Novel Echocardiographic Metrics Predict Tricuspid Insufficiency in Pediatric Ebstein Anomaly

,


Introduction
Te success of most tricuspid valve (TV) repair techniques for Ebstein anomaly (EA) rely on a mobile and large anterior leafet, especially for pediatric patients who have more fragile valves. For instance, Danielson's technique [1] uses the anterior leafet to construct a monocusp valve; Carpentier's technique [2] and cone reconstruction (CR) [3] detach the anterior leafet to achieve extensive mobilization of the leafet tissue. However, the conventional classifcation system, the Carpentier type, assesses the anterior leafet morphology qualitatively and relies too much on the evaluator's experience. Limited studies have quantitatively analyzed the anterior leafet morphology of EA and investigated the association between morphology and surgical outcomes. Although Hughes et al. [4] introduced the Ebstein valve rotation angle on cardiovascular magnetic resonance (CMR) to guide the feasibility of CR, they did not analyze the anterior leafet morphology separately. In addition, the potential need for general anesthesia in young children makes it challenging to perform CMR routinely in pediatric patients with EA [5]. Terefore, in this study, we sought to assess the anterior leafet morphology using novel 2D echocardiographic metrics quantitatively and examine the association of the novel metrics with surgical outcomes in pediatric patients with EA.

Study Design and Patient Selection.
Te study was approved by the Institutional Review Board of Shanghai Children's Medical Center on June 9, 2022 (SCMCIRBK-2022076-1), with a waiver of informed consent due to the retrospective nature. Demographic and perioperative data were collected from the electronic databases, and cardiac imaging and clinical examination details were collected from the outpatient medical records. Consecutive pediatric patients with EA undergoing CR at our center between January 2010 and December 2021 were reviewed for inclusion. Patients whose preoperative and postoperative transthoracic echocardiographic images were unavailable for remeasurement and patients who underwent palliative procedures, non-CR repair techniques, or previous TV surgery at other institutions were excluded from this study. CR should generally be avoided during the early neonatal period [6]. Terefore, patients in this cohort were relatively older. Indications for operation were conventional and included one or more of the following: symptoms of cyanosis, exercise intolerance, atrial tachyarrhythmia, heart failure, or asymptomatic patients with progressive right ventricle (RV) dilation or dysfunction.

Surgical
Technique. CR was performed according to the previous description [3]. Te frst step was to detach the anterior and posterior tricuspid leafets from their anomalous attachments in the RV. Ten, the anterior leafet was carefully mobilized by freeing its connections to the interventricular septum. Finally, the free edge of the posterior leafet was rotated clockwise and sutured to the recruited septal leafet edge, forming a new TV resembling a cone. Atrialized RV was reduced by longitudinal plication at the junction of the anterior and posterior leafets. In this cohort, bidirectional cavopulmonary shunt (BCPS) was performed intraoperatively due to hemodynamic instability after separation from cardiopulmonary bypass, and specifc indications for BCPS in our center have been previously described [7,8].

Echocardiographic
Measurements. 2D echocardiographic examination was performed using the Philips iE33 ultrasound machine (Philips, Andover, MA, USA) equipped with an X5-1 matrix-array transducer. All preoperative 2D transthoracic echocardiographic images were reviewed by two cardiologists (Lijun Chen and Yuqi Zhang). In order to avoid the bias caused by the subjective factors, the cardiologists were blinded to the results of postoperative recurrent TR. Te anterior leafet morphology was analyzed in the apical four-chamber view of preoperative echocardiographic images. Te angle α between the basal part of the anterior leafet and the tricuspid annulus plane was measured on each frame during the cardiac cycle. Te anterior leafet mobility (ALM) was defned as the maximum value minus minimum value of α (Figure 1(a)). Te anterior leafet length (ALL) was measured on the frame with the maximum value of α, and it was defned as the distance from the basal part to the apical part of the anterior leafet (Figure 1(b)). ALL was indexed to body surface area and referred to as ALL-I. Conventional echocardiographic parameters including the Carpentier type, GOSH score, right ventricle fractional area change (RV FAC), left ventricle ejection fraction, the grade of tricuspid regurgitation (TR), and the size of the TV annulus were also measured. Te severity of TR was assessed according to vena contracta width (VCW) [9]. TR was graded as mild (VCW < 0.3), mild-moderate (VCW � 0.3), moderate (0.3 < VCW ≤ 0.69), moderatesevere (VCW � 0.7), and severe (VCW > 0.7) in our center [8].

