Graft Selection in AnteriorCruciate Ligament Reconstruction

Robin V. West, MD, and Christopher D. Harner, MD

Abstract

Advances in surgical technique andrehabilitation have resulted inmarked improvement in the outcomeof anterior cruciate ligament (ACL)reconstruction over the past 10 years.Most recent series approach a 90%success rate in terms of restoration ofknee stability, patient satisfaction, andreturn to full athletic activity.1 The op-timal graft material remains contro-versial, regardless of the graft tissueselected. The graft should have struc-tural properties similar to those of thenative ACL; these properties shouldbe present at the time of graft implan-tation and persist throughout the in-corporation period. The ideal mate-rial also should allow secure fixation,permit rapid biologic incorporation,and limit donor site morbidity.

The graft options available in-clude autograft and allograft tissue

as well as synthetic ligaments. Au-tograft options include the patellar,hamstring, and quadriceps tendons(Fig. 1, A). Allograft choices consistof the quadriceps, patellar, Achilles,hamstring, and anterior and poste-rior tibialis tendons, as well as thefascia lata (Fig. 1, B). Patellar tendonautografts have been the most pop-ular graft choice because of theirstrength characteristics, ease of har-vest, rigid fixation, bone-to-bonehealing, and favorable clinical out-comes. However, donor site morbid-ity of patellar tendon autografts hasled to the investigation and use of al-ternative graft sources.

Synthetic ligaments are classifiedas scaffolds, stents, or prostheses.The prototype is the carbon fiberscaffold ligament. The scaffold stim-ulates fibrous tissue ingrowth and

contributes to the ultimate strengthof the new ligament. A stent, such asthe Kennedy ligament augmenta-tion device, supports and protectsautograft tissues during the healingphase, when the autogenous tissueis weakest. Examples of prostheticligament substances include poly-ethylene and Gore-Tex (W. L. Gore,Flagstaff, AZ).2

In selecting a graft for ACL recon-struction, results of the preoperativeexamination, as well as patient age,activity level, and comorbidities,must be considered. Graft sources canbe compared on the basis of many cri-teria, including biomechanical prop-erties, biology of healing, ease of graftharvest, fixation strength, graft sitemorbidity, and return-to-play guide-lines (Table 1).

Dr. West is Assistant Professor, Center for SportsMedicine, University of Pittsburgh Medical Cen-ter, Pittsburgh, PA. Dr. Harner is Blue Cross ofWestern Pennsylvania Professor and Director,Center for Sports Medicine, University of Pitts-burgh Medical Center, Pittsburgh.

None of the following authors or the departmentswith which they are affiliated has received anythingof value from or owns stock in a commercial com-pany or institution related directly or indirectlyto the subject of this article: Dr. West and Dr.Harner.

Reprint requests: Dr. Harner, Center for SportsMedicine, 3200 South Water Street, Pittsburgh,PA 15203.

Copyright 2005 by the American Academy ofOrthopaedic Surgeons.

The ideal graft for use in anterior cruciate ligament reconstruction should have struc-tural and biomechanical properties similar to those of the native ligament, permitsecure fixation and rapid biologic incorporation, and limit donor site morbidity. Manyoptions have been clinically successful, but the ideal graft remains controversial.Graft choice depends on surgeon experience and preference, tissue availability, pa-tient activity level, comorbidities, prior surgery, and patient preference. Patellar ten-don autograft, the most widely used graft source, appears to be associated with anincreased incidence of anterior knee pain compared with hamstring autograft. Useof hamstring autograft is increasing. Quadriceps tendon autograft is less popularbut has shown excellent clinical results with low morbidity. Improved sterilizationtechniques have led to increased safety and availability of allograft, although allograftshave a slower rate of incorporation than do most types of autograft. No graft hasclearly been shown to provide a faster return to play. However, in general, patellartendon autografts are preferable for high-performance athletes, and hamstring au-tografts and allografts have some relative advantages for lower-demand individu-als. No current indications exist for synthetic ligaments.

J Am Acad Orthop Surg 2005;13:197-207

Vol 13, No 3, May/June 2005 197

Anatomy andBiomechanics of theNormal ACL

The normal ACL functions as the pri-mary restraint to anterior translationof the tibia and as a secondary restraintto tibial rotation and varus or valgusstress.9,10 The native ACL has an av-erage cross-sectional area of 44 mm2,an ultimate tensile load of up to 2,160N, a stiffness of 242 N/mm, and astrain rate of 20% before failure.3,4,11

The forces in the intact ACL range

from 100 N during passive knee ex-tension to approximately 400 N onwalking and up to 1,700 N with cut-ting and acceleration-deceleration ac-tivities.6,12 The ACL experiences loadsexceeding its failure capacity only withunusual loading patterns on the knee.

Biomechanical Propertiesof ACL Grafts

The biomechanical properties of thevarious graft materials have been stud-

ied extensively. In one early study,Noyes et al3 subjected human ligamentgraft tissues of varying sizes to high-strain-rate failure tests. To assess therelative strength of these grafts, theresults of the failure tests were com-pared with the mechanical propertiesof normal ACLs in young adults. Themean ultimate tensile strength andmean stiffness of the normal ACLswere 1,725 N and 182 N/mm, respec-tively. The bone–patellar tendon–bonegraft (14 mm) had 168% the tensilestrength and almost four times the

Figure 1 A, Autografts. Hamstring (top) and patellar (bottom) tendons. B, Allografts. Patellar (top) and Achilles (bottom) tendons.

Table 1Comparison of Anterior Cruciate Ligament Graft Types

Biomechanical Property

GraftTensile

load (N)Stiffness(N/mm)

BiologicIncorporation

Method ofFixation

Graft SiteMorbidity

Outcomes/Return to

Play (months)

Patellar tendonautograft3,4

2,977 620 Bone-to-bonehealing (6 wks)

Interferencescrew

Anterior kneepain; larger

incision

4-6

Quadruplesemitendinosus/gracilis5

4,090 776 Soft-tissue healing(8-12 wks)

Variable Hamstringweakness

Increasedlaxity/6

Patellar tendonallograft6

Similar topatellar tendon

autograft

Similar topatellar tendon

autograft

Bone-to-bonehealing, slowincorporation

(>6 mos)

Interferencescrew

None >6

Quadricepstendon7,8

2,352 463 Bone-to-boneand soft-tissue

(6-12 wks)

Variable Similar topatellar tendon

autograft

Limited data

Graft Selection in Anterior Cruciate Ligament Reconstruction

198 Journal of the American Academy of Orthopaedic Surgeons

stiffness of the normal ACL. The semi-tendinosus and gracilis tendons pro-vided 70% and 49% of the reportedstrength of the normal ACL, respec-tively. The patellar tendon graft wasthe only one of those tested (of thegracilis tendon, semitendinosus ten-don, quadriceps tendon, fascia lata,iliotibial band) that was stronger thanthe normal ACL. Because the mechan-ical properties of the grafts are alteredduring incorporation into the hostbone, these results represent thestrength only at the time of initial im-plantation. Currently, the results of thegrafts tested have little clinical signif-icancebecausesingle-bundlehamstringand 14-mm patellar tendon grafts arerarely used. Smaller patellar tendongrafts (<12 mm) and quadrupled ham-string grafts are more commonly usedin ACL reconstruction.

