Antepartum Extracorporeal Life Support: A Viable Way to Improve Fetal Outcomes

Extracorporeal life support (ECLS) has become increasingly common in the United States to manage refractory cardiopulmonary failure and has been used in complex populations including pregnant women. Application of ECLS to antepartum cardiopulmonary arrest is particularly unique because therapy affects two patients, the mother and the fetus. Therefore, evaluating the outcomes for both the mother and fetus is important to assess the potential benefits and harms of ECLS in this population. Previous data on the use of ECLS for antepartum shock have shown favorable short-term outcomes in pregnant women, with lower in-hospital morbidity and mortality compared to other patients undergoing ECLS, although in-hospital mortality in the antepartum period remained high at approximately 30%.1 Regarding short-term neonatal outcomes, data collected predominantly from case reports identified that the use of maternal ECLS at a median gestational age of 24 weeks resulted in approximately 35% in-hospital infant mortality.2 Although such short-term evaluation of in-hospital outcomes is important, evaluation of posthospitalization outcomes among children born to mothers requiring antepartum ECLS is likewise critical. In this issue of ASAIO Journal, Seadler et al.3 describe a case series of seven women undergoing ECLS during pregnancy, with a specific focus on in-hospital and posthospital neonatal outcomes. Indication for ECLS was cardiogenic shock in one woman and respiratory failure in six women, with four undergoing veno-venous extracorporeal membrane oxygenation (V-V ECMO), two undergoing veno-pulmonary arterial (V-A ECMO), and one undergoing V-A ECMO with a percutaneous left ventricular assist device. The median gestational age at ECLS initiation was 27 weeks and median gestational age at delivery was 29 weeks, with 6 of 7 neonates delivered while on maternal ECLS. All neonates underwent operative delivery and had a median birth weight of 1.4 kg. Neonatal predischarge morbidity included the need for invasive mechanical ventilation in five neonates, bronchopulmonary dysplasia in three neonates, and patent ductus arteriosus in one neonate. All mothers and all neonates survived to discharge with a median neonatal hospitalization duration of 56 ± 21 days. The authors were able to follow the neonates for a mean postdischarge duration of 1.4 ± 1.2 years. During follow-up, one child required rehospitalization, for otitis media. Four children underwent neurodevelopmental testing, with two children demonstrating mild-moderate delays related to speech and language. The authors concluded that maternal ECLS is associated with high maternal and neonatal survival, and with favorable intermediate-term neonatal morbidity possibly as expected for gestational age. Descriptive studies, such as this case series, are an important first step to tap into an unexplored area. However, several limitations are worth noting, including those acknowledged by the authors. The case series is small, from a single center, and was collected over an 8 year period. The number of total pregnant women undergoing ECLS at the study institution, including those with pre-viable fetuses where termination may have occurred, is not reported. Longer follow-up and perhaps standardized developmental testing of children born to mothers requiring ECLS will also be beneficial to assess for longer-term developmental or health consequences. Two infants out of four tested for neurocognitive development demonstrated delays, especially in language and speech. Based on the small sample size, it is not possible to determine whether this is high or expected. Prematurity by itself predisposes to similar neurocognitive delays4,5 and severe maternal respiratory failure is associated with similar deficits in neonates.6 So, should we ascribe the developmental delays to prematurity, maternal respiratory failure, or ECLS, especially since the duration of the ECMO support in these cases is not reported? Going forward, additional studies will be needed on antepartum ECLS. The threshold for mechanical cardiopulmonary support may vary from case to case, and objective criteria for antepartum ECLS initiation would be important given the potential for significant morbidity from the treatment itself. In addition, the optimal duration of ECLS to balance maternal and fetal risks while allowing sufficient time for fetal development before delivery will require further evaluation. Further study will also be needed to evaluate optimal hemodynamics for fetal development in women on ECLS with regard to target blood pressure, oxygenation, hemodynamic medication use, and potential consequences of continuous flow on fetal development. In these ways, the ECLS literature mirrors the limited data on pregnancy outcomes in patients supported with durable ventricular assist devices, where significant fetal morbidity and mortality have been reported.7 In conclusion, Seadler et al.3 should be congratulated for focusing on antepartum ECLS beyond in-hospital outcomes, which has important clinical implications. An appreciation of favorable prognosis may lead to ECLS being seen as an increasingly viable option in the antepartum period in select women undergoing cardiopulmonary failure, both to save the life of the mother but also allow for additional time for fetal development. Given the rarity of antepartum cardiopulmonary arrest, highlighting successful outcomes may ultimately lead to wider application of ECLS in this population through additional focus on early identification of cardiopulmonary failure, on wider availability of ECLS, and on standardized multidisciplinary ECLS management. Unfortunately, ECLS is a life-saving technology which is extremely difficult to study in a randomized controlled fashion. Several recently published studies on V-A ECMO delivered negative results.8 By design, such studies may be either proactive, where patients randomized in the ECLS arm are potentially placed on support while not sick enough and suffer from adverse events, or reactive, with an inevitably high crossover rate from the medical arm. Paradoxically, databases and registries may be better suited to provide much-needed answers. As each registry starts with a case series, future efforts to collect multicenter data in this rare population may provide further clarity on neonatal outcomes postmaternal ECLS. Although antepartum cardiopulmonary failure and the need for ECLS are likely to remain rare, the study by Seadler et al.3 on infrequent but catastrophic events, which may nevertheless result in favorable outcomes, highlights antepartum ECLS as a miracle of modern medicine.