Targeting Hemodynamics of Cardiopulmonary Resuscitation to Cardiac Physiology—The Next Frontier for Resuscitation Science?*

In this issue of Pediatric Critical Care Medicine, Yates et al (1) investigate the association of hemodynamics and end-tidal CO2 (ETco2) during early in-hospital cardiac arrest (IHCA) and neurologic outcomes in 97 children rescued with extracorporeal cardiopulmonary resuscitation (ECPR). The study represents a preplanned secondary analysis of data from the Improving Outcomes from Pediatric Cardiac Arrest—the ICU-Resuscitation Project (ICU-RESUS) (2). The cohort, with contributions from 18 centers during 2016–2021, included predominantly infants (71/97 73% < 1 yr old), children with congenital heart disease (CHD; 82/97 85%) and postoperative cardiac surgical patients (62/97 64%). The investigators compared patients who survived with a favorable neurologic outcome—defined as a Pediatric Cerebral Performance Category (PCPC) score less than or equal to 3 or no increase from baseline PCPC—and those who did not. While cardiopulmonary resuscitation (CPR) duration was shorter for patients who survived with a favorable neurologic outcome (36.5 min 21–52 min vs. 47 min 34–61 min; p = 0.015), there was no identified difference in median diastolic or systolic blood pressure (BP) or number achieving target systolic or diastolic BP during the initial 10 minutes of resuscitation. Similarly, metrics of compression quality including chest compression fraction greater than 0.9 and chest compression rate were not different between the two groups. Finally, the average or maximum ETco2, available for 35 of the 97 patients, were not significantly different between survivors with good neurologic outcomes and others. Interestingly, the authors identified a small subgroup of four patients who survived with good neurologic outcome despite documentation of average ETco2 less than 10 mm Hg during the first 10 minutes of CPR preceding ECPR. As the majority of pediatric IHCA occurs in the ICU (3), leveraging invasive monitoring to understand the direct effect of CPR has recently become an important aspect of resuscitation management. In fact, multiple animal models have demonstrated improved survival with real-time titration of chest compressions and vasoactive medications based on invasive BP response and ETco2 in comparison to guideline-only directed care (4–6). In the wake of these findings, investigators in the Pediatric Intensive Care Quality of CPR (PICqCPR) study sought to better understand the significance of invasive hemodynamic data in pediatric CPR recipients. Berg et al (7) showed that sustaining a mean diastolic BP greater than or equal to 25 mm Hg in infants younger than 1 year and greater than or equal to 30 mm Hg in children 1 year old or older was associated with improved survival and survival with favorable neurologic outcome. Notably, in both the ICU-RESUS and PICqCPR cohorts, patients with preexisting cardiac diagnoses were predominant, particularly those with CHD (85% and 79%, respectively). It is therefore critical to characterize the unique challenges associated with resuscitation in this population of patients. Children with cardiac disease are at higher risk of cardiac arrest when compared with children without cardiac diagnosis (8). A recent meta-analysis showed that about 5% of children with cardiac disease admitted to an ICU suffer IHCA and, of those, 22% were supported with ECPR (9). Importantly, the proportion of patients who did not achieve return of spontaneous circulation is unchanged over the last 20 years, suggesting a crucial need for developing resuscitation strategies and support tools specific to children with cardiac disease (9). The heterogeneity of cardiac physiologies associated with underlying cardiac anatomy and surgical palliation stages present unique challenges that significantly impact the effectiveness of resuscitation (8,10). For instance, during the arrest of an infant with interstage single ventricle (SV) physiology, chest compression into parallel systemic and pulmonary circulations results in a distinctly different hemodynamic effect, with systemic output partly diverted toward the lower resistance pulmonary circulation. For patients with cavopulmonary connections—bidirectional Glenn (BDG) or Fontan circulation—characterized by passive pulmonary blood flow, chest compression produces systemic output, but venous return must flow through the lungs, limiting systemic preload and consequently systemic output during resuscitation (8). Not surprisingly, SV physiology has been associated with unfavorable outcomes in multiple studies of pediatric ECPR outcomes (8,9,11). Detailed studies of hemodynamics during cardiac arrest for palliated cardiac anatomy phenotypes have not been undertaken, but a secondary analysis of the PICqCPR study compared cardiac surgical and cardiac medical patients (12). The authors found that the number of patients achieving diastolic BP goal was similar between the two groups (cardiac surgical 58% vs. cardiac medical 68%; p = 0.488); however, survival to discharge was higher among cardiac surgical patients compared with medical (56% vs. 32%; p = 0.013), and achievement of goal diastolic BP was associated with improved survival only in cardiac surgical patients (p = 0.018) and not in cardiac medical ones (12). Similar findings were shown in a prior analysis of the ICU-RESUS cohort (13). In this setting, it is clear how a single universal dichotomic cutoff of diastolic or systolic BP may easily fail in predicting adverse outcomes in the heterogeneous population of cardiac phenotypes. Most likely, either different cutoffs or combination of hemodynamic values, specific for each cardiac physiology or phenotype, may offer more insight and possibly stronger predictive value. A substantial proportion of patients included in the study by Yates et al (1) had SV physiology, and many other had cardiac surgical or medical disease. This represents an important confounding factor when interpreting their results. Last, Yates et al (1) report no association between ETco2 measurements, available in 35 patients, and survival with favorable neurologic outcome among the ICU-RESUS ECPR cohort. In a recent publication of the entire ICU-RESUS cohort (234 patients), Morgan et al (14) found that average ETco2 greater than or equal to 20 mm Hg during CPR was associated with improved survival to hospital discharge but was also not associated with favorable neurologic outcome. There was a high prevalence of CHD in the ICU-RESUS ECPR cohort with 85% diagnosed with CHD, compared with 66% cardiac medical or surgical patients in analysis by Morgan et al (14). Indeed, the ICU-RESUS cohort included 12 infants with interstage SV physiology, four with BDG, nine patients with systemic to pulmonary artery shunt. Children palliated by such circulations have starkly different means of sustaining pulmonary blood flow compared with patients with structurally normal, two-ventricle circulation. Therefore, there may be other confounding physiologic factors impacting the ETco2 during resuscitation unrelated to CPR quality which must be considered for ECPR candidacy decision-making. In this population, it is especially important to recall the author's conclusion that low ETco2 should not preclude extracorporeal membrane oxygenation cannulation in patients otherwise felt to be appropriate candidates. Although the study by Yates et al (1) was limited by small sample size with descriptive univariate analysis only, the authors continue to characterize important constructs for clinical care. The use of invasive monitoring to personalize and titrate CPR has become a focus during resuscitation for IHCA in children. However, it is crucial that these research efforts continue to better understand the titration of resuscitative measures, with specific or composite hemodynamic goals according to cardiac physiology phenotype in the high-risk palliated CHD population. Prevention of cardiac arrest and early initiation of high-quality CPR targeted to process hemodynamic and ETco2 metrics remain imperative for improving outcomes of high-risk patients. However, these may be further enhanced through research associating the cardiac physiologies of palliated CHD patients during IHCA to improve outcomes of these vulnerable patients.