Magn Reson Mater Phy (2008) 21:177–185DOI 10.1007/s10334-008-0108-4

RESEARCH ARTICLE

Scaffold-aided repair of articular cartilage studied by MRI

Tomasz Zalewski · Przemysław Lubiatowski ·Jacek Jaroszewski · Eugeniusz Szczesniak ·Sławomir Kusmia · Jacek Kruczynski · Stefan Jurga

Received: 24 September 2007 / Revised: 15 February 2008 / Accepted: 15 February 2008 / Published online: 13 March 2008© ESMRMB 2008

AbstractObjective The objective of the study was to evaluate theability of the noninvasive magnetic resonance techniquesto monitor the scaffold-aided process of articular cartilagerepair.Materials and methods Defects of 4 mm in diameter and3 mm in depth were created in right knees of 30 adolescentwhite New Zealand rabbits. Fourteen rabbits were implantedwith poly(lactide-co-glycolic acid) (PLGA) scaffold trim-med to match the size and the shape of the defect (PLGA+group). No procedure was applied to the remaining 16 ani-mals (PLGA− group). Animals were sacrificed sequentiallyat 4, 12, and 24 weeks after the surgery and magnetic reso-nance T2-weighted images (400 MHz) of the dissected boneplugs at eight different echo times were taken to derive T2

relaxation time. The images and the T2 time dependencies

This work was supported by Government Granted Project KBN 3 PO5E 049 23.

T. Zalewski (B) · E. Szczesniak · S. Kusmia · S. JurgaDepartment of Macromolecular Physics,Adam Mickiewicz University, Ul. Umultowska 85,Poznan 61-614, Polande-mail: tomekz@amu.edu.pl

P. LubiatowskiDepartment of Traumatology, Orthopaedic and Hand Surgery,University of Medical Science in Poznan, Poznan, Poland

J. JaroszewskiDepartment of Spine Surgery,Oncology Orthopaedics and Traumatology,University of Medical Science in Poznan, Poznan, Poland

J. KruczynskiDepartment of Orthopaedics, Collegium Medicum,Nicolaus Copernicus University in Torun, Torun, Poland

versus the tissue depth were statistically analyzed. Histolo-gical results of bone plugs were evaluated using semiquanti-tative histological scales.Results The results obtained for PLGA repair tissue wereevaluated versus the PLGA− group and the healthy tissueharvested from the opposite knee (reference group), and com-pared with histological results (hematoxylin and eosin stai-ning). The magnetic resonance images and T2 relaxation timeprofiles taken 4 weeks after surgery for both the PLGA− andPLGA+ group did not reveal the tissue reconstruction. After12 weeks of treatment T2 time dependence indicates a slightreconstruction for PLGA+ group. The T2 time dependenceobtained for PLGA+ samples taken after 24 weeks of treat-ment resembled the one observed for the healthy cartilage,indicating tissue reconstruction in the form of fibrous carti-lage. The tissue reconstruction was not observed for PLGA−samples.Conclusion The study revealed correlation between magneticresonance and histology data, indicating the potential valueof using MRI and spatial variation of T2 as the noninvasivetools to evaluate the process of articular cartilage repair. Italso suggested, that the PLGA scaffold-aided treatment couldhelp to restore the proper architecture of collagen fibrils.

Keywords Cartilage repair · T2 spatial variation · MRI ·T2 anisotropy · PLGA

AbbreviationsPGA Poly(glycolic acid)PLA Poly(lactic acid)PLGA Poly(lactide-co-glycolic acid)ROI Region of interestFOV Field of viewSFM Serum-free mediumPLGA+ Group treated with PLGA scaffoldPLGA− Group treated without PLGA scaffold

123

178 Magn Reson Mater Phy (2008) 21:177–185

Introduction

Articular cartilage being a nonvascularized tissue, whichconsists of chondrocytes [1,2], surrounded by extracellularmatrix formed of proteoglycan and of well-organized colla-gen [3] fibrils has very limited potential for self-repair.

Some of the surgical strategies proposed for treating car-tilage defects make use of scaffolds fabricated of syntheticpolymers or naturally occurring tissues. The latter featurebiocompatibility and fewer regulatory constraints. Unfortu-nately, autograft materials are in limited supply, and allografttissue can transmit disease. On the other hand, syntheticpolymers are available in unlimited supply and their proper-ties, such as mechanical strength, porosity, cell adhesiveness,degradation rate can be modeled to fit for the clinical appli-cation. It is hoped that the scaffold, which should fill enti-rely the defected site, will direct and organize the process ofcartilage regeneration, serving as a mechanical substrate forcells.

