Meniscal Allograft Transplantation in the Sheep Kneecsu-cvmbs.colostate.edu/Documents/...sheep-knee...SURGICAL INTERVENTIONS Meniscectomy Before surgery, 10 sheep cadaveric knees were

1464

Meniscal transplantation has become a viable surgicaloption for patients with symptomatic meniscal deficiency.It is thought that the predictable long-term development ofosteoarthritic degeneration in patients subjected to total

meniscectomy can be decreased if the joint is protected bya meniscal substitute. Unfortunately, scientific evidenceregarding the effectiveness of this surgical interventionis lacking. There has been limited long-term clinical inves-tigation on the success of meniscal allografts.13-15,28,33

Furthermore, animal studies have demonstrated varyingresults, and to date, a reliable animal model has not beendeveloped.7,10,16,31

Although the optimal timing for meniscal allograft trans-plantation has not been well defined, it has become evidentthat surgical intervention must be performed before thedevelopment of advanced joint degeneration. The results ofthis procedure are suboptimal if transplantation is delayed

Meniscal Allograft Transplantationin the Sheep KneeEvaluation of Chondroprotective Effects

Bryan T. Kelly,*† MD, Hollis G. Potter,† MD, Xiang-Hua Deng,† MD, Andrew D. Pearle,† MD,A. Simon Turner,‡ BVSc, MS, Russell F. Warren,† MD, and Scott A. Rodeo,† MDFrom the †Hospital for Special Surgery, New York, New York, and ‡Colorado State University,Fort Collins, Colorado

Background: Early protection of articular cartilage, before degenerative changes appear on radiographs, should result in betterlong-term results, but scientific evidence regarding the effectiveness of meniscal transplantation is lacking.

Purpose: To assess the chondroprotective effects of a new meniscal allograft transplantation animal model and evaluate a mag-netic resonance imaging parameter, T2 mapping, in articular cartilage after meniscectomy and meniscal transplantation.

Study Design: Controlled laboratory study.

Methods: Forty-five skeletally mature sheep were placed into 3 surgical groups: lateral meniscectomy (n = 24), meniscal allografttransplant (n = 17), and sham (n = 4). Animals were sacrificed at 2, 4, or 12 months. Cartilage was assessed by gross inspection,magnetic resonance imaging, T2 mapping, biomechanical testing, and semiquantitative histologic analysis.

Results: There were no differences between the sham operation and nonoperated control limbs. Compared with control limbs,meniscectomy resulted in significant increases in cartilage degeneration by all objective criteria (P < .01). Compared with menis-cectomy, meniscal allograft transplantation resulted in significant decreases in cartilage degeneration (P < .02). There were sig-nificant correlations between T2-mapping data and all other traditional outcomes measures (P < .05, r2 = 0.37-0.67). Comparedwith the nonoperated control limbs, allograft transplants demonstrated no significant differences at 2 months in any category,except magnetic resonance imaging data. By 4 months, nonoperated control limbs demonstrated significantly less wear com-pared to allograft limbs in all categories except modified Mankin scores.

Conclusion: This model demonstrated significant chondroprotection compared with meniscectomy but demonstrated morecartilage wear at 4 months compared to control limbs. A high degree of allograft cell viability and vascular ingrowth was seen inallograft explants. T2 mapping may provide an accurate noninvasive measure of early cartilage degeneration after meniscectomy,as well as cartilage protection after meniscal allograft transplantation.

Clinical Relevance: This study establishes a reliable animal model for meniscal allograft transplantation and provides evidencefor the utility of T2 mapping at clinically relevant magnetic resonance imaging field strengths for evaluation of early cartilagedegeneration.

Keywords: meniscus; allograft; transplantation; T2 mapping; cartilage

*Address correspondence to Bryan T. Kelly, MD, Hospital for SpecialSurgery, 535 East 70th Street, New York, NY 10021 (e-mail: [email protected]).

Presented at the 51st annual meeting of the Orthopaedic ResearchSociety in Washington, DC.

No potential conflict of interest declared.

The American Journal of Sports Medicine, Vol. 34, No. 9DOI: 10.1177/0363546506287365© 2006 American Orthopaedic Society for Sports Medicine

Vol. 34, No. 9, 2006 Meniscal Allograft 1465

until advanced changes on radiographs are present.28 Theability to better identify early signs of articular degenera-tion in a noninvasive manner would improve our ability torecommend reconstructive procedures in a timely fashion.

Our primary hypothesis was that a more anatomicallybased meniscal allograft transplantation procedure wouldbe effective in reducing the degree of joint degenerationcompared with meniscectomy as demonstrated by MRI, T2mapping, biomechanical testing, and histologic analysis.Our secondary hypothesis was that measurements of T2relaxation time could be used as a noninvasive measure ofearly cartilage degeneration in the postmeniscectomymodel and that it could be used to monitor cartilage protec-tion in the meniscal allograft transplantation model. Thus,the purposes of this study were (1) to assess the chon-droprotective effects of meniscal allograft transplantationin the sheep postmeniscectomy model and (2) to evaluatethe utility of T2 relaxation time mapping as a noninvasivemeasure of early articular cartilage degeneration.

MATERIALS AND METHODS

After Institutional Animal Care and Use Committeeapproval was obtained, 45 skeletally mature Columbian XRambouillet ewes weighing approximately 70 to 80 kg wereallocated for use. On arrival at the Colorado State UniversityVeterinary Teaching Hospital, the sheep were placed in thelarge-animal research barn. They were dewormed and eartagged for identification. Physical examination was per-formed, and only healthy animals with no clinically evidentmedical or orthopaedic problems were selected for the study.

Thirty nonstudy animals were allocated for meniscal allo-graft donation. These animals had been involved in unre-lated research involving the spine or cranium and werescheduled for sacrifice on the day before meniscal allografttransplantation. During the harvest procedure, the entirelateral meniscus, including the full length of the anteriorand posterior meniscal ligaments with a small amount ofassociated bone, was aseptically removed from the donorknees and stored in an antibiotic/nutrient medium accord-ing to the guidelines of the American Association of TissueBanks.12 All donor knee menisci were size matched by theweight of the sheep as well as precise intraoperative meas-urements of length, width, and height of the meniscus. Theallografts were stored in the antibiotic/nutrient medium fora maximum of 48 hours before transplantation. The animalshad a high degree of genetic similarity, thus decreasing therisk for rejection of the tissue transplant.

STUDY GROUPS

The animals were randomly assigned to 1 of 3 groups. Fouranimals underwent a sham operation involving exposure ofthe lateral joint with a release and repair of the lateral col-lateral ligament and popliteus insertion. Twenty-four ani-mals underwent a lateral meniscectomy after exposure ofthe lateral compartment. Seventeen animals underwentlateral meniscal allograft transplantation immediatelyafter meniscectomy with size-matched donor menisci. For

the sham and the meniscectomy groups, half of the animalswere sacrificed at 2 months, and the other half were sacri-ficed at 4 months. For the meniscal allograft transplanta-tion group, 8 were sacrificed at 2 months, 8 were sacrificedat 4 months, and the final animal was sacrificed at 1 year.

