Discussion

The present study evaluated the influence of antegrade pulmonary blood flow on the development of pulmonary arteriovenous malformations, systemic oxygenation, and haemodynamic suitability for Fontan completion following bidirectional Glenn surgery. The principal finding of this study is that preservation of antegrade pulmonary blood flow did not significantly reduce the incidence of pulmonary arteriovenous malformations when compared with interruption of antegrade pulmonary blood flow. Although patients with preserved antegrade pulmonary blood flow demonstrated higher oxygen saturations at follow-up, this did not translate into a measurable reduction in pulmonary arteriovenous malformation burden, supporting the concept that factors other than pulsatile pulmonary blood flow are primarily responsible for pulmonary arteriovenous malformation formation after cavopulmonary connection. ^{Reference Chen, Tulloh, Caputo, Stoica, Kia and Parry10,Reference Baek, Park and Choi12,Reference Schreiber, Cleuziou, Cornelsen, Hörer, Eicken and Lange14} However, the management of additional pulmonary blood flow at the time of bidirectional superior cavopulmonary anastomosis remains controversial, with studies demonstrating both potential benefits and disadvantages of preserving antegrade flow. ^{Reference Zahr, Kirshbom and Kopf16–Reference Alghamdi19}

Pulmonary arteriovenous malformations are a well-recognised late complication following bidirectional superior cavopulmonary anastomosis and are an important contributor to progressive cyanosis and exercise intolerance. ^{Reference Kavarana, Jones, Stroud, Bradley, Ikonomidis and Mukherjee2,Reference Duncan and Desai4,Reference McFaul, Tajik, Mair, Danielson and Seward5} The prevailing hypothesis attributes pulmonary arteriovenous malformation development to exclusion of hepatic venous effluent from the pulmonary circulation, resulting in loss of hepatic-derived angiogenesis inhibitors and subsequent abnormal pulmonary vascular remodelling. ^{Reference Srivastava, Preminger and Lock3,Reference Vettukattil, Slavik and Lamb6,Reference Gossage and Kanj7} This mechanism has been supported by both experimental and clinical observations, including resolution of pulmonary arteriovenous malformations following redirection of hepatic venous blood to the pulmonary arteries. ^{Reference Srivastava, Preminger and Lock3,Reference Vettukattil, Slavik and Lamb6} The findings of the present study are consistent with this paradigm, as pulmonary arteriovenous malformation incidence remained similar in both groups despite differences in pulmonary blood flow patterns, reinforcing the concept that hepatic factor deprivation rather than flow pulsatility is the dominant driver of pulmonary arteriovenous malformation development. ^{Reference Srivastava, Preminger and Lock3,Reference Vettukattil, Slavik and Lamb6,Reference Gossage and Kanj7}

Patients in whom antegrade pulmonary blood flow was preserved demonstrated higher resting oxygen saturations both at baseline and at follow-up. This is an expected physiological consequence of increased pulmonary blood flow and has been reported consistently in previous studies. ^{Reference Allgood, Alejos, Drinkwater, Laks and Williams9–Reference Baek, Park and Choi12} However, improved oxygenation did not correlate with a lower prevalence of pulmonary arteriovenous malformations, suggesting that higher saturations reflect increased pulmonary perfusion rather than improved pulmonary vascular integrity. This distinction is clinically important, as reliance on oxygen saturation alone may obscure ongoing pulmonary vascular pathology. ^{Reference Kavarana, Jones, Stroud, Bradley, Ikonomidis and Mukherjee2,Reference Duncan and Desai4,Reference Vettukattil, Slavik and Lamb6} While improved saturations may offer symptomatic benefit in the intermediate term, they do not appear to confer protection against long-term complications related to abnormal pulmonary vascular development. ^{Reference Chen, Tulloh, Caputo, Stoica, Kia and Parry10,Reference Baek, Park and Choi12,Reference Schreiber, Cleuziou, Cornelsen, Hörer, Eicken and Lange14}

The modest and statistically non-significant increment in oxygen saturation observed in the antegrade pulmonary blood flow-preserved group at follow-up should not be considered problematic. Prior studies have demonstrated variability in the magnitude and durability of oxygenation benefit associated with antegrade pulmonary blood flow, influenced by patient selection, timing of bidirectional superior cavopulmonary anastomosis, and evolving pulmonary vascular physiology. ^{Reference Chen, Tulloh, Caputo, Stoica, Kia and Parry10,Reference Frommelt, Frommelt and Berger13} Moreover, additional pulmonary blood flow after bidirectional superior cavopulmonary anastomosis has been associated with increased ventricular volume load, prolonged pleural effusions, and higher central venous pressures in some reports, potentially offsetting the perceived benefit of higher saturations. ^{Reference Frommelt, Frommelt and Berger13,Reference Schreiber, Cleuziou, Cornelsen, Hörer, Eicken and Lange14} Recent studies have also shown that although additional pulmonary blood flow may improve oxygenation in the short term, its effect on long-term Fontan outcomes and pulmonary vascular development remains uncertain, and individualised decision-making is generally recommended. ^{Reference Chen, Yuan and Liu18–Reference Lee, Amabile, Geirsson, Gruber and Kopf21} Accordingly, the lack of a large saturation increase in our cohort does not contradict the short-term physiological effects of antegrade pulmonary blood flow but highlights that any oxygenation advantage is limited and does not equate to protection from pulmonary arteriovenous malformation formation. ^{Reference Chen, Tulloh, Caputo, Stoica, Kia and Parry10,Reference Baek, Park and Choi12,Reference Schreiber, Cleuziou, Cornelsen, Hörer, Eicken and Lange14}

