Discussion

In this analysis of registry data of patients with single ventricle anatomy who were supported with extracorporeal membrane oxygenation prior to first-stage palliation, the survival rate to hospital discharge was 42%. Variables significantly associated with hospital mortality included renal injury requiring renal replacement therapy and the presence of equal to or greater than four complications. This suggests that renal compromise may be a marker of poor pre extracorporeal membrane oxygenation perfusion, and may also be associated with worse outcomes in this specific cohort. ^{Reference Mascio, Austin and Jacbos9,Reference López-Herce, Casado and Díez10,Reference Kulik, Moler and Palmisano11} To our knowledge, this is the first study that has reported single ventricle mortality outcomes after pre-operative extracorporeal membrane oxygenation support ads a bridge to first-stage palliation.

Bautista-Hernandez et al., found that extracorporeal membrane oxygenation may be useful in bridging patients with CHD to a definitive vs. palliative repair. ^{Reference Bautista-Hernandez, Thiagarajan and Fynn-Thompson6} Although, their study had a limited single ventricle cohort, 62% of patients survived hospital discharge compared to 42% in our study. ^{Reference Bautista-Hernandez, Thiagarajan and Fynn-Thompson6} In comparison to Bautista Hernandez’ study where 62% of patients survived hospital discharge, only 42% of our patients survived. Bautista Hernandez’s group did not appreciate a significant difference in outcome between single and two ventricle patients, perhaps because the cohort was limited to 6. This may be a reason as to why we saw greater mortality, as we were exclusively looking at the single ventricle population. ^{Reference Bautista-Hernandez, Thiagarajan and Fynn-Thompson6} A population that is generally sicker and more complex. ^{Reference Aydin, Duffy and Rodriguez12} Within Bautista’s study of the three patients with hypoplastic left heart syndrome, two who died had intact atrial septum. ^{Reference Bautista-Hernandez, Thiagarajan and Fynn-Thompson6} Our hypoplastic left heart syndrome population also carried a higher mortality rate, although we were unable to stratify whether the patients had intact atrial septum which could influence outcome. Of the 20 patients who had a diagnosis of hypoplastic left heart syndrome, 14 died (70%). This may highlight that patients with hypoplastic left heart syndrome who suffer cardiac arrest or have haemodynamic deterioration prior to first-stage palliation may not be optimal candidates for pre-operative extracorporeal membrane oxygenation support.

A paper by Fernandez et al., looked at interstage (between stages 1 and 2) survival for patients with hypoplastic left heart syndrome after extracorporeal membrane oxygenation. The study found that 18% of interstage patients requiring extracorporeal membrane oxygenation died. ^{Reference Fernandez, Joy and Allen13} This discrepancy between pre-first-stage palliation and interstage may speak to a more critically ill and unstable neonatal population. ^{Reference Aydin, Duffy and Rodriguez12,Reference Fernandez, Joy and Allen13} Our study’s survival to discharge of 42% mimics the Extracorporeal Life Support Registry report published in 2016. ^{Reference Barbaro, Paden and Guner4} Our study gathered data between 2016 and 2021 and still carried the same mortality rate, indicating no improvement over the course of 5 years. This may be because patients with single ventricle physiology who had myocardial decompensation requiring extracorporeal membrane oxygenation support prior to their surgical palliation had complex haemodynamics that were not suitable for surgical palliation.

