The _Ligamentization_ Process In

1166

Collagen constitutes up to 65% to 80% of the dry weightmass of tendons and ligaments. It is responsible for func-tional integrity, and it contributes to the structural frame-work of the tissue because of its unique molecular coil con-figuration and its ability to form covalent intramolecularand intermolecular crosslinks. Crosslinks render collagenfibers stable and provide them with an adequate degree oftensile strength and viscoelasticity. Similar types of

The “Ligamentization” Process in

Human Anterior Cruciate Ligament

Reconstruction With Autogenous

Patellar and Hamstring Tendons

A Biochemical Study

Keishi Marumo,*† MD, Mitsuru Saito,† MD, Tsuneo Yamagishi,‡ MD, and Katsuyuki Fujii,† MDFrom the †Department of Orthopaedic Surgery, The Jikei University School of Medicine, Tokyo,Japan, and the ‡Department of Orthopaedic Surgery, Mutual Benefit Association for TokyoMetropolitan Employees, Aoyama Hospital, Tokyo, Japan

Background: There is little information documenting whether the phenomenon of “ligamentization,” as proposed by Amiel,occurs in the human anterior cruciate ligament after clinically effective reconstruction. To clarify this point, we analyzed bio-chemical differences between the native anterior cruciate ligament; the patellar, semitendinosus, and gracilis tendons; and ante-rior cruciate ligaments reconstructed with autografts.

Study Design: Cohort study; Level of evidence, 2.

Methods: Fifty patients who underwent arthroscopically assisted anterior cruciate ligament reconstruction using either semi-tendinosus and gracilis tendon or bone–patellar tendon–bone autografts were selected for the study. Samples of grafted tissuewere collected during arthroscopy and quantitatively analyzed for collagen content and the amount of reducible and nonre-ducible crosslinks at 4 to 6 postoperative months in patients with semitendinosus and gracilis tendon grafts and at 11 to 13months in all patients with semitendinosus and gracilis tendon or bone–patellar tendon–bone grafts.

Results: The total collagen content and nonreducible/reducible crosslink ratios increased significantly during the postoperativeperiod (P < .05). The dihydroxylysinonorleucine/hydroxylysinonorleucine ratio was 3.11 ± 0.56 in the native anterior cruciate lig-ament, 1.21 ± 0.47 in the patellar tendon, and 3.59 ± 1.58 in the anterior cruciate ligaments reconstructed with bone–patellartendon–bone autografts 1 year after surgery. The dihydroxylysinonorleucine/hydroxylysinonorleucine ratio in both semitendi-nosus and gracilis tendons was less than 1.0. However, in anterior cruciate ligaments reconstructed with semitendinosus andgracilis tendon autografts, it was 2.34 ± 0.98 at 4 to 6 months and 3.43 ± 1.61 at 11 to 13 months after the operation.

Conclusions: After anterior cruciate ligament reconstruction with autografts, biochemical characteristics of the graft resembledthose of the native anterior cruciate ligament. These findings suggest that, regarding the amount of collagen crosslinks and theirarchitecture, the phenomenon of ligamentization occurs in the successfully reconstructed human anterior cruciate ligament within1 year after operation.

Keywords: anterior cruciate ligament (ACL); ligamentization; collagen crosslinks; patellar tendon; semitendinosus tendon; gra-cilis tendon; biochemical study; reconstruction

*Address correspondence to Keishi Marumo, MD, Department ofOrthopaedic Surgery, The Jikei University School of Medicine, 3-25-8Nishi-Shinbashi, Minato-ku, Tokyo, 105-8461 Japan (e-mail:[email protected]).

No potential conflict of interest declared.

The American Journal of Sports Medicine, Vol. 33, No. 8DOI: 10.1177/0363546504271973© 2005 American Orthopaedic Society for Sports Medicine

Vol. 33, No. 8, 2005 The Ligamentization Process in ACL Reconstruction 1167

crosslinks and crosslink molecular loci have been found invarious tissues; however, their content, ratios, and quanti-tative molecular distribution vary. Tendinous tissue has aconsiderably different collagen-reducible crosslinks elu-tion profile compared to ligamentous tissue. It containsrelatively large amounts of hydroxylysinonorleucine(HLNL) and only a small amount of dihydroxylysinonor-leucine (DHLNL). Ligaments, on the other hand, containlittle HLNL, contain larger amounts of DHLNL, and havea higher DHLNL/HLNL ratio than tendons.2 In the kneejoint, the cruciate ligaments show the highestDHLNL/HLNL ratio when compared to patellar tendonsor semitendinosus and gracilis tendons, in which this pro-portion is reversed.2,14

Extensive animal3,5,16 and clinical studies§ have demon-strated that bone–patellar tendon–bone (BPTB) and semi-tendinosus and gracilis tendon (StG) autografts are themost appropriate sources of graft tissue for ACL recon-struction, since their morphologic, histologic, and biome-chanical parameters undergo the most clinically desirablefunctional adaptation. Both autografts have commonlybeen used, even though differences between the 2 tissueshave been pointed out: on the basis of the results of ameta-analysis, patients with BPTB grafts had slightly bet-ter clinical outcomes in terms of reduction of anterior tib-ial translation, side-to-side differences, and return topreinjury sports activity.33 This finding was attributed tovarying fixation techniques, surgical methods, rehabilita-tion programs, and accompanying intra-articular lesions.