SURGICAL INTERVENTIONS

Meniscectomy

Before surgery, 10 sheep cadaveric knees were dissected tobetter define the surgical anatomy (Figure 1).1 A lateralincision was made over the lateral collateral ligament of theknee joint. The femoral attachment of the lateral collateralligament and popliteus insertion site was elevated using anosteotome along with a small wedge of bone. The lateral col-lateral ligament and popliteus were then reflected distallyto expose the entire lateral meniscus. Varus stress wasapplied to the knee joint to fully expose the meniscus. Theperipheral rim of the meniscus was dissected sharply fromthe capsule, dividing the coronary ligament. The anteriorand posterior horn attachments of the meniscus were thendetached, and the meniscus was removed from the knee.Special care was taken to avoid any articular cartilagedamage during the lateral meniscectomy. With detachmentof the anterior meniscotibial ligament attachment, specialcare was taken to preserve the tibial insertion of the ACL,which lies in close proximity. With detachment of the poste-rior meniscofemoral ligament, the popliteal artery wascarefully retracted to avoid vascular injury. The osseousfemoral attachment of the lateral collateral ligament andpopliteus was then reattached to the femur using a 4.0-mmcancellous bone screw. After thorough irrigation of the jointwith saline solution, the joint capsule, overlying fascia ofthe tensor fascia lata, subcutaneous tissues, and skin wereclosed in layers.

Allograft Transplantation

Before the transplant procedure was performed in live ani-mals, 10 sheep cadaveric knees were dissected. A surgicalprocedure was developed using anatomical placement of theanterior meniscotibial and posterior meniscofemoral attach-ment sites (Figure 1) with fixation through bone tunnels. Atthe start of the transplant procedure, lateral meniscectomieswere performed using the technique described above. Thepreviously harvested menisci for transplantation were thenwashed thoroughly with normal saline solution before trans-plantation. Accurate size matching was performed using 5measurements: length from anterior horn to the posteriorhorn; width at the anterior horn, posterior horn, and centralmeniscus; and central height. The allograft meniscus wasthen prepared for transplantation by placing No. 2 Fiberwire(Arthrex, Naples, Fla) Krackow stitches through the anteriorand posterior meniscal ligaments. The anterior meniscotibialattachment site was then exposed, and a drill tunnel wasmade entering at the anatomical insertion site and exitingthe tibia on the medial side of the knee. The posterior menis-cofemoral attachment site was then palpated, and a second

1466 Kelly et al The American Journal of Sports Medicine

drill tunnel was made entering the anatomical insertion siteon the posterior aspect of the medial femoral condyle andexiting the femur on the medial side of the knee. The ante-rior and posterior meniscal allograft ligaments were thenanchored to the anterior and posterior meniscal insertionsites by passing the Fiberwire sutures through the bone tun-nels. The sutures were then fixed over buttons on the medialside of the knee. Tracking and appropriate sizing of themeniscus were confirmed after fixation. The outer rim of themeniscus was then reattached to the lateral capsule using

2 or 3 interrupted absorbable sutures. At the conclusion ofthe transplant fixation, the osseous femoral attachment ofthe lateral collateral ligament was reattached to the femurusing a 4.0-mm cancellous bone screw, and the wound wasclosed in layers, as described above.

Sham Operation

The sham operation was performed using the sameapproach to the lateral compartment required for themeniscectomy procedure and transplant procedure. In thisgroup, no further surgery was performed after exposure ofthe joint.

POSTOPERATIVE CARE

After surgery, postoperative activity was restricted for thefirst 2 weeks by housing the animals in closed confinement.The limbs were not immobilized. At 2 months, 2 animalsfrom the sham group, 12 animals from the meniscectomygroup, and 8 animals from the allograft group were sacri-ficed. Both the operated and nonoperated limbs were har-vested for analysis. At 4 months, the remaining animalsin the sham and meniscectomy groups and all but 1 of theremaining transplant animals were sacrificed, and bothlimbs were sent for analysis (2 sham, 12 meniscectomy,8 allograft).The final allograft animal was sacrificed at 1 year.

POSTSACRIFICE ANALYSIS

After sacrifice, both the operative and nonoperative kneeswere detached at the level of the midfemur and midtibia.All but 7 of the limbs were immediately frozen and storedfor analysis. The remaining 7 limbs were kept fresh for cellviability analysis of the allograft menisci using a paravitalstaining technique. The time between euthanasia andanalysis of the frozen limbs was between 5 and 7 days. Thefresh limbs were evaluated within 24 hours after sacrifice.The frozen knees were thawed before analysis and thenassessed sequentially with cartilage-sensitive MRI, MRIwith T2 mapping, gross inspection with india ink staining,biomechanical testing of cartilage stiffness, and semiquan-titative histology. Analysis of the vascularity of the trans-planted menisci was performed using the Spälteholztechnique on 3 of the specimens from the 4-month trans-plant group. Complete analysis of each limb was performedwithin a 24-hour period to avoid any potential degenera-tive effects from exposure to air. The specimens were keptrefrigerated between tests during that 24-hour period, andall tests were performed in a temperature-controlled labo-ratory set at 65° F. None of the tests resulted in heating ofthe tissue. All limbs were treated with the same protocol,eliminating the potential for any significant differences dueto prolonged exposure. Seven of the limbs (5 at 2 monthsand 2 at 4 months) were delivered fresh and analyzedwithin 24 hours of sacrifice. In addition to the routine analy-ses described above, the fresh menisci from these animalswere evaluated for cell viability using paravital stainingtechniques.

Figure 1. Normal anterior tibial (A) and posterior femoral (B)meniscal attachment sites.


MAGNETIC RESONANCE IMAGING

All animals were imaged in a clinical 1.5-T superconductingmagnet (Signa Horizon LX, General Electric MedicalSystems, Milwaukee, Wis) with a 5-inch, curved, receive-onlylinear shoulder coil (IGC-Medical Advances, Milwaukee,Wis). Animals were imaged using a fast spin echo sequencein the coronal and sagittal planes with a repetition time of4200 to 5000 milliseconds, effective echo time of 17 millisec-onds, 31.2-kHz receiver bandwidth, and field of view of 8 × 6cm with a matrix of 512 × 384, yielding an in-plane resolu-tion of 156 µm × 156 µm × 1 mm slice resolution. Imageswere obtained with 3 excitations. Tailored radiofrequency(GE Health Care, Milwaukee, Wis) was used to reduceinterecho spacing, with echo train lengths ranging between8 and 12.

Subsequent spin echo T2 relaxation maps were obtainedusing repetition time of 500 to 650 milliseconds; echo timesof 11, 40, and 80 milliseconds; at one excitation. Field ofview was 8 × 6 cm with a matrix of 256 × 160, yielding anin-plane resolution of 312 µm × 375 µm with 2-mm slices.Receiver bandwidth was 15.6 kHz (over the entire fre-quency range). T2 relaxation maps were then recon-structed using a monoexponential fit model to calculate T2relaxation pixel by pixel (Functool 3.1.10, AdvantageWindows work station, GE Health Care).

