SCCM Resource Library-ECMO During Pregnancy

Video Transcription

Hi, to everyone watching, my name is Emily Nahum. I'm a critical care physician and obstetric anesthesiologist at Massachusetts General Hospital in Boston.  I will be speaking about a thankfully rare but very important population for ECMO and that is the parturant.  My objectives to cover in this short time include reviewing the epidemiology of peripartum  ECMO, indications for use, specific considerations for pregnant patients, maternal and fetal outcomes, as well as future areas of research.  So we'll start by reviewing how often peripartum ECMO is really needed and how this has evolved over the last decades.  This first figure comes from a systematic review in 2020 looking at peripartum ECMO in 358 women.  It demonstrates there was an exponential increase in reporting of cases since 2009 where there was an important inflection point at the time of the H1N1 pandemic.  As you may recall, pregnant women had more severe courses of H1N1 and clinicians began to share their experience with peripartum ECMO at that time.  Maternal and fetal survival rates turned out to be reasonably favorable and so groups began publishing their data to provide more substantial evidence to support this practice.  This figure comes from a retrospective study reviewing ECMO hospitalizations using the National Inpatient Sample.  It found that during the 16-year study period, 331 cases were performed in obstetric patients.  The rate of obstetric ECMO utilization increased significantly during that study period from  1.1 cases per million in 1999 to 2002 to 11.2 in 2011 to 2014, corresponding to 145% increase per four-year period.  The overall patient population we're talking about here is unique. They're way less likely to have significant comorbidities.  When it comes to rescue therapies prior to ECMO, most studies show that the use is similar to non-pregnant patients.  This table here comes from the same National Inpatient Sample study and it compares patient characteristics of obstetric ECMO with non-obstetric ECMO.  Notably, you can see here that there is a statistically significant difference in the percentage of non-Hispanic black patients treated for peripartum ECMO when compared  with non-obstetric patients. This table comes from a study looking at ECMO for COVID-19 ARDS in the peripartum population.  This was a retrospective cohort study using the ELSO registry. As you can see, the patients were less likely to have obesity as a diagnosis corresponding  with prior studies suggesting these patients are healthier than non-pregnant patients. It also demonstrated that peripartum patients were more likely to be Hispanic, black, or  Asian when compared with the non-pregnant cohort. We know that black and Hispanic women are more likely to experience severe maternal  morbidity and we're starting to see this disparity reflected in the literature on ECMO in pregnancy. Let's move into the indications for peripartum ECMO.  The pulmonary and cardiac indications for ECMO in pregnant patients are similar to the non-pregnant population with a few obstetric-specific causes noted here.  ARDS does account for the bulk of respiratory ECMO in this group, but hypertensive disorders  of pregnancy and fluid overload, often described in the context of large-volume blood loss and resuscitation, are described in this population as well.  Cardiac indications certainly overlap with those described in the non-pregnant ECMO population,  but notably, amniotic fluid embolism, which is a syndrome associated with profound inflammation, hemodynamic collapse, and coagulopathy, is a unique entity to the peripartum period.  Hi to everyone. This nice graphic comes from a case series out of the University of Maryland that reports a decade of their institutional experience with peripartum ECMO.  They had a total of 21 patients, a third of which were cannulated antepartum and two-thirds were postpartum, with one cannulated intraoperatively during a cesarean delivery.  Their indications for VV support included aspiration, pneumonia, and sepsis. BA indications were largely pulmonary embolism and cardiomyopathy. Hi.  When I started looking into ECMO in pregnancy, it seemed to me that the timing of when women require support is really an important concept.  It's likely that different time periods in pregnancy have different pathologies. Sure enough, when we broke this down in our study, we found key differences in the groups  for indications, type of support, and duration. Antepartum ECMO includes those who were cannulated while still pregnant.  The most common indications are ARDS, heart failure, pulmonary hypertension, and cardiac arrest. Most of the cases are VV, and the duration of support was between one to two weeks.  The survival in this group was quite high, and fetal survival in this group was 65%. I'll talk more about fetal outcomes later on.  Intrapartum ECMO is anyone cannulated on postpartum day zero, either peri-delivery or within the first day after delivery.  The most common diagnoses include cardiac arrest in more than half the cases, heart failure, amniotic fluid embolism, pulmonary hypertension, MPE.  Three quarters of these women had VA cannulation, and the duration was much shorter, with less than a week on average.  The survival is actually highest in this group, and this supports the use of ECMO in these catastrophic events, because you can see they actually do very well.  Postpartum ECMO is a bit more of an amorphous group. It includes anyone from postpartum day one to day 42.  The majority of cases were ARDS, followed by peripartum cardiomyopathy and heart failure. Most of these patients had VA cannulation. The duration was longest in this group.  Unfortunately, survival was lowest. This group of women is really the most heterogeneous, and so the poorer outcomes are closer to those of the general up-adult population.  That may be because these patients are suffering non-pregnancy-related issues, like heart failure and underlying cardiac disease.  