| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Experimental Studies |
1 From the Orthopaedic Research Laboratory, Departments of Biomedical Engineering and Orthopaedic Surgery (J.S.W., K.J.S., C.Y., J.R.O.) and Department of Radiology (K.A.K., D.G.D.), Virginia Commonwealth University, 1112 E Clay St, 325 McGuire Annex, PO Box 980694, Richmond, VA 23298; and Commonwealth Radiology, Richmond, Va (D.G.D.). Received December 7, 2001; revision requested January 19, 2002; final revision received December 12; accepted January 14, 2003. Supported in part by a grant from Virginia’s Commonwealth Health Research Board. Address correspondence to J.S.W. (e-mail: jswayne@vcu.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Normal articular cartilage from 12 porcine patellae was evaluated biomechanically, biochemically, and with MR imaging (with and without gadolinium enhancement). The patellae were then enzymatically treated to deplete the matrix of either collagen or proteoglycan and then reevaluated biomechanically, biochemically, and with MR imaging. Correlations between cartilaginous tissue function and MR imaging parameters were made. Analysis of variance was performed to assess the effect of enzymatic treatment on measured parameters. Linear correlations among the MR imaging, biochemical, and biomechanical parameters were performed to determine the strengths of the relationships. P < .05 indicated statistically significant differences.
RESULTS: Biochemical, biomechanical, and MR analyses enabled detection of changes caused by matrix depletion (P < .05). T2 was the most useful MR imaging parameter for distinguishing proteoglycan loss from collagen loss. T2 correlated significantly with both biomechanical modulus (indicative of cartilage stiffness; P < .001, R2 = 0.51) and biochemical proteoglycan content (P < .001, R2 = 0.44). Differentiation between proteoglycan loss and collagen loss in terms of T1 improved with gadolinium enhancement. With gadolinium enhancement, proteoglycan depletion was associated with a greater decrease in T1 than collagen depletion (P < .05).
CONCLUSION: An association between biochemical and biomechanical functional status and MR imaging parameters of articular cartilage was demonstrated. Linear correlations existed between modulus and proteoglycan content in terms of T2. Additionally, proteoglycan loss and collagen loss had differing effects on gadolinium-enhanced T1 when it was expressed as the ratio of T1 after gadolinium enhancement/T1 before gadolinium enhancement.
© RSNA, 2003
Index terms: Experimental study • Cartilage, MR, 453.121411, 453.121413, 453.121416, 453.12143, 453.12144, 453.12146 • Magnetic resonance (MR), experimental studies, 453.121411, 453.121413, 453.121416, 453.12143, 453.12144, 453.12146 • Magnetic resonance (MR), tissue characterization, 453.12146 • Patella
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
The use of MR imaging as a noninvasive tool for determining the functional status of cartilage has great potential for scientific and clinical applications. The imaging protocols determined to enable the best depiction of cartilage function could be applied to in vivo models to study, for example, the progression of cartilage degeneration with a specific stimulus or applied to reparative models to study the healing quality of cartilage. These same protocols could also be applied in the study of human articular surfaces to noninvasively detect and quantify the extent of degenerative changes and thus acquire information to help initiate early interventions for delaying or reversing the degeneration.
Therefore, the purpose of this study was to correlate articular cartilage function, as reflected in biomechanical properties and biochemical composition, with MR imaging parameters of normal articular cartilage and cartilage partially depleted of matrix components.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
Seven of the 12 patellae were randomly selected for contrast agent–enhanced MR imaging (described later). Quantitative evaluation of each quadrant from the remaining five patellae then proceeded as follows: A 2-mm central transverse section, representing normal cartilage composition, was removed from the patellae for later biochemical analysis. Biomechanical testing and MR imaging of each quadrant were performed to determine the characteristics of normal articular cartilage. The proximal quadrants were then exposed to chondroitinase ABC for proteoglycan depletion, and the distal quadrants were exposed to collagenase for collagen depletion. After matrix depletion, biomechanical testing and MR imaging of each quadrant were repeated. Finally, the untreated 2-mm central transverse sections and the enzymatically treated quadrants were analyzed biochemically.