In that initial study, Noyes et al3

tested graft strength without address-ing the anatomic position of the nor-mal ACL or the direction of tensionon the tested fibers. Both the knee flex-ion angle and the direction of the ap-plied tensile load affect the strengthof the femur–ACL–tibia complex. Wooet al4 evaluated the biomechanicalproperties of the femur–ACL–tibiacomplex in specimens of various agesto determine the effects of donor ageand the direction of applied load. Theyreported a substantially higher stiff-ness (242 N/mm) and ultimate loadto failure (2,160 N) of the ACL thanhad previously been reported. The val-ue for ultimate load to failure was 30%higher when tested in the anatomicorientation of the ACL, and there wasa notable age-related decrease in thestructural properties of the ACL.

Staubli et al7 studied the mechan-ical properties of 16 10-mm wide full-thickness patellar tendon and quad-riceps tendon grafts from eight pairedmale donors. They found a Young’smodulus of 200 ± 47 N/mm2 forquadriceps tendons and 363 ± 94N/mm2 for patellar tendons whentested at 200 N (P < 0.024). Young’smodulus was 304 ± 70 N/mm2 for

quadriceps tendons and 459 ± 83N/mm2 for patellar ligaments whentested at 800 N (P = 0.016). In theirstudy of the biomechanical propertiesof quadriceps and patellar tendongrafts, Harris et al8 reported a meanload to ultimate failure of 1,075 N and876 N, respectively.

Comparing studies of the biome-chanical properties of ACL grafts isdifficult because results can varymarkedly depending on the age of thedonor, size of the graft, and methodsof testing (Table 2). Additionally,clamp design can result in slippageof the graft during testing, with sub-sequent errors in elongation measure-ments. Alternatively, the clamp cancrush the graft, resulting in prematurefailure and lower strength values.

Historically, the high-dose radia-tion used for sterilization of allograftsresulted in weakened structural prop-erties of the graft tissue. The alterna-tive use of ethylene oxide sterilizationresulted in adverse surgical reactions,most commonly chronic effusions.Ethylene oxide does not alter the me-chanical properties of the graft andcan effectively remove microorgan-isms. However, the chemical residuethat ethylene leaves behind can causechronic synovitis and subsequentgraft failure.13

The current sterilization techniquesare cryopreservation and gamma ra-diation. Cryopreservation has beenshown to have no effect on the struc-tural properties of ligaments, tendons,or meniscal tissue. Gamma radiationis an effective method of sterilization,

but doses >3.0 Mrad, which are nec-essary to kill viruses, have detrimen-tal effects on graft strength. Therefore,aseptic harvest and a cleaning processconsisting of antibiotic soaks, multi-ple cultures, and low-dose radiation(<3.0 Mrad) is the most commonlyused method for producing a sterileACL graft.14 Low-temperature chem-ical sterilization methods with goodtissue penetration seem to be spori-cidal and do not adversely affect thebiomechanical properties of tissue.Other sterilization techniques, such assupercritical CO2 and the use of an-tioxidants in combination with gam-ma irradiation, are being developed.15

The Biology of Healing

All ACL grafts undergo a sequentialprocess of incorporation into the hostknee.16-19 The first phase of incorpo-ration consists of an inflammatory re-sponse, during which the graft un-dergoes degeneration. The donorfibroblasts undergo cell death, andthe remaining tissue serves as a scaf-fold for host cell migration and ma-trix production. The second phaseconsists of a period of revasculariza-tion, with migration of host fibro-blasts into the graft tissue. This phasebegins within 20 days of implantationand usually is complete 3 to 6 monthsafter surgery.16 The material proper-ties of the graft change as the revas-cularization (ligamentization) processoccurs. During graft maturation, graftstrength drops to as low as 11% of that

Table 2Biomechanical Properties of Selected ACL Graft Tissues

Tissue

UltimateTensile

Load (N)Stiffness(N/mm)

Cross-sectionalArea (mm2)

Intact anterior cruciate ligament3 2,160 242 44Bone–patellar tendon–bone (10 mm)6 2,977 620 35Quadruple hamstring5 4,090 776 53Quadriceps tendon (10 mm)7,8 2,352 463 62

Robin V. West, MD, and Christopher D. Harner, MD

Vol 13, No 3, May/June 2005 199

of the normal ACL and stiffness to aslow as 13%.17 The final phase consistsof graft healing, with remodeling ofthe collagen structure into a more or-ganized pattern. As the graft heals, itsmechanical properties improve, butit never reaches the stiffness andstrength of the graft material at thetime of implantation.17

Healing of the graft attachment sitemay be responsible for most of thegraft strength after transplantation.From a biologic standpoint, patellartendon grafts have the advantage ofbone-to-bone healing compared withsoft-tissue grafts. Bone-to-bone heal-ing is similar to fracture healing, andit is stronger and faster than soft-tissue healing. With bone-to-bonehealing, the graft can be healed to thehost bone within 6 weeks. Soft-tissuegrafts usually incorporate into thehost bone within 8 to 12 weeks.18

Autografts and allografts undergoa similar process of incorporation, in-cluding graft necrosis, cellular repopu-lation, revascularization, and collagenremodeling. However, allografts havea slower rate of biologic incorpora-tion. Jackson et al19 compared the his-tologic and microvascular status of pa-tellar tendon autografts and allograftsin a goat model. At 6 weeks, the al-lografts demonstrated increased vas-cularity and inflammatory response.The autografts had a notably greatercross-sectional area compared with theallografts at 6 weeks. The allograftshad smaller cross-sectional areas, per-sistence of more original large collagenfibers, and fewer small-diameter col-lagen fibers. The autografts had a ro-bust response and proliferation ofsmall-diameter fibers. At 6 months,the allografts had a persistence of largecollagen fibers, indicating a delay inremodeling. Mechanical testing of theallograft and autograft groups showeda statistically significant (P < 0.01) dif-ference in anteroposterior translationat 6 months. The autograft displace-ment was 3.4 ± 0.5 mm, comparedwith 5.3 ± 1.0 mm in the allograftgroup (P < 0.01). The patellar tendon

autograft demonstrated a more robustbiologic response, improved stabili-ty, and increased strength-to-failurevalues. The authors suggested a long-er period of protection for patientswith allograft ACL reconstructionsthan for those with autograft recon-structions.

Nikolaou et al20 compared thestrength, histology, and revasculariza-tion of patellar tendon allografts andautografts. The mechanical integrityof the allografts was similar to thatof the autografts; both achieved near-ly 90% of control ligament strengthby 36 weeks. Revascularization ap-proached normal by 24 weeks in boththe autograft and allograft.