I want to talk now about pregnancy-specific considerations in this unique group of patients. This talk would be remiss without a brief discussion of the maternal physiologic changes  that have to be considered when you manage these patients in the ICU, and especially on ECMO. Cardiovascular changes reflect a hyperdynamic state.  The increase in blood volume and cardiac output is accommodated by an increase in heart rate, stroke volume, and contractility.  Cardiac and pulmonary vascular resistance drop, with an overall decrease in blood pressure, which reaches its nadir in the second trimester.  Progesterone mediates the majority of the respiratory changes of pregnancy. This notably includes an increase in minute ventilation, driven by an increase in tidal volume.  This is due to an expansion of the anterior-posterior diameter of the chest wall throughout the course of pregnancy.  The result of all of these changes is a compensated respiratory alkalosis, with an increase in renal bicarbonate excretion.  There's also an increase in oxygen consumption and a decrease in functional residual capacity, both of which contribute to a much smaller reserve and poor tolerance for hypoxia.  Managing these patients comes with nuances. Pregnant patients receiving ECMO should maintain a PaO2 of greater than 70 and an oxygen saturation  goal of greater than 95% in order to ensure adequate fetal oxygenation. Adequate PaO2 is also required to maintain fetal acid-base status.  Fetal CO2 clearance is dependent on a transplacental gradient. As baseline PaCO2 in pregnancy is 30 to 32 millimeters of mercury, sweeps should be titrated  to maintain a mild respiratory alkalosis while avoiding significant hypocapnia, which can result in fetal alkalosis, uterine artery constriction, and a leftward shift of the  fetal oxygen dissociation curve, with resultant fetal hypoxia. Severe hypercapnia with a PaCO2 greater than 60 should also be avoided, since this limits  oxygen binding to fetal hemoglobin due to a rightward shift of the dissociation curve. A technical consideration related to pregnancy, particularly in the third trimester, is that  of large vessel compression from the gravid uterus. This figure comes from a case report where surgeons found it challenging to cannulate  a woman who was 32 weeks pregnant with septic shock and ARDS. They had difficulty in advancing a cannula through the guide wire in the vena cava.  This is an important point for any pregnant patient greater than 20 weeks gestation in an ICU.  Uterine displacement 15 to 30 degrees is imperative to prevent aortic caval compression and resulting hypotension.  Even once the cannulas are placed, the uterus may continue to impair adequate flow rates through this aortic caval compression, so lateral tilt should be maintained.  These guidelines come from the Society for Maternal Fetal Medicine, or SMFM, for management of COVID ARDS, but the principles can really be applied to ARDS and the pregnant patient  in general. As you can see, proning, neuromuscular blockade, and inhaled pulmonary vasodilators are all appropriate rescue therapies in pregnant patients.  For ARDS, most centers use gas exchange criteria aligned with those of the AEOLIA trial to determine candidacy for ECMO.  Pregnant patients should be evaluated with the same lens with the added consideration of gestational age.  After 32 weeks, delivery may be considered if it will allow further optimization of care of the mother, and there may be a benefit in reducing the physiologic demands in select  patients. If the patient is less than 32 weeks gestation, delivery is unlikely to provide substantial benefit to the respiratory mechanics, so ECMO should be considered.  High patient stability and adequacy of physiologic support is often reflected in the fetal status.  Following initiation of ECMO, there should be routine fetal monitoring, and practices in frequency vary institutionally.  Daily checks for fetal viability are the most frequently used method of fetal monitoring. Continuous fetal heart rate monitoring can offer reassurance if it's normal, but it's  not routinely performed since emergent delivery for acute fetal concerns is often not an option in critically ill patients.  The frequency of hemodynamic fluctuations in these patients may warrant more frequent fetal  assessment for reassurance of well-being, assessment for evidence of labor, and to titrate hemodynamic parameters to ensure adequate fetal perfusion.  One also has to keep in mind the effect of sedative medications to the mother on this assessment of fetal well-being.  After 22 weeks gestation, there should be a discussion with the neonatal team as to whether or not resuscitation is appropriate.  And beyond 24 weeks gestation, an experienced neonatal team should be present. When delivery is indicated, a vaginal delivery is generally prioritized for all patients  who lack an obstetric indication for a cesarean delivery. The decision to deliver requires close communication between MFM, ICU, and neonatal teams.  Equipment for delivery and obstetric medications should be readily available in any ICU that manages pregnant patients on ECMO.  So how do these patients do, and how does that compare with the non-pregnant population?  Pregnancy introduces a complexity around the potential need for delivery, which could have major hemorrhagic complications.  And in addition, clinicians have historically raised concern about the hypercoagulability of pregnancy and thrombotic complications.  Thankfully, the most recent large retrospective reviews have been able to delineate these risks with slightly more clarity than the prior smaller studies.  Reported bleeding complications are slightly higher in the peripartum population. These range from 30% to 50% compared to the reported risk in general adults that's closer to 30%.  This really shouldn't be that surprising given the clinical scenarios that peripartum patients present with and the peri-delivery nature of these patients.  Thankfully, a large retrospective study with 280 peripartum patients showed that despite an increased risk of hemorrhage in this group, bleeding complications on ECMO was not found  to be a risk factor for mortality. The incidence of thrombotic events ranges from 3% to 7% across the literature, which  is comparable to the rate reported in the general population that ranges widely from 2% to 14%.  