Molecular imaging of murine intestinal inflammation with 2-deoxy-2-[18F]fluoro-D-glucose and positron emission tomography

Abstract

Background & aims: 2-Deoxy-2-[(18)F]fluoro-d-glucose (FDG) uptake by positron emission tomography (PET), a measure of glucose transporter activity, has been used to detect mucosal inflammation. However, there is limited understanding of the biologic basis of mucosal FDG uptake.

Methods: A contrast-based computed tomographic isocontour method was developed to identify intestinal anatomic regions, and FDG uptake was integrated over these regions to achieve reproducible quantification during longitudinal assessment of individual mice. Intestinal FDG uptake was compared with histologic scores and with glucose transporter 1 levels in mucosal immune cells by flow cytometry.

Results: Intestinal FDG uptake quantitatively correlated with disease activity in mild (C3H/HeJ.IL-10(-/-)) and severe (129.Galphai2(-/-), CD4(+) CD45RB(high), and Galphai2(-/-) CD3(+) transfer) murine colitis models at all time points examined (P < .05) and was sufficiently sensitive to detect preclinical inflammation. FDG uptake was correlated by flow cytometric detection of glucose transporter 1 levels in mucosal CD4(+) T lymphocyte but not other intestinal immune cell types. CD4(+) T-cell transfer in vivo confirmed that mucosal FDG uptake was associated with the activated but not quiescent state. When intestinal inflammation was increased by treatment with piroxicam and decreased with anti-TL1A treatment, FDG uptake was correspondingly altered.

Conclusions: This study clarifies the cellular basis of FDG signal in intestinal inflammation and introduces computed tomographic isocontour analysis of FDG-PET imaging for standardized quantitation of immune colitis.

Figure 1

Definition of intestine ROI using AMIDE isocontour function and contrast agent. (A) CT imaging of mice with and without contrast agent. (B) Default CT isocontour ROI generated with contrast agent. (C) Processed whole intestines ROI separated into large and small intestines’ ROIs.

Figure 2

FDG-PET in Gαi2^−/− and control mice. (A) Coronal and (B) sagittal views of CT-PET overlays and PET images of Gαi^−/− and heterozy-gote control mice with large intestine contrast agent. (C) Quantitation of mean SUV and mean percent ID/g per voxel in small and large intestine ROIs for individual mice. Gαi2^−/−, squares; Gαi2^+/− controls, triangles.

Figure 3

FDG-PET in Gd2^−/−CD3+ cell transfer colitis. (A) Coronal and (B) sagittal views of PET-CT overlays and PET images of mice treated with anti-TL1A or control antibody. (C) Quantification of mean percent ID/g and mean SUV per voxel in the large intestine for individual mice. (D) Values for histology score and percent ID/g in the large intestine were plotted and analyzed by linear regression. (C and D) Anti-TL1A and control antibody values are represented by closed and open squares, respectively.

Figure 4

FDG-PET in IL-10^−/− and control mice. (A) Coronal and (B) sagittal views of CT and PET overlays of L-10^−/− and C3H/OuJ control mice with large intestine contrast agent. Average (±SEM) of (C) mean percent D/g and (D) mean SUV for all mice scanned in indicated time interval. IL-10^−/−, open squares; C3H/OuJ control, closed circles.

Figure 5

Effect of piroxicam-in-duced inflammation on FDG uptake by IL-10^−/− mice. (A) Quantitation of mean SUV and (B) percent ID/g in small and large intestine ROIs for IL-10^−/− mice treated or untreated with piroxicam. (C) Coronal, (D) sagittal, and (E) transverse views of piroxicam-treated mice. (F) Correlation of histology and mean percent ID/g for individual C3H/OuJ (closedcircles), IL-10^−/− (open squares), and piroxicam-treated L-10^−/− mice (closedsquares). ns, not significant.

Figure 6

Glut-1 surface expression by intestinal cell types. Intestinal cells were isolated, and flow cytometry was performed on cells stained for lineage markers and Glut-1 in Gαi2^−/− (blue) and Gd2^+/− (red) mice; cells stained with negative control for Glut-1 (black). (A) CD3⁺CD4⁺ T cells, (B) macrophages (CD11b⁺F4/80⁺), (C) CD3⁺CD8⁺ T cells. Glut-1 levels in intraepithelial mononuclear cell populations. Intraepithelial lymphocyte mononuclear cells were isolated from the large intestine of individual IL-10^−/− mice with or without 2 weeks of piroxicam treatment and stained for lineage markers and Glut-1. Cells were gated on (D) CD3⁺CD4⁺ T cells, (E) F4/80⁺CD11b⁺ cells, or (F) CD3⁺CD8⁺ T cells. Glut-1 levels (S mean fluorescence intensity) were tabulated. Glut-1 levels and percent ID/g for individual mice were graphed, and correlation analysis was used to calculate P values. n/s, not significant.

Figure 7

FDG-PET in CD4⁺ CD45RB^high transfer colitis. (A) Coronal and sagittal views of CD4⁺ CD45RB^high, CD4⁺ CD45RB^{high + low}, and nontransferred C.B-17 scid mice. (B) Body weight over time (days) for CD4⁺ CD45RB^high (open squares), CD4⁺CD45RB^{high + low} (closed triangles), and C.B-17 scid (closed circles) mice. (C) Peripheral blood neutrophil numbers over time. (D) Mean corpuscular hemoglobin levels over time. (E) Mean percent ID/g over time. (F) Mean SUV over time. (G) Comparison of histology score and percent ID/g (day 42), analyzed by linear regression. For C–F, statistical comparison of each group by Student t test to the scid group is shown by asterisks: *P < .02; **P < .001; ***P < .0001. Comparisons for CD4⁺ CD45RB^{high + low} mice were nonsignificant for all assays and time points.