Introduction
Tetralogy of Fallot, the most prevalent cyanotic CHD, constitutes 5–7% of all CHDs, affecting approximately 0.34 per 1000 live births. Reference Kliegman, Behrman, Jenson and Stanton1,Reference van der Ven, van den Bosch, Bogers and Helbing2 Current 30-year life expectancy for Tetralogy of Fallot patients ranges from 68.5 to 90.5%, but long-term outcomes decline significantly beyond three to four decades post-repair. This decline is often linked to comorbidities, including progressive limitations in physical activity, right ventricular outflow obstruction, pulmonary regurgitation, arrhythmias, and right heart failure. Reference van der Ven, van den Bosch, Bogers and Helbing2,Reference Dennis, Moore, Kotchetkova, Pressley, Cordina and Celermajer3
Controversy surrounds the surgical management of symptomatic neonates with Tetralogy of Fallot, with conflicting evidence on the efficacy of either primary repair or staged repair. A staged approach, involving modified Blalock-Taussig shunt followed by total Tetralogy of Fallot correction, aims to secure pulmonary blood flow, alleviate hypoxaemia, and facilitate pulmonary artery growth. Total Tetralogy of Fallot correction involves septal defect closure, widening the subpulmonary obstruction, and reconstruction. Proponents of primary repair argue for its ability to mitigate prolonged exposure to elevated right ventricular pressure and decreased oxygen saturation, thereby averting adverse impacts on cardiovascular and cerebral function, Reference Bailliard and Anderson4 while the potential advantages of staged approach include reducing the exposure of volatile anaesthetic agents and cardiopulmonary bypass during neonatal period, and better pulmonary artery growth to preserve pulmonary valve. Reference van der Ven, van den Bosch, Bogers and Helbing2
To the best of our knowledge, this meta-analysis was the first to compare outcomes of primary versus staged repair for symptomatic neonatal Tetralogy of Fallot populations. Studies on the surgical management of symptomatic neonates with Tetralogy of Fallot are highly needed to improve the outcomes of the Tetralogy of Fallot patient population and determine the criteria for at-risk populations that may benefit from either primary or staged repair, ultimately reducing post-operative morbidity and increasing the quality of life for this Tetralogy of Fallot population.
Methods
This systematic review and meta-analysis was conducted and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 guidelines. Reference Page, McKenzie and Bossuyt5 This meta-analysis was enlisted in the International Prospective Register of Systematic Reviews (PROSPERO) with ID register number CRD42023491829.
Literature search
The databases Embase, EBSCOHost, Medline, Sage, Science Direct, and Scopus were queried on November 30, 2023, for articles comparing primary repair and staged repair for Tetralogy of Fallot in symptomatic neonates. The articles that were searched were not restricted by language or year of publication. A search method was employed in MEDLINE as follows: ((((((neonates[MeSH Terms]) OR (neonates)) OR (newborn)) OR (newborn[MeSH Terms])) AND ((tetralogy of fallot) OR (tetralogy of fallot[MeSH Terms]))) AND ((surgery) OR (repair) OR (staged repair) OR (primary repair) OR (complete repair) OR (palliative repair) OR (Procedure, Cardiac Surgical) OR (Procedures, Cardiac Surgical) OR (Surgical Procedure, Cardiac) OR (Surgical Procedures, Cardiac) OR (Surgical Procedures, Heart) OR (Cardiac Surgical Procedure) OR (Heart Surgical Procedures) OR (Procedure, Heart Surgical) OR (Procedures, Heart Surgical) OR (Surgical Procedure, Heart) OR (Heart Surgical Procedure))) AND ((((Mortality[MeSH Terms]) ) OR (Mortality) (((Stay Length) OR (Stay Lengths) OR (Hospital Stay) OR (Hospital Stays) OR (Stay, Hospital) OR (Stays, Hospital))) OR (transannular patch) OR ((Outcome, Treatment) OR (Patient-Relevant Outcome) OR (Outcome, Patient-Relevant) OR (Outcomes, Patient-Relevant) OR (Patient Relevant Outcome) OR (Patient-Relevant Outcomes) OR (Clinical Effectiveness) OR (Effectiveness, Clinical) OR (Treatment Effectiveness) OR (Effectiveness, Treatment) OR (Rehabilitation Outcome) OR (Outcome, Rehabilitation) OR (Treatment Efficacy) OR (Efficacy, Treatment) OR (Clinical Efficacy) OR (Efficacy, Clinical)) OR ((Postoperative Complication) OR (Complications, Postoperative) OR (Complication, Postoperative)))). The adjusted keywords were then implemented in search strategies for other databases.
