Cardiovascular Ultrasound (Sep 2020)

Quantification of left ventricular mass by echocardiography compared to cardiac magnet resonance imaging in hemodialysis patients

  • Sören Jendrik Grebe,
  • Uwe Malzahn,
  • Julian Donhauser,
  • Dan Liu,
  • Christoph Wanner,
  • Vera Krane,
  • Fabian Hammer

DOI
https://doi.org/10.1186/s12947-020-00217-y
Journal volume & issue
Vol. 18, no. 1
pp. 1 – 9

Abstract

Read online

Abstract Background Left ventricular hypertrophy (LVH), defined by the left ventricular mass index (LVMI), is highly prevalent in hemodialysis patients and a strong independent predictor of cardiovascular events. Compared to cardiac magnetic resonance imaging (CMR), echocardiography tends to overestimate the LVMI. Here, we evaluate the diagnostic performance of transthoracic echocardiography (TTE) compared to CMR regarding the assessment of LVMI in hemodialysis patients. Methods TTR and CMR data for 95 hemodialysis patients who participated in the MiREnDa trial were analyzed. The LVMI was calculated by two-dimensional (2D) TTE-guided M-mode measurements employing the American Society of Echocardiography (ASE) and Teichholz (Th) formulas, which were compared to the reference method, CMR. Results LVH was present in 44% of patients based on LVMI measured by CMR. LVMI measured by echocardiography correlated moderately with CMR, ASE: r = 0.44 (0.34–0.62); Th: r = 0.44 (0.32–0.62). Compared to CMR, both echocardiographic formulas overestimated LVMI (mean ∆LVMI (ASE-CMR): 19.5 ± 19.48 g/m2, p < 0.001; mean ∆LVMI (Th-CMR): 15.9 ± 15.89 g/m2, p < 0.001). We found greater LVMI overestimation in patients with LVH using the ASE formula compared to the Th formula. Stratification of patients into CMR LVMI quartiles showed a continuous decrease in ∆LVMI with increasing CMR LVMI quartiles for the Th formula (p < 0.001) but not for the ASE formula (p = 0.772). Bland-Altman analysis showed that the Th formula had a constant bias independent of LVMI. Both methods had good discrimination ability for the detection of LVH (ROC-AUC: 0.819 (0.737–0.901) and 0.808 (0.723–0.892) for Th and ASE, respectively). Conclusions The ASE and Th formulas overestimate LVMI in hemodialysis patients. However, the overestimation is less with the Th formula, particularly with increasing LVMI. The results suggest that the Th formula should be preferred for measurement of LVMI in chronic hemodialysis patients. Trial registration The data was derived from the following clinical trial: NCT01691053 , registered on 19 September 2012 before enrollment of the first participant.

Keywords