Follow-Up and
Outcomes. All patients were required to undergo follow-up examinations including echocardiography and cardiogram every 3 months postoperatively. If a scheduled visit was missed, we would interview the patients by telephone for confrmation of their condition. Follow-up in local hospitals was advised for patients living far from our center, and echocardiographic reports were sent to us through a mobile phone app (WeChat). If moderate or greater TR was suspected, patients were required to come to our center for further evaluation. Te primary endpoint of the study was the presence of signifcant recurrent TR, defned as TR of moderate-severe or greater grade within one year after CR.

Statistical Analysis.
Categoric data were presented as counts and percentages, and continuous data were presented as means ± standard deviations or medians and interquartile ranges (IQR) as appropriate. Correlation coefcients are derived using the Pearson method for continuous variables and Spearman rank-order correlation for ordinal variables. Univariable logistic regression analyses were performed to determine the risk factors for recurrent TR. Multivariable analysis was not performed given the limited sample size. To analyze the predictive power of the two novel echocardiographic metrics for recurrent TR, receiver operating characteristic (ROC) curves were performed and the area under the curve (AUC) together with a 95% confdence interval (CI) was calculated. Te calculation formula of the combined ALM and ALL-I values was shown as follows: −0.035 − 0.103 × ALM + 0.268 × ALL-I. Te coefcients and constant term were derived from multivariable logistic regression analysis which incorporated both ALM and ALL-I. Te optimal cutof values of ALM, ALL-I, and ALM + ALL-I to predict recurrent TR were evaluated using the maximum value of the Youden index. Time-to-event analysis was studied using Kaplan−Meier estimates, and comparisons were performed using the log-rank test. All statistical analyses were performed using R version 4.1.1 (R Foundation for Statistical Computing, Vienna, Austria), and a p value of <0.05 was considered to be signifcant.

Patients' Characteristics.
In total, 74 pediatric patients with EA were ft for inclusion. Baseline characteristics of the entire cohort are presented in Table 1. Median age at operation was 3.3 (IQR, 1.9-7.1) years, and 10 patients were aged less than 1 year. Most of the cohorts were classifed as Carpentier type A (63.5%). GOSH scores < 1 were reported in 80% of the patients. Te mean of ALM and ALL-I was 43.8 ± 12.4 and 6.4 ± 2.9 cm/m 2 , respectively, and there was no statistical diference in ALM and ALL-I among patients with diferent grades of TR (p � 0.40 and 0.176, respectively).
Correlations of ALM and ALL-I with conventional echocardiographic parameters are demonstrated in Table 2. Both ALM and ALL-I were signifcantly related to the Carpentier type and GOSH score but were not related to preoperative TR severity. In addition, ALM was also associated with RV FAC.

