ACCURACY OF MYOCARDIAL NATIVE T2 FOR THE DIAGNOSIS OF ACTIVE CARDIAC SARCOIDOSIS

Abstract

SESSION TITLE: Late-breaking Abstract Posters SESSION TYPE: Original Investigation Posters PRESENTED ON: October 18-21, 2020 PURPOSE: To determine the accuracy of native T2 on cardiac magnetic resonance (CMR) in the diagnosis of active cardiac sarcoidosis when compared to combined clinical diagnosis and cardiac positron emission tomography (PET) as the gold standard. METHODS: 24 patients were identified who had received both cardiac MRI and PET within a 6-month period between May 2017 and May 2020. 6 patients were excluded due to lack of active cardiac sarcoidosis at the time of imaging. T2 and PET American Heart Association (AHA) left ventricular segment mapping was loaded onto polar plots and compared with the gold standard. Statistical analysis was performed in XLSTAT including logistic regression and two-tailed T tests. RESULTS: A total of 18 patients were included in the analyses. Five models were found that adequately predicted the diagnosis of active cardiac sarcoidosis: PET SUV > 2.5 (R2 = 0.132, AUC = 0.722), max SUV (R2 = 0.358, AUC = 0.833), T2 > 55 msec (R2 = 0.294, AUC = 0.722), T2 >59.4 msec (R2 = 0.258, AUC = 0.736), and max T2 (R2 = 0.408, AUC = 0.875). The best model was using max T2 value as a predictor of active cardiac sarcoidosis, which suggested a 95% probability of a positive test at 59.4 msec. The odds ratio for every integer increase in max T2 msec was 2.098 (95% CI 0.92 – 4.78). CONCLUSIONS: The use of T2 imaging is at least as predictive of active cardiac sarcoidosis as PET in three different models. The best fit model was using T2 max. T2 > 55 msec was the most sensitive model (91.7%), while T2 > 59.4 was the most specific (100%). At our institution, the optimal diagnostic threshold lies between these two values. CLINICAL IMPLICATIONS: Quantifiable T2 parameters indicative of active cardiac sarcoid can be coupled with late gadolinium enhancement on CMR to diagnose both active and inactive cardiac sarcoidosis using a single imaging modality. This would ultimately allow for fewer imaging studies and lower patient radiation exposure. DISCLOSURES: Owner/Founder relationship with YellowDot Innovations, LLC Please note: $20001 - $100000 Added 07/16/2020 by Jeremy Burt, source=Web Response, value=Ownership interest No relevant relationships by Jordan Chamberlin, source=Web Response Research Grant relationship with Medtronic Please note: $5001 - $20000 Added 07/16/2020 by Brian Houston, source=Web Response, value=Grant/Research Support No relevant relationships by Ennis James, source=Web Response No relevant relationships by Madison Kocher, source=Web Response No relevant relationships by William Rieter, source=Web Response No relevant relationships by Madalyn Snoddy, source=Web Response No relevant relationships by Natalie Stringer, source=Web Response Grant recipient relationship with Siemens Healthcare Please note: >$100000 Added 07/17/2020 by Akos Varga-Szemes, source=Web Response, value=Grant/Research Support Consultant relationship with Elucid Bioimaging Please note: $5001 - $20000 Added 07/17/2020 by Akos Varga-Szemes, source=Web Response, value=Consulting fee Consultant relationship with Bayer Healthcare Please note: $1001 - $5000 Added 07/17/2020 by Akos Varga-Szemes, source=Web Response, value=Consulting fee No relevant relationships by Jeffrey Waltz, source=Web Response