Graft Harvesting

The goal of graft harvest should beto obtain an adequate graft specimenwhile minimizing donor site mor-bidity. The ease of graft harvest issurgeon-dependent. Each graft rep-resents a unique set of technical chal-lenges and potential pitfalls thatshould be thoroughly understood be-fore harvest.

Patellar tendon autografts requireharvesting of a tibial tubercle and pa-tellar bone plug. Large, deep cuts canincrease the risk for stress fracture inthe proximal tibia and patella, as wellas endanger the patellar articular car-tilage. Trapezoidal bone cuts, insteadof triangular ones, have been de-scribed that reduce the risk of artic-ular cartilage penetration.

The quadriceps tendon is more dif-ficult to harvest than the patellar ten-don because of its denser corticalbone, curved proximal surface, andclose adherence to the suprapatellarpouch. Fulkerson and colleagues21,22

have described a technique to harvestthe quadriceps tendon safely and ef-ficiently. Through a short midline in-cision, starting mid-patella and ex-tending proximally, a 10-mm wideand 20-mm long bone plug is harvest-ed from the proximal patella. The

bone plug should be 6 to 8 mm thick,leaving a tendon graft approximate-ly 6 mm thick. The tendon is then har-vested about 7 cm proximally, takingcare to avoid entering the suprapa-tellar pouch. A drill hole is then madein the bone plug, and a no. 5 sutureis passed through the plug.

Hamstring tendon harvest requireselevating the sartorius to expose theunderlying semitendinosus and grac-ilis tendons. The tendons can be ei-ther left on their insertion or detachedfrom the tibia during harvest.Aclosedor open tendon stripper then is usedto retrieve the tendons. Care shouldbe taken to harvest the entire tendonand not amputate it prematurely.

Mechanics of InitialFixation

Graft fixation is crucial in ACL recon-struction and is the weakest link inthe initial 6- to 12-week period, dur-ing which healing of the graft to thehost bone occurs. The graft must beable to withstand early rehabilitation,which can consist of forces as high as450 to 500 N.11 If fixation is poor, thegraft may slip or the fixation may failaltogether, resulting in an unstableknee. Fixation failure usually occurson the tibial side.23

It is difficult to compare fixationstudies because of the wide variationin fixation and testing methods (Ta-ble 3). Interference screw fixation forbone block patellar tendon grafts hasdemonstrated fixation strength andstiffness superior to that of alterna-tive fixation methods. The averagefailure strength of bone block fixationwith an interference screw is 423 to558 N.26 Factors that may affect fix-ation strength of interference screwsinclude screw diameter and screw di-vergence from the bone block. Screwdiameter, usually 7 or 9 mm, has beenshown to have a negative effect on thepull-out strength only when the gapbetween the bone block and the tun-nel wall exceeds 2 mm.27 Placement

Graft Selection in Anterior Cruciate Ligament Reconstruction

200 Journal of the American Academy of Orthopaedic Surgeons

of interference screws with a diver-gence angle >30° from the bone blockresults in failure at lower loads com-pared with screw placement parallelto the bone block.28

Fixation strength of soft-tissuegrafts varies depending on the choiceof fixation; different methods are avail-able for the tibial and femoral sides.Commonly used tibial fixation devic-es include suture posts, staples, screw-and-washer constructs, and interfer-ence screws (Fig. 2, A). Femoralfixation devices are similar to thosefor the tibial side but also includecross-pins and buttons (Fig. 2, B).

Kousa et al24,25 reviewed the fixa-tion strength of six hamstring tendongraft-fixation devices for the femoraland the tibial sides during ACL recon-struction. On the femoral side, the BoneMulch Screw offered superior fixation(P < 0.001) compared with the otherdevices in the single-cycle load-to-failure test (Table 3). The yield loadof the Bone Mulch Screw was signif-icantly greater (P < 0.001) than that

of the BioScrew and the RCI Screw.The stiffness of the Bone Mulch Screwfixation was significantly (P < 0.001)greater than the stiffness of the En-doButton, RigidFix, BioScrew, and theRCI screw. The stiffness of the BoneMulch Screw was not significantlygreater than that of the SmartScrewACL.24

On the tibial side, the Intrafix wasthe strongest (P < 0.01) in the single-cycle load-to-failure test (Table 3). TheIntrafix also had a significantly great-er stiffness (P < 0.01) than the Wash-erLoc, the tandem spiked washer, theSmartScrew, the BioScrew, and theSoftSilk.

Graft Morbidity

Graft morbidity is caused by associ-ated symptoms (eg, anterior kneepain, quadriceps and hamstringweakness), donor site and syntheticgraft complications, strength deficits,stiffness, and infection.

Anterior Knee PainAnterior knee pain is common af-

ter ACL reconstruction, with symp-toms occurring anywhere along theextensor mechanism (ie, patellar orquadriceps tendons, patellofemoraljoint, peripatellar soft tissues). Thepain may be related to the choice ofgraft material; most studies show atendency for decreased pain with theuse of hamstring autografts comparedwith patellar tendon autografts.1,29-34

However, there is no difference in theincidence of anterior knee pain be-tween patients with patellar tendonautografts and allografts.35 Some au-thors have suggested that anteriorknee pain is related more to loss ofmotion and poor rehabilitation tech-niques than to graft choice. Sachs etal36 demonstrated a correlation be-tween the development of patello-femoral symptoms and the presenceof a flexion contracture and quadri-ceps weakness. Shelbourne and Nitz37

noted a decrease in patellofemoralsymptoms after an accelerated reha-bilitation protocol; they attributed thedecrease in symptoms to the early res-toration of range of motion (ROM)and quadriceps strength. In a meta-analysis of 21 studies using patellartendon autografts and 13 studies us-ing hamstring grafts, the incidence ofanterior knee pain was 17.4% in thepatellar tendon group versus 11.5%in the hamstring group.1

No specific studies compare ante-rior knee symptoms between quad-riceps tendon, patellar tendon, andhamstring tendon autografts. How-ever, Chen et al38 reported only mildharvest site tenderness in 12 patientsat an average of 18 months after ACLreconstruction with quadriceps au-tograft. Fulkerson and Langeland21

reported no early quadriceps morbid-ity in their series of 28 patients.