In-Hospital and Follow-Up
Results. Operative and early postoperative data are summarized in Table 3 Figure 3). ROC curve analyses demonstrated that the optimal cutof values of ALM, ALL-I, and ALM + ALL-I for predicting recurrent TR were 39.5°(sensitivity, 69.2%; specifcity, 77.8%), 5.6 cm/m 2 (sensitivity, 50.8%; specifcity, 100%), and −2.16 (sensitivity, 75.2%; specifcity, 88.9%), respectively. When an ALM of >39.5°was used as the cutof value to defne patients with a fexible anterior leafet, there were 47 patients with an ALM of >39.5°(fexible group) and 27 patients without an ALM of >39.5°(infexible group). Recurrent TR developed in 2 patients in the fexible group and in 7 patients in the infexible group (log-rank p � 0.006, Figure 4(a)). When an ALL-I of >5.6 cm/m 2 was used as the cutof value to defne patients with a redundant anterior leafet, there were 41 patients with an ALL-I of >5.6 cm/m 2 (redundant group) and 33 patients without an ALL-I of >5.6 cm/m 2 (nonredundant group). Recurrent TR developed in 9 patients in the redundant group and in no patients in the nonredundant group (log-rank p � 0.005, Figure 4(b)).
When an ALM + ALL-I of > −2.16 was used as the cutof value to defne patients with a dysfunctional anterior leafet, there were 24 patients with an ALM + ALL-I of > −2.16 (dysfunctional group) and 50 patients without an ALM + ALL-I of > −2.16 (functional group). Recurrent TR developed in 8 patients in the dysfunctional group and in 1 patient in the functional group (log-rank p < 0.001, Figure 4(c)).

Discussion
To the best of our knowledge, this is the frst study that investigated the anterior leafet morphology on 2D echocardiography in pediatric Ebstein anomaly (pEA). Te main fndings of our study were as follows: (1) ALM and ALL-I were well associated with the Carpentier type and GOSH score, and ALM was also correlated to RV FAC; (2) ALM and ALL-I were risk factors for recurrent moderate-severe or greater TR after CR in pEA and were more reliable metrics to predict recurrent TR compared with the Carpentier type, GOSH score, and preoperative TR severity.
In most TV repair techniques involving EA, the anterior leafet is the fundamental tissue for the reconstruction of a competent TV. Patients with restricted or insufcient anterior leafets may need modifed techniques, such as valve enlargement [10] and the Spinnaker repair [11]. In addition, the anterior leafet morphology is an important criterion for grading the severity of EA according to the Carpentier type. Terefore, the anterior leafet morphology may theoretically have an impact on surgical strategy and outcomes. Hughes et al. [4] found that 3 of the 4 patients with surgical dehiscence in their cohort exhibited thickened, rolled edges of the anterior leafet and were noted to have numerous leafet attachments to the RV anterior wall and poor leafet mobility. None of the 16 patients without dehiscence exhibited thickened, rolled anterior leafet margins. Similar circumstances have been found in mitral valve disease. Brescia et al. [12] assessed mitral anterior leafet mobility and calcifcation to determine mitral repair or replacement in patients with rheumatic heart disease; Gupta et al. [13] found that mitral anterior leafet length was a strong predictor of mitral valve repairability, and a value of 26 mm or more was associated with successful repair. Under this context, we remeasured the tricuspid anterior leafet mobility and length on 2D echocardiography in a pediatric cohort with EA, and our result showed that measuring ALM and ALL might help identify Ebstein valves that were at risk of recurrent TR after CR in pEA.
Compared with the Carpentier type and GOSH score, ALM and ALL-I had better performance for predicting recurrent TR within one year after CR and the combinations of ALM and ALL-I increased the predictive performance compared with ALM or ALL-I alone. Te possible explanations for this could include the following: (1) varying degrees of failed delamination of TV leafets are the underlying pathology in EA. Specifc to the anterior leafet, it manifests a wide spectrum of morphologic variability, ranging from normal formation to sail-like redundance, to muscularization [14]; the failure of delamination also results in adherence of the anterior leafet to the underlying myocardium, and the leafet free edge shows varying degrees of limited mobility [15]. Terefore, in our cohort, where almost all patients' conditions are Carpentier types A or B, a sail-like redundant and less mobile anterior leafet may suggest less delamination of the anterior leafet and severer attachment of the leafet body to RV walls, so the length and mobility of the anterior leafet in EA may help assess the disease severity. Tis hypothesis was based on our results that ALM and ALL-I correlated well with the two conventional classifcation systems: the Carpentier type and GOSH score, and ALM also correlated with RV FAV; (2) the Carpentier type stratifes EA qualitatively according to severity of TV and RV, thus relying too much on the evaluator's experience. Te GOSH score [16] is more frequently used for neonatal patients with EA and do not involve the function of TV. Both of them have limitations and cannot accurately evaluate EA patients who manifest a wide variety in clinical presentation and anatomic severity. In contrast, ALM and ALL-I stratify the vast variability in the anterior leafet morphology quantitatively and can help discriminate the morphological diference among patients with the same Carpentier type or grade of the GOSH score.
With Geerdink et al., conficting results have been reported about whether severe TR is associated with the prognosis in pediatric patients [17] and it was found that severe TR was a predictor of mortality in a multicenter cohort of 168 pediatric patients, whereas Prota et al. [18] concluded from a cohort of 50 pediatric patients that severe TR had no prognostic value in cardiac adverse events. Our study showed that preoperative TR severity was not a risk factor for recurrent TR. Some may question whether these negative results were due to the small number of enrolled patients; however, we should notice that quantitative parameters including RV FAC and right atrium peak atrial longitudinal strain were included in Prota's study [18]. In this study, ALM and ALL-I were associated with prognosis signifcantly, and our correlation analysis showed that ALM and ALL-I were not associated with preoperative TR severity. Tis may refect the better competency of ALM and ALL-I as novel echocardiographic metrics to optimize evaluation in pEA and predict surgical outcomes. Although prospective testing is needed, our fndings can help surgeons better comprehend the morphological heterogeneity of the anterior leafet in pEA and assist in surgical decision-making. First, surgeons can assess the anterior leafet morphology on the basis of ALM and ALL-I preoperatively. Second, repair modifcations such as leafet augmentation may be considered if preoperative ALM and ALL-I suggest a high risk of recurrent TR after CR.
Tis study has several limitations. First, this cohort was a small number of pediatric patients with EA, and almost all of the patients' conditions were Carpentier types A or B. Further investigation of a larger patient population including various age subgroups and a wider spectrum of anterior leafet morphology is needed to defnitively confrm our results. Second, this is a retrospective, single-center study. Further multicenter or registry study with a cohort of normal patients as a comparison is undoubtedly necessary and useful. Tird, the application of ALM and ALL-I in surgical decision-making is only a reasonable speculation because we did not analyze the association of the two metrics with repair modifcations. Further prospective study with an appropriate design may help provide more scientifc conclusions. Finally, the goal of CR is to create a cone-shaped TV from all available leafet tissue including the posterior and septal tissue. However, the posterior and septal leafets are hypoplastic and displaced downward in EA, which make it difcult to introduce appropriate metrics from 2D echocardiography to assess the morphology. A future cohort with 3D echocardiography may be helpful to provide more precise assessment of all leafets.

