal and clinical data sug-gest that, as with autograft tissue, allograft tissue revas-cularizes and becomes viable after implantation.9,54,98,99

There are a few studies demonstrating that the rate ofgraft incorporation and remodeling are slower for allograftthan autograft reconstructions.28,54 In addition, there aresome animal studies demonstrating a deterioration of al-lograft tissues after implantation.24,54,99 However, clini-cal studies have demonstrated minimal differences be-tween allograft and autograft reconstruction at 3 to 5years of follow-up.39,97

Graft Placement

An important aspect in ACL reconstruction is the place-ment of the soft tissue graft. Significant research has beendevoted to identifying the ideal position for graft place-ment with the hope of being able to best reproduce theanatomy and function of the intact ACL. The cross-sec-tional area of the intact ACL changes between its midsub-stance and at the insertion sites, where the ligamentbroadly inserts over the lateral intercondylar notch prox-imally and over the anterior tibial plateau distally.40 Be-cause there are broad insertion sites, a number of loca-tions for graft placement have been evaluated in anattempt to obtain the best replication of the in vivo func-tion of the ACL. Isometric placement of the ACL replace-ment graft was favored over other methods in thepast.42,43 Isometric graft placement limits changes ingraft length and tension during knee flexion and exten-sion. This concept was initially introduced to avoid signif-icant changes in graft tension during flexion and exten-sion, which may possibly lead to overstretching or failureof the graft. However, the concept of isometry is nowconsidered oversimplified, as recent basic science studieshave shown that the normal ACL is not isometric. Thefiber bundles of the ACL are under variable stress duringknee motion. For example, the anteromedial bundle of theACL experiences higher stress during flexion, and theposterolateral bundle experiences higher stress duringextension.31

TABLE 1Ultimate Tensile Load of the Intact Human ACL and a Few

Common ACL Replacement Graftsa

Graft type Ultimate tensile load (N)

Intact ACL 2160 6 157114

Bone-patellar tendon-bone (10 mm) 2376 6 15188

Single-strand semitendinosus 1216 6 5018

Quadruple hamstring graft 4108 6 20016

Quadriceps tendon (10 mm) 2352 6 49588

a Measurements in parentheses are widths of the graft. Refer-ence numbers refer to studies revealing higher limits of the ulti-mate tensile load.

Vol. 27, No. 6, 1999 Current Trends in ACL Reconstruction 823

With the use of robotic technology, the in situ force inthe ACL has been identified in response to an anteriortibial load. The in situ force in the ligament is defined asthe force carried by the ligament in its natural environ-ment. The in situ force in the ACL in response to ananterior tibial load is highest at 30° of knee flexion andslowly decreases with increasing knee flexion. The in situforce in the posterolateral bundle shows a similar trend asthe intact ACL, while the in situ force in the anteromedialbundle remains relatively unchanged throughout therange of knee flexion and extension (Fig. 1).85,86

Graft placement in ACL reconstructions has changedwith advances in understanding the importance of eachbundle of the ACL. A number of investigators advocategraft placement at the posterior portion of the ACL tibialinsertion site, near the posterolateral bundle position, forbest reproduction of the function of the intact ACL.5,48,50

Besides best reproducing the in situ force, this placementdecreases graft impingement with knee extension (Fig. 2).An anteriorly placed tibial tunnel can limit knee extensionwhen the graft impinges on the roof of the intercondylarnotch. While it is now generally accepted that placementof the tibial tunnel should be at the central or posteriorportion of the tibial insertion site of the intact ACL tibialinsertion, the ideal position for graft placement is stillunknown.

Initial Graft Tension

Besides graft placement, the tension applied to the re-placement graft at the time of graft fixation, or initialgraft tension, can significantly alter joint kinematics andin situ forces in the graft during knee motion. A low initialgraft tension will not provide joint stability. An exceed-ingly high initial graft tension can restrain joint motionand compromise graft survivability. Clinically, many in-vestigators have suggested an initial tension near 44 N, or

10 pounds, but this tension is empirically chosen and isnot scientifically based.

The effect of initial graft tension was first studied in acanine ACL reconstruction study.118 Initial graft tensionof 1 versus 39 N on patellar tendon autografts was exam-ined. At 3 months, no differences were found in antero-posterior joint stability or in tensile properties of the bone-graft-bone complex. However, the reconstructed kneesthat had an initial graft tension of 39 N did demonstratesome histologic evidence of degeneration of the articularcartilage. The effect of initial graft tensioning was alsoexamined in a prospective, randomized study of humanACL reconstructions.117 In this study, patients were ran-domly divided into three groups based on initial grafttension of 20, 40, and 80 N. At a minimum 2-year follow-up, the group with an initial graft tension of 80 N hadsignificantly less anterior laxity than the group with aninitial graft tension of 20 N. There were no reported dif-ferences in clinical symptoms for the three groups. How-ever, this study is a short-term follow-up report, and long-term effects of initial graft tension on the outcome of ACLreconstruction remains a subject to be investigated.107,117

As mentioned earlier, the in situ forces in the ACLchange with flexion and extension. A recent cadavericstudy compared the kinematics of the ACL-reconstructedknee with the graft tensioned and fixed at full extensionand at 30° of knee flexion.113 In response to a simulatedanterior drawer test, the anterior translation was signif-icantly less in the latter group (Fig. 3). This study dem-onstrated that the flexion angle of the knee at the time ofgraft fixation can significantly affect the resulting kine-matics of the reconstructed knee and should be consideredwhen performing ACL reconstructions.

