Abstract
Purpose:
This study aimed to establish the topographical and zonal T2 patterns of multi-resolution MRI in medial tibial cartilage in a canine model of osteoarthritis (OA), initiated by the anterior cruciate ligament (ACL) transection surgery, and studied after 8-weeks and 12-weeks post-surgery.
Methods:
Articular cartilage from healthy, two stages of contralateral, and of OA knees were quantitatively imaged by the MRI T2 protocols at two imaging resolutions (100 and 17.6 μm/pixel). The zonal T2 changes at five topographical locations (anterior (AMT), exterior (EMT), posterior (PMT), central (CMT) and interior (IMT) medial tibia) and subsequent two averaged regions (covered by meniscus and exposed) were analyzed. At each location, full-thickness cartilage was studied in four sub-tissue zones (superficial, transitional, upper and lower radial zones).
Results:
Tissue degradation can be detected by measurable changes of T2, which is resolution- and orientation-dependent. T2 changes ranging from +28.82% increase (SZ, PMT) to −23.15% decrease (RZ1, AMT) in healthy to disease (8C), with the largest increase of T2 in the surface tissue. Various location-dependent patterns of degradation are found over the tibial surface, most commonly shown in early-stage OA (8C) on the anterior site, different from the posterior. Finally, the contralateral cartilage has specific degradation patterns, different from those in OA cartilage.
Conclusions:
This is the first quantitative and highest multi-resolution characterization of cartilage at five topographical locations over the medial tibial plateau with fine zonal resolution in an animal model of OA, which would benefit future investigation of human OA in clinics.
Keywords: microscopic MRI (μMRI), T2, cartilage, osteoarthritis, ACL transection
1. Introduction
Articular cartilage is a thin layer of avascular tissue covering the ends of long bones in diarthrodial joints. The tissue in large animals and human is mostly extracellular and composed primarily of water (~70%), collagen (~20%) and proteoglycans (~5%) [1, 2]. Cartilage possesses a unique depth-dependent zonal structure, which is largely determined by the orientation of its collagen fibers at different tissue depths. Three sub-tissue zones commonly divide the thickness of cartilage: the superficial zone (SZ) where the collagen fibers are parallel with the tissue surface, the transitional zone (TZ) where the fibers are mostly in random orientation, and the radial zone (RZ) where the fibers are more densely packed and perpendicular to the surface [3–5]. In addition, it has been known that the zonal structures and molecular properties of cartilage are topographically different among the tissue harvested from different surface locations of a single joint [6–9], which is likely due to the demands of the complex loading patterns and force dissipations for the particular joint [10–12].
The degradation of cartilage is the hallmark of osteoarthritis (OA), which is the most common musculoskeletal disease that affects millions of people in our society [13–17]. Given the zonal and topographical natures of cartilage properties, cartilage degradation must have zonal and topographical signatures at different stages of OA during its slow progression [18]. The fact that articular cartilage is a thin layer of curved tissue with both topographical and zonal variabilities makes the detection of early OA progression challenging. Among the non-invasive imaging tools in clinics, magnetic resonance imaging (MRI) is the only technique that has sensitivity to the motion of molecules in soft tissue [19]. Several MRI parameters, in particular, T2 relaxation time, have been widely used for the diagnostic assessment of OA cartilage [20–22]. In biological tissues, T2 relaxation has complex mechanisms and reflect the mobility of water molecules in their local environments, which show anisotropic properties in cartilage and other collagen-rich tissues [23, 24]. For the simplest notion in relation to osteoarthritis cartilage, a high T2 value implies a tissue having less proteoglycans, less organized collagen matrix, more extracellular water, and looser interactions between water and macromolecules [25–29].
In this project of canine model of OA, the degradation of cartilage was triggered by anterior cruciate ligament (ACL) transection surgery [30, 31] and studied at 8-weeks and 12-weeks after the surgery. There were five types of tissue: healthy cartilage, OA cartilage 8-weeks and 12-weeks after surgery, and contralateral cartilage 8-weeks and 12-weeks after surgery. (Note that the contralateral joint should not be considered as a true normal control, due to the changes in movement pattern of any surgery animal.) Cartilage from five topographical locations on the medial tibial plateau was studied by MRI T2 protocols, each with two zonal resolutions (100μm/pixel and 17.6μm/pixel). We aimed to investigate the interactions on the MRI T2 OA detection by different disease progression, at different topographical locations, in different sub-tissue zones, and using different imaging protocols.
2. Methods
2.1. Sample Preparation
With the approval of the institutional review committees, nineteen mature canines were acquired from a dedicated facility, then quarantined onsite for two weeks with the proper care. Prior to surgery, the conformation and gait of each animal was studied to exclude animals that had pre-existing anatomic or gait abnormalities. The surgical and care procedures, which are available upon request, satisfied the guidelines of the related regulatory agencies. Seven of the nineteen canines were non-operated healthy controls (N, 14 knees), and the rest of the twelve canines had the anterior cruciate ligament (ACL) in one knee transected 8 weeks (8OA, 6 knees) and 12 weeks (12OA, 6 knees) prior to sacrifice. The other knees in the surgical groups were untouched and labeled as the contralateral knee (8C and 12C, each 6 knees). After the excess tissues were trimmed off at sacrifice, all unopened knee capsules were imaged in a 7T MRI. Then each joint capsule was opened to harvest five rectangular cartilage-bone blocks from its medial tibia. Each block was about 3×3×5 mm3 in size, representing a specific topographical location on the medial joint surface (Fig 1).
Fig 1.
An illustrative topographic map of medial tibia joint divided into five locations (thick solid rectangular) representing each of the five cartilage bone plugs selected for μMRI studies. The five locations are labelled as Anterior (AMT), Exterior (EMT), Posterior (PMT), Central (CMT) and Interior (IMT) respectively. The dotted curved line represents the meniscus covered area in the medial tibial surface. Three mMRI slices are marked by the dot-dash lines that cover the five locations.
2.2. MRI Protocols
All unopened knee capsules were imaged in quantitative MRI T2 experiments on a Varian MRI system with a 7T/20cm horizontal-bore magnet (Santa Clara, CA) and used the multi-slice multi-echo (MSME) pulse sequence. The FOV was 5cm with a matrix size of 256×256, which yielded a 200μm pixel size. These images were then interpolated in post-processing into 100μm per pixel size, which are labeled as macro-MRI (mMRI) in the remainder of this report. The repetition time was 3 sec; 10 echo times were used, with the minimum echo time of 10ms and each increment at 10ms with a total scan time of about 52 minutes. At each echo time, 10 slices were acquired, each with a slice thickness of 1mm (approx. 2.5mm apart from each other, running at interleaving sequences). Each joint thus had 100 T2*-weighted images, whose time-domain data were reconstructed (including zero padding) from a matrix size of 256×256 to 512×512, yielding the final pixel size of 100μm [32]. Ten quantitative T2* images at 10 different slice locations were calculated based on these weighted images. The locations of three mMRI slices and the subsequent five μMRI cartilage-bone plug locations are shown in Fig 1.
