Longitudinal assessment of synovial lymph node and bone volumes in inflammatory arthritis in mice by in vivo magnetic resonance imaging and microfocal computed tomography.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 56, No. 12, December 2007, pp 4024–4037 DOI 10.1002/art.23128 © 2007, American College of Rheumatology Longitudinal Assessment of Synovial, Lymph Node, and Bone Volumes in Inflammatory Arthritis in Mice by In Vivo Magnetic Resonance Imaging and Microfocal Computed Tomography Steven T. Proulx,1 Edmund Kwok,1 Zhigang You,1 M. Owen Papuga,1 Christopher A. Beck,1 David J. Shealy,2 Christopher T. Ritchlin,1 Hani A. Awad,1 Brendan F. Boyce,1 Lianping Xing,1 and Edward M. Schwarz1 measures were used to assess the natural history of erosive inflammatory arthritis. We also performed antiTNF versus placebo efficacy studies in TNF-Tg mice in which treatment was initiated according to either age (4–5 months) or synovial volume (3 mm3 as detected by CE-MRI). Linear regression was performed to analyze the correlation between synovitis and focal erosion. Results. CE-MRI demonstrated the highly variable nature of TNF-induced joint inflammation. Initiation of treatment by synovial volume produced significantly larger treatment effects on the synovial volume (P ⴝ 0.04) and the lymph node volume (P < 0.01) than did initiation by age. By correlating the MRI and micro-CT data, we were able to demonstrate a significant relationship between changes in synovial and patellar volumes (R2 ⴝ 0.75, P < 0.01). Conclusion. In vivo CE-MRI and micro-CT 3-D outcome measures are powerful tools that accurately demonstrate the progression of erosive inflammatory arthritis in mice. These methods can be used to identify mice with arthritis of similar severity before intervention studies are initiated, thus minimizing heterogeneity in outcome studies of chronic arthritis seen between genetically identical littermates. Objective. To develop longitudinal 3-dimensional (3-D) measures of outcomes of inflammation and bone erosion in murine arthritis using contrast-enhanced magnetic resonance imaging (CE-MRI) and in vivo microfocal computed tomography (micro-CT) and, in a pilot study, to determine the value of entry criteria based on age versus synovial volume in therapeutic intervention studies. Methods. CE-MRI and in vivo micro-CT were performed on tumor necrosis factor–transgenic (TNFTg) mice and their wild-type littermates to quantify the synovial and popliteal lymph node volumes and the patella and talus bone volumes, respectively, which were validated histologically. These longitudinal outcome Supported by research grants from Centocor and by grants from the NIH (USPHS grants AR-43510, AR-46545, AR-48697, AR-51469, AR-54041, and DE-17096). 1 Steven T. Proulx, MS, Edmund Kwok, PhD, Zhigang You, MS, M. Owen Papuga, MS, Christopher A. Beck, PhD, Christopher T. Ritchlin, MD, Hani A. Awad, PhD, Brendan F. Boyce, MD, Lianping Xing, PhD, Edward M. Schwarz, PhD: University of Rochester, Rochester, New York; 2David J. Shealy, PhD: Centocor Research and Development, Radnor, Pennsylvania. Dr. Shealy owns stock or stock options in Johnson & Johnson, of which Centocor is a subsidiary. Dr. Ritchlin has received grants from Centocor for research related to disease mechanisms of inflammatory arthritis. Dr. Boyce has received consulting fees, speaking fees, and/or honoraria from Merck and Amgen (less than $10,000 each) and from Ariad Pharmaceuticals (more than $10,000) and has provided expert testimony for Ariad Pharmaceuticals. Dr. Schwarz has received consulting fees, speaking fees, and/or honoraria (less than $10,000 each) from Centocor and Amgen, owns stock or stock options in Amgen, and has provided expert testimony for Amgen. Address correspondence and reprint requests to Edward M. Schwarz, PhD, The Center for Musculoskeletal Research, University of Rochester Medical Center, 601 Elmwood Avenue, Box 665, Rochester, NY 14642. E-mail: Edward_Schwarz@URMC.Rochester.edu. Submitted for publication April 9, 2007; accepted in revised form August 31, 2007. Although murine models of inflammatory arthritis have significantly advanced our understanding of erosive inflammatory arthritis (1,2), they are limited by a lack of longitudinal translational outcome measures of disease progression or interventional therapy. This issue presents 3 problems for prototypical preclinical investigations of drug effects on inflammation, erosion, and healing (3,4). First, most cross-sectional studies in mice with “established arthritis” do not include an objective 4024 MRI AND MICRO-CT ASSESSMENTS IN MICE WITH INFLAMMATORY ARTHRITIS assessment of disease severity prior to treatment. Thus, neither rates of change nor healing responses can be assessed. Second, the commonly used end points (i.e., histology and ex vivo molecular analyses) require the death of the mice, thus markedly increasing the number of animals needed to assess efficacy at multiple time points. Third, although the incidence and severity of arthritis vary among genetically identical littermates, there are no established scoring criteria to stratify different groups of mice based on disease activity in an intervention study. Therefore, the development of imaging techniques that could assess disease activity and progression in vivo would greatly enhance the utility of these preclinical studies. Magnetic resonance imaging (MRI) has become the “gold standard” for the assessment of joint inflammation and damage in inflammatory arthritis (5–8). Several studies have demonstrated the value of MRI in the detection of synovial inflammation and bone marrow edema before irreversible joint damage occurs (9–14). While quantitative measures of these imaging biomarkers have been developed, their validation has been difficult to perform during clinical studies (15–18). Conversely, while histologic analysis of arthritis in animal models is readily available to validate imaging biomarkers, initial attempts to use this longitudinal in vivo imaging modality in mice have fallen short of the desired 3-dimensional (3-D) quantitative outcome measure (19– 22). The purpose of the current study was to develop and validate in vivo quantitative 3-D imaging biomarkers of erosive inflammatory arthritis in mice. Although many different animal models exist (1,2), we chose the human tumor necrosis factor–transgenic (TNF-Tg) mouse largely because it is a model of chronic disease of known etiology that is completely ameliorated by antiTNF therapy (23). Furthermore, a critical role of TNF in rheumatoid arthritis (RA) has been firmly established by the success of treatment with TNF antagonists (24). Thus, achievement of predicted outcomes in this wellestablished model serves as a validation of the novel 3-D imaging biomarkers and has important translational value as well, given the histologic findings that are similar to those in RA in humans. Using a custom-designed murine knee coil for MRI, we have developed volumetric quantifications for 2 outcome measures: synovial inflammation and popliteal lymph node enlargement. We demonstrate the progression of these biomarkers in the disease, as well as the reversal of their progression after anti-TNF therapy. We also validate these measurements with histologic 4025 analyses and demonstrate significant correlations between MRI measurements and novel microfocal computed tomography (micro-CT) measurements of bone erosion. MATERIALS AND METHODS Animals and anti-TNF treatment. The 3647 line of TNF-Tg mice were originally obtained from Dr. G. Kollias (Institute of Immunology, Biomedical Sciences Research Center Alexander Fleming, Vari, Greece) and are maintained as heterozygotes in a CBA ⫻ C57BL/6 background (25). Experiments were performed with sex-matched TNF-Tg and wildtype (WT) littermate