Since a clinically successful ACL graft usually appearssimilar to a mature native ligament enveloped by vascu-larized synovial tissue—that is, the graft appears to havenative ACL morphology—it has been believed that incor-porated grafts transform into morphologically “normal”ligaments, but the term was not specific or clearly defined.Morphologic changes might reflect histologic and biochem-ical remodeling of the grafts, and in fact, Amiel et al4

demonstrated in a rabbit model that autografts undergo“ligamentization”—transition of the biochemical and his-tologic parameters of the graft from tendinous to ligamen-tous in appearance. Such properties become similar tothose of the native ACL, even though the 2 tissues are bio-chemically distinct (different proportions of collagen, gly-cosaminoglycans, and collagen-reducible crosslinks). Theprocess of ligamentization is thought to be a part of grafttissue adaptation, which consists of necrosis, swelling,revascularization, fibroblastic invasion, and synthesis ofcollagen with ligament reformation (ligamentization).

Biomechanical studies in different animal models haveshown that the results vary depending on the model andthat the strength and stiffness of the grafted tissues neverreach those of the native ACL.8-10,18,23 In other words,although morphologically resembling normal ligaments,the grafts may not function exactly alike. When interpret-ing these data with regard to changes occurring in humanautografts, we must consider many clinically importantfactors such as graft isometricity, anatomical positioning,patient compliance, healing response, vascularity, biochem-

ical strength, and postoperative rehabilitation, all of whichwere,and still are,difficult to control in animals.Nevertheless,the results of animal studies are of great importancebecause human research has been limited only to postmortemand second-look arthroscopy evaluation.1,12,13,19,21,27,28

The present study attempted to clarify whether, and towhat extent, the process of ligamentization occurs in thesuccessfully reconstructed ACL in humans. Biochemicaland histologic characteristics of the native ACL; the patel-lar, semitendinosus, and gracilis tendons; and the ACLreconstructed with BPTB and StG autografts are evaluatedand compared. As far as we know, this is the first studyaddressing biochemical properties of autografts in livingsubjects after ACL reconstruction surgery.

MATERIALS AND METHODS

Fifty of 855 patients who underwent ACL reconstructionsat our institutions from 1996 to 2002 were selected for thestudy. An arthroscopically assisted double-incision tech-nique with either BPTB or distally based StG graft wasapplied in each case. The BPTB grafts were used in high-performance athletes and heavy workers (this group ofpatients is later referred to as the BPTB group), and StGgrafts were used in other patients (later referred to as theStG group).

Patient selection was based on postoperative assess-ment using a modified version of the International KneeDocumentation Committee’s (IKDC’s) Knee LigamentStandard Evaluation questionnaire that included 7 crite-ria: (1) patient’s subjective assessment, (2) symptom eval-uation (pain, swelling, partial giving way, full giving way),(3) range of motion, (4) ligament examination (Lachmantest using KT-1000 arthrometer (MedMetic Corp, SanDiego, Calif), total anteroposterior translation, medial jointopening, and pivot shift tests), (5) compartmental crepitus(medial, lateral, and patellofemoral), (6) radiographic eval-uation (medial, lateral, and patellofemoral joint space nar-rowing), and (7) a 1-legged hop test for distance as a per-centage of the distance jumped on the contralateral leg.Only patients with the overall grade of A, which reflectedclinically effective ACL reconstructions, were chosen. Allpatients had less than 3-mm differences on manual maxi-mum side-to-side testing. The mean side-to-side differencewas 0.03 ± 1.09 mm (range, –2 to 2 mm) in the BPTB groupand 0.55 ± 1.05 mm (range, –2 to 2 mm) in the StG group.

Primary arthroscopically assisted ACL reconstructionsusing BPTB or StG autografts were performed in 30 and20 patients, respectively, for ACL injuries. Associatedmeniscal damage requiring partial meniscectomy wasfound during surgery in 5 patients of the BPTB group;medial meniscal tears were corrected in 3 cases and lateralmeniscal tears in 2 cases. No additional surgical interven-tion was required in patients with StG grafts, and no obvi-ous cartilage lesions were found in either group.

