Molecular imaging of inflammation in the brain-heart axis after ischemic stroke: comparison of two murine stroke models.

Introduction: In addition to neurologic dysfunction, cerebral stroke increases risk of acute cardiac events and chronic heart failure. We hypothesized that inflammation is a linking mechanism, which can be interrogated by systems-based, multi-organ targeted positron emission tomography (PET).

Methods: Cerebral ischemic stroke was induced in C57Bl6 mice by intraluminal transient 45 minute middle cerebral artery occlusion (MCAo, n=32) or after craniotomy using topical application of vasoconstrictor endothelin-1 (ET-1) (n=23). In controls, the intraluminal filament was immediately withdrawn (n=8) or a vehicle was applied after craniotomy (n=7). PET imaging of translocator protein (TSPO) using ¹⁸F-GE180 assessed brain and heart inflammation at 24h, 7d and 21d after stroke. Magnetic resonance imaging (MRI) at 1wk and 3 wk determined stroke size and cardiac function.

Results: MCAo mice exhibited more severe stroke-related behaviour than ET-1. Nonetheless, Stroke-induced cellular hypoxia could be macroscopically defined in the acute phase after MCAo and ET-1 by triphenyltetrazolium chloride staining. MRI in MCAo denoted a median stroke size of 36.86 (5-82 mm³) at wk1, which declined to 5.73 (2-22 mm³) at wk3. By contrast, ET-1 did not result in an MRI-discernible stroke area. In MCAo, neuroinflammation was identified in the stroke region relative to contralateral at 7d (%ID/g_max: 3.8±0.8 vs 2.5±0.3 for contralateral region, vs 2.6±0.8 for ipsilateral sham, p<0.001). This was associated with a persisting increase in cardiac TSPO signal (%ID/g_mean 8.6±1.9 vs 6.8±1.2 for controls, p=0.036 at 24 hrs and 8.7±2.5 vs 5.6±0.8, p<0.001, at 3wk), and with a persisting decline in contractile function (% ejection fraction, 50±18 vs 63±5% for controls, p=0.078 at 1 wk and 55±6 vs 63±6%, p=0.013 at 3wk). Ipsilateral TSPO signal in the stroke region was proportional to the cardiac signal at 7d, suggesting systemic organ interaction following brain injury. Histopathology confirmed diffuse inflammatory cell infiltration in the heart early after cerebral stroke. ET-1, on the other hand, resulted in a localized higher ¹⁸F-GE180 signal in the cerebral cortex at the site of application relative to contralateral (4.3±1.0 vs 2.4±0.7, p<0.001), but was poorly distinguished from craniotomy alone (3.7±1.0, p=0.16), and did not show cardiac effects (% ejection fraction, 59±2 vs 60±3, p=0.78).

Conclusions: In contrast to topical ET-1 application after craniotomy, the intravascular MCAo model enables imaging studies of brain-heart networking after stroke, which evokes elevated TSPO PET signal in both brain and heart. The severity of cerebral ischemic damage may contribute to cardiac dysfunction via systemic inflammatory networking.