Introduction

Tetralogy of Fallot (ToF) is the most common cyanotic congenital cardiac disease, affecting around 5 of 10 000 live newborns [1,2]. Tetralogy of Fallot total correction (TFTC) comprises ventricular septal defect closure and right ventricular (RV) muscle shave with or without patch enlargement to relieve the obstruction of the right ventricular outflow tract (RVOT). Postoperative complications include chronic volume overload, RVOT obstruction, ongoing pulmonary regurgitation and stenosis, RV extension, and tricuspid regurgitation. A previous long-term follow-up study also identified aortic root dilatation, left ventricular (LV) dysfunction, and arrhythmias [3].

Myocardial wall motion analysis is crucial in determining ventricular contractile activity [4]. Myocardial strain is defined as the ratio of length change from the resting state to the contracting state of the cardiac muscle [5] In contrast to ejection fraction (EF), measuring global longitudinal strain (GLS), global circumferential strain (GCS), or global radial strain (GRS) directions allows for a more comprehensive and detailed examination of the spatial components of contractility [5]. Quantifying myocardial deformation has provided additional insight into cardiac function in a range of subclinical cardiac diseases [4].

Time is of the essence for pulmonary valve replacement (PVR) to prevent RV and LV dysfunction [3,6]. RV function and geometry assessments are somehow challenging, with atrioventricular, ventriculoatrial, and interventricular interactions thus far suggested for predicting RV dysfunction [6]. Nevertheless, the mechanism and predictors of RV remodeling in TFTC remain poorly understood. Reduced GLS and GCS strain values of both LV and RV in repaired ToF were associated with adverse events [7].

Ventriculoarterial coupling measures the ventricle’s response to its afterload [8]. In patients with pulmonary hypertension, the ratio of stroke volume(SV) to end systolic volume(ESV) has been reported to be a more sensitive volume measure of ventriculoarterial coupling ratio than ventricular EF to detect alterations in function at the earliest stages of the disease. It also has been found as a prognostic ratio [9,10].

Drawing upon cardiac magnetic resonance imaging (CMR) indices, in the present study, we sought to highlight the association between CMR indices and the midterm outcomes of patients with TFTC. We also aimed to assess changes in the follow-up parameters and compare the differences between patients with PVR and those without it.

Methods

The present longitudinal study identified and recruited all patients with a history of TFTC referred to Rajaie Cardiovascular Medical and Research Center for CMR from March 21, 2007, through March 21, 2017. The inclusion criteria were a history of TFTC with high-quality cine images and at least two CMR studies conducted at the aforementioned center. Based on previous studies, patients with an interval of less than a year between ToF repair and CMR or an interval of less than three months between the two CMR studies were excluded. The baseline characteristics of the patients, changes in their CMR induces, and adverse events during the postoperative follow-up period were recorded. The primary outcomes were PVR, sustained ventricular tachyarrhythmia requiring cardiovascular implantable electronic devices and death. The secondary outcomes were changes in RV, LV systolic function and volumes, pulmonary regurgitation. The ventricular end-diastolic volume index (EDVI) and EF changes were assessed in a subgroup of patients with PVR and compared with those without PVR.

Results

Fifty-six patients with TFTC at a mean age of 15.23 ± 11.66 years (range: 1–41) were observed during the present study. The mean duration of the whole follow-up (the time between surgery and the last follow-up) and the mean duration between the two follow-ups (between the two CMRs) were 159.92 ± 109.34 months and 28.29 ± 17.66 months, respectively. The baseline characteristics and the prevalence of the outcomes in the studied patients are summarized in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. The baseline characteristics and outcomes in the studied patients.

https://doi.org/10.1371/journal.pone.0308362.t001

Regarding the outcomes, one patient (2%) experienced ventricular tachyarrhythmia, and two (3.6%) received a permanent pacemaker. PVR was performed on 18 patients (32%), of whom 13 (72%) received a metallic valve, and 5 (28%) received a biological valve. The mean duration from TFTC to PVR was 12.66 ± 7.58 years, and the mean age at PVR was 15.68 ± 9.01 years. While assessing the outcomes in the CMRs, we detected reduced RVEF lower than 50% in 49 patients (87.5%) in both follow-ups. However, the prevalence of LV failure rose from 11 (19.6%) to 18 (32.1%) in the second follow-up. Mortality did not occur in any patient until the telephone follow-up. No patient showed LV myocardial fibrosis.