The three general strategies for defect filling are [4]:

− implantation of the sole scaffold into a full-thicknessdefect to incite mesenchymal elements to attach to thescaffold, populate, and differentiate into chondrocytes

− use of scaffolds which contain bioactive factors, facili-tating migration of native cells into the treated area

− implementation of scaffolds with chondrocytes expan-ded in culture

The polymers that are currently the most frequently usedas the scaffolds for treating cartilage defects include poly(glycolic acid)-PGA, poly(lactic acid)-PLA, and PLGAcopolymers. They feature proper porous structure, biocom-patibility, and cell adhesiveness [5–8]. Moreover, theirbiodegradation products exert no adverse effects on the sur-rounding cartilage tissue [5,6,9–11], so no inflammatory orimmune responses occur.

Information on the quality of repair is of crucial impor-tance for evaluation of the undertaken therapy and forappraisal of the anticipated results of treatment. However,evaluation means are rather limited. Needle biopsy [12], pro-bably the most often used method for histological evaluation,is an invasive procedure yielding information restricted to asmall area of the treated site.

Recent studies of the healthy tissue have shown thatmagnetic resonance techniques are capable of providing infor-mation on structural organization of collagen fibers in thisheterogeneous and highly ordered tissue [13,14,17,23].

The aim of this work was to evaluate the ability of MRI andT2 techniques to follow up the repair process of the articularcartilage by means of PLGA scaffold.

Materials and methods

Poly(lactide-co-glycolic acid) copolymers (InnoPol, Inno-tech Medical Inc., South Korea) were used in the study. Inno-Pol produced by means of the gas-foaming salt method [5]has a well-defined, highly porous structure and can be easilyshaped into scaffolds of various geometries and sizes. Thedesired properties of scaffold depend on the copolymer com-position [4], i.e., on the content ratio of PLA and PLG. Thescaffolds used in this investigation of rabbit knee cartilageregeneration were composed of 50% PLA and 50% PLG.Pores of this copolymer are interconnected and their size isof about 200 µm, providing optimum conditions for chon-drocyte culture. The scaffold was applied in its pure form.

Before implantation PLGA scaffolds were immersed inserum free medium (SFM). To make sure that properties ofthe scaffold do not conceal information on the regenerationprocess we measured the T2 relaxation times of the scaffoldalone [15]. The measured T2 times did not show depthwisedependence and were one order of magnitude longer thanthose in the cartilage repaired with PLGA. For the scaffoldalone we also studied the translational diffusion and foundthat this process was isotropic and not influenced by the scaf-fold porous structure.

Operative procedure

The experiments were carried out on 30 adolescent whiteNew Zealand rabbits weighing 2.2–2.7 kg. Their right kneeswere operated following general and local anesthesia. Opera-tions were performed in an operating theater for experimentalresearch at the Department of Traumatology and Hand Sur-gery, University of Medical Sciences in Poznan. Medial jointexposure was performed and patella was dislocated laterallyto expose the distal end of the femur. A full-thickness carti-lage defect was created down to the subchondral bone with a4 mm drill. The depth of the defect was 3 mm. The scaffoldwas then inserted into the defect using the press-fit technique.Patella was reduced and wound sutured. No immobilizationwas applied after operation. No procedure was performed onthe opposite knee. The study was approved by the Commit-tee of Ethics of the University of Medical Sciences in Poznan(protocol no. 27/2003).

Animals were sacrificed sequentially at 4, 12, and24 weeks after the surgery and the repair quality was eva-luated using MRI and histology methods. The number ofanimals studied at the given time points was: 4 weeks (PLGA+5, PLGA− 7), 12 weeks (PLGA+ 6, PLGA− 6), 24 weeks(PLGA+ 3, PLGA− 3).

