Abstract
BACKGROUND: Fibrosis, with accumulation of type I collagen, is a hallmark of postthrombotic change after deep vein thrombosis (DVT), but tools for its direct detection are lacking. Here, we investigate whether molecular magnetic resonance imaging (MRI) using a collagen-specific gadolinium-based probe can detect and measure changes in collagen during thrombus resolution and in response to treatment in a mouse model of DVT. METHODS: Venous thrombus was induced in the inferior vena cava of BALB/c mice (n=45), and MRI was performed at day 2 (n=3) and weeks 1, 2, and 3 post-surgery using the collagen-specific probe, EP-3533 (10 μmol/kg; n=11-13/group). A subgroup of mice with DVT (n=7) was treated with pravastatin in drinking water (40 mg/kg per day) for 3 weeks post-DVT. Pre- and post-EP-3533 MRI scans were performed. Magnetic resonance venography was used to measure thrombus volume. Inversion recovery T1-weighted images and T1 maps, pre- and post-contrast, were used to calculate the percent change (%) in Δ contrast-to-noise ratio, Δ signal-to-noise ratio, and Δ relaxation rate. Tissues were used for ex vivo analyses. RESULTS: EP-3533 uptake increased during thrombus organization and resolution, resulting in MRI signal enhancement, with % Δ contrast-to-noise ratio, % Δ signal-to-noise ratio, and % Δ relaxation rate peaking at 3 weeks after DVT. MRI measurements of collagen accumulation quantified as an increase in % Δ contrast-to-noise ratio (ρ=0.89; P=0.012) and % Δ relaxation rate (ρ=0.80; P=0.029) correlated positively with collagen histology. The spatial distribution of gadolinium in the tissue colocalized with collagen type I based on immunohistochemistry (ρ=0.95; P<0.001). Statin treatment decreased both collagen accumulation and vein wall thickness, without affecting thrombus size. CONCLUSIONS: Molecular MRI using a collagen-targeting probe made collagenous thrombus visible on MRI and detected changes in collagen content during thrombus resolution.