controls. The University of Rochester Committee for Animal Resources approved all animal studies. An initial natural history study examined TNF-Tg mice and their WT littermates (n ⫽ 5 per group) from the age of 2 months until the age of 5 months. MRI scans were performed twice a month at ⬃2-week intervals, except at 4 months of age because of temporary technical problems with the MR scanner. At 5 months of age, mice were killed, and the knee joints and popliteal lymph nodes were harvested for histology. In the drug-intervention studies, mice received either murine monoclonal anti-human TNF IgG1 antibody or an irrelevant murine IgG1 placebo control (Centocor R&D, Radnor, PA) at a dosage of 10 mg/kg/week by intraperitoneal injection, as previously described (3). Two studies of anti-TNF versus placebo were performed. Study 1 controlled for the age of the mice, and study 2 controlled for MRI-based evidence of synovitis. The age-controlled study consisted of 4 groups of 5–6-month-old mice (n ⫽ 4 per group) at baseline. Groups 1 and 2 were WT littermates that received placebo or anti-TNF, respectively. Groups 3 and 4 were TNF-Tg mice that received placebo or anti-TNF, respectively. Treatments were administered for 16 weeks. MRI scans were performed at baseline and every 4 weeks thereafter in the TNF-Tg groups, and at baseline and 16 weeks in the WT groups. At 16 weeks, mice were killed, and the knee joints and popliteal lymph nodes were harvested for histologic assessment. The synovitis-controlled anti-TNF study consisted of 2 groups, each containing 4 female TNF-Tg mice. These mice were scanned every 2 weeks starting at 3 months of age and were entered into the study when it was determined that they had synovial volumes ⬎3 mm3 (see below). The mean ⫾ SD age at initiation of treatment was 3.88 ⫾ 0.79 months. One group received weekly anti-TNF injections, while the second group was given placebo, as described above. MRI scans were performed at baseline and every 2 weeks for 8 weeks. In vivo micro-CT scans of the knees and ankles were performed at baseline and at 8 weeks. At 8 weeks, mice were killed, and the knee joints, ankle joints, and popliteal lymph nodes were harvested for histologic assessment. Contrast-enhanced MRI (CE-MRI). MR scans were performed with a 3T Siemens Trio MRI (Siemens Medical Solutions, Erlangen, Germany). Mice were anesthetized by intraperitoneal injection of a ketamine/xylazine mixture at a dosage of 110 g/ml. Mice were then positioned on an imaging platform with the right leg inserted through the custom 4026 PROULX ET AL Figure 1. Contrast-enhanced magnetic resonance imaging (CE-MRI) of the mouse knee. A, Anesthetized mouse positioned in the knee surface coil prior to MRI scanning. B and C, Sagittal MR images of a 5-month-old tumor necrosis factor–transgenic mouse obtained precontrast (B) and postcontrast (C). Note the high-resolution of the tibia (t), femur (f), synovium (s), popliteal lymph node (ln), and gastrocnemius muscle (m). D and E, Threedimensional (3-D) imaging using Amira 3.1 software. For 3-D imaging and volumetric quantitation, the precontrast image is first registered to and subtracted from the postcontrast image (D). Then a limit line surrounding the synovium is manually drawn around the region of interest (ROI) on the postcontrast image, and copied to the subtracted image (yellow line). The lymph node is manually segmented as it is clearly visualized on postcontrast images. This is performed on all slices encompassing the knee and lymph node, and the segmented labels are reconstructed as volumes (E) for visualization and quantification. F and G, Determination of the threshold value used to quantify synovial volume in the ROI based on the delivered dose of gadodiamide (Gd-DTPA-BMA). A dosage study was performed to establish a direct linear relationship between the Gd-DTPA-BMA dose and contrast enhancement of the gastrocnemius muscle (F). At constant synovial volume, the threshold used to segment the synovial volume was curve-fitted to the contrast enhancement of the muscle (G). This curve is used to threshold and segment the enhanced synovial area in the drawn ROI from the surrounding tissues, using muscle as a normalization tissue to determine the delivered dose of contrast agent. designed knee coil (Figure 1A). The coil is composed of a 1.5-cm–diameter circular loop consisting of 2 parallel 14-gauge copper wires. This design was found to give optimal signal-tonoise ratio (SNR) while providing sufficient volume coverage of the joint. An imaging template assisted in reproducible positioning between scans and between animals. After a series of three 10-second localization scans, a fat-suppressed T1-weighted high-resolution scan was performed (sagittal T1-weighted fast low-angle shot (FLASH) sequence, consisting of a recovery time of 45 msec, an echo time of 9.03 msec, a 192 ⫻ 192–pixel matrix, with a 20 ⫻ 20–mm field of view, 32 slices of 0.16-mm slice thickness, a flip MRI AND MICRO-CT ASSESSMENTS IN MICE WITH INFLAMMATORY ARTHRITIS angle of 25°, 1 signal average, and a scan time of 8 minutes 28 seconds). Gadodiamide (Gd-DTPA-BMA) contrast agent (Omniscan; Amersham Health, Oslo, Norway) at a dose 0.500 ml/kg diluted in sterile saline was injected via the retroorbital venous plexus. After 3 minutes to allow for Gd-DTPA-BMA to equilibrate with the joint fluid, a second high-resolution scan was performed to image the knee with contrast enhancement. Imaging sessions took ⬃30 minutes per mouse. MRI analysis. Amira 3.1 software (TGS/Mercury Computer Systems, San Diego, CA) was used for image segmentation and volume computations of synovium and popliteal lymph nodes. The 3-D stacks of images for the precontrast (Figure 1B) and postcontrast (Figure 1C) scans were loaded into the software. An automatic registration module (Registration 3 LineSearch 3 Correlation) was used to align the images in 3-D. The precontrast scan was then subtracted from the postcontrast scan using the arithmetic module, resulting in a 3-D stack of images of contrast enhancement (Figure 1D). A segmentation and threshold procedure using the Amira Segmentation Editor was used to determine synovial and lymph node volumes. The editor allows the user to create 3-D “labels” of each tissue of interest. First, limit lines corresponding to the regions of interest (ROIs) enclosing the synovium, but excluding enhancing tissues such as the subcutaneous layer and the popliteal vessels, were drawn on each slice. For visualization purposes, these lines were drawn on the postcontrast image stack, and the labels were copied onto the subtracted image. Next, a section (⬎15 mm3) of muscle tissue was labeled on the subtracted image stack. The TissueStatistics module was used to determine the contrast enhancement of the muscle as an estimate of the delivered dosage of GdDTPA-BMA. Based on the level of muscle enhancement, a threshold value corresponding to synovial enhancement was found using the equation determined by a dosage study (see next section). All voxels above the threshold value within the limit lines were labeled as synovium. Enhanced voxels above this threshold that were within the bone marrow were then subtracted from the label. The lymph node, which is much easier to visualize than synovium, was segmented by manually drawing ROIs on postcontrast images and were thresholded based on signal intensity ⱖ1,500 arbitrary units (AU) to define the boundary between the lymph node and the fat pad surrounding the node. These labels were also copied onto the subtracted image. Rendering as a 3-D image (Figure 1E) was performed by applying the SurfaceGen module to reconstruct the tissue labels into volumes, and the SurfaceView module was used for visualization. The TissueStatistics module was used to quantify