Quadriceps and HamstringTendon Strength

Persistent quadriceps and ham-string weakness is another problemafter graft harvest. Rosenberg et al39

Table 3Failure Strength of Various Techniques of Graft Fixation

Fixation

UltimateFailure Load

(N)Stiffness(N/m)

Patellar TendonMetal interference screw19 558 —Bioabsorbable interference screw19 552 —

Soft Tissue (Femoral)Bone Mulch Screw (Arthrotek, Warsaw, IN)24 1,112 115EndoButton (Smith & Nephew Endoscopy,

Andover, MA)241,086 79

RigidFix (Ethicon, Somerville, NJ)24 868 77SmartScrew ACL (Linvatec, Largo, FL)24 794 96BioScrew (Linvatec)24 589 66RCI Screw (Smith & Nephew Endoscopy)24 546 68

Soft Tissue (Tibial)Intrafix (Ethicon)25 1,332 223WasherLoc (Arthrotek)25 975 87Tandem spiked washer25 769 69SmartScrew ACL25 665 115BioScrew25 612 91SoftSilk (Acufex Microsurgical, Mansfield, MS)25 471 61

Robin V. West, MD, and Christopher D. Harner, MD

Vol 13, No 3, May/June 2005 201

evaluated extensor mechanism func-tion in 10 randomly selected patients12 to 24 months after ACL reconstruc-tion using the central third of the pa-tellar tendon. Although all patientswere satisfied with their results andconsidered their knees to be stable,many had subjective complaints ofanterior knee pain and weakness.Only 3 of the 10 patients returned toparticipation in all of their preinjurysports. Isokinetic testing at 60°/secangular velocity showed an averagequadriceps deficit of 18% comparedwith the normal extremity.Axial com-puted tomography scans revealed amarked decrease in the quadricepscross-sectional area.

Lephart et al40 compared objectivemeasurements of quadriceps strengthand functional capacity in athletesafter patellar tendon autograft andallograft ACL reconstruction. Theyreported no significant differencebetween the autograft and allograftgroups with regard to thigh circum-ference, quadriceps strength andpower, and functional performancetests. The quadriceps index, which iscalculated by comparing the involved

leg with the uninvolved leg at 60°/sec angular velocity, indicated simi-lar results between the allografts andautografts, with the indexes averag-ing 90% to 95%. The findings suggestthat harvesting the central third of thepatellar tendon does not diminishquadriceps strength or functional ca-pacity in highly active patients whoundergo intense rehabilitation.

Carter and Edinger41 comparedhamstring and quadriceps isokinetictest results 6 months postoperativelyin 106 randomly selected patientswho had ACL reconstruction with ei-ther patellar tendon or hamstring au-tografts. No statistically significantdifferences were found between thedifferent graft sources with regard toknee extension or flexion strength.However, at 6 months postoperative-ly, most patients had not achieved ad-equate strength to safely partake inunlimited activities.

Nakamura et al42 evaluated ham-string strength at 2 years postopera-tively in 74 consecutive patients whohad had hamstring ACL reconstruc-tion. Similar to the results of otherstudies, recovery of peak flexion

torque was more than 90%. Howev-er, the recovery was less at 90° kneeflexion. These results suggest that lossof hamstring strength after harvest-ing may be more prominent at deepflexion angles.

Therearenostudiesevaluatingham-string or quadriceps tendon strengthrecovery after reconstruction with aquadriceps tendon autograft. Becausequadriceps harvest is similar to pa-tellar tendon harvest in regard to ex-tensor mechanism disruption, simi-lar strength testing results have beeninferred.

Donor Site ComplicationsAlthough infrequent, reported com-

plications with patellar tendon au-tograft include patellar fractures,43 pa-tellar tendon ruptures,44 tendinitis,localized tenderness, and numbness.Numbness also can occur with ham-string autograft as a result of injuryto the superficial branch of the saphe-nous nerve.45 The infrapatellar branchof the saphenous nerve passes on theinferior medial aspect of the knee;therefore, the risk of nerve injurywould seem to be low after quadri-

Figure 2 A, Tibial side hamstring fixation devices. A = WasherLoc, B = spiked washer, C = Intrafix, D = BioScrew, E = SoftSilk, F = Smart-Screw. B, Femoral side hamstring fixation devices. A = EndoButton, B = Bone Mulch Screw, C = RigidFix, D = Bioscrew, E = RCI Screw, F= SmartScrew. (Panel A reproduced with permission from Kousa P, Järvinen TL, Vihavainen M, Kannus P, Järvinen M: The fixation strengthof six hamstring tendon graft fixation devices in anterior cruciate ligament reconstruction: II. Tibial site. Am J Sports Med 2003;31:182-188.Panel B reproduced with permission from Kousa P, Järvinen TL, Vihavainen M, Konnus P, Järvinen M: The fixation strength of six hamstringtendon fixation devices in anterior cruciate ligament reconstruction: I. Femoral revision. Am J Sports Med 2003;31:174-181.)

Graft Selection in Anterior Cruciate Ligament Reconstruction

202 Journal of the American Academy of Orthopaedic Surgeons

ceps tendon harvest. The risk of pa-tellar fracture also should be low be-cause the proximal patellar bone ismore dense.

A serious potential problem withthe use of allografts is disease trans-mission, although the problem haslargely been eliminated with the de-velopment of better donor screeningand testing procedures. Guidelines is-sued by the American Association ofTissue Banks have been revised andupdated six times since they were in-stituted in 1986 to ensure sterility andquality during allograft processing. Adetailed medical, social, and sexualhistory must be obtained for each po-tential cadaveric donor. Extensive test-ing includes blood cultures, harvest-ed tissue cultures, and screening forantibodies to human immunodefi-ciency virus (HIV)-1 and HIV-2, hep-atitis B surface antigen, hepatitis C,syphilis, and human T-cell lympho-tropic virus.14

To decrease the window of vulner-ability between the infection and theproduction of antibodies by the do-nor, more than half of tissue banksnow use polymerase chain reactiontesting to directly detect viral antigens.Polymerase chain reaction testing issensitive enough to detect as few as5 to 20 viral DNA copies of HIV persample tested.14 To date, there havebeen three reported cases of diseasetransmission from bone–patellar ten-don–bone allografts used to recon-struct the ACL.14 The first case of HIVwas reported in 1985; the other twowere cases of hepatitis C, reported in1991.46

Synthetic GraftsSynthetic ligaments have a higher

complication rate than do allograftand autograft reconstructions. Car-bon fiber scaffolds have had very lim-ited success. They have been associ-ated with a high rate of synovitis, andan in vitro study in pigs found thatingrowth into the scaffold does notoccur.2 Also, the carbon fibers do notadhere to the bony channels, and the

peak tensile strength of the ligamentsafter reconstruction with a carbonscaffold was one third the strength ofa normal ACL at 4 months postoper-atively.2,47

Prosthetic implants also have hadvery limited success. They have beenassociated with recurrent instability,chronic effusions, and synovitis. TheKennedy ligament-augmentation de-vice (LAD) was devised to protect thebiologic graft during the early phas-es of weakness and degeneration. TheLAD goes through a revasculariza-tion process similar to that of biolog-ic grafts, but the collagen producedis abnormal. The connective tissuehas an insufficient exposure to tensileforces, which causes failure of the col-lagen alignment. LAD–related com-plications range from 0% to 63% inthe literature.2 Effusions and reactivesynovitis are the most commonly re-ported complications. The LAD alsomay have a higher association withinfections.2,48

Clinical Outcome byGraft Type

Freedman et al1 performed a meta-analysis of 21 studies (1,348 patients)from 1966 to 2000 (minimum follow-up, 24 months) and reported a signif-icantly (P < 0.001) lower rate of fail-ure in the patellar tendon autograftgroup than in the hamstring tendonautograft group (1.9% versus 4.9%).Laxity was measured clinically withthe KT-1000 arthrometer and pivot-shift testing. Asignificantly (P = 0.017)higher proportion of the patellar ten-don group (79%) had a side-to-sidedifference of <3 mm compared withthe hamstring group (73.8%). The pa-tellar tendon group had a significant-ly (P = 0.009) higher rate of subjectsrequiring lysis of adhesions than didthe hamstring group (6.3% versus3.3%). The patellar group also dem-onstrated a higher rate (P = 0.007) ofanterior knee pain (17.4% versus11.5%). The hamstring group had a

higher rate of hardware removal af-ter reconstruction (5.5% versus 3.1%)(P = 0.017). Infection rates were notsignificantly different between the twogroups (0.5% in the patellar group ver-sus 0.4% in the hamstring group).

