In vivo molecular imaging of experimental joint inflammation by combined 18F-FDG positron emission tomography and computed tomography

Introduction The purpose of this work was to establish and validate combined small animal positron emission tomography - computed tomography (PET/CT) as a new in vivo imaging method for visualisation and quantification of joint inflammation. Methods Signalling of radioisotope 18F labelled Fluorodeoxyglucose (18F-FDG) injected in mice with glucose-6-phosphate isomerase (G6PI)-induced arthritis was analysed by PET/CT. Accumulation of 18F-FDG in tissue was quantified by PET measurement, whereas high definition CT delivered anatomical information. The fusion of both images revealed in detail spatial and temporal distribution and metabolism of 18F-FDG. Results A distinct 18F-FDG signal could be measured by PET in carpal and tarsal joints, from mice with early or established arthritis. In contrast, no accumulation of 18F-FDG was detectable before arthritis onset. Comparison of 18F-FDG joint uptake with histopathological evaluation revealed a significant correlation of both methods. Conclusions Small animal PET/CT using 18F-FDG is a feasible method for monitoring and, more importantly, quantitative assessment of inflammation in G6PI-arthritis. Since it is possible to perform repeated non-invasive measurements in vivo, not only numbers of animals in preclinical studies can markedly be reduced by this method, but also longitudinal studies come into reach, e. g. for individual flare-up reactions or monitoring therapy response in progressive arthritis.


Introduction
The current gold standard for the assessment of inflammation in small-rodent models of arthritis is histopathological evaluation of joint sections. Apart from semiquantitative grading, limitations of this method include the impossibility of performing in vivo or longitudinal studies and the need for relatively large numbers of animals. Non-invasive imaging techniques allowing an objective quantitative assessment of inflammation would, therefore, be a most welcome tool for arthritis research.
Positron emission tomography (PET), predominantly using the radiopharmaceutical 18 F-labelled fluorodeoxyglucose ( 18 F-FDG) as metabolic tracer, is a functional in vivo imaging technique that is clinically used mainly for tumor diagnosis, therapy monitoring, and experimental cancer research. 18 F-FDG is a glucose analog in which the 2'-OH has been replaced by 18 F. Consequently, 18 F-FDG cannot be further metabolized after phosphorylation and is trapped and enriched within the cell. This offers the opportunity of a quantifiable 18 F-FDG PET signal from sites of pathological increased glucose metabolism in the tissue (that is, sites of inflammation). In addition, glucose metabolism is affected by proinflammatory tumor necrosis factor-alpha (TNF-α) and characteristically increased in inflamed tissue [1,2], making PET a potentially interesting technique for the detection and quantification of inflammation. Analyses of radioisotope-labelled FDG uptake in vitro revealed a distinct accumulation in fibroblasts and neutrophils, whereas resting macrophages incorporated only small amounts [3]. Stimulation of cells with TNF-α revealed a shift to strong glucose uptake in macrophages and even more in fibroblasts, whereas accumulation in neutrophils remained stable at the level of stimulated macrophages. The accumulation of 18 F-FDG in cells contributing to synovial inflammation [4][5][6] could provide a sensitive and non-invasive tool for visualization and quantification of joint inflammation in vivo.
In rheumatoid arthritis (RA), only a few studies thus far have examined the use of PET, proposing quantitative 18 F-FDG PET as a promising technique to measure disease activity and to monitor the efficacy of anti-inflammatory drugs without restriction to morphology-based information [7,8]. PET is also increasingly used in preclinical research [3,9]. Since computed tomography (CT) has a higher resolution than PET, one approach of modern imaging systems is the combination of PET and CT. Thus, in humans, the detailed morphological information provided by CT can be used to localize the anatomical source of PET signals exactly [10].
The goal of our work was to examine, in well-described glucose-6-phosphate isomerase (G6PI)-induced murine arthritis [11][12][13][14][15][16][17], whether 18 F-FDG micro-PET/CT can be used for the quantitative in vivo assessment of inflammation in acute and chronic stages of experimental arthritis. To validate the method, results were systematically compared with semiquantitative histopathological analyses. Our findings revealed a linear and statistically significant correlation of 18 F-FDG PET/CT quantification and histopathological evaluation of inflammatory experimental arthritis. In the field of arthritis research, small-animal PET/CT imaging thus offers new opportunities for a quantitative assessment of inflammation and even can be performed in vivo.