Experimental

Magnetic resonance experiments were performed with aBruker Avance DMX 400 MHz spectrometer equipped with

123

Magn Reson Mater Phy (2008) 21:177–185 179

Table 1 Semiquantitativemicroscopic evaluation scale(modified O’Driscoll)

Category Subcategory Grading

1. Nature of the Cellular morphology Mature hyaline articular cartilage 4

predominant tissue Immature hyaline articular cartilage 3

Incompletely differentiated mesenchyme 2

Fibrous cartilage 1

Fibrous tissue 0

2. Structural characteristics Surface regularity Smooth and intact 3

of repair tissue Superficial horizontal lamination 2

Fissures: 25–100% of the thickness 1

Severe disruption, including fibrillation 0

Structural integrity Normal 2

Slight disruption including cysts 1

Severe disintegration 0

Repair tissue thickness 100% of normal adjacent cartilage 2

50–100% of normal cartilage 1

0–50% of normal cartilage 0

Integration with Bonded at both ends of graft 2

surrounding cartilage Bonded at one end, or partially 1

at both endsNot bonded 0

4. subchondral bone Subchondral bone 100% 2

formation reconstruction 50–100% 1

0–50% 0

a 9.4-T vertical bore magnet (diameter 89 mm), a microima-ging probe head of birdcage type, and an orthogonal gradientsystem capable of generating gradients up to 1 mT/mm. Theinvestigated samples of the cartilage-bone cylindrical plugs,harvested from the same location of joints, were placed intoglass tubes (diameter 15 mm), which were closed to avoiddehydration. The samples were oriented so that the anglebetween the normal to the cartilage surface and the directionof B0 field was 90◦ to keep identical experimental condi-tions and minimize the influence of the magic-angle effectreported previously [16–18]. All experiments were perfor-med at a constant temperature of approximately 20◦C usingthe conventional spin echo pulse sequence (multislice multi-echo). After the precise localization of the repaired site, toensure that measurements of different samples refer to sameregion of interest (ROI), cross-sectional images of repair tis-sue perpendicular to the direction of the magnetic field weretaken. T2 times were calculated from eight images taken atdifferent echo times (TE), varying from 6.6 to 53 ms. Therepetition time (TR) was set to 10 s. Images were reconstruc-ted on 256×256 pixel matrices with a field of view (FOV) of15 mm and slice thickness of 300 µm, which yielded a voxelsize of 59×59×300 µm. The T2 relaxation time was obtai-ned by a single exponential fitting of the data on a pixel-by-

pixel basis to get a better insight into anisotropy effect [19].Relaxation times were calculated using the image sequenceanalysis tool included with the Paravision software packageprovided by the scanner manufacturer. The T2 profiles acrosscartilage thickness were calculated by averaging 15 pixelcolumns, corresponding to the central part of the repair tis-sue (five columns in three adjacent layers) perpendicular tothe cartilage surface, which resulted in a 59 × 295 × 300 µmaveraged voxel. To avoid the partial-volume error when ana-lysing the data we ignored the first pixel corresponding tothe cartilage surface (59 µm). T2 relaxation times refer-ring to a specific depth in a given group of animals wereanalyzed using the one-way analysis of variance (ANOVA)test. The T2 dependencies of PLGA repair tissue were eva-luated versus the PLGA− group and the healthy tissue bymeans of Pearson correlation test. Statistical analysis wasperformed using the program InStat 3.06, GraphPad SoftwareInc. For histological assessment the sections of PLGA+ andPLGA− repair tissue, as well as of the healthy cartilage werestained with hematoxylin–eosin. In histological evaluationspecial attention was paid to the quality and organization ofthe repair tissue, surface regularity, thickness of repair, andsigns of disintegration, using the semiquantitative scale byO’Driscoll with own modification (Table 1) [14].

123

180 Magn Reson Mater Phy (2008) 21:177–185

Table 2 Semiquantitative microscopic evaluation scale (modified O’Driscoll): histological mean result values for each group in all categories

Category Subcategory Grading

PLGA Control

4 weeks 12 weeks 24 weeks 4 weeks 12 weeks 24 weeks

1. Nature of predominant tissue Cellular morphology 2.2 3.6 3.9 1.7 1.4 1.2

2. Structural characteristics of the repair tissue Surface regularity 2.6 2.6 2.9 1.7 1.8 1.7

Structural integrity 1.2 1.8 1.9 1.0 1.4 1.3

Repair tissue thickness 1.9 1.9 2.0 1.4 1.6 1.7

Integration with surrounding cartilage 1.6 1.6 1.9 1.8 1.6 1.4

3. Subchondral bone formation 0.4 1.3 2.0 0.3 1.2 2.0

Fig. 1 The T2-weighted proton image of normal cartilage

Results

Table 2 presents mean histological results of all samples in thegiven group evaluated using the modified O’Driscoll scale.

