ASAIO Journal 2017
Clinical Critical Care
Predictors of Survival for Nonhighly Selected Patients Undergoing Resuscitation With Extracorporeal Membrane Oxygenation After Cardiac Arrest DIRK PABST,* ALY EL-BANAYOSY,† BEHZAD SOLEIMANI,* AND CHRISTOPH E. BREHM*
In several case reports and case series, extracorporeal membrane oxygenation during chest compression (CPR) has been shown to be a reasonable tool to improve outcome of patients under resuscitation. Although recommendations for ECPR include younger patients with shockable rhythm and short previous CPR-time, it remains unclear if nonhighly selected patients have a similar outcome. Aim of this study was to determine outcome in our nonhighly selected patient population treated with ECPR and investigate possible predictors of survival. We made a retrospective single-center study of adults who underwent ECPR for in-hospital cardiac arrest between June 2008 and September 2016. Outcome and predictors of survival were identified. In this period of time, 59 patients underwent ECPR due to cardiac arrest. Fifteen patients (25.4%) survived discharge of which all had a good neurological outcome (cerebral performance category, ≤ 2). Survival to discharge of patients with shockable rhythm (ventricular fibrillation or ventricular tachycardia) was 40.7%. Serum lactate ≥ 8, pulseless electrical activity (PEA) or asystole and male gender could be identified as predictors for low survival rate. Age, body mass index, renal replacement–dependent kidney injury had no significant influence on survival outcome. Mean CPR-time was 41.1 minutes (interquartile range, ±29.25 minutes). Extracorporeal membrane oxygenation seems to be a useful tool to improve the outcome of CPR also in nonhighly selected patients when compared with CPR alone and could be considered in patients with refractory cardiac arrest also after longer previous CPR-time. Serum lactate and heart rhythm should be taken into account for patient selection. ASAIO Journal 2017; XX:00–00.
survival benefits.1,2 A strategy is debated in recent discussions proposing extracorporeal membrane oxygenation as a tool to improve the outcome of CPR after cardiac arrest (ECPR) that might show better short- and long-term survival over conventional cardiopulmonary resuscitation (CCPR).3,4 Although highquality randomized control trial data do not exist to prove the better outcome after using ECPR, a number of observational studies suggest a benefit for these patients.5 A recently published meta-analysis including 3,098 cardiac arrest patients confirmed this suggestion by showing an increase of 30 day survival of up to 13%.6 Kim et al.7 showed in another metaanalysis with high-quality data that the survival to hospital discharge, when using ECPR almost doubled when compared with patients who underwent CCPR. Despite these benefits of ECPR, it also has significant ethical implications,8 and it remains debatable which patient should receive this invasive and costly technique. Many ECPR programs exclude patients from ECPR with presumably weak outcome criteria like initial rhythm asystole, total arrest time of more than 60 minutes, and suspicion of shock due to sepsis. However, most of these criteria are not supported by significant study data. Conclusive predictors for better outcome in ECPR could simplify decision which patients to treat with ECPR. We retrospectively examined outcome of a single-center trial on a group of “less high selected” patients who received ECPR after cardiac arrest and were trying to determine predictors for good outcome. Methods This retrospective single-center study includes 59 patients who received placement of VA extracorporeal membrane oxygenation (ECMO) during resuscitation for refractory cardiac arrest between June 2008 and September 2016. Refractory cardiac arrest was defined as a sudden cessation of normal cardiac function in which mechanical CPR did not lead to a return of spontaneous circulation within the first 10 minutes of being initiated. The procedure of the further treatment was discussed with the family and/or power of attorney directly after ECPR was done and patients advance directive wishes were addressed. Only patients who received an ECMO cannula placement during chest compressions were included in this study. Information obtained through the “Cerner Health Facts® database” (Cerner Corporation, Kansas City, MO) including age, sex, body mass index (BMI), initial heart rhythm during arrest, reason for cardiogenic shock, time of chest compressions, location of cardiac arrest, maximum lactate levels, creatinine levels before ECMO placement, maximum troponin levels and ECMO complications were compiled and analyzed with regard to patients survival to discharge.
Key Words: ECPR, ECMO, survival, outcome
C
urrent resuscitation strategies for cardiac arrest include electrical defibrillation, chest compression (CPR), and ACLS medications. Although electrical defibrillation and CPR are fairly effective in the early minutes after cardiac arrest, the addition of ACLS medications seem not to show significant
From the *Heart and Vascular Institute, Penn State Milton S. Hershey Medical Center, Hershey, Pennsylvania; and †Cardiothoracic Surgery, INTEGRIS Baptist Medical Center, Oklahoma City, Oklahoma. Submitted for consideration February 2017; accepted for publication in revised form June 2017. Disclosure: The authors have no conflicts of interest to report. Correspondence: Dirk Pabst, Heart and Vascular Institute, Penn State Milton S. Hershey Medical Center, 500 University Drive, Hershey, PA 17033-0850. Email: [email protected]. Copyright © 2017 by the ASAIO DOI: 10.1097/MAT.0000000000000644
1
Copyright © American Society of Artificial Internal Organs. Unauthorized reproduction of this article is prohibited.
