Tickle me ECMO…Differences and Outcomes Unearthed for Kids requiring ECMO in SARS-CoV-2 Associated Disease

After three years into the pandemic, pediatricians have grown accustomed to the fact that SARS-CoV-2 causes either acute COVID-19 disease or multisystem inflammatory syndrome in children (MIS-C). At their core, these are two different disease processes with the former a disorder of primary respiratory pathophysiology and the latter, a syndrome characterized by hyperinflammation and multiorgan dysfunction, especially cardiac disease [1,2]. Unsurprisingly, both can lead to fulminant organ failure requiring extracorporeal mechanical support, but as the disease processes are driven by different organ systems, the indications and modalities of support are different [2,3]. This is in stark contrast to adults with acute SARS-CoV-2 infection, who required extracorporeal membrane oxygenation (ECMO) almost exclusively for respiratory failure [4,5]. Sadly, there is a large gap in knowledge for ECMO in pediatric patients with SARS-CoV-2 associated disease and the literature is extremely limited [6]. Thus, it is imperative for this disease and potentially future pandemics with divergent pathophysiologic processes to best understand who, when, and how to deploy a scarce and important resource to guide clinical decision-making.

In this issue of Pediatric Critical Care Medicine, Bembea et al [7] describe the characteristics of children and adolescents with SARS-CoV-2 who required ECMO support. While there have been several large database studies and meta-analyses examining adults treated with ECMO for acute COVID-19, this is the first study to examine a large cohort of pediatric patients [4,5,8]. This is an important undertaking, and the results are informative.

Overall, Bembea et al. describe ECMO as rarely used for acute COVID-19 (5.9%) or MIS-C (2.4%); numbers comparable to adult ICU patients [9]. Patients with acute COVID-19 disease were supported with veno-venous (VV) strategies, while patients with MIS-C were supported using veno-arterial (VA) cannulation. This is not surprising, as MIS-C is a disease of inflammation and hemodynamic compromise [2]. These patients had an overall low mortality rate compared to other diagnoses requiring VA ECMO; very similar to the mortality rate in myocarditis [10,11]. Furthermore, 11% of patients in the MIS-C ECMO group had an oncologic diagnosis or immune compromise, which is a population with historically poor survival with mechanical support [12]. The aggregate of this data informs clinicians that MIS-C patients with refractory cardiogenic shock can have good outcomes when supported by ECMO, especially with VA ECMO.

Pediatric patients supported with ECMO for acute COVID-19, like their adult counterparts, had mostly respiratory disease and thus were most often supported with VV ECMO [7]. In contrast to adults, the overall rate of VV ECMO (vs VA) was much lower (59% vs 95–99%) [5,8]. Why was there a higher use of VA ECMO in pediatric ECMO patients? Perhaps pediatric patients requiring ECMO were more likely to be hemodynamically affected by the inflammatory milieu often seen in acute COVID-19 and subsequently developed vasodilatory or cardiogenic shock in addition to their respiratory failure, or their refractory hypoxemia led to significant hemodynamic compromise. Furthermore, it is possible that the infrastructure for placing children on VV ECMO is not available immediately or 24/7 in every pediatric ECMO center, leading to some patients who could have been supported with VV ECMO being placed on VA support due to cannulation urgency.

Intriguingly, despite the higher use of VA ECMO in pediatric patients with acute COVID-19, the mortality rate compares favorably to adult studies, especially when we consider the increasing mortality of adults supported on ECMO in the later stages of the pandemic. Pediatric patients in this study had a 37% mortality, while adults on ECMO for COVID-19 had 36.9–41.2% mortality in the first half of 2020, increasing up to >50% later in the pandemic [4,5]. Why are children faring better than their adult counterparts? Adult COVID-19 ECMO mortality is not well-understood, as reasons for this rising mortality are not clear. Speculations include a higher rate of steroid use, more inexperienced centers employing ECMO, and a higher rate of pre-ECMO non-invasive positive pressure ventilation delaying deployment [4]. Adult ECMO mortality in COVID-19 has been associated with increasing age, increasing time from intubation to ECMO, VA support (vs VV), pre-ECMO arrest, AKI, and chronic lung disease [4,8]. Children are obviously younger, but children supported with ECMO for acute COVID-19 had a higher rate of VA support (41% vs <5%), underlying lung disease (31% vs 3–4%), and ECPR (13% vs 1%) [4,5,7,8]. In many ways these children have more potent underlying illness and are sicker at the time of cannulation yet are still surviving at a higher rate. Perhaps age is more important than all these factors? Considering these data, pediatric patients with acute COVID-19 appear to be excellent ECMO candidates in the right circumstances.

