International Cartilage Repair Society

Transcription

1 Osteoarthritis and Cartilage (2009) 17, 1583e1588 ª 2009 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved. doi: /j.joca Simultaneous computed tomography of articular cartilage and subchondral bone A. S. Aulay*, J. S. Jurveliny and J. Töyräsyz y Department of Physics, University of Kuopio, P.O. Box 1627, FI-70211, Kuopio, Finland z Department of Clinical Neurophysiology, Kuopio University Hospital, P.O. Box 1777, FI-70211, Kuopio, Finland Summary International Cartilage Repair Society Objective: Contrast agent enhanced computed tomography (CECT) may be used to detect depletion of superficial proteoglycans in articular cartilage. In principle, computed tomography can also be used to diagnose the state of subchondral bone. Because osteoarthritis affects both cartilage and bone, we aimed to evaluate the feasibility of the CECT to simultaneously assess the state of these tissues. Further, we studied the spatial variation of contrast agent content in the CECT, properties of subchondral bone and the mechanical stiffness of articular cartilage across the bovine medial tibial plateau. Methods: Osteochondral samples (n ¼ 10) were prepared from fresh and visually intact bovine medial tibial plateaus. The dynamic mechanical modulus of the samples was measured using an arthroscopic indentation device. Subsequently, the samples were scanned with a peripheral quantitative computed tomography (pqct) device prior to and after 35.5 h of immersion in an anionic iodinated contrast agent, (ioxaglate) solution. The thickness of the cartilage was measured with ultrasound and pqct. Results: In the medial tibial plateau, the site-dependent variation in the dynamic modulus, thickness of the cartilage, thickness of the subchondral plate and contrast agent content in the superficial cartilage was statistically significant (P < 0.01). The linear correlation between the superficial contrast agent content and the dynamic modulus was r ¼ 0.80 (P < 0.01). Discussion: The CECT enabled simultaneous analysis of the contrast agent penetration into cartilage, cartilage thickness and subchondral bone density and thickness. The contrast agent content in cartilage depends on the composition and structure of the tissue. Thereby the CECT also provided indirect information on the mechanical properties of the tissue, analogously to the dgemric. Thus, the CECT may provide means to diagnose simultaneously the integrity of cartilage and subchondral bone in vivo. ª 2009 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved. Key words: Articular cartilage, Subchondral bone, Contrast agent, Computed tomography, Indentation, Dynamic modulus, Mineral density. Introduction Articular cartilage forms the connective tissue layer that covers the ends of articulating bones. Healthy cartilage distributes the mechanical loads applied to articulating joints and provides nearly frictionless surfaces 1. The solid matrix of articular cartilage consists mainly of type II collagen and negatively charged proteoglycans (PG). Due to their electronegativity, PGs repel each other and attract water, which creates a swelling pressure that controls the compressive properties of cartilage 2. The collagen network balances the swelling tendency of cartilage and provides it with good tensile and dynamic characteristics 3. Osteoarthritis (OA) is a severe joint disease characterized by disruption of the collagen network, depletion of the PGs and impaired mechanical properties of cartilage 4. Contrast agent enhanced computed tomography (CECT), an analogous technique to the delayed gadolinium enhanced magnetic resonance imaging of cartilage (dgemric) 5e7, has been suggested to quantify the PG *Address correspondence and reprint requests to: Dr Antti S. Aula, Department of Physics, University of Kuopio, P.O. Box 1627, FI-70211, Kuopio, Finland. Tel: ; Fax: ; antti.aula@uku.fi, Received 20 March 2009; revision accepted 26 June content of articular cartilage 8. The quantification is based on the assumption that an anionic contrast agent diffuses into the tissue and equilibrates inversely to the local, depth-dependent PG content 5. Thereby, the equilibrium concentration of the contrast agent may provide a surrogate marker for the integrity and mechanical properties of cartilage 9. However, as suggested by earlier studies, the contrast agent distribution may also be affected by other factors (e.g., water and collagen contents) than PGs or their fixed charge density (FCD) 10,11. In vitro, the diffusion process takes a long time and the equilibrium may not be reached until over 21 h of