Quantitative interpretation of FDG PET for cardiac sarcoidosis reclassifies visually interpreted exams and potentially impacts downstream interventions

Background: FDG PET is used in cardiac sarcoidosis (CS) diagnosis and management, including decisions about initiation and titration of immunosuppression. However, optimal methods to identify sarcoidosis-related inflammation on these scans is unknown. Traditional interpretive methods for FDG PET rely on qualitative visual analysis, but quantitative techniques including standardized uptake values (SUVs) may be more specific. This study evaluated the diagnostic reclassification of FDG PET studies using quantitative versus qualitative analysis and evaluated the potential impact of reclassification on downstream management and events. Methods: Cardiac-focused FDG PET examinations performed for the evaluation of CS were analyzed, comparing results from the clinically reported visual analysis to quantitative re-analysis using left ventricular maximal SUV values (SUVmax). Net diagnostic reclassification index (NDI) was calculated and compared to admissions, deaths, ICD placements, immunosuppression initiation/escalation. Of 154 exams, 22 were reclassified from positive to negative using quantitative re-analysis whereas only 2 clinically reported negative exams were quantitatively reclassified to positive, leading to a NDI of -13.0%. In the quantitatively negative/clinically reported positive group, 11 patients had immunosuppression adjusted after 22 exams and 4 ICDs were placed. Conclusions: Quantitative re-analysis of FDG PET for CS led to an overall negative diagnostic reclassification from positive to negative. Studies that were clinically reported as positive by visual analysis but reclassified as negative by quantitative analysis had numerous medical interventions but few clinical events. The low event rate suggests the use of quantitative interpretation of FDG PET for CS may help in providing providers with a more targeted therapeutic framework. (Sarcoidosis Vasc Diffuse Lung Dis 2018; 35: 342-353)