the volumes of the tissues. Total time for image analysis by an experienced operator was ⬃20 minutes per mouse. To determine the threshold parameters for longitudinal analyses of synovial volume, we performed a dose-response study in which 4 TNF-Tg mice (3–5 months old) were scanned on 3 consecutive days using incremental dosages of Gd-DTPABMA (dose 1 ⫽ 0.167 ml/kg, dose 2 ⫽ 0.333 ml/kg, and dose 3 ⫽ 0.500 ml/kg). The concentrations of contrast agent in saline were adjusted, such that the bolus injected into each animal was at a constant volume between doses. After 24 hours, no trace of contrast agent was detected on the precon- 4027 trast scans obtained on day 2 or day 3 in the animals. A strong linear relationship between muscle enhancement and dosage was found using a linear mixed-effects regression model (R2 ⫽ 0.87) (Figure 1F). Data from dose 2 were analyzed with the above-described method; however, an initial synovial threshold equal to muscle enhancement plus 1,000 AU was used to generate a synovial volume. The data from day 1 and day 3 were then analyzed, and the threshold value needed to generate the same synovial volume was recorded. The threshold values versus muscle enhancement values for all 4 mice at all 3 doses fit the increasing form of a 2-phase exponential-decay curve (Figure 1G), with R2 ⫽ 0.98. The values from this curve were used to determine synovial threshold values that were adjusted for Gd-DTPA-BMA dosage as measured by contrast enhancement of muscle. Reproducibility for synovial and lymph node volume measurements was assessed for intrareader, interreader, and inter-MRI variations. For determining intrareader variation, 10 CE-MRI scans (5 TNF-Tg and 5 WT mice) were analyzed by the same operator (STP) on 2 different days in random order. The same 10 scans were analyzed by a second operator (MOP) after a training period of 2 hours to determine interreader variation. For inter-MRI variation, 5 TNF-Tg mice with disease of varying severity were scanned twice within 48 hours, and the same operator analyzed the scans in random order. Volumetric assessment of bone erosion via micro-CT. High-resolution in vivo micro-CT was used to scan the mouse knee and ankle joints (VivaCT 40; Scanco, Southeastern, PA). Animals were anesthetized with isoflurane. Each joint was scanned at an isotropic resolution of 17.5 m in a custom sample holder at 55 keV, with cone beam mode. The data were reconstructed via Scanco software into Dicom files for analysis. Amira 3.1 software was used to segment and visualize the bones of the knee and ankle on the micro-CT scans. For the knee, the Segmentation Editor was used to label the femur, tibia, patella, and menisci (Figure 2A). A density threshold ⬎11,000 AU was set as representing “bone,” and the labels were reconstructed using the SurfaceGen module to visualize the joint (Figure 2B). The threshold was kept constant throughout the study. Since the entire patella is scanned, the volume of this bone, as determined from the TissueStatistics module, was used as a quantitative measure of bone erosion for the knee (Figure 2C). For the ankle, the talus was specifically labeled, and the volume of this bone was used as the measure of erosion in this joint (Figure 2D). Analysis typically takes an experienced operator ⬃15 minutes per joint. Histologic assessment. The right knee joint was removed and fixed in 10% neutral buffered formalin. The joints were then decalcified in 14% EDTA at room temperature (pH adjusted to 7.2) for 21 days. The joint was then carefully embedded in paraffin for sectioning into 3-m slices. Sections were then stained with orange G–Alcian blue for histologic examination. The popliteal lymph node was also harvested and prepared in a similar manner, omitting the decalcification step. The lymph node area was quantified using ImageJ software (NIH Image, National Institutes of Health, Bethesda, MD; online at: http://rsbweb.nih.gov/ij/) at the slice found to have maximal cross-sectional area. Statistical analysis. Linear mixed-effects regression models, with mouse as a random effect and time (treated as a 4028 PROULX ET AL analysis of variance models. All underlying assumptions of the parametric methods were checked, and no serious violations were detected. P values less than 0.05 were considered significant, and P values less than 0.01 were considered highly significant. RESULTS Figure 2. Three-dimensional reconstruction and quantification of bone volume from microfocal computed tomography (micro-CT) imaging. A, A representative sagittal slice from a micro-CT scan of the knee of a wild-type mouse, demonstrating the density-based segmentation that is performed on the bone to generate labels for the patella (yellow), femur (light blue), tibia (dark blue), and menisci (red), as described in Materials and Methods. B, Reconstruction of the bones in 3 dimensions. The labels generated from the micro-CT sagittal slice are then used to reconstruct the bones in 3 dimensions. C, Threedimensional image of the patella. Due to its ease of reconstruction and proximity to inflamed synovium, the volume of the patella is used as a quantitative measure of bone volume at the knee joint. D, Threedimensional image of the ankle joint. The ankle joint is reconstructed in a manner similar to that of the knee. The volume of the talus (yellow) is used as a quantitative measure of bone volume in the ankle joint. Inset, Three-dimensional view of the talus, rotated and enlarged. continuous covariate) as a fixed effect, were used to assess changes over time based on longitudinal data. Similar models used age or dose instead of time to assess changes over age or linear dose-response relationships based on repeated measurements. Analyses based on cross-sectional data used standard linear regression models. A nonlinear mixed-effects model, with a random effect for mouse to account for the repeated measures design, was used to fit the 2-phase exponential-decay curve. Two-sided t-tests assuming unequal variances were used to make comparisons with the micro-CT data or with the CE-MRI data between groups at the same time point or age. Correlations between measures were estimated using Pearson’s correlation coefficient and were tested for significance using a 2-sided t-test. Interreader, intrareader, and inter-MRI reliability was estimated using coefficients of variation and intraclass correlation coefficients based on random-effects Longitudinal CE-MRI biomarkers of TNFinduced arthritis in mice. High-resolution MR images of mouse knees clearly defined contrast-enhanced inflamed synovium and enlarged lymph node in TNF-Tg mice (Figure 3A) and permitted volumetric quantification (Figure 3B). These findings were absent in WT littermates (Figures 3E and 3F). The remarkable differences between the joints of the TNF-Tg and WT mice were confirmed histologically (Figures 3C and D and Figures 3G and H, respectively). Before we used these novel imaging outcome measures in prospective studies, we assessed the validity of our lymph node volume calculations by demonstrating correlations with the maximum lymph node area as determined by MRI and histology. (Figures illustrating the validation of the CE-MRI lymph node volume measurements by correlation with the 2-D MRI and histology are available upon request from the corresponding author). Linear regression analyses demonstrated significant relationships (P ⬍ 0.0001) between the 2-D and 3-D measurements and substantiated the validity of our volumetric CE-MRI measurements for use in outcomes studies. Intrareader, interreader, and inter-MRI reliability was assessed with intraclass correlation coefficient (ICC) analysis and found to be excellent for all variables. Synovial volume measurements were found to have ICC values of 0.996 for intrareader, 