There were 28 men and 2 women, with a mean age of28.5 ± 11.1 years (range, 15.1-49.0 years), in the BPTBgroup; the StG group included 3 men and 17 women, witha mean age of 24.5 ± 9.1 years (range, 13.3-44.0 years).§References 1, 6, 7, 11, 13, 17, 19, 20, 24-26, 28, 31, 32.

1168 Marumo et al The American Journal of Sports Medicine

At 4 to 6 months after the initial surgery, secondaryarthroscopy and graft biopsy were performed in 8 patientsof the StG group after obtaining their informed consent.Out of these patients, 6 volunteered to undergo biopsy and2 required arthroscopy for recurrent effusions of the affectedknee joints. In the latter 2 cases, a slightly inflamed syn-ovium was found at arthroscopy; however, no abnormali-ties in the reconstructed ligaments were observed.Effusions subsided soon after the arthroscopy.Unfortunately, biopsy specimens could not be obtainedfrom patients of the BPTB group at 4 to 6 months after thefirst surgery.

Patients from both therapeutic groups underwentarthroscopic biopsy when hardware (screws and staples)was removed, 11 to 13 months after the primary operation.Arthroscopic findings of the grafts showed they were alltaut and surrounded by thin and smooth synovial tissuewith relatively poor blood vessels. The synovial tissueenveloped a thick ligamentous one, similar to that of thenative ACL.

Tissue samples obtained from cadaveric specimens wereselected for the controls. Native ACL as well as patellar,semitendinosus, and gracilis tendons were harvested from7 male cadavers with a mean age of 36 years (range, 20-69years), within 48 hours after their accidental death. Thecollected samples were examined histologically for degen-erative changes, which were not found.

Informed consent for the secondary biopsy and use of thecollected material for research was given by the patientsthemselves or, in the case of the postmortem samples, bythe subjects’ relatives, in accordance with proceduresrequired by the ethics committee and the hospital trust.

SURGICAL PROCEDURE

StG Group

The semitendinosus and gracilis tendons were harvestedproximally at the musculotendinous junction through a 3-cm longitudinal skin incision 4 cm below the joint line overthe pes bursa with an open-loop tendon stripper. After allmuscle was removed from the tendons without detach-ment of the tibial insertion, a ligament augmentationdevice (LAD; Kennedy LAD, 3M, Minneapolis, Minn), a 6-mm wide × 1-mm thick ×18-cm long strip of biodegradable,braidlike material consisting of about 5000 fibers of 5 to 6µm of poly-L-lactic acid yarn heat-sealed at both ends, wasanchored to the distal ends of both tendons with severalinterrupted sutures of a nonabsorbable material. The LADwas used to provide early tensile strength to a compositegraft, protecting it from disruption and attenuation untilrevascularization and remodeling had occurred in thereconstructed ACL. The LAD had a failure strength of1180 N and a failure strain of 35%. Hydrolysis testsshowed that maximum tensile strength was maintaineduntil 3 months, after which it gradually decreased to 50%of the initial strength at 12 months and reached zero by 24months.22 The tendons and LAD were sutured togetherproximally using absorbable materials under even ten-

sion, thereby totally encompassing the augmentationdevice with both tendons. The 2 tendons were folded backat 11 cm and then 3 cm from the distal end of the LAD andwhipstitched together to make 6-bundle semitendinosusand gracilis composite grafts with the LAD. The diameterof the graft was generally 9 mm in the intra-articular por-tion and proximal tibial tunnel and 8 mm in the distalfemoral tunnel.

A tibial tunnel was drilled to exit from the center of theACL anatomical insertion site. The tunnel was carefullychamfered to avoid disruption of the ACL stump. A lateralfemoral incision about 4 cm in length was used to place therear-entry femoral aiming device (Acufex MicrosurgicalInc, Mansfield, Mass), to allow femoral tunnel reamingfrom “outside-in.” A hook was placed through the inter-condylar notch and positioned 6 mm anterior to the top ofthe femoral condyle at the 11 o’clock position on the rightknee and the 1 o’clock position on the left knee.

After the femoral and tibial osseous tunnels were drilledwith an 8-mm and a 9-mm cannulated reamer, respectively,and the sharp edge of the femoral drill hole was cham-fered, the graft was passed from distal to proximal using agraft passer. The graft was pulled firmly and the knee wascycled several times to give pretension to the graft, to con-firm isometricity throughout the range of motion, and toensure that no impingement occurred. All grafts becametight when the knee approached full extension, but nonewere tight in increased flexion angles. The LAD was fixedto the femur separately with spiked staples under maxi-mum tension and the knee at 20° of flexion (Figure 1).