The fast spin echo images were analyzed to assess mor-phologic changes in cartilage, bone, and bone marrowedema, as well as the presence or absence of osteophytes.Maximum cartilage thickness in the nonosteoarthriticsheep model was 0.8 mm based on MRI measurements. AnMRI score was calculated based on the observed changes incartilage, subchondral bone, and bone marrow edema, aswell as the presence or absence of osteophytes. A similarsystem for cartilage degeneration has been previously vali-dated in a clinical knee study using arthroscopy as thestandard.27 Cartilage was scored at the anterior, central,posterior, and peripheral margin over the lateral condyleand lateral plateau using a 1 to 4 grading system: 1,increased signal intensity; 2, <50% cartilage loss; 3, >50%cartilage loss; 4, exposed subchondral bone. The scoring wasperformed by a single attending MRI radiologist who wasblinded to the surgical procedure. A sample of the imageswere scored a second time to confirm intraobserver relia-bility; no significant differences were found between scoringsessions. The subchondral bone was also assessed in each ofthese locations for the presence of sclerosis, scoring 1 (noneapparent), 2 (mild), 3 (moderate), and 4 (severe). Bone mar-row edema and osteophyte formation were both scored asabsent (0) or present (1). A total MRI score for the lateralcompartment was calculated by summing the mean valuesfor each of the 4 categories. With this system, the minimumvalue of 2 corresponded to normal joint characteristics withno evidence of degeneration, whereas the maximum score of10 corresponded to advanced degenerative changes.

Spin echo T2 maps of the lateral tibial plateau werecreated in this study to reduce the potential error introducedby stimulated echo formation generated in most multiechofast spin echo techniques, yielding additive T1 contrast.20

Quantitative T2 values were then obtained from the anterior,

central, and posterior regions of the central weightbearingzone of the lateral tibial plateau using a standardized regionof interest analysis. These areas of interest corresponded tothe areas in which maximal cartilage degeneration was seenin all animals due to the loading characteristics of the post-meniscectomy ovine knee. They also corresponded to the3 areas of interest identified for mechanical testing usingindentation probe testing. Subsequent color maps were gener-ated using the same Functool program in which T2 valuestratification was depicted as prolonged values in green to blueand shorter values in orange to red, using an expanded colorscale stratified from a minimum echo time of 0 milliseconds toa maximum echo time of 150 milliseconds.

GROSS INSPECTION

After MRI, all limbs were dissected, and the tibial plateauswere cleared of all surrounding soft tissue. The cartilage sur-faces were stained with a dilute (1%) india ink solution(Higgins waterproof black india ink), and all specimens werephotographed for determination of gross evidence of cartilagedegeneration as demonstrated by uptake of india ink stain.Average stain areas overlying the tibial plateaus were meas-ured using a Metamorph computer analysis (square mil-limeters; Universal Imaging Corporation, Downington, Pa).

BIOMECHANICAL TESTING

Cartilage stiffness was measured over the central weight-bearing zone of the lateral tibial plateau. The area testedrepresented the area of greatest wear in the postmeniscec-tomy knees. Three measurements were collected from 3 dif-ferent locations along the central contact area of the tibia:anterior, central, and posterior. The location of these datapoints was the same in all specimens and was verified bydividing the central weightbearing zone into thirds andfinding the center point of each section. The location ofthese data points also corresponded to the areas sampledfor the T2 analysis. The 3 data points from each locationwere averaged. Stiffness was measured using a cartilageindentation probe specifically designed for thin cartilage(Artscan Inc, Helsinki, Finland). The probe consisted of ameasurement rod with a 100-µm-diameter indenter locatedin the center of a reference plate at the end of the probe.The value of the indenter force (newtons) reflects the forcethat the cartilage exerts against the indenter and can reli-ably be used as an index of cartilage stiffness. Several stud-ies have demonstrated the association between cartilagedegeneration and decreased cartilage stiffness.2,6,11,17,19,26,30

HISTOLOGIC ANALYSIS

Histologic sections were subsequently prepared and scoredby an attending pathologist blinded to the procedure per-formed.The specimens were sectioned in the coronal plane atthe midpoint of the weightbearing zone of the tibial plateau.The osteochondral specimens were then fixed in 10% neutralbuffered formalin (Sigma Diagnostics, St Louis, Mo) and then


decalcified in 5% nitric acid (50 mL concentrated nitric acidto 950 mL dH2O). The tissues were checked daily; as soon asthe decalcification was complete, the tissues were removed,so as to avoid overdecalcification. Tissue processing andparaffin embedding were then performed using the TissueTek VIP1000 tissue processor. Five-micrometer-thick sec-tions were cut and then stained with H&E and safranin O.The histologic sections were taken from the middle third ofthe tibia and were directed from the lateral edge (peripheralzone) to the intercondylar notch (central zone). The centralzone corresponded to the area of greatest wear that was eval-uated with the biomechanical testing.

The histologic sections were graded using the modifiedMankin grading scale for hyaline cartilage degenera-tion.21,32 This semiquantitative analysis assessed cartilagestructure (0-6), cellular abnormalities (0-3), matrix staining(0-4), and tidemark integrity (0-1). A minimum score of0 denotes no cartilage degeneration, and a maximum scoreof 14 indicates severe cartilage destruction. Regional differ-ences between the central zone and peripheral zone wereevaluated.

Polarized light microscopy was performed to furtherassess collagen organization within 4 layers of the matrix:superficial layer, transitional zone, radial zone, and calci-fied zone. The collagen was scored as organized (0) or dis-organized (1) in each of these 4 zones, resulting in a scorerange between 0 and 4.

Paravital staining to assess cell viability of the trans-planted menisci was performed on 7 fresh specimens (5 fromthe 2-month allograft transplant group and 2 from the4-month allograft transplant group).24 A full detail of thetechnique for cell viability paravital staining has been pre-viously reported.18 The paravital technique was made up ofa combination of 2 indicators. Propidium iodide is a cellimpermeable dye that labels nucleic acids only in dead cellswith a bright red indicator. Fluorescein diacetate is a cellpermeable dye that labels cytoplasmic activity of live cellswith a green indicator. The staining technique results in aclear contrast between viable (green indicator) and nonvi-able (red indicator) cells.

Vascularity of the transplanted menisci was assessed in3 of the 4-month allograft transplant animals using stan-dard Spälteholz technique.3 After MRI was completed inthese specimens and before gross dissection, the limbswere injected with india ink solution (Higgins waterproofblack india ink) via the femoral artery under maximalmanual pressure to perfuse the entire knee joint. The ves-sels distal to the knee were occluded by ligation andtourniquet application. A minimum of 600 mL of india inkwas injected into each specimen. Capillary perfusion withindia ink was demonstrated by visible black mottling ofthe skin. The specimens were then frozen, and 3-mm sagit-tal and coronal sections were cut using a band saw. Thesections were then cleared by a modified Spälteholz tech-nique3,4 involving (1) fixing the sections in 10% bufferedformalin for 3 days, (2) decalcification using 5% nitric acidfor 3 days (changing the solution daily), (3) dehydration inethyl alcohol for 6 days, and (4) defatting in chloroform for2 days. The tissue was then cleared by changing its refrac-tive index with the Spälteholz solution. This solution is

made up of 3 parts benzyl benzoate and 5 parts methylsalicylate. The specimens were then visualized under directmagnification and photographed using high-resolution filmfor further review.

STATISTICAL METHODS

Means and standard deviations were calculated for allmeasurements. A paired Wilcoxon nonparametric test wasused to make direct comparisons between sham operationanimals and nonoperated control animals, between the cen-tral tibial plateaus and peripheral tibial plateaus of theallograft transplant animals, between the 2-month menis-cectomy and 4-month meniscectomy groups, and betweenthe 2-month allograft transplant and the 4-month allografttransplant groups. Significant differences (P < .05) betweenallograft animals, meniscectomy animals, and nonoperatedcontrol animals were calculated with a 1-way analysis ofvariance followed by t tests with a Bonferroni correction ora Kruskal-Wallis test followed by Mann-Whitney tests.Correlations between T2 data and all other variables werecalculated with Spearman rank correlation (r2 values) testsfor nonparametric data. The collected data were analyzedand interpreted by each of the investigators to ensure accu-racy, and all data were analyzed in a blinded fashion.

