Determination of scar area using native and post contrast T1 mapping: a validation study

Abstract

\n \n \n Type of funding sources: Public Institution(s). Main funding source(s): Goethe University Hospital\n \n \n \n Late gadolinium enhancement (LGE) with Full Width Half Maximum (FWHM) is the standard imaging method for the measurement of the scar area in ischemic heart disease, a major determinant of adverse outcome. T1 mapping techniques have become increasingly applied in everyday clinical practice. Agreement in area quantification between T1 mapping and LGE for the scar extent has not been systematically assessed.\n \n \n \n This study aims to ascertain concordance between the area determined by native and post contrast T1 mapping against LGE to determine scar area in patients with a prior myocardial infarction. It also examines intra and inter-observer reproducibility for these two methods.\n \n \n \n Subendocardial scar was evaluated in 132 patients with ischemic heart disease using LGE, native and post contrast T1 mapping in 3 short axis slices. Native and postcontrast T1 mapping was performed using GoetheCVI®MOLLI sequence (Gadovist® 0.1mmol/kg). FWHM method was applied for LGE and compared with the manually delineated area of scar in native and postcontrast T1 mapping acquisitions within the identical slices. The presence of the scar was defined as LGE >3% of the myocardial mass. Bland-Altman methods were employed to assess concordance between techniques and reproducibly between observers.\n \n \n \n In comparison to LGE, area measurements in native T1 acquisitions underestimated infarct size (9,1% vs 12,6%, p\u2009=\u20090.01), while postcontrast T1 overestimated it (19,4% vs 12,6%, p\u2009<\u20090.001). The disparity between the measurements was accentuated with larger scar areas (Figure 1). Intraobserver reproducibility (native T1: intra/inter MD\u2009±\u2009SD, CoV, postcontrast T1: intra/inter) was similar with both methods, whereas interobserver variability for native T1 mapping acquisition was worse.\n \n \n \n We demonstrate that measurements of the area of myocardial scar by T1 mapping acquisitions yield considerably different results in comparison to the standard LGE-based FWHM method. High observer agreement indicates that these differences are genuinely related to the type of the underlying acquisition and the differences in detected signal. The approaches are not interchangeable. However, the importance of these results for predicting patient outcome is yet to be elucidated. Nonetheless, these results were consistent in the inter and intraobserver analysis for post contrast T1 mapping, although native T1 mapping failed to show similar consistency.\n Figure 1: Bland-Altmann plot of differences (mean ± upper and lower limit of agreement-LOA) in scar area extent (percentage) between LGE and native T1 mapping (A), LGE and post contrast T1 mapping (B). The difference increases linearly for larger scar areas when using post contrast mapping, which does not apply for native mapping.\n Abstract Figure 1\n