2
PABST ET AL.
All 59 patients received mechanical CPR in accordance with the ACLS guidelines of the American Heart Association before receiving ECMO support. The ECMO circuits that were utilized in this study consisted of a Quadrox oxygenator (Maquet Cardiovascular; Wayne, NJ) and either a Rotaflow® centrifugal pump (Maquet Cardiovascular; Wayne, NJ) or a Centrimag® centrifugal pump (Levitronix LLC, Waltham, MA). Attending intensivists or attending cardiac surgeons in the hospital setting performed all ECMO cannulations. The cannulation technique utilized for each patient was at the discretion of the performing physician, with a peripheral catheter placement in both the femoral artery and the femoral vein by percutaneous Seldinger technic being the preferred cannulation method. Subsequent management and care of the patient was performed by the Heart and Vascular Institute Critical Care Unit team of ECMO specialized nurses, nurse practitioners, and intensivists. Extracorporeal membrane oxygenation flows and inotrope dosing (epinephrine, dobutamine, or milrinone) were adjusted appropriately to maintain mean arterial pressures > 65 mm Hg and arterial saturations > 93%. Most patients were treated with “target temperature management” (TTM) after resuscitation for a goal temperature of 32– 34°C for 24 hours according to the institutional TTM protocol. Neurological status was described with the cerebral performance category (CPC). Good neurological outcome was defined as CPC 1 and CPC 2.9 Results were reported as percentages, means ± standard deviations, and/or medians and interquartile ranges. Categorical data were analyzed using a Pearson’s χ2 test of independence or a Fisher’s exact test when appropriate. Conversely, continuous data were analyzed using a Mann–Whitney U test. Kaplan–Meier survival curves were constructed to examine differences in survival over time between patient groups with heart rhythm previous to ECMO implantation as PEA/asystole, (shockable) ventricular rhythm, or unknown rhythm as well as survival over time for all patients. Subsequently, values between each pair were compared for statistical significance using the log-rank test. All tests were two-tailed, and statistical significance was defined by a p value of < 0.05. IBM SPSS Statistics (IBM Corporation, Armonk, NY) was used for the statistical analyses in this study. ECPR score developed by Park et al.10 was identified. For this score, identification of age, shockable rhythm, CPR time, arterial pulse pressure, and Sequential Organ Failure Assessment (SOFA) score was done. Results In the time period between June 2008 and September 2016, we performed 59 VA ECMOs under mechanical resuscitation for refractory cardiac arrest; characteristics of these patients are summarized in Table 1. The mean patient age was 57 years (range, 19–80 years). Forty patients (67.8%) were male. The mean patient BMI was 32.8 (SD ± 10.7). Twenty patients (47.5%) presented with ventricular tachycardiac rhythm (ventricular tachycardia or ventricular fibrillation) at the time of decision for VA ECMO placement was made. Sixteen patients (27.1%) had PEA, seven patients (11.9%) had an asystolie, and in 8 patients (13.6%) initial rhythm was not reliably documented. Twenty-nine patients (49.2%) had a myocardial infarction that was proven in ECG and/or in coronary angiography
Table 1. Patient Characteristics (n = 59) Variables Age (y), mean ± SD Male sex, n (%) BMI (kg/m2), mean Presenting rhythm, n (%) PEA Asystole Ventricular arrhythmia Unknown Duration of CCPR before ECMO flow or ROSC (min), mean (IQR) CPR < 40 min, n (%) CPR ≥ 40 min, n (%) CPR time not documented, n (%) Location of cardiac arrest, n (%) Emergency room Intensive care unit Operation room Cardiac catheter laboratory Ordinary ward CT scan Other hospital Type of cannulation, n (%) Peripheral Central Reason for cardiogenic shock, n (%) Proven myocardial infarction Suspicion for myocardial infarction Pulmonary embolism Allograft rejection Nonischemic cardiomyopathy Septic shock Other
Value 57.02 (±14.69) 40 (67.8) 32.84 (±10.71) 16 (27.1) 7 (11.9) 28 (47.5) 8 (13.6) 41.47 (±17.99) 15 (25.4) 27 (45.8) 17 (28.8) 4 (6.8) 18 (30.5) 2 (3.4) 16 (27.1) 8 (13.6) 2 (3.4) 9 (15.3) 58 (98.3) 1 (1.7) 29 (49.2) 10 (16.9) 4 (6.8) 5 (8.5) 2 (3.4) 2 (3.4) 7 (11.9)
CCPR, conventional cardiopulmonary resuscitation; ECMO, extracorporeal membrane oxygenation; IQR, interquartile range; PEA, pulseless electrical activity; ROSC, return of spontaneous circulation; SD, standard deviation.