All small specimens were immersed in physiological saline solution, which also contained 1mM Gd-DTPA2− contrast agent and 1% protease inhibitor cocktail (Sigma, MO). The specimens were maintained at 4°C (never frozen) and imaged quantitatively in a different 7T micro-MRI (Bruker AVANCE II 300 micro-imager). The acquisition matrix was 256×128, which was reconstructed into 256×256 matrix, resulting in a pixel size of 17.6 μm (labeled as μMRI). The field of view was 4.5 mm with a slice thickness of 0.8 mm. The repetition time TR was set at 500 as all samples were soaked in Gd-DTPA2− (for T1 experiments). The T2-weighted images were acquired using a CPMG magnetization-prepared imaging sequence with five echo times (2, 8, 20, 50, and 100 ms) and total scan time of approximately 1 hour and 30 minutes [23, 33].
Two sets of μMRI images were acquired for each specimen with the normal axis to the cartilage surface being oriented at 0° and 55° to the external magnetic B0 field, which were used to calculate two orientation-dependent T2 images. These two orientations in the quantitative T2 imaging were necessary since T2 characteristic in articular cartilage is known to have strong magic angle effect [23, 33], which has its origin in the dipolar interaction in spin dynamics. The angles of 0° and 55° to the external magnetic field are the two orientations that give the maximum and minimum influence of the dipolar interaction to the depth-dependent T2 characteristics.
2.3. Image and Statistical Analysis
The high-resolution T2 profiles in μMRI were divided into four zones based on the established zonal division method [4, 34], where based on the bell-shaped curve of the T2 profile (0°) the depth-dependent profile was divided into individual zones. The division for the low-resolution profile in mMRI considered the fact that the average thickness of SZ and TZ was each 80–120μm for this type of canine tissue [8, 35], hence assigning the first and second pixels as SZ and TZ for mMRI respectively, with the rest of the profiles as the two equal parts of RZ (RZI and RZII). Both mMRI and μMRI quantitative T2 maps were calculated using a single exponential decay on pixel-by-pixel basis with the aid of a MATLAB code. The exponential decay equation used in the code allows for a more rigorous calculation of T2 and has been used in previous studies of high-resolution MRI of cartilage.
One-way ANOVA with Bonferroni correction test using KaleidaGraph (Reading, PA) was performed on all specimens from each of the five topographical locations (AMT, EMT, PMT, CMT and IMT) of the medial tibia, comparing T2 relaxation times among healthy (N), contralateral (8C and 12C) and osteoarthritic (8OA and 12OA) cartilage in each of four unequal-thickness zones (SZ, TZ, RZI and RZII) by mMRI and μMRI (0° and 55°). Pairwise statistical tests were also carried out between healthy (N) and both contralateral joints (8C and 12C), and between healthy (N) and both osteoarthritic joints (8OA and 12OA). Pairwise statistical comparisons were also carried out between comparable disease groups. A resultant p-value of less than 0.05 was considered significant.
3. Results
3.1. The mMRI and μMRI images and profiles
Figure 1 shows the locations of the mMRI and μMRI slices, from which the T2 of cartilage was measured quantitatively. The central mMRI slice (#2) captured three locations from anterior to posterior (AMT, CMT and PMT), where each topographical location was studied individually in μMRI. Fig 2a/2d/2g show three identically-located mMRI slices from three specimen groups (N, 8OA, 12OA); all had the colored T2 images of cartilage overlaid on the grey-scale intensity images of knee joint, together with the corresponding μMRI T2 maps at 0° and 55° orientations. With the progression of the tissue lesion from healthy (N) to two OA stages (8OA, 12OA), an incremental increase in both T2 and tissue thickness can be seen clearly, with the 12OA tissue showing the biggest T2 increases (Fig 2g) and the clearest swelling in cartilage across the entire tibial sagittal slice (Fig 2h, 2i). These mMRI changes vary topographically from one surface location to another, illustrating the fact that not all cartilage degrades simultaneously.
Fig 2.
The mMRI images (a, d, g) of three intact knee joints representing the Normal, 8OA and 12OA animals, overlaid by their quantitative T2 maps (colored, 0–100ms). The mMRI ROI (the white box) shows the approximate region where the cartilage-bone plugs were harvested for the μMRI 0° and 55° imaging, which were shown under each knee image. The arrow heads in (b) and (c) mark the thickness of articular cartilage. The direction of the magnetic field B0 is pointing vertically up in the figure (represented with an arrow); the angles of 0° and 55° are between the normal axis of the articular surface and the magnetic field direction.
Fig. 3 shows the depth-dependent T2 profiles at one tibial site (CMT), comparing healthy cartilage with the OA cartilage. The comparative plots used a relative thickness to accommodate varying tissue thickness due to tissue swelling. The region of interest (ROI) on both mMRI and μMRI T2 maps had a consistent width (~200μm). In the low-resolution mMRI profiles (Fig. 3a), three features can be observed. First, a significant elevation of T2 can be seen in the surface tissue between the normal and the OA cartilage (0 – 0.15 depth). Second, there was no difference in T2 between the 8-weeks and 12-weeks after the surgery. Finally, for most of the deep cartilage (0.15 – 1.0 relative depth), there is little change in T2 values, regardless of whether the tissue is healthy, 8-weeks or 12-weeks after the surgery.
Fig 3.
Quantitative T2 profiles of the entire cartilage from the images shown in Fig 2, (a) mMRI profiles at 100 μm per pixel, (b) μMRI profiles at 17.6 μm per pixel. The vertical dotted lines mark the sub-tissue zones.
In the high-resolution μMRI profiles (Fig 3b), two specimen orientations in the magnetic field modulated the T2, where the dipolar interaction has the strongest effect on T2 at 0° orientation and the least effect on T2 at the 55° (the magic angle). At 0°, the μMRI T2 profiles appear nearly identical to the mMRI profiles (Fig 3a), where the only sensitivity to tissue degradation was in the surface tissue. When the specimen was oriented at 55° (Fig 3b), the progression of cartilage lesion was clearly different from the corresponding 0° profiles. Two additional features of these μMRI T2 55° profiles should be noted. First, there was a lack of major differences in the surface tissue (0 – 0.1 relative depth) between the 8OA and 12OA groups, which signals the lack of further degradation in the surface tissue from 8 weeks to 12 weeks after the ACL transection. Second, despite the lack of progression in the surface tissues between 8OA and 12OA, there were significant T2 elevations for the deep tissue (0.2 – 1.0 relative depth) between 8OA and 12OA, which indicates that the degradation in the deep cartilage happened later than the degradation in the surface cartilage.
3.2. Topographical and zonal T2 patterns among different lesion groups
Quantitative T2 maps for three imaging protocols (mMRI, μMRI 0°, μMRI 55°) were constructed individually in all five topographical locations on the medial tibial surface (c.f. Fig 1). At each location, a depth-dependent profile (c.f. Fig 3) was obtained by averaging all the T2 profiles among the animals. These profiles were sub-divided into the four zones and the quantitative T2 values within each zone were averaged [4, 35]. The progressions of tissue degradation were analyzed among the five topographical sites (AMT, EMT, PMT, CMT and IMT), among the four sub-tissue zones (SZ, TZ, RZI, RZII), among the five disease groups (N, 8C, 12C, 8OA, 12OA), and among the three imaging protocols (mMRI, μMRI 0°, μMRI 55°).
Fig 4 shows the topographical T2 patterns in the superficial zone of cartilage measured by three imaging protocols at one specific topographical site. Since T2 values of healthy cartilage at different sites of the knees are known to be different [8], simply comparing an averaged T2 value would not reflect the true trend due to the degradation. For this reason, we also calculated the topographical and zonal percentage differences of T2 between various osteoarthritic cartilage and the healthy counterpart, and used the percentages to color these topographical patterns, where darker red/blue signifies increased/decreased percentage difference with the normal (uncolored). This use of color ‘hot/cold’ maps highlights any trend of positive/negative percent changes respectively that is significant to the disease progression.