0.984 for interreader, and 0.968 for inter-MRI reliability. Lymph node volume quantifications had ICC values of 0.999, 0.997, and 0.994, respectively. Additionally, the inter-MRI reliability was assessed with a coefficient of variation and found to be 4.53% for synovial volume and 3.95% for lymph node for n ⫽ 5 TNF-Tg mice. In our first longitudinal study of 3-D CE-MRI biomarkers of inflammatory arthritis in mice, we assessed the progression of disease in 2 month-old TNF-Tg versus WT littermates over a 3-month period. Figure 3I shows that TNF-induced knee synovitis began at ⬃3 months of age and steadily increased thereafter (slope ⫽ 0.60 mm3/month; P ⬍ 0.0001). In contrast, the synovial volume in the WT mice significantly decreased over time (slope ⫽ –0.26 mm3/month; P ⫽ 0.01), likely MRI AND MICRO-CT ASSESSMENTS IN MICE WITH INFLAMMATORY ARTHRITIS 4029 Figure 3. Identification, quantification, and validation of synovial and lymph node volume as longitudinal outcome measures of inflammatory knee arthritis in mice. A–D, A representative 5-month-old transforming growth factor–transgenic (TNF-Tg) mouse. E–H, A 5-month-old wild-type (WT) littermate. Postcontrast magnetic resonance images show enhancing synovium (arrows). Bone marrow edema is present in the TNF-Tg mouse, as indicated by the high signal intensity in the bone marrow space (A), but is absent in the WT mouse (E). Corresponding 3-dimensional reconstructions and calculated synovial (yellow) and popliteal lymph node (red) volumes (B and F) are also shown. The dramatic differences in these quantitative imaging biomarkers are validated in the corresponding orange G–Alcian blue–stained histology sections shown at 40⫻ magnification (C and G); boxed areas of C and G are shown at 200⫻ magnification (D and H, respectively). The TNF-Tg mouse displays thickened synovial lining (#) and infiltration into subchondral bone (arrowhead). I and J, Disease progression in TNF-Tg and WT mice as a function of synovial volume (I) and lymph node volume (J). Mice were scanned bimonthly from the age of 2 months until the age of 5 months (see Materials and Methods), and the synovial and lymph node volumes for each scan were calculated. Values are the mean ⫾ SD of 5 mice per group. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01, versus WT mice of the same age, by 2-sided t-test. Linear mixed-effects regression analysis revealed a highly significant difference in the slopes for both the synovial and lymph node volumes in TNF-Tg mice versus WT mice (P ⬍ 0.0001 for both comparisons). due to growth effects. This resulted in a highly significant difference in slope for the TNF-Tg mice versus the WT mice (P ⬍ 0.0001). While a highly significant difference in synovial volume between TNF-Tg and WT mice was observed at 4.5 months (mean ⫾ SD 3.32 ⫾ 0.93 mm3 versus 1.44 ⫾ 0.42 mm3; P ⬍ 0.01), the significance of this difference decreased at 5 months due to the increased variability in the TNF-Tg group (4.26 ⫾ 1.57 mm3 versus 1.89 ⫾ 0.41 mm3; P ⬍ 0.05). Interestingly, TNF-induced changes in popliteal lymph node volumes preceded knee synovitis (Figure 3J), since the steadily significant increase (slope ⫽ 2.95 mm3/month; P ⬍ 0.0001) in TNF-Tg animals began at the age of 2 months (2.43 ⫾ 0.76 mm3) and plateaued at the age of 4.5 months (10.17 ⫾ 5.69 mm3). Despite the large variability at the end of the study, we still observed a highly significant 10-fold difference in lymph node volume between the TNF-Tg mice and the WT mice at 5 months (10.41 ⫾ 5.57 mm3 versus 1.20 ⫾ 0.34 mm3; P ⬍ 0.01). As expected, the popliteal lymph node volume in the WT mice did not change throughout the study (slope ⫽ –0.10 mm3/month; P ⫽ 0.71). However, 4030 we detected a highly significant difference in slope for the lymph node volume in TNF-Tg mice versus WT mice (P ⬍ 0.0001). Changes in longitudinal 3-D biomarkers of erosive inflammatory arthritis following effective anti-TNF therapy. In order to validate the CE-MRI and micro-CT 3-D outcome measures as in vivo biomarkers of erosive inflammatory arthritis in an intervention study, we used a model of known etiology (TNF-Tg) (23,25), and a proven treatment (anti-TNF) whose efficacy in this model has been demonstrated by several groups of investigators (3,23,26). Since we observed that the onset of knee synovitis in TNF-Tg mice varies from 3 months to 5 months of age (Figure 3I), we aimed to determine whether entering the mice into the study based on synovial volume rather than age would improve the statistical outcome of a small number of animals. Initiation of therapy based on age. In a traditional intervention study based on age, we randomized 5–6month-old TNF-Tg and WT mice to receive either anti-TNF or placebo treatment (Figure 4). The baseline CE-MRI data for the TNF-Tg mice demonstrated remarkable variability in synovial volume (5.66 ⫾ 2.46 mm3) and lymph node volume (9.14 ⫾ 3.12 mm3). Due to this variability, we did not observe a significant decrease in synovial volume with anti-TNF treatment (slope ⫽ –0.17 mm3/week; P ⫽ 0.06), despite a 44% reduction from baseline values at 16 weeks (Figure 4A). However, in comparison with placebo treatment (slope ⫽ 0.20 mm3/week; P ⫽ 0.02), the difference in slope was highly significant (P ⬍ 0.01). In contrast, we observed a significant difference in lymph node volume at the first time point following treatment (3.71 ⫾ 2.11 mm3 in the anti-TNF group versus 8.91 ⫾ 1.50 mm3 in the placebo group; P ⬍ 0.01). Overall, anti-TNF therapy demonstrated a significant 67% reduction from baseline (slope ⫽ –0.33 mm3/week; P ⬍ 0.0001) by the end of the study (Figure 4B). Consistent with the lymph node volume plateau observed in the natural history study after 4.5 months (Figure 3H), no changes were observed in placebotreated TNF-Tg mice over the course of this study (slope ⫽ 0.08 mm3/week; P ⫽ 0.30), but the slopes for placebo versus anti-TNF treatment were highly significantly different (P ⬍ 0.001). No changes or drug effects were observed in the WT groups. One surprising outcome of this study was the synovial volume peak at 12 weeks (122% increase from baseline), which decreased to a 52% change from baseline at 16 weeks (Figure 4A). To better understand this, we investigated the areas of decreased Gd-DTPA-BMA PROULX ET AL Figure 4. Effects of anti–tumor necrosis factor (anti-TNF) therapy in 5–6-month-old TNF-transgenic (TNF-Tg) and wild-type (WT) mice. Mice underwent contrast-enhanced magnetic resonance imaging (CEMRI) at baseline and were then randomized to receive anti-TNF or placebo treatment. TNF-Tg mice underwent CE-MRI every 4 weeks thereafter; WT mice underwent CE-MRI at 16 weeks after treatment. A and B, Synovial and lymph node volumes, respectively. Data from each scan of the TNF-Tg placebo-treated (pink line) versus anti-TNF– treated (blue line) groups and of the WT placebo-treated (dark green line) versus anti-TNF–treated (red line) groups were calculated and plotted. Note the decrease in synovial volume between 12 and 16 weeks in the placebo-treated TNF-Tg group. Anti-TNF treatment had no effects in WT mice, and no changes in synovial or lymph node volumes were detected in these animals after 16 weeks. Values are the mean ⫾ SD of 4 mice per group. