Quantifying right ventricular diffuse fibrosis in tetralogy of Fallot - a novel customised approach for the challenges of the right ventricle

NIHR Cardiovascular Biomedical Research Unit, Royal Brompton & Harefield NHS Foundation Trust and Imperial College London, London, UK Full list of author information is available at the end of the article Figure 1 Representative short-axis images from subjects scanned for RV T1 mapping by fat-water separated, MSPrep dark blood imaging. Images for five subjects with repaired tetralogy of Fallot and five healthy volunteers shown (one subject per column): Top row MoCo averaged water-only image at Ts 600 ms, Second row MoCo averaged water-only anchor image at same window/level, Third row MoCo averaged fat only image, Bottom row T1 map generated from registration and 2-parameter fit of the six images per sampling scheme. Heng et al. Journal of Cardiovascular Magnetic Resonance 2016, 18(Suppl 1):O26 http://www.jcmr-online.com/content/18/S1/O26

Quantifying right ventricular diffuse fibrosis in tetralogy of Fallot -a novel customised approach for the challenges of the right ventricle Ee Ling Heng 1,2* , Peter Kellman 3 , Michael A Gatzoulis 1,2 , James Moon 4 , Peter Gatehouse 1,5 , Sonya V Babu-Narayan 1,2  Figure 1 Representative short-axis images from subjects scanned for RV T1 mapping by fat-water separated, MSPrep dark blood imaging. Images for five subjects with repaired tetralogy of Fallot and five healthy volunteers shown (one subject per column): Top row -MoCo averaged water-only image at Ts 600 ms, Second row -MoCo averaged water-only anchor image at same window/level, Third row -MoCo averaged fat only image, Bottom row -T1 map generated from registration and 2-parameter fit of the six images per sampling scheme.
Heng et al. Journal of Cardiovascular Magnetic Resonance 2016, 18(Suppl 1):O26 http://www.jcmr-online.com/content/18/S1/O26 adjacent strong signals from blood and epicardial fat. We prospectively aimed to explore the possible clinical significance of RV diffuse fibrosis in rTOF compared to health.

Clinical
Data for 22 subjects (11 patients with rTOF and 11 age and gender-matched healthy volunteers) were acquired. Two-parameter fit pixelwise T1 maps were generated for each run. Mean RV free wall T1s were measured by two independent observers using CMR42. Interobserver reproducibility was calculated by coefficient of variation (CoV%) = (within-subject standard deviation/mean) x100%.

Results
RV T1 maps were obtained in all subjects, with interobserver reproducibility of native RV myocardial T1 (CoV 1.8%) and RV ECV (CoV 6.8%). There was no significant difference in RV native T1 and ECV of patients with rTOF compared to the controls who had thin RV wallls (Figure 2). This may reflect the modest sample size, or the inclusion of clinically stable patients with rTOF with minimal residual haemodynamic lesions, or may also reflect technical limitations. Saturation was optimised to <1% in the RV but non-uniformity over the heart requires investigation. The MSPrep required subject-specific optimisation for minimal partial volume contamination by blood which may explain high RV ECV. Known underestimation of native T1 by MOLLI compared to SASHA may also explain some of the RV-LV difference.

Conclusions
RV T1 and ECV quantification is possible with the proposed technique but requires further development. The diagnostic value in this patient population merits further work towards a larger study and to explore histological correlation.