Fig 4.
Topographical T2 maps of the superficial zone cartilage in the medial tibias from three imaging protocols: (a) μMRI at 0°, (b) μMRI at 55°, and (c) mMRI. Each T2 value is averaged among the T2 of all specimens at the same zone and same topographical location. The three blocks inside the dotted line represent the meniscus covered region, while the other two blocks represent the exposed region. All T2 maps were plotted on the same scale (0 – 100ms). (N represents healthy cartilage, 8C and 12C are the contralateral cartilage at 8-weeks and 12-weeks post the ACL surgery, 8OA and 12OA are the OA cartilage at 8-weeks and 12-weeks post the ACL surgery.) The darkness of blue and red colors is based on the calculated percentage difference between each disease stage to the healthy counterpart, where darker red/blue means higher/lower percentage change than the normal, which is uncolored. The max and min % changes were ±30 % in μMRI 0° (a) and 55° (b), and ±40% in mMRI (c).
Several trends can be noted for the superficial cartilage. First, a large cluster of blue colors were noted in the anterior and exterior sites of the contralateral cartilage (AMT and EMT at 8C and 12C), where both are covered by the meniscus; and a higher cluster of red and more dark-red colors could be seen in OA cartilage (8OA and 12OA). This illustrates that T2 generally increased as the disease progressed from healthy to severe lesions. Second, although three imaging protocols (mMRI, μMRI 0°, μMRI 55°) measured approximately the same tissue, they showed different sensitivities towards the T2 changes. For example, both μMRI 0° and μMRI 55° at the AMT location showed lower T2 in contralateral cartilage and modest high T2 in 12OA cartilage; in contrast, mMRI showed significantly higher T2 (darker red) for both contralateral and OA cartilage at the same location (consistent with the information in Fig 3). Finally, each topographical location showed different trends of T2. For example, at the CMT location, the biggest increases in T2 was in the 8C cartilage (μMRI 0°), which reduced towards later degradation. At the same time, AMT and PMT locations (which are the most anterior and posterior locations) have the opposite trends in 8C cartilage.
All other sub-tissue zones were studied similarly as in the SZ maps above. We found that cartilage in each topographical zone degrades independently. One essential finding was that T2 changes in μMRI 0° is more evident in SZ and TZ than the deep zones, as 0° orientation showed the weakened sensitivity of T2 for the deep cartilage due to the increased influence of the dipolar interaction on water mobility.
3.3. Zonal T2 patterns at all topographical locations
Topographically distributed zonal T2 data in all five locations and by three imaging protocols were summarized as values with statistically significant differences from the healthy (N) counterparts emphasized by bold and underline, and the color hot/cold maps were scaled based on percentage differences shown in Fig 5. The relevant percentage differences of T2 between various osteoarthritic degradation and the healthy counterpart were presented in Table 1, together with statistical analysis where any significant differences were made in bold and underlined in Fig 5 and marked with an asterisk in Table 1. The averaged data was also presented as the correlated linear regression plots in the Supplement Fig 1, which presents consistent overall conclusions, shows the effects of orientational and resolution dependency. Some of the major observations are:
Fig 5.
Mean T2 values of all imaging protocols, all topographical locations, starting with the normal to each of the disease stages through all histological zones (SZ, TZ, RZI and RZII). Covered marks the meniscus-covered region on tibial surface, which has three topographic locations (AMT, EMT, PMT); Exposed marks the tibial region that is not covered by the meniscus, which has two topographic location (CMT and IMT). The mean T2 values with underlined bold font imply statistical significance in pair-wise comparison. Color coding (blue white red) is based on the calculated percentage difference between each disease stage to the healthy counterpart: μMRI 0° and μMRI 55° are −30% to +30%, mMRI is −40% to +40%. The actual percentage differences are in Table 1.
Table 1.
Quantitative T2 percentage differences of μMRI 0° (a), μMRI 55° (b) and mMRI (c) of the individual topographical locations, which were used to color Fig 5. The table starts with the differences of each disease stage (8C, 12C, 8OA and 12OA) from the normal (N) followed by the disease group comparison, through all of the histological zones (SZ, TZ, RZI and RZII). AMT, EMT, PMT mark the three meniscus-covered locations on tibial surface, while CMT and IMT mark the two tibial locations that is not covered by the meniscus. Any number marked with * signifies the statistical significance (p<0.05) within the pairwise comparison.
| a | site | Zone | N v. 8C | N v. 12C | N v. 8OA | N v. 12OA | 8C v. 8OA | 12C v.12OA | 8C v. 12C | 8OA v.12OA |
|---|---|---|---|---|---|---|---|---|---|---|
| μMRI 0° | AMT | SZ | −14.28 | −4.63 | 13.17* | 16.46* | 27.32* | 21.05* | 9.67 | 3.31 |
| TZ | −17.18* | 1.52 | 4.81 | 5.99 | 21.95* | 4.48 | 18.69* | 1.18 | ||
| RZI | −23.15* | 10.01 | 1.51 | 4.27 | 24.64* | −5.75 | 32.97* | 2.76 | ||
| RZII | −18.12* | 9.47 | 0.52 | 15.28* | 18.63* | 5.83 | 27.47* | 14.76 | ||
| EMT | SZ | 4.76 | −12.55* | 18.82* | 13.93* | 14.09* | 26.37* | −17.28* | −4.92 | |
| TZ | 4.59 | −7.53 | 2.73 | −1.64 | −1.85 | 5.89 | −12.11* | −4.38 | ||
| RZI | 2.55 | −1.70 | −0.25 | 13.28* | −2.80 | 14.97* | −4.25 | 13.53* | ||
| RZII | 3.53 | 14.95* | 0.61 | 28.30* | −2.93 | 13.49* | 11.43* | 27.71* | ||