afterward. In 10 patients (16.9%), myocardial infarction was assumed due to history and/or elevated troponine but could not be proven by EKG signs or coronary angiography. Four of these patients had a very high troponine level (> 16.24), four patients had a significant elevated level (> 0.13), and in one patient the troponine was not determined. In four patients (6.8%) pulmonary embolism was diagnosed as reason for cardiogenic shock. Five patients (8.5%) had an acute allograft rejection after heart transplantation. Two patients (3.4%) had an acute heart failure with history of nonischemic cardiomyopathy. Two patients (3.4%) were diagnosed with a septic shock in the aftermaths. Seven patients (11.9%) had other reason for cardiac arrest or a reliable diagnose could not be accomplished. Eighteen patients (30.5%) were in the intensive care unit during ECPR, and 16 patients (27.1%) were in the cardiac catheter laboratory. Other locations were the emergency department, operation room, ordinary ward, or CT scan. In nine patients (15.3%), resuscitation and ECMO placement were done in an outside hospital where more detailed information of the exact location could not be reliably evaluated. In the majority of patients (98.3%), a peripheral cannulation was done. Table 2 shows the outcome variables. The survival to discharge was 25.4%. In 14 cases (23.7%), the patients passed away during the first day after resuscitation. In five patients (8.5%), ECMO flow could not be established under CPR due
Copyright © American Society of Artificial Internal Organs. Unauthorized reproduction of this article is prohibited.
ECMO AND IMPROVED OUTCOME OF PATIENTS UNDER RESUSCITATION Table 2. Patient Outcome (n = 59)
Variables Survival to discharge Early death (during first day), n (%) Total ECLS days, mean (SD) ICU days, median (IQR), mean (SD) Days on ventilator support, mean (SD) Postdecannulation support, n (%) Left ventricular assist device Total artificial heart AB5000 ventricular assist device Acute kidney injury needed renal replacement therapy, n (%) Initial ECMO flow, mean (SD) ECMO support could not be established, n (%)
Value 15 (25.4) 14 (23.7) 6.00 (±6.75) 13.42 (±20.92) 9.54 (±12.76) 2 (3.4) 1 (1.7) 1 (1.7) 27 (45.8) 3.99 (±1.18) 5 (8.5)
ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit; IQR, interquartile range; SD, standard deviation.
to technical problems with cannulation in the first 75 minutes after cardiac arrest, and these patients were pronounced dead. Mean initial ECMO flow was 3.99 L/min (SD ± 1.18 L/min). The average time on ECLS was 6 days (range, 0–33 days). Two patients (3.4%) were bridged to a left ventricular assist device, and one patient was bridged to a total artificial heart (TAH). One patient was intermittently bridged on “AB5000 Ventricular Assist Device” (Abiomed Inc, Danvers, MA) that could be explanted after the heart function improved. One of the two LVAD patients passed away during the same hospital admission. The other LVAD patient and the TAH patient could be successfully heart transplanted during the course of the following 2 years. Daily plasma hemoglobin (pHb) was taken to detect hemolysis. Nineteen patients (32.2%) had hemolysis during their ECMO run, defined as plasmohemoglobin level > 50 mg/dL.11 There was no significant difference in survival rate between
3
patients with proven hemolysis and ECMO patients without hemolysis. In 36 patients (61.0%), blood culture was taken during ECMO treatment. In three patients (5.1%), bacteremia could be discovered. Other ECMO-related complications were significant groin bleedings leading to blood transfusion in seven patients. Two patients had severe groin bleeding in which surgical intervention was needed. In one patient, a surgical vessel repair was done, the other patient needed a fasciotomy due to compartment syndrome. Limb ischemia occurred in three patients. In one patient, thrombectomy was done. In no patient, amputation was necessary. Twenty-seven patients (45.8%) had to be treated with renal replacement therapy (RRT). Four patients (6.8%) already had indication for RRT before their cardiac arrest. Seven patients (46.7%) of the survived discharged patients went on CRRT during the hospital admission. In three patients (18.8%), RRT-dependent terminal kidney failure (CKD stage 5) remained as diagnose in the following up visits. Of the 16 patients, which survived hospital discharge, all patients had a good neurological outcome defined as CPC score 1 or 2. Figure 1 shows the Kaplan–Meier analysis survival over the first year of the patient group in which ECMO was explanted at least 1 year before data were collected. The 1-year survival of the discharged patients was 73.3%. Twenty-three patients had asystolie or PEA as initial rhythm. The survival rate of these patients was significantly lower (n = 2; 8.7%). In 27 patients, a shockable rhythm (ventricular tachycardia or ventricular fibrillation) was seen as initial rhythm. These patients showed a survival to discharge rate of 40.7% (n = 11). Figure 2 shows the Kaplan–Meier 1-year survival curve separated by initial heart rhythm. Patients with serum lactate level of ≥ 8 mmol/L had a significant lower survival to discharge when compared with patients with lower lactate levels (13.3% vs. 45.0%).
Figure 1. Kaplan–Meier survival curve of 59 patients after ECPR.
Copyright © American Society of Artificial Internal Organs. Unauthorized reproduction of this article is prohibited.
4
PABST ET AL.
Figure 2. Kaplan–Meier survival curve separated by initial heart rhythm.
