Reproducibility and Prognostic Value of 4D Flow CMR Quantification of AV Valve Regurgitation Post-Fontan. A FORCE registry study

Presenting Author(s)

• 

  Tarek Alsaied, MD

  Associate Professor of Pediatric Cardiology
  Dallas Children's Medical Center
  Irving, Texas, United States

Primary Author(s)

• 

  Tarek Alsaied, MD

  Associate Professor of Pediatric Cardiology
  Dallas Children's Medical Center
  Irving, Texas, United States

Co-Author(s)

• HL

  Husayn M. Ladha

  Undergraduate Student
  Case Western Reserve University
  Parsippany-Troy Hills, New Jersey, United States
• JW

  Justin Wang, BA

  Medical Student
  University of Pittsburgh Medical Center, United States
• 

  Adam B. Christopher, MD

  Assistant professor
  UPMC Children's Hospital of Pittsburgh
  Pittsburgh, Pennsylvania, United States
• 

  Rajesh Krishnamurthy, MD

  Professor
  The Ohio State University
  Columbus, Ohio, United States
• 

  Francesca Raimondi, MD

  Consultant
  ASST-Papa Giovanni XXIII Bergamo
  Bergamo, Lombardia, Italy
• 

  Joshua D. Robinson, MD, FSCMR

  Section Head, Cardiac Imaging
  Lurie Children's / Northwestern University
  Chicago, Illinois, United States
• 

  Yue-Hin Loke, MD

  Associate Professor of Pediatrics
  Children's National Hospital
  Washington DC, District of Columbia, United States
• LO

  Laura Olivieri, MD

  Associate Professor
  Children's National Hospital
  Washington, District of Columbia, United States
• 

  Mark A. Fogel, MD

  Professor
  University of Pennsylvania
  PHILADELPHIA , Pennsylvania, United States
• RC

  Rafael Ceschin

  Associate Professor
  Children's Hospital of Pittsburgh of UPMC
  Pittsburgh, United States
• 

  Rahul H. Rathod, MD, MBA

  Associate Chair of Cardiology
  Boston Children's Hospital
  Boston, Massachusetts, United States
• 

  Tarek Alsaied, MD

  Associate Professor of Pediatric Cardiology
  Dallas Children's Medical Center
  Irving, Texas, United States

Background: Atrioventricular valve regurgitation (AVVR) remains a common complication in post-Fontan patients, yet accurate quantification is hindered by limitations in echocardiography and 2D CMR. 4D flow CMR may overcome these challenges, but its clinical performance in this population remains unexplored.

Methods: In this multicenter study from the Fontan Outcome Registry using CMR (FORCE), we assessed AVVR using two 4D flow CMR methods to calculate regurgitant fractions (RFs): AVVR_{AVV-Ao (4D)}, calculated from atrioventricular valve (AVV) and aortic flows/ AVV flow, and AVVR_Jet, which directly measures AVVR jet volume/ AVV flow (Figure). We compared these against categorical AVVR severity by CMR and echo, and examiner-reported RF (AVVR_Report-RF). Correlation, kappa agreement, and Bland-Altman analyses were performed. Logistic regression assessed associations with a composite adverse clinical outcome. The composite clinical outcome included transplant listing, atrial arrhythmia, and mortality.

Results: The study included 218 post-Fontan patients (median age 16.2 years) (Table1). Both 4D methods were highly reproducible (interobserver intraclass correlation coefficient ≥ 0.98 interobserver). AVVR_{AVV-Ao (4D)} agreed moderately with examiner-reported RFs in classifying the severity of AVVR(κ = 0.50), while AVVR_Jet showed better agreement with echo-based assessment (κ = 0.34). Both 4D methods underestimated regurgitant fraction by an average of 7% compared to AVVR_Report-RF. AVVR_{AVV-Ao (4D)} was significantly associated with the composite adverse clinical outcomes (OR = 1.53, p = 0.0045; AUC = 0.71). Similarly, AVVR_Jet associated with the composite adverse outcome (OR = 1.74, p = 0.002; AUC = 0.65). Conventional examiner-reported RFs were not associated with outcomes although it was available in the report in only ~25% of patients (Table2).

Conclusion: 4D flow CMR provides reproducible, quantitative assessment of AVVR in post-Fontan patients, with significant prognostic value. AVVR_{AVV-Ao (4D)} and AVVR_Jet methods had moderate agreement with conventional assessments and predict adverse outcomes, supporting their integration into routine post-Fontan surveillance.

Table 1: Demographic, clinical, and CMR characteristics of study cohort. Variables were presented in Median (IQR) or n(%).

Patient characteristics	N= 218
Age at CMR (year), Median (IQR)	16.2 (12.8, 22.5)
Female, n (%)	79 (36.2%)
Body mass index (Kg/m2)	21.1 (18.0, 25.9)
Dominant ventricular morphology
Right	96 (44.0%)
Left	83 (38.1%)
Balanced / mixed	37 (17.0%)
Genetic diagnosis	12 (5.5%)
Fontan type
Lateral tunnel	64 (29.4%)
Extracardiac	139 (63.8%)
Atrial-pulmonary	1 (0.5%)
Other	14 (6.4%)
Age at Fontan (year)	2.8 (2.2, 3.9)
Four or more total cardiac surgeries	43 (19.7%)
AV valve procedure prior to CMR	20 (9.2%)
Surgical valve repair	1 (0.5%)
Catheter balloon dilation	0 (0%)
Surgical valve replacement	16 (7.3%)
Other	3 (1.4%)
CMR characteristics
Single ventricle EDV (ml/m2)	96.7 (79.6, 123.6)
Single ventricle ejection fraction (%)	50.4 (44.2, 55.3)
Single ventricle ESV (ml/m2)	50.1 (36.1, 66.2)
Single ventricle mass (gm/m2)	43.5 (31.9, 54.8)
Ascending aorta flow rate(mlL/m2)	2.8 (2.4, 3.3)
Qp/Qs (number)	0.98 (0.9, 1.1)

Table2. Association of AVVR Quantification Methods with Composite Adverse Outcomes

 

Method (AVVR)	OR (95% CI)	p-value	AUC	Sensitivity	Specificity
AVVR-AVV-Ao (4D)	1.53 (1.13 – 2.07)	0.004	0.714	0.769	0.663
AVVR-Jet (4D)	1.74 (1.21 – 2.51)	0.002	0.649	0.462	0.844
Examiner-Reported RF	1.13 (0.55 – 2.21)	0.711	0.560	0.600	0.638

Figure: Multiple AVVR regurgitant fraction (RFs) and corresponding severity classification were calculated for each heart using two 4D flow-based methods. 〖AVVR〗_(AVV-Ao(4D)): Difference of AVV diastolic flow and net aorta flow divided by AVV diastolic flow 〖AVVR〗_jet: Total AVV jet flow divided by AVV diastolic flow