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01, versus placebo-treated TNF-Tg mice at the same time point, by 2-sided t-test. Linear mixed-effects regression analysis revealed a highly significant difference in the slopes for both the synovial (P ⬍ 0.01) and lymph node (P ⬍ 0.001) volumes in anti-TNF–treated versus placebo-treated TNF-Tg mice. C–F, CE-MR image (C) and tissue sections (D–F) from the knee of a representative placebo-treated TNF-Tg mouse. The linear progression of inflammatory arthritis was limited by tissue fibrosis, as shown by nonenhancing synovial regions on CE-MRI (arrows) and on the corresponding histology section (#) shown at 40⫻ magnification (D); boxed areas at the bottom and top of C are shown at 200⫻ magnification (E and F, respectively), illustrating pannus fibrosis during the end-stage of arthritis in this animal. enhancement (Figure 4C), with corresponding histologic assessment of the knees from these mice, and we found large regions of pannus fibrosis (Figures 4D–F). This MRI AND MICRO-CT ASSESSMENTS IN MICE WITH INFLAMMATORY ARTHRITIS 4031 Figure 5. Effects of anti–tumor necrosis factor (anti-TNF) therapy in TNF-transgenic (TNF-Tg) mice with established synovitis. Mice underwent contrast-enhanced magnetic resonance imaging (CE-MRI) bimonthly from 3 months of age until a synovial volume ⬎3 mm3 was achieved. They then underwent in vivo microfocal computed tomography (micro-CT) scanning and were randomized to receive anti-TNF or placebo treatment. Mice underwent CE-MRI every 2 weeks for 8 weeks, when a followup micro-CT scan was performed. A and B, Synovial and lymph node volumes, respectively. Data from each scan were calculated and plotted. Values are the mean ⫾ SD of 4 mice per group. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01, versus placebo-treated mice at the same time point, by 2-sided t-test. Linear mixed-effects regression analysis revealed a highly significant difference in the slopes for both the synovial (P ⬍ 0.001) and lymph node (P ⬍ 0.0001) volumes in anti-TNF–treated versus placebo-treated mice. C and D, Three-dimensional (3-D) reconstructions and calculated synovial (yellow) and popliteal lymph node (red) volumes at baseline (C) and after 8 weeks of anti-TNF therapy (D). The protective effects of anti-TNF therapy are also apparent from these 3-D reconstructions of images from a representative mouse. E and F, Treatment effect evaluated by a measure of the difference in the slopes between treatment groups. When treatment was initiated based on synovial volume rather than age as the entry criterion, there was a significantly larger treatment effect on the synovial volume (E) (0.92 mm3/week versus 0.37 mm3/week; P ⫽ 0.04) and the lymph node (LN) volume (F) (1.26 mm3/week versus 0.41 mm3/week; P ⫽ 0.04). Values are the mean ⫾ SD of 4 mice per group. fibrosis is consistent with a loss of CE-MRI signal enhancement as a result of the loss of vascularity. Initiation of therapy based on synovial volume. To determine if a more significant therapeutic effect of anti-TNF therapy on synovial and lymph node volumes could be observed in TNF-Tg mice with established arthritis, we repeated the placebo-controlled study by randomizing the animals to the treatments at the point 4032 PROULX ET AL Figure 6. Arrested bone erosion after anti–tumor necrosis factor (anti-TNF) therapy and correlation with sustained synovial inflammation in TNF-transgenic (TNF-Tg) mice. A, Change in cortical bone volume of the patella and talus, as determined from the baseline and 8-week microfocal computed tomography (micro-CT) scans in the anti-TNF and placebo treatment groups. While there was no significant difference in the change in patellar bone volume between groups (P ⫽ 0.14), anti-TNF treatment had a significant effect on bone loss in the talus (ⴱ ⫽ P ⬍ 0.02). Values are the mean ⫾ SD of 4 mice per group. B and C, Area under the curve (AUC) measurement as a function of joint inflammation and time (B) and regression analysis of change in patellar volume versus synovial AUC (C). The contrast-enhanced magnetic resonance imaging data for the synovial volume in a representative placebo-treated TNF-Tg mouse (Figure 5) were plotted to show the AUC measurement as a function of joint inflammation and time. This measurement was then used to perform regression analyses of the change in patellar volume versus the synovial volume AUC in all 8 mice in the study described in Figure 5. D–I, Micro-CT reconstructed images of the patella at baseline (D) and at 8 weeks (E), with accompanying histology section (F) (at 10⫻ magnification) and of the talus at baseline (G) and at 8 weeks (H), with accompanying histology section (I) (at 4⫻ magnification). Boxed area in I shows the area of severe synovitis in the talus, which had reduced the volume of bone to 20% of its original size. Erosion data from the representative placebo-treated TNF-Tg mouse shown in B were used to reconstruct the images shown in D, E, G, and H. when CE-MRI first showed a knee synovium volume of ⬎3 mm3. This value represents a 50% increase over that in WT mice at 3–5 months of age, confirming the presence of synovitis (Figures 3I and 4A). In this study, anti-TNF therapy had dramatic effects on synovial volume (Figure 5A) and lymph node volume (Figure 5B) over time as compared with placebo. Synovial volumes did not demonstrate a significant decrease during the course of the 8-week study with anti-TNF treatment (slope ⫽ –0.20 mm3/week; P ⫽ 0.21). However, this was due to the rapid response in these animals to anti-TNF therapy, reaching levels in the WT mice within 4 weeks, as visualized in volumetric reconstructions of the baseline (Figure 5C) and 4-week (Figure 5D) data from a representative animal. A significant decrease (–49%) from baseline to 4 weeks (slope ⫽ –0.50 mm3/week; P ⫽ 0.01) was found, and there was no further change from 4 weeks to 8 weeks (3%) (slope ⫽ 0.03; P ⫽ 0.91). In contrast, synovial volumes in the placebo-treated mice showed a highly significant increase (169%) throughout the 8 weeks of study (slope ⫽ 0.72 mm3/week; P ⬍ 0.0001). This resulted in highly significantly different slopes for placebo versus anti-TNF treatment (P ⬍ 0.001). Another advantage of this study design is that no pannus fibrosis effects were seen in any of the mice. MRI AND MICRO-CT ASSESSMENTS IN MICE WITH INFLAMMATORY ARTHRITIS Lymph node volume significantly decreased (–73%) with anti-TNF therapy (slope ⫽ –0.84 mm3/ week; P ⬍ 0.0001) (Figure 5B). Lymph node volumes also showed a highly significant increase (60%) with placebo therapy (slope ⫽ 0.43 mm3/week; P ⬍ 0.01). This resulted in highly significantly different slopes for placebo versus anti-TNF treatment (P ⬍ 0.0001). Interestingly, the lymph node volume had wide variability in placebo-treated mice throughout the study (Figure 5B), suggesting that synovial inflammation and lymph node volume are not directly linked. More significant treatment effect with therapy initiated based on constant synovial volume rather than with therapy initiated based on age. The difference in slopes of volume versus time between anti-TNF and placebo treatment groups is a measure of the treatment effect of anti-TNF therapy. Initiation of treatment based on synovial volume resulted in a significantly larger treatment effect on synovial volumes than initiation based on age alone (difference in slopes of 0.92 mm3/week versus 0.37 mm3/week; P ⫽ 0.04) (Figure 5E). The treatment effect on lymph node volume was also significantly larger when the entry criterion was based on synovial volume as opposed to age (difference in slopes of 1.26 mm3/week versus 0.41 mm3/week; P ⬍ 0.01) (Figure 5F). Effect of anti-TNF therapy on bone erosion. By incorporating longitudinal micro-CT analysis into this study, in which focal erosions were assessed by subtracting the baseline bone volumes from the 8-week bone volumes, we were able to evaluate changes in bone volume of the patella and talus in TNF-Tg mice treated with anti-TNF or placebo (Figure 6A). While half of the placebo-treated mice had a decreased patella volume due to erosions and all of the anti-TNF–treated mice showed an increase or no change in patellar volume, these changes were small and failed to demonstrate a significant drug effect (P ⫽ 0.14). However, all of the placebo-treated mice had a markedly decreased