Conclusions
Easily-acquired ALM and ALL-I from 2D echocardiography can help optimize evaluation of the anterior leafet morphology and predict recurrent TR after CR in pEA.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request.

Ethical Approval
Te study was approved by the Institutional Review Board of Shanghai Children's Medical Center on June 9, 2022 (SCMCIRBK-2022076-1), with a waiver of informed consent due to the retrospective nature.

Disclosure
Wei Liu and Chen Wen shared co-frst authorship.

Conflicts of Interest
Te authors declare that they have no conficts of interest.

Authors' Contributions
Wei Liu was responsible for data analysis/interpretation, drafting the article, critical revision of the article, and approval of the article. Chen Wen was responsible for drafting the article, critical revision of the article, approval of the article, and data collection. Jin Shentu was responsible for data analysis/interpretation, critical revision of the article, and approval of the article. Yuqi Zhang was responsible for data analysis/interpretation, critical revision of the article, approval of the article, and statistics. Zhongqun Zhu was responsible for critical revision of the article, approval of the article, and statistics. Lijun Chen was responsible for concept/design, drafting the article, critical revision of the article, approval of the article, and funding security. Huiwen Chen was responsible for concept/design, drafting the article, critical revision of the article, approval of the article, and funding security. Wei Liu and Chen Wen contributed equally to this work and shared co-frst authorship.