Serial Assessment of Right Ventricular Deformation in Patients With Hypoplastic Left Heart Syndrome: A Cardiovascular Magnetic Resonance Feature Tracking Study.

Background As right ventricular dysfunction is a major cause of adverse outcome in patients with hypoplastic left heart syndrome, the aim was to assess right ventricular function and deformation after Fontan completion by performing 2-dimensional cardiovascular magnetic resonance feature tracking in serial cardiovascular magnetic resonance studies. Methods and Results Cardiovascular magnetic resonance examinations of 108 patients with hypoplastic left heart syndrome (female: 31) were analyzed. Short-axis cine images were used for right ventricular volumetry. Two-dimensional cardiovascular magnetic resonance feature tracking was performed using long-axis and short-axis cine images to measure myocardial global longitudinal, circumferential, and radial strain. All patients had at least 2 cardiovascular magnetic resonance examinations after Fontan completion and 41 patients had 3 examinations. Global strain values and right ventricular ejection fraction decreased from the first to the third examination with a significant decline in global longitudinal strain from the first examination to the second examination (median, first, and third quartile: -18.8%, [-20.5;-16.5] versus -16.9%, [-19.3;-14.7]) and from the first to the third examination in 41 patients (-18.6%, [-20.9;-15.7] versus -15.8%, [-18.7;-12.6]; P-values <0.004). Right ventricular ejection fraction decreased significantly from the first to the third examination (55.4%, [49.8;59.3] versus 50.2%, [45.0;55.9]; P<0.002) and from the second to the third examination (53.8%, [47.2;58.7] versus 50.2%, [45.0;55.9]; P<0.0002). Conclusions Serial assessment of cardiovascular magnetic resonance studies in patients with hypoplastic left heart syndrome after Fontan completion demonstrates a significant reduction in global strain values and right ventricular ejection fraction at follow-up. The significant reduction in global longitudinal strain between the first 2 examinations with non-significant changes in right ventricular ejection fraction suggest that global longitudinal strain measured by 2-dimensional cardiovascular magnetic resonance feature tracking might be a superior technique for the detection of changes in myocardial function.

2‐dimensional cardiovascular magnetic resonance feature tracking

body surface area

global circumferential strain

global longitudinal strain

global radial strain

hypoplastic left heart syndrome

neo‐aortic valve regurgitation

right ventricular end‐diastolic volume index

right ventricular ejection fraction

right ventricular end‐systolic volume index

right ventricular myocardial mass index

right ventricular stroke volume index

total cavopulmonary connection

tricuspid valve regurgitation

Clinical Perspective

What Is New?

Most hypoplastic left heart syndrome studies have assessed right ventricle (RV) function and size during staged palliation but serial data about RV deformation after Fontan completion are rare.

There is evidence that myocardial longitudinal strain and strain rate decrease during Fontan follow‐up before a decline in RV ejection fraction manifests.

Combined volumetry and analysis of long‐axis function using cardiovascular magnetic resonance feature tracking seem to be suitable to detect RV dysfunction early, but further research is warranted to develop protocols and functional analysis tools best suited to assess the abnormally loaded RV in hypoplastic left heart syndrome.

What Are the Clinical Implications?

Regular and life‐long follow‐up of patients with hypoplastic left heart syndrome in specialized congenital heart centers including monitoring of RV size and function is mandatory.

Current practice aims for a repeated cardiovascular magnetic resonance scan every 3 to 5 years depending on the patient’s condition and future serial studies may be supplemented by additional cardiac markers, such as cardiac laboratory markers.

The findings of this study encourage to proceed with further research to detect reliable factors that contribute to early myocardial RV dysfunction in patients with hypoplastic left heart syndrome.

Hypoplastic left heart syndrome (HLHS) is one of the most severe forms of congenital heart disease with fatal outcome if untreated. It is characterized by hypoplasia of the left sided heart structures, including mitral and aortic valvular atresia or stenosis as well as hypoplasia of the ascending aorta. The established surgical treatment strategy involves a 3‐stage palliation with creation of a total cavopulmonary connection (TCPC) being the third step. Although survival rates for patients born in the 21st century compared with those born at the early 90s increased, , information about right ventricular (RV) function and deformation in patients with HLHS in Fontan circulation is still sparse. ,

In a recent study, our group showed a significant increase in RV volumes with only mild reduction in RV ejection in patients with HLHS during serial follow‐up.

