Hybrid PET/MR imaging for the prediction of left ventricular recovery after percutaneous revascularisation of coronary chronic total occlusions

To evaluate myocardial viability assessment with hybrid 2-deoxy-2-[${}^{18}$F]fluoro-d-glucose positron emission tomography/magnetic resonance imaging ([${}^{18}$F]FDG-PET/MR) in predicting left ventricular (LV) wall motion recovery after percutaneous revascularisation of coronary chronic total occlusion (CTO). Forty-nine patients with CTO and corresponding wall motion abnormality (WMA) underwent [${}^{18}$F]FDG-PET/MR imaging for viability assessment prior to percutaneous revascularisation. After 3–6 months, 23 patients underwent follow-up MR to evaluate wall motion recovery. In total, 124 segments were assigned to the CTO territories, while 80 segments displayed impaired wall motion. Of these, 68% (54) were concordantly viable in PET and MR; conversely, only 2 segments (2%) were assessed non-viable by both modalities. However, 30% showed a discordant viability pattern, either PET non-viable/MR viable (3 segments, 4%) or PET viable/MR non-viable (21 segments, 26%), and the latter revealed a significant wall motion improvement at follow-up (p = 0.033). Combined imaging by [${}^{18}$F]FDG-PET/MR showed a fair accuracy in predicting myocardial recovery after CTO revascularisation (PET/MR area under ROC curve (AUC) = 0.72, p = 0.002), which was superior to LGE-MR (AUC = 0.66) and [${}^{18}$F]FDG-PET (AUC = 0.58) alone. Hybrid PET/MR imaging prior to CTO revascularisation predicts more accurately the recovery of dysfunctional myocardium than PET or MR alone. Its complementary information may identify regions of viable myocardium with increased potential for functional recovery.