Male gender also showed significant lower survival to discharge rate (17.5% vs. 42.1%). Maximum lactate values of ≥ 8 mmol/L and PEA or asystole as heart rhythm by time of decision were significant risk factors for in-hospital mortality. In this patient cohort, we could not find significance to mortality in age, troponine level, baseline plasma creatinine, BMI, location of CPR, or reasons for cardiogenic shock. Surprisingly, there was no significant difference between groups of shorter and longer CPR-time in the outcome of survival. This was tested for groups more and less than 30 and 60 minutes of CPR-time. ECPR score10 could be identified in 33 patients. The mean ECPR score was 9.06 (minimum 5; maximum 13). Forty percentage of the patients with an ECPR score of > 10 survived when compared with 17.4% to patients with an ECPR score of ≤ 10. Discussion In the last years, ECMO treatment has grown widely as a rescue therapy with several indications like cardiogenic shock, ARDS, and other reasons. Table 3. Patients Survived Discharge (n = 15) Variables Good neurological outcome (CPC ≤ 2), n (%) Renal replacement therapy during admission, n (%) Stayed on RRT after admission, n (%) One year survival, n (%)
Value 15 (100) 7 (46.7) 3 (18.75) 11 (73.3)
CPC, cerebral performance category; RRT, renal replacement therapy.
It was described that VA ECMO placement during CPR could improve the outcome of patients undergoing cardiac arrest.3,4 According to the registry from the Extracorporeal Life Support Organization (ELSO) as of July 2016, 2,885 adult patients have received VA ECMO during ECPR. Survival to discharge rate was 29% in these patients.11 We found a survival rate of 25.4% in our institution and therefore are in the range of the documented ELSO data. Other publications describe a survival to discharge rate from 13% to 54%.12–20 In many publications, inclusion criteria are well selected and exclude patients with asystolie or PEA as initial heart rhythm.15,20 Bellezzo et al.20 defined not only persistent cardiopulmonary arrest as inclusion criteria but also shock (SBP < 70 mm Hg) refractory to standard therapies. Other reasons for high variation of the outcome between studies could be different comorbidities, different percentage of VT/pVT. In our study, all patients were placed on ECMO under ongoing mechanical compression (ECPR) and a high number of patients with asystolie or PEA were included. The survival rate of patients without asystolie or PEA as initial rhythm in our study is 40.7% and is significantly different from the survival to discharge from the patients with asystolie or PEA (8.7%; p = 0.04). ECPR is an invasive procedure, which requires resources including costs for staff and equipment. It should be clearly defined when patients likely would benefit from this procedure and when the resources, risk, and poor outcome does not justify the initiation of VA ECMO during CPR. Which outcome in a certain patient group justifies the initiation of this treatment needs to be discussed. However, knowing predictors for poor outcome might help to simplify the decision in which patient ECPR should be tried. Even if some studies show some
Copyright © American Society of Artificial Internal Organs. Unauthorized reproduction of this article is prohibited.
5
ECMO AND IMPROVED OUTCOME OF PATIENTS UNDER RESUSCITATION Table 4. Survival to Discharge and Potential Predictors for Survival (n = 59)
Variables Baseline plasma creatinine ≥ 2.0 mg/dL Baseline plasma creatinine < 2.0 mg/dL Unknown BMI ≥ 30 kg/m2 BMI < 30 kg/m2 Male Female Age ≥ 75 y Age < 75 y Age ≥ 65 y Age < 65 y Lactate ≥ 7 mmol/L Lactate < 7 mmol/L Unknown lactate Lactate ≥ 8 mmol/L Lactate < 8 mmol/L Unknown lactate Troponine I ≥ 10 ng/ml Troponine I < 10 ng/ml Unknown troponine PEA or asystole Ventricular arrhythmia Unknown rhythm CPR-time ≥ 30 min CPR-time < 30 min CPR-time unknown CPR-time ≥ 60 min CPR-time < 60 min CPR-time unknown AKI needed renal replacement therapy No need for RRT Location cath laboratory No cath laboratory Reason for cardiogenic shock Proven MI No proven MI Pulmonary embolism No pulmonary embolism Allograft rejection No allograft rejection Non ischemic CM No nonischemic CM Septic shock No septic shock ECPR score > 10 ECPR score ≤ 10
n/N (% survival)
p
2/13 (15.4) 10/38 (26.3) 3/8 (37.5) 7/36 (19.4) 8/23 (34.8) 7/40 (17.5) 8/19 (42.1) 0/5 (0.0) 15/54 (27.8) 3/20 (15.0) 12/39 (30.8) 7/36 (19.4) 6/14 (42.9) 2/9 (22.2) 4/30 (13.3) 9/20 (45.0) 2/9 (22.2) 9/27 (33.3) 4/20 (20.0) 2/12 (16.7) 2/23 (8.7) 11/27 (40.7) 2/9 (22.2) 6/32 (18.8) 3/10 (30.0) 6/17 (35.3) 1/9 (11.1) 8/33 (24.2) 6/17 (35.3) 7/27 (25.9) 8/32 (25.0) 6/16 (37.5) 9/43 (20.9)
0.42
10/29 (34.5) 5/30 (16.7) 1/4 (25.0) 14/55 (25.5) 2/5 (40.0) 13/54 (24.1) 0/2 (0.0) 15/57 (26.3) 0/2 (0.0) 15/57 (23.3) 4/10 (40.0) 4/23 (17.4)
0.12
0.19
Table 5. Complications Related to ECMO Treatment (n = 59) Variables
Value
Groin bleeding (nonsevere), n (%) 7 (11.9) Severe vessel bleeding (surgical intervention needed), n (%) 2 (3.4) Limb ischemia without treatment indication, n (%) 2 (3.4) Limb ischemia needed thrombectomy, n (%) 1 (1.7) Positive blood cultures, n (%) 4 (6.8) Hemolysis (plasmohemoglobin > 50 mg/dL), n (%) 19 (32.2)
0.04 0.17 0.19 0.09 0.01 0.31 0.04 0.42 0.37 0.94 0.19
0.98 0.43 1.00 1.00 0.16
BMI, body mass index; CPR, chest compression; PEA, pulseless electrical activity; RRT, renal replacement therapy.
