The prognostic value of multiparametric cardiac magnetic resonance in patients with systemic light chain amyloidosis

Front Oncol. 2023 May 3:13:1069788. doi: 10.3389/fonc.2023.1069788. eCollection 2023.

Abstract

Background: Late gadolinium enhancement (LGE) is a classic imaging modality derived from cardiac magnetic resonance (CMR), which is commonly used to describe cardiac tissue characterization. T1 mapping with extracellular volume (ECV) and native T1 are novel quantitative parameters. The prognostic value of multiparametric CMR in patients with light chain (AL) amyloidosis remains to be thoroughly investigated.

Methods: A total of 89 subjects with AL amyloidosis were enrolled from April 2016 to January 2021, and all of them underwent CMR on a 3.0 T scanner. The clinical outcome and therapeutic effect were observed. Cox regression was used to investigate the effect of multiple CMR parameters on outcomes in this population.

Results: LGE extent, native T1 and ECV correlated well with cardiac biomarkers. During a median follow-up of 40 months, 21 patients died. ECV (hazard ratio [HR]: 2.087 for per 10% increase, 95% confidence interval [CI]: 1.379-3.157, P < 0.001) and native T1 (HR: 2.443 for per 100 ms increase, 95% CI: 1.381-4.321, P=0.002) were independently predictive of mortality. A novel prognostic staging system based on median native T1 (1344 ms) and ECV (40%) was similar to Mayo 2004 Stage, and the 5-year estimated overall survival rates in Stage I, II, and III were 95%, 80%, and 53%, respectively. In patients with ECV > 40%, receiving autologous stem cell transplantation had higher cardiac and renal response rates than conventional chemotherapy.

Conclusion: Both native T1 and ECV independently predict mortality in patients with AL amyloidosis. Receiving autologous stem cell transplantation is effective and significantly improves the clinical outcomes in patients with ECV > 40%.

Keywords: T1 mapping; cardiac magnetic resonance imaging; late gadolinium enhancement; light chain amyloidosis; prognosis.

Grants and funding

This work was supported by the National Natural Science Foundation of China (82270767) and the Key Research and Development Plan Project of Jiangsu Province - Social Development Projects (BE2017721).