talus volume, and 75% of the anti-TNF–treated mice showed an increase in the size of this bone, which resulted in a significant difference between these groups (P ⬍ 0.02). Correlation of longitudinal 3-D biomarkers of inflammation and bone erosion. As a final validation of our 3-D biomarkers, we assessed the relationship between synovial inflammation and bone erosion using the data from the study of the initiation of treatment based on synovial volume (n ⫽ 8 mice). Since the erosion data are a function of the change in bone volume over time, we first had to convert the synovial volume data to reflect the severity of synovitis over the study period. This was done by computing the area under the curve 4033 from the plot of the synovial volume versus time for each animal (Figure 6B). Using this approach, we found a highly significant correlation between our measures of synovial volume and patellar erosion (R2 ⫽ 0.7469, P ⬍ 0.01) (Figure 6C). The impact of unchecked synovitis on bone was demonstrated by a longitudinal analysis of the patella (Figures 6D and E). The dramatic loss of bone as quantified by micro-CT was associated with widespread synovial infiltration into the bone as visualized in the corresponding histology section (Figure 6F). An even more dramatic loss of bone was present at the talus (Figures 6G and H), in which severe synovitis, as visualized histologically (Figure 6I), had reduced the volume of bone to 20% of its original size. DISCUSSION In vivo imaging measurements have emerged as the outcome of choice for translational research in preclinical studies, based on their potential to objectively quantify change and their compatibility with modalities used in clinical trials (27). Since MRI and micro-CT are widely accepted as the “gold standards” for the assessment of soft tissue and bone volumes, respectively, we aimed to adapt these methods as longitudinal outcome measures of erosive inflammatory arthritis in mice. In the case of CE-MRI for the mouse knee, several innovations were required (Figure 1). These included first, the generation of a mouse knee–specific coil that can interface with a clinical 3T MRI; second, the establishment of pulse sequences that produce high-resolution images (105 m) with minimal slice thickness (160 m) to reduce partial volume effects; third, a method by which to normalize for the variability of Gd-DTPABMA administration using muscle contrast enhancement; and fourth, standardization for thresholding and segmentation of biomarkers for longitudinal quantitative 3-D analyses. Although commercial small-animal MRI instruments may become more popular in the future, we chose to use a clinical MRI to ensure that all of the biomarkers we identified could be used to study arthritis in humans with readily available pulse sequences. In addition, the quantification methods developed could easily be adapted to data collected with small-animal scanners. Several quantification methods have been developed to assess synovial volumes in humans by CE-MRI (28,29). However, no quantification methods have been established for clinical trials, for which there is a great demand. The currently accepted evaluation determined by the Outcome Measures in Rheumatology Clinical 4034 Trials Group is the RA MRI Scoring system RAMRIS, which consists of a semiquantitative global scoring system (scores of 0–3) based on synovial tissue thickness and Gd-DTPA enhancement (30). The current gold standard for quantification of synovitis is a manual segmentation technique (14,31). Although this technique has been validated in intervention studies and has demonstrated correlation with an aspirated volume of synovial fluid, the time required for analysis (0.75–2 hours) and the technical expertise necessary have limited its use in clinical trails. To address these limitations, automated segmentation methods have been evaluated (32,33). This 2-step segmentation process consists of a limited manual segmentation to remove enhancing vessels and skin (similar to the limit lines used in the current study) and application of a threshold on subtracted images. Methods by which to threshold enhancing synovial tissue remain a subject of controversy. Østergaard et al (32) evaluated several different thresholds based on the percentage of enhancement of synovium compared with the manual segmentation technique and determined that a 45% enhancement threshold was optimal. Although this reduced the time required for analysis to 20 minutes, there was increased inter-MRI variation as compared with manual segmentation techniques, particularly when misalignment occurred between precontrast and postcontrast scans. Palmer et al (33) used a threshold value based on the signal intensity differences between several regions of nonenhancing tissue (suppressed fat) and enhancing pannus. This method has an advantage in that the threshold is determined after each individual scan; however, the time required for analysis is ⬃45 minutes. The threshold is also sensitive to misregistration artifacts and the consistency of fat suppression. To our knowledge, the current study is the first to use adjacent muscle as a normalization tissue with which to determine the synovial threshold. This approach is attractive because muscle is an enhancing tissue that is not implicated in the pathologic changes of inflammatory arthritis and because its tissue properties do not change significantly between scans; variations in values in muscle will reflect the signal variations that are present between scanning sessions. In small-animal studies, this approach is especially warranted due to the inconsistencies in dosage delivery that are inherent during administration via intravenous injection. During our dosage study, we found that muscle was linearly enhanced with increasing dosages of Gd-DTPA-BMA. Even more importantly, there was a direct relationship PROULX ET AL between muscle contrast enhancement and the synovial threshold that was used to maintain a constant synovial volume between doses. The use of this optimized threshold, combined with the ability to minimize motion artifacts with anesthesia and standardized positioning, has allowed reproducible measurements of synovial volume in mice (4.5% coefficient of variation) that far exceed those reported in humans (22% median relative variation as measured by the manual segmentation method ). Whether a similar threshold approach based on muscle enhancement could be adapted to clinical studies warrants further investigation. Considering that micro-CT has been used to examine focal erosions in murine arthritis models for several years (35), it is somewhat surprising that this approach has yet to evolve into a longitudinal quantitative outcome measure. Based on our experience, we found that this is likely due to the difficulty in registering the baseline and outcome 3-D micro-CT images so that the erosion as a negative change in volume could be accurately assessed. It is also clear from our work (Figure 2), that segmentation can only be readily performed on small bones that are clearly defined by soft-tissue boundaries (e.g., the patella and talus). This is because minor imperfections in the registration of large bones and subjective segmentation of bone parts (e.g., the distal femur and proximal tibia) can result in significant measurement errors. In this study, we focused our attention on the synovial and lymph node volume as the primary outcome measures of inflammation, based on their facile segmentation and quantification from CE-MR images. Interestingly, our findings indicated that these tissues behave differently during the onset of inflammatory arthritis and its amelioration following effective therapy. We found that the popliteal lymph node volume was the most sensitive biomarker of lower limb