The subject of initial graft tension remains controver-sial, as the in situ forces in the ACL during daily activitiesare unknown. In addition, the significance of the vis-coelastic behavior of ACL replacement grafts has not beenentirely characterized. Graft tension can significantly de-

Figure 1. The in situ force in the ACL in response to ananterior displacement load of 134 N at various knee flexionangles. AM, anteromedial bundle; PL, posterolateral bundle.(Used with permission from Sakane et al.85)

Figure 2. A schematic diagram indicating the desired posi-tion of the tibial tunnel to prevent graft impingement duringknee extension. (From Howell and Barad.48)

824 Fu et al. American Journal of Sports Medicine

crease with viscoelastic elongation of the replacementgrafts. It has been demonstrated that the stress in theprimate patellar tendon can decrease by up to 70% of theinitial stress within 30 minutes.37 Further studies areneeded to determine the importance of initial graft tensionon the success of ACL reconstructions.

BIOLOGY OF HEALING OF THE ACLREPLACEMENT GRAFT

After transplantation, tendon grafts undergo biologicmodifications before they form a strong fibrous tissue.Initially, the tendon undergoes inflammation and necro-sis. The graft then undergoes revascularization and re-population with fibroblasts. The last stage is marked by agradual remodeling of the graft and continuous modifica-tion of its collagenous structure.8,27 There is evidence thatautograft as well as allograft transplants are repopulatedwith extrinsic fibroblasts within 4 weeks.65 In an experi-mental dog model using allograft patellar tendon, it hasbeen shown that an allograft is repopulated within 4weeks with fibroblastic cells from the insertion sites. After4 to 6 weeks, the graft is completely repopulated. Donorfibroblasts undergo cell death and are not detectable afterthis period of time. The tendon structure, however, servesas a template for soft tissue remodeling.54,99

The synovium also plays an important role in ACLreconstruction. It has been demonstrated that a synoviallayer grows around the graft, forming a sheath that pro-vides blood supply to the transplanted tissue.45,99 Revas-cularization of the ACL graft from the proximal insertionsite results in increased blood supply to the graft for up to18 months.98,116 Biopsies of human ACL grafts haveshown that there is an early inflammatory response andneovascularization within 20 days after ACL reconstruc-tion. At 3 months, complete revascularization of the graft

takes place, initiated by the surrounding vascular syno-vial layer.1,58 These clinical findings underline the impor-tance of the well-vascularized synovial layer to the successof ACL reconstruction.

After ACL reconstruction, the graft undergoes restruc-turing of its collagen fibers and proteoglycan content. Thisprocess is known as remodeling. Whereas histologic sec-tions of ACL grafts show a normal crimp pattern after 6 to8 weeks, the ligament ultrastructure, the size of the col-lagen fibrils, and the distribution of glycosaminoglycansare significantly different from the original tissue. In bothautograft and allograft tissue, the extracellular matrixshows a distinctly different pattern between the normalACL and the patellar tendon graft.14 This process hasbeen observed up to 8 years after allograft ACL replace-ment in humans100 and corresponds with the findingsreported in sheep PCL replacement with a bone-patellartendon-bone autograft.14,15 Although the transplantedgraft shows ability to adapt to the mechanical demands, itnever reaches the structural identity of either the originaltendon tissue or the native ACL.

GRAFT FIXATION

Graft fixation is another important factor in cruciate lig-ament reconstruction. The bone-patellar tendon-bonegraft allows rigid fixation of the bony ends within theosseous tunnels, and healing is believed to be similar tofracture healing. Soft tissue tendon grafts, such as thequadruple semitendinosus/gracilis tendon graft, have adifferent healing process, with tendon healing within anosseous tunnel.38,83 A rabbit study using the semitendi-nosus tendon for ACL reconstruction found that the struc-tural properties of the replacement graft increased withtime; however, at 1 year after surgery, the ultimate tensileload was still only 25% of the intact ACL.13 While bone-tendon healing is generally believed to be slower and tohave inferior strength as compared with bone-bone heal-ing, there have been no reports definitively studying thedifference in biomechanical properties between bone-ten-don and bone-bone healing in the ACL-reconstructedknee.

With the advent of accelerated rehabilitation after ACLreconstruction, the demand for higher fixation strength towithstand early mobilization has also increased.95 Fixa-tion of replacement grafts can generally be divided intodirect and indirect methods. Direct fixation devices in-clude interference screws, staples, washers, and cross-pins. Indirect fixation devices include polyester tape-tita-nium button and suture-post. Significant research hasbeen performed to determine the stiffness and ultimatetensile load of these devices. The ultimate tensile loads ofthese fixations range between 200 and 1600 N (Table 2).

Interference screw fixation is the most popular fixationmethod for bone-patellar tendon-bone grafts. Many re-search studies have been performed to evaluate the sig-nificance of screw size, divergence, and direction of place-ment. Studies have shown that screw divergence cansignificantly affect the ultimate failure load. Screwsplaced parallel to the bone block have higher ultimate

Figure 3. The difference in kinematics in response to ananterior displacement load of 134 N at various knee flexionangles for an ACL reconstruction with the graft fixed at fullextension and at 30° of knee flexion. Intact, intact ACL; ACLdef, ACL deficient; Extension, ACL graft tensioned and fixedin extension; 30 deg, ACL graft tensioned and fixed at 30° ofknee flexion.

Vol. 27, No. 6, 1999 Current Trends in ACL Reconstruction 825

tensile loads than diverging screws.82 In a porcine study,screw divergence of more than 15° lowered the ultimatetensile load up to 50%. On the other hand, no significantdifferences in ultimate failure load have been demon-strated with changes in interference screw diameter.75,92

In recent years, bioabsorbable screws have been intro-duced. With the improvement in the material propertiesas well as screw design, the pull-out strength of bioabsorb-able screws is comparable with their metal counterparts,providing an alternative for graft fixation.20,59

Significant research has been performed on the ultimatetensile load of various fixation devices. Readers are cau-tioned that some of these studies were performed on cadav-eric knees from elderly donors or animal models that maynot represent the ultimate tensile load of the device whenused clinically. Nevertheless, these studies provide compar-ison data regarding different fixation methods (Table 2).

