Response to the commentary on “determining the optimal timing for tetralogy of Fallot management: A meta-analysis of neonatal versus postneonatal repairs”

We thank the authors for their interest and for raising three main issues: (1) the definition of “post-neonatal” and its alignment with guidelines, (2) our focus on peri-operative rather than long-term outcomes, and (3) the impact of heterogeneity. We address each directly below.

DEFINITION OF “POST-NEONATAL” AND ALIGNMENT WITH GUIDELINES

We defined neonatal repair as surgery performed within the first 28–30 days of life and postneonatal repair as surgery performed after that period. This threshold was prespecified and applied consistently across all included studies. The mean age at repair in the postneonatal cohorts was consistently more than 1 month, commonly between 2 and 6 months, confirming that the grouping was not arbitrary but aligned with the standard neonatal definition used in pediatric cardiac surgery and epidemiology.[1]

This approach aligns with contemporary consensus. The 2023 American Association for Thoracic Surgery Expert Consensus recommends elective complete repair at approximately 3–6 months for stable infants with tetralogy of Fallot (TOF) rather than in the neonatal period.[2] Recent reviews similarly suggest that repair during early infancy, at around three to 6 months of age, is preferred when clinically feasible.[3] We agree that peer-reviewed society guidelines and consensus statements are the most appropriate sources to guide clinical practice, and we therefore prioritized these over general web pages. Our postneonatal group was intentionally defined to capture the early infancy repair window supported by expert guidance, and the observed between center variation is more likely to reflect the real-world practice patterns than definitional inconsistency.

FOCUS ON PERI-OPERATIVE OUTCOMES GIVEN DATA LIMITATIONS ON LONG-TERM SEQUELAE

The commentary highlights the lack of very long-term endpoints such as chronic pulmonary regurgitation, adult right ventricular function, late pulmonary valve replacement (PVR), and exercise capacity. This gap reflects the limitations in the published literature rather than an omission specific to our work. Most comparative studies stratified by timing of repair focus on perioperative and early outcomes, while long-term follow-up stratified by repair timing remains scarce and largely observational. Even large longitudinal series reporting more than 900 TOF repairs over approximately 30 years describe excellent late survival and PVR rates but do not consistently attribute these outcomes to neonatal repair compared with infant repair timing.[4] There are no randomized trials with timing-stratified long-term data; all available comparisons are retrospective.

Given this evidence base, we focused on peri-operative outcomes that are both consistently reported and mechanistically tied to late sequelae. The need for a transannular patch (TAP) or right ventricle (RV) to pulmonary artery conduit at the index repair is a well-established driver of chronic pulmonary regurgitation, progressive RV dilation, and future PVR in many patients.[5] In our meta-analysis, postneonatal repair was associated with a lower need for TAPs/conduits, plausibly because slightly larger annuli in older infants permit valve-sparing repairs more often. We also observed lower rates of early arrhythmia and pacemaker implantation outside the neonatal period. These peri-operative differences are not trivial surrogates; they are causal precursors of many late problems in repaired TOF. Until timing-stratified long-term data are available, clinicians must infer long-term risk from intermediate surgical realities. On that basis, deferring repair beyond the neonatal period in stable infants is expected to reduce later valve-related morbidity, even if current publications cannot yet quantify the long-term delta by timing.[2,5]

ADDRESSING HETEROGENEITY THROUGH METHODOLOGICAL RIGOR AND CONSISTENCY OF FINDINGS

Heterogeneity is inherent in multi-center, multi-era surgical meta-analyses using the observational data. We anticipated this and used random-effects models whenever between-study variability was present, yielding conservative pooled estimates and appropriately wider confidence intervals. We also performed Risk of Bias in Nonrandomized Studies of Interventions (ROBINS-I) assessments for all included studies, with overall study quality rated as high, and judgments reported transparently.[6]

Critically, heterogeneity did not obscure the overall signal. The direction of effect was consistently aligned across endpoints and across most studies, favoring postneonatal repair. Pooled early mortality was lower outside the neonatal period, with odds ratios indicating higher risk with neonatal repair, and adverse perioperative events such as delayed chest closure and pacemaker insertion were also more frequent in neonates. Although absolute event rates varied, reflected by moderate I-squared heterogeneity for some outcomes, the consistency in direction across the diverse settings argues against an artifact of pooling. If the literature were truly conflicted, pooled effects would be expected to trend toward the null with clear subgroup divergence. We observed the opposite. Benefits of postneonatal repair were coherent across institutions, techniques, and eras.

CONCLUSION

Our definitions were explicit and consistent with authoritative consensus that, when clinically feasible, elective TOF repair in early infancy at three to 6 months of age is preferable to neonatal repair.[2,3] The literature does not yet provide the decades long outcomes stratified by repair timing, so our analysis focused on robust perioperative endpoints that plausibly shape long-term trajectories, including TAP or conduit use and perioperative arrhythmias. Across 19 studies including 28,968 patients, random effects pooling with ROBINS-I appraisal indicated a consistent pattern in which repair after the neonatal period was associated with lower early mortality and fewer complications, together with operative strategies that may reduce later valve related morbidity. Specifically, neonatal repair was associated with higher mortality with odds ratio 1.68 and 95% confidence interval 1.47–1.92, higher odds of delayed chest closure with odds ratio 2.52 and 95% confidence interval 2.06–3.09, and higher odds of pacemaker implantation with odds ratio 3.68 and 95% confidence interval 2.89–4.70, consistently favoring postneonatal repair. Neonatal repair was also associated with longer intensive care unit stay with standard mean difference 1.58 and 95% confidence interval 1.04–2.12, and longer hospital stay with standard mean difference 5.18 and 95% confidence interval 3.54–6.82.

These findings are biologically plausible, methodologically cautious, and clinically actionable now, while underscoring the need for prospective cohorts with long-term follow-up stratified by repair timing. The available evidence supports deferring complete TOF repair beyond the neonatal period in stable infants to optimize perioperative outcomes and, by reducing TAP and conduit use, to plausibly improve long-term risk profiles. This stance is consistent with current expert consensus and the best comparative data presently available. We welcome further prospective, timing aware studies to quantify the long-term benefits more precisely.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.