Materials and methods
Glucose-6-phosphate isomerase-induced arthritis DBA/1 mice were bred at the animal facility of Jena University (Jena, Germany). All animal studies were approved by the local commission for animal protection (registered number 02-045/08). Arthritis was induced as described elsewhere [11]. In brief, DBA/1 mice were immunized subcutaneously with 400 μg of recombinant human G6PI in emulsified complete Freund's adjuvant (Sigma-Aldrich, Taufkirchen, Germany). Macroscopic evaluation of arthritis was performed as previously described [11]. A score of 0 indicates no macroscopically recognizable signs of arthritis, 1 indicates swelling and redness, 2 means strong swelling and redness, and 3 indicates massive swelling and redness. For the total clinical score per animal, results from all paws were summed. In the experiment assessing practicability of small-animal PET/CT in monitoring therapeutical intervention, 150 μg of soluble tumor necrosis factor receptor (sTNFR) fusion protein (etanercept; Wyeth, Münster, Germany) in 100 μL of saline or saline alone as control was daily administered intraperitoneally, beginning 1 day after immunization.
Histopathological assessment of glucose-6-phosphate isomerase-induced arthritis For histopathological examination, mice were sacrificed immediately after PET/CT measurement. Sections of fixed and decalcified joints were evaluated by a pathologist. Arthritis severity of each paw was graded according to a histopathological scoring system, with consideration of acute and chronic inflammatory parameters as described before [18]. In brief, acute inflammation reflects pathological alterations by infiltration of neutrophil granulocytes in the synovial membrane (0 = no, 1 = slight, 2 = moderate, and 3 = severe) and by exudate (0 = no, 1 = slight, 2 = moderate, and 3 = severe), with an additional scoring for the presence of fibrin and the affection of periarticular tissue. The maximal possible value per paw for acute inflammation according to this scoring was 8. Chronic inflammation denotes pathological alterations by synovial hyperplasia, infiltration of mononuclear cells, and fibrosis (0 = no, 1 = slight, 2 = moderate, and 3 = severe), with a maximum of 9. Total inflammation score reflects the sum of acute and chronic inflammation in the joint and periarticular tissue. For individual histopathological arthritis evaluation, scores from all paws were summed.

Positron emission tomography-computed tomography in vivo imaging
A Siemens Inveon small-animal multimodality PET/CT system (Preclinical Solutions; Siemens Healthcare Molecular Imaging, Knoxville, TN, USA) was used for in vivo imaging. This PET/CT system is characterized by the combination of two independently operating PET and CT scanners. Radial, tangential, and axial resolutions at the center of the field of view of the PET module are better than 1.5 mm [19,20]. PET acquisitions were carried out with default settings of coincidence timing window of 3.4 ns and energy window of 350 to 650 keV. Images were reconstructed using Fourier rebinning and the two-dimensional ordered-subset expectation maximization (OSEM 2D) algorithm. Attenuation was corrected on the basis of the CT measurements. The CT module consists of a cone beam micro-x-ray source (50-μm focal spot size) and a 2,048 × 3,072 pixel x-ray detector. Our standard micro-CT imaging protocol used 80 kVp at 500 μA, 360°of rotation, and 200 projections per bed position. Micro-CT images were reconstructed using a Shepp-Logan filter and cone-beam filtered backprojection.
Quantitative analysis of 18 F-FDG accumulation in the period from 40 to 60 minutes after injection was enabled by the image fusion technology of Siemens Inveon Research Workplace software 2.2 (Siemens Healthcare Molecular Imaging). In the kinetic analyses, the 60-minute dataset was divided in 39 time frames; otherwise, data from 40 to 60 minutes after 18 F-FDG injection were analyzed in a single time frame. Volumes of interest (VOIs) used for quantification of fore and hind limb inflammation were 2.5 and 3.1 mm 3 , respectively. Inflammatory tissue uptake of the glucose analog was measured as percentage of injected dose per gram (% ID/g). Three-dimensional (3D) surface rendering (SR) of areas with increased 18 F-FDG uptake was carried out with PMOD software 3.0 (PMOD Technologies Ltd., Zurich, Switzerland) by means of SR techniques using an 8% ID/g threshold. Mice were anesthetized with 1.5% to 2% vaporized isoflurane (DeltaSelect, Dreieich, Germany) in oxygen (1.5 L/minute) to prevent animal movement and reduce imaging artefacts. 18 F (half-life 109 minutes)-labelled FDG (Eckert & Ziegler, Berlin, Germany) with an activity of 9.6 ± 0.6 MBq was injected intravenously in the lateral tail vein.