GRAFT TUNNEL MOTION

Traditional focus on the biomechanical properties of fixa-tion devices has been on the ultimate tensile load andpull-out strength. However, the stiffness, the amount ofgraft motion within the osseous tunnels, and the effects ofcyclic loading on the graft construct are also importantvariables.

Longitudinal Graft Tunnel Motion

The cyclic behavior of the soft tissue graft construct isanother variable in ACL reconstruction. A fixation devicecan have adequate ultimate tensile load but a low tensilestiffness that may compromise healing. Longitudinal grafttunnel motion (bungee effect) describes motion of the graftalong the axis of the osseous tunnel (Fig. 4). A recent

cadaveric study characterized hamstring tendon grafttunnel motion when using a polyester tape-titanium but-ton technique for femoral fixation.46 Graft tunnel motionof 1 to 3 mm occurred under physiologic loading of 100 to300 N, with the majority of motion contributed from thepolyester tape loop. Graft tunnel motion can be reducedwith the use of a shorter length of polyester tape loop anda longer portion of graft material within the osseoustunnel.

Sagittal Graft Tunnel Motion

Graft motion can also occur in the sagittal plane when thegraft moves anteriorly and posteriorly within the osseoustunnel, especially during knee flexion and extension. Thismotion is also known as the windshield wiper effect. Themagnitude of this sagittal motion can theoretically beincreased with fixation of the graft distant to the joint lineor osseous tunnel entrance. A number of studies have beenperformed to demonstrate the importance of this variablein ACL reconstruction (Fig. 5).12 In a porcine ACL recon-struction study, graft fixation that was distant to the jointline resulted in significantly more anterior laxity than didgraft fixation near the joint line.52 On the contrary, in aprospective clinical study, anatomic fixation near the jointline for bone-patellar tendon-bone graft ACL reconstruc-tion was compared with nonanatomic fixation that wasdistant to the joint line.3 This 18-month study showed nodifference in objective and subjective stability; however,distant fixation did result in a higher incidence of bone-tunnel widening. Long-term follow-up is needed to dem-onstrate the significance of these findings.

Tunnel Expansion

Tunnel expansion seems to occur with both bone-patellartendon-bone and hamstring tendon grafts (Fig. 6). Theradiolucency associated with this phenomenon is the ap-parent result of bone resorption in the insertion site area.This finding is supported by MRI data showing fluid sig-nal rather than bone signal surrounding the tendon.67 Todate, the clinical significance of this finding is unknown.The presence of bone resorption at the graft insertion siteas early as 3 months after ACL reconstruction is a dis-turbing phenomenon that needs to be followed closely.

BIOLOGY OF INSERTION SITE HEALING

The normal insertion site anatomy of the ACL has a spe-cific arrangement of collagen fibers, fibroblasts, fibrochon-droblasts, and osteoblasts forming a direct ligament inser-tion. The typical architecture of a direct ligamentinsertion consists of four layers (Fig. 7). The first layercomprises the ligament proper. The second layer is char-acterized as a nonmineralized cartilage zone containingfibrocartilaginous cells aligning themselves within the col-lagen bundles. The third layer is the mineralized cartilagezone. In this region, the fibrocartilage is mineralized andinserts into the subchondral bone plate, the fourth layer,to which the ligament is attached. This specific insertion

TABLE 2Ultimate Tensile Load of Various Fixation Devicesa

Type of fixation device Ultimate tensileload (N) Ref.

IndirectSingle polyester tape loop 375 6 8 104Double polyester tape loop 612–651 84, 104Single loop 5 Ethibond 238 6 3 104Double loop 5 Ethibond 463 6 18 104

Direct soft tissueMetal interference screw (7 mm) 242 6 90 20Bioabsorbable screw (7 mm) 341 6 163 20Bone mulch screw 1126 6 80 72Tandem soft tissue washers 768 72Cross-pin technique (animal) 725–1600 22Suture-post (animal) 374 78

Direct boneMetal interference screw (7 mm) 640 6 201 81Metal interference screw (9 mm) 276–436 59, 75Metal interference screw (11 mm) 302 75Metal interference screw (13 mm) 328 75Metal interference screw (15 mm) 328 75Bioabsorbable screw (7 mm) 330–418 81Bioabsorbable screw (9 mm) 565 59Staples 588 32a Experiments were performed on human cadaveric knees un-

less specified.

826 Fu et al. American Journal of Sports Medicine

site anatomy is designed to distribute longitudinal andshear forces from the ligament proper into the subchon-dral bone plate, thus minimizing stresses on individualcollagen bundles.11 This complex anatomy, however, is notrestored by conventional free tendon transfers within thefirst 6 months after graft implantation. Direct implanta-tion of tendon within a bone tunnel yields a soft tissuefixation predominantly composed of fibrous tissue aligned

along the load axis. After 6 weeks, collagenous fibers canbe seen protruding from the collagen fibers of the graftthrough the fibrous tissue into bone, thus anchoring thetendon transplant directly into the cancellous bone.68,83

Anterior cruciate ligament insertion site healing hasbeen evaluated in animal studies. Studies involving ACLreconstruction in rabbits using a free semitendinosus ten-

Figure 4. A, a titanium button-polyester tape construction for soft tissue fixation. B, a schematic diagram illustrating longitudinalgraft tunnel motion, or the bungee effect.

Figure 5. A schematic diagram demonstrating anatomicversus distant tibial-sided fixation of the ACL replacementgraft. (Modified from Ishibashi et al.52)

Figure 6. Anteroposterior (left) and lateral (right) radio-graphs illustrating bone-tunnel widening after ACLreconstruction.

