CXCR4-targeted imaging of leukocyte mobilization after myocardial infarction

Objectives: : The severity and duration of inflammation after myocardial infarction (MI) is thought to impact infarct size and contractile dysfunction. We aimed to non-invasively quantify leukocyte infiltration in the infarct using CXCR4-targeted imaging, and to determine the effect on infarct stability and heart failure progression.

Methods: C57Bl6 mice underwent permanent occlusion (n=50) or 60min ischemia/reperfusion (n=9) of the left coronary artery, or sham surgery (n=5) to obtain a broad range of inflammation severity. Subgroups of animals underwent macrophage depletion by clodronate-loaded liposomes (n=6) or targeted anti-inflammatory treatment with CXCR4 antagonist AMD3100 (n=8). Inflammation was assessed by serial PET imaging using the CXCR4-ligand ⁶⁸Ga-pentixafor within the first week after MI. LV function and infarct size were assessed at 6wks by cardiac magnetic resonance and perfusion SPECT. Results: In vitro assays in isolated leukocytes demonstrated comparable uptake of ⁶⁸Ga-pentixafor by neutrophils, monocytes, and macrophages. In vivo, ⁶⁸Ga-pentixafor signal was elevated in the infarct territory as early as 1h post-MI (1.2±0.04%ID/g vs 0.5±0.1 for sham, p<0.005), consistent with acute neutrophil infiltration. This signal remained elevated at 1d and 3d (1d:1.0±0.2, 3d:0.8±0.1, p<0.005 vs sham), and declined by 7d (0.5±0.1, p=0.9 vs sham), paralleling total leukocyte content in the infarct territory. CXCR4 signal in the spleen peaked at d1 (%ID/g, d1: 1.0±0.4, d3: 0.7±0.3, d7: 0.6±0.2, p=0.02) and correlated with the infarct signal (r=0.79, p<0.001), suggesting systemic inflammatory activation. Mice that died of LV rupture (n=4) showed sustained inflammation at 3d post-MI compared to survivors (1.1±0.2 vs 0.8±0.1, p<0.001). LVEF was impaired at 6wks compared to sham (45±16 vs 70±7, p=0.001), and associated with significant ventricular dilatation (µl; end systolic volume: 60±35 vs 15±5; p=0.005). ⁶⁸Ga-Pentixafor uptake in the infarct on d1 and d3 was proportional to 6wk infarct size, ESV and LVEF. On stepwise multivariate analysis, 3d ⁶⁸Ga-pentixafor uptake predicted depressed late LVEF, independent of infarct size (r_partial=-0.50, p<0.03). Macrophage depletion led to sustained CXCR4 upregulation over 3d post-MI, and resulted in increased acute LV rupture rate (50% vs 22%, p<0.001). By contrast, targeted CXCR4 blockade at the peak of PET-defined upregulation (125µg 1h post-MI) reduced subsequent LV rupture rate (0% vs 22%, p<0.001), and improved function (LVEF, 40±13 vs 29±10, p=0.04). Conclusions: CXCR4 is a suitable target for leukocyte imaging, which is upregulated in the infarct territory within 1h of ischemia. Intensity and persistence of the CXCR4 signal predict acute LV rupture and subsequent LV remodelling. Targeted CXCR4 blockade, but not gross macrophage depletion, improves outcome. This experimental work provides the foundation for novel image-guided, inflammation targeted strategies to improve infarct healing.