Multimodality approach by cardiac magnetic resonance and biological markers in left ventricular non-compaction with heart failure with preserved ejection fraction - revealing the unknown

Type of funding sources: Public grant(s) – National budget only. Main funding source(s): PN-III-P1-1-TE-2016-0669, within PNCDI III Left ventricular non-compaction (LVNC) is associated with an increased risk of heart failure (HF). The presence of a real LVNC with HF with preserved ejection fraction (HFpEF), is still controverted. We evaluated prospectively 42 patients with HFpEF, 21 with LVNC (61 ± 9 years) and 21 without LVNC (LVC), aged and risk factor matched, by cardiac magnetic resonance (CMR) 1.5T. LVNC diagnosis was confirmed by Petersen and Jacquier criteria (NC/C ratio and the percentage of NC myocardium). We performed myocardial T1 mapping (normal value of 950 ± 21ms). We calculated a mean value of all native T1 (T1mean), and also for apical (apicalT1) and basal segments (basalT1). We also calculated ECV mean, basal and apical. All patients had NTproBNP and biomarkers for systemic inflammation (hsCRP, IL6, cystatin C and sST2), endothelial dysfunction: VCAM, von Willebrand factor (vWf), vWF metalloproteinase-ADAMTS13, and myocardial fibrosis: vascular peroxidase (VPO), and Galectin-3. In the LVNC, mean NC/C ratio was 2.9 ± 0.5 mm and the percentage of NC myocardium was 24.41 ± 8.8%. LVNC patients had significantly higher T1apical, higher ECVmean, ECV basal and apical (Table) by comparison with LVC group, suggesting an extensive fibrosis in LVNC group with significantly higher apical fibrosis.  Inflammatory markers were similar between groups, LVNC patients had lower values of ADAMTS13, suggesting endothelial dysfunction, and higher values of Galectin-3, suggesting increased myocardial fibrosis (Table). Galectin-3 correlated positively only with apicalT1 (R = 0.49, p = 0.04). NTproBNP significantly correlated with VPO, a promotor of fibrosis (r = 0.61, p = 0.009) in LVNC group, whereas in LVC group correlated with cystatin C (r = 0.62, p = 0003) and VCAM (r = 0.4, p = 0.05). Native apical T1 cut off >1021 ms provided the highest sensibility and specificity to differentiate segments with and without NC in HFpEF (p = 0.002) (Figure).  HFpEF patients with LVNC have significant higher NTproBNP, higher fibrosis than patients without LVNC, more extensive in non-compacted apical segments. Galectin-3 level correlates only with apical fibrosis on CMR, expressed by apicalT1 time. Moreover, endothelial dysfunction seems to play an important role in HFpEF generation in LVNC. All findings suggests that LVNC is a stand alone condition, not an adaptive hyper-trabeculation in HFpEF. Table.Comparison between groups NTproBNP (pg/ml) Galectin3 (ng/ml) ADAMTS13 (ng/ml) T1mean (ms) basalT1 (ms) apicalT1 (ms) ECV mean (%) ECV basal (%) ECV apical (%) LVNC 294 ± 282 8.44 ± 3.45 767.35 ± 335.56 1013.8 ± 31.8 1002.8 ± 27.2 1059 ± 73 27.2 ± 2.9 26.2 ± 2.9 29.6 ± 3.9 LVC 163 ± 71 6.67 ± 2.88 962.33 ± 253.78 1003.2 ± 28.1 1004.3 ± 29.5 1007 ± 40 24.3 ± 2.5 24.2 ± 2.7 25.2 ± 2.8 P value 0.031 0.048 0.049 0.26 0.865 0.007 0.002 0.033 <0.001 Abstract Figure