Figure 1. The semitendinosus and gracilis tendons harvestedwithout detachment of tibial insertion. The tendons and liga-ment augmentation device (LAD) are sutured together, and acomposite graft is fixed to the femur with spiked staples onlythrough the LAD. The bone tips are filled up in the space ofthe femoral bone tunnel.


BPTB Group

The central third (9-10 mm) of the patellar tendon wasremoved with bone blocks at each end through a 5-cmmedial parapatellar incision. The patellar bone block wastrapezoidally shaped (2 mm wider proximally than distally),while the tibial bone block had the same width proximallyand distally.

The graft was introduced via the femoral tunnel throughthe joint and into the tibial tunnel with the patellar boneblock first, and it was fixed with Kurosaka interferencescrews (De Puy, Warsaw, Ind). The details of the operationare the same as described above, except for the drilling ofboth osseous tunnels with 9- to 10-mm diameter cannulat-ed screws.

After immobilization of the operated knee in a cast at 0°of knee flexion for 1 week, passive and active motion of theknee was begun. Flexion to 90° was usually achieved by 3weeks and advanced to 120° by 6 weeks. Particular empha-sis was placed on obtaining full extension in the early post-operative phase. Partial weightbearing was allowed at 3weeks, with full weightbearing allowed by 6 weeks.Quadriceps and hamstrings muscle-strengthening exerciseswere initiated on the first postoperative day. Active flexionand hamstrings isometrics were encouraged, but quadri-ceps muscle strengthening in the final 30° of knee flexionwas not allowed until 8 weeks. Bicycling, half-squats, andbalance training were begun at 5 to 6 weeks. Straight-ahead jogging and swimming were allowed at 5 to 6months, and a gradual return to sports was allowedbetween 10 to 12 months postoperatively if rehabilitationcriteria were met. A hinged knee brace was discontinued 4months postoperatively.

SAMPLE PREPARATION

The biochemical and histologic evaluations were carried outas in our previous study on age-related changes in biochem-ical characteristics of collagen fibers in human subjects.29

An autograft tissue sample was collected from itsapproximately most central part during arthroscopy usinga 2.5-mm basket forceps via the anteromedial portal. Asingle bundle was isolated and removed from the middlesegment of the graft in its longitudinal and cross-sectionaldimension. A single specimen was approximately 2 to 3mm thick and 1.2 to 1.6 mm long (Figure 2). It was subse-quently divided into 2 pieces.

One piece was fixed in periodate-lysine-paraformaldehyde(PLP) for 6 hours at 4°C. It was then dehydrated in alcoholand embedded in paraffin. Sections 5 to 6 µm thick werecut and stained with hematoxylin-eosin for lightmicroscopy observations.

The other piece was used for quantitative analysis ofcollagen-reducible and -nonreducible crosslinks by high-performance liquid chromatography (HPLC) using a fluo-rescence detection method established previously in ourlaboratory.30 Sample preparation for HPLC was as follows.The adherent tissues of each sample were removed byscraping. All tissues were minced and suspended in 500

vol (v/wt) of 0.05 M potassium phosphate buffer, pH 7.6(ionic strength = 0.15), at 4°C and continuously stirred for72 hours under a vacuum. A 1/30 volume of sodium boro-hydride (NaBH4) was added to the solution. The reactionwas allowed to proceed for 60 minutes at 37°C and was ter-minated by the addition of 3 N acetic acid to decrease thepH value to 4.0. The solution residue was collected by cen-trifugation (3000g, 15 minutes), washed with deionizedwater, and lyophilized. For the HPLC analysis, lyophilizedsamples were dissolved in 0.2 N sodium citrate buffer (pH2.20) and filtered through a 0.45-µm filter (GelmanSciences Japan Ltd, Tokyo, Japan).

The obtained results were compared with those of thenative ACL and the patellar, semitendinosus, and gracilistendons harvested from 7 men within 48 hours after theiraccidental deaths.

Statistical analyses were carried out using the Mann-Whitney test or the Wilcoxon signed rank test, and statis-tical significance was set at P < .05.

RESULTS

The histologic light-microscopy appearance of the trans-planted tissue 4 to 6 months after ACL reconstruction sur-gery (StG group) showed spindle-shaped cells oriented lon-gitudinally along with collagen fibers and a moderatenumber of mononuclear cells. The graft remained vascular,blood vessels within the graft substance were occasionallyseen, and the highest cellularity was observed in zones ofvascular invasion (Figure 3A). Cellularity and vascularitywere graded with time, and the majority of graftsappeared to have nearly normal ACL morphology 1 yearafter ACL reconstruction (Figures 3 B and C). For bio-chemical analyses, samples with as little as 10% metapla-sia or myxoid degeneration (a hazy, avascular, acellular,and disorganized alignment of tissue fibers) were selected.

