Cardiac MRI parametric mapping aids cardiac amyloidosis diagnosis in advanced renal dysfunction cohorts.

This single-institution cohort study assessed the diagnostic performance of cardiac MRI (CMR) in 65 patients with advanced renal dysfunction (ARD), defined as a GFR <30 mL/min/1.73 m², dialysis dependence, or renal transplant. The population had suspected cardiac amyloidosis (CA), and CMR assessment included T1 relaxation time, extracellular volume (ECV), T1 scout, late gadolinium enhancement (LGE), and overall reader likelihood. Diagnosis was established via PYP scintigraphy grade ≥2, positive endomyocardial biopsy, or positive extracardiac biopsy with clinical features.

Among the 65 patients, 14 (22%) received a confirmed CA diagnosis. CMR parametric mapping showed significantly higher T1 times and ECV in patients with CA compared to the cohort (p<0.001). ECV reliably predicted CA with an area under the curve (AUC) of 0.87, while T1 time yielded an AUC of 0.88. Using an ECV cutoff of ≥45% provided 75% sensitivity and 80% specificity, whereas a T1 time cutoff of ≥1390 ms offered 75% sensitivity and 85% specificity. LGE was observed in 86% of patients with CA and 84% of those without, indicating limited discriminatory power for this specific metric in this population.

Safety and tolerability data were not reported, and no adverse events or discontinuations were documented. Key limitations include the single-institution setting and the lack of prior data on CMR utility in this specific ARD population. The study suggests CMR parametric mapping exhibits high negative predictive value for CA, with improved positive predictive value when higher cutoffs are applied. However, the overall reader impression showed high negative predictive value but low positive predictive value. These findings are observational and may not generalize beyond the studied ARD cohort.