Vol. 27, No. 6, 1999 Current Trends in ACL Reconstruction 827

don graft show a firm junction between the graft and thebone tunnel after 6 weeks.13,38 This junction, however,shows longitudinally directed fibrous tissue that does notresemble restoration of a normal ACL insertion. In biome-chanical tensile tests, the semitendinosus tendon graftwas pulled out of the tunnel, leaving a fibrous liningwithin the bone tunnel. This suggests that the graft hasbeen pulled out of the fibrous tissue envelope rather thanout of a firm bony attachment.13,38 The work of Shino etal.,99 however, suggests that, in dogs, the fibrous tissueformation within the bone tunnel transforms into a directinsertion demonstrating the four-zonal structure after 30to 52 weeks. These studies suggest that insertion sitehealing may undergo a progression from unstructuredfibrous tissue to structured tissue and to the reestablish-ment of a normal insertion morphology.

A recent rabbit study examined the incorporation of thebone-patellar tendon-bone graft into the bone tunnel. Inthis study, the incorporation of the bone block was de-scribed as early as 16 weeks after transplantation, and anormal four-zonal insertion site was reestablished after 6to 9 months.90 These data are surprising since the regulartime for bone healing is usually faster than 16 weeks. Theresults of this study suggest that the healing response ofthe bone-patellar tendon-bone autograft may be morecomplex than a direct bone-to-bone healing response. Fu-ture studies will be needed to determine the differences inhealing response and ultrastructural quality of bone-pa-tellar tendon-bone versus free tendon grafts.

FUTURE DIRECTIONS OF ACLRECONSTRUCTION

Recent emphasis in research has been directed towardunderstanding the importance of the ACL in vivo as wellas the rehabilitation and healing process after ACL recon-struction. Biomechanical analysis of the ACL and ACL-reconstructed knees has traditionally been evaluated incadaveric studies. The results of these studies are ex-tremely useful; however, their application can be limited

without the information of their behavior in vivo. Significanteffort has been directed toward obtaining in vivo biomechan-ics regarding the ACL and ACL-reconstructed knees. In vivostrain measurements12 and gait analysis17,66,109 can pro-vide valuable information that can be used in modifyingrehabilitation protocols, surgical approaches, and preopera-tive planning. This information may also allow us to havemore thorough evaluation of ACL-reconstructed knees.

Another focus of orthopaedic research has been the ap-plication of technology in molecular biology to acceleratingand promoting soft tissue healing. Growth factors havebeen identified as specific signaling molecules that lead tochanges in the proliferative and migratory behavior ofvarious cells. A recent study has shown that growth fac-tors such as the basic fibroblast growth factor, transform-ing growth factor-b, and the platelet-derived growth factorcan promote the healing response of ligaments both invitro and in vivo.44,89 Promising reports have also beenpublished on the enhancement of ligament insertion sitehealing after the application of growth factors known asgrowth differentiation factors.112

Although some success has been demonstrated withdirect application of growth factors to ligament injuries inanimal experiments, there are problems associated withdirect delivery. These problems include short-term actionand an exceedingly high local concentration of proteins.The development of delivery techniques has been aimed atachieving continuous, local delivery of growth factors.Gene therapy has emerged as one of the more promisingapproaches to overcome this problem. By using viral ornonviral vectors, the genetic information encoding theprotein of interest is inserted into a live cell. The geneti-cally modified cell has the potential to express the proteinof interest in a sustained manner. This gene-transfermethod is a potential vehicle for targeted, long-term de-livery of protein to the healing ACL.26,33

REFERENCES

1. Abe S, Kurosaka M, Iguchi T, et al: Light and electron microscopic studyof remodeling and maturation process in autogenous graft for anteriorcruciate ligament reconstruction. Arthroscopy 9: 394–405, 1993

2. Aglietti P, Buzzi R, Zaccherotti G, et al: Patellar tendon versus doubledsemitendinosus and gracilis tendons for anterior cruciate ligament re-construction. Am J Sports Med 22: 211–218, 1994

3. Aglietti P, Zaccherotti G, Simeone AJV, et al: Anatomic versus non-anatomic tibial fixation in anterior cruciate ligament reconstruction withbone-patellar tendon-bone graft. Knee Surg Sports Traumatol Arthrosc6: S43–S48, 1998

4. Amiel D, Kleiner JB, Roux RD, et al: The phenomenon of “ligamentiza-tion”: Anterior cruciate ligament reconstruction with autogenous patellartendon. J Orthop Res 4: 162–172, 1986

5. Amis AA, Jakob RP: Anterior cruciate ligament graft positioning, tension-ing and twisting. Knee Surg Sports Traumatol Arthrosc 6: S2–S12, 1998

6. Arnoczky SP, Marshall JL: The cruciate ligaments of the canine stifle: Ananatomical and functional analysis. Am J Vet Res 38: 1807–1814, 1977

7. Arnoczky SP, Matyas JR, Buckwalter JA, et al: Anatomy of the anteriorcruciate ligament, in Jackson DW, Arnoczky SP, Woo SL-Y, et al. (eds):The Anterior Cruciate Ligament: Current and Future Concepts. NewYork, Raven Press, 1993, pp 5–22

8. Arnoczky SP, Tarvin GB, Marshall JL: Anterior cruciate ligament replace-ment using patellar tendon. An evaluation of graft revascularization in thedog. J Bone Joint Surg 64A: 217–224, 1982

9. Arnoczky SP, Warren RF, Ashlock MA: Replacement of the anteriorcruciate ligament using a patellar tendon allograft. An experimentalstudy. J Bone Joint Surg 68A: 376–385, 1986

Figure 7. A histologic section of a direct insertion showingthe four layers of insertion: ligament proper, uncalcified fibro-cartilage layer, calcified fibrocartilage layer, and bone.

