Myocardial deformation and acute cellular rejection after heart transplantation: Impact of inter‐vendor variability in diagnostic effectiveness

• Published in: Echocardiography (Mount Kisco, N.Y.) Vol. 36; no. 12; pp. 2185 - 2194
• Main Authors: Ruiz‐Ortiz, Martín, Rodriguez‐Diego, Sara, Delgado, Mónica, Kim, Jiwon, Weinsaft, Jonathan W., Ortega, Rosa, Carnero, Lucía, Sánchez, José J., Carrasco, Francisco, López‐Aguilera, José, López‐Granados, Amador, Arizón, José M., Paredes, Nick, Oneto‐Fernandez, Jesús, Pan, Manuel, Mesa, Dolores
• Format: Journal Article
• Language: English
• Published: United States 01.12.2019
• Subjects: echocardiography; heart transplantation; strain; strain; heart transplantation; echocardiography
• ISSN: 0742-2822; 1540-8175
• DOI: 10.1111/echo.14544

Purpose Our objective was to investigate the impact of inter‐vendor variability in the ability of myocardial strain analysis to detect acute cellular rejection (ACR) in heart transplant recipients. Methods We performed serial echocardiographic examinations in 18 consecutive adult heart transplanted patients, in their first year post‐transplantation, within 3 hours of the routine surveillance endomyocardial biopsies (EMB) in a single center. Myocardial strain was analyzed using two software in two different institutions, and inter‐vendor variability of strain values and its association with ACR (any grade or grade ≥2R) was investigated. The parameter of comparison was the peak value of the average curve of strain during the entire cardiac cycle. Results A total of 147 pairs of EMB‐echocardiogram were performed, 65 with no ACR, 63 with ACR grade 1R, and 19 with ACR grade ≥2R. Intra‐class correlation coefficients for left ventricle longitudinal, radial, and circumferential strain were 0.38, 0.39, and 0.77, respectively, and 0.32 for right ventricular longitudinal strain. Neither software found significant association of left ventricular longitudinal strain with rejection. Grade ≥2R ACR was associated with left ventricular circumferential strain measured with the first software and with left ventricular radial strain with the other; and ACR of any grade was only significantly associated with right ventricle longitudinal strain measured with the first software. Conclusions Inter‐vendor reproducibility of strain values was low in this study. Some strain parameters were associated to ACR, although these results were inconsistent between two commercially available software. Specific validation of each software is warranted for this clinical indication.