| PMT | SZ | 28.82* | 6.67 | 18.60* | 21.53* | −10.36* | 14.92* | −22.27* | 2.96 | |
| TZ | 17.83* | −2.82 | −2.59 | −8.25* | −20.40* | −5.44 | −20.63* | −5.67 | ||
| RZI | 18.51* | −5.99 | 5.42 | −13.62* | −13.12* | −7.64 | −24.43* | −19.01* | ||
| RZII | 11.94* | −4.79 | −2.26 | 1.34 | −14.20* | 6.12 | −16.71* | 3.60 | ||
| CMT | SZ | 17.88* | −14.21* | 8.52 | −8.43 | −9.39 | 5.79 | −31.89* | −16.93* | |
| TZ | 16.83* | −9.33* | 3.11 | −1.68 | −13.74* | 7.66 | −26.06* | −4.79 | ||
| RZI | 18.68* | 1.19 | 7.52 | 3.29 | −11.20 | 2.10 | −17.50* | −4.24 | ||
| RZII | 14.44* | 10.71 | −2.05 | 8.49 | −16.48* | −2.23 | −3.74 | 10.53 | ||
| IMT | SZ | 7.79 | −3.63 | 30.95* | 14.18* | 23.30* | 17.79* | −11.41 | −16.95* | |
| TZ | 4.82 | 2.51 | 17.62* | −0.74 | 12.83* | −3.26 | −2.30 | −18.36* | ||
| RZI | 11.93* | 4.93 | 28.11* | 0.81 | 16.31* | −4.12 | −7.02 | −27.32* | ||
| RZII | 5.43 | 5.35 | 14.80* | −3.00 | 8.36 | −8.35 | −0.08 | −16.76* |
| b | site | Zone | N v. 8C | N v. 12C | N v. 8OA | N v. 12OA | 8C v. 8OA | 12C v.12OA | 8C v. 12C | 8OA v.12OA |
|---|---|---|---|---|---|---|---|---|---|---|
| μMRI 55° | AMT | SZ | −11.96* | −4.62 | −11.33* | 8.50 | 0.64 | 13.11* | 7.36 | 19.78* |
| TZ | −13.97* | 3.99 | −9.84 | 9.43* | 4.14 | 5.45 | 17.93* | 19.23* | ||
| RZI | −11.17* | 12.21* | −13.88* | 7.99 | −2.71 | −4.23 | 23.30* | 21.80* | ||
| RZII | −15.17* | 9.58 | −25.98* | −0.57 | −10.92 | −10.14 | 24.66* | 25.42* | ||
| EMT | SZ | 7.53 | −18.41* | 29.34* | 28.55* | 21.93* | 46.35* | −25.85* | −0.81 | |
| TZ | 4.89 | −24.36* | 23.63* | 23.87* | 18.80* | 47.54* | −29.16* | 0.24 | ||
| RZI | 1.48 | −27.19* | 18.87* | 12.25* | 17.41* | 39.11* | −28.63* | −6.66 | ||
| RZII | 4.40 | −31.78* | 10.73* | −8.03 | 6.34 | 23.90* | −36.05* | −18.72* | ||
| PMT | SZ | 19.83* | −1.25 | 5.60 | 6.21 | −14.27* | 7.46 | −21.06* | 0.62 | |
| TZ | 19.88* | −0.32 | 8.65* | 1.57 | −11.28* | 1.89 | −20.20* | −7.09 | ||
| RZI | 15.90* | 1.79 | 5.35 | −4.96 | −10.58* | −6.75 | −14.12* | −10.30* | ||
| RZII | 11.85* | 0.21 | 0.55 | −13.17* | −11.30* | −13.38* | −11.65* | −13.72* | ||
| CMT | SZ | 11.12* | −9.97* | 4.52 | −3.84 | −6.61 | 6.14 | −21.03* | −8.35 | |
| TZ | 6.25 | −8.12 | 4.33 | −2.79 | −1.92 | 5.33 | −14.35* | −7.12 | ||
| RZI | −0.10 | −1.28 | 6.81 | 7.12 | 6.91 | 8.40 | −1.18 | 0.32 | ||
| RZII | −1.82 | 10.56* | 14.36* | 27.25* | 16.17* | 16.80* | 12.38 | 13.02* | ||
| IMT | SZ | 7.27 | 0.04 | 20.24* | 23.92* | 13.02* | 23.89* | −7.23 | 3.73 | |
| TZ | 4.38 | −5.55 | 18.77* | 11.37* | 14.42* | 16.89* | −9.92 | −7.43 | ||
| RZI | 4.04 | 0.85 | 17.63* | −0.60 | 13.62* | −1.45 | −3.18 | −18.23* | ||
| RZII | 0.94 | 4.50 | 3.25 | −5.96 | 2.25 | −10.45* | 3.56 | −9.15* |
| c | site | Zone | N v. 8C | N v. 12C | N v. 8OA | N v. 12OA | 8C v. 8OA | 12C v.12OA | 8C v. 12C | 8OA v.12OA |
|---|---|---|---|---|---|---|---|---|---|---|
| mMRI | AMT | SZ | 27.54 | 36.29* | 42.93* | 59.76* | 15.86 | 24.81 | 8.97 | 17.98 |
| TZ | −0.60 | 4.41 | 27.94* | 46.72* | 28.52* | 42.53* | 5.01 | 19.42 | ||
| RZI | 1.12 | 0.04 | 28.31* | 31.56* | 27.21 | 31.52* | −1.08 | 3.33 | ||
| RZII | −12.88 | −6.01 | 20.10 | 13.57 | 32.77* | 19.54 | 6.88 | −6.58 | ||
| EMT | SZ | 6.78 | 10.20 | 43.35* | 29.13* | 36.84* | 19.07 | 3.42 | −14.68 | |
| TZ | 9.64 | 7.60 | 31.00* | 19.15 | 21.51 | 11.60 | −2.05 | −12.02 | ||
| RZI | 1.22 | 0.66 | 15.20 | 15.42 | 13.98 | 14.76 | −0.56 | 0.22 | ||
| RZII | 9.90 | 1.81 | 24.49 | 25.49 | 14.68 | 23.71 | −8.09 | 1.02 | ||
| PMT | SZ | −6.21 | 6.09 | 11.07 | 19.89 | 17.25 | 13.84 | 12.29 | 8.87 | |
| TZ | −9.77 | −12.93 | 1.17 | −2.88 | 10.94 | 10.06 | −3.18 | −4.05 | ||
| RZI | −19.55* | −30.60* | 4.39 | −18.41* | 23.89* | 12.36 | −11.22 | −22.76* | ||
| RZII | −24.68 | −31.22 | −3.45 | −17.34 | 21.28 | 14.07 | −6.66 | −13.91 | ||
| CMT | SZ | 2.91 | 11.24 | 36.49* | 37.03* | 33.67* | 26.06* | 8.34 | 0.56 | |
| TZ | −3.10 | −11.15 | 16.92 | 13.64 | 19.99 | 24.70 | −8.06 | −3.30 | ||
| RZI | −9.61 | −9.27 | 8.77 | 8.64 | 18.33 | 17.87 | 0.34 | −0.13 | ||
| RZII | −23.78 | 0.62 | 2.81 | 10.24 | 26.55 | 9.62 | 24.39 | 7.44 | ||
| IMT | SZ | 39.90* | 10.44 | 21.10 | 31.47* | −19.20 | 21.20 | −29.76* | 10.54 | |
| TZ | 6.52 | −13.97 | 6.60 | 2.18 | 0.08 | 16.14 | −20.45 | −4.43 | ||
| RZI | −13.27 | −13.66 | 2.08 | −3.89 | 15.34 | 9.78 | −0.39 | −5.97 | ||
| RZII | −5.52 | 7.46 | −8.85 | −5.54 | −3.75 | −12.98 | 12.96 | 3.72 |
When comparing 8C to N, the most striking observation was that in μMRI 0° data, T2 in AMT and PMT (which are positioned as the most anterior and posterior locations of the tibial surfaces – both are covered by the meniscus) had the opposite trends. While T2 changes in AMT statistically decreased in all zones, T2 changes in PMT statistically increased in all zones. More interestingly, despite the opposite changes between AMT and PMT, the T2 in SZ at both locations ended with statistically significant increases at the late stage of the degradation (8OA and 12OA). In addition, the T2 in CMT, which is between AMT and PMT, share the same significance as in PMT. Most mMRI T2 data in the surface tissue also showed significant increases at advanced OA.
When comparing 12C to N, μMRI data showed more significant T2 decreases in EMT and CMT than mMRI data, especially for EMT where μMRI 55° T2 in all zones statistically decreased, which was followed by significant increases in 8OA and 12OA. Between (8C – N) and (12C – N), there were more decreases (more blues) in the (12C – N) comparisons.
When comparing 8OA or 12OA to N, more significant increases in T2 could be found in all three protocols – more larger increases in the superficial tissue (SZ, TZ) and more changes in several locations (EMT, IMT), with the exception of 8OA μMRI 55° data in AMT, where significant T2 reductions were measured.
