Tracking multi-organ treatment response in systemic AL amyloidosis with cardiac magnetic resonance derived extracellular volume mapping

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Background Systemic light-chain (AL) amyloidosis commonly involves the liver, spleen and heart. Cardiac magnetic resonance (CMR) with extracellular volume (ECV) mapping has demonstrated accuracy in measuring cardiac, hepatic and splenic amyloid infiltration. Purpose We sought to: (1) assess the association between baseline multi-organ ECVs and prognosis (2) assess the multi-organ response to treatment using ECV mapping, and (3) assess the association between multi-organ treatment response and prognosis. Methods We identified 351 patients with a confirmed diagnosis of systemic AL amyloidosis who underwent baseline serum amyloid P component (SAP) scintigraphy and CMR at diagnosis, of which 171 had follow-up imaging. We also recruited 20 healthy volunteers who underwent CMR with ECV mapping, without corresponding SAP scintigraphy, to allow calculation of the ECV normal ranges. Results At diagnosis, ECV mapping demonstrated that 304(86.7%) had cardiac involvement, 114(32.5%) significant hepatic involvement and 147(41.9%) significant splenic involvement. Baseline myocardial and liver ECV independently predict mortality (myocardial: HR=1.05,95CI%[1.03–1.07],P<0.001; liver: HR=1.03,95%CI[1.01–1.05],P<0.001). Liver and spleen ECV correlated with amyloid load assessed by SAP scintigraphy (R=0.751,P<0.001; R=0.765,P<0.001, respectively). Serial multi-organ ECV measurements accurately tracked treatment response as validated against serial SAP scintigraphy (the current reference standard). Multi-organ ECV regression was observed as early as 6-months in patients with a good haematological response (liver=15%, spleen=15%, heart=5%). The remaining patients with a good haematological response had stable liver and spleen ECVs, but 20% had cardiac progression. By 12-months more patients with a good haematological response demonstrated cardiac regression (liver=30%, spleen=36%, heart=32%), and this trend was maintained at 24-months. Multi-variable analysis adjusting for haematological response, change in myocardial, liver and spleen ECV demonstrated that haematological response, change in myocardial ECV (HR=1.11, 95%CI[1.02–1.19], P=0.011) and liver ECV (HR=1.06, 95%CI[1.01–1.11], P=0.015) remained independent predictors of prognosis at 6-months. Conclusions Multi-organ ECV quantification accurately tracks treatment response, and demonstrates different rates of organ regression, with the liver and spleen regressing more rapidly than the heart. A good haematological response alone is likely to induce visceral organ stabilisation/regression, but may not be sufficient to induce myocardial stabilisation/regression. Liver and myocardial ECV at diagnosis and changes in ECV at 6-months independently predict mortality. ECV mapping offers a comprehensive multi-organ assessment of treatment response and accurate prognostication.