Use of Impella heart pump for management of women with peripartum cardiogenic shock

Abstract Background Percutaneous mechanical circulatory support (MCS), such as the Impella heart pump is a valuable option for cardiogenic shock (CS), although the use of Impella in CS due to peripartum cardiomyopathy (PPCM) is limited. Objective To assess outcomes in women with PPCM supported with an Impella device from the global catheter‐based ventricular assist device (cVAD) Registry. Methods and Results A total of 15 women with PPCM supported with Impella devices between November 2008 and October 2015 were included. Of the 15 women, five were treated at Hannover medical school and have been reported previously, the rest were managed at various US hospitals. The mean age was 30.0 ± 7.34 years, eight women were Caucasian, and seven were African‐American. The occurrence of PPCM was post‐delivery in eight (53.3%), at delivery in one (6.7%), and during gestation in four women (26.7%). At admission, all women had severe heart failure with a mean ejection fraction of 14.7 ± 6% and 13 women (86.7%) presented with CS. Prior to Impella, 100% were mechanically ventilated, 79% received inotropes/vasopressors, 20% supported with IABP, and 27% received veno‐arterial extracorporeal membrane oxygenation (VA ECMO) during Impella support. Two women (13.3%) died, and 13 (87.7%) survived to discharge. Eight women (53.3%) had a recovery of native heart function and six (40%) were bridged to durable left ventricular assist device (LVAD). Conclusion MCS with Impella devices can be successfully used as a bridge to early improvement, heart recovery, or successful implantation of durable LVAD in women with PPCM complicated by severe LV dysfunction.

age, increased rate of multifetal pregnancies, and increased recognition of the disease. 2 PPCM can be associated with severe complications including cardiogenic shock (CS), cardiac arrest, and death. 2,3 A recent meta-analysis of 46 studies from 13 countries on women with PPCM reported a global mortality rate of 9%. 4 Although recovery of LV function is frequently observed, the continued deterioration of cardiac function is reported in about 5%-20% and can lead to severe heart failure, cardiogenic shock, and mortality. 3 Early identification and hemodynamic stabilization is needed to increase the likelihood of myocardial recovery, particularly in PPCM complicated by CS. While medical management using catecholamines may be considered, their effect may be limited in patients with CS, and their use is often associated with adverse events effects including tachycardia, hypotension, myocardial ischemia, and arrhythmias. 5 Percutaneous mechanical circulatory support (MCS) devices are a promising therapeutic option given their ability to improve hemodynamics and tissue perfusion. 6 Recent reports suggest significant recovery with MCS in patients with CS due to PPCM not responding to medical therapy.
The current American and European guidelines suggest early use of MCS in patients with CS. 7,8 The Impella devices (Abiomed, Danvers, Massachusetts) are transvalvular microaxial pumps that can be rapidly deployed in the catheterization lab and provide LV unloading with forward blood flow of up to 5 L/min. 6 Given the low prevalence, published reports on the use of Impella devices in PPCM is limited. 9, 10 We report on 15 women with PPCM supported with an Impella device from the global catheter-based ventricular assist device (cVAD) Registry. This study, to our knowledge, is the largest reported series to date on the use of Impella devices in women with severe heart failure and CS due to PPCM.

| Study population
The current study is a retrospective analysis of 15 patients with PPCM undergoing LV unloading with Impella devices between November 2008 and October 2015, identified in the global cVAD registry. Of the 15 patients, five patients were treated at Hannover medical school (a cVAD site) and have been previously described. 10 The remaining patients were enrolled at eight cVAD sites within the United States.
The global cVAD registry is an ongoing registry of patients who received Impella support or attempted support in North America and Europe since the inception of the registry in 2009 to date. 11 The cVAD Registry was designed by an Executive Steering Committee that oversees its ongoing conduct. The registry protocol was reviewed and approved by the Institutional Review Board at each participating site.
Sites are invited to report all consecutive Impella cases without preselection of indication or patients. In order to avoid patient selection bias, a process was developed to reconcile the site's utilization of the device (commercial database) with reporting in the clinical cVAD registry. Patients that were identified as having received an Impella device in the commercial database were expected to be reported in the cVAD Registry database; otherwise, sites were notified of the obligation to enter and report the cases to ensure consecutiveness.