Summarizing Fig 5, the T2 in SZ had the highest increases among all zones (comparing all rows) when the disease progressed, which was evident also from the profiles in Fig 3. Among different imaging protocols, μMRI 55° had the most consistent sensitivities, reflected also by the nearly consistent slopes of the linear regression lines in the Supplement Fig 1. We found that T2 changes measured for each topographical site have no direct or indirect correlation to the disease progression but directly correlated based on the zonal depth, resolution and orientation of articular cartilage.
3.4. Zonal T2 OA progression between the covered and exposed regions
To correlate with some previously published regional analysis [20, 36], we grouped all T2 data at five locations (as in Fig 5) into two regions, the meniscus-covered and exposed regions, and showed them in Fig 6. In essence, Fig 5 was condensed into Fig 6 with the same use of colormaps based on the percentage significance on the regional analysis (Table 2). This average from 5 locations to 2 regions strengthened the statistical power of the data analysis, at the price of losing the ability to identify the trends in individual topographical locations. The major observations in this regional analysis (Fig 6) agree with the topographical location analysis (Fig 5). Some of the highlights are:
Fig 6.
Regional average T2 values of cartilage in all imaging protocols and two grouped topographical regions, starting with the normal to each of the disease stages through all of the histological zones (SZ, TZ, RZI and RZII). Other labels, font definitions and colormaps are the consistent with Fig 5. The mean T2 values with underlined bold font imply statistical significance in pair-wise comparison. Color coding (blue white red) is based on the calculated percentage difference between each disease stage to the healthy counterpart: μMRI 0° and μMRI 55° are −20% to +20%, mMRI is −30% to +30%. The actual percentage differences are in Table 2.
Table 2.
Quantitative T2 percent differences of all imaging protocols of two topographical regions (covered and exposed), which were pooled from the five location-data in Table 1. The values in this table were used to color Fig 6. Any number marked with * signifies the statistical significance (p<0.05) within the pairwise comparison.
| site | Zone | N v. 8C | N v. 12C | N v. 8OA | N v. 12OA | 8C v. 8OA | 12C v.12OA | 8C v. 12C | 8OA v.12OA | |
|---|---|---|---|---|---|---|---|---|---|---|
| μMRI 0° | Covered | SZ | 6.96* | −3.92 | 16.84* | 17.13* | 9.91* | 21.02* | −10.88* | 0.30 |
| TZ | 2.36 | −3.30 | 1.82 | −0.95 | −0.53 | 2.35 | −5.65 | −2.77 | ||
| RZI | 0.59 | 0.61 | 2.11 | 2.60 | 1.52 | 1.99 | 0.02 | 0.49 | ||
| RZII | −0.23 | 7.09* | −0.33 | 16.16* | −0.10 | 9.10* | 7.31* | 16.49* | ||
| Exposed | SZ | 11.89* | −9.60* | 18.71* | 1.78 | 6.85 | 11.38* | −21.43* | −16.94* | |
| TZ | 10.22* | −3.71 | 10.07* | −1.25 | −0.16 | 2.46 | −13.92* | −11.31* | ||
| RZI | 14.47* | 2.88 | 17.32* | 2.18 | 2.87 | −0.70 | −11.60* | −15.15* | ||
| RZII | 9.39* | 8.22* | 5.77 | 3.21 | −3.63 | −5.01 | −1.18 | −2.57 | ||
| μMRI 55° | Covered | SZ | 5.42 | −8.12* | 7.87* | 14.21* | 2.45 | 22.26* | −13.52* | 6.35 |
| TZ | 3.99 | −6.51 | 7.37* | 11.60* | 3.38 | 18.08* | −10.49* | 4.25 | ||
| RZI | 2.48 | −3.24 | 3.27 | 5.08 | 0.79 | 8.31* | −5.71 | 1.81 | ||
| RZII | 0.43 | −5.78 | −5.15 | −6.91* | −5.58 | −1.13 | −6.21 | −1.76 | ||
| Exposed | SZ | 7.83* | −5.69 | 11.38* | 8.71* | 3.56 | 14.38* | −13.50* | −2.68 | |
| TZ | 5.43 | −6.97* | 11.00* | 3.73 | 5.58 | 10.70* | −12.39* | −7.27 | ||
| RZI | 1.91 | −0.23 | 12.25* | 3.43 | 10.34* | 3.66 | −2.14 | −8.83* | ||
| RZII | 0.56 | 7.33* | 8.47* | 10.72* | 7.91* | 3.40 | 6.76 | 2.25 | ||
| mMRI | Covered | SZ | 7.24 | 16.37* | 31.01* | 34.80* | 23.91* | 18.70* | 9.16 | 3.90 |
| TZ | −0.84 | −1.20 | 19.14* | 20.52* | 19.97* | 21.71* | −0.37 | 1.40 | ||
| RZI | −7.06 | −11.53 | 14.64* | 7.92 | 21.64* | 19.41* | −4.48 | −6.74 | ||
| RZII | −10.72 | −14.02 | 11.68 | 5.07 | 22.34 | 19.05 | −3.30 | −6.63 | ||
| Exposed | SZ | 21.72* | 10.82 | 28.87* | 33.84* | 7.26 | 23.23* | −10.96 | 5.10 | |
| TZ | 1.89 | −12.57 | 11.83 | 8.01 | 9.95 | 20.52* | −14.46 | −3.84 | ||
| RZI | −11.46 | −11.49 | 5.40 | 2.43 | 16.84* | 13.91 | −0.03 | −2.98 | ||
| RZII | −13.50 | 4.39 | −3.55 | 2.00 | 9.95 | −2.39 | 17.86 | 5.55 |
The effect of location average can be seen clearly when comparing the μMRI 0° 8C regional data (the top two 2nd columns in Fig 6) with the same 8C location data (the top five 2nd columns in Fig 5). The opposite trends in two 8C locations (AMT and PMT) in Fig 5 were averaged to result in a mostly non-significant 8C covered data in Fig 6. At the same time, the consistent trend in both 8C CMT and IMT data in Fig 5 were averaged to result in more significant data across all four sub-tissue zones in the exposed region. Similar observations can be found in other disease progression and other imaging protocols, for example, the μMRI 0° 8OA data between the location analysis (Fig 5) and the regional analysis (Fig 6).
When comparing two contralateral cartilages (8C and 12C in Fig 6), the most striking feature is mostly T2 increases (red) in 8C and mostly T2 decreases (blue) in 12C, in both μMRI 0° and μMRI 55°. Interestingly, the mMRI sensitivity was very different from μMRI (due to the lack of resolution in mMRI), which had consistent results between 8C and 12C.
When comparing two OA cartilages (8OA and 12OA in Fig 6), 8OA had on average a higher T2 than 12OA, especially in the exposed regions, which solidified the observation of different trends in the zonal changes in the T2 profiles in Fig 2 and 3b.