Statistical analysis
Statistical differences between groups were evaluated using the non-parametric Mann-Whitney U test. Mean scores of fore and hind limbs or individual total scores as the sums of all paws were used for analyses of relationships between histopathological parameters and 18 F-FDG uptake (Spearman test). Statistical significance was accepted for P values of less than 0.05 (*P < 0.05; **P < 0.01; ***P < 0.001). All calculations were performed using the software package SPSS version 16.0 (SPSS Inc., Chicago, IL, USA). In bar charts and text, data are presented as arithmetic mean and standard error of the mean.

Histopathological assessment of arthritis severity
To examine the usability of state-of-the-art smallanimal PET/CT molecular imaging as an objective method to quantify inflammation in experimental arthritis research, severity of inflammation was analyzed at six different time points (days 2, 6, 9, 13, 21, and 35) in the course of G6PI-induced arthritis, using both micro-PET/CT and histology. Histopathological assessment of the degree and components of immune cell infiltration, synovial hyperplasia, and exudate in arthritic joints enabled a semiquantitative scoring of experimental arthritis severity ( Figure 1a). Histology revealed no pathological changes in joints and periarticular joint tissue 2 and 6 days after G6PI immunization (Figure 1b). At day 9, histopathological signs of inflammation coincided with the onset of clinical arthritis. Total individual scores as the sum of acute and chronic parameters revealed a maximum of arthritis severity at days 13 (24.8 ± 0.6) and 21 (23.9 ± 1.4) and a subsequent decrease at day 35 (8.1 ± 1.9).
Scoring of macroscopically visible signs of arthritis (redness and swelling) showed a similar course of experimental arthritis severity (Figure 1c). The first clinical signs appeared at day 9 (1.7 ± 0.6), and the maximal clinical score (8.9 ± 0.5) was obtained at day 13. Clinical signs of arthritis were attenuated at day 21 (4.9 ± 0.4) and day 35 (2.4 ± 0.2).
Small-animal positron emission tomography/computed tomography molecular imaging Scanning of mice with micro-PET/CT generated highresolution 3 D images with information on temporal and spatial biodistribution of 18 F-FDG, reflecting the intensity of tissue glucose metabolism. Precise anatomical localization of VOIs was given by high-resolution CT imaging ( Figure 2a). The combination of PET and CT data enabled detailed visualization of 18 F-FDG signalling in the body (Figure 2b). 18 F-FDG reached heart and lung via the vena cava within the first 0-to 10-second time frame after injection and subsequently was distributed in the gastrointestinal tract and cranially via the carotid arteries. Before arthritis onset (day 2), tracer enrichment was restricted to tissue with high basal metabolic activity (in particular, heart muscle and eyes) and organs of the excretory system (kidneys and bladder; Figure 2c; Additional file 1). In acute experimental arthritis at day 13, additional hot spots of 18 F-FDG signalling could be detected in the joints of fore and hind paws, indicating a specific 18 F-FDG uptake in inflamed joints (Figure 2d; Additional file 2).