immersion in the contrast agent solution, depending on the thickness of the cartilage 12,13. Gadolinium-based contrast agents, e.g., gadopentetate, gadobenate and gadoteridol, have usually one gadolinium-atom per molecule. Gadolinium-based contrast agents are widely used in magnetic resonance imaging (MRI) studies, but are not very efficient as X-ray contrast agents. Iodine-based contrast agents often contain several iodine atoms per molecule. For example, ioxaglate contains six and iothalamate three iodine atoms per molecule. This makes them more suitable as X-ray contrast agents, compared to e.g., gadopentetate, as they provide much higher X-ray absorption and, thus, image contrast. Both iodineand gadolinium-based contrast agents are available in different charges and molecular masses. 1583

2 1584 A. S. Aula et al.: CECT of cartilage and subchondral bone Palmer et al. introduced a contrast agent enhanced microcomputed tomography method for monitoring the morphological and compositional properties of cartilage in small-animal models 8. They used relatively high concentration (128 mm) of anionic iodinated contrast agent, and stated that further dilution may reduce its sensitivity to the FCD and hence to the PGs. However, in our earlier study we compared feasibility of iodine- and gadolinium-based contrast agents in detection of PG loss in cartilage and suggested that 21 mm ioxaglate provides sufficient contrast with a clinical peripheral quantitative computed tomography (pqct) scanner 12. In earlier studies the CECT has shown good potential for detection of cartilage degeneration 14. CECT also enables assessment of the morphology and thickness of cartilage, as well as tissue degeneration, in small-animal joints 15,16. Importantly, it has been reported that spontaneous PG depletion in bovine 17 and human 13 articular cartilage can be detected with CECT using a clinical pqct scanner. The pqct allows high-resolution in vivo imaging of extremities like the wrist, ankle and knee. Moreover, it may enable simultaneous measurement of cartilage and subchondral bone. This is important as OA-induced changes appear in both tissues 4,18. The pqct is traditionally used to measure bone mineral density (BMD) and cortical bone thickness in extremities. Thus, analysis of the thickness and BMD of the subchondral plate are hypothesized to be accurate from the CECT images. In this study, we evaluate the feasibility of the CECT to simultaneously assess the status of articular cartilage and subchondral bone. Further, we attempt to use the CECT to reveal impairment of mechanical properties of cartilage. Earlier, Samosky et al. proposed that a topographical dgemric map could, due to the PG loss detected with the technique, indirectly reveal softening of articular cartilage 19. Specifically, we compare the site-dependent relationship between the near equilibrium partitioning of the contrast agent and the indentation stiffness in bovine tibial cartilage. Materials and methods tetraacetic acid disodium salt (VWR International, Fontenay, France) and 5 mm benzamidine hydrochloride (SigmaeAldrich Inc., St. Louis, MO, USA). INDENTATION The dynamic indentation stiffness of the tibial cartilage was determined with an arthroscopic indentation instrument (ArtScan 200, Helsinki, Finland) optimized for the measurement of thin cartilage 20,21. The diameter and length of the indenter were 0.6 mm and 0.14 mm, respectively. The instrument is described in more detail in our earlier publication 22. The force experienced by the indenter (cartilage stiffness) was recorded when the contact force of the reference plate (pressing force) reached 5 N. The values of three repeated measurements were averaged and the protocol was repeated four times for each location (locations 1e4; Fig. 1). The final value for cartilage stiffness was determined as the average of the four measurements. The thickness of the cartilage (th cartus ) was estimated with an ultrasound system (UltraPAC, Physical Acoustics Corporation, Princeton, NJ, USA) using a focused 30 MHz transducer (V327-SU, Olympus NDT Inc., MA, USA) in the pulse-echo measurement geometry 23. The dynamic modulus (E ) could be calculated from the indentation measurements as the thickness of the cartilage was known 24. After the indentation measurements, the samples were cut into two halves in the medio-lateral direction. The samples were stored in a freezer ( 20 C) for further analyses. The first half was stored for the present contrast agent enhanced pqct experiment and the other to be analyzed in an another study. CONTRAST AGENT ENHANCED pqct Each sample was first scanned