Abstract 13823: Clinical Presentation Rather Than Imaging Findings Determine Outcomes in Cardiac Sarcoidosis

Abstract

Background: Cardiac PET and MRI have facilitated increased diagnosis of Cardiac Sarcoidosis (CS). Varied clinical presentations such as advanced heart block, ventricular arrhythmias and heart failure (HF) are initiating this diagnostic pathway but it is unclear if prognosis is dictated by clinical presentation or imaging findings. In our quaternary care center, we conducted a retrospective review of patients diagnosed with CS by cardiac MRI/PET and examined outcomes in relation to imaging findings and clinical presentation. Methods: 223 patients underwent evaluation for suspected CS by cardiac MRI/PET between 2016-2019 and 62 were identified as having CS based on clinical and imaging criteria. We divided this cohort based on their primary clinical indication for imaging- arrhythmia, HF or screening due to history of pulmonary sarcoidosis. Outcome defined as MACE included death, ICD-shock or admission for HF. Results: Study population was predominantly male (65%), Caucasian (79%) with mean age 60±10 years, Referral diagnosis was HF in 37 (59%), arrhythmia (defined as high grade AV block/ventricular arrhythmias) in 15 (25%) and screening in 10 (16%). While there were no differences in the CMRI LGE findings or degree of inflammation by FDG-PET, there was a higher incidence of resting anteroseptal perfusion on PET, with a higher summed rest score in the arrhythmia group (mean 2 vs 0 in other groups; p<0.05). Based on imaging findings, 39% were treated with steroids, 25% with immunosuppressant, 10% were on combination. Over mean follow-up of 3.5 years, there were 26 events- 6 deaths, 10 ICD shocks 10 heart failure admissions. Event-free survival was 53%, 76% and 80% in the arrhythmia, HF and screening groups respectively. In multivariable analysis, a model that included VT presentation (OR 11; p<0.05) and global hypokinesis on PET images (OR 10; p<0.05) had a sensitivity of 86% and specificity of 63% for MACE prediction. Compared to the model, neither Inflammation on FDG-PET or LGE (MRI) predicted outcomes and treatment with steroids and/or immunosuppressant did not alter prognosis. Conclusions: In this single center study of CS, a primary diagnosis of ventricular arrhythmia and global LV systolic dysfunction, regardless of LVEF, had the worst prognosis.