One study looking at 88 patients who underwent ECMO for peripartum cardiomyopathy found that  there was a slightly higher incidence of circuit clotting when compared to non-pregnant patients. However, this also was not associated with an increase in mortality.  This table is, again, from the National Inpatient Sample Study. Notably, in this study, they found that compared with non-obstetric ECMO, peripartum ECMO is  associated with a significantly increased risk of non-traumatic hemoperitoneum and, when looking at the adjusted odds ratio, an increased risk of venous trauma embolism.  Despite this, their finding was that obstetric ECMO is still associated with decreased in-hospital  mortality compared with non-obstetric ECMO. Overall, for all peripartum patients who require ECMO support, the maternal survival ranges from 70% to 77%.  This is distinctly higher than the survival of adults overall. This may be because, on average, these patients are younger, have fewer comorbidities, and  it may also be that the pathologies for which they require ECMO are more reversible than the general population.  Maternal survival rates are a bit lower, and this should not come as a surprise. Women who require ECMO while pregnant are critically ill, and the fetus is exposed to  extreme physiologic derangements during this time period. These pregnancies are also more likely to end with preterm delivery and a higher incidence of NICU stay.  Over the years, multiple studies have shown increasing use and improving survival in cohorts of maternal patients requiring ECMO.  This figure on the left comes from the NIS study demonstrating decreased mortality in both non-obstetric and obstetric ECMO between 1999 and 2014.  The figure on the right comes from a retrospective study using the ELSO registry that identified 280 peripartum patients over a 21-year time frame.  These studies demonstrate decreasing maternal mortality over the last two decades. This figure comes from a retrospective study utilizing the ELSO registry looking at 88  women who received VA ECMO for peripartum cardiomyopathy. It demonstrated that maternal survival has improved over the last decade and that our  use of mechanical circulatory support has expanded. This table comes from that ELSO registry study.  They found that patients who were cannulated for pulmonary indications did better than those who were cannulated for cardiac indications.  And as expected, patients who required eCPR had the lowest survival. Having said that, the survival in their study was also quite high for eCPR at almost 55%,  compared with all comers in the ELSO registry where the survival to discharge for eCPR is only 29%.  So for reference, the survival to discharge for general adult ECMO for cardiac indications is 44%. In this population, it was found to be 67.5%.  Survival to discharge for general adults for pulmonary indications for ECMO is 59% versus 76% in this study.  One note from this study, patients who had renal complications did have a higher risk of mortality than those who did not.  We can't ignore the impact of COVID-19 on peripartum patients. This figure comes from the retrospective cohort study of ECMO for COVID-19 ARDS using the  ELSO registry that identified 100 peripartum women. It demonstrated higher survival and fewer ECMO-related renal complications than the  non-pregnant propensity score-matched cohorts. When you look at the survival based on indications for ECMO, it can provide insight into which  conditions may have better outcomes in this population. This table demonstrates that survival can vary based on the indication.  Women who were cannulated for cardiac arrest had a survival rate of nearly 88%. When you compare this with one of the largest population-based studies on maternal cardiac  arrest, the survival in that study was only 59%. ECMO may have a role in improving clinical outcomes in this context with eCPR.  Pulmonary arterial hypertension has a very poor prognosis in pregnant women. This review found a survival of 50%.  Anticipatory interventions may improve outcomes in these high-risk patients. There are several reports of antepartum placement of venous and arterial sheaths for the potential  rapid initiation of ECMO in patients with severe pulmonary hypertension. The current literature regarding amniotic fluid embolism reports a wide range of survival,  anywhere from 40% to 90%. This study found an overall survival of 52%, which is consistent with those data.  However, the rate of neurologically intact survival was significantly higher compared to the non-ECMO literature.  Before ending, I want to touch on areas of research to better inform us on this group of patients.  There is still a huge amount of work to be done looking at ECMO in pregnancy, given how rare this is.  There is a move to create a centralized registry of peripartum patients who've required ECMO.  Among many questions of interest, we're looking at where patients are cannulated, how they deliver, anticoagulation, and a better review of the neonatal outcomes.  Another important topic of interest is the concept of anticipated ECMO, women at risk of decompensation who may not tolerate the stress of delivery and require ECMO support,  hopefully transiently. These patients may include those with pulmonary hypertension or underlying heart disease.  It would seem that women who have been preemptively evaluated and have ECMO on standby may do better than those who crash onto ECMO.  But of course, this hasn't been formally studied. In addition to potentially making it smoother to transition to ECMO support, just like in  the CSER trial, it may help to have a team of ECMO experts consult to prevent a need for ECMO support. Please feel free to contact me at my email listed here.  I welcome any questions or input about this topic. I would love to meet anyone who shares my passion for ECMO and pregnancy. Hi. Thank you so much for taking your time to listen.  I look forward to meeting more SCCM members at our next in-person Congress.