The characteristics of the patients with and without PVR are compared in Table 2. This comparison demonstrated a significantly higher RVESVI (P <0.05 in both follow-ups), RVEDVI (P <0.05), and pulmonary pressure gradient (P <0.01) in the patients with PVR than in those without PVR. Compared with the patients without PVR, those with PVR had a notably lower RVEF (P <0.01), right-sided coupling ratio (P <0.05), and weight (P <0.05). Other differences between the two groups were not statistically significant. The incidence of tachyarrhythmia and permanent pacemaker insertion was too small to perform any subgroup analysis.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Comparisons of the indices in the patients with and without PVR.

https://doi.org/10.1371/journal.pone.0308362.t002

A comparison between the two CMRs demonstrated a considerable increase in RV and LVSV in the second follow-up (P <0.01 and P <0.05, respectively).

The correlations between the variables of function, strain, and coupling ratio are summarized in Table 3. LVEF was significantly correlated with LVGCS (P <0.01 in both follow-ups), LVGRS (P <0.05 in the first follow-up), and LVGLS (P <0.01 in the second follow-up and P = 0.052 in the first follow-up). LVSV was negatively correlated with LVGLS and LVGCS in the second follow-up (P <0.05), although the mentioned correlation was not observed in the first follow-up. RVEF and RVSV negatively correlated with RVGLS in the second follow-up (P <0.05). The right-sided coupling ratio correlated with RVEF (P <0.01) and RVGRS (P <0.05) in the first follow-up and with RVGCS (P <0.05) in the second follow-up. RVOT fibrosis did not correlate with any of the variables of function, strain, or coupling ratio.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Correlations between the variables of function, strain, and ventriculoarterial coupling.

https://doi.org/10.1371/journal.pone.0308362.t003

Discussion

Our longitudinal study found a meaningful correlation between LV and RV volumes and the global strain values of two CMR evaluations. LVGCS was correlated to LVEF, RVEF correlated with RV coupling and RVGLS, and RV coupling correlated with RVGRS and RVGCS. Consistent with an earlier study [13], patients who had increased RVESVI and RVEDVI were more likely to require PVR.

GCS, GLS, and GRS have been employed before to evaluate the function of the ventricles [6,14]. Still, their predictive value in repaired ToF remains uncertain [15]. There is also disagreement regarding the relationship between EF and strain patterns [6]. However, the strain could indicate myocardial dysfunction even with a normal EF [16]. Our study confirmed the correlation between all LV strain parameters and LVEF. LVGCS is the most relevant parameter to LVEF, as the ejection of a sizable amount of LVSV results from torsional and circumferential shape alterations [17]. Contrary to the other strain parameters, the correlation between LVEF and LVGCS was present in both CMRs. Thus, using LVGCS over LVEF in early myocardial dysfunction detection seems preferable.

While a prior investigation found GCS the most reproducible strain measurement on CMR feature tracking [18], we support the findings of Menting et al. [19] and Therrien et al. [20] in so far as they contended that GLS was a significant reflector of the overall function of the RV inflow and apical parts. In contrast to Orwat et al. [21] and Kalaitzidis et al. [22], we found no correlation between RVGCS and RVEF. Possible explanations for this observation include that, under normal conditions, the RV ejects blood generally by longitudinal shortening and that, under increasing RV pressure loading, the RV ejects blood primarily through circumferential contraction [23]. Similar to the findings of Orwat et al. [21], we confirmed that strain parameters offered more information than EF regarding their use in the early detection and prediction of dysfunction and adverse outcomes in patients with repaired ToF.