RESULTS

Operative Observations

At the time of the initial surgery, there was no evidence ofgross cartilage degeneration within the lateral compart-ment of the joint. Complete visualization of the lateral tib-ial plateau was feasible in all operative specimens usingthe approach described.

Sham Operation

No statistical differences were found between any of thesham operation animals and the nonoperated controllimbs with regard to MRI, T2 mapping, gross inspectionwith india ink, biomechanical testing, and histologicanalysis. Thus, the nonoperated limbs were subsequentlyused as the control group in all comparisons between themeniscectomy and allograft groups.

STUDY GROUPS

Postmeniscectomy Animals

There were statistically significant differences (P < .05)between the nonoperative control limbs and the meniscec-tomy limbs in all outcome measures at both 2 months and4 months after meniscectomy (MRI, T2 mapping, grossinspection with india ink, biomechanical testing, and his-tologic analysis) (Tables 1 and 2).

Magnetic Resonance Imaging. The MRI appearance of thenonoperated control limbs demonstrated good preservation


of the lateral compartment joint space, no cartilage wear(mean cartilage thickness was approximately 0.8 mm), nosubchondral sclerosis or edema, and no peripheral osteo-phyte formation (mean score, 2.0) (Figure 2A). By2 months after meniscectomy, MRI demonstrated moderatecartilage degeneration, with significant focal cartilage wearover the central weightbearing zone of the tibial plateau,mild to moderate subchondral sclerosis, and the presenceof osteophytes and subchondral edema in the majorityof the specimens (mean score, 7.1 ± 0.6). By 4 months,advanced degeneration was clearly seen with large areas offull-thickness cartilage wear, significant sclerosis, and thepresence of osteophytes and edema in all specimens (meanscore, 8.4 ± 0.7) (Figure 2B).

T2 Mapping. T2 maps demonstrated significant prolonga-tion (P > .05) of T2 relaxation times in the lateral tibialplateau after meniscectomy compared with nonoperatedcontrols at both 2 months (Figure 3) and 4 months (Figure 4;Tables 1 and 2). The mean T2 relaxation time for the non-operated controls was 38.7 ± 10.0 milliseconds. There was nosignificant difference between T2 relaxation times in thecontrol animals at 2 months (41.6 ± 7.0) compared with4 months (35.7 ± 13.0). At 2 months after meniscectomy, themean T2 relaxation time had increased to 57.2 ± 11.1 milli-seconds, and by 4 months after meniscectomy, it was alsoelevated at 55.1 ± 13.0 milliseconds. This finding confirmedthe osteoarthritis noted on traditional MRI, with increased

signal intensity seen at both 2 months and 4 months com-pared with control limbs. There was no significant differencebetween the increased T2 relaxation time seen at 2 monthsand 4 months, suggesting that a threshold of cartilagedegeneration was reached by 2 months that was beyond thelimits of resolution of the T2 mapping to detect further car-tilage deterioration. Little to no stratification of T2 relax-ation times was noted over the plateau, likely because of theadvanced degree of degeneration and the limitations ofthe in-plane resolution of the T2 mapping sequence, giventhe time requirements for obtaining 3 echo samples.

In an attempt to identify correlations between the T2mapping and the other outcome measures, Spearman rankcorrelations were calculated and subjected to post hocanalysis. There was no assumption of a linear relationshipbetween variables at 2 months and 4 months (Table 3).Significant correlations were identified between prolonga-tion of T2 relaxation times and all other outcome measuresat 4 months (P < .05). At 2 months, significant correlationswere identified between prolongation of T2 relaxation timesand all other outcome measurements except the morpho-logic MRI score and the india ink staining area (P < .05).

Gross Inspection. Representative gross specimens at4 months stained with 1% dilute india ink are seen inFigures 5 A and B. The nonoperated control limbs (Figure5A) demonstrated minimal staining over the lateralplateau at 4 months (mean stain area, 3.0 ± 6.2 mm2). This

TABLE 1Evaluation of Tibial Plateaus in Postmeniscectomy Animals and Nonoperated Controls at 2 Months After Surgerya

Meniscectomy Control

Variable n Mean SD n Mean SD P

MRI score 12 7.1 0.6 12 2.0 0.0 <.0001T2 map, ms 6 57.2 11.1 6 41.6 7.0 .01India ink, mm2 12 36.1 7.9 12 3.0 6.2 <.0001Biomechanics, N 12 0.37 0.13 12 0.63 0.08 .0001Mankin 12 7.3 1.5 12 0.25 0.45 <.0001Polarized 12 2.2 0.8 12 0.0 0.0 <.0001

aModified Mankin scores are between 0 and 14; polarized light scores are between 0 and 4; MRI scores are between 2 and 10. Highlysignificant differences were seen in all outcome measures (P < .05).

TABLE 2Evaluation of Tibial Plateaus in Postmeniscectomy Animals and Nonoperated Controls at 4 Months After Surgerya

Meniscectomy Control


MRI score 12 8.4 0.7 12 2.0 0.0 <.0001T2 map, ms 6 55.1 13.0 6 35.7 13.0 .004India ink, mm2 12 46.7 13.0 12 2.8 7.0 <.0001Biomechanics, N 12 0.25 0.08 12 0.63 0.08 <.0001Mankin 12 11.3 1.6 12 0.67 0.89 <.0001Polarized 12 3.5 0.8 12 0.0 0.0 <.0001

aModified Mankin scores are between 0 and 14; polarized light scores are between 0 and 4; MRI scores are between 2 and 10. Highlysignificant differences were seen in all outcome measures (P < .05).


finding was consistent across all animals, with virtually noevidence of baseline cartilage degeneration in the lateralplateau. At 2 months after meniscectomy, there was a sig-nificantly increased stain consistent with moderate carti-lage degeneration (mean stain area, 36.1 ± 7.9 mm2). By 4months, advanced cartilage wear was evident by grossinspection (mean stain area, 46.7 ± 13.0 mm2) (Figure 5B).

Biomechanical Testing. The central weightbearing zoneof the tibial plateau represented the area of most rapiddegeneration and was the area that was focused on for bio-mechanical and histologic testing. The biomechanical test-ing was performed in the anterior, central, and posteriorregions of the weightbearing zone depicted by the areas of

Figure 2. A, magnetic resonance imaging of the nonoperatedcontrol limbs demonstrated no evidence of cartilage wear orsubchondral bony changes. B, by 4 months after meniscec-tomy, advanced changes were seen in all specimens, includ-ing subchondral depression, edema pattern, and markedthinning of the cartilage.

