Cardiac magnetic resonance T1 imaging for the diagnosis of cardiac amyloidosis among patients with symptomatic heart failure

Abstract Background Cardiac amyloidosis (CA) is a significant cause of heart failure in elderly patients. However, specific signs in cardiac imaging are lacking, hence bone scintigraphy in combination with screening for monoclonal protein, or myocardial biopsy is required for the diagnosis. Cardiac magnetic resonance (CMR) imaging with T1 mapping is a promising diagnostic tool, which may allow to differentiate causes of left ventricular (LV) hypertrophy, particularly CA. Purpose The aim of this study was to test the diagnostic accuracy of T1 mapping for the diagnosis of CA in elderly patients with symptomatic heart failure and red flags for CA. Methods We prospectively enrolled 120 patients aged above 60 years with symptomatic heart failure NYHA II-IV, LV ejection fraction ≥ 40%, increased LV wall thickness of ≥ 12 mm end-diastolic on echocardiography, and elevated cardiac biomarkers (N-terminal pro-B-type natriuretic peptide (NT-proBNP) ≥ 1,000 ng/L, high-sensitivity cardiac troponin T (hs-cTnT) ≥ 14 ng/L). All patients underwent standardized quantitative CMR with a 3.0T scanner. 99mTc-DPD bone scintigraphy, blood screening for monoclonal protein, and, where applicable, cardiac biopsy, were performed as reference methods in all patients. Primary endpoint was diagnostic accuracy of native T1 mapping for the diagnosis of CA. Results 112 patients were included in the final analysis. Median age was 81.0 years (interquartile range 74.0–85.0 years), and 45% were female. CA was diagnosed in 38 patients (34%), of which 32 suffered from a Transthyretin (ATTR) and 6 patients from light chain (AL) amyloidosis. The areas under the receiver operating characteristic curves for native T1 were 0.713, 0.772, 0.836 for global, septal short-axis (SAX), and septal 4-chamber (4Ch) view measurements. The corresponding value of extracellular volume fraction (ECV) was 0.904 (Figure 1). Based on these results, we developed a new diagnostic algorithm for the diagnosis of CA including T1 septal 4Ch and ECV, with a sensitivity and positive predictive value of both 92.9% and a specificity and negative predictive value of 95.8% (Figure 2). Thereby, contrast application can be avoided in 33% of patients. A definite diagnosis could be provided in 81%. Conclusion CMR with T1 imaging provides a high diagnostic accuracy to diagnose CA in patients with symptomatic heart failure at risk for CA. These data may help to identify underlying causes in heart failure which require specific treatment. An earlier diagnosis of CA may improve outcome of these patients.Figure 1Figure 2