Ferumoxytol-enhanced MRI assessment of venous Thrombus resolution and macrophage content in a murine deep vein thrombosis model

Abstract

Background: Imaging evaluation of acute deep vein thrombosis (DVT) or post-thrombotic syndrome (PTS) in animal or clinical models is limited to anatomical assessment of the location and extent of thrombi. We hypothesize that Fe-MRI, used to evaluate macrophage content in other inflammatory diseases, can be useful to evaluate the thromboinflammatory features after DVT over time.

Methods: Nineteen wild-type CD-1 mice underwent surgical IVC ligation to induce DVT. Mice received either saline or 5 mg/kg of 14E11, a Factor XI inhibitor, before the procedure. Fe-MRI was performed on days 6-7 after ligation to evaluate thrombus volume, perfusion, and macrophage content via T₂-weighted images. Mice were euthanized at days 3-15 after surgery. The thrombi and adjacent vein walls were excised, weighed, formalin-fixed, and paraffin-embedded for immunohistological analysis. Specimens were stained with specific antibodies to evaluate macrophage content, collagen deposition, neovascularization, and recanalization. Significance was determined using the Mann-Whitney U or Student's t-test.

Results: After IVC-ligation in control mice, thrombus weights decreased by 59 % from day 3 to 15. Thrombus volumes peaked on day 5 before decreasing by 85 % by day 13. FXI inhibition led to reduced macrophage content in both thrombi (p = .008) and vein walls (p = .01), decreased thrombus volume (p = .03), and decreased thrombus mass (p = .01) compared to control mice. CCR2+ staining corroborated these findings, showing significantly reduced macrophage presence in the thrombi (p = .002) and vein wall (p = .002).

Conclusions: Fe-MRI T₂ relaxation times can be used to characterize and quantify post-thrombotic changes of perfusion, macrophage content, and thrombus volume over time in a surgical mouse model of venous thrombosis. This approach could lead to better quantification of in vivo inflammation correlating monocyte and macrophage content within resolving thrombi and veins and may serve as a useful tool for research and clinically in the evaluation of the post-thrombotic environment.

Keywords: FXI inhibition; Ferumoxytol; Magnetic resonance; Mouse; thrombus resolution.

Figure 1.

Axial T₂ images were obtained after inferior vena caval ligation and thrombus formation. (a) Pre-contrast injection T₂-weighted image. (b) Post-contrast injection T₂-weighted image. (c) 24-hour post-contrast injection T₂-weighted image. Arrows delineate the IVC.

Figure 2.

Thrombus weight changes over time following inferior vena cava (IVC) ligation over 15 days. Thrombus weights normalized to body weight post-IVC ligation in saline-treated mice showed a significant decrease (**p = .0002) in thrombus mass over time from post-op day 3 (n= 2, 1.37±.24) to day 15 (n=6, .56±.09), with a peak mean thrombus weight at post-op day 5 (n=2, 1.63±.23).

Figure 3.

Thrombus volume changes over time following inferior vena cava (IVC) ligation over 13 days. Thrombus volume in saline-treated mice showed a decrease in thrombus volume from post-op day 3 (n= 1, 5.85) to post-op day 13 (n=1, 1.12) with a peak mean thrombus weight at post-op day 5 (n=1, 7.52).

Figure 4.

Effect of pharmacological modulation of factor XI (FXI) by 14E11 on thrombus resolution following inferior vena cava (IVC) ligation. (a) Thrombus weights normalized to body weight 7 days post IVC ligation in saline- and 14E11-treated mice (n = 9–10, 1.3±0.2 vs. 1.0±0.2, **p = .01) showed a significant decrease in thrombus mass of 14E11-treated mice. (b) Thrombus volumes 7 days post IVC ligation in saline- and 14E11-treated mice (n = 9–10, 6.4±1.9 vs. 4.9±1.2, **p = .03) showed a significant reduction in volume in 14E11-treated mice.

Figure 5.

Effect of pharmacological inhibition of factor XI (FXI) by 14E11 on both thrombus formation and vein wall ferumoxytol (Fe) penetration, as assessed by ferumoxytol enhanced magnetic resonance imaging (Fe-MRI). T₂ Fe-MRI analysis of tissue penetration shows no significant difference in the delta T₂ value related to thrombus formation between saline- and 14E11-treated mice (n = 9–12, 2.8±1.4 vs. 1.7±1.4, p = .12). No statistically significant difference was observed in the analysis of vein wall ferumoxytol penetration (p = .16)

Figure 6.