Colombet et al49 reported good clin-ical results in a series of 200 patientswho underwent ACL reconstructionwith a four-strand hamstring autograftand metal interference screw fixation.Eighty-six percent of the patients re-turned to high-performance athleticsat a national or international compe-tition level.

Revision Surgery

RevisionACLreconstruction is essen-tially a salvage procedure; the causeof failure of the initial reconstruction(eg, trauma, poor surgical technique,fixation failure) must be established.Also, it is critical to determine the de-sires and aspirations of the patient.Meticulous preoperative planning isessential, including weight-bearingradiographs and a thorough knee ex-amination to determine associated in-stabilities. Possible available graft do-nor sites include the ipsilateral orcontralateral hamstrings, patellar ten-don, quadriceps tendon, fascia lata,and allograft.

Noyes and Barber-Westin50 pro-spectively studied the outcomes ofpatellar tendon allografts and au-tografts used for revision ACL sur-gery in 65 patients (mean follow-up,42 months). Notable improvement insymptoms, function, anteroposteriordisplacement, and overall ratings wasnoted in all patients. Although notstatistically significant, the KT-2000results showed <3 mm increased dis-placement in 53% of the allograftgroup and in 67% of the autograftgroup. Overall failure rates were 33%for the allografts and 27% for the au-tografts.50

Kartus et al51 compared 12 patientswho underwent revision ACL sur-gery with a reharvested ipsilateral pa-

Robin V. West, MD, and Christopher D. Harner, MD

Vol 13, No 3, May/June 2005 203

tellar tendon autograft with 12 pa-tients who underwent the revisionreconstruction with a contralateralpatellar tendon graft. PostoperativeKT-1000 arthrometer measurementsshowed no significant differences be-tween the two groups. However, theLysholm scores were significantly (P= 0.002) higher in the patients withthe contralateral patellar tendon graft.Functional outcomes and the Tegneractivity levels were similar for bothgroups. Two major complications oc-curred—one patellar fracture and onepatellar tendon rupture—both in theipsilateral patellar tendon harvestgroup.

Uribe et al52 reported the results ofrevision ACL reconstruction in 54 pa-tients with use of ipsilateral patellartendon autograft (31%), contralater-al patellar tendon autograft (30%), pa-tellar tendon allograft (35%), andhamstring autograft (4%). At a meanfollow-up of 32 months, autogenousgrafts provided greater objective sta-bility compared with allografts, withaverage KT-1000 readings of 2.2 and3.3, respectively. There were no func-tional differences between the groups,and harvesting the contralateral ten-don had no adverse long-term effects.The subjective results were marked-ly worse depending on the degree ofarticular cartilage degeneration. Only54% of the patients returned to theirpre-ACL injury activity level.

Reharvesting the ipsilateral patel-lar tendon has been recommended.Woods et al53 reported on long-termresults in 10 patients who underwent13 revision ACL reconstructions us-ing the lateral third of the ipsilateralpatellar tendon as a graft. All of theprimary reconstructions had used thecentral third of the ipsilateral patel-lar tendon as autograft. At a meanfollow-up of 43 months, the averageKT-1000 difference between kneeswas 2.4 mm. All patients had a neg-ative pivot shift, and the mean bilat-eral comparison ratios of isokineticstrength testing were 84% for exten-sion and 96% for flexion. All patients

returned to their previous work lev-els, and 80% returned to moderatesports activities.

Colosimo et al54 reviewed 13 pa-tients who underwent revision ACLreconstruction with the reharvestedipsilateral patellar tendon. At an av-erage follow-up of 39.4 months, 11 pa-tients had good or excellent resultsand 2 had fair results. PostoperativeKT-1000 testing showed an averageside-to-side difference of 1.92 mm.The average extension and flexiondeficits were 10% and 1.7%, respec-tively. Only one patient reportedmoderate patellofemoral problems.

Several studies have addressed thebiomechanical and histologic proper-ties of the patellar tendon after rehar-vest. In their evaluation of the harvestsite in 14 patients, Nixon et al55 foundthat the central third of the patellartendon reconstituted itself and wasnearly identical to normal tendon tis-sue by 24 months postoperatively.Burks et al56 examined the biomechani-cal and histologic changes in remain-ing tissue following a central third pa-tellar tendon harvest. They found amarked increase in tendon cross-sectional area at 3 months, with a fur-ther increase at 6 months. However,in a canine study, LaPrade et al57 dem-onstrated that the average load to fail-ure of reharvested patellar tendon was54% that of contralateral controls af-ter 1 year. Proctor et al58 reported thatthe maximum force to failure and ul-timate stress of reharvested tendonsin a goat model had decreased by 51%and 65%, respectively, compared withthe contralateral side.

Return to Play

Functional testing, clinical evaluation,and subjective assessment should beused to determine when a patientmay return to full activity. CompleteROM is needed prior to return tosports; regaining less than completeROM places the extremity at a me-chanical disadvantage and increases

the risk for reinjury. Muscle strengthand balance must be achieved to pro-vide the required dynamic stability.Determination of successful return toplay is subjective. It may be based onan earlier return to play, level of com-petition, or stability and strength ofthe knee. Most data, which aresurgeon- and technique-dependent,support the use of patellar tendon au-tograft for a quicker return to play.

The bone-to-bone healing of patel-lar tendon autografts offers the quick-est incorporation into the host bone.Soft-tissue grafts, such as hamstringautografts, can take up to 6 weekslonger to incorporate than patellartendon autografts. Allografts have alonger biologic incorporation—some-times up to 1 year for complete inte-gration.18,19 However, return to playis based not only on graft healing butalso on strength, agility, and stabilityof the knee.

O’Neill59 performed a prospectiverandomized analysis of 125 patientswho underwent one of three auto-graft ACL reconstruction techniques:hamstring autograft, two-incision pa-tellar tendon autograft, or single-incision patellar tendon autograft. Re-turn to preinjury level of activity was88% in the hamstring group, 95% inthe two-incision group, and 89% inthe single-incision group. In anotherprospective randomized compari-son,31 there was a notable reductionin activities after ACL reconstructionwith both the patellar tendon andhamstring autografts. Activity reduc-tion was 45% in the patellar groupand 37% in the hamstring group.However, many patients reported adecrease in activity because they wereno longer participating in high schoolor college sports.