Enzymatic Treatment
The proximal quadrants were selected for proteoglycan depletion, and the distal quadrants were selected for collagen depletion. This process resulted in 10 quadrants, or specimens, per enzymatic treatment for biomechanical, biochemical, and MR imaging evaluations. For proteoglycan depletion, the proximal medial and lateral quadrants were placed in a 25°C buffer solution of 1 unit per milliliter of chondroitinase ABC (5) in 0.15 mol/L of sodium chloride and 0.05 mol/L of Tris(hydroxymethyl)aminomethane hydrochloride for 36 hours. This solution also contained protease inhibitors (1 mmol/L of phenylmethylsulfonyl fluoride, 10 mmol/L of N-ethylmaleimide, and 5 mmol/L of benzamidine–hydrochloride hydrate) to inhibit nonproteoglycan matrix digestion. For collagen depletion, the distal medial and lateral quadrants were placed in a similar buffer solution that contained collagenase (50 mg/100 mL of buffer solution) (6,7) instead of chondroitinase. All chemicals were purchased from Sigma-Aldrich (St Louis, Mo).
After depletion of the target component (ie, proteoglycan or collagen), repeat biomechanical and MR imaging analyses of the quadrants were performed. The same location that was marked on the quadrant before matrix depletion was identified by the initially placed fiducial marks and used for biomechanical testing. Likewise, pre- and postenzymatic depletion MR images of the same anatomic section of each quadrant were obtained. In this way, each quadrant served as its own control specimen.
Biochemical Evaluation
Biochemical assays of the quadrants were performed to determine the hydration, proteoglycan content, and collagen content of the cartilage after the specified enzymatic depletion treatment. Normal cartilage values were determined from the 2-mm central transverse section harvested from each patella.
To determine water content, the cartilage was blotted to remove excess solution and weighed before and after dehydration (24 hours at 80°C). The cartilage was further analyzed to determine the concentrations of hydroxyproline—a measure of collagen content—and sulfated glycosaminoglycan—a measure of proteoglycan content. For assessment of hydroxyproline content, the cartilage was hydrolyzed with 6N hydrochloric acid at 130°C for 6 hours, neutralized, and analyzed by using a colorimetric procedure at a wavelength of 557 nm (8). For analysis of sulfated glycosaminoglycan content, the cartilage was digested in a papain solution and then assessed with dimethylmethylene blue binding assay at a wavelength of 535 nm (9).
Biomechanical Evaluation
The mechanical properties of the patellar cartilage in each quadrant were determined by performing indentation experiments. The quadrants were tested in situ (with intact cartilage and underlying bone) in a custom-designed indentation apparatus. This device was used to measure the instantaneous and time-dependent creep (ie, change in surface compression with time) behaviors of the cartilage when a constant load was applied to it. The quadrant was mounted and submerged in a bath of 0.15 mol/L of saline solution with protease inhibitors (1 mmol/L of phenylmethylsulfonyl fluoride, 10 mmol/L of N-ethylmaleimide, and 5 mmol/L of benzamidine–hydrochloride hydrate) for 10 minutes. The testing site was oriented in a three-dimensional space through a ball hinge so that the 1.5-mm-diameter cylindric indenter was perpendicular to the articular surface.
A 2-g preload was applied to the cartilaginous tissue by using a loading platform attached to the indenter to ensure full contact between the indenter and the cartilage surface. The creep deformation of the cartilage surface was monitored by using a linear variable displacement transducer. After the surface had equilibrated for 15 minutes, a 10-g weight was applied to the loading platform by using a solenoid, and the instantaneous and creep deformations were measured for up to 2 hours or until equilibrium was reached. Equilibrium was set as a small value of the slope of the creep curve (10-06 mm/sec).
Loading was controlled and surface compression data were acquired by using a personal computer (MacIntosh IIci; Apple, Cupertino, Calif) and data acquisition-display software (LabVIEW; National Instruments, Austin, Tex). Paired sets of data—namely, time and amount of creep (ie, surface compression)—from each indentation test were collected for use in property determination. Time points at appropriate intervals were chosen to adequately reflect the rate of creep of the cartilage.
The experimental data from the indentation tests were then analyzed by using a nonlinear least-squares curve-fitting algorithm (10). This approach provided a means of determining the intrinsic biomechanical properties of the cartilaginous tissue—specifically, the aggregate modulus as a measure of matrix stiffness and the permeability as a measure of the ease with which fluid could flow within the tissue. The following equation is used to calculate the aggregate modulus, Ha:
|
), the displacement of the cartilage surface at the end of creep; h, the thickness of the tissue; 
s, the Poisson ratio; and 
(a/h, s), a dimensionless parameter that accounts for certain geometric quantities. Permeability was evaluated in terms of the displacement behavior over time.