Figure 2. A routine observation revealed that the ACL graftwas taut and functional. The synovium was carefully clearedfrom the graft at the biopsy site, and a cylindrical specimen(2-3 mm thick and 1.2-1.6 mm long) was procured from themiddle segment of the graft in its longitudinal and cross-sectional dimension.


Biochemical studies showed that the total collagen con-tent, expressed as a percentage of the ACL tissue’s dryweight, increased significantly during the postoperativeperiod (P < .05). In the StG group, collagen content rosefrom 78.5% ± 9.1% at 4 to 6 months to 92.2% ± 7.2% at 11to 13 months after the initial surgery. In the BPTB group,

it was 95.3% ± 4.1% at 11 to 13 months after the recon-struction. By comparison, the mean value of total collagencontent in native ACL was 83.0% ± 5.3% (Table 1).

The amount of total reducible crosslinks, that is, theamount of DHLNL, HLNL, and lysinonorleucine (LNL) ininsoluble collagen (mol/mol ratio), was higher at 4 to 6months than at 11 to 13 months after the primary opera-tion, with a ratio of 0.76 between the 2 measured time val-ues. In contrast, this ratio was 2.07 for nonreduciblecrosslinks pyridinoline and deoxypyridinoline when theirlevels were examined (Table 2).

The DHLNL/HLNL ratio was 3.11 ± 0.56 in native ACL,1.21 ± 0.47 in the patellar tendon, and 3.59 ± 1.58 in theBPTB group of patients 1 year after surgery. TheDHLNL/HLNL ratio in both semitendinosus and gracilistendons was less than 1.0; however, in the StG group ofpatients, it was 2.34 ± 0.98 at 4 to 6 months and 3.43 ± 1.61at 11 to 13 months after the operation (Table 3).

DISCUSSION

To our knowledge, this is the first study on the biochemi-cal properties of human autografts in living patients afterclinically successful ACL reconstructions. One previous

Figure 3. A, a 5-month biopsy from a hamstring tendonautograft. Spindle-shaped cells are oriented longitudinallyand aligned with the longitudinally oriented collagen fibers.A moderate number of mononuclear cells are present. Thegrafts remain vascular, and blood vessels within the graftsubstance are occasionally seen. The most cellular areas ofeach graft are the zones of vascular invasion. B, a 12-monthbiopsy from the hamstring tendon autografts. By this time,the cellular counts and collagen maturation are close tothose of native ACL controls. C, the native ACL from a freshcadaver. Notice the relatively low fibroblast nuclear count,linear and spindle nuclear morphology, and orderly collagenmatrix. Magnification ×200 for A-C.

A B

C

TABLE 1Collagen Content of Native Versus Reconstructed ACLa

Tissue Total Collagen, %

Native ACL (n = 7) 83.0 ± 5.3StG at 4-6 mo (n = 8) 78.5 ± 9.1StG at 11-13 mo (n = 20) 92.2 ± 7.2BPTB at 11-13 mo (n = 30) 95.3 ± 4.1

aData are expressed as a percentage of the tissue’s dry weight,with the means ± SDs given for the subjects in each group. StG,semitendinosus and gracilis tendon; BPTB, bone–patellar ten-don–bone.

bA significant increase (P < .05) between the 2 groups indicatedusing the Mann-Whitney test.

cA significant increase (P < .05) between the 2 groups indicatedusing the Wilcoxon signed rank test.

]c ]b ]b


case report described a crosslink profile of a semitendi-nosus autograft; it was found that 4 years after the recon-struction, the graft had a much higher DHLNL/HLNLratio than did a hamstring tendon, simulating the proper-ties of the native ACL.21

Our study focused on the biochemical characteristics ofthe BPTB and StG grafts commonly used in ACL recon-structions. We found that after successful ACL reconstruc-tions, total collagen content was relatively lower in StGgrafts than in the native ACL 4 to 6 months after the oper-ation; however, it increased significantly in the interven-ing 5 to 7 months. The amount of reducible crosslinksdecreased 24%, and the amount of nonreducible crosslinksincreased approximately 2-fold in StG grafts during a 6-month interval between 4 to 6 and 11 to 13 months afterthe primary operation. In the BPTB grafts at 11 to 13months postoperatively, the collagen content was evenhigher than in StG grafts and native ACL. The differencein the DHLNL/HLNL ratios between native ACL andautogenous graft tissues decreased, and chromatographicpatterns of reducible crosslinks in the grafted tissuesclosely resembled the pattern found in the native ACL.Our findings suggest that the process of ligamentizationtakes place in successfully reconstructed human ACLwithin 1 year after surgery.

