Experience of extracorporeal cardiopulmonary resuscitation in a refractory cardiac arrest patient at the emergency department

Background Extracorporeal cardiopulmonary resuscitation (ECPR) is a method to improve survival outcomes in refractory cardiac arrest. Hypothesis This study aimed to determine the associated factors related to outcome and to analyze the post‐ECPR management in patients who received ECPR due to nonresponse to advanced cardiac life support (ACLS). Methods This was a retrospective analysis based on a prospective cohort. Cardiac arrest patients who received ECPR in our emergency department from May 2006 to December 2017 were selected from the prospective cohort. Patients who received ECPR for rearrest were excluded. The primary outcome was survival to discharge. Results ECPR was attempted in 100 patients who did not respond to ACLS. Fourteen patients survived to discharge, and 12 (85.7%) patients showed good neurologic outcomes. The rate of survival to discharge decreased according to increasing age and ACLS duration. Age, presence of any return of spontaneous circulation (ROSC) during ACLS, and prolongation of ACLS were associated factors for survival discharge in ECPR patients. Fourteen patients required distal perfusion catheters, and 35 patients received continuous renal replacement therapy (CRRT). The proportion of death was the highest within 24 hours after ECPR as 57.0%. Conclusions The early transition from ACLS to ECPR may improve the ECPR outcomes. In addition, good outcomes are expected for ECPR performed after refractory arrest if the patient is young and experiences an ROSC event during ACLS. In post ECPR management, the majority of mortality events were occurred in the early period, and distal perfusion catheter and CRRT were frequently required.

patients receiving conventional CPR. Refractory cardiac arrest was defined as no survival event within 30 minutes of advanced cardiac life support (ACLS). 4 Many studies have suggested the indicators of extracorporeal CPR (ECPR) in refractory cardiac arrest, including brief no-flow time, reversible cause of arrest, and optimal transition time from ACLS to ECPR. [5][6][7][8][9] Because ECPR is a highly invasive procedure that requires many medical resources and well-coordinated hospital system, the application of ECPR in arrest patients without sufficient medical information on the presumed etiology of arrest at the emergency department (ED) should be carefully considered. 10 Therefore, in addition to the reported indicators of ECPR, post-ECPR management is also important for good outcome of ECPR. The ED is the most common site of CPR for patients with out-of-hospital cardiac arrest (OHCA) or arrest occurring within a short time after ED presentation.
This study aimed to determine the factors associated with good outcome and to analysis the post-ECPR management in patients who received ECPR due to nonresponse to ACLS to further elucidate the usefulness of ECPR for refractory arrest in ED.

| Study design and setting
This was a retrospective analysis based on a prospective cohort study.
We reviewed the CPR registry that prospectively collected the data of cardiac arrest patients who underwent CPR at the ED of one tertiary hospital from May 2006 to December 2017. Patients who received ECPR due to cardiac arrest were selected from the cohort. Patients who received ECPR for rearrest were excluded ( Figure 1).
The Institutional Review Board (IRB) of the Korea University Medical Center approved this retrospective analysis (IRB no. 2018AN0451).
The IRB of the Korea University Medical Center waived the requirement for informed consent. The data were also analyzed anonymously.

| Data collection of CPR registry
A CPR coordinator prospectively collected data for the CPR registry based on the Utstein style guidelines. 11,12 The registry included demographic data, if the arrest was witnessed; location of arrest; incidence of suspected or confirmed trauma; presumed arrest time; presence of bystander CPR; first documented arrest rhythm; any return of spontaneous circulation (ROSC); presence of survival event after ACLS; presence of ECPR; presumed cause of arrest; application of therapeutic hypothermia, coronary angiography (CAG), or percutaneous coronary intervention (PCI); and cerebral performance category (CPC) score at discharge.

| Indications and management of ECPR at the ED
Our indications for ECPR for adult cardiac arrest in the ED were (a) age ≥ 18 years; (b) sudden cardiac arrest with presumed correctable causes; (c) witnessed arrest with or without bystander CPR; or (d) no-flow time (time interval from presumed arrest to CPR started) was expected to be short, even for unwitnessed arrest. The contraindications for ECPR were (a) arrest due to clearly uncorrectable causes; (b) presence of a terminal illness or malignancy; (c) suspected or confirmed traumatic origin of arrest; and (d) no informed consent from the family. The ECPR team was activated by the emergency physician in cases when the patient met the inclusion criteria and required ACLS duration >20 minutes.
Depending on the patient's body size, a 15-to 17-Fr arterial catheter and a 21-to 23-Fr venous catheter were inserted into the femoral artery and vein while maintaining chest compressions. The flow rate was initially set at 2.5 to 3.0 L/minute. Anticoagulation with heparin was administered immediately after initiation of ECPR. After implementation of ECPR, CAG was performed as soon as possible in cases of suspected acute coronary syndrome, and insertion of bridge catheter for the prevention of lower limb ischemia and continuous renal replacement therapy (CRRT) were performed if required. Severe complications of ECPR were defined as intracranial hemorrhage, massive hemothorax, malposition of catheter, and severe local bleeding or lower limb ischemia.