MR Imaging Evaluation
All MR imaging experiments were performed by using a 2.35-T MR imager-spectrometer (Biospec ABX; Bruker BioSpin, Billerica, Mass) with a horizontal-bore magnet. A 12-cm (inner diameter) self-shielded gradient insert allowed a gradient strength of up to 260 mT/m per axis, with a corresponding maximum slew rate of 1,180 T/m/sec.
A birdcage transmit radio-frequency coil and a 30-mm-diameter surface receive radio-frequency coil were used in this study. The surface coil was mounted on a horizontal acrylic tray, so that it could be fixed at the magnet isocenter within the birdcage coil. Patellar samples were imaged in pairs (ie, two quadrants side by side) and enveloped in plastic wrap to prevent water loss during imaging. The samples were positioned with the cartilage facets flat against the horizontal plane of the surface coil.
Transverse and coronal scout MR images were acquired to ensure that sample placement was reproducible with use of the fiducial marks. All subsequent MR imaging for quantitative analysis was carried out in the transverse plane. Use of the standardized parameters of a 2.5-mm section thickness, a 3.0 x 1.5-cm field of view, and a 256 x 128 matrix yielded two-dimensional MR image pixel dimensions of 117 x 117 µm. These parameters yielded approximately 15–20 pixels across the thickness of the porcine patellar cartilage.
MR imaging of each quadrant was performed to assess T1, T2, and the apparent diffusion coefficient. Between the biomechanical and MR imaging evaluations, the specimens were wrapped in saline-soaked gauze, sealed in airtight plastic bags, and frozen for no longer than 1 week. They were thawed at room temperature for at least 1 hour before evaluation. In vitro T1-weighted MR imaging of the quadrants was accomplished by using a segmented echo-planar inversion-recovery sequence with five to eight inversion times. The imaging parameters were as follows: repetition time of 5,000 msec, effective echo time of 24 msec, 16 phase-encoding steps per segment, one signal acquired, and inversion times of 30–2,500 msec. An integrated three-parameter exponential fitting function was applied to each group of inversion-recovery MR images to obtain pure T1-weighted images, on which each pixel reflected the T1 of the corresponding tissue voxel. T2 was measured by using an eight-echo spin-echo sequence with 4,000/24–192 (repetition time msec, echo time msec) and one signal acquired. T2-weighted MR images were likewise calculated by using a two-parameter exponential fit of the decay curves.
A series of diffusion-weighted spin-echo images (1,500/36, one signal acquired) also were acquired by using four diffusion factors: 10.6, 205.0, 633.0, and 1,108.0 sec/mm2. Apparent diffusion coefficient maps were created by fitting the signal intensity of each pixel to a decaying exponential, with appropriate correction for the effect of gradient cross terms (11). To test for possible diffusion anisotropy in cartilage, diffusion data were acquired with the diffusion gradient applied independently along the section, frequency, and phase gradient axes.
Contrast-enhanced MR Imaging
The changes resulting from enzymatic matrix depletion in the separate set of seven patellae, each sectioned into four quadrants, were assessed by using MR imaging enhanced with an anionic contrast agent (gadopentetate dimeglumine2-, Magnevist; Berlex Laboratories, Wayne, NJ) (1). Separate patellae were used for these experiments because preliminary study results showed that the contrast agent did not readily diffuse out of the cartilage so that depletion or further imaging analyses could be performed. The quadrants were placed in a 2 mmol/L solution of gadopentetate dimeglumine2- with Hanks normal saline. MR imaging was then performed by using the segmented echo-planar inversion-recovery sequence described earlier, and T1 maps were created from the inversion-recovery images.
Each quadrant was imaged immediately before it was bathed in the gadolinium solution and 1, 3, 7, 15, 30, and 60 minutes after this bathing to determine the time course of the change in T1 due to the uptake of the contrast agent. This time course was fit to an exponentially decreasing function to yield a time constant of the decay. For assessment of the characteristics of normal cartilage enhanced with contrast agent, eight quadrants from two of the seven patellae were imaged without having been depleted of matrix. The quadrants of the remaining five patellae were imaged after chondroitinase (10 proximal quadrants) or collagenase (10 distal quadrants) matrix depletion to identify the contrast-enhanced characteristics of matrix depletion.