with a clinical pqct device (XCT 2000, Stratec Medizintechnik GmbH, Pforzheim, Germany) and then immersed in 500 ml of PBS. The PBS was supplemented with 21 mm of an anionic contrast agent (charge 1, mass 1269 g/mol) ioxaglate (HexabrixÔ, Mallinckrodt, St. Louis, MO, USA) and metalloproteinase inhibitors. The osmolarity of the solution was adjusted to 310 mosmol/l and ph to 7.4. After 35.5 h of immersion in the room temperature the samples were removed from the immersion solution and scanned again with the pqct (Fig. 2). The solution was gently stirred during the immersion. The thickness of the pqct slice was 2.3 mm and the in-plane pixel size was mm 2. The X-ray tube voltage and current were 58 kvp and 175 ma, respectively. Five fully overlapping pqct slices were scanned from the center of the sample, without moving the sample, and averaged to reduce noise in the images. The distance from the sample edges to the pqct slice was approximately 5 mm. This distance is greater than the thickness of the cartilage ( mm). Since the diffusion time is proportional to the diffusion distance to the power of two 25, the diffusion through the sample sides can be considered negligible compared to diffusion through the surface. During the pqct scans, full hydration of the samples was achieved by sealing the sample in a plastic container with a PBS-soaked swab. The present pqct scanner quantifies X-ray attenuation as BMD values. BMD values of the pqct images were converted to contrast agent concentrations by SAMPLE PREPARATION Osteochondral samples (n ¼ 10; Fig. 1) were prepared from visually intact bovine (age 18 months) knees, obtained from a local slaughterhouse (Atria Oyj, Kuopio, Finland), within 6 h postmortem. Rectangular medio-lateral samples (width 30 mm, length 50 mm and height 20 mm) were cut from the visually intact medial tibial plateaus with a band saw. During the preparation, the samples were kept moist with phosphate buffered saline (PBS). The samples were stored in a freezer ( 20 C) and thawed prior to the measurements. During the experiments and storing, the samples were immersed in PBS containing inhibitors of metalloproteinases, 5 mm ethylenediamine Fig. 1. The anatomical locations of the indentation (circles) and pqct (dot line) measurements in the medial tibial plateau of bovine knee. The edge of the meniscus is illustrated with a dash line. Fig. 2. The samples were measured with a pqct device before and after the contrast agent immersion. Five fully overlapping pqct slices were scanned from the center of the sample and averaged to reduce noise in the images. Full hydration of the samples was achieved by sealing the sample in a plastic container with a PBS-soaked swab.

3 Osteoarthritis and Cartilage Vol. 17, No measuring a set of known contrast agent concentrations (11 concentrations between 0.42 and 42 mm) in PBS 17. The linear correlation between the contrast agent concentration and the BMD values was 1.00 and the calibration equation was c ¼ BMD To determine the contrast agent partition in cartilage, the contrast agent concentration was normalized with the concentration of the immersion solution. Following the protocol applied in a previous study 19, the contrast agent content in the samples was calculated for the three most superficial pixels (600 mm, c surf ) and for the full-thickness cartilage (c full ). The edges of the samples were excluded and the rest was divided into four sections, corresponding to the indentation sites (Fig. 3). The thickness of the cartilage (th cartct ) at each location was determined from the pqct images. The mean BMD of the subchondral plate (BMD plate ) was calculated from the pqct images. For each location, a 1 mm thick region of interest (ROI), immediately below the cartilageebone boundary, was selected and averaged. Thickness of the subchondral plate (th plate ) was calculated from the images using local thresholding to separate the subchondral plate and trabecular bone. The threshold level, 75% of the maximum BMD, could effectively discern subchondral plate and trabecular bone. Same local threshold criterion was applied to all analyzed areas. The BMD of the subchondral trabecular bone (BMD trab ) was calculated as the mean of a 10 mm thick ROI immediately below the subchondral plate (Fig. 3). STATISTICAL ANALYSIS The data was analyzed with Matlab 7.4 (MathWorks Inc., Natick, MA, USA). Normality of the data was tested with the ShapiroeWilk test. The KruskaleWallis posthoc test was used for testing the significance of the differences between the measurement sites. Spearman s rho was determined to quantify the correlations