4. Discussion
To the best of our knowledge, this was the first highest multi-resolution MRI study of five topographical patterns of T2 relaxation in tibial surface during OA progression in a well-established animal model, which is a major step above the previous studies where only two regional differences were analyzed [4, 8]. The use of multi-resolution enabled the correlation between the zonal characteristics at high μMRI resolution and the intact joint features at clinical MRI resolution. The degradation was initiated by the transection surgery of ACL in one knee with two waiting periods, which precisely defined the degradation period. A number of significant results were observed in the imaging study, including the overall positive correlation between the T2 values and a number of representative parameters (e.g., number of weeks after surgery) in tissue degradation, the topographical nature of tissue degradation over the tibial surface, and the depth-dependent nature of morphological degradation where the surface tissue had the most changes. The complexity of these findings were based on multiple parameters that interplay with the data and their interpretation, which included the disease progression over time, the disease detection based on imaging protocols, the disease progression per location or region, and the disease progression in depth over the zones. This study enlightened the importance and drawbacks to the study of OA progression comparing factors of overall tissue versus topographical and depth dependent.
4.1. Influences on disease detection due to imaging protocols
As shown clearly in Fig 3, tissue degradation can be detected by the MRI T2 protocols. There are two subtle influences on the sensitivity of T2 detection to cartilage degradation. First, articular cartilage is known to have a T2 anisotropy in MRI, due to the influences of the collagen orientation in the external magnetic field [23, 37]. When the surface normal of cartilage is in parallel with B0, the dipolar interaction has the strongest attenuation effect on T2, making T2 have the shortest value and reduce its sensitivity to tissue degradation. In contrast, when the surface normal of cartilage is at 55° with B0 (the magic angle), the dipolar interaction is minimized, making T2 have the highest sensitivity to tissue degradation. This orientational influence on T2 sensitivity was the reason to measure T2 at two different orientations (0° and 55°) in this study. In human MRI (and similarly in our mMRI), the orientation of an intact joint in the magnet is about 0° and the topographical ROIs selected from mMRI were predominately at 0° orientation to the magnetic field, which explains the similarity between the mMRI profiles (Fig 3a) and the μMRI 0° profiles (Fig 3b).
Second, for a thin layer of cartilage with a strong depth dependence across its thin thickness, the imaging resolution significantly impacts the outcome of OA detection by any imaging method. At low resolution, a large voxel contains more tissue, not only different structured cartilage but also neighboring tissues (synovial fluid, meniscus, membranes, etc.) – all having very different MRI properties than cartilage. For these canine tissues, the zonal thicknesses have been determined in several previous studies using optical imaging, as approximately SZ 80μm, TZ 120μm, and RZ 600μm [4, 35]. At the mMRI resolution of 100μm/pixel, both surface zones (SZ, TZ) were just one pixel each. At the resolution of 17.6μm/pixel in μMRI, there were about 5 pixels in SZ and 7 pixels in TZ, which were adequate to resolve these thin zones. This was the reason to quantify cartilage at two different MRI resolutions in this study, since clinical MRI typically resolves the whole thickness of human cartilage (which is much thicker) in 3–5 pixels. Consequently, our mMRI of canine cartilage mimics almost exactly the tissue averaging scenario in clinical MRI of human cartilage with an improved unequally divided zonal division of cartilage [32]; at the same time, our μMRI has the ability to examine cartilage degradation in a more controlled manner than what is available for clinical study [38]. Hence, the results from our multi-resolution MRI methodology can provide an invaluable translational pathway, from bench to bedside.
Third, to avoid major effects in tissue averaging, the slice dimension was selected tangentially along the surface of the joint, which recognized the fact that the topographical variation over the surface has the smallest variation over a millimeter scale. Since the fastest variation of cartilage is its depth-dependent zonal changes, we set the higher resolution dimension (100μm and 17.6μm) over the depth and capitalized on the SNR improvement of a thicker (0.8mm and 1mm) slice. We consider this orientation to be the best choice in cartilage imaging. Still, the voxel size in μMRI was merely 2.5% of the voxel size in mMRI. Having a finer resolution also increases the experimental noises in quantitative imaging, which complicates a straightforward comparison of T2 changes by different resolutions. For this reason, we developed an average-based percentage difference to compare each group of disease stage from healthy or otherwise within the disease group (8C and 8OA; 12C and 12OA), which proves beneficial in the major differences posed when comparing higher resolution with low resolution data, as well as with gradient-echo based sequence to a quantitatively improved spin-echo sequence. We noticed in this analysis that one should be careful in comparing the data from different imaging methods, not represented the data without a percentage-difference based comparison.
4.2. Variations in disease progression.
OA in humans can take multiple years to progress, which begins with increased synovial inflammation and an upregulation of degradative enzymes. The avascular and aneural natures of cartilage structure keeps the early degradation silent to the patients, who only feel the degradation at its later stages when the pain and stiffness occur. The use of animal models offers a unique advantage for the study of OA, since the degradation initiation event can be precisely defined [39]. The data in this project showed that cartilage at different post-surgical times had different signatures of the degradation, which had inconsistent topographical T2 patterns between contralateral and OA knees. The swelling seen in the images of Fig 2 clearly showed the disruption of the collagen network in both stages of OA cartilage in the presence of sufficient PG in the tissue (otherwise the swelling won’t be clear). Between the 8-weeks and 12-weeks post-surgery, it seemed the 8-wk OA had fairly incremental changes when compared with the later timepoint (12 weeks). This observation was consistent with an earlier anatomic/morphologic report [18] that was part of this study with similar time points, where the 8-week OA knees were found to have worse OA changes than the 12-week or even 24-week knees post-surgery. These findings could be caused by the overcompensating damages (e.g., swelling) in the 8wk OA joint as well as the 8wk C joint, which recovered partially at the later time points.
Coming back to the biology of cartilage degradation, the onset of OA causes the tissue to have more water and less PG. However, any change to a complex tissue would have consequences. More water and less PG also change the local orientation and concentration of the collagen fibers, both of which strongly influence the quantitative measurement of T2. As shown clearly in the magic angle effect of cartilage in MRI, the fibril orientation can both reduce or increase the T2 values in cartilage tissue, depending upon the precise modification to the local fibril matrix [4, 8]. There are therefore multiple factors that can vary the quantative measurement, towards opposite directions, on a site-specific, resolution-dependent and orientation dependent manner. Further studies are needed to verify the biological and biochemical origins of this observation.
4.3. Variations in zonal disease progression.
T2 in the surface cartilage has showed high sensitivity towards the disease progression. As evident from Fig 3 and other parts of the results in this report, the top ¾ of cartilage had a rapid degradation progression during the first 8 weeks post-surgery (Fig 3b), consistent with the finding in previous studies [8]. Between 8-weeks and 12-weeks post-surgery, most of additional degradation occurs in the middle tissue (0.15 – 0.8 relative depth). Given the fact that the early lesions in human OA are known to localize near/around the surface of the tissue [40–42], the use of high resolution in imaging will increase the detection sensitivity and differentiation of different tissue degradation. With the improvement of clinical MRI technology, resolving cartilage thickness in more pixels could become possible for human MRI. It should be noted that since the imaging pixel size in MRI has a consistent size while the sub-tissue zones in cartilage have variable thickness, it is the thinnest zone in cartilage that places the requirement for the minimum resolution that is adequate to resolve the zones in MRI.
4.4. Variations in topographical disease progression.
This study demonstrates clearly that the tissue degradation is not uniform across the joint surface; instead, different trends of degradation occur at different locations over a joint surface. This location variability is compounded with the zonal variability, making the patterns of OA degradation in the tibial cartilage a time-dependent three-dimensional matrix. As evident from Fig 5 and Table 1 in this report, the degradation measured in 8C between AMT and PMT had opposite trends by both μMRI 0° and 55° protocols. While T2 showed significant decreases in AMT, the trends in PMT (and CMT) were the opposite – both locations had consistent and statistically significant increases over all zones. This is likely caused by the differences in the local loading pattern on the tibial surface as well the early degradation measured at 8-week time.
