3D Cine Circumferential Strain Better Predicts Cardiomyopathy Progression in Duchenne Muscular Dystrophy Compared to 2D Circumferential Strain (RF_TH_200)

Presenting Author(s)

• CE

  Conner C. Earl, PhD

  MD/PhD Student
  Purdue University, Indiana University School of Medicine

Primary Author(s)

• CE

  Conner C. Earl, PhD

  MD/PhD Student
  Purdue University, Indiana University School of Medicine

Co-Author(s)

• AJ

  Alexa M. Jauregui, BSc

  Medical Student
  Indiana University School of Medicine
• MP

  Marco A. Prado, BSc

  PhD Candidatee
  Purdue University
• 

  Kan N. Hor, MD

  Director, Advanced Cardiovascular Imaging and Fellowship Program, Associate Division Chief for Research
  Nationwide Children's Hospital and The Ohio State University
• LM

  Larry Markham, MD

  Professor
  Riley Hospital for Children at IU Health
• 

  Jonathan Soslow, MD, MSCI

  Professor
  Vanderbilt University Medical Center
  Vanderbilt
• CG

  Craig Goergen, PhD

  Professor
  Purdue University
• GL

  Guang Lin, PhD

  Professor
  Purdue University

Background: Cardiomyopathy (CM) is the primary cause of mortality in Duchenne muscular dystrophy (DMD). The onset and clinical progression of CM are highly variable, making it difficult to characterize the disease trajectory accurately. 2D Circumferential strain (E_cc) has been shown to be an effective metric for predicting outcomes in DMD cardiomyopathy but may be insufficient for assessing the pace of CM progression. We hypothesized that novel 4D (3D cine) cardiovascular magnetic resonance (CMR) E_cc could more effectively differentiate slow vs fast progression of CM in DMD compared to conventional 2D feature tracking E_cc methods.

Methods: We analyzed stacks of short-axis cine CMR images for n=66 DMD patients (13.3[10.8-16.5] years, median[interquartile range]) imaged annually for three consecutive years. We performed 4D strain analysis using a custom-built MATLAB (Mathworks, Natick, MA) graphical user interface by compiling short-axis slices. Regional circumferential strain (E_cc) was measured by determining the relative change in the localized 4D endocardial surface mesh compared to end-diastole. Conventional 2D strain was measured using feature tracking software (Medis, Leiden, Netherlands). For group comparison, we used a Wilcoxon Rank Sum Test with a Bonferroni correction for multiple comparisons. For the classification of slow vs fast progression, we used the area under the receiver operating characteristic curve (AUC).

Results: Using LVEF decrease, measured two years following the initial scan, we classified patients into slow (DLVEF%£5; n=37) or fast (ΔLVEF% >5; n=28) progressing groups. We also subclassified patients into cohorts based on initial LVEF and compared slow and fast progression in patients with normal (LVEF≥55%; n=23 slow; n=21 fast) and abnormal (LVEF< 55%; n=14 slow; n=7 fast) function (Figure 1C). In patients with either normal or abnormal function, there was no statistical difference between the slow and fast-progressing groups in terms of standard metrics such as age(p>0.99), LVEF(p>0.59), late gadolinium enhancement presence (p>0.36), or 2D strain (p>0.37) at the initial visit. However, 4D basal E_cc magnitude was significantly decreased in fast progressors in both normal and abnormal cohorts (p< 0.001 and p=0.02, respectively; Fig. C, D). The max AUC value for classifying slow vs fast progression was 0.87[0.76-0.97] (AUC[95% CI]) and 0.84[0.76-0.97] for basal 4D CMR strain in the normal and abnormal function groups respectively, compared to 0.59[0.42-0.76] and 0.62[0.35-0.89] in 2D basal strain (Figure 2, Table 1).

Conclusion: Regional E_cc from 4D CMR provides a clear and substantial improvement for classifying fast vs slow progression of cardiomyopathy in DMD patients compared to traditional 2D feature tracking strain.

Figure 1: 4D vs. 2D circumferential strain (Ecc) for cardiomyopathy progression in Duchenne muscular dystrophy (DMD) A) Schematic depicting study design and 4D CMR feature tracking. B) 4D circumferential strain analysis. C) Initial visit progression analysis for normal function (LVEF≥55%) comparing regional Ecc of 4D and 2D strain methods. D) Initial visit progression analysis for abnormal function (LVEF <55%) comparing regional Ecc of 4D and 2D strain methods.

Figure 2: Receiver operator curve (ROC) analysis comparing slow (LVEF%5) vs fast (LVEF%>5) progression between 4D and 2D circumferential strain (Ecc) in DMD cardiomyopathy. A) ROC curve showing regional Ecc using 4DCMR vs B) 2D CMR in DMD patients with initially normal function (LVEF≥55%). C) ROC curve showing regional Ecc using 4DCMR vs B) 2D CMR in DMD patients with initially abnormal function (LVEF <55%).