Molecular Magnetic Resonance Imaging of Aneurysmal Inflammation Using a Redox Active Iron Complex.

Abstract

Inflammation plays a key role in driving brain aneurysmal instability and rupture, but clinical tools to noninvasively differentiate between inflamed and stable aneurysms are lacking. We hypothesize that imaging oxidative changes in the aneurysmal microenvironment driven by myeloid inflammatory cells may represent a noninvasive biomarker to evaluate rupture risk. In this study, we performed initial evaluation of the oxidatively activated probe Fe-PyC3A as a tool for magnetic resonance imaging (MRI) of inflammation in a rabbit model of saccular aneurysm. The difference in longitudinal relaxivity ( r1 ) in reduced and oxidized states of Fe-PyC3A was measured in water and blood plasma phantoms at 3 T. A rabbit saccular aneurysm model was created by endovascular intervention/elastinolysis with subsequent decellularization in situ. Rabbits were imaged at 4 weeks (n = 4) or 12 weeks (n = 4) after aneurysmal induction, when luminal levels of inflammation reflected by the presence of myeloperoxidase positive cells are relatively high and low, respectively, using a 3 T clinical scanner. Both groups were imaged dynamically using a 2-dimensional T1-weighted fast field echo pulse MRI sequence before and up to 4 minutes postinjection of Fe-PyC3A. Dynamic imaging was then repeated after an injection of gadobutrol (0.1 mmol/kg) as negative control probe. Rabbits from the 12-week aneurysm group were also imaged before and 20 minutes and 3 hours after injection of Fe-PyC3A using an axial respiratory gated turbo-spin echo (TSE) pulse sequence with motion-sensitized driven equilibrium (MSDE) preparation. The MSDE/TSE imaging was repeated before, immediately after dynamic acquisition (20 minutes postinjection), and 3 hours after injection of gadobutrol. Aneurysmal enhancement ratios (ERs) were calculated by dividing the postinjection aneurysm versus skeletal muscle contrast ratio by the preinjection contrast ratio. After imaging, the aneurysms were excised and inflammatory infiltrate was characterized by fluorometric detection of myeloperoxidase activity and calprotectin immunostaining, respectively. In vitro relaxometry showed that oxidation of Fe-PyC3A by hydrogen peroxide resulted in a 15-fold increase of r1 at 3 T. Relaxometry in the presence of blood plasma showed no more than a 10% increase of r1 , indicating the absence of strong interaction of Fe-PyC3A with plasma proteins. Dynamic imaging with Fe-PyC3A generated little signal enhancement within the blood pool or adjacent muscle but did generate a transient increase in aneurysmal ER that was significantly greater 4 weeks versus 12 weeks after aneurysm induction (1.6 ± 0.30 vs 1.2 ± 0.03, P < 0.05). Dynamic imaging with gadobutrol generated strong aneurysmal enhancement, but also strong enhancement of the blood and muscle resulting in smaller relative ER change. In the 12-week group of rabbits, MSDE/TSE imaging showed that ER values measured immediately after dynamic MRI (20 minutes postinjection) were significantly higher ( P < 0.05) in the case of Fe-PyC3A (1.25 ± 0.06) than for gadobutrol injection (1.03 ± 0.03). Immunohistochemical corroboration using anticalprotectin antibody showed that leukocyte infiltration into the vessel walls and luminal thrombi was significantly higher in the 4-week group versus 12-week aneurysms (123 ± 37 vs 18 ± 7 cells/mm 2 , P < 0.05). Magnetic resonance imaging using Fe-PyC3A injection in dynamic or delayed acquisition modes was shown to generate a higher magnetic resonance signal enhancement in aneurysms that exhibit higher degree of inflammation. The results of our pilot experiments support further evaluation of MRI using Fe-PyC3A as a noninvasive marker of aneurysmal inflammation.