| Patient treatment and follow-up
Data were abstracted from the medical record to a standard electronic case report form by the sites study coordinators who were centrally trained. Information was collected on the patient's demographic characteristics, medical history, clinical presentation, hemodynamic, echocardiographic, and angiographic characteristics, treatment during hospitalization, hospital discharge status, and follow-up status.
CS was defined as hypotension (systolic blood pressure < 90 mm Hg or need for inotropes/vasopressors to maintain systolic blood pressure > 90 mm Hg) and/or end-organ hypoperfusion indicated by altered mental status, clammy skin, or lactate >2 mmol/L after adequate correction of preload. 10 All patients were admitted to the intensive care unit where hemodynamic support with an Impella device was initiated. Impella devices were implanted via either femoral or axillary access and the correct position was verified every 8 hours by transthoracic echocardiography. Hemodynamic measurements were performed using a Swan-Ganz catheter. Clinical decisions regarding device repositioning due to reduced blood flow, device explanation, and early weaning of device were made by the treating physician. defined PPCM as idiopathic cardiomyopathy presenting with HF secondary to LV systolic dysfunction towards the end of pregnancy or in the months following delivery, where no other cause of HF is found. 11 Recovery was defined as successful weaning of Impella, followed by survival to discharge without the need for additional MCS support or heart transplant.

| Statistical analysis
Data are represented as n (%), mean ± SD for variables with normal distribution or median and range (minimum-maximum) for non-normal distribution. Change in hemodynamics before and during Impella support and in LVEF from baseline to discharge and post-discharge follow-up was analyzed using the paired t test. A two-sided P-value of less than .05 was considered statistically significant and all statistical tests were performed using SAS version 9 (SAS Institute, Cary, North Carolina) software.
Prior to Impella support, 78.6% patients (11/14) were supported with inotropes or vasopressors and three patients with an intra-aortic balloon pump (IABP). Among the 15 women receiving Impella support, Impella 2.5 was used in 33.3%, Impella CP in 53.3%, and Impella 5.0 in 13.3% (Table 3). The timing of onset of CS to Impella support was 15.6 ± 7.7 hours and duration of Impella support was 265.5 ± 460.6 hours. Four women received biventricular unloading with VA-ECMO and Impella and two of them received VA-ECMO prior to Impella without significant improvement in hemodynamics necessitating concomitant therapy with Impella.
T A B L E 3 Admission, procedural, and on-support characteristics and outcomes

| Clinical outcomes
Survival to discharge was 86.6% (13/15) ( Table 3). Recovery of native heart functioning occurred in 53.3% (8/15), 40.0% (6/15) were bridged to another ventricular assist device, and two patients expired in the hospital (one patient who was admitted with acute liver failure developed irreversible disseminated intravascular coagulopathy and another patient died of septic shock after bridged to an LVAD).
Among the 11 survivors to discharge with data available, the average LVEF increased from the baseline value of 15 ± 7% to 28 ± 18% at discharge (P = .04). Follow-up at 6 months was available in five patients who were treated medically in addition to Impella support.
No myocardial infarction or stroke occurred until discharge. Two of 15 women (13.3%) had bleeding requiring transfusion and three additional women (20.0%) received transfusion due to anemia (Table 5). Need for new renal replacement therapy, infection, and hypotension during support was observed in three women each.
Hemolysis occurred in three patients (20.0%) and limb ischemia in one patient (6.7%) who also received ECMO support. Increase in device purge pressure was noted in one patient with no consequence as the patient was successfully bridged to a durable LVAD. As described previously, 9 adherent thrombotic material was found at the impeller housing in one case after device explanation, although without evidence of thromboembolism. One device malfunction was reported in a woman supported with an Impella 5.0 beyond the approved duration of support of 6 days (error message of "Impella sensor fail" displayed on the console on day 20 of support). No intervention was required and the Impella support was successfully continued for 43 days when the patient was transitioned to a surgical VAD.