Several studies have demonstrated that echocardiography and two‐dimensional speckle tracking echocardiography are suitable to evaluate RV function through staged palliation. , ,

2D‐CMR‐FT can be used in patients with HLHS. It has been shown to be comparable to 2‐dimensional speckle tracking echocardiography for the assessment of longitudinal strain, allows an exact measurement of ventricular size and ejection fraction (EF) and has the advantage of unlimited imaging windows.

The aim of this study was to use 2D‐CMR‐FT to investigate global and regional right ventricular myocardial deformation in a large cohort of patients with HLHS during protocolized serial follow‐up. Furthermore, we aimed to evaluate the value of deformation parameters to detect early RV dysfunction in patients with HLHS.

Methods

The data that support the findings of this study are available from the corresponding author upon reasonable request within the limits of ethical and legal restrictions.

Ethical Statement

The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the local ethics committee (ID Nr.: D503/20, date of the approval: 10th August 2020). Informed consent was obtained from the parents or guardians of the children enrolled into the study.

Patients

One hundred and eight pediatric and adult patients with HLHS [median age at first scan 4.5 years (1st and 3rd quartile: 3.9;6.4 years)] who underwent clinical CMR imaging as part of routine clinical follow‐up between December 2005 and July 2021 were included. Inclusion criteria were: (1) a completed Fontan circulation, (2) the availability of at least 2 serial CMR studies after Fontan completion, and (3) no contraindications for CMR. Patients were excluded if: (1) Fontan circulation was not completed, (2) only one CMR study after Fontan completion was performed, and (3) if CMR data sets were of insufficient quality.

CMR studies with axial long‐axis cine views and in most cases a complete stack of short‐axis cines were required to measure RV deformation, volumes, and RVEF. Patients with insufficient data sets were excluded. Patient characteristics were obtained from medical records and included underlying diagnosis, variables related to surgical palliation and history of complications (protein‐losing enteropathy, plastic bronchitis, arrhythmias, thromboembolic events). Data from the clinical follow‐up included oxygen saturation and current medication.

CMR Acquisition

CMR was performed using a 3T or 1.5T scanner. In small children conscious sedation with midazolam and propofol in younger children was needed (154/257 CMR studies). In sedated patients, heart rate, respiratory motion, oxygen saturation and noninvasive blood pressure were monitored.

Short‐axis cine stacks were acquired using gradient echo or steady‐state free precession cine imaging with retrospective ECG gating, to measure RV volumes, myocardial mass, and function. The scan parameters were as follows: field of view 175 to 450 mm, slice thickness 5 to 8 mm, 20 to 30 cardiac phases, no slice gap, non‐breath‐hold in sedated children, breath‐hold in awake patients. In addition, axial long‐axis cine images showing the atria and right ventricle in a similar manner as a standard 4‐chamber view was obtained with the following scan parameters: field of view 175 to 400 mm, slice thickness 5 to 8 mm, 20 to 30 cardiac phases, non‐breath‐hold in sedated children, breath‐hold in awake patients.

CMR Analysis

Post‐processing was performed using commercially available software (cvi42 for Cardiovascular MRI, Circle Cardiovascular Imaging, Calgary, Canada; Medis Suite Solutions, Medical Imaging Software, Leiden, the Netherlands). All CMR measurements were performed by the same observer (LK) and carefully checked by a senior observer specializing in congenital CMR (IV).

Assessment of RV end‐diastolic and end‐systolic volumes (RVEDV, RVESV), myocardial mass and RVEF was performed by manual tracing of endocardial contours for each slice from the stack of short‐axis cine images at end‐systole and end‐diastole using Simpson’s method. The position of the tricuspid valve was confirmed by linking the short‐axis stack to a long‐axis view of the RV. Trabeculations and papillary muscles were included into the ventricular volume. All volumes and ventricular mass were indexed to body surface area (BSA).