predictors for poor survival rates (for example age ≤ 66 years, shockable arrest rhythm, lactate acid, CPR to ECMO pump-on time between (≤ 45 minutes) and comorbidity10), guidelines that propose in which patient groups ECMO placement during CPR should be initiated are still missing. In our patient cohort, lactate ≥ 8 mmol/L and as mentioned above cardiac rhythm of PEA or asystole were significant predictors for low survival rate. We did not find a significant age boarder, which limited the outcome in our study group. We could show that patients even with PEA and asystole at the time of ECPR could survive. Therefore, a nonshockable rhythm by itself should not be an exclusion criterion for ECMO placement and other criteria should be taken into account and also have to be identified in further studies.
Questionable is still after how many minutes of CPR outcome is futile and when attempt of ECPR should be avoided or withdrawn. Kelly et al.21 described two cases with good neurological outcome after 176 minutes and 97 minutes cardiopulmonary resuscitation before extracorporeal rescue. The CPR time in our study cohort was up to 75 minutes. Surprisingly, we did not find a significant correlation between CPR time and survival rate. Although ECPR can increase survival rates after cardiac arrest, it could also result in poor postresuscitation neurological status, and the data for this are still poor. Studies reported high serum lactic acid levels before ECMO not only as predictors for poor survival rate but also for poor neurological outcome.22 In our data, all survivors of discharge had a good neurological outcome with CPC of 1 or 2. It is still unclear which role acute kidney injury might play in terms of survival rate. Oliguria seen during the first 24 hours after ECMO may be an independent predictor of mortality.23 In our patient cohort, there was no significant higher mortality for patient needed CRRT during admission. The reason for the cardiac arrest also did not show any differences in outcome. Some studies mention that the outcome of ECPR in adult patients for septic shock is controversial.24,25 In our cohort, we had two patients in whom septic shock was diagnosed as reason for cardiac arrest in the aftermath. Both patients did not survive hospital discharge. Patients undergoing ECMO support through femoral vessels, especially under CPR are prone to complications.26–28 The incidence of complications in our cohort was comparable with complications reported in other publications.26,28,29 Most complications include femoral artery injuries and were manageable. Severe complication was limb ischemia needed fasciectomy and severe groin bleeding needing immediate intervention, but no amputation. Hemolysis during extracorporeal membrane oxygenation is a known complication.30 Daily pHb levels were measured. Elevated pHb could be seen in 19 patients (32.2%), which had no significant influence on survival to discharge rate. Experience of the provider (especially in terms of cannulation) could influence a successful placement of the cannulas and therefore increase the outcome.29 Technical challenge during CPR is particularly to place the arterial cannula because due to low flow the vessel is often small and under ongoing CPR difficult to detect. In this study, in five patients (8.5%) ECMO support could not be established. Most of our patients who were placed on ECMO in the cardiac catheter laboratory already had femoral arterial access that were used for placing the ECMO cannula. Anyway, there was no significant difference in mortality between patients who were cannulated in the cardiac catheter laboratory and patients who were placed somewhere else on ECMO.
Copyright © American Society of Artificial Internal Organs. Unauthorized reproduction of this article is prohibited.
6
PABST ET AL.
