Microstructural Abnormalities in Heart Failure with Improved Ejection Fraction: Insights from Cardiac Diffusion Tensor Imaging

Presenting Author(s)

• 

  May T. Lwin, MB

  Clinical Research Fellow
  University of Leeds
  Leeds, England, United Kingdom

Primary Author(s)

• 

  May T. Lwin, MB

  Clinical Research Fellow
  University of Leeds
  Leeds, England, United Kingdom

Co-Author(s)

• MA

  Mehak Asad, MD, MPH

  Clinical research fellow
  University of Leeds, Multidisciplinary Cardiovascular Research Centre and Biomedical Imaging Science Department, Leeds Institute of Cardiovascular and Metabolic Medicine, LS2 9JT, United Kingdom
  Leeds, England, United Kingdom
• RT

  Raluca Tomoaia, MD, PhD

  Research Fellow
  University of Leeds
  Leeds, England, United Kingdom
• 

  Masafumi Takafuji, MD, PhD

  Clinical Research Fellow
  Leeds Institute of Cardiovascular and Metabolic Medicine
  Leeds, England, United Kingdom
• TL

  Tin Lwin, MB

  Cardiology Registrar
  Leeds Institute of Cardiovascular and Metabolic Medicine, United Kingdom
• SC

  Sam Coveney, PhD

  Research Fellow
  University of Leeds
  Leeds, England, United Kingdom
• 

  Maryum Farooq, MB

  Research Fellow
  Leeds Institute of Cardiovascular and Metabolic Medicine, University of Leeds
  Leeds, England, United Kingdom
• CS

  Chin Soo, MB

  CRF
  LICAMM
  Leeds, England, United Kingdom
• TA

  Thomas Anderton, MB

  CRF
  LICAMM
  Leeds, England, United Kingdom
• IT

  Irvin Teh, PhD

  Senior Research Fellow
  Leeds Institute of Cardiovascular and Metabolic Medicine
  Leeds, England, United Kingdom
• lb

  lisa borg, MB

  Cardiology Consultant
  Leeds Institute of Cardiovascular and Metabolic Medicine, United Kingdom
• 

  Bradley Chambers

  Clinical Research Fellow
  University of Leeds
  Harewood, England, United Kingdom
• DB

  David L. Buckley, PhD

  Professor of Medical Physics
  Leeds Institute of Cardiovascular and Metabolic Medicine, United Kingdom
• 

  Peter P. Swoboda, PhD

  Consultant Cardiologist
  Leeds Institute of Cardiovascular and Metabolic Medicine
  Leeds, England, United Kingdom
• 

  Sven Plein, MD, PhD

  Professor of Cardiology
  Leeds Institute of Cardiovascular and Metabolic Medicine
  Leeds, England, United Kingdom
• 

  Jurgen E. Schneider, PhD

  Professor of Biomedical Imaging
  Leeds Institute of Cardiovascular and Metabolic Medicine
  Leeds, England, United Kingdom
• ED

  Erica Dall'Armellina, PhD

  Associate Professor and Honorary Consultant Cardiologist
  University of Leeds, England, United Kingdom

Background: Non-ischaemic cardiomyopathy (NICM) is a significant cause of heart failure (HF) worldwide, associated with substantial mortality and morbidity. HF is associated with microstructural alterations on cardiac diffusion tensor magnetic resonance imaging (cDTI). Heart failure with improved ejection fraction (HFimpEF) has been identified as a novel clinical entity significantly different from heart failure with reduced or preserved EF. It is unknown whether functional recovery in HFimpEF is associated with normalisation of the myocardial microstructure. The study aims to compare microstructural cardiomyocyte changes between patients with HFimpEF and healthy volunteers (HV) assessed by cDTI.

Methods: Seventeen HFimpEF patients and 14 HV underwent 3T cDTI (Siemens Prisma). HFimpEF was defined as EF improved following prior severely impaired LVEF (< 40%) by CMR or echocardiography. Cine, tissue mapping pre- and post-contrast, free-breathing spin-echo DTI (3 slices, 18 diffusion directions at b-values 100s/mm2, 200s/mm2 and 500s/mm2) using second-order motion-compensated (M2) gradients, acquired resolution 2.20*2.27*8mm³ and late gadolinium enhancement (LGE) was acquired. HV had the same protocol without contrast. Global measurement for mean diffusivity (MD) (measure the freedom of water diffusion in the myocardium; high MD represents increase in interstitial fibrosis and/or oedema), fractional anisotropy (FA) (measure the directionality variability of water diffusion, low levels represent collagen infiltration and cardiomyocyte disorganisation), secondary eigenvector angle (E2A) (a marker of sheetlet orientation) were derived.

Results: Left ventricular mass did not differ significantly between groups (101.6 ± 37g vs 88.4 ± 25 g, p=0.28). LVEF remained lower in HFimpEF compared with HV (51.9 ± 4.4% vs 62.4 ± 4.5%, p< 0.0001), with a trend toward reduced RVEF (50.9 ± 9% vs 58.8 ± 6%, p=0.054). Native T1 in HFimpEF was 1319 ± 36 ms, extracellular volume (ECV) of 29 ± 5 % and T2 value of 41 ± 5 ms. Nine patients had midwall LGE positive. Despite improvement of LVEF, HFimpEF patients showed continuing elevated MD relative to healthy controls (1.56 ± 0.12 vs 1.49 ± 0.04 ×10⁻³ mm²/s, p=0.0329), and FA was also reduced in that group (0.31 ± 0.03 vs 0.35 ± 0.03, p=0.004). E2A was significantly lower in the group of HFimpEF (37.7 ± 8.2° vs 44.7 ± 7.3°, p=0.019).

Conclusion: cDTI demonstrates distinct myocardial microstructural changes in HFimpEF, suggesting incomplete myocardial recovery despite improvement of global LV systolic function. This indicates persistent tissue-level remodelling and highlights cDTI as a marker for characterising NICM beyond measurement of LVEF. Further evaluation in larger cohort studies will be needed to validate clinical utility.

Table 1: Demographic, CMR and DTI parameters of Participants

Variable	NICM with recovered EF (HFimpEF)	Healthy volunteers (HV)	P value
Age (years)	63 ±14	25.9 ± 5	< 0.001
Male (%)	58.82	28.57	0.15
BSA (m2)	1.98± 0.20	1.8± 0.20	0.0183
LV mass (g)	101.56 ±37	88.43 ±25	0.2761
LVEF (%)	51.9 ±4.4	62.4 ±4.5	< 0.0001
Native T1 (ms)	1319 ±36	-	-
ECV (extracellular volume) (%)	29 ± 5	-	-
T2 (ms)	41 ± 5	-	-
MD (x 10-3 mm2/s)	1.56 ± 0.12	1.49 ±0.04	0.0329
FA (scalar between 0 & 1)	0.31 ±0.03	0.35 ±0.03	0.0041
E2A (‘)	37.7 ±8.2	44.7 ±7.3	0.0193

Figure 1: Panel A, B, C: Comparison of MD, FA, and E2A in HFimpEF and healthy volunteers (HV)
Figure 2. Example DTI parameter maps in a healthy volunteer (HV) and a patient with HFimpEF. Maps include (left to right): mean diffusivity (MD), fractional anisotropy (FA), and second sheetlet orientation (E2A) maps.