between the parameters. SPSS 14.0 (SPSS Inc., Chicago, IL, USA) and Matlab 7.4 were used for the statistical analyses. Results The CECT enabled simultaneous analysis of the contrast agent penetration into the cartilage, the cartilage thickness, BMD and thickness of the subchondral plate and BMD of trabecular subchondral bone. The lateral (mid-knee) side of the medial tibial plateau showed higher thickness of cartilage and subchondral plate, higher contrast agent intake (P < 0.01) and lower values of the dynamic modulus (P < 0.01) than the medial side of the tibial plateau (Fig. 4, Tables IeII). The linear correlation between the contrast agent content in the superficial cartilage and the dynamic modulus was significant (r ¼ 0.80, n ¼ 40, P < 0.001; Fig. 5). Further, the thicknesses of the cartilage (r ¼ 0.82, n ¼ 40, P < 0.001) and subchondral plate (r ¼ 0.57, n ¼ 40, P < 0.001) correlated significantly with the dynamic modulus of the cartilage. Moreover, the thickness of the subchondral plate correlated with the thickness of the cartilage (r ¼ 0.50, n ¼ 40, P < 0.001), the contrast agent content in the superficial cartilage (r ¼ 0.54, n ¼ 40, Fig. 3. A typical contrast agent distribution map and ROIs in the bovine medial tibial plateau. The BMD values of the cartilage were converted into the contrast agent concentrations. After discarding the edges of the sample, the central region of the sample was divided into four sections (1e4) that match the locations of the dynamic indentation measurements. The mean BMD was also calculated for the subchondral plate (blue) and trabecular bone (green) at each location. Fig. 4. Normalized superficial contrast agent concentration showed higher contrast agent intake in the lateral side of the medial tibial plateau. The dynamic modulus decreased along with the increase of the contrast agent intake. P < 0.001) and the BMD of the subchondral plate (r ¼ 0.44, n ¼ 40, P < 0.01). The correlation between the cartilage thickness values, determined using ultrasound and CECT, was r ¼ 0.84 (P < 0.01) (Table III). The contrast agent content in the full-thickness cartilage did not correlate significantly with any of the measured parameters. Moreover, the BMD of the subchondral plate did not vary significantly between the locations. Coefficient of variation (CV) for the repeated indentation measurements was 9.3%. Discussion In the present study, for the first time, simultaneous evaluation of the articular cartilage and subchondral bone, and topographical agreement of the contrast agent content and the dynamic indentation modulus were investigated with a clinical pqct instrument. The contrast agent intake into the superficial cartilage correlated significantly with the mechanical properties of the tissue. The pqct also enabled the determination of the thickness and BMD of the subchondral plate and BMD of subchondral trabecular bone. Thus, the CECT may provide a means to diagnose the integrity of cartilage and enable simultaneous quantitative evaluation of the status of the subchondral bone. This study focuses on the superficial cartilage as reaching the diffusion equilibrium may take several days in thick cartilage 26. Only the superficial, and not full thickness, contrast agent content was found to correlate with the dynamic modulus of the cartilage. This is in line with the earlier dgemric study by Samosky et al. 19. The mechanical indenter produces both tensile and compressive stresses in cartilage. The collagen architecture dominates the tensile behavior while the PG content is responsible of the compressive behavior 27,28. In the indentation geometry, properties of the superficial tissue control the mechanical response of the tissue 29. The BMD of the subchondral plate was found to vary insignificantly in the bovine tibial plateau. This is in line with the current understanding that the mineralization of the cortical bone tissue is rather independent of the skeletal location. However, the BMD and thickness of the subchondral plate exhibited a positive correlation. Partial volume effect occurs when the volume element is located at a boundary of tissues. For example, when the voxel includes 50% of bone and 50% of bone marrow, the