Our study has some limitations. Nine patients were cannulated in other hospitals and were transported for further care to our institution. Values for variables like first detected rhythm on CPR, baseline creatinine, troponine were sometimes difficult to evaluate in these patients. Chest compression time was well documented in 42 patients. In 17 patients, no reliable CPR time was documented. Male gender was a significant predictor for poor survival in our cohort. Reasons for this remain unclear, and it might be reevaluated in cohorts of bigger studies. Park et al.10 developed a score to monitor survival to discharge in patients after ECPR. The score ranges from 3 to 15 and includes the age of the patient, shockable arrest rhythm, CPR time, post-ECMO arterial pulse pressure, and the SOFA score. When the ECPR score was > 10, the sensitivity and specificity for prediction of survival to discharge in Parks study were 89.6% and 75.0%, respectively. The ECPR score could be identified in 33 patients in this cohort. Ten patients had an ECPR score of > 10. Forty percentage of these patients survived to discharge when compared with 17.4% of the patient with an ECPR score less than 10. Although there was a remarkable difference in survival between these two groups, our patients did not reach the significant difference in outcome expected by ECPR score. One reason might be the smaller cohort in our study. Anyhow, the ECPR score has already shown in a bigger cohort that it can be a useful tool for clinical providers.10 Although this study did not find any significant difference in the survival outcome between groups with longer and shorter CPR time, we believe that shortening time to establish a sufficient blood circle is important in patients with cardiac arrest. Logistic structures might improve survival rate by shortening time of CPR. Several case series could even show that prehospital placement of ECMO is successfully possible27,31,32 and might be considered as a challenge for future ECMO programs. The results of this study do not deviate much from the published ELSO data. Slightly lower survival to discharge results can be explained mainly with the higher number of patients having PEA/asystole while ECMO placement. Demographic differences in matters of age do not seem to play an important role in our study since we did not see a clear cutoff point for age. In conclusion, we showed that ECLS after cardiac arrest could be successful even in a nonhighly selected group of patients. In patients with high serum lactate levels, PEA or asystole presenting as initial heart rhythm, ECPR should be carefully considered since these were predictors for poor outcome in our study. References 1. Stiell IG, Wells GA, Field B, et al; Ontario Prehospital Advanced Life Support Study Group: Advanced cardiac life support in outof-hospital cardiac arrest. N Engl J Med 351: 647–656, 2004. 2. Olasveengen TM, Sunde K, Brunborg C, Thowsen J, Steen PA, Wik L: Intravenous drug administration during out-of-hospital cardiac arrest: A randomized trial. JAMA 302: 2222–2229, 2009. 3. Chen YS, Lin JW, Yu HY, et al: Cardiopulmonary resuscitation with assisted extracorporeal life-support versus conventional cardiopulmonary resuscitation in adults with in-hospital cardiac arrest: An observational study and propensity analysis. Lancet 372: 554–561, 2008. 4. Blumenstein J, Leick J, Liebetrau C et al. Extracorporeal life support in cardiovascular patients with observed refractory in-hospital
cardiac arrest is associated with favourable short and long-term outcomes: A propensity-matched analysis. Eur Heart J Acute Cardiovasc Care 2015. 5. Ortega-Deballon I, Hornby L, Shemie SD, Bhanji F, Guadagno E: Extracorporeal resuscitation for refractory out-of-hospital cardiac arrest in adults: A systematic review of international practices and outcomes. Resuscitation 101: 12–20, 2016. 6. Ouweneel DM, Schotborgh JV, Limpens J, et al: Extracorporeal life support during cardiac arrest and cardiogenic shock: A systematic review and meta-analysis. Intensive Care Med 42: 1922–1934, 2016. 7. Kim SJ, Kim HJ, Lee HY, Ahn HS, Lee SW: Comparing extracorporeal cardiopulmonary resuscitation with conventional cardiopulmonary resuscitation: A meta-analysis. Resuscitation 103: 106–116, 2016. 8. Riggs KR, Becker LB, Sugarman J: Ethics in the use of extracorporeal cardiopulmonary resuscitation in adults. Resuscitation 91: 73–75, 2015. 9. Cummins RO, Chamberlain DA, Abramson NS, et al: Recommended guidelines for uniform reporting of data from out-of-hospital cardiac arrest: the Utstein Style. A statement for health professionals from a task force of the American Heart Association, the European Resuscitation Council, the Heart and Stroke Foundation of Canada, and the Australian Resuscitation Council. Circulation 84: 960–975, 1991. 10. Park SB, Yang JH, Park TK, et al: Developing a risk prediction model for survival to discharge in cardiac arrest patients who undergo extracorporeal membrane oxygenation. Int J Cardiol 177: 1031–1035, 2014. 11. Registry from the Extracorporeal Life Support Organization (ELSO). 2017. Available at elso.org. 12. Johnson NJ, Acker M, Hsu CH, et al: Extracorporeal life support as rescue strategy for out-of-hospital and emergency department cardiac arrest. Resuscitation 85: 1527–1532, 2014. 13. de Chambrun MP, Bréchot N, Lebreton G, et al: Venoarterial extracorporeal membrane oxygenation for refractory cardiogenic shock post-cardiac arrest. Intensive Care Med 42: 1999–2007, 2016. 14. Mazzeffi MA, Sanchez PG, Herr D, et al: Outcomes of extracorporeal cardiopulmonary resuscitation for refractory cardiac arrest in adult cardiac surgery patients. J Thorac Cardiovasc Surg 152: 1133–1139, 2016. 15. Stub D, Bernard S, Pellegrino V, et al: Refractory cardiac arrest treated with mechanical CPR, hypothermia, ECMO and early reperfusion (the CHEER trial). Resuscitation 86: 88–94, 2015. 