Bach et al60 retrospectively studiedthe results of 97 patients who under-went ACL reconstruction with a two-incision patellar tendon autografttechnique. Fifty-three patients (55%)returned to their preinjury sports lev-els; 18 (19%) indicated a decrease inactivity level because of the knee.

Graft Selection in Anterior Cruciate Ligament Reconstruction

204 Journal of the American Academy of Orthopaedic Surgeons

Fourteen patients (14%) reported adecrease in activity unrelated to theknee, and four (4%) stopped all sportsfor problems unrelated to the knee.Three patients (3%) stopped all sportsbecause of knee problems.

Yunes et al61 compared patellartendon and hamstring autografts intheir meta-analysis of four studieswith at least a 2-year follow-up. Thepatellar tendon group had signifi-cantly (P = 0.01) better results: a 75%return to preinjury activity level com-pared with a 64% return with thehamstring reconstruction. No real dif-ferences could be demonstrated be-tween the types of autografts.

Shino et al62 reported that 79 of 84patients who had ACL reconstructionwith soft-tissue allografts returned totheir desired activity levels. No directcomparisons have been done withother types of grafts.

For patients whose priority is aquick return to sports, Shelbourneand Urch63 recommend ACL recon-struction with a contralateral patel-lar tendon autograft. They reviewedpatients who had had primary ACLreconstruction with either the con-tralateral (434 patients) or ipsilateral(228 patients) patellar tendon to de-termine the difference between thegroups regarding ROM, quadricepsstrength, and return to sports. The pa-tients were divided into three sub-groups. The competitive subgroupconsisted of 260 athletes who compet-ed in twisting, pivoting, or jumpingsports at either a high school, colle-giate, or professional level. The rec-reational subgroup consisted of 172patients older than age 26 years whocompeted in similar activities at a rec-reational level. The third subgroupconsisted of 230 patients who were ei-ther participating in sports that didnot involve jumping or pivoting, orwho did not meet the age require-ment for the recreational subgroup.The average times to return to fullpreinjury sport participation for thecontralateral and ipsilateral groupswere 4.9 and 6.1 months, respec-

tively. The competitive subgroup re-turned to full preinjury sports ac-tivity at an average of 4.1 months(contralateral group) and 5.5 months(ipsilateral group).

Return-to-play results after revisionACL surgery are much more variable.Only a few studies have been pub-lished on outcomes following revisionsurgery, and many of them do not doc-ument return-to-play results. Eber-hardt et al64 reviewed 44 patients whounderwent a revisionACLreconstruc-tion with patellar tendon autograft (av-erage follow-up, 41 months). Sixteenpatients (36%) did not reach their pre-injury activity level; 4 were unable toreturn to play, and 10 elected not to.In one study of 54 patients who un-derwent revision ACL reconstruction,only 46% returned to their preinjuryactivity level.52

Many of the reported ACL seriesdo not address preinjury or postin-jury activity level. After ACL recon-struction, many patients do not returnto their preinjury levels because of avariety of factors, such as lifestylechanges (out of school, no longer com-peting), fear of recurrent knee injury,instability, and pain. Return-to-playcriteria should incorporate factors suchas agility, strength, clinical stability,and ROM. These criteria are difficultto correlate with a particular graft typeor surgical technique.

Summary

The patellar tendon autograft has thelargest number of reported outcomesin the literature and is the most wide-ly used graft source.45 Patellar tendonautografts may have some relativeadvantages for high-demand patientswho participate in cutting, pivoting,or jumping sports or who desire aquick return to play. Preexisting an-terior knee pain and certain lifestyleactivities (eg, kneeling for work orreligion) are relative contraindica-tions to the patellar tendon autograft.Quadriceps tendon autografts are less

commonly used but have been re-ported to have excellent results witha low rate of morbidity.

Hamstring grafts are increasingin popularity, primarily because oftheir excellent stiffness and tensileload properties, improved fixationtechniques, reduced harvest mor-bidity, and excellent outcome andpatient satisfaction scores. However,there is a higher reported degree ofinstrumented (KT-1000) tested laxityfor hamstring reconstruction and alower return to preinjury activity lev-els.29-32,49 Quadruple hamstring au-tograft is preferable in lower-demandpatients, recreational athletes, youn-ger patients with open growth plates,and patients who are concerned aboutcosmesis. Contraindications to ham-string autograft include generalizedincreased ligamentous laxity, compet-itive sprinters (terminal flexion weak-ness), or previous hamstring injury.

Allografts have had a recent resur-gence. Improved sterilization tech-niques, along with a wide range ofgraft sources (eg, tibialis, Achilles, andpatellar tendons), have led to increasedsafety and availability. However, thebenefits of decreased surgical morbid-ity and easier rehabilitation must beweighed against the higher costs ofthe allografts and the slower periodof incorporation.19 Patellar tendonallografts have some advantage inlower-demand patients, in older pa-tients who prefer an easier rehabili-tation, and for patients with knees thathave multiple injured ligaments.

There is no evidence that routineuse of synthetic ligaments shouldbe advocated. Synthetic grafts areassociated with a higher risk of com-plications, including recurrent in-stability, effusions, and, possibly, in-fections.2,47,48

Graft selection for revisionACLre-construction depends on the etiolo-gy of the failure and on patient pref-erence. Patellar tendon allografts maybe used in patients who have failedautograft (particularly patellar ten-don) and in complex revisions with

Robin V. West, MD, and Christopher D. Harner, MD

Vol 13, No 3, May/June 2005 205

secondary restraint failures (meniscalor collateral injury). Autografts alsoare optimal for revision when the pri-mary etiology of failure is failure ofan allograft.

The ideal graft forACLreconstruc-tion should have biomechanical prop-

erties similar to those of the nativeACL, enable stable initial fixation andrapid biologic incorporation, and of-fer a low rate of morbidity. Under-standing the technical challenges andpotential pitfalls involved with eachgraft is important. The morbidity of

the harvest is surgeon- and technique-dependent. Graft selection dependson many factors, including the sur-geon’s philosophy and experience,tissue availability, patient activity lev-el, comorbidities, prior surgery, andpatient preference.

References1. Freedman KB, D’Amato MJ, Nedeff

DD, Kaz A, Bach BR Jr: Arthroscopicanterior cruciate ligament reconstruc-tion: A metaanalysis comparing patel-lar tendon and hamstring tendon au-tografts. Am J Sports Med 2003;31:2-11.

2. Zoltan DJ, Reinecke C, Indelicato PA:Synthetic and allograft anterior cruciateligament reconstruction. Clin SportsMed 1988;7:773-784.

3. Noyes FR, Butler DL, Grood ES, ZernickeRF, Hefzy MS: Biomechanical analysisof human ligament grafts used in knee-ligament repairs and reconstructions.J Bone Joint Surg Am 1984;66:344-352.