Two different topographical analyses were used in this project, the location analysis where each of the five particular locations on the medial tibial surface were studied individually, and the regional analysis where the medial tibial surface was divided into two regions, covered by the meniscus and not covered by the meniscus. The regional analysis certainly benefits from a stronger statistical power in data analysis, which was used in several of the previous publications [20, 21, 36]. The price to pay for a regional analysis is that T2 trends are averaged over a much larger area on the tibial surface, where different locations within one area are known to progress differently during the tissue degradation. Previous studies have also shown that, after OA progression, the meniscus-covered area changes from the defined healthy meniscal and exposed regions on the tibia [7, 18], causing the grouping of sites to move from covered to exposed regions, affecting the interpretation of disease progression in covered and exposed regions. This project kept the same group from healthy to each disease stage for consistency whereas the effect due to OA may change the assignment of the site from one group to another. This project clearly demonstrates the beneficial value from a more detailed location analysis, and perhaps a dynamic grouping from locations to regions. The results from this project provide a background knowledge required for future study of localized knee OA.
5. Conclusion
The influences of several experimental parameters and methods to the detection of early OA in cartilage by the ACL transection have been studied in this project. We observed the positive correlation between the T2 values and tissue degradation, the depth dependence of tissue degradation where the surface tissue had the most changes, and the location-dependence of tissue degradation over the tibial surface. To the best of our knowledge, the μMRI results are the first quantitative and multi-resolution characterization of cartilage at five independent locations over the medial tibial surface in an animal model of OA. We showed that in the early degradation of OA, the changes in cartilage properties are influenced strongly by the ability to image small location topographically and small tissue depths zonally. A topographical view with fine zonal resolution is crucial for diagnostic detection of early OA by any imaging tool. In addition, the contralateral cartilage can have different degradation patterns than seen in OA cartilage. The use of MRI imaging protocols can also influence the data interpretation. The use of dual resolutions in MRI in this project provides a scalable translational pathway between bench and bedside.
Supplementary Material
Linear regression plots of zonal averaged T2 values, comparing healthy (N, horizontal axis) and disease (vertical axis). (Green 8C, blue 8OA, yellow 12C, and red 12OA; solid square AMT, solid circle EMT, diamond PMT, up triangle CMT, down triangle IMT). For each row of data, the differences between the min T2 and max T2 are kept the same.
Acknowledgements
Y. Xia is grateful to the National Institutes of Health for the R01 grants (AR052353, AR069047). The authors are in debt to Dr J Matyas (Faculty of Veterinary Medicine, University of Calgary, Canada) for providing the animal tissues, and grateful to Drs. J R Ewing, G Ding, and Q Jiang (Neurology Dept, Henry Ford Hospital, Michigan) for access and assistance to the 7T/20cm MRI scanner. The authors thank Ms. Carol Searight (Dept of Physics, Oakland University) for the linguistic editing of the final manuscript.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Supplemental Analysis
The data shown in the colored ‘heat’ maps as in Fig 5 were also analyzed using the linear regression analysis, by comparing the T2 values between healthy and disease tissues, in all zones and at all topographical locations. The supplement Fig 1 shows the summary of this analysis effort. In these regression plots, the solid diagonal lines indicate that the T2 of the disease tissues (vertical axis) is the same as the healthy tissues (horizontal axis). Each data point represents the average T2 value, which are shape-coded, as well as color-coded. When the data points are above/below the diagonal lines, the disease tissues have a higher/lower T2 than the healthy tissue respectively. The trends of T2s for each category were fitted linearly and shown as the colored dotted lines. The major findings from this analysis support the observations from the color maps in Fig 5.
Conflicts of Interest
None.
Ethical Approval
This study was approved by the local institutional review board, IACUC.
References:
- [1].Maroudas A Physicochemical properties of articular cartilage. In: Freeman MAR, editor Adult articular cartilage, 2nd ed. Kent, England: Pitman Medical; 1979, p. 215–90. [Google Scholar]
- [2].Buckwalter JA, Mankin HJ, Grodzinsky AJ. Articular cartilage and osteoarthritis. Instructional Course Lectures-American Academy of Orthopaedic Surgeons 2005;54:465. [PubMed] [Google Scholar]
- [3].Broom ND. The collagen framework of articular cartilage: Its profound influence on normal and abnormal load-bearing function. In: Nimni ME, editor Collagen - Biochemistry and Biomechanics. Boca Raton, Florida: CRC Press; 1988. [Google Scholar]
- [4].Xia Y, Moody J, Burton-Wurster N, Lust G. Quantitative In Situ Correlation Between Microscopic MRI and Polarized Light Microscopy Studies of Articular Cartilage. Osteoarthritis Cartilage 2001;9(5):393–406. [DOI] [PubMed] [Google Scholar]
- [5].Bijlsma JW, Berenbaum F, Lafeber FP. Osteoarthritis: an update with relevance for clinical practice. Lancet 2011;377:2115–26. [DOI] [PubMed] [Google Scholar]
- [6].Xia Y Heterogeneity of Cartilage Laminae in MR Imaging. J Magn Reson Imaging 2000;11(6):686–93. [DOI] [PubMed] [Google Scholar]
- [7].Appleyard RC, Burkhardt D, Ghosh P, Read R, Cake M, Swain MV, et al. Topographical analysis of the structural, biochemical and dynamic biomechanical properties of cartilage in an ovine model of osteoarthritis. Osteoarthritis Cartilage 2003;11(1):65–77. [DOI] [PubMed] [Google Scholar]
- [8].Lee JH, Badar F, Kahn D, Matyas J, Qu X, Chen CT, et al. Topographical Variations of the Strain-dependent Zonal Properties of Tibial Articular Cartilage by Microscopic MRI. Connect Tissue Res 2014;55(3):205–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Makela JT, Rezaeian ZS, Mikkonen S, Madden R, Han SK, Jurvelin JS, et al. Site-dependent changes in structure and function of lapine articular cartilage 4 weeks after anterior cruciate ligament transection. Osteoarthritis Cartilage 2014;22(6):869–78. [DOI] [PubMed] [Google Scholar]
- [10].McLeod WD, Moschi A, Andrews JR, Hughston JC. Tibial plateau topography. The American Journal of Sports Medicine 1977;5(1):13–8. [DOI] [PubMed] [Google Scholar]
- [11].Hall FM, Wyshak G. Thickness of articular cartilage in the normal knee. J Bone Joint Surg [Am] 1980;62(3):408–13. [PubMed] [Google Scholar]
- [12].Xia Y, Momot KI, Chen Z, Chen CT, Kahn D, Badar F. Introduction to Cartilage. In: Xia Y, Momot KI, editors. Biophysics and Biochemistry of Cartilage by NMR and MRI. The Royal Society of Chemistry; 2017, p. 1–43. [Google Scholar]