An important observation in this study is the comparable pulmonary artery dimensions and pulmonary artery pressures between the two groups. Patients with interrupted antegrade pulmonary blood flow did not demonstrate inferior pulmonary artery growth or unfavourable haemodynamics, indicating that elimination of antegrade flow does not compromise pulmonary vascular development or Fontan candidacy. In fact, mean pulmonary artery pressures tended to be lower in the antegrade pulmonary blood flow-interrupted group, a finding that may be advantageous in the context of Fontan physiology, where elevated pulmonary vascular resistance is a key determinant of long-term failure. ^{Reference Mainwaring, Lamberti, Uzark, Spicer, Cocalis and Moore11,Reference Frommelt, Frommelt and Berger13} These findings align with prior reports suggesting that ventricular unloading achieved by eliminating antegrade pulmonary blood flow may offset the theoretical benefits of increased pulmonary blood flow. Long-term follow-up studies of cavopulmonary connections have similarly demonstrated acceptable survival and functional outcomes with different sources of pulmonary blood flow, further supporting an individualised approach. ^{Reference Zahr, Kirshbom and Kopf16,Reference Steele, Zahr, Kirshbom, Kopf and Karimi17,Reference Delikatnyi, Sekelyk, Maksymenko and Kurkevych20}

With regard to age at bidirectional superior cavopulmonary anastomosis, all patients in the present study were younger than 15 years at follow-up, reflecting a cohort in which Glenn procedures were performed later than current international practice, where bidirectional superior cavopulmonary anastomosis is typically undertaken in early infancy. ^{Reference Uemura, Yagihara, Kawashima, Okada, Kamiya and Anderson8,Reference Chen, Tulloh, Caputo, Stoica, Kia and Parry10} Delayed presentation and staged palliation are common in low- and middle-income settings due to referral patterns, socio-economic factors, and limited access to specialised paediatric cardiac care. Although timing of bidirectional superior cavopulmonary anastomosis may influence pulmonary vascular adaptation, both antegrade pulmonary blood flow groups in our cohort were demographically and haemodynamically comparable, supporting the validity of intergroup comparisons. Importantly, the persistence of similar pulmonary arteriovenous malformation rates across both groups despite differences in antegrade pulmonary blood flow status further supports the hypothesis that hepatic venous exclusion, rather than age at bidirectional superior cavopulmonary anastomosis or the presence of pulsatile flow, is the principal determinant of pulmonary arteriovenous malformation development. ^{Reference Srivastava, Preminger and Lock3,Reference Vettukattil, Slavik and Lamb6,Reference Gossage and Kanj7} The higher incidence of cyanosis observed in the antegrade pulmonary blood flow-interrupted group should be interpreted in the context of lower baseline oxygen saturation and underlying haemodynamic differences rather than as a direct effect of antegrade pulmonary blood flow interruption, reflecting the retrospective and non-randomised nature of the study.

The clinical implications of these findings are particularly relevant in resource-limited settings, where timing of staged palliation and long-term follow-up may be variable. Preservation of antegrade pulmonary blood flow may offer short-term improvement in oxygenation but at the potential cost of increased ventricular volume load. ^{Reference Frommelt, Frommelt and Berger13,Reference Schreiber, Cleuziou, Cornelsen, Hörer, Eicken and Lange14} Conversely, interruption of antegrade pulmonary blood flow appears to provide a haemodynamically favourable profile without increasing the risk of pulmonary arteriovenous malformation formation. ^{Reference Allgood, Alejos, Drinkwater, Laks and Williams9,Reference Schreiber, Cleuziou, Cornelsen, Hörer, Eicken and Lange14} These observations support an individualised surgical strategy rather than a uniform approach, with consideration of ventricular function, pulmonary artery pressures, and clinical status when deciding on antegrade pulmonary blood flow management. ^{Reference Kavarana, Jones, Stroud, Bradley, Ikonomidis and Mukherjee2,Reference Srivastava, Preminger and Lock3,Reference Vettukattil, Slavik and Lamb6,Reference Mahle, Wernovsky, Bridges, Linton and Paridon15}

Several limitations must be acknowledged. The retrospective design introduces potential selection bias, particularly as the decision to preserve or interrupt antegrade pulmonary blood flow was surgeon-dependent. The sample size, while comparable to other single-centre studies, may be underpowered to detect small differences in pulmonary arteriovenous malformation severity. Quantitative assessment of shunt fraction and exercise capacity was not available, and follow-up duration, although sufficient to detect early pulmonary arteriovenous malformation formation, may not capture late-onset changes. Additionally, the absence of biochemical or molecular markers of hepatic factor activity limits mechanistic interpretation.

Despite these limitations, this study contributes meaningful data to an area where evidence remains limited and conflicting. By directly comparing pulmonary arteriovenous malformation incidence between antegrade pulmonary blood flow-preserved and antegrade pulmonary blood flow-interrupted cohorts using imaging-based assessment, the study provides clinically relevant insights into the long-standing debate regarding optimal Glenn physiology. ^{Reference Chen, Tulloh, Caputo, Stoica, Kia and Parry10,Reference Baek, Park and Choi12–Reference Schreiber, Cleuziou, Cornelsen, Hörer, Eicken and Lange14} The findings reinforce the concept that prevention of pulmonary arteriovenous malformations is primarily related to restoration of hepatic venous flow rather than maintenance of antegrade pulmonary flow.