Figure 3. Quantitative T2 relaxation time maps comparing thenonoperated control (A) and the 2-month postmeniscectomy(B) tibial plateau. The color maps are coded to capture T2 val-ues ranging from 0 to 150 milliseconds, with green and bluereflecting longer T2 values, yellow intermediate values, andorange/red the shorter values. The apparent prolongation inT2 values to the posterior margin (green regions) is owing tothe magic angle effect, yielding the expected prolongation ofT2 when the collagen is oriented close to 55° relative to themagnetic field. Note diffuse prolongation of T2 values over theanterior margin (arrow; blue region in B) after meniscectomy(B) compared with the nonoperated control (A).


highest stain uptake (Figure 5B). There was a significantdecrease in cartilage stiffness at both 2 months (0.37 ± 0.13 N)and 4 months (0.25 ± 0.08 N) compared with controls

(0.64 ± 9 N). Significant differences between the 2- and4-month animals were also detected (P < .05). Biomechanicaltesting of 4-month meniscectomy knees demonstrated sig-nificantly decreased articular cartilage stiffness over theentire lateral tibial plateau (anterior, central, and posteriordata points) compared with controls (P = .006), whereas at2 months, stiffness was significantly decreased over theanterior lateral tibial plateau only (P = .01).

Histologic Analysis: H&E. Histology specimens were cutin the coronal plane and extended from the periphery ofthe tibial plateau to the central (intercondylar notch)region at the midpoint of the tibial plateau. Normal artic-ular cartilage was seen in all nonoperated controls andsham animals and was consistent with the absence of anybaseline cartilage wear over the lateral plateau (meanmodified Mankin score was 0.46 ± 0.67). At 2 months,histologic evidence of moderate cartilage degenerationwas seen (modified Mankin score was 7.3 ± 1.5), and by4 months, advanced cartilage degeneration was confirmed

Figure 4. Quantitative T2 relaxation time maps of the lateral tibial plateaus comparing the nonoperated control (A), the4-month postmeniscectomy (B), and the 2-month (C) and 4-month (D) postmeniscal allograft transplantation animals. The colormaps are coded to capture T2 values ranging from 0 to 150 milliseconds, with green and blue reflecting longer T2 values, yellowintermediate values, and orange/red the shorter values. Note diffuse prolongation of T2 values after meniscectomy (B) comparedwith the nonoperated control (A), particularly over the anterior margin (arrow). There is comparatively diffuse prolongation of T2relaxation times (arrows) 4 months after meniscal allograft transplantation (D) compared with the 2-month allograft specimen(C), which cannot be attributed to the magic angle phenomenon.

TABLE 3Correlation of T2 With Other Variables

at 2 Months and 4 Monthsa

2 Months 4 Months

Variable r2 P r2 P

MRI 0.37 .07b 0.63 <.05India ink 0.43 .06b 0.55 <.05Biomechanics –0.58 <.05 –0.67 <.05Mankin 0.69 <.05 0.51 <.05Polarized 0.64 <.05 0.55 <.05

aSpearman rank correlation was used, nonlinear.bNot significant.


in all specimens (mean modified Mankin score was11.2 ± 1.6). Characteristic histologic findings included dis-organization of cartilage structure with fissure formation,hypocellularity, reduced matrix staining, and loss of tide-mark integrity.

Histologic Analysis: Polarized Light Microscopy.Polarized light microscopy was used to further evaluatethe collagen organization. The control animals demon-strated normal cartilage with polarized light analysis,which was reflected by maintenance of collagen organiza-tion within the 4 layers of the matrix: superficial layer,transitional zone, radial zone, and calcified zone (polarizedlight score of 0). Significant loss of collagen organizationwas seen at both 2 months (polarized light score of 2.2 ±0.8) and 4 months after meniscectomy (polarized lightscore of 3.5 ± 0.8). Polarized light scores demonstrated sig-nificantly decreased collagen organization at 4 monthscompared with 2 months (P = .002). Collagen disorganiza-tion was evident in the superficial and transitional zonesby 2 months, with extension into the deeper layers (radialzone and calcified zone) at 4 months.

Allograft

There were statistically significant differences between theallograft limbs and the meniscectomy limbs in all objectivecategories at 2 months (MRI, T2 mapping, gross inspectionwith india ink, biomechanical testing, and histologic analy-sis). At 4 months, allograft limbs demonstrated significantdifferences in all objective categories except the T2mapping (Table 4). In comparing the allograft limbs to thenonoperated control limbs, no significant differences wereseen at 2 months in any of the categories except the MRIdata. At 4 months, the nonoperated control limbs demon-strated significantly less wear compared to the allograftlimbs in all categories except the modified Mankin scores(Table 5). One animal was sacrificed at 12 months after sur-gery. The scores in all categories for this animal remainedbetter than the mean scores at 2 months in the postmenis-cectomy animals.

Magnetic Resonance Imaging. The MRI appearance of theallograft limbs demonstrated significant improvements inall morphologic criteria compared with the meniscectomylimbs at both 2 and 4 months, although there was a statis-tically significant increase in the MRI score between 2 and4 months (2.7-4.2, P < .05). Compared with the nonoperatedcontrol limbs, however, MRI scores were significantly worseat both 2 and 4 months after allograft transplantation(Table 5). Magnetic resonance imaging further evaluatedthe transplanted meniscus with regard to intrameniscalsignal, extrusion in the sagittal and coronal planes, andmeniscal morphologic characteristics. At 2, 4, and 12 months,the transplanted allografts demonstrated minimal intra-meniscal signal, minimal extrusion in either plane, andpreservation of normal meniscal shape. The allografts werehealed at the capsular attachment site. Several specimensdemonstrated evidence of fibrous tissue ingrowth into theanterior and posterior meniscal-ligament attachment sites.No meniscal tears were seen on MRI.

Figure 5. Gross inspection with 1% dilute india ink compar-ing the nonoperated control tibial plateau (A), the 4-monthpostmeniscectomy tibial plateau (B), and the 4-month menis-cal allograft transplant tibial plateau (C).


T2 Mapping. T2 maps demonstrated improved mainte-nance of T2 relaxation times at 2 months and 4 monthscompared with meniscectomy (Table 4 and Figure 4). Thisimprovement was statistically significant at 2 months butnot at 4 months. This finding suggests improved preserva-tion of collagen organization resulting from the meniscustransplant and is correlated with the findings observedunder polarized light microscopy. There were no significantdifferences in the T2 maps between the nonoperated con-trol limbs and the allograft limbs at 2 months. However, by4 months, the allograft limbs demonstrated a significantprolongation in T2 relaxation times compared with thenonoperated controls (Table 5).

Gross Inspection. A representative gross specimen4 months after allograft transplantation is depicted inFigure 5C. Compared with the meniscectomy limbs, therewere significantly decreased india ink stain areas over the lat-eral plateau at both 2 and 4 months. However, there was asignificant increase in stain area between 2 and 4 months

(6-18 mm2, P < .05).The wear pattern, as evidenced by the indiaink staining, demonstrated additional wear along the periph-ery of the tibial plateau in some specimens. This wear patterndiffered from the meniscectomy animals, in which the wear wasisolated to the central weightbearing zone. At 2 months, therewas no difference in the gross appearance between the allograftanimals and control animals (Table 5). By 4 months, however,the allograft animals demonstrated significantly increasedstain area compared with the control animals.