828 Fu et al. American Journal of Sports Medicine

10. Ballmer PM, Jakob RP: The non operative treatment of isolated completetears of the medial collateral ligament of the knee. A prospective study.Arch Orthop Trauma Surg 107: 273–276, 1988

11. Benjamin M, Evans EJ, Copp L: The histology of tendon attachments tobone in man. J Anat 149: 89–100, 1986

12. Beynnon BD, Fleming BC: Anterior cruciate ligament strain in-vivo. Areview of previous works. J Biomech 31: 519–525, 1998

13. Blickenstaff KR, Grana WA, Egle D: Analysis of a semitendinosus au-tograft in a rabbit model. Am J Sports Med 25: 554–559, 1997

14. Bosch U, Gassler N, Decker B: Alterations of glycosaminoglycans duringpatellar tendon autograft healing after posterior cruciate ligament re-placement. A biochemical study in a sheep model. Am J Sports Med 26:103–108, 1998

15. Bosch U, Kasperczyk WJ: Healing of the patellar tendon autograft afterposterior cruciate ligament reconstruction—a process of ligamentiza-tion? An experimental study in a sheep model. Am J Sports Med 20:558–566, 1992

16. Brown CH Jr, Steiner ME, Carson EW: The use of hamstring tendons foranterior cruciate ligament reconstruction: Technique and results. ClinSports Med 12: 723–756, 1993

17. Bulgheroni P, Bulgheroni MV, Andrini L, et al: Gait patterns after anteriorcruciate ligament reconstruction. Knee Surg Sports Traumatol Arthrosc5: 14–21, 1997

18. Butler DL: Anterior cruciate ligament: Its normal response and replace-ment. J Orthop Res 7: 910–921, 1989

19. Butler DL, Noyes FR, Grood ES: Ligamentous restraints to anterior-posterior drawer in the human knee: A biomechanical study. J Bone JointSurg 62A: 259–270, 1980

20. Caborn DNM, Coen M, Neef R, et al: Quadrupled semitendinosus-gracilis autograft fixation in the femoral tunnel: A comparison between ametal and a bioabsorbable interference screw. Arthroscopy 14: 241–245,1998

21. Cameron M, Buchgraber A, Passler H, et al: The natural history of theanterior cruciate ligament-deficient knee. Changes in synovial fluid cy-tokine and keratan sulfate concentrations. Am J Sports Med 25: 751–754, 1997

22. Clark R, Olsen RE, Larson BJ, et al: Cross-pin femoral fixation: A newtechnique for hamstring anterior cruciate ligament reconstruction of theknee. Arthroscopy 14: 258–267, 1998

23. Denti M, Bigoni M, Dodaro G, et al: Long-term results of the Leeds-Keioanterior cruciate ligament reconstruction. Knee Surg Sports TraumatolArthrosc 3: 75–77, 1995

24. Drez DJ Jr, DeLee J, Holden JP, et al: Anterior cruciate ligament recon-struction using bone-patellar tendon-bone allografts. A biological andbiomechanical evaluation in goats. Am J Sports Med 19: 256–263, 1991

25. Drongowski RA, Coran AG, Wojtys EM: Predictive value of meniscal andchondral injuries in conservatively treated anterior cruciate ligamentinjuries. Arthroscopy 10: 97–102, 1994

26. Evans CH, Robbins PD: Possible orthopaedic applications of gene ther-apy [current concept review]. J Bone Joint Surg 77A: 1103–1114, 1995

27. Frank C, Amiel D, Woo SL-Y, et al: Normal ligament properties andligament healing. Clin Orthop 196: 15–25, 1985

28. Fu FH, Jackson DW, Jamison J, et al: Allograft reconstruction of theanterior cruciate ligament, in Jackson DW, Arnoczky SP, Woo SL-Y, etal. (eds): The Anterior Cruciate Ligament. Current and Future Concepts.New York, Raven Press, 1993, pp 325–338

29. Fukubayashi T, Torzilli PA, Sherman MF, et al: An in vitro biomechanicalevaluation of anterior-posterior motion of the knee. Tibial displacement,rotation, and torque. J Bone Joint Surg 64A: 258–264, 1982

30. Fulkerson JP, Langeland R: An alternative cruciate reconstruction graft:The central quadriceps tendon [technical note]. Arthroscopy 11: 252–254, 1995

31. Furman W, Marshall JL, Girgis FG: The anterior cruciate ligament. Afunctional analysis based on postmortem studies. J Bone Joint Surg 58A:179–185, 1976

32. Gerich TG, Cassim A, Lattermann C, et al: Pullout strength of tibial graftfixation in anterior cruciate ligament replacement with a patellar tendongraft: Interference screw versus staple fixation in human knees. KneeSurg Sports Traumatol Arthrosc 5: 84–88, 1997

33. Gerich TG, Kang R, Fu FH, et al: Gene transfer to the rabbit patellartendon: Potential for genetic enhancement of tendon and ligament heal-ing. Gene Ther 3: 1089–1093, 1996

34. Gibbons MJ, Butler DL, Grood ES, et al: Dose-dependent effects ofgamma irradiation on the material properties of frozen bone-patellartendon-bone allografts. Trans Orthop Res Soc 14: 513, 1989

35. Girgis FG, Marshall JL, Al Monajem ARS: The cruciate ligaments of theknee joint: Anatomical, functional and experimental analysis. Clin Orthop106: 216–231, 1975

36. Goertzen MJ, Clahsen H, Burrig KF, et al: Sterilisation of canine anteriorcruciate allografts by gamma irradiation in argon. Mechanical and neu-rohistological properties retained one year after transplantation. J BoneJoint Surg 77B: 205–212, 1995