4. Woo SL-Y, Hollis JM, Adams DJ, LyonRM, Takai S: Tensile properties of thehuman femur-anterior cruciate liga-ment-tibia complex: The effects of spec-imen age and orientation. Am J SportsMed 1991;19:217-225.

5. Hamner DL, Brown CH Jr, Steiner ME,Hecker AT, Hayes WC: Hamstring ten-don grafts for reconstruction of the an-terior cruciate ligament: Biomechanicalevaluation of the use of multiplestrands and tensioning techniques.J Bone Joint Surg Am 1999;81:549-557.

6. Markholf KL, Burchfield DM, ShapiroMM, Cha CW, Finerman GAM, Slaut-erbeck JL: Biomechanical consequencesof replacement of the anterior cruciateligament with a patellar ligament al-lograft: II. Forces in the graft comparedwith forces in the intact ligament. J BoneJoint Surg Am 1996;78:1728-1734.

7. Staubli HU, Schatzmann L, Brunner P,Rincon L, Nolte LP: Mechanical tensileproperties of the quadriceps tendonand patellar ligament in young adults.Am J Sports Med 1999;27:27-34.

8. Harris NL, Smith DAB, Lamoreauz L,Purnell M: Central quadriceps tendon foranterior cruciate ligament reconstruction.Am J Sports Med 1997;25:23-28.

9. Butler DL, Noyes FR, Grood ES: Liga-mentous restraints to anterior-posteriordrawer in the human knee. J Bone JointSurg Am 1980;62:259-270.

10. Markolf KL, Mensch JS, Amstutz HC:Stiffness and laxity of the knee: The

contributions of the supporting struc-tures. A quantitative in vitro study.J Bone Joint Surg Am 1976;58:583-593.

11. Frank CB, Jackson DW: The science ofreconstruction of the anterior cruciateligament. J Bone Joint Surg Am 1997;79:1556-1576.

12. Butler DL, Grood ES, Noyes FR, SoddAN: On the interpretation of our ante-rior cruciate ligament data. Clin Orthop1985;196:26-34.

13. Jackson DW, Windler GE, Simon TM:Intraarticular reaction associated withthe use of freeze-dried, ethylene oxide-sterilized bone-patella tendon-bone al-lografts in the reconstruction of the an-terior cruciate ligament. Am J SportsMed 1990;18:1-11.

14. Shelton WR, Treacy SH, Dukes AD,Bomboy AL: Use of allografts in kneereconstruction: I. Basic science aspectsand current status. J Am Acad OrthopSurg 1998;6:165-168.

15. Vangsness CT Jr, Garcia IA, Mills CR,Kainer MA, Roberts MR, Moore TM:Allograft transplantation in the knee:Tissue regulation, procurement, pro-cessing, and sterilization. Am J SportsMed 2003;31:474-481.

16. Falconiero RP, DiStefano VJ, Cook TM:Revascularization and ligamentizationof autogenous anterior cruciate liga-ment grafts in humans. Arthroscopy1998;14:197-205.

17. Beynnon BD, Johnson RJ: Anterior cru-ciate ligament injury rehabilitation inathletes: Biomechanical considerations.Sports Med 1996;22:54-64.

18. Rodeo SA, Arnoczky SP, Torzilli PA,Hidaka C, Warren RF: Tendon-healingin a bone tunnel: A biomechanical andhistological study in the dog. J BoneJoint Surg Am 1993;75:1795-1803.

19. Jackson DW, Grood ES, Goldstein JD, etal: A comparison of patellar tendon au-tograft and allograft used for anteriorcruciate ligament reconstruction in thegoat model. Am J Sports Med 1993;21:176-185.

20. Nikolaou PK, Seaber AV, Glisson RR,Ribbeck BM, Bassett FH III: Anterior

cruciate ligament allograft transplanta-tion: Long-term function, histology, re-vascularization, and operative tech-nique. Am J Sports Med 1986;14:348-360.

21. Fulkerson JP, Langeland R: An alterna-tive cruciate reconstruction graft: Thecentral quadriceps tendon. Arthroscopy1995;11:252-254.

22. Theut PC, Fulkerson JP, Armour EF,Joseph M: Anterior cruciate ligamentreconstruction utilizing central quadri-ceps free tendon. Orthop Clin North Am2003;34:31-39.

23. Scheffler SU, Sudkamp NP, GockenjanA, Hoffman RFG, Weiler A: Biomechan-ical comparison of hamstring and pa-tellar tendon graft anterior cruciate lig-ament reconstruction techniques: Theimpact of fixation level and fixationmethod under cyclic loading. Arthrosco-py 2002;18:304-315.

24. Kousa P, Jarvinen TL, Vihavainen M,Kannus P, Jarvinen M: The fixationstrength of six hamstring tendon graftfixation devices in anterior cruciate lig-ament reconstruction: I. Femoral site.Am J Sports Med 2003;31:174-181.

25. Kousa P, Jarvinen TL, Vihavainen M,Kannus P, Jarvinen M: The fixationstrength of six hamstring tendon graftfixation devices in anterior cruciate lig-ament reconstruction: II. Tibial site. AmJ Sports Med 2003;31:182-188.

26. Caborn DNM, Urban WP Jr, Johnson DL,Nyland J, Pienkowski D: Biomechanicalcomparison between BioScrew and tita-nium alloy interference screws for bone-patellar tendon-bone graft fixation in an-terior cruciate ligament reconstruction.Arthroscopy 1997;13:229-232.

27. Butler JC, Branch TP, Hutton WC: Op-timal graft fixation—the effect of gapsize and screw size on bone plug fixa-tion in ACL reconstruction. Arthroscopy1994;10:524-529.

28. Lemos MJ, Jackson DW, Lee TQ, SimonTM: Assessment of initial fixation of en-doscopic interference femoral screwswith divergent and parallel placement.Arthroscopy 1995;11:37-41.

29. Aune AK, Holm I, Risberg MA, Jensen

Graft Selection in Anterior Cruciate Ligament Reconstruction

206 Journal of the American Academy of Orthopaedic Surgeons

HK, Steen H: Four-strand hamstringtendon autograft compared with patel-lar tendon-bone autograft for anteriorcruciate ligament reconstruction: A ran-domized study with two-year follow-up. Am J Sports Med 2001;29:722-728.

30. Ejerhed L, Kartus J, Sernet N, Kohler K,Karlsson J: Patellar tendon or semiten-dinosus tendon autografts for anteriorcruciate ligament reconstruction? Aprospective randomized study with atwo-year follow-up. Am J Sports Med2003;31:19-25.

31. Shaieb MD, Kan DM, Chang SK, Maru-moto JM, Richardson AB: A prospectiverandomized comparison of patellar ten-don versus semitendinosus and gracilistendon autografts for anterior cruciateligament reconstruction. Am J Sports Med2002;30:214-220.

32. Jansson KA, Linko E, Sandelin J, Har-ilainen A: A prospective randomizedstudy of patellar versus hamstring ten-don autografts for anterior cruciate lig-ament reconstruction. Am J Sports Med2003;31:12-18.