The need for renal replacement therapy has been associated with poor outcomes and increased mortality in paediatric patients requiring extracorporeal membrane oxygenation support. ^{Reference Pettit, Selewski and Askenazi14,Reference Wolf, Chanani and Heard15,Reference Morris, Ittenbach and Godinez16} Several studies have highlighted the association between renal replacement therapy and adverse outcomes in single ventricle patients on extracorporeal membrane oxygenation. Renal injury can manifest as acute kidney injury with or without anuria, characterised by a rapid decline in renal function. ^{Reference Wolf, Chanani and Heard15} The underlying mechanisms contributing to renal injury in these patients on extracorporeal membrane oxygenation include haemodynamic instability, decreased renal perfusion, micro emboli, inflammatory response, and nephrotoxic medications. ^{Reference Pettit, Selewski and Askenazi14,Reference Wolf, Chanani and Heard15} A study by Wolf et al., found that paediatric patients with a cardiac diagnosis of extracorporeal membrane oxygenation were significantly more likely to die if they required continuous venovenous hemofiltration, in addition, they had longer hospital stays. ^{Reference Wolf, Chanani and Heard15} Wolf et al.’s population included single and two ventricle patients. Of whom, nearly 20% of patients had hypoplastic left heart syndrome. ^{Reference Wolf, Chanani and Heard15} Another study by Pettit et al., examined the impact of acute kidney injury and fluid overload on paediatric patients receiving extracorporeal membrane oxygenation for cardiac disease. ^{Reference Pettit, Selewski and Askenazi14} They found a significant association between peak fluid overload status and mortality in the presence of acute kidney injury. ^{Reference Pettit, Selewski and Askenazi14} Our study was unable to assess for fluid overload given the limitations of the database. None of the above-mentioned studies exclusively explored renal injury in the context of neonates with single ventricle physiology requiring extracorporeal membrane oxygenation prior to stage one palliation. Our study highlights that renal injury and the use of renal replacement therapy continue to be associated with poor outcomes in this vulnerable population.

Multiple complications on extracorporeal membrane oxygenation have been associated with poor outcomes in neonates with single ventricle physiology. ^{Reference Raithel, Pennington and Boegner17} These patients are at higher risk of developing complications due to the complex nature of their cardiac anatomy and the challenges in managing their pre-operative physiology. ^{Reference Raithel, Pennington and Boegner17,Reference Jaggers, Forbess and Shah18,Reference Perry, Brown and Misfeldt19} These complications can include acute kidney injury, neurologic injury, infections, coagulopathy, and cardiac dysfunction and have been associated with increased mortality. While prolonged duration of extracorporeal membrane oxygenation support and longer hospital length of stay are linked to higher rates of long-term morbidity. ^{Reference Jaggers, Forbess and Shah18,Reference Perry, Brown and Misfeldt19} A study by Misfeldt et al., found that acute renal failure was independently associated with mortality in neonates with single ventricle physiology during extracorporeal membrane oxygenation. ^{Reference Misfeldt, Kirsch and Goldberg7} Another study by Morris et al. reported that both renal and hepatic injury for paediatric post-cardiac surgery patients on extracorporeal membrane oxygenation was associated with increased mortality. ^{Reference Morris, Ittenbach and Godinez16} Our study was looking at preoperative patients; however, one could extrapolate that any evidence of multi-organ dysfunction such as bleeding, anuria, or seizure could be a negative prognostic marker for our patients. The presence of multiple complications on extracorporeal membrane oxygenation may reflect the severity of the underlying disease. Our study similarly found that patients with four or more complications were significantly more likely to die on extracorporeal membrane oxygenation and before hospital discharge.

There were several limitations to this retrospective database study. The variables incorporated into the analysis were limited to discrete data fields collected in the Extracorporeal Life Support Registry. The analysis of specific outcome variables was impacted by incomplete data field completion and only focused on limited variables with an association with mortality. For example, we were unable to determine which patients were placed on prostin prior to extracorporeal membrane oxygenation cannulation, which may have influenced outcome variables. Additionally, the Extracorporeal Life Support Registry includes multiple centres and does not account for centre level-variability related to extracorporeal membrane oxygenation cannulation strategies, circuit architectures, and anticoagulation practices which may have confounded the results. The small sample size and limited power allow us to infer generalisable impact on outcomes but future studies should include more patients and more granular data.

In conclusion, patients supported with extracorporeal membrane oxygenation prior to their first stage of surgical palliation are at high risk for mortality, without improvement in the last decade. Renal injury requires renal replacement therapy, and those who suffer from multiple complications are at increased risk of mortality. This study may serve to set expectations for clinicians as they counsel families on the risks and benefits of extracorporeal membrane oxygenation support prior to their first-stage, single ventricle palliation.