- [13].Buckland-Wright JC, Macfarlane DG, Lynch JA, Jasani MK, Bradshaw CR. Joint space width measures cartilage thickness in osteoarthritis of the knee: high resolution plain film and double contrast macroradiographic investigation. Ann Rheum Dis 1995;54(4):263–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Watson PJ, Carpenter TA, Hall LD, Tyler JA. Cartilage swelling and loss in a spontaneous model of osteoarthritis visualized by magnetic resonance imaging. Osteoarthritis Cartilage 1996;4(3):197–207. [DOI] [PubMed] [Google Scholar]
- [15].Matyas JR, Ehlers PF, Huang D, Adams ME. The early molecular natural history of experimental osteoarthritis. I. Progressive discoordinate expression of aggrecan and type II procollagen messenger RNA in the articular cartilage of adult animals. Arthritis Rheum 1999;42(5):993–1002. [DOI] [PubMed] [Google Scholar]
- [16].Vignon E, Piperno M, Le Graverand MP, Mazzuca SA, Brandt KD, Mathieu P, et al. Measurement of radiographic joint space width in the tibiofemoral compartment of the osteoarthritic knee: comparison of standing anteroposterior and Lyon schuss views. Arthritis Rheum 2003;48(2):378–84. [DOI] [PubMed] [Google Scholar]
- [17].Huang M, Schweitzer ME. The role of radiology in the evolution of the understanding of articular disease. Radiology 2014;273(2 Suppl):S1–22. [DOI] [PubMed] [Google Scholar]
- [18].Kahn D, Mittelstaedt D, Matyas J, Qu X, Lee JH, Badar F, et al. Meniscus Induced Cartilaginous Damage and Non-linear Gross Anatomical Progression of Early-stage Osteoarthritis in a Canine Model. The open orthopaedics journal 2016;10:690–705. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Xia Y MRI of articular cartilage at microscopic resolution. Bone and Joint Res 2013;2(1):9–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Alhadlaq H, Xia Y, Moody JB, Matyas J. Detecting Structural Changes in Early Experimental Osteoarthritis of Tibial Cartilage by Microscopic MRI and Polarized Light Microscopy. Ann Rheum Dis 2004;63(6):709–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Lee JH, Badar F, Kahn D, Matyas J, Qu X, Xia Y. Loading-induced changes on topographical distributions of the zonal properties of osteoarthritic tibial cartilage - A study by magnetic resonance imaging at microscopic resolution. J Biomech 2015;48(13):3625–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Kurkijarvi JE, Nissi MJ, Kiviranta I, Jurvelin JS, Nieminen MT. Delayed gadolinium-enhanced MRI of cartilage (dGEMRIC) and T2 characteristics of human knee articular cartilage: topographical variation and relationships to mechanical properties. Magn Reson Med 2004;52(1):41–6. [DOI] [PubMed] [Google Scholar]
- [23].Xia Y Relaxation Anisotropy in Cartilage by NMR Microscopy (μMRI) at 14 μm Resolution. Magn Reson Med 1998;39(6):941–9. [DOI] [PubMed] [Google Scholar]
- [24].Xia Y, Zheng S, Szarko M, Lee J. Anisotropic properties of bovine nasal cartilage. Microsc Res Tech 2012;75(3):300–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Li X, Benjamin Ma C, Link TM, Castillo DD, Blumenkrantz G, Lozano J, et al. In vivo T(1rho) and T(2) mapping of articular cartilage in osteoarthritis of the knee using 3 T MRI. Osteoarthritis Cartilage 2007;15(7):789–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Stahl R, Blumenkrantz G, Carballido-Gamio J, Zhao S, Munoz T, Hellio Le Graverand-Gastineau MP, et al. MRI-derived T2 relaxation times and cartilage morphometry of the tibio-femoral joint in subjects with and without osteoarthritis during a 1-year follow-up. Osteoarthritis Cartilage 2007;15(11):1225–34. [DOI] [PubMed] [Google Scholar]
- [27].Millington SA, Li B, Tang J, Trattnig S, Crandall JR, Hurwitz SR, et al. Quantitative and topographical evaluation of ankle articular cartilage using high resolution MRI. J Orthop Res 2007;25(2):143–51. [DOI] [PubMed] [Google Scholar]
- [28].Hannila I, Raina SS, Tervonen O, Ojala R, Nieminen MT. Topographical variation of T2 relaxation time in the young adult knee cartilage at 1.5 T. Osteoarthritis Cartilage 2009;17(12):1570–5. [DOI] [PubMed] [Google Scholar]
- [29].Hanninen N, Rautiainen J, Rieppo L, Saarakkala S, Nissi MJ. Orientation anisotropy of quantitative MRI relaxation parameters in ordered tissue. Scientific reports 2017;7(1):9606. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].McDevitt C, Gilbertson E, Muir H. An experimental model of osteoarthritis; early morphological and biochemical changes. J Bone Joint Surg Br 1977;59(1):24–35. [DOI] [PubMed] [Google Scholar]
- [31].Matyas JR, Atley L, Ionescu M, Eyre DR, Poole AR. Analysis of cartilage biomarkers in the early phases of canine experimental osteoarthritis. Arthritis Rheum 2004;50(2):543–52. [DOI] [PubMed] [Google Scholar]
- [32].Badar F, Xia Y. Image interpolation improves the zonal analysis of cartilage T2 relaxation in MRI. Quantitative imaging in medicine and surgery 2017;7(2):227–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [33].Zheng S, Xia Y. Multi-components of T2 relaxation in ex vivo cartilage and tendon. J Magn Reson 2009;198(2):188–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [34].Lee JH, Xia Y. Quantitative zonal differentiation of articular cartilage by microscopic magnetic resonance imaging, polarized light microscopy, and Fourier-transform infrared imaging. Microsc Res Tech 2013;76:625–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [35].Lee J, Xia Y. Quantitative in situ correlation between μMRI, PLM, and FTIRI studies of articular cartilage. the 57th Conference of Orthopaedic Research Society. Long Beach, California; 2011:1609. [Google Scholar]
- [36].Lee JH, Badar F, Matyas J, Qu X, Xia Y. Topographical variations in zonal properties of canine tibial articular cartilage due to early osteoarthritis: a study using 7-T magnetic resonance imaging at microscopic resolution. MAGMA 2016;29(4):681–90 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [37].Xia Y, Farquhar T, Burton-Wurster N, Lust G. Origin of cartilage laminae in MRI. J Magn Reson Imaging 1997;7(5):887–94. [DOI] [PubMed] [Google Scholar]
- [38].Xia Y. Resolution ‘scaling law’ in MRI of articular cartilage. Osteoarthritis Cartilage 2007;15(4):363–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [39].Pond MJ, Nuki G. Experimentally-induced osteoarthritis in the dog. Ann Rheum Dis 1973;32(4):387–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [40].Hwang WS, Li B, Jin LH, Ngo K, Schachar NS, Hughes GN. Collagen fibril structure of normal, aging, and osteoarthritic cartilage. J Pathol 1992;167(4):425–33. [DOI] [PubMed] [Google Scholar]
- [41].Squires GR, Okouneff S, Ionescu M, Poole AR. The pathobiology of focal lesion development in aging human articular cartilage and molecular matrix changes characteristic of osteoarthritis. Arthritis Rheum 2003;48(5):1261–70. [DOI] [PubMed] [Google Scholar]
- [42].Pritzker KP, Gay S, Jimenez SA, Ostergaard K, Pelletier JP, Revell PA, et al. Osteoarthritis cartilage histopathology: grading and staging. Osteoarthritis Cartilage 2006;14(1):13–29. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Linear regression plots of zonal averaged T2 values, comparing healthy (N, horizontal axis) and disease (vertical axis). (Green 8C, blue 8OA, yellow 12C, and red 12OA; solid square AMT, solid circle EMT, diamond PMT, up triangle CMT, down triangle IMT). For each row of data, the differences between the min T2 and max T2 are kept the same.