Figure 1 shows the proton T2-weighted image taken acrossthe cartilage bone of the healthy tissue. A laminar appearanceobserved in early works [17,20–22] is clearly seen. Alsothe transverse relaxation time T2 data presented in Fig. 2demonstrate noticeable anisotropic behavior, showing the

Fig. 2 Cross-sectional profile of T2 for normal cartilage. The profileexhibits noticeable anisotropic behavior showing the dependence of T2on the tissue depth

dependence on the tissue depth [21–23] and the tissue orien-tation with respect to the magnetic field [19,22,23].

Micro magnetic resonance images of the cartilage acqui-red after 4 weeks of regeneration for the PLGA+ and PLGA−groups are shown in Fig. 3a and b, respectively. It can be seenthat after 4 weeks of regeneration the defects were filled withthe repair tissue. In the case of the cartilage treated with thescaffold the filling is homogenous over the whole volume.Histology results of the same samples showed that the defec-ted site in both groups was predominantly filled with notfully differentiated fibrous cartilage. In the case of the scaf-fold group the defect filling almost fully restored the normallevel of articular surface. No signs of organization typicalfor normal cartilage could be observed and the tissue qualitywas similar in both groups (Figs. 4 and 5).

After 12 weeks of regeneration with PLGA a strong NMRsignal was observed, corresponding to the surface of the car-tilage, and no signal was seen for the zone at the boundarywith the subchondral bone (Fig. 6b). For the same follow-up

123

Magn Reson Mater Phy (2008) 21:177–185 181

Fig. 3 The T2-weighted proton images of cartilage repaired for4 weeks: without scaffold (a) and with PLGA scaffold (b). The lightrectangle shows the pixels taken to calculate T2 profiles

Fig. 4 Microhistological images of cartilage repaired for 4 weekswithout scaffold. ×50. The arrows indicate the borders between repairsite and healthy tissue. The line in the lower right-hand corner corres-ponds to a length of 1 mm

time, the cartilage treated without PLGA yielded a strongsignal of similar intensity over the whole volume of interestand the defects were shallower than those observed in thecartilage sample after 4 weeks of regeneration (Fig. 6a). His-tology revealed a significant improvement in the tissue qua-lity towards formation of the normal cartilage in the PLGA+group treated for 12 weeks, as compared to deteriorating stateof repair tissue in the PLGA− group. The PLGA+ group sho-wed some signs of organization, as indicated by columnar

Fig. 5 Microhistological images of cartilage repaired for 4 weeks withPLGA scaffold. ×50. The arrows indicate the borders between repairsite and healthy tissue. The line in the lower right-hand corner corres-ponds to a length of 1 mm

alignment of chondrocytes in deep layer of regenerated tis-sue and small tangentially organized cells in the superficialarea (Fig. 7 and 8).

In the group regenerating for 24 weeks the magnetic reso-nance image of tissue treated with PLGA showed a thincartilage with an outline of laminar appearance, whereasthe image of tissue treated without PLGA was homogenous(Fig. 9a, b, respectively).

The reconstruction in PLGA+ group was corroborated byhistology results, which showed a pronounced alignment ofchondrocytes after 24 weeks of regeneration (Fig. 10). On thecontrary, in the PLGA− group histology revealed some signsof organization occurring only focally, which were overw-helmed by disintegration of the regenerate (Fig. 11). Fibroustissue or fibrocartilage dominated in site of repair in the lattergroup. Although no collagen-specific staining was applied,the absence of isogenic groups and contrast from distinct lightrefraction of collagen bundles and intercellular matrix allo-wed the arrangement of collagen fibers and type of cartilageformed to be monitored.