By ligamentization, we mean the phenomenon proposedby Amiel et al 4 in 1986, that is, the biochemical andhistologic remodeling of the graft tissue from tendinous toligamentous-appearing in the new intra-articular environ-ment specific to the native ACL. The present studyaddresses ligamentization in its narrow sense, not as awhole process of the graft tissue adaptation but rather asan integral part of it. A “normal” ligament never formsafter ACL reconstruction, despite morphologic and histo-logic findings suggesting that it does. Tendons used forACL reconstructions appeared similar to the native ACLat arthroscopy and by observation under an optical micro-scope (Figure 3); however, an electron microscopic studydemonstrated that collagen fibrils in the reconstructed

ligament are differently organized than those of the nativeACL.1

Changes in collagen concentrations and biochemicalprofiles might be explained as follows. Since the center ofthe transplanted tissue is initially avascular and containsrelatively low numbers of viable cells, collagen synthesiscannot be very active in the early postoperative months,even though vascular invasion from the surface of thegraft occurs within 3 to 8 weeks after the reconstructionand is followed by the repopulation phase.13,27,28 Enoughrevascularization, release of growth factors by viable cellsthat enter the graft tissue through the newly formed ves-sels, and mechanical forces all stimulate collagen produc-tion; that is why, as observed in our study, collagen contentincreased with time and became even higher than in thenative ACL, probably because of collagen overexpression.The conversion of collagen crosslinks from reducible intononreducible occurred simultaneously with collagen syn-thesis and mechanical stress, as well as with other intra-articular factors that might have contributed to thisrearrangement (and thus ligamentization). Both collagencontent and number of nonreducible crosslinks might sub-due later, eventually becoming lower than in the originalACL and/or demonstrating different profiles depending onthe graft region (peripheral to the central region gradient),but these changes could not be clarified because of limita-tions in the study: the restricted number of biopsies andspecimen collection sites.

Clinical outcomes and biomechanical/biochemical char-acteristics of the grafts might be influenced by surgicaltechniques and postoperative rehabilitation programs. Inthe present study, the hamstring tendons were harvestedwithout detachment of their tibial insertions. The tendonsand LAD were sutured together, and the composite graftwas fixed to the femur with spiked staples only throughthe LAD. The advantage of preserving the tibial insertionof hamstring tendons was to ensure (1) less injury, (2)

TABLE 2Quantitative Changes in CollagenCrosslinks in Reconstructed ACLa

DHLNL + HLNL Pyr + Dpyr,+ LNL, mol/mol mol/mol

Tissue of collagen of collagen

Native ACL (n = 7)b 0.433 ± 0.093 0.159 ± 0.079Reconstructed ACLStG at 4-6 mo (n = 8)b 0.627 ± 0.123 0.144 ± 0.093StG at 11-13 mo (n = 20)b 0.477 ± 0.243 0.298 ± 0.029

StG at 11-13 mo/StG at 4-6 moc 0.76 2.07

aDHLNL, dihydroxylysinonorleucine; HLNL, hydroxylysinonor-leucine; LNL, lysinonorleucine; Pyr, pyridinoline; Dpyr,deoxypyridinoline; StG, semitendinosus and gracilis tendon.

bData are expressed as means ± SDs of subjects.cData are expressed as the mean of StGs at 11 to 13 months

divided by that of StGs at 4 to 6 months.

TABLE 3Relative Amounts of Collagen Crosslinks in Native

Versus Reconstructed ACLa

Tissue DHLNL/HLNL

Native ACL (n = 7) 3.11 ± 0.56Native PT (n = 7) 1.21 ± 0.47Native St (n = 7) 0.63 ± 0.15Native G (n = 7) 0.78 ± 0.28Reconstructed ACLStG at 4-6 mo (n = 8) 2.34 ± 0.98StG at 11-13 mo (n = 20) 3.43 ± 1.61BPTB at 11-13 mo (n = 30) 3.59 ± 1.58

aData are expressed as means ± SDs of subjects. DHLNL, dihy-droxylysinonorleucine; HLNL, hydroxylysinonorleucine; PT;patellar tendon; St, semitendinosus tendon; G, gracilis tendon;BPTB, bone–patellar tendon–bone.

bA significant increase (P < .05) in the DHLNL/HLNL ratiosbetween the 2 groups indicated using the Wilcoxon signed ranktest.