Assessments of RV myocardial peak strain, strain rate, velocity and displacement were performed based on manual tracing of the endocardial and epicardial contours in ECG gated CMR images (QStrain, Medis Medical Imaging Systems BV). Peak strain and strain rate values were measured as these parameters are relatively load independent or less sensitive to expected ventricular dyssynchrony and temporal resolution. , , , , Global circumferential (GCS) and radial strain (GRS) and strain rates (GCSR and GRSR) were measured from short‐axis cine images at 3 ventricular levels (basal, mid‐ventricular and apical); see Figure 1. Global longitudinal strain and strain rate values (GLS and GLSR; Figure 1) were analyzed from axial long‐axis cine images showing the atria and right ventricle. Simultaneously, velocity and displacement were measured in longitudinal, radial, and circumferential direction. Arithmetic means of segmental values in long‐axis images (7 segments) and short‐axis images (16 segments) were calculated for the analysis of global deformation parameters.

Figure 1

Segmental CMR feature tracking in the short‐ and long‐axis of the RV.

CMR indicates cardiovascular magnetic resonance; and RV, right ventricle.

The degree of tricuspid and neo‐aortic valve regurgitation (TR, neo‐AVR) was documented from CMR and echocardiographic reports and was classified as none, trivial (regurgitant fraction <5%), mild (regurgitant fraction <20%), moderate (regurgitant <40%), and severe (regurgitant fraction 40% and higher).

In 9/257 CMR studies imaging data sets were incomplete or of reduced quality and did not allow short‐axis deformation analysis. In 7 cases RV volumetry was performed from transaxial cine stacks.

To assess inter‐observer variability, myocardial strain parameters (SBG, LK) and RV volumetry parameters (IV, LK) were measured twice by 2 experienced operators in 30 patients.

Statistical Analysis

Statistical analysis was performed using MedCalc version 19.3.1 (MedCalc Software Ltd, Belgium) and SPSS version 25. Normal distribution was assessed using the Shapiro‐Wilk test and checked visually from histograms. Normally distributed continuous variables are presented as mean and standard deviation or as median with first and third quartile (Q1;Q3) in non‐normally distributed data. Categorical variables are expressed as total counts with percentages. Paired samples Wilcoxon test was performed to compare results between serial CMR examinations. Correlations of RVEF and strain parameters were assessed using Spearman’s coefficient of rank correlation for non‐normally distributed data.

An explorative factor analysis based on a principal components method (Varimax) was performed to find the variables best suitable to detect serial RV changes in patients with HLHS. The number of factors was determined by exploiting the decrease in the eigenvalues indicating the explained overall variance.

Inter‐observer variability was assessed with the intraclass correlation coefficient (ICC). ICC values between 0.5 and 0.75 indicate moderate, values between 0.75 and 0.9 indicate good and values >0.9 indicate excellent reliability. Considering multiple testing the significance level was adjusted by Bonferroni correction to 0.006. The global significance level of 5% was divided by the number of comparisons at 3 time points which were additionally multiplied by 3 as considering the 3 main post‐processing categories (volumetry and feature tracking parameters in short‐axis and long‐axis views, respectively). This calculation resulted in a corrected significance level of 0.006.

Results

Patient Characteristics

Data could be obtained from 108 patients with HLHS. A short overview is shown in Table 1 and additional patient characteristics are displayed in Supplemental Material (Table S1).