16. Maekawa K, Tanno K, Hase M, Mori K, Asai Y: Extracorporeal cardiopulmonary resuscitation for patients with out-of-hospital cardiac arrest of cardiac origin: A propensity-matched study and predictor analysis. Crit Care Med 41: 1186–1196, 2013. 17. Kagawa E, Inoue I, Kawagoe T, et al: Assessment of outcomes and differences between in- and out-of-hospital cardiac arrest patients treated with cardiopulmonary resuscitation using extracorporeal life support. Resuscitation 81: 968–973, 2010. 18. Mégarbane B, Leprince P, Deye N, et al: Emergency feasibility in medical intensive care unit of extracorporeal life support for refractory cardiac arrest. Intensive Care Med 33: 758–764, 2007. 19. Sakamoto T, Morimura N, Nagao K, et al; SAVE-J Study Group: Extracorporeal cardiopulmonary resuscitation versus conventional cardiopulmonary resuscitation in adults with outof-hospital cardiac arrest: A prospective observational study. Resuscitation 85: 762–768, 2014. 20. Bellezzo JM, Shinar Z, Davis DP, et al: Emergency physician initiated extracorporeal cardiopulmonary resuscitation. Resuscitation 83: 966–970, 2012. 21. Kelly RB, Porter PA, Meier AH, Myers JL, Thomas NJ: Duration of cardiopulmonary resuscitation before extracorporeal rescue: How long is not long enough? ASAIO J 51: 665–667, 2005. 22. Ryu JA, Cho YH, Sung K, et al: Predictors of neurological outcomes after successful extracorporeal cardiopulmonary resuscitation. BMC Anesthesiol 15: 26, 2015. 23. Lee JJ, Han SJ, Kim HS, et al: Out-of-hospital cardiac arrest patients treated with cardiopulmonary resuscitation using extracorporeal membrane oxygenation: Focus on survival rate and
Copyright © American Society of Artificial Internal Organs. Unauthorized reproduction of this article is prohibited.
ECMO AND IMPROVED OUTCOME OF PATIENTS UNDER RESUSCITATION
neurologic outcome. Scand J Trauma Resusc Emerg Med 24: 74, 2016. 24. Bréchot N, Luyt CE, Schmidt M, et al: Venoarterial extracorporeal membrane oxygenation support for refractory cardiovascular dysfunction during severe bacterial septic shock. Crit Care Med 41: 1616–1626, 2013. 25. Cheng A, Sun HY, Lee CW, et al: Survival of septic adults compared with nonseptic adults receiving extracorporeal membrane oxygenation for cardiopulmonary failure: A propensity-matched analysis. J Crit Care 28: 532.e1–532.10, 2013. 26. Bisdas T, Beutel G, Warnecke G, et al: Vascular complications in patients undergoing femoral cannulation for extracorporeal membrane oxygenation support. Ann Thorac Surg 92: 626– 631, 2011. 27. Lamhaut L, Jouffroy R, Soldan M, et al: Safety and feasibility of prehospital extra corporeal life support implementation by non-surgeons for out-of-hospital refractory cardiac arrest. Resuscitation 84: 1525–1529, 2013.
7
28. Aziz F, Brehm CE, El-Banyosy A, Han DC, Atnip RG, Reed AB: Arterial complications in patients undergoing extracorporeal membrane oxygenation via femoral cannulation. Ann Vasc Surg 28: 178–183, 2014. 29. Rupprecht L, Lunz D, Philipp A, Lubnow M, Schmid C: Pitfalls in percutaneous ECMO cannulation. Heart Lung Vessel 7: 320– 326, 2015. 30. Palanzo DA, El-Banayosy A, Stephenson E, Brehm C, Kunselman A, Pae WE: Comparison of hemolysis between CentriMag and RotaFlow rotary blood pumps during extracorporeal membrane oxygenation. Artif Organs 37: E162–E166, 2013. 31. Grunau B, Scheuermeyer FX, Stub D, et al: Potential candidates for a structured Canadian ECPR Program for out-of-hospital cardiac arrest. CJEM 18: 453–460, 2016. 32. Kippnich M, Lotz C, Kredel M, et al: [Venoarterial extracorporeal membrane oxygenation for out-of-hospital cardiac arrest. Case series of prehospital and in-hospital therapies]. Anaesthesist 64: 580–585, 2015.
Copyright © American Society of Artificial Internal Organs. Unauthorized reproduction of this article is prohibited.
Clinical Critical Care
Predictors of Survival for Nonhighly Selected Patients Undergoing Resuscitation With Extracorporeal Membrane Oxygenation After Cardiac Arrest DIRK PABST,* ALY EL-BANAYOSY,† BEHZAD SOLEIMANI,* AND CHRISTOPH E. BREHM*
In several case reports and case series, extracorporeal membrane oxygenation during chest compression (CPR) has been shown to be a reasonable tool to improve outcome of patients under resuscitation. Although recommendations for ECPR include younger patients with shockable rhythm and short previous CPR-time, it remains unclear if nonhighly selected patients have a similar outcome. Aim of this study was to determine outcome in our nonhighly selected patient population treated with ECPR and investigate possible predictors of survival. We made a retrospective single-center study of adults who underwent ECPR for in-hospital cardiac arrest between June 2008 and September 2016. Outcome and predictors of survival were identified. In this period of time, 59 patients underwent ECPR due to cardiac arrest. Fifteen patients (25.4%) survived discharge of which all had a good neurological outcome (cerebral performance category, ≤ 2). Survival to discharge of patients with shockable rhythm (ventricular fibrillation or ventricular tachycardia) was 40.7%. Serum lactate ≥ 8, pulseless electrical activity (PEA) or asystole and male gender could be identified as predictors for low survival rate. Age, body mass index, renal replacement–dependent kidney injury had no significant influence on survival outcome. Mean CPR-time was 41.1 minutes (interquartile range, ±29.25 minutes). Extracorporeal membrane oxygenation seems to be a useful tool to improve the outcome of CPR also in nonhighly selected patients when compared with CPR alone and could be considered in patients with refractory cardiac arrest also after longer previous CPR-time. Serum lactate and heart rhythm should be taken into account for patient selection. ASAIO Journal 2017; XX:00–00.