Video Summary

Dr. Emily Nahum, a critical care physician and obstetric anesthesiologist, discusses the use of ECMO (extracorporeal membrane oxygenation) in pregnant women. Peripartum ECMO is a rare but important treatment option for pregnant patients with severe respiratory or cardiac failure. The use of ECMO in pregnancy has increased significantly over the past decade, especially during the H1N1 pandemic. Pregnant patients requiring ECMO are often healthier than non-pregnant patients and have different indications, such as ARDS, hypertensive disorders, and amniotic fluid embolism. Maternal physiological changes during pregnancy must be taken into account when managing these patients. ECMO support should aim to maintain adequate oxygenation and fetal wellbeing. The survival rates for peripartum patients on ECMO are higher than in the general adult population, although bleeding and thrombotic complications are slightly increased in this group. Future research aims to create a centralized registry of peripartum ECMO patients and explore the concept of anticipated ECMO for high-risk individuals.

Asset Subtitle

Procedures, Obstetrics, 2022

Asset Caption

Extracorporeal life support (ECLS) has been established as an advanced therapeutic alternative for respiratory failure, cardiogenic shock, and failure to wean off cardiopulmonary bypass. With a global increase in expertise, extracorporeal membrane oxygenation (ECMO) and other ECLS options have been explored in other patients’ indications. This session aims to increase understanding of future possibilities and indications for ECMO and ECLS toward improving individual outcomes.

Meta Tag

Content Type Presentation

Knowledge Area Procedures

Knowledge Area Obstetrics

Knowledge Level Advanced

Learning Pathway Cardiothoracic Critical Care

Membership Level Select

Tag Extracorporeal Membrane Oxygenation ECMO

Tag Maternal Fetal Medicine

Tag Cardiothoracic Critical Care

Year 2022

Keywords

ECMO

peripartum ECMO

pregnant patients

respiratory failure

cardiac failure

	Society of Critical Care Medicine 500 Midway Drive Mount Prospect, IL 60056 USA Phone: +1 847 827-6888 Fax: +1 847 439-7226 Email: support@sccm.org	Contact Us About SCCM Newsroom Advertising & Sponsorship DONATE	MySCCM LearnICU Patients & Families Surviving Sepsis Campaign Critical Care Societies Collaborative	GET OUR NEWSLETTER
© Society of Critical Care Medicine. All rights reserved. \| Privacy Statement \| Terms & Conditions The Society of Critical Care Medicine, SCCM, and Critical Care Congress are registered trademarks of the Society of Critical Care Medicine.