Table 1

Patient Characteristics

Parameters	1st CMR (n=108)	2nd CMR (n=108)	3rd CMR (n=41)
Age, y	4.5 [3.9;6.4]	10.1 [9.1;12.5]	15.3 [14.0;17.2]
Female/male (n, %)	31/77 (28.7%/71.3%)	31/77 (28.7%/71.3%)	11/30 (26.8%/73.2%)
Weight, kg	17.0 [15.6;20.0]	31.0 [26.9;37.6]	54.0 [44.8;63.3]
Height, cm	104.5 [100.0;114.0]	137.0 [130.0;150.5]	163.0 [154.5;173.0]
BMI, kg/m²	15.7 [14.7;16.6]	16.2 [15.3;18.1]	19.5 [16.9;23.1]
Heart rate, bpm	83.0 [75.0;89.0]	80.0 [70.0;90.0]	78.1 [72.3;86.3]
DBP, mm Hg	45.0 [41.0;51.8]	56.0 [47.0;66.0]	67.0 [58.0;78.0]
SBP, mm Hg	85.0 [80.0;90.8]	100.0 [90.0;116.0]	117.0 [106.8;129.5]
Oxygen saturation (%)	90.0 [87.0;94.0]	91.0 [88.5;95.0]	92.0 [88.0;95.0]
Intraatrial lateral tunnel (n, %)	103 (95.4%)	103 (95.4%)	41 (100%)
Extracardiac conduit (n, %)	5 (4.6%)	5 (4.6%)	0 (0%)
Open Fenestration (n, %)	79 (73.1%)	70 (64.8%)	22 (53.7%)

Data are presented as frequencies (%) or median [first and third quartile]. BMI indicates body mass index; bpm, beats per minute; CMR indicates cardiovascular magnetic resonance; DBP, diastolic blood pressure; SBP, systolic blood pressure; and y, years.

The first scan was performed at a median age of 4.5 (3.9;6.4) years and 1.8 (1.3;3.2) years after TCPC completion. TCPC was completed at a median age of 2.5 (2.2;2.9) years. The median interval between the first and second examination was 5.2 (4.8;6.0) years and between the first and third scan 10 (8.8;11.7) years respectively.

Most patients had an intraatrial lateral tunnel (n=103, 95.4%; extracardiac conduit: n=5, 4.6%). Mild TR and trivial neo‐AVR were common (third CMR examination: 63.4% and 48.8%, respectively; Tables 1 and S1).

CMR Results

There was a significant decrease in GLS and myocardial GLSR from the first to the second and the third examination (all P‐values <0.004, Table 2 and Figure 2A). GCSR decreased significantly from the first to the third scan (P=0.0039, Table 2). GCS, GRS, and GRSR did not change significantly between examinations. RVEF decreased significantly from the first to the third and from the second to the third examination (P‐values <0.002, Table 3 and Figure 2B); the slight decrease in RVEF between the first and second scan did not reach statistical significance.

Table 2

Results From CMR Feature Tracking

Parameters	1st CMR (n=108)	2nd CMR (n=108)	3rd CMR (n=41)
GLS (%)	−18.8 [−20.5;−16.5]	−16.9 [−19.3;−14.7]	−15.8 [−18.7;−12.6]
GCS (%)	−23.0 [−25.4;−19.7]	−22.1 [−24.8;−19.3]	−21.3 [−24.7;−18.5]
GLSR (1/s)	−1.2 [−1.4;−1.0]	−1.0 [−1.2;−0.9]	−0.9 [−1.1;−0.8]
GCSR (1/s)	−1.3 [−1.5;−1.1]	−1.2 [−1.4;−1.0]	−1.1 [−1.3;−1.0]
GRS (%)	58.5 [47.9;72.9]	57.1 [45.0;71.2]	53.8 [42.3;68.2]
GRSR (1/s)	2.7 [2.3;3.1]	2.6 [2.2;3.2]	2.4 [2.0;2.5]
Longitudinal velocity, cm/s	1.7 [1.4;2.1]	1.8 [1.6;2.3]	2.0 [1.7;2.5]
Longitudinal displacement, mm	1.6 [1.1;2.2]	1.9 [1.3;2.4]	1.8 [1.3;2.9]
Rotation velocity, deg/s	37.2 [31.5;45.6]	33.9 [28.5;41.6]	30.2 [24.5;39.0]
Rotation displacement, deg	3.7 [2.4;5.6]	4.3 [2.8;5.7]	4.0 [2.4;5.6]
Radial velocity, cm/s	2.5 [2.2;2.9]	2.7 [2.4;3.1]	2.8 [2.5;3.3]
Radial displacement, mm	5.3 [4.9;6.0]	5.8 [5.1;6.3]	6.2 [5.0;6.8]

Parameters are myocardial arithmetic means; Data are presented as median [first and third quartile]. CMR indicates cardiovascular magnetic resonance; GCS, global circumferential strain; GCSR, global circumferential strain rate; GLS, global longitudinal strain; GLSR, global longitudinal strain rate; GRS, global radial strain; and GRSR, global radial strain rate.