Biomechanical Testing. Biomechanical testing was per-formed in the same locations along the central weight-bearing zone of the lateral plateau as was done for themeniscectomy animals. There was significantly improvedmaintenance of cartilage stiffness at both 2 and 4 monthscompared with the meniscectomy animals (Table 4); how-ever, stiffness values were significantly lower at 4 monthscompared with 2 months (0.4 N vs 0.6 N, P < .05). Similarto the other outcome measures, 2-month allograft limbsdemonstrated no differences compared with control limbs,

TABLE 4Allograft Transplant and Postmeniscectomy Outcome Measures Evaluating

the Tibial Plateaus at 2 and 4 Months After Surgerya

2 Months 4 Months

Meniscectomy Allograft Meniscectomy Allograft

Variable n Mean SD n Mean SD P n Mean SD n Mean SD P

MRI 12 7.1 0.6 7 2.7 1.0 <.001 12 8.4 0.7 5 4.2 0.9 <.001T2 map, ms 12 57.2 11.1 5 46 2 .02 11 55.1 13.0 5 50 5 .08India ink, mm2 12 36.1 7.9 8 6 7 <.001 12 46.7 13.0 8 18 6 <.001Biomechanics, N 12 0.37 0.13 8 0.6 0.1 <.001 12 0.25 0.08 8 0.4 0.06 .001Mankin 12 7.3 1.5 8 0.3 1.0 <.001 12 11.3 1.6 8 0.7 1.0 <.001Polarized 12 2.2 0.8 8 0 0 <.001 12 3.5 0.8 8 0.9 0.7 <.001

aModified Mankin scores are between 0 and 14; polarized light scores are between 0 and 4; MRI scores are between 2 and 10. All outcomemeasures demonstrated significant differences between the 2 groups except the T2 mapping at 4 months (P < .05).

TABLE 5Allograft Transplant and Nonoperated Control Outcome Measures Evaluating the Central

Weightbearing Zone of the Tibial Plateaus at 2 and 4 Months After Surgerya

2 Months 4 Months

Control Allograft Control Allograft

Variable n Mean SD n Mean SD P n Mean SD n Mean SD P

MRI 12 2.0 0.0 7 2.7 1.0 .045 12 2 0 5 4.2 0.9 .01T2 map, ms 6 41.6 7.0 5 46 2 .3 6 29 3 5 50 5 .005India ink, mm2 12 3.0 6.2 8 6 7 .4 12 3 7 8 18 6 .01Biomechanics, N 12 0.63 0.08 8 0.6 0.1 .9 12 0.6 0.09 8 0.4 0.06 .01Mankin 12 0.25 0.45 8 0.3 1.0 .7 12 0.67 0.89 8 0.7 1.0 .9Polarized 12 0.0 0.0 8 0 0 .9 12 0 0 8 0.9 0.7 .01

aModified Mankin scores are between 0 and 14; polarized light scores are between 0 and 4; MRI scores are between 2 and 10. At 2 months,the MRI score was the only outcome measure that was statistically different from the control animals. At 4 months, all values except themodified Mankin scores were significantly worse compared with the control animals (P < .05).


but by 4 months, the allografts were significantly worsecompared with controls (Table 5).

Histologic Analysis: H&E. Compared with the menis-cectomy animals, there were statistically significantimprovements in the modified Mankin scores over the cen-tral weightbearing zone at both 2 months and 4 months.Excellent preservation of the cartilage structure, cellular-ity, matrix staining, and tidemark integrity was observedin this central zone. However, there was significantly morecartilage wear in the peripheral zone beneath the allograftmeniscus compared with the central weightbearing zone ofthe tibial plateau (Table 6). This finding was true for themodified Mankin scores at both 2 and 4 months (P = .007and P = .02, respectively). Scores along the periphery of thetibial plateau for the allograft animals were still lower thanwere the Mankin scores for the meniscectomy animals inthe central weightbearing zone. No significant differenceswere noted between the Mankin scores at 2 months and 4months in the allograft animals. In comparing the centralzone histologic findings of the meniscal allograft animalsand the nonoperated control animals, no significant differ-ences were noted at 2 or 4 months (Table 5).

Histologic Analysis: Polarized Light. Polarized lightmicroscopy demonstrated maintenance of collagen organiza-tion after allograft transplantation at both 2 and 4 monthscompared with the meniscectomy animals (Table 4); how-ever, there was loss of collagen organization at 4 months com-pared with 2 months (P < .05), thus correlating with theprolonged T2 relaxation times noted in the 4-month allograftgroup. Compared with the control limbs, polarized lightscores were not significantly different at 2 months, but by4 months, allograft limbs demonstrated significantlydecreased collagen organization. In comparing the periph-eral zone to the central weightbearing zone of the tibialplateau, the polarized light scores were significantly differ-ent at 2 months only (P = .006) (Table 6).

Allograft Explant Analysis

All allograft explants were evaluated by gross examination(Figure 6). There was tissue ingrowth into the allograft at

the anterior and posterior attachment sites, as well asalong the capsular periphery. The allografts appeared to befirmly fixed within the joint, with no evidence of lateralextrusion or any evidence of rupture of the attachmentsites. No meniscal tears were seen at any location alongthe length of the allograft at any of the sacrifice timepoints (2, 4, and 12 months).

Routine histologic examination of the meniscus was per-formed on 3 of the specimens and demonstrated normal-appearing fibrocartilage that was indistinguishable fromthe native medial meniscus. Histologic evaluation of theanterior and posterior attachment sites demonstratedfibrous tissue loosely integrated into the recipient bone.Cell viability of the allograft menisci was evaluated usingparavital staining techniques in 7 allograft specimens (5 at2 months and 2 at 4 months). These specimens were

TABLE 6Comparison of the Regional Histology (Modified Mankin Scores and Polarized Light Scores) of the

Central Zone and Peripheral Zone After Meniscal Allograft Transplantation at 2 and 4 Monthsa

Central Zone Peripheral Zone


2 moMankin 9 0.3 1.0 9 3.78 1.79 .007Polarized 9 0 0 9 1.33 0.5 .006

4 moMankin 7 0.7 1.0 7 3.86 1.52 .02Polarized 7 0.9 0.7 7 1.29 0.49 .08

aModified Mankin scores are between 0 and 14; polarized light scores are between 0 and 4. Significantly greater wear was noted at theperiphery compared with the central zone by modified Mankin scores at both time points. The polarized light scores were significantly worseat the periphery for the 2-month animals only (P < .05).

Figure 6. Gross examination of the allograft transplantsdemonstrated good fixation at the anterior and posterior lig-ament insertion sites, as well as along the capsular periphery.Good preservation of the central weightbearing cartilage onthe tibial plateau was also seen by gross examination.


evaluated fresh and not subjected to a freeze-thaw cycle aswere the other specimens. At both 2 and 4 months, therewas greater than 90% cell viability in all allograft menisciexamined. Again, the allograft tissue was indistinguish-able from the native medial meniscus. Evaluation of thevascularity of the allograft meniscus was assessed usingthe Spälteholz technique in 3 of the 4-month specimens.The nonoperated limbs of 2 animals were also assessedwith the vascular injection study to determine the vascu-larity of the native lateral meniscus. Extensive vascularingrowth was observed at both the anterior and posteriorattachment sites as well as along the periphery. The allo-grafts were hypervascular compared with the nativemeniscus (Figure 7).