33. Beynnon BD, Johnson RJ, Fleming BC,et al: Anterior cruciate ligament re-placement: Comparison of bone-patellar tendon-bone grafts with two-strand hamstring grafts. A prospective,randomized study. J Bone Joint Surg Am2002;84:1503-1513.

34. Aglietti P, Buzzi R, D’Andria S, Zac-cherotti G: Patellofemoral problems af-ter intraarticular anterior cruciate liga-ment reconstruction. Clin Orthop 1993;288:195-204.

35. Shelton WR, Papendick L, Dukes AD:Autograft versus allograft anterior cru-ciate ligament reconstruction. Arthros-copy 1997;13:446-449.

36. Sachs RA, Daniel DM, Stone ML, GarfeinRF: Patellofemoral problems after ante-rior cruciate ligament reconstruction. AmJ Sports Med 1989;17:760-765.

37. Shelbourne KD, Nitz P: Accelerated re-habilitation after anterior cruciate liga-ment reconstruction. Am J Sports Med1990;18:292-299.

38. Chen CH, Chen WJ, Shih CH: Arthro-scopic anterior cruciate ligament recon-struction with quadriceps tendon-patellar bone autograft. J Trauma 1999;46:678-682.

39. Rosenberg TD, Franklin JL, BaldwinGN, Nelson KA: Extensor mechanismfunction after patellar tendon graft har-vest for anterior cruciate ligament re-construction. Am J Sports Med 1992;20:519-526.

40. Lephart SM, Kocher MS, Harner CD,Fu FH: Quadriceps strength and func-tional capacity after anterior cruciate

ligament reconstruction: Patellar ten-don autograft versus allograft. Am JSports Med 1993;21:738-743.

41. Carter TR, Edinger S: Isokinetic evalu-ation of anterior cruciate ligament re-construction: Hamstring versus patel-lar tendon. Arthroscopy 1999;15:169-172.

42. Nakamura N, Horibe S, Sasaki S, et al:Evaluation of active knee flexion andhamstring strength after anterior cruci-ate ligament reconstruction using ham-string tendons. Arthroscopy 2002;18:598-602.

43. Viola R, Vianello R: Three cases of pa-tella fracture in 1,320 anterior cruciateligament reconstructions with bone-patellar tendon-bone autograft. Arthros-copy 1999;15:93-97.

44. Marumoto JM, Mitsunaga MM, Rich-ardson AB, Medoff RJ, Mayfield GW:Late patellar tendon ruptures after re-moval of the central third for anteriorcruciate ligament reconstruction: A re-port of two cases. Am J Sports Med 1996;24:698-701.

45. Miller SL, Gladstone JN: Graft selectionin anterior cruciate ligament recon-struction. Orthop Clin North Am 2002;33:675-683.

46. Simonds RJ, Holmberg SD, HurwitzRL, et al: Transmission of human im-munodeficiency virus type 1 from a se-ronegative organ and tissue donor. NEngl J Med 1992;326:726-732.

47. Makisalo S, Skutnabb K, Holmstrom J,Gronblad M, Paavolainen P: Recon-struction of anterior cruciate ligamentwith carbon fiber: An experimentalstudy on pigs. Am J Sports Med 1988;16:589-593.

48. Kumar K, Maffulli N: The ligamentaugmentation device: An historical per-spective. Arthroscopy 1999;15:422-432.

49. Colombet P, Allard M, Bousquet V, deLavigne C, Flurin PH, Lachaud C: An-terior cruciate ligament reconstructionusing four-strand semitendinosus andgracilis tendon grafts and metal inter-ference screw fixation. Arthroscopy2002;18:232-237.

50. Noyes FR, Barber-Westin SD: Revisionanterior cruciate ligament surgery: Ex-perience from Cincinnati. Clin Orthop1996;325:116-129.

51. Kartus J, Stener S, Lindahl S, ErikssonBI, Karlsson J: Ipsi- or contralateral pa-tellar tendon graft in anterior cruciateligament revision surgery: A compari-son of two methods. Am J Sports Med1998;26:499-504.

52. Uribe JW, Hechtman KS, Zvijac JE, Tjin-A-Tsoi EW: Revision anterior cruciateligament surgery: Experience from Mi-ami. Clin Orthop 1996;325:91-99.

53. Woods GW, Fincher AL, O’Connor DP,Bacon SA: Revision anterior cruciateligament reconstruction using the later-al third of the ipsilateral patellar tendonafter failure of a central-third graft: Apreliminary report on 10 patients. Am JKnee Surg 2001;14:23-31.

54. Colosimo AJ, Heidt RS Jr, Traub JA,Carlonas RL: Revision anterior cruciateligament reconstruction with a rehar-vested ipsilateral patellar tendon. Am JSports Med 2001;29:746-750.

55. Nixon RG, SeGall GK, Sax SL, Cain TE,Tullos HS: Reconstitution of the patel-lar tendon donor site after graft harvest.Clin Orthop 1995;317:162-171.

56. Burks RT, Haut RC, Lancaster RL: Bio-mechanical and histological observa-tions of the dog patellar tendon after re-moval of its central one-third. Am JSports Med 1990;18:146-153.

57. LaPrade RF, Hamilton CD, Montgom-ery RD, Wentorf F, Hawkins HD: Thereharvested central third of the patellartendon: A histologic and biomechan-ical analysis. Am J Sports Med 1997;25:779-785.

58. Proctor CS, Jackson DW, Simon TM:Characterization of the repair tissue af-ter removal of the central one-third ofthe patellar ligament: An experimentalstudy in a goat model. J Bone Joint SurgAm 1997;79:997-1006.

59. O’Neill DB: Arthroscopically assistedreconstruction of the anterior cruciateligament: A prospective randomizedanalysis of three techniques. J Bone JointSurg Am 1996;78:803-813.

60. Bach BR Jr, Tradonsky S, Bojchuk J,Levy ME, Bush-Joseph CA, Khan NH:Arthroscopically assisted anterior cru-ciate ligament reconstruction using pa-tellar tendon autograft: Five- to nine-year follow-up evaluation. Am J SportsMed 1998;26:20-29.

61. Yunes M, Richmond JC, Engels EA,Pinczewski LA: Patellar versus ham-string tendons in anterior cruciate lig-ament reconstruction: A meta-analysis.Arthroscopy 2001;17:248-257.

62. Shino K, Inoue M, Horibe S, HamadaM, Ono K: Reconstruction of the ante-rior cruciate ligament using allogeneictendon: Long-term followup. Am JSports Med 1990;18:457-465.

63. Shelbourne KD, Urch SE: Primary ante-rior cruciate ligament reconstructionusing the contralateral autogenous pa-tellar tendon. Am J Sports Med 2000;28:651-658.

64. Eberhardt C, Kurth AH, Hailer N, JagerA: RevisionACLreconstruction using au-togenous patellar tendon graft. Knee SurgSports Traumatol Arthrosc 2000;8:290-295.

Robin V. West, MD, and Christopher D. Harner, MD

Vol 13, No 3, May/June 2005 207