]b

]b ]b]b

b

b

b


more viability, and (3) more stable distal fixation of thegraft. In the StG group of patients, the biodegradable LADwas employed for the following reasons: (1) to protect theautograft tissue during the early postoperative period bypromoting load sharing, (2) to avoid stress shielding, (3) toget enough diameter of the graft in the intra-articular por-tion, and (4) to obtain rigid graft fixation to the femur withspiked staples through the LAD without damaging theautograft by fixation device. Although the direct effects ofthe biodegradable LAD on clinical results in ACL recon-structions are not precisely understood, the LAD theoreti-cally promotes optimal load sharing with the autograftand may facilitate ligamentization.22

Arthroscopic findings and clinical results of ACL recon-structions with BPTB or StG grafts are usually found to besatisfactory when either aggressive or low-aggressiverehabilitation programs are used.6,7,11,17,20,24,26,31 However,histologic assessments of the grafted tissue in patientsundergoing aggressive rehabilitation have shown degener-ative changes in the reconstructed ligaments (Yamagishiet al, unpublished data, 2000), and clinical studies haveindicated that an early return to vigorous physical activitymay cause or increase the risk of greater knee laxity afterACL reconstruction with either a BPTB (Beynnon et al,unpublished data, 2002) or StG graft.15 These facts haveraised the issue of the long-term effects of the moreaggressive approach. Although the direct influence ofrehabilitation methods on biochemical properties of thegrafted tissue was not examined in the present study, ourdata indicated that ligamentization continues for at least1 year in the successfully reconstructed ACL, and duringthat period, the graft does not gain sufficient mechanicalfunctionality. Because histologic degeneration was scarceand the grafts were undergoing ligamentization in the suc-cessfully reconstructed ACL, a low-aggressive rehabilita-tion program might seem to be more biochemically, histo-logically, and clinically desirable.

REFERENCES

1. Abe S, Kurosaka M, Iguchi T, Yoshiya S, Hirohata K. Light and elec-tron microscopic study of remodeling and maturation process inautogenous graft for anterior cruciate ligament reconstruction.Arthroscopy. 1993;9:394-405.

2. Amiel D, Kleiner JB. Biochemistry of tendon and ligament. In: NimniME, ed. Collagen: Biotechnology. Vol 3. Boca Raton, Fla: CRC Press;1988:223-251.

3. Amiel D, Kleiner JB, Akeson WH. The natural history of anterior cru-ciate ligament autograft of patellar tendon origin. Am J Sports Med.1986;14:449-462.

4. Amiel D, Kleiner JB, Roux RD, Harwood FL, Akeson WH. The phe-nomenon of “ligamentization”: anterior cruciate ligament reconstruc-tion with autogenous patellar tendon. J Orthop Res. 1986;4:162-172.

5. Arnoczky SP, Tarvin GB, Marshall JL. Anterior cruciate ligamentreplacement using patellar tendon: an evaluation of graft revascular-ization in the dog. J Bone Joint Surg Am. 1982;64:217-224.

6. Bach BR Jr, Levy ME, Bojchuk J, Tradonsky S, Bush-Joseph CA,Khan NH. Single-incision endoscopic anterior cruciate ligamentreconstruction using patellar tendon autograft: minimum two-yearfollow-up evaluation. Am J Sports Med. 1998;26:30-40.

7. Bach BR Jr, Tradonsky S, Bojchuk J, Levy ME, Bush-Joseph CA,Khan NH. Arthroscopically assisted anterior cruciate ligament recon-

struction using patellar tendon autograft: five- to nine-year follow-upevaluation. Am J Sports Med. 1998;26:20-29.

8. Ballock RT, Woo SL, Lyon RM, Hollis JM, Akeson WH. Use of patel-lar tendon autograft for anterior cruciate ligament reconstruction inthe rabbit: a long-term histologic and biomechanical study. J OrthopRes. 1989;7:474-485.

9. Butler DL, Grood ES, Noyes FR, et al. Mechanical properties of pri-mate vascularized vs. nonvascularized patellar tendon grafts:changes over time. J Orthop Res. 1989;7:68-79.

10. Clancy WG Jr, Narechania RG, Rosenberg TD, Gmeiner JG,Wisnefske DD, Lange TA. Anterior and posterior cruciate ligamentreconstruction in Rhesus monkeys. J Bone Joint Surg Am.1981;63:1270-1284.

11. Corry IS, Webb JM, Clingeleffer AJ, Pinczewski LA. Arthroscopicreconstruction of the anterior cruciate ligament: a comparison ofpatellar tendon autograft and four-strand hamstring tendon autograft.Am J Sports Med. 1999;27:444-454.

12. Delay BS, McGrath BE, Mindell ER. Observations of a retrieved patel-lar tendon autograft used to reconstruct the anterior cruciate liga-ment: a case report. J Bone Joint Surg Am. 2002;84:1433-1438.

13. Falconiero RP, DiStefano VJ, Cook TM. Revascularization and liga-mentization of autogenous anterior cruciate ligament grafts inhumans. Arthroscopy. 1998;14:197-205.