DISCUSSION

Currently, there is no consistently reproducible animalmodel for meniscal allograft transplantation in the litera-ture. Previous animal models have demonstrated variableresults; however, these earlier studies have not empha-sized anatomical reconstruction of the native meniscalinsertion sites, nor have they used bone tunnel fixation as

is typically performed in humans. Arnoczky et al5 andMikic et al22 reported some degree of protection of the car-tilage after meniscal allograft transplantation in dogs;however, these findings were not compared with meniscec-tomy. Edwards et al10 found no radiographic differencesbetween meniscal transplants and those of meniscectomyat 21 months and concluded that meniscal autogenousgrafts and allografts did not protect cartilage. Cumminset al8 found that meniscal allografts offered some histologicevidence of articular cartilage protection in a rabbit model.Szomor et al31 reported perhaps the best published large-animal model for the protective effects of autograft andallograft meniscal transplantation. They demonstrated a50% decrease in cartilage wear as determined by macro-scopic grading of the cartilage; however, they found no dif-ferences between the meniscectomy and transplant groupswhen histologic criteria were examined. No further out-come measures were used in any of these studies.

The surgical technique used in this study attempted torestore the anatomical anterior tibial and posterior femoralmeniscal attachments. In our dissections and cadaverictransplant trials, re-creation of the normal, anatomicalanterior and posterior lateral meniscal attachmentsgrossly reproduced normal meniscal kinematics. Adheringto these anatomical considerations resulted in significantlydecreased cartilage wear compared with meniscectomy andmay provide a more accurate evaluation of the chondropro-tective effect of a meniscus transplant. We used a widearray of outcome measures to provide a comprehensiveevaluation of the chondroprotective effects of this surgicalprocedure. We were able to demonstrate the protectiveeffects of allograft transplantation by gross inspection, histo-logic analysis, biomechanical testing, MRI, and T2 mappingof the tibial plateau to 4 months. A single animal was sacri-ficed at 1 year with promising results; however, further stud-ies with larger numbers of long-term survival animals arenecessary. Each of these outcome measures provides usefulinformation regarding the status of the underlying cartilageand subchondral bone. Although this model demonstratedsignificant improvements compared with meniscectomy,progressive cartilage degeneration still occurred. As earlyas 4 months after allograft transplantation, we noted sig-nificantly more cartilage degeneration compared with con-trol animals in all outcomes except the modified Mankinscores, even in this “best-case scenario” of transplantationimmediately after meniscectomy.

Our outcome analyses focused on the central tibialplateau. This was the area of greatest cartilage degenerationnoted after meniscectomy. In the absence of the meniscus,the lateral femoral condyle focally loaded the central tibialplateau, resulting in rapid cartilage destruction. Theanatomical allograft transplantation resulted in a redistrib-ution of the load away from the central tibia, so that the focalwear was eliminated. Histologic analysis of the central tibiaas well as biomechanical testing of the central cartilage stiff-ness demonstrated significant “chondroprotection” aftertransplantation of the meniscus. However, we did observeincreased wear along the periphery, where the allograftmeniscus had direct contact with the peripheral tibialplateau. We believe that this increased peripheral wear was

Figure 7. Vascularity of the normal control meniscus (A) com-pared with the 4-month allograft transplant meniscus (B).


in part a result of the surgical approach (release of thelateral collateral ligament and popliteus with open exposureto the joint); however, it may have also been a result of allo-graft size mismatch and subsequent increased cartilage load-ing along the periphery. Despite the increased wear along theperiphery, this area of cartilage wear in the allograft animalswas still less than the cartilage wear seen in the centralweightbearing zone of the meniscectomy animals.

The second main purpose of this study was to evaluatethe utility of T2 relaxation time mapping as a noninvasivemeasure of early articular cartilage degeneration and todetermine if it could be used to monitor cartilage protectionin the meniscal allograft transplantation model using clini-cally relevant MRI field strengths of 1.5 T. Althoughcartilage-sensitive techniques are helpful in assessing mor-phologic changes,27 a study of healthy bovine and degenera-tive human cartilage samples indicated that routine clinicalimages do not reveal early degenerative changes, thusemphasizing the need for more sensitive MRI techniques.29

T2 mapping techniques have been shown in very high-fieldmagnetic resonance microscopy systems to reflect the struc-ture and orientation of the collagen component of the extra-cellular matrix, exploiting the depth-sensitive internucleardipolar interaction of the hydrogen nuclei.25,34

Clinically, T2 relaxation time is a quantifiable MRIparameter that at high-field strengths of 3 T has demon-strated a relationship between the water content of hyalinecartilage, as well as the relative restricted mobility of waterin the cartilage within an anisotropic solid matrix.9 Furtherwork at 3 T has demonstrated that aging is associated withan asymptomatic increase in T2 relaxation times in thetransitional zone of articular cartilage, compatible with arelative increase in water mobility.23 In this study, we wereable to demonstrate maintenance of T2 mapping times 2months after allograft transplantation compared with 2months after meniscectomy. This difference was no longersignificant after 4 months (Table 5), whereas all other out-come measures were significantly different between thetransplant and meniscectomy groups. This finding suggeststhat T2 mapping may be able to detect more minor changesin collagen organization that are not detectable with theother traditional outcome modalities and, therefore, hasimportant clinical applications. We are currently using T2mapping in select patients in an attempt to identify earlycartilage degeneration that is not appreciated with tradi-tional MRI techniques.

We found significant correlations between T2 mappingvalues and our other more traditional outcome measures(routine MRI, gross inspection, biomechanical testing, andhistology). Perhaps the most important correlation seen inthis study was between the T2 mapping and the polarizedlight scores. Polarized light microscopy evaluates collagenorganization and provides a histologic corollary to thestructure and orientation of the collagen component of theextracellular matrix reflected by the T2 mapping.

Of interest, there was a significant 30% increase in sur-face area of cartilage degeneration from 2 to 4 monthsdetected by india ink staining that was not detected by T2relaxation times. This finding suggests that a threshold ofdetectable degeneration reflected in the T2 relaxation times

was met and exceeded by 2 months after meniscectomy.This threshold is likely owing to the limits of the resolutionof T2 mapping in the context of the moderate to severedegree of cartilage degeneration seen in these animals atthe sacrifice time points analyzed. Further cartilage degen-eration beyond the 2-month time point was not reflected inadditional changes in the T2 mapping score. This is a fairlimitation of this technique for looking at differencesbetween moderate to severe cartilage degeneration as seenin the postmeniscectomy animals but does not reflect aninability of T2 mapping to detect more subtle changes inearly cartilage degeneration.

In summary, this meniscal allograft transplantationmodel demonstrated significantly decreased cartilagedegeneration compared with meniscectomy, as well as ahigh degree of allograft cell viability and vascular ingrowth.We developed a more anatomical surgical technique thataccurately reproduces normal meniscus anatomy and moresecurely fixes the transplanted meniscus compared withpreviously described techniques. We used a wide array ofoutcome measures to more fully and accurately characterizethe cartilage surfaces after meniscal transplantation, aswell as the allograft tissue after transplantation, and haveprovided a comprehensive baseline against which futureanimal studies investigating this procedure may be com-pared. We recognize the importance of early detection of car-tilage wear to optimize the results of meniscal allografttransplantation, and our findings help to validate, by stan-dard MRI, histology, and biomechanical testing, the use ofT2 mapping of cartilage to assess collagen in a noninvasivemanner. In this study, MRI with T2 mapping proved to bevery sensitive for the early detection of hyaline cartilagematrix changes and promises to be a valuable clinical toolfor the early detection of degenerative joint disease. Byincreasing the sensitivity and sophistication of our testingparameters through noninvasive imaging of early cartilagebreakdown, and integrating these tools with surgical plan-ning and timing, the outcome of these surgical interventionsshould be improved. Future studies will be aimed at evalu-ating the efficacy of delayed allograft transplantation aswell as the use of novel biomaterials to provide cartilageprotection after meniscectomy.