Study selection
A total of four examiners (R.S, S.S, A.R.S., and M.R.B.) conducted the study selection process independently, according to pre-established criteria. Discussion was done to resolve any disagreements that arose among the examiners. Titles and abstracts were utilised to assess the studies in the initial phase of the study selection process; the complete texts were examined in the subsequent phase. Selection criteria were based on the following criteria: (1) Enrolled patients who were neonates with symptomatic Tetralogy of Fallot, (2) compared the clinical outcomes of primary repair and staged repair, and (3) randomised controlled trial or cohort studies. Studies were excluded if enrolled patients were Tetralogy of Fallot with major aortopulmonary collateral arteries undergoing unifocalization, Tetralogy of Fallot with atrioventricular canal, and subjects with double outlet right ventricle. The most appropriate article was selected in cases where redundant publications containing overlapping data were found.
Data extraction
The demographic and study characteristics were extracted separately by two reviewers (R.S. and S.S.). Similarly, two reviewers (M.R.B. and D.G.F.) extracted the data pertaining to the study outcomes. Through discussion, disagreements among reviewers were resolved. The outcomes data were rate of mortality, reintervention, cardiac complication, transannular patch usage, hospital, and ICU length of stay duration. The outcomes data in the staged repair group are cumulative for both the initial palliative procedure and the subsequent complete repair.
Quality assessment
The study’s quality was independently assessed by four reviewers (P.W.B., R.S., S.S., and A.R.S.). Disagreements among the reviewers were settled through discussion. This study used the Newcastle-Ottawa Quality Assessment Scale for cohort studies, which consists of eight questions graded from zero to nine and assesses three elements of assessment: selection, comparability, and outcome. A methodological quality score of six or more was considered high quality, three to five was considered fair, and less than three was considered low quality. Reference Wells, Shea, O’Connell and Peterson6
Statistical analysis
The data were retrieved and recorded utilising Microsoft Excel. Continuous outcomes were presented in mean ± standard deviation (SD) and the mean difference (MD) with 95% confidence interval (CI) was calculated. The odds ratio (OR) with a 95% CI was utilised to assess categorical outcomes expressed as a proportion of the total participants (percentage). To assess the impact of surgical technique and study heterogeneity, the analysed data were graphically represented through forest plots. The data were analysed utilising random (significant heterogeneity) or fixed (no significant heterogeneity) effects. Statistical heterogeneity was assessed utilising the I2 statistic and the χ2 test. The I2 values of 25, 50, and 75% have been suggested to be indicators of low, moderate, and high heterogeneity, respectively. Meta-analysis was conducted utilising RevMan (RevMan) version 5.4.
Results
Study selection
The literature search results, which were obtained from EBSCOhost, MEDLINE, Embase, Scopus, Science Direct, and Sage Journal, are illustrated in Figure 1. In total, 1950 articles were retrieved from the search. Out of the 46 articles that underwent purview extraction and subsequent analysis to meet the research criteria, two were not found in full text. Case reports or correspondence comprised six articles, subjects other than neonates comprised twelve, subjects with Tetralogy of Fallot in conjunction with other cardiac abnormalities comprised seven articles, and eleven articles did not compare primary repair and staged repair. Therefore, a total of eight studies were included in this review. Reference Kanter, Kogon, Kirshbom and Carlock7–Reference Meadows, Bauser-Heaton and Petit14
This systematic review and meta-analysis was conducted in accordance to the preferred reporting items for systematic reviews and meta-analyses flow chart. In total, 1393 articles were identified from six databases (Embase, Scopus, Science Direct, Sage Journal, EBSCOhost, MEDLINE) and eight studies were included involving 4464 subjects who met all inclusion criteria.