37. Graf BK, Vanderby R Jr, Ulm MJ, et al: Effect of preconditioning on theviscoelastic response of primate patellar tendon. Arthroscopy 10: 90–96,1994

38. Grana WA, Egle DM, Mahnken R, et al: An analysis of autograft fixationafter anterior cruciate ligament reconstruction in a rabbit model.Am J Sports Med 22: 344–351, 1994

39. Harner CD, Olson E, Irrgang JJ, et al: Allograft versus autograft anteriorcruciate reconstruction. 3- to 5-year outcome. Clin Orthop 324: 134–144,1996

40. Harner CD, Xerogeanes JW, Livesay GA, et al: The human posteriorcruciate ligament complex: An interdisciplinary study. Ligament morphol-ogy and biomechanical evaluation. Am J Sports Med 23: 736–745, 1995

41. Hefti FL, Kress A, Fasel J, et al: Healing of the transected anteriorcruciate ligament in the rabbit. J Bone Joint Surg 73A: 373–383, 1991

42. Hefzy MS, Grood ES: Sensitivity of insertion locations on length patternsof anterior cruciate ligament fibers. J Biomech Eng 108: 73–82, 1986

43. Hefzy MS, Grood ES, Noyes FR: Factors affecting the region of mostisometric femoral attachments. Part II: The anterior cruciate ligament.Am J Sports Med 17: 208–216, 1989

44. Hildebrand KA, Woo SL-Y, Smith DW, et al: The effects of platelet-derived growth factor-BB on healing of the rabbit medial collateral liga-ment. An in vivo study. Am J Sports Med 26: 549–554, 1998

45. Hoffmann MW, Wening JV, Apel R, et al: Repair and reconstruction ofthe anterior cruciate ligament by the “sandwich technique.” A compara-tive microangiographic and histological study in the rabbit. Arch OrthopTrauma Surg 112: 113–120, 1993

46. Hoher J, Livesay GA, Ma CB, et al: Hamstring graft motion in the femoralbone tunnel when using titanium button/polyester tape fixation. KneeSurg Sports Trauma Arthrosc 7: 215–219, 1999

47. Howe JG, Johnson RJ, Kaplan MJ, et al: Anterior cruciate ligamentreconstruction using quadriceps patellar tendon graft. Part I. Long-termfollowup. Am J Sports Med 19: 447–457, 1991

48. Howell SM, Barad SJ: Knee extension and its relationship to the slope ofthe intercondylar roof. Implications for positioning the tibial tunnel inanterior cruciate ligament reconstructions. Am J Sports Med 23: 288–294, 1995

49. Howell SM, Taylor MA: Brace-free rehabilitation, with early return toactivity, for knees reconstructed with a double-looped semitendinosusand gracilis graft. J Bone Joint Surg 78A: 814–825, 1996

50. Howell SM, Taylor MA: Failure of reconstruction of the anterior cruciateligament due to impingement by intercondylar roof. J Bone Joint Surg75A: 1044–1055, 1993

51. Indelicato PA, Bittar ES: A perspective of lesions associated with ACLinsufficiency of the knee: A review of 100 cases. Clin Orthop 198: 77–80,1985

52. Ishibashi Y, Rudy TW, Livesay GA, et al: The effect of anterior cruciateligament graft fixation site at the tibia on knee stability: Evaluation usinga robotic testing system. Arthroscopy 13: 177–182, 1997

53. Jackson DW, Grood ES, Cohn BT, et al: The effects of in situ freezing ofthe anterior cruciate ligament. An experimental study in goats. J BoneJoint Surg 73A: 201–213, 1991

54. Jackson DW, Grood ES, Goldstein JD, et al: A comparison of patellartendon autograft and allograft used for anterior cruciate ligament recon-struction in the goat model. Am J Sports Med 21: 176–185, 1993

55. Jackson DW, Simon TM, Kurzweil PR, et al: Survival of cells afterintra-articular transplantation of fresh allografts of the patellar and ante-rior cruciate ligaments. DNA-probe analysis in a goat model. J Bone JointSurg 74A: 112–118, 1992

56. Jackson DW, Windler GE, Simon TM: Intraarticular reaction associatedwith the use of freeze-dried, ethylene oxide-sterilized bone-patella ten-don-bone allografts in the reconstruction of the anterior cruciate liga-ment. Am J Sports Med 18: 1–11, 1990

57. Jarvinen M, Jozsa L, Johnson RJ, et al: Effect of anterior cruciateligament reconstruction with patellar tendon or prosthetic ligament on themorphology of the other ligaments of the knee joint. An experimentalstudy in dogs. Clin Orthop 311: 176–182, 1995

58. Johnson LL: The outcome of a free autogenous semitendinosus tendongraft in human anterior cruciate reconstructive surgery: A histologicalstudy. Arthroscopy 9: 131–142, 1993

59. Johnson LL, Dyk GY: Metal and biodegradable interference screws:Comparison of failure strength. Arthroscopy 12: 452–456, 1996

60. Johnson RJ, Beynnon BD, Nichols CE, et al: The treatment of injuries tothe anterior cruciate ligament [current concept review]. J Bone Joint Surg74A: 140–151, 1992

61. Kannus P, Jarvinen M: Conservatively treated tears of the anterior cruciateligament: Long-term results. J Bone Joint Surg 69A: 1007–1012, 1987

62. Kaplan MJ, Howe JG, Fleming B, et al: Anterior cruciate ligament recon-struction using quadriceps patellar tendon graft. Part II. A specific sportreview. Am J Sports Med 19: 458–462, 1991

63. King G, Edwards P, Brant R, et al: Freeze-thawing impairs long termhealing of a rabbit medial collateral ligament autograft model. TransOrthop Res Soc 17: 660, 1992