ACKNOWLEDGMENT

Research assistance was provided by Peter Torzilli,Manjula Bansal, Stephen Lyman, Li Foong Foo, and ChrisChen. This research was supported by grants from theAmerican Orthopaedic Society for Sports Medicine YoungInvestigator’s Award, National Football League Charities,and Aircast Foundation.

REFERENCES

1. Allen MJ, Houlton JE, Adams SB, Rushton N. The surgical anatomyof the stifle joint in sheep. Vet Surg. 1998;27:596-605.

2. Appleyard RC, Burkhardt D, Ghosh P, et al. Topographical analysis ofthe structural, biochemical and dynamic biomechanical properties ofcartilage in an ovine model of osteoarthritis. Osteoarthritis Cartilage.2003;11:65-77.


3. Arnoczky SP, Warren RF. Microvasculature of the human meniscus.Am J Sports Med. 1982;10:90-95.

4. Arnoczky SP, Warren RF. The microvasculature of the meniscus andits response to injury: an experimental study in the dog. Am J SportsMed. 1983;11:131-141.

5. Arnoczky SP, Warren RF, McDevitt CA. Meniscal replacement using acryopreserved allograft: an experimental study in the dog. Clin OrthopRelat Res. 1990;252:121-128.

6. Bae WC, Temple MM, Amiel D, Coutts RD, Niederauer GG, Sah RL.Indentation testing of human cartilage: sensitivity to articular surfacedegeneration. Arthritis Rheum. 2003;48:3382-3394.

7. Canham W, Stanish W. A study of the biological behavior of themeniscus as a transplant in the medial compartment of a dog’s knee.Am J Sports Med. 1986;14:376-379.

8. Cummins JF, Mansour JN, Howe Z, Allan DG. Meniscal transplanta-tion and degenerative articular change: an experimental study in therabbit. Arthroscopy. 1997;13:485-491.

9. Dardzinski BJ, Mosher TJ, Li S, Van Slyke MA, Smith MB. Spatial vari-ation of T2 in human articular cartilage. Radiology. 1997;205:546-550.

10. Edwards DJ, Whittle SL, Nissen MJ, Cohen B, Oakeshott RD, KeeneGC. Radiographic changes in the knee after meniscal transplantation:an experimental study in a sheep model. Am J Sports Med. 1996;24:222-226.

11. Franz T, Hasler EM, Hagg R, Weiler C, Jakob RP, Mainil-Varlet P. Insitu compressive stiffness, biochemical composition, and structuralintegrity of articular cartilage of the human knee joint. OsteoarthritisCartilage. 2001;9:582-592.

12. Friedlaender GE, Mankin HJ, Sell KW. Osteochondral Allografts:Biology, Banking and Clinical Applications. Boston, Mass: Little, Brownand Co; 1983.

13. Garrett JC. Meniscal transplantation. Am J Knee Surg. 1996;9:32-34.14. Garrett JC. Meniscal transplantation: a review of 43 cases with 2- to

7-year follow-up. Sports Med Arthrosc Rev. 1993;1:164-167.15. Garrett JC, Steensen RN, Stevensen RN. Meniscal transplantation in

the human knee: a preliminary report. Arthroscopy. 1991;7:57-62.16. Jackson DW, McDevitt CA, Simon TM, Arnoczky SP, Atwell EA,

Silvino NJ. Meniscal transplantation using fresh and cryopreservedallografts: an experimental study in goats. Am J Sports Med. 1992;20:644-656.

17. Korhonen RK, Wong M, Arokoski J, et al. Importance of the superfi-cial tissue layer for the indentation stiffness of articular cartilage. MedEng Phys. 2002;24:99-108.

18. Lucchinetti E, Adams CS, Horton WE Jr, Torzilli PA. Cartilage viabilityafter repetitive loading: a preliminary report. Osteoarthritis Cartilage.2002;10:71-81.

19. Lyyra T, Kiviranta I, Vaatainen U, Helminen HJ, Jurvelin JS. In vivocharacterization of indentation stiffness of articular cartilage in thenormal human knee. J Biomed Mater Res. 1999;48:482-487.

20. Maier CF, Tan SG, Hariharan H, Potter HG. T2 quantitation of articu-lar cartilage at 1.5 T. J Magn Reson Imaging. 2003;17:358-364.

21. Mankin HJ, Dorfman H, Lippiello L, Zarins A. Biochemical and meta-bolic abnormalities in articular cartilage from osteo-arthritic humanhips. J Bone Joint Surg Am. 1971;53:523.

22. Mikic ZD, Brankov MZ, Tubic MV, Lazetic AB. Transplantation offresh-frozen menisci: an experimental study in dogs. Arthroscopy.1997;13:579-583.

23. Mosher TJ, Dardzinski BJ, Smith MB. Human articular cartilage: influ-ence of aging and early symptomatic degeneration on the spatial vari-ation of T2. Preliminary findings at 3T. Radiology. 2000;214:259-266.

24. Muldrew K, Chung M, Novak K, et al. Evidence of chondrocyte repop-ulation in adult ovine articular cartilage following cryoinjury and long-term transplantation. Osteoarthritis Cartilage. 2001;9:432-439.

25. Nieminen MT, Rieppo J, Toyras J, et al. T2 relaxation reveals spatialcollagen architecture in articular cartilage: a comparative quantitativeMRI and polarized light microscopic study. Magn Reson Med. 2001;46:487-493.

26. Peterson L, Brittberg M, Kiviranta I, Akerlund EL, Lindahl A.Autologous chondrocyte transplantation: biomechanics and long-term durability. Am J Sports Med. 2002;30:2-12.

27. Potter HG, Linklater JM, Allen AA, Hannafin JA, Haas SB. Magneticresonance imaging of articular cartilage in the knee: an evaluationwith use of fast-spin-echo imaging. J Bone Joint Surg Am. 1998;80:1276-1284.

28. Rodeo SA. Meniscal allografts: where do we stand? Am J SportsMed. 2001;29:246-261.

29. Rubenstein JD, Li JG, Majumdar S, Henkelman RM. Image resolutionand signal-to-noise ratio requirements for MR imaging of degenera-tive cartilage. AJR Am J Roentgenol. 1997;169:1089-1096.

30. Swann AC, Seedhom BB. The stiffness of normal articular cartilageand the predominant acting stress levels: implications for the aetiol-ogy of osteoarthrosis. Br J Rheumatol. 1993;32:16-25.

31. Szomor ZL, Martin TE, Bonar F, Murrell GA. The protective effects ofmeniscal transplantation on cartilage: an experimental study insheep. J Bone Joint Surg Am. 2000;82:80-88.

32. van der Sluijs JA, Geesink RG, van der Linden AJ, Bulstra SK, KuyerR, Drukker J. The reliability of the Mankin score for osteoarthritis.J Orthop Res. 1992;10:58-61.

33. Veltri DM, Warren RF, Wickiewicz TL, O’Brien SJ. Current status ofallograft meniscal transplantation. Clin Orthop Relat Res. 1994;303:44-55.

34. Xia Y, Moody JB, Burton-Wurster N, Lust G. Quantitative in situ correla-tion between microscopic MRI and polarized light microscopy studies ofarticular cartilage. Osteoarthritis Cartilage. 2001;9:393-406.