Study and demographic characteristic
All studies were published in English with full texts and were all nonrandomised cohort retrospective studies. Six studies were multicentre studies, and two were single-centre. A total of eight studies were performed in America and involved a total of 4464 subjects. Of 4464 subjects, 2025 (45.4%) underwent primary repair, and 2439 (54.6%) underwent staged repair, which consisted of palliative at the neonatal period and complete repair later in life. Palliative strategies consisted of systemic to pulmonary shunts in all studies except Goldstein et al., Meadows et al., and O’Byrne et al., which included transcatheter procedure as palliative strategy in the staged repair group. Reference Kanter, Kogon, Kirshbom and Carlock7–Reference Meadows, Bauser-Heaton and Petit14 All studies were published from 2009 to 2023 with cases collected from 2000 to 2017. The pooled mean patient age from five studies was 8.68 (±7.38) and 8.56 (±6.8) days for primary repair and staged repair, respectively. One study did not report the reintervention rate, duration of hospital, and ICU length of stay. Reference Savla, Faerber and Huang9 Two studies did not report the complication rate outcome, Reference Kanter, Kogon, Kirshbom and Carlock7,Reference Steiner, Tang and Gossett8 while three studies did not report the transannular patch usage rate outcome. Reference Steiner, Tang and Gossett8,Reference Savla, Faerber and Huang9,Reference O’Byrne, Glatz and Yuan-Shung12 The study characteristics are summarised in Table 1.
Study characteristic
Quality assessment
The quality assessment result of studies is shown in Table 2. The eight included studies were subjected to critical analysis and assessment of their biased risk using Newcastle-Ottawa Quality Assessment Scale, a critical appraisal tool for retrospective cohort studies. Five high-quality articles and three fair-quality articles were identified in this study.
Quality assessment of included studies using Newcastle-Ottawa Quality Assessment Scale (NOS) for cohort studies
*, star system in NOS which will be calculated into the total score; -, counted as 0 in NOS. Description for Selection: 1 = Is the case definition adequate? 2 = Representativeness of the cases; 3 = Selection of controls; and 4 = Definition of controls. Description for second element is comparability of cases and controls based on the design or analysis and each study can have up to two stars. Outcome element contains: 1 = Assessment of outcome; 2 = Was follow-up long enough for outcomes to occur? and 3 = Adequacy of follow-up of cohorts.
Risk of mortality
A pooled analysis of 4464 neonates with symptomatic Tetralogy of Fallot from eight studies found no statistically significant differences in mortality risk between neonates who received primary repair and staged repair strategies (OR = 1.26; 95% CI confidence interval = 0.72–2.19; p = 0.42) (Figure 2). Mortality rates displayed substantial statistical heterogeneity with I2 = 65%.
Forest plot for mortality rate, reintervention rate, cardiac complication rate, transannular patch usage rate. Solid squares represent an odd ratio that is proportional to the weights employed in the meta-analysis. Zero effect is denoted by the solid vertical line. The horizontal lines represent a 95% confidence interval (CI). The diamond symbolises the pooled odds ratio, while the lateral edges of the diamond indicate the corresponding CI.
Risk of reintervention
Similarly to the risk of mortality, seven studies included in the forest plot showed no significant difference between both groups (OR = 0.86; 95% CI = 0.67–1.09; p = 0.21) (Figure 2). Low heterogeneity was found for statistical heterogeneity of risk of reintervention (I2 = 15%).
Risk of cardiac complication
The risk of cardiac complication post Tetralogy of Fallot repair in symptomatic neonates was 1.5 times higher in the primary repair group (OR = 1.50; 95% CI = 1.07–2.10; p = 0.02) compared to the staged repair group (Figure 2). The pooled analysis of 3793 subjects showed a substantial heterogeneity (I2 = 70%) for statistical heterogeneity of risk of cardiac complication post Tetralogy of Fallot repair in symptomatic neonates.