Statistical Analyses
Univariate repeated measures analysis of variance was performed (with SAS for Windows; SAS, Cary, NC) to determine whether differences in biomechanical characteristics (ie, aggregate modulus, permeability), biochemical characteristics (ie, hydration, collagen content, proteoglycan content), and/or MR imaging parameters (ie, T1, T2, apparent diffusion coefficient, gadolinium enhancement) existed among normal and matrix-depleted specimens. Additional paired t tests of the T1 of the normal, chondroitinase-treated, and collagen-treated cartilaginous tissues were individually performed before and after gadolinium enhancement to assess changes in T1 due to contrast agent enhancement. Percent changes, where indicated, were calculated as follows: (value after matrix depletion - value before matrix depletion)/value before matrix depletion. Finally, linear correlations among the MR imaging parameters and biomechanical and biochemical characteristics (Excel; Microsoft, Redmond, Wash) were made. All values are presented as means ± SDs. P < .05 indicated a statistically significant difference.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
|
|
|
|
|
Correlations
Linear correlations among the different functional (ie, biomechanical and biochemical) parameters and MR imaging characteristics (Fig 6) were made. A significant correlation between modulus and proteoglycan content was noted (P < .001, R2 = 0.89): An increase in proteoglycan content resulted in an increase in the modulus that was similar to that seen in other studies (12). Permeability was inversely related to proteoglycan content (P < .01), but the correlation was not as strong (R2 = 0.32). Modulus and permeability had a moderate inverse relationship (P < .03, R2 = 0.23). The relationship between modulus and T2 was significant (P < .001, R2 = 0.51), as was the correlation between proteoglycan content and T2 (P < .001, R2 = 0.44).
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
Both enzymatic treatments caused a decrease in the modulus (indicative of cartilage stiffness); however, the effect was more pronounced with chondroitinase treatment. This is consistent with the belief that proteoglycans, by virtue of their large negatively charged status, have a primary role in resisting compression while collagen has a primary role in resisting tension. The biomechanical evaluation in this study was performed with compression. As in other studies (13,14), in this study, cartilaginous tissue permeability increased with chondroitinase treatment; this indicates a decreased resistance of the solid matrix to fluid flow within the tissue. Both of these changes are consistent with alterations in biomechanical function with the initiation of arthritic diseases. Changes in these biomechanical properties have long been associated with altered biomechanical function (15–17).
At MR imaging, T2 was the most useful parameter for distinguishing between proteoglycan loss and collagen loss. T2 increased with both enzymatic treatments, but it increased more from the normal levels after chondroitinase treatment than after collagenase treatment. Additionally, T2 correlated significantly with both biomechanical modulus and biochemical proteoglycan content (also linked to cartilage stiffness). Proteoglycan depletion resulted in a greater gadolinium-induced decrease in T1 than collagen depletion. The shorter T1 may have been due to a reduced electrostatic repulsion of gadolinium as the proteoglycans were depleted from the tissue.
The results of this study also reemphasize the need for a comprehensive evaluation of cartilaginous tissue. Literature data on the effects of proteoglycan depletion or collagen depletion must be carefully interpreted to ensure that the described effects are truly due to proteoglycan or collagen loss. In this study, chondroitinase and collagenase treatments were studied to glean information to help decipher differences between the two treatments in terms of depletion of various matrix components. Other enzymes such as trypsin may be considered for proteoglycan depletion, but the full effects of treatments with these enzymes on both proteoglycan and collagen must be verified first.
MR imaging relaxation properties, magnetization transfer–weighted MR sequences, and MR imaging contrast agents have been promoted as tools for detecting variations in cartilage signal intensity (4,18–26). The qualitative spatial variations in T2, as seen by others (4,18,23,27), were verified in this study, although quantification throughout the thickness of the patellar cartilage was not performed, because this would have required increased spatial resolution. In some previous investigations (19,25), relaxation properties—specifically, T1 and T2—were not shown to be reflective of changes in the proteoglycan content of cartilage, and some believe that these properties are more reflective of the collagen content (22,24,26). However, others have linked T2 to proteoglycan distribution (4,18,23). This latter finding is supported by the results presented in this study.
