Late gadolinium enhancement (LGE) cardiac magnetic resonance (CMR) is the gold-standard modality to assess myocardial scar. The extent of myocardial fibrosis is proportional to the risk of adverse outcomes and LGE is ≥15% of LV mass is one of the criteria for prophylactic implantable cardioverted defibrillator. Currently, there are various methods of quantifying the size of LGE which are unlikely to be interchangeable; data about the most reproducible one in Hypertrophic Cardiomyopathy (HCM) are sparse.
The aim of our study was to compare the intra-observer and inter-observer reproducibility of 5 LGE quantification techniques in patients with HCM.
100 patients with HCM (apical pattern, n = 34; non-apical patterns, n = 66) undergoing CMR were assessed. A European Association of Cardiovascular Imaging (EACVI) level III operator quantified LGE using 5 techniques: 2SD, 5SD, 6SD, full width half maximum (FWHM) and manual thresholding methods and repeated the same measurements after 1 month, in order to assess intra-observer variability. The analysis was also performed by one EACVI level I operator and one EACVI level III operator in order to assess inter-observer variability. LGE volume was calculated in all patients and in patients with apical HCM and intra-observer and inter-observer variability were assessed.
Assessment of LGE volume was highly reproducible when measured by two expert operators with all methods apart from the 2SD method, in which the size of LGE was greater. There was high intra-observer agreement with all methods, the manual one being that with the best intra-observer reproducibility (ICC 0.97 (95% CI 0.95-0.98). Both the FWHM (ICC 0.93, 0.87-0.96) and the 6SD methods (ICC 0.93, 0.88-0.96) yielded the highest inter-observer reproducibility between the expert operators. Inter-observer reproducibility between the EACVI level III and the EACVI level I operators was poor across quantification methods. Intra-observer and inter-observer reproducibility were high with manual and FWHM methods in the apical variant.
In patients with HCM, LGE quantification techniques are not interchangeable and reproducibility depends on the method and operator expertise. The FWHM and the 6SD methods were the most reproducible methods in all patients; the first one was also the most reproducible method in apical HCM alongside the manual method. Whether various LGE quantification methods have different diagnostic and prognostic impact should be further investigated.