Transannular patch usage
Forest plots from five studies that included the transannular patch usage data have shown that the rate of transannular patch usage in Tetralogy of Fallot repair for symptomatic neonates was 2.62 times significantly higher in the primary repair group (95% CI = 2.02–3.40; p < 0.001) compared to staged repair group (Figure 2). There was low statistical heterogeneity of transannular patch usage, which might not be important (I2 = 29%).
Length of stay (LOS) and economic burden
The duration of either hospital length of stay or ICU length of stay was significantly longer in the staged repair group than in the primary repair group. The length of hospital and ICU stay was 11.84 days longer (95% CI = 9.59–14.10; p = <0.001) and 3.06 days longer (95% CI = 1.64–4.47; p < 0.001) in the staged repair group than primary repair group, respectively (Figure 3). However, the statistical heterogeneity was considered moderate (I2 = 59%) for the hospital length of stay and was considered substantial (I2 = 76%) for ICU length of stay.
Forest plot for hospital length of stay and ICU length of stay. The solid squares represent the proportional mean difference concerning the weights employed in the meta-analysis. The vertical line represents no effect. The horizontal lines represent the confidence interval (CI) at a level of 95%. The diamond represents the weighted mean difference, with the corresponding CI denoted by the diamond’s lateral edges. SD = standard deviation.
Although in this study the economic burden was not included in the forest plot, two out of eight included studies in this meta-analysis had presented data regarding the total cost needed for each surgical strategy in neonates with symptomatic Tetralogy of Fallot. Both studies showed that the staged repair group had a higher total cost needed than in primary repair group, $203,917 (±95,693) versus $169,344 (±106,656) in study by O’Bryne et al. Reference Goldstein, Petit and Qureshi11 and $464,618 (±284,750) versus $392,915 (±271,346) in study by Steiner et al. Reference Steiner, Tang and Gossett8
Discussion
The mortality rate for primary repair in patients with Tetralogy of Fallot is generally low, with a surgical mortality rate of 0–2% in infants. Reference Egbe, Mittnacht, Nguyen and Joashi15 In comparison, a study reported an in-hospital mortality rate of 4.1% following Tetralogy of Fallot repair, which occurred more commonly in infants and patients weighing less than 2.5 kg. The long-term survival rates for patients with Tetralogy of Fallot after surgical repair are quite high, with 10-, 20-, and 30-year survival rates of 95.8, 92.7, and 90.5%, respectively. Reference Qureshi, Caldarone and Romano13,Reference Smith, McCracken and Thomas16 The study could not offer precise comparative data on death rates between the two surgical techniques when comparing the mortality rate between primary repair and staged repair. However, it mentioned that staged repair and non-valve-sparing operations show an increased risk of early mortality, and genetic abnormalities increase the risk at≥6 years post-surgery. Reference Smith, McCracken and Thomas16 Meanwhile, other studies linked primary repair with potentially greater initial mortality for patients with high-risk features. Reference Savla, Faerber and Huang9 However, based on the results of this meta-analysis, there was no difference in mortality between the two procedures with substantial statistical heterogeneity in mortality rates. This heterogeneity can result from a number of things, including the distribution and variances in treatment strategies across all centres, the facilities used for the treatments, the operator’s skill, and the changes in the timing of surgical procedures. For a comprehensive comparison of mortality rates between primary repair and staged repair in Tetralogy of Fallot patients, further specific studies directly comparing the two approaches would be needed. Reference Meadows, Bauser-Heaton and Petit14,Reference Persson, Gyllencreutz Castellheim and Dellborg17