The parameters of RV function measured by CMR or echocardiography are load-sensitive, meaning they cannot tell us whether the RV is functioning effectively in response to a particular afterload. A higher RV afterload should result in greater contractility to preserve function and coupling [24]. Small ventricular function changes may be more easily detected with the coupling ratio (= SV/ESV) and may be superior to EF for quantifying RV contractile reserve [25]. Therefore, based on the approach of Sanz et al. [26] and supporting studies [25,27], we performed a noninvasive assessment of RV function using the ventriculoarterial coupling ratio. Patients with relatively preserved RV volumes and EFs may develop symptoms, which explains why identifying those with uncoupling may show a group that might profit from therapies before the onset of symptoms or borderline RV dilation and impairment [24].

CMR studies of volume and EF are considered the standard method for determining the risk level and optimal timing of PVR for individuals with chronic PR. The prognostic value of CMR has been extensively studied in the TFTC population. Nevertheless, CMR-based indices solely incorporate the preload with RVEDV and a load-dependent measure of systolic function as EF. These measurements do not consider the impact of abnormal RV afterload, which is present in TFTC patients [28]. Additionally, most TFTC patients exhibit normal or near-normal RV systolic function, signifying that RV systolic dysfunction is a delayed manifestation in the pathophysiological progression of TFTC-associated PR [29]. This statement suggests that a normal RV systolic function may not adequately reveal the RV’s adaptation to a given afterload [29]. In this regard, several studies found that the ventriculoarterial coupling ratio and myocardial strain indicate myocardial changes in TFTC patients despite having a normal or relatively preserved ventricular volumes and EF [16,21,24,25,28,29]. These markers suggest a significantly higher magnitude for coupling ratio and strain parameters.

In-depth, Orwat et al. [21] reported that combining CMR assessments of LV GCS and RV GLS identified a subgroup of TFTC patients with a significantly higher risk of adverse outcomes. They also found that RV GCS was significantly associated with symptomatic deterioration of at least one NYHA class during the follow-up period. Besides, our study uncovered a significant association between RV strain and the ratio of the right ventriculoarterial coupling. Therefore, it seems that RV strain serves as an indicator of PVR. Overall, these findings imply that evaluating strain and ventriculoarterial coupling ratio in cases where the EF is normal could provide valuable insights to make well-informed choices regarding the timely PVR.

Buddhe et al. [30] found a significant correlation between RV coupling and RVEDVI, ESVI, EF, and LVEF. In this regard, Panaioli et al. [29] demonstrated that RV coupling was worse in patients with repaired ToF requiring PVR despite normal RV systolic function. Panaioli and colleagues also concluded that before the manifestation of clinical symptoms and RV systolic dysfunction, RV coupling might be a reliable indicator of a continual deleterious RV reaction to enduring volume overload in repaired ToF. Since our population was slightly larger than this study, further studies are needed to address the controversy.

Similar to the studies of Berganza et al. [15] and Kalaitzidis et al. [22], the present study recruited a young sample at a mean age of 15.23 ± 11.66 years. Although Bokma et al. [13] reported that patients with TFTC at a younger age were more prone to undergo PVR, our analysis did not support this finding. Since the rate of adverse events rises with patient age, the relationships between other variables may reach statistical significance in samples with older patients. Furthermore, even in young individuals with repaired ToF, most physiological parameters may still be generally normal alongside a higher capacity of RV remodeling, highlighting the need for more sensitive diagnostic measures to assess young patients [31,32].

Conclusion

Our study found that, except for the SV of the ventricles, other CMR parameters may not change significantly at a mean follow-up duration of about 28 months. It also suggests that CMR derived strain and ventriculoarterial coupling parameters could be potential personalized-based indicators of PVR requirement in patients with repaired ToF. Further studies are warranted to confirm the findings of the present study.

Supporting information