Representative T₂ relaxation maps and segmentation of regions of interest using Fe-MRI. Axial T₂ relaxation maps were obtained after inferior vena caval ligation and thrombus formation. Using MATLAB, segmentations of the thrombus separate from the vein wall were performed. (a) Saline-treated mouse pre-contrast injection. (b) Saline-treated mouse 24 hours post-contrast injection. (c) 14E11-treated mouse pre-contrast injection. (d) 14E11-treated mouse 24 hours post-contrast injection. Representative pre-post contrast images are each for the same mouse in the respective condition.

Figure 7.

Effect of pharmacological inhibition of factor XI (FXI) by 14E11 on macrophage content. (a) T₂ Fe-MRI analysis of macrophage content in the vein wall 7 days post inferior vena cava (IVC) ligation in saline- and 14E11-treated mice demonstrated significantly lower delta T₂ values in 14E11-treated mice (n = 9–12, 17.4±4.8 vs. 12.7±3.9, **p = .01). T₂ Fe-MRI analysis of macrophage content in the thrombus 7 days post inferior vena cava (IVC) ligation in saline- and 14E11-treated mice also demonstrated significantly lower delta T₂ values in 14E11-treated mice and thrombus (n = 9–12, 13.4±4.2 vs. 8.8±3.7, **p = .008). (b) Quantification of the percent area of ionized calcium binding adaptor molecule 1 (Iba1) in saline- and 14E11-treated mice suggested a significant decrease in macrophage recruitment in the thrombus of 14E11 treated mice (n = 9–10, 18±8.4 vs. 8.3±6.9, *p = .02). No significant difference was found between groups in analysis of the vein wall (n = 9–10, 23±18 vs. 13±12, p = .11). (c) Quantification of pro-inflammatory macrophages by CCR2+ in the thrombus of saline- and 14E11-treated mice (n = 9–10, 39.5±17.9 vs. 18.8±7.3, **p = .002) and vein wall (n = 9–10, 56.4±12.4 vs. 33.1±18.3, **p = .002)

Figure 8.

(a) Effect of pharmacological inhibition of factor XI (FXI) by 14E11 on the estimation of collagen deposition in the thrombus and vein wall, as assessed by ferumoxytol-enhanced magnetic resonance imaging (Fe-MRI). T₂ Fe-MRI analysis of collagen deposition demonstrates a significant decrease signal in the thrombus of 14E11-treated (n = 9, 26.54±6.70, **p = .005) versus saline-treated mice (n = 12, 31.32±7.45). No statistically significant difference was observed in the analysis of vein wall collagen deposition (p = .06). (b) Effect of pharmacological inhibition of factor XI (FXI) by 14E11 on collagen deposition in the vein wall and thrombus on day 7 post inferior vena cava (IVC) ligation as determined by Masson’s trichrome staining in saline- and 14E11-treated mice suggested a significant increase in collagen deposition in the vein walls (n = 9–10, 5.9±9.0 vs. 48±34, **p = .0009) and thrombus (n = 9–10, 8.4±4.0 vs. 20±12, *p = .03) of 14E11-treated mice compared to saline-treated. (c) Trichrome immunostaining of thrombus and vein wall in saline treated animal at 20x. (d) Trichrome immunostaining of thrombus and vein wall in 14E11 treated animal at 20x.

Figure 9.

Effect of pharmacological inhibition of factor XI (FXI) by 14E11 on endothelial cell counts in the vein wall and thrombus on day 7 post inferior vena cava (IVC) ligation. (a) Quantification of endothelial cells in the thrombus of saline- and 14E11-treated mice (n = 9–10, 19±11 vs. 31±11, *p = .05) and vein wall (n = 9–10, 27±13 vs. 29±7.8, p = .56). (b) CD31 immunostaining of saline treated mice at 20x. (c) CD31 immunostaining of 14E11 treated mice at 20x.

Figure 10.

Effect of pharmacological inhibition of factor XI (FXI) by 14E11 on vascular and tissue remodeling through the expression of smooth muscle actin (SMA) in the vein wall and thrombus on day 7 post inferior vena cava (IVC). Quantification of percent area of smooth muscle actin of the thrombus in saline- and 14E11-treated mice (n = 9–10, 20±7.9 vs. 13±5.8, p = .11) and in the vein wall (n = 9–10, 91±8.9 vs. 92±9.2, p = >.99)