arthritis. This conclusion comes from the observation that lymph node volumes significantly increased when TNF-Tg mice were 2.5 months-old (Figure 3H), which correlated with the point at which increased TNF serum levels and changes in peripheral blood mononuclear cell populations are first detected in this model (36). Our finding that increases in lymph node volume precede the occurrence of knee synovitis is consistent with the function of popliteal lymph nodes in draining both the knee and ankle joint tissues and the fact that arthritis first occurs in the ankle joint in this model (23). This conclusion is corroborated by similar results in the K/BxN serum-transfer model of arthritis (37), in which the popliteal lymph node volume increased concurrently MRI AND MICRO-CT ASSESSMENTS IN MICE WITH INFLAMMATORY ARTHRITIS with ankle inflammation in the absence of knee synovitis. (Figures illustrating the assessment of K/BxN seruminduced synovitis and lymph node inflammation by longitudinal CE-MRI are available upon request from the corresponding author.) Thus, the introduction of proinflammatory mediators into the joint alone is not sufficient for the initiation of pannus formation. Our future efforts will focus on understanding the mechanism that initiates this destructive process. This issue is paramount, since synovitis correlates with joint destruction in inflammatory arthritis (Figure 6C). In addition, many group of investigators have generated transgenic animals that can be used to dissect these pathologies (35,38–41), and these transgenic animals can now be assessed with our novel longitudinal outcome measures. By using a validated drug therapy in our model, we were able to address the major concern with all preclinical intervention studies of arthritis that are performed with a single cross-sectional end point, namely, whether or not the differences observed are due to drug effects or to variability between the initiation and/or progression of disease in the individual mice. Using our previous study of anti-TNF treatment of TNF-Tg mice as an example (3), we were unable to draw definitive conclusions about the healing effects on bone and cartilage lesions since there were no baseline assessments. When we repeated this experiment using age as the enrollment criterion (Figure 4), these concerns regarding interanimal variability were realized based on the broad range of baseline synovitis levels (3.00–9.85 mm3). Furthermore, during this study, we saw a dramatic decrease in synovial volume in the placebo group from age 8 months to age 9 months that was due to pannus tissue fibrosis. Thus, by modifying the study design to include an entry criterion based on synovial volume (Figure 5), we were able to make several remarkable observations. First, we found that the treatment effect of anti-TNF therapy on synovial volume and lymph node volume is significantly greater than when the entry criterion is based on age. Second, we noted that synovial volumes were significantly reduced to levels in the WT mice after only 4 weeks of anti-TNF therapy. Third, we found that initiation of the study earlier in the disease process combined with an 8-week time course avoided the pannus tissue fibrosis effects previously observed. Fourth, a healing response was noted in which the mean lymph node volumes in the TNF-Tg mice treated with anti-TNF therapy were reduced but were still 2.5 times higher than those in the WT mice at the end of the study. 4035 Fifth, we found that bone erosion was arrested, as demonstrated by a significant difference in the change in talus volume between anti-TNF–treated and placebotreated animals. In summary, we found that CE-MRI and micro-CT are very useful longitudinal outcome measures of the severity of inflammatory arthritis since they demonstrated significant results with as few as 4 mice per group. We are also pursuing additional outcomes with the CE-MRI, including quantifications of bone marrow edema, which appears as a bright signal with this MR pulse sequence in the bones of arthritic animals as compared with a dark signal in the bones of WT controls. (Figures illustrating bone marrow edema as determined by CE-MRI and its correspondence to osteitis as determined histologically are available upon request from the corresponding author.) This biomarker has previously been demonstrated to be a faithful predictor of focal erosions in RA (15), and its nature has recently been demonstrated to be osteitis (42). Moreover, bone marrow edema measured by CE-MRI has been effectively used to assess disease severity and response to therapy in patients with ankylosing spondylitis (43,44). Thus, the hope is that an edema outcome measure could be developed as the long sought-after surrogate for arthritic pain in animal models and/or a predictor of joint destruction in humans. ACKNOWLEDGMENTS The authors would like to thank Patricia Weber for technical assistance with the MRI, Laura Yanoso for technical assistance with the micro-CT, Colleen Hock for assistance with animal breeding and genotyping, and Krista Scorsone for technical assistance with the histology. AUTHOR CONTRIBUTIONS Mr. Proulx and Dr. Schwarz had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study design. Proulx, Kwok, You, Shealy, Ritchlin, Awad, Boyce, Xing, Schwarz. Acquisition of data. Proulx, Kwok, You, Awad, Boyce. Analysis and interpretation of data. Proulx, Kwok, Papuga, Beck, Shealy, Ritchlin, Awad, Boyce, Xing, Schwarz. Manuscript preparation. Proulx, Kwok, You, Papuga, Beck, Shealy, Ritchlin, Awad, Boyce, Xing, Schwarz. Statistical analysis. Proulx, Beck, Schwarz. ROLE OF THE STUDY SPONSOR Centocor facilitated the study design, writing of the manuscript, and the decision to submit the manuscript for publication. Centocor had no role in the data collection or interpretation of the data. 4036 PROULX ET AL REFERENCES 1. Dustin ML. In vivo imaging approaches in animal models of rheumatoid arthritis. Arthritis Res Ther 2003;5:165–71. 2. Van den Berg WB. Animal models of arthritis: what have we learned? J Rheumatol Suppl 2005;72:7–9. 3. Shealy DJ, Wooley PH, Emmell E, Volk A, Rosenberg A, Treacy G, et al. Anti-TNF-␣ antibody allows healing of joint damage in polyarthritic transgenic mice. Arthritis Res 2002;4:R7. 4. Li P, Schwarz EM, O’Keefe RJ, Ma L, Boyce BF, Xing L. RANK signaling is not required for TNF␣-mediated increase in CD11hi osteoclast precursors but is essential for mature osteoclast formation in TNF␣-mediated inflammatory arthritis. J Bone Miner Res 2004;19:207–13. 5. Conaghan PG, McQueen FM, Peterfy CG, Lassere MN, Ejbjerg B, Bird P, et al. The evidence for magnetic resonance imaging as an outcome measure in proof-of-concept rheumatoid arthritis studies. J Rheumatol 2005;32:2465–9. 6. Haavardsholm EA, Ostergaard M, Ejbjerg BJ, Kvan NP, Uhlig TA, Lilleas FG, et al. Reliability and sensitivity to change of the OMERACT rheumatoid arthritis magnetic resonance imaging score in a multireader, longitudinal setting. Arthritis Rheum 2005;52:3860–7. 7. Martinez-Martinez MU, Cuevas-Orta E, Reyes-Vaca G, Baranda L, Gonzalez-Amaro R, Abud-Mendoza C. Magnetic resonance imaging in patients with rheumatoid arthritis with complete remission treated with disease-modifying antirheumatic drugs or antitumour necrosis factor ␣ agents. Ann Rheum Dis 2007;66:134–5. 8. McQueen FM. Magnetic resonance imaging in early inflammatory arthritis: what is its role? Rheumatology (Oxford) 2000;39:700–6. 9. Kirkham BW, Lassere MN, Edmonds JP, Juhasz KM, Bird PA, Lee CS, et al. Synovial membrane cytokine expression is predictive of joint damage progression in rheumatoid arthritis: a two-year prospective study (the DAMAGE Study cohort). Arthritis Rheum 2006;54:1122–31. 10. McQueen FM, Ostendorf B. What is MRI bone oedema in rheumatoid arthritis and why does it matter? Arthritis Res Ther 2006;8:222. 