Similar to us, Nieminen and colleagues (27) also observed a decrease in the biomechanical modulus following either collagenase or chondroitinase ABC treatment. However, Nieminen et al observed a longer T2 only with collagenase treatment, in contrast to the findings in this study. Recently, the T1
relaxation of articular cartilage has been described as a more sensitive parameter for detecting smaller changes in the proteoglycan content of articular cartilage (28).
Because each patella was quantitatively evaluated by using various methods in this study, only one level of matrix depletion could be evaluated. The decreased proteoglycan or collagen content caused by incubating the quadrants in either chondroitinase ABC or collagenase, respectively, for 36 hours comprised more than 50% of the target component. This amount of depletion does not mirror the smaller matrix changes—particularly in proteoglycan content—seen in the early stages of arthritic diseases (29). However, the observed correlation between T2 and function parameters (ie, modulus or proteoglycan content) was evident among a range of values that resulted from individual specimen variations.
Some early physiologic changes, such as fibrillation of the superficial zone, which is composed predominantly of tangentially oriented collagen fibrils, also tend to be localized in specific regions of the cartilage layer. With higher strength magnets and specialized radio-frequency and gradient coils, differentiation of imaging parameters between zones may be feasible and informative in future studies. Finally, enzymatic treatment to produce functionally and compositionally altered tissue that simulates physiologic deterioration is in wide use (5,6,19–22,25,27,28). This technique represents a systematic and quantifiable approach to altering cartilage structure and determining the effects of this alteration on specific parameters. Additional evaluations of physiologically altered cartilage from in vivo models or clinically degenerated surfaces should be performed.
Practical application: Establishing a noninvasive means of determining the mechanical function of cartilage in a diarthrodial joint would affect the detection of arthritic changes in tissue and aid in determining the quality of reparative cartilage generated from in vivo reparative models. This study was performed to increase understanding of how imaging information yields details about underlying function. We believe that an extension of these results to studies of physiologically degenerated surfaces is warranted. This information can also be applied to early interventions aimed at delaying or reversing the degeneration of articular cartilage in osteoarthritis. These additional studies can be focused on the MR imaging parameters judged to be the most efficient at depicting cartilage matrix deterioration. In this way, MR imaging findings may prove to be surrogate markers of biomechanical and biochemical measures and thus be used for the early detection of articular disorders and to facilitate successful surgical interventions and drug therapies.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
-relaxation of articular cartilage at 4T. Magn Reson Med 2001; 46:419-423.[CrossRef][Medline]
This article has been cited by other articles:
|
|
 |
|
  G. H. Welsch, T. C. Mamisch, S. E. Domayer, R. Dorotka, F. Kutscha-Lissberg, S. Marlovits, L. M. White, and S. Trattnig Cartilage T2 Assessment at 3-T MR Imaging: In Vivo Differentiation of Normal Hyaline Cartilage from Reparative Tissue after Two Cartilage Repair Procedures--Initial Experience Radiology, April 1, 2008; 247(1): 154 - 161. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. G. Disler, Y. Y. Ho, A. J. Stanley, J. H.-P. Hui, and S.-C. Wang Invited Commentary * Authors' Response RadioGraphics, January 1, 2007; 27(1): 221 - 222. [Full Text] [PDF]
|
|
|
|
|
|
L. M. White, M. S. Sussman, M. Hurtig, L. Probyn, G. Tomlinson, and R. Kandel Cartilage T2 Assessment: Differentiation of Normal Hyaline Cartilage and Reparative Tissue after Arthroscopic Cartilage Repair in Equine Subjects. Radiology, November 1, 2006; 241(2): 407 - 414. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H Graichen, D Al-Shamari, S Hinterwimmer, R von Eisenhart-Rothe, T Vogl, and F Eckstein Accuracy of quantitative magnetic resonance imaging in the detection of ex vivo focal cartilage defects Ann Rheum Dis, August 1, 2005; 64(8): 1120 - 1125. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Watrin-Pinzano, J.-P. Ruaud, P. Olivier, L. Grossin, P. Gonord, A. Blum, P. Netter, G. Guillot, P. Gillet, and D. Loeuille Effect of Proteoglycan Depletion on T2 Mapping in Rat Patellar Cartilage Radiology, January 1, 2005; 234(1): 162 - 170. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