Figure 2

A and B, Box and whisker plots illustrating results for myocardial GLS (A) and RVEF (B) across the 3 CMR scans.

*n=108 at first first and second scan, n=41 at third scan; time points (T). CMR indicates cardiovascular magnetic resonance; GLS, global longitudinal strain; and RVEF, right ventricular ejection fraction.

Table 3

Results From CMR volumetry

Parameters	1st CMR (n=108)	2nd CMR (n=108)	3rd CMR (n=41)
RVEDV, mL	65.8 [53.9;79.8]	104.1 [89.4;132.1]	171.6 [145.6;198.8]
RVESV, mL	29.8 [22.7;37.8]	49.4 [39.4;64.0]	83.0 [63.8;105.7]
RVSV, mL	33.8 [30.3;40.9]	54.4 [47.3;66.9]	80.7 [73.1;92.9]
RVEF, %	54.4 [49.7;58.7]	53.4 [48.9;58.6]	50.2 [45.0;55.9]
RVEDVi, mL/m²	88.3 [74.1;104.9]	95.9 [82.8;113.5]	107.9 [94.2;132.5]
RVESVi, mL/m²	40.8 [30.2;50.7]	44.0 [34.8;57.3]	52.0 [43.5;71.1]
RVSVi, mL/m²	47.7 [42.7;56.1]	52.7 [43.2;59.9]	53.3 [46.1;60.4]
RVMM, g	34.0 [28.3;40.4]	56.3 [43.9;73.3]	76.1 [63.0;94.4]
RVMMi, g/m²	47.3 [39.0;54.9]	50.3 [42.7;59.8]	49.6 [44.2;58.4]
Cardiac index, L/min per m²	4.1 [3.2;4.6]	4.0 [3.3;4.8]	4.2 [3.5;4.9]

Data are presented as median [first and third quartile]. CMR indicates cardiovascular magnetic resonance; EDV, end‐diastolic volume; EF, ejection fraction; ESV, end‐systolic volume; MM, myocardial mass; RV, right ventricle; RVEDVi, right ventricular end‐diastolic volume index; RVEF, right ventricular ejection fraction; RVESVi, right ventricular end‐systolic volume index; RVMMi, right ventricular myocardial mass index; RVSVi, right ventricular stroke volume index; and SV, stroke volume.

Indexed RV end‐diastolic and end‐systolic volumes increased significantly from the first examination to the second and to the third examination (Table 3). Median right ventricular myocardial mass index (RVMMi) increased significantly between the first and second scan but not significantly between the first and third examination. GCS and GLS correlated negatively with RVEF (GCS: Spearman’s r=−0.57 to −0.69; GLS: −0.32 to −0.61; all P‐values <0.0008; Figure 3).

Figure 3

Spearman’s rank correlation exemplarily for myocardial GLS and RVEF at second scan.

GLS indicates, global longitudinal strain; and RVEF, right ventricular ejection fraction.

There were no significant changes for the motion parameters (displacement and velocity) across the 3 examinations (Table 2).

Factor Analysis

Factor analysis was performed to find the CMR parameters best suited to detect serial RV changes in HLHS patients. For this analysis, 99 pairs of complete first and second CMR studies were used.

With 3 factors 54% of the overall variance could be explained. The first domain contained RVEF, and feature tracking parameters derived from the short axis. The second domain included RV volumes and the third domain comprised feature tracking parameters from long‐axis views. Results are shown in Supplemental Material (Table S2).

Reproducibility of RV Measurements

Good to excellent inter‐observer agreement was shown for right ventricular volumetric parameters. The intraclass correlation coefficient ranged from 0.90 to 0.99.