Figures 12 and 13 show the T2 profiles across the thick-ness of the cartilage corresponding to different times of tissueregeneration for the PLGA− and PLGA+ groups,

123

182 Magn Reson Mater Phy (2008) 21:177–185

Fig. 6 The T2-weighted proton images of cartilage repaired for12 weeks: without scaffold (a) and with PLGA scaffold (b). In (a) thearrow shows disintegration of the tissue that progresses from the surfaceto deep layers. The light rectangle shows the pixels taken to calculateT2 profiles

Fig. 7 Microhistological images of cartilage repaired for 12 weekswithout scaffold. ×50. The arrows indicate the borders between repairsite and healthy tissue. The line in the lower right-hand corner corres-ponds to a length of 1 mm

respectively. The points listed in the figures are the averageT2 taken from two-dimensional (2D) T2 images of the sameposition of the knee cartilage of all rabbits in a given group. Inthe figures we have not listed data concerning the surface of

Fig. 8 Microhistological images of cartilage repaired for 12 weekswith PGLA scaffold. ×50. The arrows indicate the borders betweenrepair site and healthy tissue. The line in the lower right-hand cornercorresponds to a length of 1 mm

Fig. 9 The T2-weighted proton images of cartilage repaired for24 weeks: without scaffold (a) and with PLGA scaffold (b). The lightrectangle shows the pixels taken to calculate T2 profiles

cartilage, because such data were encumbered with an appre-ciable error due to the partial-volume effect. We noticed asteady decrease in the thickness of the cartilage layer withtime. Yet, for a given group the layer thickness differed by amaximum of two pixels (118 µm). The profiles statisticallycompared differed by one pixel, therefore when analyzingthe data we took into account only seven pixels. We ignoredthe eighth pixel of the healthy tissue (tissue harvested fromthe unoperated knee), whose T2 time value was close to thevalue of pixel 7. The profiles were compared using ANOVA

123

Magn Reson Mater Phy (2008) 21:177–185 183

Fig. 10 Microhistological images of cartilage repaired for 24 weekswithout scaffold. ×50. The arrows indicate the borders between repairsite and healthy tissue. The line in the lower right-hand corner corres-ponds to a length of 1 mm

test, which makes use of the Bonferroni correction. We cor-related the T2 time dependence for the healthy tissue anddependence for the tissue treated with PLGA scaffold. Thecorrelation coefficient obtained amounted to 0.9.

Analysis of the obtained data revealed that only the depth-wise dependencies of T2 for the healthy tissue (p < 0.0001)and for the PLGA+ group at the time point of 24 weeks(p < 0.0001) showed significant statistical variation. ThePearson correlation test indicated that the results in thePLGA+ group after 24 weeks of regeneration compared wellwith the results obtained for the healthy tissue (p = 0.006).The standard deviation (SD) of the T2 value correspondingto a given depth in the group investigated ranged between 10and 20%.

It was evident that T2 profiles referring to the tissue repai-red without PLGA showed no dependence on tissue depth,in contrast to the data obtained for cartilage treated with theuse of the PLGA scaffold. In the latter case the dependenceof T2 on the depth became increasingly pronounced as theregeneration time increased.

Discussion

The spatial arrangement of collagen fibers influences motionof water molecules, which in healthy cartilage results in astrong depthwise dependence of T2 and accounts for themagic-angle effect in magnetic resonance images [16,17].

Fig. 11 Microhistological images of cartilage repaired for 24 weekswith PGLA scaffold (b). ×50. The arrows indicate the borders betweenrepair site and healthy tissue. The line in the lower right-hand cornercorresponds to a length of 1 mm

With the hope that magnetic resonance techniques mayprovide information on the tissue reconstruction process westudied the characteristics of magnetic resonance images andT2 profiles in the cartilage repaired by means of PLGA scaf-fold at different stages of treatment and compared the resultswith those obtained for cartilage treated without PLGA andfor healthy cartilage, as well as with the histology data.

The T2 dependencies observed by us for the healthy arti-cular cartilage of rabbit were consistent with those reportedin literature [17–23]. The anisotropic behavior of T2 versusthe depth of the tissue has been accounted for by the arran-gement of water molecules in ordered collagen structures[22,24,25]. For a sample oriented as in our studies, a shortT2 can be expected for protons from the deep cartilage zone,in which collagen fibrils are perpendicular to the Bo field. Inthe intermediate zone, with randomly organized fibrils, T2

reaches a maximum in the region with the maximum numberof fibrils at the magic angle. Moving away from this regiontowards superficial or deep zones the T2 decreases.

123

184 Magn Reson Mater Phy (2008) 21:177–185

Fig. 12 The mean cross-sectional profiles of T2 for full-thickness arti-cular cartilage repaired without PLGA for (a) 4, (b) 12, and (c) 24 weeks.The points in the figures include all samples in the given group

Because of the ability of the T2 time to reflect collagenfibrils ordering we decided to make use of this parameterto follow up the restoration of fibril arrangement in repairedtissue.

