Determination of scar area using native and post-contrast T1 mapping: Agreement with late gadolinium enhancement. Issue 150 (May 2022)

• Record Type: Journal Article
• Authors: Kolentinis, Michael; Carerj, Ludovica M.; Vidalakis, Eleftherios; Giokoglu, Eleni; Martin, Simon; Arendt, Christophe; Vogl, Thomas J.; Nagel, Eike; Puntmann, Valentina O.
• Is Part Of: European journal of radiology. Issue 150(2022)
• Publication Date: 2022-05
• Subjects: T1 mapping -- Scar -- Late gadolinium enhancement; Medical radiology -- Periodicals; Radiology -- Periodicals; Radiologie médicale -- Périodiques; Medical radiology; Periodicals; 616.075705
• DOI: 10.1016/j.ejrad.2022.110242
• Languages: English
• ISSNs: 0720-048X

Abstract

Graphical abstract: Highlights: Late gadolinium enhancement remains the method of choice for ischemic scar quantification. Native T1 mapping underestimates scar area compared to late gadolinium enhancement. Post contrast T1 mapping systematically overestimates scar area in proportion to scar area extent. Post contrast T1 mapping is highly reproducible. Abstract: The purpose of this study is to ascertain agreement in measurements of the scar area between late gadolinium enhancement (LGE), native and post-contrast T1 mapping in patients with known ischemic heart disease. 132 patients (age 60 ± 11 yrs, male 82%) were included in the study. Corresponding 3 short axis slices images of LGE, native and post contrast T1 mapping were used. Scar area was evaluated semi- quantitatively with FWHM methods, in which scar is automatically determined by specialized post-processing software. Agreement per culprit vessel was also assessed. Concordance and inter- intraobserver reproducibility were assessed with Bland-Altman analysis. The results show that scar area amounted to 12.6% of myocardium for LGE, 9.1% for native (p < 0.05) and 19.4% (p < 0.05) for post-contrast T1 mapping. LAD and RCA territory infarcts showed statistical discrepancy for both T1 acquisitions. Intraobserver differences in infarct size were comparable at 0.39% ± 0.28, 2.93% ± 0.03 and 0.97% ± 0.01 respectively (p≫0.05). Interobserver differences were 5.56% ± 0.91 for LGE, 11.87% ± 3.21 (p < 0.05) for native and 5.55% ± 2.87 (p≫0.05) for post-contrast T1 mapping. In conclusion, native T1 acquisitions systematically underestimated infarct size in comparison to LGE, while post-contrast T1 overestimated it. Variances in measurements were most pronounced for LAD and RCA territory infarcts. Intraobserver reproducibility was similar with both methods, whereas interobserver variability for native T1 mapping acquisition was worse.