Most interestingly, separate from the technical aspects of ECMO deployment, delving deeper into the story by Bembea et al. tells a demographic tale and unearths a fascinating story of the risk factors for severe SARS-CoV-2-associated disease in children. First, those that required ECMO were older than expected with a high median age (>13.6 years in acute COVID-19 and >9.9 years in MIS-C). Second, the authors found a higher BMI percentile (>95) associated with acute COVID-19, with even higher percentiles in those requiring ECMO. Third, Hispanic and Black non-Hispanic children comprised a disproportionate majority of both ICU patients and patients supported by ECMO compared to the US population. For instance, 64% of black children with SARS-CoV-2 associated disease had MIS-C vs 50% for white children and 53% of white Hispanic or Latino children. The largest racial group in the MIS-C ECMO cohort was black children, while white children were the largest group in the acute COVID-19 ECMO cohort [7]. These data seem to suggest that white and black children are manifesting severe illness associated with SARS-CoV-2 differently, whether due to epigenetic differences, or social determinants of health (SDoH), or both. Studying more children with these diseases could clarify these associations and guide future research that could allow clinicians to risk stratify patients to better predict and respond to their severe illness. Understanding this discordance with disease mechanistic and causal pathways in addition to potential SDoH modifiers would be quite informative to the field.

Undeniably, acute COVID-19 and MIS-C are both caused by infection with SARS-CoV-2, but as described, they are discrete conditions. The two diseases attack a child in different ways, and lead to decompensations that necessitate different ECMO support. The work from Bembea et al. further clarify and refine the consideration that acute COVID-19 and MIS-C, despite their shared heritage, caused different illnesses and should be considered distinct entities, including in how to support them extracorporeally.

Inasmuch, this study has several important limitations. ECMO management was not standardized, and there are few details reported about the ECMO run course, such as cannulation strategies, ECMO flow rates, mobilization, ventilator support, and airway management. During the pandemic, several adult centers employed ECMO strategies that included aggressive rehabilitation, patient awakening, right ventricular support, and extubation without tracheostomy; many modifications with good results [13,14]. It would be interesting to see how these strategies fared in children, especially with the developmental considerations of awake ECMO in younger children.

Another limitation is that ECMO outcome associations were only reported for acute COVID-19 and MIS-C. Further elucidation of how outcomes were associated with risk factors such as obesity, race, vaccination status, and socioeconomic factors could offer helpful clinical insight. Comparing outcomes for variables from adult outcome scores, such as the RESP score, could be illuminating as well [15]. Furthermore, adult centers used BMI as a cutoff for allowing cannulation, whereas in the report from Bembea et al. patients with a higher BMI fared quite well on ECMO. This beckons the question of whether the adult community should reconsider BMI as a contraindication for cannulation. Likewise, the field needs to further refine which ECMO patients are dying and why, so we can create a model to risk stratify and better predict who will need mechanical support and which strategies can help them. This report only touches the surface of answering such challenging but needed future explorations.

Nonetheless, further reports for pediatric ECMO are grossly needed to identify granular data in acute COVID-19 disease and within its sister disease, MIS-C. The virus manifests differently in children based on age and other patient factors, and the result is two distinct disease processes, each with different indications and support types for rescue therapy. Additionally, the impact on need for ECMO after vaccination or variant mutation will further need to be delineated to see if these two disease entities remain discordant in pathophysiology or in severity. Despite this incongruous pathophysiology, children with SARS-CoV-2 related illness appear to be excellent ECMO candidates with ‘better than their adult’ counterpart outcomes. Further research is needed to best understand how to deploy this heroic technology and refine the best populations for its use with epigenetic and SDoH factors, all in an effort to improve outcomes. Some children will invariably need to ‘tickle me ECMO,’ and understanding factors that lead to this therapy’s success will allow for more children to survive the disease.