The controversy surrounding reintervention rates in primary versus staged Tetralogy of Fallot repair procedures persists in current research. Some studies propose a higher reintervention rate in cohorts employing a palliative or staged repair approach, Reference Hennein, Mosca, Urcelay, Crowley and Bove18 while others suggest an increased reintervention rate in the primary repair group. Reference Tamesberger, Lechner, Mair, Hofer, Sames-Dolzer and Tulzer19,Reference Wilder, Van Arsdell and Benson20 A retrospective study conducted from 2010 to 2022, including 171 patients with complete Tetralogy of Fallot, reported that the risk of reinterventions ranges from 3 to 16%, and several strategies have been proposed to decrease the risk of reinterventions after Tetralogy of Fallot repair. Reference Mashali, Yousef and Elmahrouk21 A retrospective cohort study of symptomatic neonates with Tetralogy of Fallot found no difference in the number of subjects who required unanticipated re-interventions between those who had neonatal primary repair and those who had staged repair. Reference Bailey, Elci, Mascio, Mercer-Rosa and Goldmuntz10 A 26-year experience with surgical management of Tetralogy of Fallot found that early mortality after primary repair has significantly improved, and primary repair in infancy does not increase the risk for reintervention on the right ventricular outflow tract. Reference Knott-Craig, Elkins, Lane, Holz, McCue and Ward22 Our analysis has shown that there was no statistically significant difference between the two approaches. These findings validate the diverse outcome variabilities observed in preceding studies. The variability in outcomes is attributed to the absence of a standardised definition for reintervention actions. Certain studies exclusively focus on surgical reinterventions, whereas others encompass post-operative catheterisation within the reintervention scope. Furthermore, the anatomical complexity of patients and the timing of surgery initiation are posited as factors influencing reintervention rates in patients. Reference Bailey, Elci, Mascio, Mercer-Rosa and Goldmuntz10,Reference Mashali, Yousef and Elmahrouk21,Reference Knott-Craig, Elkins, Lane, Holz, McCue and Ward22
Cardiac complications in patients with Tetralogy of Fallot can occur both in the immediate or long-term post-operative period. Some examples of complications include residual ventricular septum, persistent right ventricular outflow tract, pulmonary regurgitation, right heart failure, infections in the heart layers, and heart valves, as well as arrhythmias. Reference van der Ven, van den Bosch, Bogers and Helbing2 Arrhythmias and cardiac arrest remain common in operated-Tetralogy of Fallot cases, even in patients who have undergone adequate correction. Reference Krieger, Zeppenfeld and Dewitt23 According to Savla et al., Reference Savla, Faerber and Huang9 cardiac complications are more frequent in primary repair, and this could also be a cause of the differences in 30-day and two-year mortality between the two methods. This is consistent with the findings of Bailey et al., Reference Bailey, Elci, Mascio, Mercer-Rosa and Goldmuntz10 where primary repair also has a higher prevalence of cardiac complications. One consideration that determines the chosen method is the McGoon Ratio. The McGoon ratio is a critical factor in Tetralogy of Fallot management. A normal McGoon ratio would typically favour the primary repair approach, while an inadequate ratio would indicate the need for palliative surgery first. Reference Moustafa, Hussein, Sultan, Bilal, El Gamal and Sobh24 In cases with a borderline McGoon ratio, there is no difference in the complication rates between the two methods. Reference Abdelgawad, Elshafie, Bayoumy and Elatafy25 The results of a meta-analysis of six studies regarding the relationship between cardiac complications and the choice of Tetralogy of Fallot correction methods in neonates indicate a statistically significant difference between the two procedures, with primary repair having a higher rate of complications compared to staged repair. The primary repair in symptomatic neonatal patients with Tetralogy of Fallot has a higher rate of post-operative and ICU cardiac complications compared to staged repair. The reasons and mechanisms behind this have not been explicitly mentioned in the available search results, but it can be hypothesised that a staged repair approach allows for better adaptation of the neonatal cardiovascular system to surgical changes, potentially reducing the risk of post-operative cardiac complications. Further research is needed to understand the underlying mechanisms and other contributing factors. Reference van der Ven, van den Bosch, Bogers and Helbing2,Reference Savla, Faerber and Huang9,Reference Bailey, Elci, Mascio, Mercer-Rosa and Goldmuntz10,Reference Krieger, Zeppenfeld and Dewitt23–Reference Abdelgawad, Elshafie, Bayoumy and Elatafy25