The first evaluation of treatment efficiency was carried outafter 4 weeks of scaffold implantation. The magnetic reso-nance images corresponding to the ROI in the cartilage rege-nerated with PLGA were homogenous, while the T2 timeincreased with the tissue depth. These observations may indi-cate that a 4 week period is too short for the copolymer todecompose and make it possible for the tissue to fill the deeplayers entirely [8].

In samples regenerated for the same time without PLGAthe T2 time did not change down to a depth of 900 µm,reflecting the lack of ordered arrangement of collagen fibrils.However, as we noticed, this homogeneity did not extendthroughout the whole volume of the defect, which may beconsistent with the initial stage of tissue differentiation [25].

Magnetic resonance images of the PLGA− group samplesregenerated for 12 weeks showed disintegration of the tissue,which progressed from the surface to deep layers at the cen-tral part of the defect. The disintegration might result from

Fig. 13 The mean cross-sectional profiles of T2 for full-thickness arti-cular cartilage repaired with PLGA scaffold for (a) 4, (b) 12, and (c)24 weeks. The points in the figures include all samples in the givengroup

the lack of spatial collagen organization. This statement wascorroborated by the T2 times, which in the areas not afflictedwith disintegration showed no spatial dependence.

In the case of the cartilage repaired for 12 weeks withPLGA scaffold no disintegration of the regenerated tissuewas seen and the T2 results showed weak dependence ondepth. This provides evidence that the use of scaffold isadvantageous for restoring the order of collagen fibrils. Howe-ver, the restoration process did not include all voxels and, as aresult, the averaged T2 profile shoed only a slight depthwisedependence.

A sounder conclusion on collagen architecture was yiel-ded by MR images of samples treated with PLGA for 24weeks, from which it was evident that the restored tissuefeatures a laminar character and that the relaxation time T2

shows a distinct dependence on depth. The results resembledthose obtained for healthy tissue, indicating restoration ofcollagen order in the treated area [26].

Magnetic resonance images of the samples treated withoutPLGA were homogenous and the T2 time showed nodepthwise dependence, indicating the lack of proper collagenarrangement.

123

Magn Reson Mater Phy (2008) 21:177–185 185

The NMR results are consistent with histology observa-tion. However, we would like to acknowledge the limitationof the applied staining technique as compared to polarizedlight microscopy, which is the gold standard for collagenassessment.

In summary, the results presented herein have shown thepotential value of using MRI and T2 time studies as nonin-vasive tools to evaluate the treatment process of articularcartilage in vivo [25,27]. Moreover, they suggest that tissuerepair using PLGA scaffold helps to restore the appropriatearchitecture of collagen fibrils.

Acknowledgments The authors thank Ms. Agata Gradys from theDepartment of Traumatology and Dr. Marek Kempka from the Depart-ment of Macromolecular Physics for their technical assistance.

References

1. Buckwalter JA (1997) Tissue design and chondrocyte matrix inter-action. J Bone Joint Surg Am 79:600–611

2. Maroudas A (1975) Biophysical chemistry of cartilaginous tissueswith special reference to solute and fluid transport. Biorheology12:233–248

3. Mayne R, Irwin MH (1986) Collagen types in cartilage. In: Kuett-ner KE, Schleyerbach R, Hascall VC (eds) Articular cartilage bio-chemistry. Raven Press, New York, pp 22–23

4. Lu L, Zhu X, Valenzuela RG, Currier BL, Yaszemski MJ(2001) Biodegradable polymer scaffolds for cartilage tissue engi-neering. Clin Orthop Relat Res 391:251–270

5. Wu XS (1995) Synthesis and properties of biodegradable lac-tic/glycolic acid polymers. In: Wise et al (eds) Encyclopedic hand-book of biomaterials and bioengineering. Marcel Dekker, NewYork, pp 1015–1054

6. Lewis DH (1990) Controlled release of bioactive agents from lac-tide/glycolide polymers. In: Chasin M, Langer R (eds) Biodegra-dable polymers as drug delivery systems. Marcel Dekker, NewYork, pp 1–41

7. Moran JM, Pazzano D, Bonassar LJ (2003) Characterization ofpolylactic acid–polyglycolic acid composites for cartilage tissueengineering. Tissue Eng 9(1):63–70