4 1586 A. S. Aula et al.: CECT of cartilage and subchondral bone Table I The mean values ( standard deviation) of the measured mechanical and pqct parameters for the samples. See Table II for the statistical differences Location E (MPa) c surf (%) c full (%) th cartct (mm) th cartus (mm) th plate (mm) BMD plate (g/cm 3 ) BMD trab (g/cm 3 ) E, dynamic modulus; c surf, contrast agent content at the surface of the cartilage; c full, contrast agent content in the full-thickness cartilage; th cartct, thickness of the cartilage, calculated from the CECT images; th cartus, thickness of the cartilage, measured with ultrasound; th plate, thickness of the subchondral plate; BMD plate, BMD of the subchondral plate and BMD trab, BMD of the subchondral trabecular bone. value of the voxel is in the mean value of those two tissues. Obviously, the partial volume effect is more severe with larger voxel sizes. Because the partial volume pixels were discarded, the pqct systematically underestimated the thickness of the cartilage and subchondral plate. The thickness of the subchondral plate had a weak but statistically significant correlation with the BMD of the trabecular bone and the subchondral plate. Both BMD values were found to be independent of the properties of cartilage. The thickness of the subchondral plate at the most medial location, however, was significantly lower than that of the other three locations. Furthermore, the thickness of the subchondral plate was related to the dynamic modulus, superficial contrast agent content and thickness of the cartilage. This may be explained by the functional tissue adaptation to prevailing loading conditions in the joint. In the lateral side of the medial tibial plateau (site 1) cartilage was thick and soft, and the contrast agent penetration was high in the superficial cartilage. Also, the subchondral plate was thick. As this anatomical site is not covered by the meniscus, its mechanical load differs from the load of the medial tibial plateau. This difference in loading probably explains the detected topographical variation in the properties of cartilage and subchondral bone. In order to determine dynamic modulus from the indentation measurements, thickness of the cartilage at the indentation site must be known 24. Ultrasound was applied to Table II Statistical differences in the measured mechanical and pqct parameters among the anatomical locations. Dynamic modulus, superficial contrast agent concentration and thickness of the cartilage showed statistically significant spatial variation. Most parameters of the bone at the medial location 4 were statistically different from the other locations Locations 1e2 1e3 1e4 2e3 2e4 3e4 E e ** ** e ** e c surf e ** ** e * e c full e e e e e e th cartct e ** ** e ** e th plate e e ** e ** ** BMD plate e e e e e e BMD trab e e e e ** ** E, dynamic modulus; c surf, contrast agent content at the surface of the cartilage; c full, contrast agent content in the full-thickness cartilage; th cartct, thickness of the cartilage, calculated from the CECT images; th plate, thickness of the subchondral plate; BMD plate, BMD of the subchondral plate and BMD trab ¼ BMD of the subchondral trabecular bone. *P < 0.05, **P < obtain the thickness information exactly at the actual indentation site. It has been reported in our earlier publications, that ultrasound provides feasible tool for measurement of cartilage thickness at an indentation location 23,30. In contrast, since the present pqct scanner provides rather large in-plane pixel size ( mm 2 ) and high minimum slice thickness (2.3 mm), the partial volume artifact was significant and the voxels at the surface and cartilageebone interface had to be discarded. Evidently, this leads to underestimation of the cartilage thickness. For this reason the ultrasound was selected as the reference thickness measurement technique. Contrast agent enhanced pqct is a minimally invasive method. The effective radiation dose of one pqct crosssection is less than 2 msv. However, the contrast agent has to be injected into the subject, probably intra-articularly to obtain high enough contrast agent concentration. The related risks, for example infection, should be acceptable if the method provides a cheap and straightforward means to assess the status of articular cartilage in vivo. The osmolarity of the injectable solution should be at the level of physiological saline. If the solution is hypertonic, i.e., at the level or above the osmolality of the cartilage (350e450 mosmol/kg 31 ), the osmotic swelling pressure of the tissue diminishes and the cartilage shrinks and softens 32. Further, loading of the softened cartilage, even by means of e.g., kneeling, may cause cell death and disruption of the cartilage matrix 33. In the present study, this was taken into account as the osmolarity of the immersion solution was adjusted to 310 mosmol/l. Fig. 5. A significant linear correlation (r ¼ 0.80, n ¼ 40) was demonstrated between the dynamic modulus and normalized superficial contrast agent content.

