Single Ventricular Torsional Mechanics After Fontan Palliation and Their Impact on Outcomes

Abstract

Background: Abnormal left ventricular (LV) rotational mechanics in biventricular hearts are associated with adverse outcomes; however, these are less well characterized for hearts with functionally single ventricles.

Objectives: The purpose of this study was to characterize ventricular rotational mechanics in the Fontan circulation and their relationship to outcomes.

Methods: Single-center, retrospective analysis of magnetic resonance examinations for 329 Fontan patients (15 [IQR: 10-21] years) and 42 controls. The ventricular cine short-axis stack was analyzed to derive torsion metrics. Torsion calculated as the difference between apical and basal rotation normalized to ventricular length.

Results: Fontan patients had higher indexed ventricular end-diastolic volume (97 mL/body surface area^1.3 vs 72 mL/body surface area^1.3), lower ejection fraction (53% vs 60%), and lower proportion of basal clockwise rotation (62% vs 93%), apical counterclockwise rotation (77% vs 95%), and positive torsion (82% vs 100%); P < 0.001 for all. A composite outcome of death or heart transplant-listing occurred in 31 (9%) patients at a median follow-up of 3.9 years. Torsion metrics were associated with the outcome; although, on multivariate analysis only right ventricular (RV) morphology and indexed ventricular end-diastolic volume were independently associated. LVs with negative torsion, and RVs regardless of torsional pattern, had worse outcomes compared to LVs with positive torsion (P = 0.020).

Conclusions: Single ventricles in a Fontan circulation exhibit abnormal torsional mechanics, which are more pronounced for RV morphology. Abnormal torsion is associated with death or need for heart transplantation. Fontan patients with LV morphology and preserved torsion exhibit the highest transplant-free survival and torsion may offer incremental prognostic data in this group of patients.

Keywords: Fontan palliation; cardiac magnetic resonance imaging; congenital heart disease; torsion; ventricular mechanics.

Graphical abstract

Central Illustration

Representative Rotational Patterns for a Control LV and 2 Single Ventricles With a Fontan Circulation and Kaplan-Meier Survival Plots Stratified by Direction of Torsion and Ventricular Morphology Schematic representing rotational mechanics in left ventricle of a control patient, single ventricle in a Fontan patient with reduced magnitude but normal direction of torsion (>0 deg/cm) and, single ventricle in a Fontan patient with negative torsion. As viewed from the apex, counterclockwise rotation is displayed as positive and clockwise rotation as negative. Twist is defined as the difference between maximum global circumferential displacement for the basal and the apical most slices (x-y). Torsion is defined as twist divided by the length of the ventricle (x-y/L). Circumferential displacement in degrees on the vertical axis and slice number on the horizontal axis. Torsional gradient is defined as the slope of the linear regression line (dashed line) between points (red dots) represented by slice number on the horizontal axis and the maximum circumferential displacement of each slice on the vertical axis. Torsional concordance is the correlation coefficient of this linear regression line. (A) Compared to patients with negative torsion, those with positive torsion experienced a higher freedom from death or transplantation. (B) When further stratified by ventricular morphology, the left ventricular group with positive torsion experienced the highest freedom from the composite outcome. LV = left ventricle.

Figure 1

Boxplots Illustrating Abnormal Rotational Parameters in Fontan Patients Compared to Control LVs Boxplot comparison of torsion, apical rotation, and basal rotation between control left ventricles, single left ventricles, and single right ventricles. Rotational parameters differ between Fontan patients compared to control left ventricles with a more pronounced abnormality in the right ventricular group. LV = left ventricle; RV = left ventricle.

Figure 2

Venn Diagram Capturing the Overlap Between Patients With Ventricular Dilation, AVVR, and Torsion Notable overlap exists for patients with ventricular dilation, negative torsion, and significant atrioventricular valve regurgitation. Of the 301 patients with atrioventricular valve regurgitation data, 99 were dilated, 55 had negative torsion and 46 had >mild atrioventricular valve regurgitation. AVVR = atrioventricular valve regurgitation; BSA = body surface area.

Figure 3

Torsion Quartiles for Left and Right Ventricles and the Rate of Composite Outcome in Each Quartile Violin plots with medians show the distribution of torsion within each quartile by ventricular morphology and below (dots with lines) depicts the proportion of patients with the composite outcome. In the right ventricular group there is a linear decreasing trend in the prevalence of the outcome along the rising quartiles (P = 0.005). In the left ventricular group the lowest torsion quartile has the highest rate of composite outcome; however, a trend similar to the right ventricular group was not present (P = 0.747). LV = left ventricle; RV = right ventricle.

Figure 4

Bland Altman Plots Depicting Results of Interobserver and Intraobserver Reliability Analysis Interobserver (A) and intraobserver (B) reliability was assessed for torsion, twist, and global circumferential strain using a random set of 25 patients. LV = left ventricle; NS = non-significant; RV = right ventricle.