Vol. 27, No. 6, 1999 Current Trends in ACL Reconstruction 829

64. Klein W, Jensen KU: Synovitis and artificial ligaments. Arthroscopy 8:116–124, 1992

65. Kleiner JB, Amiel D, Roux RD, et al: Origin of replacement cells for theanterior cruciate ligament autograft. J Orthop Res 4: 466–474, 1986

66. Kowalk DL, Duncan JA, McCue FC III, et al: Anterior cruciate ligamentreconstruction and joint dynamics during stair climbing. Med Sci SportsExerc 29: 1406–1413, 1997

67. L’Insalata JC, Klatt B, Fu FH, et al: Tunnel expansion following anteriorcruciate ligament reconstruction: A comparison of hamstring and patellartendon autografts. Knee Surg Sports Traumatol Arthrosc 5: 234–238,1997

68. Liu SH, Panossian V, Al-Shaikh R, et al: Morphology and matrix com-position during early tendon to bone healing. Clin Orthop 339: 253–260,1997

69. Macnicol MF, Penny ID, Sheppard L: Early results of the Leeds-Keioanterior cruciate ligament replacement. J Bone Joint Surg 73B: 377–380,1991

70. Maeda A, Shino K, Horibe S, et al: Anterior cruciate ligament reconstruc-tion with multistranded autogenous semitendinosus tendon. Am J SportsMed 24: 504–509, 1996

71. Maekawa K, Furukawa H, Kanazawa Y, et al: Electron and immunoelec-tron microscopy on healing process of the rat anterior cruciate ligamentafter partial transection: The roles of multipotent fibroblasts in the syno-vial tissue. Histol Histopathol 11: 607–619, 1996

72. Magen HE, Howell SM, Hull ML: Structural properties of six tibial fixationmethods for anterior cruciate ligament soft tissue grafts. Am J SportsMed 27: 35–43, 1999

73. Marder TA, Raskind JR, Carroll M: Prospective evaluation of arthroscopi-cally assisted anterior cruciate ligament reconstruction: Patellar tendonversus semitendinosus and gracilis tendons. Am J Sports Med 19:478–484, 1991

74. Markolf KL, Mensch JS, Amstutz HC: Stiffness and laxity of the knee—the contributions of the supporting structures. A quantitative in vitrostudy. J Bone Joint Surg 58A: 583–594, 1976

75. Matthews LS, Parks BG, Sabbagh RC: Determination of fixation strengthof large-diameter interference screws. Arthroscopy 14: 70–74, 1998

76. Miyasaka KC, Daniel DM, Stone ML, et al: The incidence of kneeligament injuries in the general population. Am J Knee Surg 4: 3–8, 1991

77. Nagineni CN, Amiel D, Green MH, et al: Characterization of the intrinsicproperties of the anterior cruciate and medial collateral ligament cells: Anin vitro cell culture study. J Orthop Res 10: 465–475, 1992

78. Novak PJ, Wexler GM, Williams JS Jr, et al: Comparison of screw postfixation and free bone block interference fixation for anterior cruciateligament soft tissue grafts: Biomechanical considerations. Arthroscopy12: 470–473, 1996

79. Odensten M, Lysholm J, Gillquist J: The course of partial anterior cruci-ate ligament ruptures. Am J Sports Med 13: 183–186, 1985

80. Paulos LE, Cherf J, Rosenberg TD, et al: Anterior cruciate ligamentreconstruction with autografts. Clin Sports Med 10: 469–485, 1991

81. Pena F, Grøntvedt T, Brown GA, et al: Comparison of failure strengthbetween metallic and absorbable interference screws. Influence of in-sertion torque, tunnel-bone block gap, bone mineral density, and inter-ference. Am J Sports Med 24: 329–334, 1996

82. Pierz K, Baltz M, Fulkerson J: The effect of Kurosaka screw divergenceon the holding strength of bone-tendon-bone grafts. Am J Sports Med 23:332–335, 1995

83. Rodeo SA, Arnoczky SP, Torzilli PA, et al: Tendon-healing in a bonetunnel. A biomechanical and histological study in the dog. J Bone JointSurg 75A: 1795–1803, 1993

84. Rowden NJ, Sher D, Rogers GJ, et al: Anterior cruciate ligament graftfixation: Initial comparison of patellar tendon and semitendinosus au-tografts in young fresh cadavers. Am J Sports Med 25: 472–478, 1997

85. Sakane M, Fox RJ, Woo SL-Y, et al: In situ forces in the anterior cruciateligament and its bundles in response to anterior tibial loads. J Orthop Res15: 285–293, 1997

86. Sakane M, Woo SL-Y, Hildebrand K, et al: The contribution of theanterior cruciate ligament to knee joint kinematics: Evaluation of itsin-situ forces using a robot/universal force-moment sensor test system.J Orthop Sci 1: 335–347, 1996

87. Savarese A, Lunghi E, Budassi P, et al: Remarks on the complicationsfollowing ACL reconstruction using synthetic ligaments. Ital J OrthopTraumatol 19: 79–86, 1993

88. Schatzmann L, Brunner P, Staubli HU: Effect of cyclic preconditioning onthe tensile properties of human quadriceps tendons and patellar liga-ments. Knee Surg Sports Traumatol Arthrosc 6: S56–S61, 1998

89. Scherping SC Jr, Schmidt CC, Georgescu HI, et al: Effect of growthfactors on the proliferation of ligament fibroblasts from skeletally maturerabbits. Connect Tissue Res 36: 1–8, 1997

90. Schiavone Panni A, Fabbriciani C, Delcogliano A, et al: Bone-ligamentinteraction in patellar tendon reconstruction of the ACL. Knee SurgSports Traumatol Arthrosc 1: 4–8, 1993