One concern associated with the primary repair is the increased use of the transannular patch and the potential for long-term complications in patients. The effective use of the transannular patch reduces the degree of obstruction in the right ventricular outflow tract. However, the long-term outcome of using this technique may result in pulmonary valve regurgitation. Chronic regurgitation leads to right ventricular dilation, an increased risk of biventricular dysfunction, decreased exercise tolerance, arrhythmias, heart failure, and sudden cardiac arrest. Reference Romeo, Etnel and Takkenberg26,Reference Stephens, Wolfe and Talwar27 This has been demonstrated in various early cohort studies of Tetralogy of Fallot patients undergoing the transannular patch procedure, whose populations are now entering adulthood. In the long term, there is a significant increase in mortality and morbidity associated with the choice of the transannular patch procedure in Tetralogy of Fallot patients. This condition necessitates patients undergoing reoperation, typically involving pulmonary valve replacement, 10–15 years after the initial surgery. Reference Ylitalo, Nieminen, Pitkänen, Jokinen and Sairanen28 Ventriculotomy procedures can result in transmural myocardial scarring and damage to coronary arteries, which can increase the risk of arrhythmias and disrupt right ventricular function. This should be a primary consideration due to its association with long-term outcomes after Tetralogy of Fallot correction. Reference Luijten, Van den Bosch and Duppen29,Reference Apitz, Webb and Redington30 Our study indicates a significant association between primary repair and the risk of transannular patch usage. These findings underscore the advantages of a staged repair approach in managing symptomatic neonates with Tetralogy of Fallot, aiming to mitigate the necessity for transannular patch application and facilitate future valve-sparing surgical interventions. Reference Luijten, Van den Bosch and Duppen29
In theory, one of the advantages of staged repair is to avoid complex procedures as the initial treatment for neonates, leading to shorter hospital stays. However, we found that the staged repair approach is associated with longer hospitalisation durations both in the general hospital setting and, in the ICU, compared to primary repair. This discrepancy may be attributed to the total hospitalisation duration, encompassing both the palliative procedure during the neonatal period and the subsequent complete corrective procedure for Tetralogy of Fallot patients undergoing staged repair. The heterogeneity we found in our analysis may arise from various factors, such as differences in the timing of surgical interventions and variations in patient management policies across different studies. Studies conducted by O’Byrne et al. Reference O’Byrne, Glatz and Yuan-Shung12 and Steiner et al. Reference Steiner, Tang and Gossett8 have also presented data on the total cost disparities associated with each type of surgical management strategy in neonatal Tetralogy of Fallot patients. The staged repair tends to incur higher total treatment costs compared to the primary repair. However, the total treatment costs calculated were for short-term outcomes, and there is no study providing data regarding the total costs required for the long-term outcome, knowing that the primary repair was associated with a higher transannular patch usage rate.
Limitations
There were several inevitable limitations of this study. First, our study did not carry out a subgroup analysis for the existing variables due to limited data regarding study and demographic characteristics. The results of the analysis were suboptimal because there were still very limited studies comparing the two management strategies, particularly in the neonatal population. All studies included in the analysis were performed in America. It is advised that in the future, studies comparing long-term variable outcomes should be carried out, as our study was only able to analyse short-term outcome variables and did not include long-term outcome variables on both management strategies.
Conclusion
This meta-analysis provides further information regarding the controversy between primary repair and staged repair for Tetralogy of Fallot correction in the neonatal population. Our study shows three main standings. First, primary repair resulted in a higher post-operative complication rate and higher usage of transannular patches. Second, staged repair resulted in a longer hospital and ICU length of stay. Furthermore, it resulted in higher costs compared with primary repair. Third, there is no difference in mortality or reoperation rates between two groups.
Acknowledgements
None.
Financial support
All authors declare that they have no financial interest or financial involvement in any organisation with a direct financial interest in the subject matter or materials discussed in our manuscript.
Competing interests
None.
