Molecular imaging of the immune-fibrosis axis to guide therapeutic intervention after myocardial infarction

Abstract Objectives Myocardial infarction evokes inflammation and activation of quiescent cardiac fibroblasts, leading to heart failure progression. We hypothesized that targeted molecular imaging of fibroblast activation protein (FAP) on activated fibroblasts could predict subsequent remodeling and monitor fibroblast response to anti-inflammatory therapy. Methods In vitro uptake of 68Ga-FAPI-46 was assessed in human cardiac fibroblasts stimulated by supernatant from inflammatory macrophages. C57Bl/6 mice underwent 60min left coronary artery ischemia/ reperfusion (I/R) (n=22) or sham (n=7). Serial 68Ga-FAPI-46 and 18F-FDG PET imaging was performed 1d, 3d, 7d, 14d and 42d after surgery. Left ventricle (LV) function and perfusion defect size were assessed by MRI and SPECT at 6wks. Additional animals were treated with anti-inflammatory enalapril and similarly scanned. Results Exposure to M1-like macrophage supernatant increased 68Ga-FAPI46 uptake by cardiac fibroblasts (p=0.01), associated with higher expression of IL-1β (p<0.05) and TGF-β (p=0.003). I/R induced moderate infarct size at 6wks post I/R (18.1±5.3% of LV). I/R animals exhibited increased LV volumes compared to sham (systole, 38±7 vs 21±4µL, p<0.001; diastole, 69±9 vs 49±4µL, p<0.001) and decreased LV ejection fraction (EF)(46±10 vs 58±2%, p<0.001). Serial FAP imaging revealed biphasic upregulation in the infarcted myocardium. Signal intensity 3d after injury was significantly increased (%ID/g, 0.69.±0.1 vs 1.3±0.3, p<0.001). After declining to sham levels at 7d, signal was again elevated 14d after I/R(%ID/g 0.70±0.1 vs 0.94±0.1, p=0.05) before returning to sham levels at 6wks. Signal increase was not restricted to infarct area, but similarly observed in non-infarcted remote myocardium at 3d (p<0.001) and 14d (p<0.05), suggesting involvement of both replacement and reactive fibrosis. There is modest inverse correlation between 3d infarct area FAPI signal and LVEF 6wks later (r=-0.43, p<0.05), but stronger with 14d FAPI signal (r=-0.51, p<0.05). Assuming the elevation at 3d would reflect inflammatory cell infiltration and lead to replacement fibrosis, whereas the later wave would derive from interstitial fibrosis, we treated animals with ACE inhibitor enalapril for 10d, beginning 2d prior to surgery to prevent mobilization of myeloid cells from the spleen. FAPI signal was significantly reduced in the infarct area in enalapril-treated mice at 3d (p=0.05), supporting the notion that early FAP expression is regulated by inflammatory environment. Conclusions Molecular imaging of FAP provides a prognostic biomarker of subsequent remodeling reflecting both early inflammation driven replacement and later interstitial fibrosis, ultimately supporting its capacity to monitor novel therapies.