5 Osteoarthritis and Cartilage Vol. 17, No Table III The linear correlations between the measured mechanical and pqct parameters. The contrast agent content in the superficial cartilage, i.e., the first 600 mm, correlated significantly with the dynamic modulus of cartilage. The thickness of the subchondral plate correlated significantly with the superficial contrast agent content, dynamic modulus and thickness of the cartilage c surf c full th cartct th plate BMD plate BMD trab E 0.80** ** 0.57** c surf e 0.62** 0.69** 0.54** c full e e th cartct e e e 0.50** th plate e e e e 0.44** 0.60** BMD plate e e e e e 0.25 E, dynamic modulus; c surf, contrast agent content at the surface of the cartilage; c full, contrast agent content in the full-thickness cartilage; th cartct, thickness of the cartilage, calculated from the CECT images; th plate, thickness of the subchondral plate; BMD plate, BMD of the subchondral plate and BMD trab, BMD of the subchondral trabecular bone. **P < According to Fick s laws of diffusion, the time to reach the diffusion equilibrium depends on the thickness of the cartilage 25. Earlier results suggest that even more than 24 h may be needed to obtain true equilibrium of the contrast agent accumulation in bovine cartilage 17. In the present samples, the thickness of the cartilage varied from 0.8 to 3.6 mm in different anatomical sites. Consequently, a long diffusion time (35.5 h) was chosen to enable the diffusion equilibrium at all locations. The immersion time used in the present study is obviously not applicable for clinical studies. In clinical dgemric studies, 0.5e7 h delay has been used between the contrast agent administration and MRI scan 34e36, which suggests that the dynamic, physiological equilibrium may be achieved in a relatively short time. However, it must be noted that the physiological equilibrium is not equal to the diffusion equilibrium and, thus, does not enable calculation of the FCD in the cartilage. Equilibrium partitioning of the contrast agent is assumed to depend on the FCD of the tissue according to the GibbseDonnan equilibrium. Therefore, collagen has a smaller effect on the equilibrium partitioning. However, as it is a significant contributor to solid content of the tissue, it may affect the diffusion rate. Possibly, the measurement of spatial contrast agent diffusion rate could enable evaluation of the spatial solid content and even the properties of the collagen network. Quantification of the diffusion rate in cartilage may provide a feasible means to apply this technique in vivo as solutes have been reported to diffuse faster into degenerated cartilage 10,37. Even though gadopentetate and ioxaglate have showed promising results, smaller contrast agent molecules may prove to be clinically more feasible due to significantly faster diffusion into cartilage 26. According to recent studies, dynamic loading of the cartilage also increases the rate of diffusion and may lead to diffusion equilibrium in physiologically reasonable times 38,39. Although further studies are needed to verify the in vivo feasibility of CECT of articular cartilage, it is tempting to think that CECT could provide a cheap and minimally invasive means to simultaneously assess the integrity of articular cartilage and subchondral bone in patients with