8. Yoon JJ, Park TG (2001) Degradation behaviors of biodegradablemacroporous scaffolds prepared by gas foaming of effervescentsalts. J Biomed Mater Res 55(3):401–408

9. Tice TR, Cowsar DR (1984) Biodegradable controlled-releaseparenteral systems. Pharm Technol 11:26–35

10. Kitchell JP, Wise DL (1985) Poly(lactic/glycolic acid) biodegra-dable drug-polymermatrix systems. Methods Enzymol 112:436–448

11. Brady JM, Cutright DE, Miller RA, Battistone GC (1973)Resorption rate, route of elimination and ultrastructure of theimplant site of polylactic acid in the abdominal wall of the rat.J Biomed Mater Res 7:155–166

12. Brittberg M, Winalski CS (2003) Evaluation of cartilage injuriesand repair. J Bone Joint Surg Am 85-A(Suppl 2):58–69

13. Xia Y (1998) Relaxation anisotropy in cartilage by NMR micro-scopy (µMRI) at 14-µm resolution. Magn Reson Med 39:941–949

14. Rubenstein JD, Kim JK, Morava-Protzner I, Stanchev PL, Hen-kelman RM (1993) Effects of collagen orientation on MR imgingcharacteristics of bovine articular cartilage. Radiology 188:219–226

15. Marcos M, Cano P, Fantazzini P, Garavaglia C, Gomez S, Gar-rido L (2006) NMR relaxometry and imaging of water absorbed inbiodegradable polymer scaffold. Magn Reson Imaging 24:89–95

16. Grunder W, Wagner M, Werner A (1998) MR-microscopic visua-lization of anisotropic internal cartilage structures using the magicangle technique. Magn Reson Med 39:435–442

17. Xia Y (2000) Magic-angel effect in magnetic resonance imagingof articular cartilage. Invest Radiol 35:602–621

18. Mlynarik V, Szomolanyi P, Toffanin R, Vittur F, Trattnig S(2004) Transverse relaxation mechanisms in articular cartilage.J Magn Reson 169:300–307

19. Xia Y, Moody BJ, Alhadlaq H (2002) Orientational dependenceof T2 relaxation in articular cartilage: a microscopic MRI (µMRI)study. Magn Reson Med 48:460–469

20. Modl JM, Sether LA, Haughton VM, Kneeland JB(1990) Articular cartilage correlation of histologic zones withsignal intensity at MR imaging. Radiology 176:853–855

21. Mlynarik V, Degrassi A, Toffanin R, Vittur F, Cova M, Pozzi-Mucelli RS (1996) Investigation of laminar appearance of articu-lar cartilage by means of magnetic resonance microscopy. MagnReson Imaging 14:435–442

22. Xia Y, Moody JB, Burton-Wurster N, Lust G (2001) Quantitativein situ correlation between microscopic MRI and polarized lightmicroscopy studies of articular cartilage. Osteoarthr Cartil 9:393–406

23. Henkelman RM, Stanisz GJ, Kim JK, Bronskill MJ (1994)Anisotropy of NMR properties of tissues. Magn Reson Med32:592–601

24. Nieminen MT, Rieppo J, Toyras J, Hakumaki JM, SilvennoinenJ, Hyttinen MM, Helminen HJ, Jurvelin JS (2001) T2 relaxationreveals spatial collagen architecture in articular cartilage: a com-parative quantitative MRI and polarized light microscopic study.Magn Reson Med 46:487–493

25. Watrin-Pinzano A, Ruaud JP, Cheli Y, Gonord P, Grossin L,Gillet P, Blum A, Payan E, Olivier P, Guillot G, Netter P, LoeuilleD (2004) T2 mapping: an efficient MR quantitative technique toevaluate spontaneous cartilage repair in rat patella. Osteoarthr Car-til 12:191–200

26. Nieminen MT, Toyras J, Laasanen MS, Silvennoinen J, HelminenHJ, Jurvelin JS (2004) Prediction of biomechanical properties ofarticular cartilage with quantitative magnetic resonance imaging.J Biomech 37:321–328

27. Watrin-Pinzano A, Ruaud JP, Cheli Y, Gonord P, Grossin L,Bettembourg-Brault I, Gillet P, Payan E, Guillot G, Netter P,Loeuille D (2004) Evaluation of cartilage repair tissue after bio-material implementation in rat patella by using T2 mapping. MagnReson Mater Phys 17:219–228

123