14. Fujii K, Yamagishi T, Nagafuchi T, Tsuji M, Kuboki Y. Biochemicalproperties of collagen from ligaments and periarticular tendons of thehuman knee. Knee Surg Sports Traumatol Arthrosc. 1994;2:229-233.

15. Fujimoto E, Sumen Y, Urabe Y, et al. An early return to vigorous activ-ity may destabilize anterior cruciate ligaments reconstructed withhamstring grafts. Arch Phys Med Rehabil. 2004;85:298-302.

16. Goradia VK, Rochat MC, Kida M, Grana WA. Natural history of ahamstring tendon autograft used for anterior cruciate ligament recon-struction in a sheep model. Am J Sports Med. 2000;28:40-46.

17. Howell SM, Taylor MA. Brace-free rehabilitation, with early return toactivity, for knees reconstructed with a double-looped semitendi-nosus and gracilis graft. J Bone Joint Surg Am. 1996;78:814-825.

18. Jackson DW, Grood ES, Goldstein JD, et al. A comparison of patel-lar tendon autograft and allograft used for anterior cruciate ligamentreconstruction in the goat model. Am J Sports Med. 1993;21:176-185.

19. Johnson LL. The outcome of a free autogenous semitendinosus ten-don graft in human anterior cruciate reconstructive surgery: a histo-logical study. Arthroscopy. 1993;9:131-142.

20. Kleipool AE, Zijl JA, Willems WJ. Arthroscopic anterior cruciate liga-ment reconstruction with bone–patellar tendon–bone allograft orautograft: a prospective study with an average follow up of 4 years.Knee Surg Sports Traumatol Arthrosc. 1998;6:224-230.

21. Lane JG, McFadden P, Bowden K, Amiel D. The ligamentizationprocess: a 4-year case study following ACL reconstruction with asemitendinosus graft. Arthroscopy. 1993;9:149-153.

22. Marumo K, Fujii K. Minimum five-year follow-up of anterior cruciateligament reconstruction using semitendinosus and gracilis tendonautografts with poly-L-lactic acid ligament augmentation device [inJapanese]. Rheumatology. In press.

23. Ng GY, Oakes BW, Deacon OW, McLean ID, Lampard D.Biomechanics of patellar tendon autograft for reconstruction of theanterior cruciate ligament in the goat: three-year study. J Orthop Res.1995;13:602-608.

24. Noyes FR, Barber-Westin SD. A comparison of results in acute andchronic anterior cruciate ligament ruptures of arthroscopically assistedautogenous patellar tendon reconstruction. Am J Sports Med.1997;25:460-471.

25. Noyes FR, Butler DL, Grood ES, Zernicke RF, Hefzy MS. Biomechanicalanalysis of human ligament grafts used in knee-ligament repairs andreconstructions. J Bone Joint Surg Am. 1984;66:344-352.

26. O’Neill DB. Arthroscopically assisted reconstruction of the anteriorcruciate ligament: a prospective randomized analysis of three tech-niques. J Bone Joint Surg Am. 1996;78:803-813.

27. Rougraff BT, Shelbourne KD. Early histologic appearance of humanpatellar tendon autografts used for anterior cruciate ligament recon-struction. Knee Surg Sports Traumatol Arthrosc. 1999;7:9-14.


28. Rougraff B, Shelbourne KD, Gerth PK, Warner J. Arthroscopic andhistologic analysis of human patellar tendon autografts used for ante-rior cruciate ligament reconstruction. Am J Sports Med. 1993;21:277-284.

29. Saito M. Age-related changes in biochemical characteristics of colla-gen from human weight-bearing and non–weight-bearing bone [inJapanese]. Tokyo Jikeikai Medical Journal. 1999;114:327-337.

30. Saito M, Marumo K, Fujii K, Ishioka N. Single-column high-performanceliquid chromatographic–fluorescence detection of immature, mature,and senescent cross-links of collagen. Anal Biochem. 1997;253:26-32.

31. Shelbourne KD, Gray T. Anterior cruciate ligament reconstructionwith autogenous patellar tendon graft followed by accelerated reha-bilitation: a two- to nine-year followup. Am J Sports Med.1997;25:786-795.

32. Woo SL, Hollis JM, Adams DJ, Lyon RM, Takai S. Tensile propertiesof the human femur–anterior cruciate ligament–tibia complex: theeffects of specimen age and orientation. Am J Sports Med.1991;19:217-225.

33. Yunes M, Richmond JC, Engels EA, Pinczewski LA. Patellar versushamstring tendons in anterior cruciate ligament reconstruction: ameta-analysis. Arthroscopy. 2001;17:248-257.

Documents

The _Ligamentization_ Process In