a suspected joint disease. Conflict of interest The authors have no conflicts of interest. Acknowledgments Financial support from Emil Aaltonen Foundation, Finnish Cultural Foundation and Kuopio University Hospital (EVO 5221) are acknowledged. Bone and Cartilage Research Unit, University of Kuopio is acknowledged for providing the pqct instrument. The authors thank Atria Lihakunta Oyj, Kuopio, Finland for providing fresh bovine joints. References 1. Mow WC, Zhu W, Ratcliffe A. Structure and function of articular cartilage and meniscus. In: Mow WC, Hayes WC, Eds. Basic Orthopaedic Biomechanics. New York: Raven Press, Ltd.; 1991:143e Buckwalter JA, Mankin HJ. Articular cartilage: tissue design and chondrocyteematrix interactions. Instr Course Lect 1998;47:477e Buckwalter JA, Martin J. Degenerative joint disease. Clin Symp 1995; 47(2):1e Buckwalter JA, Mankin HJ. Articular cartilage: degeneration and osteoarthritis, repair, regeneration, and transplantation. Instr Course Lect 1998;47:487e Bashir A, Gray ML, Burstein D. Gd-DTPA 2 as a measure of cartilage degradation. Magn Reson Med 1996;36(5):665e Bashir A, Gray ML, Hartke J, Burstein D. Nondestructive imaging of human cartilage glycosaminoglycan concentration by MRI. Magn Reson Med 1999;41(5):857e Nieminen MT, Rieppo J, Silvennoinen J, Töyräs J, Hakumäki JM, Hyttinen MM, et al. Spatial assessment of articular cartilage proteoglycans with Gd-DTPA-enhanced T 1 imaging. Magn Reson Med 2002; 48(4):640e8. 8. Palmer AW, Guldberg RE, Levenston ME. Analysis of cartilage matrix fixed charge density and three-dimensional morphology via contrastenhanced microcomputed tomography. Proc Natl Acad Sci U S A 2006;103(51):19255e Wayne JS, Kraft KA, Shields KJ, Yin C, Owen JR, Disler DG. MR imaging of normal and matrix-depleted cartilage: correlation with biomechanical function and biochemical composition. Radiology 2003; 228(2):493e Evans RC, Quinn TM. Solute diffusivity correlates with mechanical properties and matrix density of compressed articular cartilage. Arch Biochem Biophys 2005;442(1):1e Maroudas A. Distribution and diffusion of solutes in articular cartilage. Biophys J 1970;10(5):365e Kallioniemi AS, Jurvelin JS, Nieminen MT, Lammi MJ, Töyräs J. Contrast agent enhanced pqct of articular cartilage. Phys Med Biol 2007;52(4):1209e Silvast TS, Jurvelin JS, Lammi MJ, Töyräs J. pqct study on diffusion and equilibrium distribution of iodinated anionic contrast agent in human articular cartilage e associations to matrix composition and integrity. Osteoarthritis Cartilage 2009;17(1):26e Cockman MD, Blanton CA, Chmielewski PA, Dong L, Dufresne TE, Hookfin EB, et al. Quantitative imaging of proteoglycan in cartilage using a gadolinium probe and microct. Osteoarthritis Cartilage 2006; 14(3):210e Xie L, Lin AS, Levenston ME, Guldberg RE. Quantitative assessment of articular cartilage morphology via EPIC-muCT. Osteoarthritis Cartilage 2008;17(3):313e Piscaer TM, Waarsing JH, Kops N, Pavljasevic P, Verhaar JA, van Osch GJ, et al. In vivo imaging of cartilage degeneration using microct-arthrography. Osteoarthritis Cartilage 2008;16(9):1011e Silvast TS, Jurvelin JS, Aula AS, Lammi MJ, Töyräs J. Contrast agentenhanced computed tomography of articular cartilage: association with tissue composition and properties. Acta Radiol 2009;50(1): 78e85.