Accumulation of 18 F-FDG in joints
Swelling and erythema of paws are characteristic clinical features of acute G6PI-induced arthritis (Figure 3a). 18 F-FDG accumulation measured by micro-PET/CT revealed hot spots of inflammatory metabolic activity in wrist and ankle joints and a less intense inflammation of smaller distal joints. SR of PET/CT data using thresholds (shown is 8% ID/g, according to the level of 18 F-FDG joint accumulation in acute G6PI arthritis) enabled a fusion-based 3 D visualization of morphological structures and regions, where 18 F-FDG was enriched because of inflammatory processes. Kinetic analysis of 18 F-FDG signalling in joints before G6PI immunization (Additional file 3) and at the stage of acute inflammatory arthritis (day 15; Additional file 4) revealed a beginning of 18 F-FDG accumulation approximately 5 minutes after injection (Figure 3b). In arthritic joints, 18 F-FDG uptake was still slightly increasing 60 minutes after tracer injection. In healthy joints, negligible tracer enrichment occurred during the initial 20 minutes after injection, followed by a slight decline. In arthritic animals, percentages of injected radioisotope activities in the joint regions were between 8% and 15% ID/g, whereas tracer uptake in the same joints was around 1% to 3% ID/g before immunization. Therefore, differences in 18 F-FDG uptake of arthritic and healthy joints were pronounced, could be attributed to joint tissue glucose metabolism changes caused by pathogenesis of G6PI-induced arthritis, and could easily be quantified by small-animal PET/CT techniques.
Repeated carpal and tarsal inflammation assessment in the same animals by 18   did not influence the course of experimental arthritis (Figure 3c). Both the clinical scores (data not shown) and the levels of carpal and tarsal 18 F-FDG joint uptake (Figures 3c and 4) were similar in those mice that underwent small-animal PET/CT examination  repeatedly and those that were examined by PET/CT only once and then employed for histopathological evaluation of arthritis.
In vivo quantification of joint inflammation by 18 F-FDG positron emission tomography/computed tomography To analyze the usefulness of micro-PET/CT imaging for monitoring severity of inflammatory arthritis, 18 F-FDG uptake in paws from a total of 42 mice at 6 different time points of experimental disease was quantified. In joint tissue without pathological changes (days 2 and 6), mean accumulation of 18 F-FDG in fore ( Figure 4a) and hind (Figure 4b) limbs was below 2% ID/g, the level of 18 F-FDG accumulation over time in joints of healthy animals before induction of experimental disease (Figure 3b). At the onset of clinical arthritis at day 9, emerging inflammation was associated with an increase of glucose metabolism and induced a significant increase of 18 F-FDG uptake compared with day 2 (carpal: P = 0.0002; tarsal: P = 0.02). Mean values of % ID/g were 3.8 ± 0.57 in carpal and 2.4 ± 0.55 in tarsal joints. 18 F-FDG uptake was further increased at the peak of acute arthritis at day 13 (carpal: 9.2 ± 0.38; tarsal: 12.7 ± 0.40), followed by subsequent decreases at day 21 (carpal: 7.1 ± 0.39; tarsal: 7.6 ± 0.66) and day 35 (carpal: 3.9 ± 0.29; tarsal: 4.4 ± 0.46). Hence, PET data revealed a 4-to 10-fold increase of 18 F-FDG joint uptake in inflammatory stages of arthritis. Interestingly, 18 F-FDG uptake revealed significant (P = 0.001; Figure 4a,b) deviating metabolic activity between days 13 and 21, demonstrating a reduced inflammatory activity at day 21, whereas histopathological assessment showed no difference of arthritis severity in these groups (P = 0.7; Figure 1b). This reduced inflammation activity measured by PET/CT at day 21 was confirmed by macroscopical arthritis assessment (P = 0.001; Figure 1c).

Histopathological evaluation correlates significantly with positron emission tomography quantification
To validate the method and to determine whether 18 F-FDG uptake was correlated with histopathological changes in arthritis, we performed correlation analysis of individual quantitative PET results with semiquantitative clinical and histopathological scoring data. Individual 18 F-FDG PET/CT quantification of inflammation severity correlated significantly with macroscopic clinical scoring (Figure 5a). Furthermore, there were significant correlations of 18 F-FDG uptake and histopathological data assessment regarding scoring of acute, chronic, and total inflammation (Figure 5b-d). Animals that scored in the upper range of the histopathological arthritis index also achieved maximum values in PET analysis. Interestingly, arthritis severity in animals with maximal histopathological scoring was further differentiated by 18 F-FDG PET/CT technique. This means that 25% of the animals with the highest histopathological scores ranged in the top 12% of the total bandwidth of histopathological scoring. In contrast, assessment of inflammation by PET/CT enabled a more refined staging. Here, the 25% of the animals that showed the strongest 18 F-FDG accumulation in the joints covered the top 38% of the total range of 18 F-FDG uptake. For carpal and tarsal joints and for joints total, correlations of 18 F-FDG uptake and histopathological assessment were highly significant (Table 1). These significant correlations of both methods demonstrate that small-animal PET/CT is a reliable imaging method for in vivo quantification of inflammatory arthritis. 18 F-FDG positron emission tomography/computed tomography and histopathological assessment in monitoring therapeutic intervention Experimental modulation of inflammation is an important aim of arthritis research. To examine the usability of small-animal PET/CT to monitor the outcome of a therapeutic intervention, G6PI-immuized mice were injected with human sTNFR (etanercept) or saline. We quantified 18 F-FDG enrichment in the joints of treated and control mice by PET/CT and found a significant 4.9-fold decrease of total 18 F-FDG uptake in sTNFR (etanercept)treated arthritic mice (Figure 5e). Comparable results were obtained using histopathological assessment of therapeutic intervention (Figure 5f). The correlation coefficient of histopathological assessment and 18 F-FDG uptake was 0.862, whereas a P value of less than 0.002 revealed statistical significance of correlation of both methods. Thus, PET/CT is a convenient technique for monitoring treatment modalities in experimental arthritis, even with small numbers of animals.

