LAUSR

Background Myocardial disease is a major cause of death in systemic sclerosis (SSc). We hypothesised that the in SSc patients with suspected myocarditis, the Lake Louise criteria underdiagnose myocarditis compared to diagnostic models that incorporate parametric cardiovascular magnetic resonance (CMR) indices (extracellular volume fraction [ECV], native T1- and T2-mapping). Methods 32 patients with diffuse SSc (dSSc) and a clinical suspicion of myocarditis were prospectively evaluated with a 1.5 T scanner using the Lake Louise criteria, and the aforementioned parametric indices were determined. The Lake Louise criteria were compared with parametric indices individually and in two models (with and without taking late gadolinium enhancement [LGE] into account), with regard to diagnostic agreement. Results Native T1-mapping and ECV individually had the greatest discordance with the Lake Louise criteria. The presence or absence of LGE in the proposed parametric models did not lead to changes in identified patient proportions. Native T1-mapping led to the same proportion of diagnosis of myocarditis as either parametric model, with the exception of 2 patients. Most importantly, patients identified only by parametric models, had a significantly subtler presentation than those identified by the Lake Louise criteria, as exemplified by significantly lower T2 ratio, early gadolinium enhancement and T2-mapping values. Conclusions In dcSSc patients with a clinical suspicion of myocarditis, novel parametric models including native T1-mapping, T2-mapping and ECV, identified a significantly greater proportion of patients with myocarditis compared to those identified by the Lake Louise criteria, and native T1 mapping had the greatest utility in this cohort. Disclosure of Interest None declared