Moderate to excellent inter‐ and intra‐observer agreement was found for myocardial strain parameters. ICC coefficients for inter‐observer agreement were 0.34 in GRS, 0.79 in GCS and 0.84 in GLS. Intra‐observer agreement ranged from 0.66 in GRS to 0.90 in GLS and 0.91 in GCS, respectively.

Discussion

This is the largest single center CMR study that assessed serial changes in RV deformation and function in patients with HLHS after Fontan completion. The results demonstrate that myocardial GLS and GLSR decrease earlier than RVEF suggesting that CMR feature tracking might be superior to volume based functional parameters such as RVEF in detecting RV dysfunction early.

Factor analysis demonstrated that serial RV changes over time can be best described by a combination of CMR volumetry and 2D‐CMR‐FT parameters.

Changes in RV Deformation and Motion, RV Function, and RVMMi

Only few studies have assessed longitudinal functional changes in single RV patients. In a recent study, we were able to show that in patients with HLHS RVEF remains largely unchanged over a period of 10 years after TCPC completions but that indexed RV volumes significantly increase in patients older than 10 years of age.

The present study additionally demonstrates that a decrease in RV strain occurs before a reduction in RVEF becomes apparent. Similar findings were reported by Meyer and colleagues who found a decrease in ventricular strain in Fontan patients with preserved EF over a time period of 2 years using CMR whereas Latus and colleagues did not detect significant changes in CMR‐derived myocardial strain and strain rate values. , However, different to this study, both studies included mixed cohorts of Fontan patients with only a small number of patients with HLHS and the follow‐up time was shorter. , An echocardiographic study from our institution reported a reduction in global strain rate 1.6–5.1 years after TCPC completion in patients with HLHS, but global strain did not change.

Studies in biventricular hearts have shown that GLS of the systemic ventricle is an independent and powerful predictor of outcome, , not well correlated with EF, and that strain better than EF reflects the systolic function of adult patients with heart disease and preserved EF. The lower sensitivity of EF to the reduction of longitudinal shortening is explained by possibly compensatory GCS predominance, a phenomenon, which was also observed in HLHS studies during staged palliation. , Ruotsalainen and colleagues demonstrated a significant correlation between vector‐velocity‐imaging derived strain and strain rate parameters in a horizontal long‐axis view and MRI derived EF between the palliation stages but without being a significant predictor of EF. The described independency of GLS is consistent with our findings. Myocardial GLS decreased although RVEF was still largely preserved and a correlation between RVEF and GLS was only observed (Spearman’s r=−0.61), when RVEF deteriorated at the time of the third CMR examination. The non‐significant GCS changes across the CMR examinations might reflect a shift to circumferential contraction patterns in patients with HLHS.

Although few echocardiographic studies suggest that the size of the left ventricle has no impact on RV function, , more recent studies point towards the fact that LV size matters. , , , Petko and colleagues demonstrated that patients with mitral and aortic atresia with small left ventricles had better septal strain values compared with HLHS subtypes with larger left ventricles using speckle tracking echocardiography. Similar findings were described by 2 other groups by echocardiography and CMR, suggesting that a larger left ventricle particularly impacts septal deformation. Others suggested that apical bulging of the RV is associated with impaired RV strain values and is more commonly found in patients with HLHS with hypertrophied hypoplastic left ventricle. The present study did not investigate the effect on different HLHS subtypes; however, future longitudinal studies should focus on this aspect. Beside geometric properties among HLHS subtypes, treatment strategies, hemodynamically relevant ventricular loading, heart rate, , and dyssynchrony in contraction caused by activation delay can affect myocardial deformation.

Besides that, RV myocardial mass is in a constant process of change in patients with HLHS due to alternating cardiac conditions during staged palliation and increases significantly during midterm Fontan follow‐up. The present findings might be explained by a responsive RV hypertrophy, that creates favorable conditions for ventricular dysfunction by failing to adapt to a more circumferential contraction pattern that would be typical for a left ventricle. Longitudinal shortening accounts for nearly 80% of global RV function in normal hearts and the interventricular septum is responsible for 80% of the RV performance. A relative increase in subepicardial fiber mass after TCPC completion might strengthen circumferential RV free wall shortening to sustain RV performance in patients with HLHS whereas RV longitudinal function declines.