91. Schreck PJ, Kitabayashi LR, Amiel D, et al: Integrin display increases inthe wounded rabbit medial collateral ligament but not the woundedanterior cruciate ligament. J Orthop Res 13: 174–183, 1995

92. Shapiro JD, Jackson DW, Aberman HM, et al: Comparison of pulloutstrength for seven- and nine-millimeter diameter interference screw sizeas used in anterior cruciate ligament reconstruction. Arthroscopy 11:596–599, 1995

93. Shelbourne KD, Gray T: Anterior cruciate ligament reconstruction withautogenous patellar tendon graft followed by accelerated rehabilitation.A two- to nine-year followup. Am J Sports Med 25: 786–795, 1997

94. Shelbourne KD, Klootwyk TE, Wilckens JH, et al: Ligament stability twoto six years after anterior cruciate ligament reconstruction with autoge-nous patellar tendon graft and participation in accelerated rehabilitationprogram. Am J Sports Med 23: 575–579, 1995

95. Shelbourne KD, Nitz P: Accelerated rehabilitation after anterior cruciateligament surgery. Am J Sports Med 18: 292–299, 1990

96. Shelbourne KD, Rettig AC, Hardin G, et al: Miniarthrotomy versus ar-throscopic-assisted anterior cruciate ligament reconstruction with autog-enous patellar tendon graft. Arthroscopy 9: 72–75, 1993

97. Shelton WR, Papendick L, Dukes AD: Autograft versus allograft anteriorcruciate ligament reconstruction. Arthroscopy 13: 446–449, 1997

98. Shino K, Inoue M, Horibe S, et al: Surface blood flow and histology ofhuman anterior cruciate ligament allografts. Arthroscopy 7: 171–176,1991

99. Shino K, Kawasaki T, Hirose H, et al: Replacement of the anteriorcruciate ligament by an allogeneic tendon graft: An experimental study inthe dog. J Bone Joint Surg 66B: 672–681, 1984

100. Shino K, Oakes BW, Horibe S, et al: Collagen fibril populations in humananterior cruciate ligament allografts. Electron microscopic analysis.Am J Sports Med 23: 203–209, 1995

101. Shoemaker SC, Markolf KL: In vivo rotatory knee stability: Ligamentousand muscular contributions. J Bone Joint Surg 64A: 208–216, 1982

102. Simonian PT, Harrison SD, Cooley VJ, et al: Assessment of morbidity ofsemitendinosus and gracilis tendon harvest for ACL reconstruction. Am JKnee Surg 10: 54–59, 1997

103. Smith CW, Young IS, Kearney JN: Mechanical properties of tendons:Changes with sterilization and preservation. J Biomech Eng 118: 56–61,1996

104. Spencer EE, Chissell HR, Spang JT, et al: Behavior of sutures used inanterior cruciate ligament reconstructive surgery. Knee Surg SportsTraumatol Arthrosc 4: 84–88, 1996

105. Spindler KP, Imro AK, Mayes CE, et al: Patellar tendon and anteriorcruciate ligament have different mitogenic responses to platelet-derivedgrowth factor and transforming growth factor b. J Orthop Res 14: 542–546, 1996

106. Thompson WO, Fu FH: The meniscus in the cruciate-deficient knee. ClinSports Med 12: 771–796, 1993

107. Tohyama H, Yasuda K: Significance of graft tension in anterior cruciateligament reconstruction. Basic background and clinical outcome. KneeSurg Sports Traumatol Arthrosc 6: S30–S37, 1998

108. Wallace MP, Howell SM, Hull ML: In vivo tensile behavior of a four-bundle hamstring graft as a replacement for the anterior cruciate liga-ment. J Orthop Res 15: 539–545, 1997

109. Wexler G, Hurwitz DE, Bush-Joseph CA, et al: Functional gait adapta-tions in patients with anterior cruciate ligament deficiency over time. ClinOrthop 348: 166–175, 1998

110. Wirth CJ, Kohn D: Revision anterior cruciate ligament surgery: Experi-ence from Germany. Clin Orthop 325: 110–115, 1996

111. Witkowski J, Yang L, Wood DJ, et al: Migration and healing of ligamentcells under inflammatory conditions. J Orthop Res 15: 269–277, 1997

112. Wolfman NM, Hattersley G, Cox K, et al: Ectopic induction of tendon andligament in rats by growth and differentiation factors 5, 6, and 7, mem-bers of the TGF-beta gene family. J Clin Invest 100: 321–330, 1997

113. Woo SL-Y, Fox RJ, Sakane M, et al: Force and force distribution in theanterior cruciate ligament and its clinical implications. Sportorthopaedie-Sporttraumatologie 13: 37–48, 1997

114. Woo SL-Y, Hollis JM, Adams DJ, et al: Tensile properties of the humanfemur-anterior cruciate ligament-tibia complex: The effects of specimenage and orientation. Am J Sports Med 19: 217–225, 1991

115. Woo SL-Y, Orlando CA, Camp JF, et al: Effects of postmortem storageby freezing on ligament tensile behavior. J Biomech 19: 399–404, 1986

116. Yamagishi T, Fujii K, Roppongi S, et al: Blood flow measurement inreconstructed anterior cruciate ligaments using laser Doppler flowmetry.Knee Surg Sports Traumatol Arthrosc 6: 160–164, 1998

117. Yasuda K, Tsujino J, Tanabe Y, et al: Effects of initial graft tension onclinical outcome after anterior cruciate ligament reconstruction: Autoge-nous doubled hamstring tendons connected in series with polyestertapes. Am J Sports Med 25: 99–106, 1997

118. Yoshiya S, Andrish JT, Manley M, et al: Graft tension in anterior cruciateligament reconstruction: An in vivo study in dogs. Am J Sports Med 15:464–470, 1987

830 Fu et al. American Journal of Sports Medicine