Discussion
PET data have been shown to correlate with clinical and sonographic parameters in studies on patients with RA [7,8]. The use of in vivo imaging techniques, including PET, in murine models of arthritis has been hindered by the reduced dimensions of the subjects of study. Here, we used a small-animal PET/CT system to assess arthritis severity in mice. This combination of PET with highdefinition CT in a single device allows the exact 3 D anatomical localization of PET signals (that is, 18 F-FDG accumulation in the tissue). Moreover, combined PET and CT imaging also allows the use of CT data to correct for tissue attenuation of PET signals. Our findings demonstrate the feasibility and usefulness of 18 F-FDG joint uptake for objective in vivo quantification of synovial inflammation in the course of G6PI-induced arthritis. Whereas histopathological evaluation inevitably requires sacrificing animals and therefore restricts the complete data range of individual arthritis progression to a single snapshot, 18 F-FDG PET/CT allows repeated investigations of the same subjects. This not only drastically reduces the number of animals required for a study but also offers the appealing possibility of longitudinal analyses within the same animal. Furthermore, in vivo imaging increases statistical quality of data in treatment studies because the status of an animal before and after disease-modulating intervention can be compared directly. Therefore, combined PET/CT using 18 F-FDG as a tracer is a new and powerful tool to quantify experimental joint inflammation non-invasively and accurately in vivo.
Another non-invasive technique, thermal signature analysis, has been used to analyze rat rather than mouse arthritis models [21]. Gait analysis, which has been performed in both rat and mouse models, provides indirect  but not direct measurement and quantification of inflammation [22,23]. The reduced need to sacrifice animals for preclinical arthritis studies, the option to perform intraindividual longitudinal studies, and the fact that quantitative evaluation of inflammation is obtained faster with PET/CT than with histopathological examinations all encourage the use of PET/CT. In contrast to histopathology, 18 F-FDG PET/CT cannot reveal the exact cellular composition of inflammatory infiltrates. However, the necessity to acquire this information depends upon the specific question of the study and therefore is only casually addressed in experimental arthritis studies in detail. Moreover, this shortcoming of 18 F-FDG PET/CT is compensated by functional knowledge and an increased range of values. First, histopathological assessment gives information about the number of cells present at the site of inflammation, whereas 18 F-FDG PET/CT gives information about metabolic activation of these cells. This difference facilitates the early detection of arthritis resolution. In our experiments, this is exemplified by the deviance in G6PI arthritis assessment at day 21 by using histology or PET/CT and macroscopical evaluation. Second, even among mice with maximal scores in semiquantitative histopathological assessment, 18 F-FDG PET/CT still showed quantitative differences and thus enhanced discernability. Another aspect of 18 F-FDG is the reduction of methodological variability since approval as a nuclear medicinal product for humans guarantees a constant quality of the agent and therefore reproducibility of data.
In regard to CT images in small-animal PET/CT hybrid systems, only a small technological step in image resolution is necessary to use CT data not only for gaining anatomical information but, beyond that, to assess inflammation and erosive processes in a single in vivo data acquisition, extending the usefulness of this sophisticated molecular imaging technique. The use of PET/ CT is not restricted to 18 F-FDG or glucose metabolism. In human autoimmune diseases, the approaches of other imaging techniques, in particular scintigraphy, using radiolabelled antibodies or drugs for staging and therapy decision-making, increasingly attracted interest [24,25]. Therefore, other forms of small-animal PET/CT application worth further developmental work are localizing targets of therapeutical monoclonal antibodies in situ, if these were labelled with radioisotopes, and the possibility of revealing the presence or expression level of specific receptors correlated with disease in inflammatory regions [26][27][28].