6 1588 A. S. Aula et al.: CECT of cartilage and subchondral bone 18. Day JS, Van Der Linden JC, Bank RA, Ding M, Hvid I, Sumner DR, et al. Adaptation of subchondral bone in osteoarthritis. Biorheology 2004; 41(3e4):359e Samosky JT, Burstein D, Eric Grimson W, Howe R, Martin S, Gray ML. Spatially-localized correlation of dgemric-measured GAG distribution and mechanical stiffness in the human tibial plateau. J Orthop Res 2005;23(1):93e Lyyra-Laitinen T, Niinimäki M, Töyräs J, Lindgren R, Kiviranta I, Jurvelin JS. Optimization of the arthroscopic indentation instrument for the measurement of thin cartilage stiffness. Phys Med Biol 1999; 44(10):2511e Timonen MA, Karjalainen JP, Riekkinen O, Jurvelin JS. Technical and practical improvements in arthroscopic indentation technique of articular cartilage. Trans Orthop Res Soc 2009;34: Korhonen RK, Saarakkala S, Töyräs J, Laasanen MS, Kiviranta I, Jurvelin JS. Experimental and numerical validation for the novel configuration of an arthroscopic indentation instrument. Phys Med Biol 2003;48(11):1565e Töyräs J, Rieppo J, Nieminen MT, Helminen HJ, Jurvelin JS. Characterization of enzymatically induced degradation of articular cartilage using high frequency ultrasound. Phys Med Biol 1999;44(11):2723e Hayes WC, Keer LM, Herrmann G, Mockros LF. A mathematical analysis for indentation tests of articular cartilage. J Biomech 1972;5(5):541e Crank J. The Mathematics of Diffusion. 2nd edn. Oxford: Clarendon Press; pp. 1e Silvast TS, Kokkonen H, Nieminen MT, Lammi M, Quinn TM, Jurvelin JS, et al. Diffusion rate and near-equilibrium distribution of MRI and CT contrast agents in articular cartilage. Trans Orthop Res Soc 2009;34: Mow VC, Fithian DC, Kelly MA. Fundamentals of articular cartilage and meniscus biomechanics. In: Ewing JW, Ed. Articular Cartilage and Knee Joint Function: Basic Science and Arthroscopy. New York: Raven Press Ltd.; 1990:1e Bader DL, Kempson GE, Egan J, Gilbey W, Barrett AJ. The effects of selective matrix degradation on the short-term compressive properties of adult human articular cartilage. Biochim Biophys Acta 1992; 1116(2):147e Korhonen RK, Wong M, Arokoski J, Lindgren R, Helminen HJ, Hunziker EB, et al. Importance of the superficial tissue layer for the indentation stiffness of articular cartilage. Med Eng Phys 2002;24(2): 99e Töyräs J, Lyyra-Laitinen T, Niinimäki M, Lindgren R, Nieminen MT, Kiviranta I, et al. Estimation of the Young s modulus of articular cartilage using an arthroscopic indentation instrument and ultrasonic measurement of tissue thickness. J Biomech 2001;34(2):251e Urban JP, Hall AC, Gehl KA. Regulation of matrix synthesis rates by the ionic and osmotic environment of articular chondrocytes. J Cell Physiol 1993;154(2):262e Eisenberg SR, Grodzinsky AJ. Swelling of articular cartilage and other connective tissues: electromechanochemical forces. J Orthop Res 1985;3(2):148e Töyräs J, Turunen M, Lammi M, Jurvelin JS, Korhonen RK. Hyperosmolaric contrast agents in cartilage tomography may expose cartilage to overload-induced cell death. Trans Orthop Res Soc 2009;34: Bashir A, Gray ML, Boutin RD, Burstein D. Glycosaminoglycan in articular cartilage: in vivo assessment with delayed Gd(DTPA)(2-)-enhanced MR imaging. Radiology 1997;205(2):551e Tiderius CJ, Jessel R, Kim YJ, Burstein D. Hip dgemric in asymptomatic volunteers and patients with early osteoarthritis: the influence of timing after contrast injection. Magn Reson Med 2007;57(4):803e Williams A, Sharma L, McKenzie CA, Prasad PV, Burstein D. Delayed gadolinium-enhanced magnetic resonance imaging of cartilage in knee osteoarthritis: findings at different radiographic stages of disease and relationship to malalignment. Arthritis Rheum 2005;52(11): 3528e Maroudas A. Biophysical chemistry of cartilaginous tissues with special reference to solute and fluid transport. Biorheology 1975;12(3e4): 233e Evans RC, Quinn TM. Dynamic compression augments interstitial transport of a glucose-like solute in articular cartilage. Biophys J 2006; 91(4):1541e Albro M, Li R, Yeager K, Hung C, Ateshian G. Dynamic loading of porous gels and cartilage can induce active solute transport. Trans Orthop Res Soc 2009;34:151.