Abstract
BACKGROUND: Non-invasive cardiac imaging allows detection of cardiac amyloidosis (CA) in patients with aortic stenosis (AS). Our objective was to estimate the prevalence of clinically suspected CA in patients with moderate and severe AS referred for cardiovascular magnetic resonance (CMR) in age and gender categories, and assess associations between AS-CA and all-cause mortality. METHODS: We retrospectively identified consecutive AS patients defined by echocardiography referred for further CMR assessment of valvular, myocardial, and aortic disease. CMR identified CA based on typical late-gadolinium enhancement (LGE) patterns, and ancillary clinical evaluation identified suspected CA. Survival analysis with the Log rank test and Cox regression compared associations between CA and mortality. RESULTS: There were 113 patients (median age 74 years, Q1-Q3: 62-82 years), 96 (85%) with severe AS. Suspected CA was present in 9 patients (8%) all > 80 years. Among those over the median age of 74 years, the prevalence of CA was 9/57 (16%), and excluding women, the prevalence was 8/25 (32%). Low-flow, low-gradient physiology was very common in CA (7/9 patients or 78%). Over a median follow-up of 18 months, 40 deaths (35%) occurred. Mortality in AS + CA patients was higher than AS alone (56% vs. 20% at 1-year, log rank 15.0, P < 0.0001). Adjusting for aortic valve replacement modeled as a time-dependent covariate, Society of Thoracic Surgery predicted risk of mortality, left ventricular ejection fraction, CA remained associated with all-cause mortality (HR = 2.92, 95% CI = 1.09-7.86, P = 0.03). CONCLUSIONS: Suspected CA appears prevalent among older male patients with AS, especially with low flow, low gradient AS, and associates with all-cause mortality. The importance of screening for CA in older AS patients and optimal treatment strategies in those with CA warrant further investigation, especially in the era of transcatheter aortic valve implantation.