11. Dohn UM, Ejbjerg BJ, Court-Payen M, Hasselquist M, Narvestad E, Szkudlarek M, et al. Are bone erosions detected by magnetic resonance imaging and ultrasonography true erosions? A comparison with computed tomography in rheumatoid arthritis metacarpophalangeal joints. Arthritis Res Ther 2006;8:R110. 12. Benton N, Stewart N, Crabbe J, Robinson E, Yeoman S, McQueen FM. MRI of the wrist in early rheumatoid arthritis can be used to predict functional outcome at 6 years. Ann Rheum Dis 2004;63:555–61. 13. McGonagle D, Conaghan PG, O’Connor P, Gibbon W, Green M, Wakefield R, et al. The relationship between synovitis and bone changes in early untreated rheumatoid arthritis: a controlled magnetic resonance imaging study. Arthritis Rheum 1999;42: 1706–11. 14. Ostergaard M, Hansen M, Stoltenberg M, Gideon P, Klarlund M, Jensen KE, et al. Magnetic resonance imaging–determined synovial membrane volume as a marker of disease activity and a predictor of progressive joint destruction in the wrists of patients with rheumatoid arthritis. Arthritis Rheum 1999;42:918–29. 15. McQueen FM, Benton N, Perry D, Crabbe J, Robinson E, Yeoman S, et al. Bone edema scored on magnetic resonance imaging scans of the dominant carpus at presentation predicts radiographic joint damage of the hands and feet six years later in patients with rheumatoid arthritis. Arthritis Rheum 2003;48: 1814–27. 16. Appel H, Loddenkemper C, Grozdanovic Z, Ebhardt H, Dreimann M, Hempfing A, et al. Correlation of histopathological findings and magnetic resonance imaging in the spine of patients with ankylosing spondylitis. Arthritis Res Ther 2006;8:R143. 17. Gaffney K, Cookson J, Blades S, Coumbe A, Blake D. Quantitative assessment of the rheumatoid synovial microvascular bed by gadolinium-DTPA enhanced magnetic resonance imaging. Ann Rheum Dis 1998;57:152–7. 18. Clunie G, Hall-Craggs MA, Paley MN, King A, Wilkinson ID, Ell PJ, et al. Measurement of synovial lining volume by magnetic resonance imaging of the knee in chronic synovitis. Ann Rheum Dis 1997;56:526–34. 19. Munasinghe JP, Tyler JA, Hodgson RJ, Barry MA, Gresham GA, Evans R, et al. Magnetic resonance imaging, histology, and x-ray of three stages of damage to the knees of STR/ORT mice. Invest Radiol 1996;31:630–8. 20. Kapadia RD, Stroup GB, Badger AM, Koller B, Levin JM, Coatney RW, et al. Applications of micro-CT and MR microscopy to study pre-clinical models of osteoporosis and osteoarthritis. Technol Health Care 1998;6:361–72. 21. Dardzinski BJ, Schmithorst VJ, Holland SK, Boivin GP, Imagawa T, Watanabe S, et al. MR imaging of murine arthritis using ultrasmall superparamagnetic iron oxide particles. Magn Reson Imaging 2001;19:1209–16. 22. Saidenberg-Kermanac HN, Leclercq F, Herscovici J, Scherman D, Boissier MC. High-resolution magnetic resonance imaging of knee inflammatory arthritis in mice. Clin Exp Rheumatol 2006;24: 216–7. 23. Keffer J, Probert L, Cazlaris H, Georgopoulos S, Kaslaris E, Kioussis D, et al. Transgenic mice expressing human tumour necrosis factor: a predictive genetic model of arthritis. EMBO J 1991;10:4025–31. 24. Feldmann M, Maini RN. TNF defined as a therapeutic target for rheumatoid arthritis and other autoimmune diseases [published erratum appears in Nat Med 2003;9:1433]. Nat Med 2003;9: 1245–50. 25. Li P, Schwarz EM. The TNF-␣ transgenic mouse model of inflammatory arthritis. Springer Semin Immunopathol 2003;25: 19–33. 26. Redlich K, Hayer S, Maier A, Dunstan CR, Tohidast-Akrad M, Lang S, et al. Tumor necrosis factor ␣–mediated joint destruction is inhibited by targeting osteoclasts with osteoprotegerin. Arthritis Rheum 2002;46:785–92. 27. Massoud TF, Gambhir SS. Molecular imaging in living subjects: seeing fundamental biological processes in a new light. Genes Dev 2003;17:545–80. 28. Lassere MN, Bird P. Measurements of rheumatoid arthritis disease activity and damage using magnetic resonance imaging. Truth and discrimination: does MRI make the grade? J Rheumatol 2001;28:1151–7. 29. Farrant JM, O’Connor PJ, Grainger AJ. Advanced imaging in rheumatoid arthritis. Part 1. Synovitis. Skeletal Radiol 2007;36: 269–79. 30. Ostergaard M, Peterfy C, Conaghan P, McQueen F, Bird P, Ejbjerg B, et al. OMERACT rheumatoid arthritis magnetic resonance imaging studies: core set of MRI acquisitions, joint pathology definitions, and the OMERACT RA-MRI scoring system [published erratum appears in J Rheumatol 2004;31:198]. J Rheumatol 2003;30:1385–6. 31. Ostergaard M, Stoltenberg M, Lovgreen-Nielsen P, Volck B, Jensen CH, Lorenzen I. Magnetic resonance imaging–determined synovial membrane and joint effusion volumes in rheumatoid arthritis and osteoarthritis: comparison with the macroscopic and microscopic appearance of the synovium. Arthritis Rheum 1997; 40:1856–67. 32. Ostergaard M. Different approaches to synovial membrane volume determination by magnetic resonance imaging: manual versus automated segmentation. Br J Rheumatol 1997;36:1166–77. 33. Palmer WE, Rosenthal DI, Schoenberg OI, Fischman AJ, Simon LS, Rubin RH, et al. Quantification of inflammation in the wrist MRI AND MICRO-CT ASSESSMENTS IN MICE WITH INFLAMMATORY ARTHRITIS 34. 35. 36. 37. 38. 39. with gadolinium-enhanced MR imaging and PET with 2-[F-18]fluoro-2-deoxy-D-glucose. Radiology 1995;196:647–55. Klarlund M, Ostergaard M, Lorenzen I. Finger joint synovitis in rheumatoid arthritis: quantitative assessment by magnetic resonance imaging. Rheumatology (Oxford) 1999;38:66–72. Herrak P, Gortz B, Hayer S, Redlich K, Reiter E, Gasser J, et al. Zoledronic acid protects against local and systemic bone loss in tumor necrosis factor–mediated arthritis. Arthritis Rheum 2004; 50:2327–37. Li P, Schwarz EM, O’Keefe RJ, Ma L, Looney RJ, Ritchlin CT, et al. Systemic tumor necrosis factor ␣ mediates an increase in peripheral CD11bhigh osteoclast precursors in tumor necrosis factor ␣–transgenic mice. Arthritis Rheum 2004;50:265–76. Mangialaio S, Ji H, Korganow AS, Kouskoff V, Benoist C, Mathis D. The arthritogenic T cell receptor and its ligand in a model of spontaneous arthritis. Arthritis Rheum 1999;42:2517–23. Kong YY, Feige U, Sarosi I, Bolon B, Tafuri A, Morony S, et al. Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 1999; 402:304–9. Pettit AR, Ji H, von Stechow D, Muller R, Goldring SR, Choi Y, et al. TRANCE/RANKL knockout mice are protected from bone erosion in a serum transfer model of arthritis. Am J Pathol 2001;159:1689–99. 4037 40. Redlich K, Hayer S, Ricci R, David JP, Tohidast-Akrad M, Kollias G, et al. Osteoclasts are essential for TNF-␣-mediated joint destruction. J Clin Invest 2002;110:1419–27. 41. Van Lent PL, Grevers L, Lubberts E, de Vries TJ, Nabbe KC, Verbeek S, et al. Fc␥ receptors directly mediate cartilage, but not bone, destruction in murine antigen-induced arthritis: uncoupling of cartilage damage from bone erosion and joint inflammation. Arthritis Rheum 2006;54:3868–77. 42. Jimenez-Boj E, Nobauer-Huhmann I, Hanslik-Schnabel B, Dorotka R, Wanivenhaus AH, Kainberger F, et al. Bone erosions and bone marrow edema as defined by magnetic resonance imaging reflect true bone marrow inflammation in rheumatoid arthritis. Arthritis Rheum 2007;56:1118–24. 43. Braun J, Landewe R, Hermann KG, Han J, Yan S, Williamson P, et al, for the ASSERT Study Group. Major reduction in spinal inflammation in patients with ankylosing spondylitis after treatment with infliximab: results of a multicenter, randomized, doubleblind, placebo-controlled magnetic resonance imaging study. Arthritis Rheum 2006;54:1646–52. 44. Van der Heijde D, Han C, DeVlam K, Burmester G, van den Bosch F, Williamson P, et al. Infliximab improves productivity and reduces workday loss in patients with ankylosing spondylitis: results from a randomized, placebo-controlled trial. Arthritis Rheum 2006;55:569–74.