In contrast to the only slight reduction in RVEF the first 2 CMR examinations, myocardial GLS and GLSR decrease significantly during serial follow up in patients with HLHS after Fontan circulation. Thus, caution is advised when assessing RVEF in isolation.

That there is an increase of indexed RV volumes during longitudinal follow in older patients with HLHS (>10 years) was already recently shown by this group. Potential reasons might be a higher degree of tricuspid regurgitation but also cardiovascular and metabolic changes during puberty with a raise in blood pressure, increased RV afterload and insulin resistance should be taken into account. ,

Certainly, volumetry and analysis of long‐axis function using CMR‐FT seem to be suitable to detect RV dysfunction. However, further studies are warranted to develop protocols and functional analysis tools best suited to assess the abnormally loaded systemic RV in HLHS.

An exploratory factor analysis of 35 parameters between the first and second CMR examination demonstrated high factor loading (>0.5) components revealing the most representative variables which should be considered in particular for RV assessment (Table S2, Figure 4). The results highlight a benefit to combine CMR volumetry and 2D‐CMR‐FT parameters in patients with HLHS.

Figure 4

Factor analysis between first and second scan: heatmap of factor loadings, based on relative deviations (variable at second scan−variable at first scan)×100/(variable at first scan).

a, endocardial value; age difference (between CMR scans); b, myocardial value; BSA indicates body surface area; D, displacement; EDV, end‐diastolic volume; EF, ejection fraction; ESV, end‐systolic volume; i, index; MM, myocardial mass; RV, right ventricle; S, strain; SR, strain rate; SV, stroke volume; and V, velocity.

Future confirmatory factor analyses in HLHS studies could be supplemented with additional markers to verify and improve the accuracy of the domains found.

Reproducibility Assessment

Interobserver agreement for strain values were lower compared with volumetric parameters, especially for GRS. CMR volumetry is performed from a complete short‐axis stack covering the right ventricle from the base to the apex whereas 2D‐CMR‐FT only uses 3 selected slices from the same stack. Consequently, differences in tracing may have a greater impact on 2D‐CMR‐FT results and might affect data reproducibility. Nevertheless, intra‐observer and inter‐observer agreements in 2D‐CMR‐FT were good for GLS and good to excellent for GCS.

Limitations

This study assessed a large cohort of patients with HLHS in a comparable way but the number of patients with 3 examinations was smaller than those with 2 examinations. Due to the retrospective nature of the study, few CMR data sets were incomplete. In addition, comparisons with echocardiographic measurements were not performed.

The software used for 2D‐CMR‐FT analysis is designed for hearts with a normal anatomy. , Myocardial structures are automatically assigned to particular myocardial segments which was more difficult for the single right ventricle in patients with HLHS with often coarse trabeculations. Therefore, adaption of manual tracing after automatic segmentation was performed. GLS was assessed from an axial long‐axis cine view only, therefore some segments are missing.

General disadvantages of 2D‐CMR‐FT include a susceptibility to temporal resolution problems and beat‐to‐beat differences in cine image quality and stability. In addition, accurate tracking of features can fail due to through‐plane motion, especially in long‐axis views. ,

Furthermore, reduced reproducibility of GRS is a known weakness of 2D‐CMR‐FT but it does not impact reproducibility of GCS which corresponds to our results.

The influence of comorbidities, medical therapy or sedation on RV measurements were not statistically investigated.

Conclusions

The results of this study suggest that GLS values decrease earlier than RVEF, indicating that 2D‐CMR‐FT should routinely be performed in addition to ventricular volumetry to detect early alterations in myocardial deformation and function. Factor analysis might contribute to a rational reduction of data dimension.

Sources of Funding

None.

Disclosures

None.

Tables S1–S2

Click here for additional data file.