| Data analysis
The primary end point was the rate of survival to discharge. In addition, we analyzed the incidence of organ donation as an outcome of ECPR. The resuscitation and ECPR-related factors between the survivor and nonsurvivor groups were compared. Data on mean arterial blood pressure (MAP) and serum lactate concentration were compared between the two groups. We analyzed the rate of survival to discharge based on increasing of age and increasing of ACLS duration. Selection of study patients and outcomes. Rearrest was defined as recurrent cardiac arrest within 24 hours after survival event (sustained return of spontaneous circulation >20 minutes). Refractory arrest was defined as the nonachievement of survival event within 30 minutes of advanced cardiac life support. Successful commencement of ECPR implies that spontaneous heart beating was obtained under the support of ECPR. Successful weaning from ECPR was defined as survival to 24 hours after weaning of ECPR. ED, emergency department; OHCA, out-of-hospital cardiac arrest; CPR, cardiopulmonary resuscitation; ECPR, extracorporeal CPR; CPC, cerebral performance category Multiple logistic regression analysis was performed to determine factors associated with survival among CPR-related factor before ECPR implementation. We analyzed the incidences of severe complications related with ECPR, insertion of distal perfusion catheter, and CRRT.
The changes of MAP and serum lactate during 24 hours after implementation of ECPR were analyzed. All statistical analyses were performed using SPSS version 20.0 (IBMSPSS, Chicago, Illinois), and the data were expressed as mean with SD. Continuous and categorical variables were compared using t test and Fisher exact test, respectively. Statistical significance was determined by P < 0.05.

| Study population and outcomes
During the study period, 100 patients received ECPR for sudden arrest at the ED. This included 80 patients with previously reported data 13 and an additional 20 patients that were added to this study.
Our reported study 13 has researched ECPR in both refractory arrest and rearrest, but in this study, only the refractory arrest was studied in increasing the study duration. There were 75 and 25 patients for OHCA arrest and arrest at ED, respectively. Nineteen patients (19.0%) were weaned off from the ECPR successfully, and 14 patients (14.0%) were discharged alive ( Figure 1). Twelve patients had a CPC 1 at discharge ( Figure 1). Three patients had controlled organ donation after brain death at third, sixth, 13th day after ECPR.

| Comparison with CRP-related factor between survivors and nonsurvivors
The age range and ACLS duration before ECPR were 18 to 85 years and 22 to 159 minutes, respectively. No significant differences were found in terms of sex, location of arrest, presence of a witness, presence of bystander CPR, and cause of arrest CPR between survivors and nonsurvivors (Table 1). Survivors were younger than nonsurvivors (P = 0.015). Survivors showed a high incidence of any ROSC during ACLS and short duration of ACLS compared with the nonsurvivors (P = 0.130, 0.061, respectively). The level of serum lactate before ECPR did not have any difference between the two groups; however, the rate of lactate clearance was higher in survivors than in nonsurvivors (P = 0.071).

| Multiple logistic regression on CPR-related factors associated with survival to discharge
We grouped the patients based on the quartile range of age and ACLS The odds ratio of survival discharge was significantly low in the old age group compared with that in the 18 to 45 years age group. Prolongation of ACLS also decreased the probability of survival (Table 2).

| Post-ECPR implementation management
The median duration of ECPR was of 21.2 hours (minimum to maximum: 0.3-329.1 hours). The immediate recovery of spontaneous heart beat after ECPR was a significant factor for predicting survival  The proportion of mortality cases was highest within 24 hours after ECPR as 57.0% ( Figure 2). The MAP of survivors were significantly higher than those of nonsurvivors, and survivors showed the tendency of rapid reduction in serum lactate level (Figure 3).

| DISCUSSION
The ED is the most common place to perform CPR for sudden cardiac arrest occurring at out-of-hospital setting or at the ED during evaluating the patient's condition. Despite sudden arrest, many patients do not respond to ACLS. 14 If the patient has a reversible cause of arrest and brief no-flow time, we must consider ECPR. Particularly, our study showed that if the patient was not old and had presence of any ROSC during ACLS, early transition from ACLS to ECPR should be considered for good outcome from ECPR when the patient does not respond to ACLS within an indicated time or had a refractory arrest. The metaanalysis of Kim et al 15 also showed that the studies including predefined criteria with witnessed arrest or shockable rhythm showed better outcome in ECPR, although no single criterion was a dominant factor in determining the outcome. Our study will improve the detail of these required conditions of ECPR in refractory cardiac arrest. We have previously reported on the prognostic indicators of survival following ECPR and survival prediction model deriving from it. 13 We reported a prediction model using age, initial rhythm, any ROSC, and CPR duration. 13 We have expanded our data to the post ECPR period.
In this study, the most common complication was the local bleeding from catheter site, followed by a malposition of the catheter,    However, the rapid clearance of serum lactate was related with good outcome. In addition, our study showed that a high blood pressure was related with survival to discharge. We thought that the rapid recovery from lactic acidosis under the optimal hemodynamic support will be more important than the single serum lactate level measured before ECPR for a good outcome in ECPR patients.

| Study limitation
This study has several limitations that require consideration. First, the standard CPR guideline has been changed during study period.

| CONCLUSION
ECPR in the ED may be considered as an alternative feasible method for patients who had a sudden cardiac arrest unresponsive to conventional CPR, as bridge until evaluation and definitive care for refractory sudden arrest through an early improvement in the hemodynamic and physiologic status. However, rigorous criteria for candidate selection are necessary to predict good outcomes of ECPR. Our data showed that the early transition from ACLS to ECPR may improve the outcome of ECPR. In addition, if the patient is young and has any ROSC event during ACLS, good outcome will be expected from ECPR for refractory arrest. In post ECPR management, the majority of mortality FIGURE 3 Mean arterial pressure (MAP) and serum lactate level of survivors and non-survivors according time. MAP was significantly higher in the survivors than in the nonsurvivors at 2, 6, and 24 hours post ECPR implementation. Serum lactate level was significantly lower in the survivors than in the nonsurvivors group at 24 hours post ECPR implementation. The serum lactate level tended to decrease rapidly in the surviving group. * P < 0.05. ECPR, extracorporeal cardiopulmonary resuscitation cases were occurred in the early period and distal perfusion catheter