Cardiovascular magnetic resonance to differentiate veteran athlete's heart with cavity dilatation and mild dilated cardiomyopathy

Abstract Introduction Endurance athletic training may lead to left ventricular (LV) dilatation and mildly reduced resting LV function which can be difficult to differentiate from dilated cardiomyopathy (DCM). Myocardial fibrosis is increasingly recognised in lifelong athletes and is also prevalent in DCM where it confers adverse prognosis.(1,2) However, it is unknown whether the pattern and prevalence of fibrosis in athletes with cavity dilatation differs from DCM. In this study, we compared the CMR fibrosis distribution and tissue characteristics between athletic LV dilatation and mild DCM patients. Methods We prospectively recruited 113 males; 64 endurance athletes and 49 mild DCM patients. Inclusion criteria Age 50-80 years, LVEF 45-54% and LV end-diastolic volume indexed to body surface area (LVEDVi)≥110ml/m2. Athletes trained≥10 weekly hrs for≥15 yrs. Exclusion criteria; Chest pain, prior coronary revascularisation, severe valvular disease, myocarditis, hypertrophic cardiomyopathy, inducible ischaemia or myocardial infarction on CMR. CMR protocol included volumetric assessment, T1 mapping, quantitative stress perfusion and quantitative late gadolinium enhancement. Statistical analysis between groups was performed using unpaired t-test and receiver-operator curve (ROC) analysis. Results LVEDVi was not significantly different between athletes and mild DCM patients (123.3±12.6 vs 129.8±23.1ml/m2, P=0.057). However, LVEF (52.0±6.1 vs 47.6±5.2%, P<0.001) and right ventricular (RV) EDVi (121.0±14.3 vs 97.6±25.2ml/m2, P<0.001) were both greater in athletes. There was no difference in non-ischaemic fibrosis prevalence between both groups (50.0 vs 49.0%, P=0.915) nor the burden of fibrosis (3.5±2.9 vs 7.4±12.0g, P=0.087). However, the distribution of fibrosis varied with a significantly greater prevalence of basal mid-myocardial inferolateral fibrosis amongst athletes (87.5 vs 50.0%, P=0.002) whereas basal mid-myocardial inferoseptal fibrosis was significantly more common in mild DCM (45.8 vs 9.4%, P=0.002). Native T1 (1249.0±38.1 vs 1308.3±47.1ms, P<0.001) and extracellular volume (ECV) (22.0±2.1 vs 25.9±3.5%, P<0.001) were both lower in athletes than mild DCM patients. Furthermore, athletes had higher myocardial perfusion reserve (MPR) (3.65±1.30 vs 2.76±0.92, P<0.001) and stress myocardial blood flow (MBF) (2.09±0.70 vs 1.62±0.66, P<0.001). On ROC analysis, native T1 (area under curve (AUC) 0.89, P<0.001), ECV (AUC 0.85, P<0.001) and stress MBF (AUC 0.68, P<0.001) were able to differentiate athletes and mild DCM. Native T1 and ECV were significantly better at discriminating than MPR (P<0.001). Conclusion Myocardial fibrosis was highly prevalent in both veteran endurance athlete's heart and mild DCM whilst its distribution was distinctive between the groups. Native T1 and ECV were the best discriminators. Recognition of fibrosis patterns and associated tissue characteristics may be clinically useful to differentiate these two overlapping phenotypes.Figure 1; LGE distribution in athletes cFigure 2; CMR tissue characteristics to