| DISCUSSION
PPCM is a pregnancy-associated, idiopathic form of cardiac dysfunction that can lead to severe morbidity and mortality. The incidence of this condition is on the rise, 12 and it is a leading cause of nonobstetrical maternal mortality in the United States. 13 Although the majority of women demonstrate either partial or complete recovery after the diagnosis, the adverse outcome with severe and lasting morbidity and mortality remain unacceptably high. A subgroup of patients, mostly with a large myocardial insult (Ejection Fraction (EF) < 30%) at the time of diagnosis demonstrate further deterioration leading to severe heart failure, cardiogenic shock, and mortality. [14][15][16] A recent algorithm for the management for severe acute PPCM, suggests optimization of oxygenation using noninvasive or invasive ventilation in patients with hypoxemia followed by the use of inotropic support. 5 Analysis of the German PPCM registry suggested an unfavorable effect of dobutamine (a β1-adrenergic receptor agonist) in women with PPCM. 17  and headache without reduction of all-cause mortality with levosimendan compared to dobutamine. 21 The use of inotropic drugs in patients with CS is often limited due to the development of tachycardia, hypotension, and arrhythmias. 21,22 In addition, these drugs should be used with caution during pregnancy as fetal safety is not known due to limited information. 23 The recommendations of the PPCM working group for the man- Given the hypotensive and prothrombotic effect of bromocriptine, its use in women with CS complicating PPCM may be limited. 5,9,27 Four of the women in this study presented with severe heart failure prior to the delivery. The management of a woman presenting with severe PPCM prior to fetal maturation is challenging. The European guidelines recommend immediate delivery, 5 which may be associated with a high risk of fetal mortality or life-long complications. While the wellbeing of the mother is a top priority, in this condition, maternal stabilization may allow for delaying of delivery and improving fetal prognosis. 28 Percutaneous MCS devices are increasingly used in the treatment of CS. The Impella devices are transvalvular micro-axial flow pumps that entrain blood from the LV and pump it into the aorta in series, thus unloading the LV. These devices are a valuable tool in the management of CS due to their ability to maintain vital organ perfusion, decrease LV wall stress and myocardial oxygen consumption, thereby enhancing the likelihood of ventricular recovery. 6,29 In fact, multiple studies have demonstrated that early initiation of Impella support within 90 minutes of CS onset was associated with higher survival. [30][31][32] An additional advantage of early Impella use in CS is the decrease in the use of inotropic agents, thus, reducing the effects of these cardiotoxic drugs. 33 In comparison to the growing evidence base of benefits of the Impella support in CS due to myocardial infarct, very few reports exist on the use of Impella in CS due to PPCM. In this study, the majority of women were diagnosed with PPCM post-delivery with a mean LVEF of 14%. This group of women was previously shown to have a low likelihood of recovery. 34 In addition, most of the women in the present study were in CS and received mechanical ventilation and inotropic support. Despite the critically ill nature of these women, >85% survived to discharge and 1 in two women had recovery of their native heart function, demonstrating the potential of MCS with Impella for hemodynamic stabilization and as a bridge to recovery or implantation of a durable cardiac assist device in women with CS due to PPCM.
Published case reports also suggest the potential stabilization of patients with CS due to PPCM with VA-ECMO. 35 In contrast to VA-ECMO, a relatively low incidence of manageable adverse events was observed with the use of the Impella device in the 15 patients reported in this study.

| Limitations
Limitations of this study include missing data in some patients, lack of a control group, and the use of multiple and non-uniform therapeutic interventions, all inherent to any retrospective registry-based analysis.
Given the observational nature of this study, modest series of patients, and the concomitant use of inotropes, bromocriptine and multiple MCS, it was not possible to evaluate the effect of the Impella alone and hence the results should be considered hypothesis-generating. Since PPCM is a rare disease and CS complicating PPCM occurs in very few women, the results of this study provide valuable insights into the use of MCS in this group of critically ill women.

| CONCLUSIONS
To the best of our knowledge, this report is the largest series to date of hemodynamic support with Impella heart pump in women with CS complicating PPCM. The results of this study are encouraging and demonstrate that the use of Impella for LV unloading can be used for hemodynamic stabilization and as a bridge to recovery or successful implantation of durable MCS in patients with severe myocardial insult and CS due to PPCM.

ACKNOWLEDGMENTS
The cVAD registry is funded by Abiomed Inc. (Danvers, MA).