survival benefits.1,2 A strategy is debated in recent discussions proposing extracorporeal membrane oxygenation as a tool to improve the outcome of CPR after cardiac arrest (ECPR) that might show better short- and long-term survival over conventional cardiopulmonary resuscitation (CCPR).3,4 Although highquality randomized control trial data do not exist to prove the better outcome after using ECPR, a number of observational studies suggest a benefit for these patients.5 A recently published meta-analysis including 3,098 cardiac arrest patients confirmed this suggestion by showing an increase of 30 day survival of up to 13%.6 Kim et al.7 showed in another metaanalysis with high-quality data that the survival to hospital discharge, when using ECPR almost doubled when compared with patients who underwent CCPR. Despite these benefits of ECPR, it also has significant ethical implications,8 and it remains debatable which patient should receive this invasive and costly technique. Many ECPR programs exclude patients from ECPR with presumably weak outcome criteria like initial rhythm asystole, total arrest time of more than 60 minutes, and suspicion of shock due to sepsis. However, most of these criteria are not supported by significant study data. Conclusive predictors for better outcome in ECPR could simplify decision which patients to treat with ECPR. We retrospectively examined outcome of a single-center trial on a group of “less high selected” patients who received ECPR after cardiac arrest and were trying to determine predictors for good outcome. Methods This retrospective single-center study includes 59 patients who received placement of VA extracorporeal membrane oxygenation (ECMO) during resuscitation for refractory cardiac arrest between June 2008 and September 2016. Refractory cardiac arrest was defined as a sudden cessation of normal cardiac function in which mechanical CPR did not lead to a return of spontaneous circulation within the first 10 minutes of being initiated. The procedure of the further treatment was discussed with the family and/or power of attorney directly after ECPR was done and patients advance directive wishes were addressed. Only patients who received an ECMO cannula placement during chest compressions were included in this study. Information obtained through the “Cerner Health Facts® database” (Cerner Corporation, Kansas City, MO) including age, sex, body mass index (BMI), initial heart rhythm during arrest, reason for cardiogenic shock, time of chest compressions, location of cardiac arrest, maximum lactate levels, creatinine levels before ECMO placement, maximum troponin levels and ECMO complications were compiled and analyzed with regard to patients survival to discharge.
Key Words: ECPR, ECMO, survival, outcome
C
urrent resuscitation strategies for cardiac arrest include electrical defibrillation, chest compression (CPR), and ACLS medications. Although electrical defibrillation and CPR are fairly effective in the early minutes after cardiac arrest, the addition of ACLS medications seem not to show significant
From the *Heart and Vascular Institute, Penn State Milton S. Hershey Medical Center, Hershey, Pennsylvania; and †Cardiothoracic Surgery, INTEGRIS Baptist Medical Center, Oklahoma City, Oklahoma. Submitted for consideration February 2017; accepted for publication in revised form June 2017. Disclosure: The authors have no conflicts of interest to report. Correspondence: Dirk Pabst, Heart and Vascular Institute, Penn State Milton S. Hershey Medical Center, 500 University Drive, Hershey, PA 17033-0850. Email: [email protected]. Copyright © 2017 by the ASAIO DOI: 10.1097/MAT.0000000000000644
1
Copyright © American Society of Artificial Internal Organs. Unauthorized reproduction of this article is prohibited.
2
PABST ET AL.
All 59 patients received mechanical CPR in accordance with the ACLS guidelines of the American Heart Association before receiving ECMO support. The ECMO circuits that were utilized in this study consisted of a Quadrox oxygenator (Maquet Cardiovascular; Wayne, NJ) and either a Rotaflow® centrifugal pump (Maquet Cardiovascular; Wayne, NJ) or a Centrimag® centrifugal pump (Levitronix LLC, Waltham, MA). Attending intensivists or attending cardiac surgeons in the hospital setting performed all ECMO cannulations. The cannulation technique utilized for each patient was at the discretion of the performing physician, with a peripheral catheter placement in both the femoral artery and the femoral vein by percutaneous Seldinger technic being the preferred cannulation method. Subsequent management and care of the patient was performed by the Heart and Vascular Institute Critical Care Unit team of ECMO specialized nurses, nurse practitioners, and intensivists. Extracorporeal membrane oxygenation flows and inotrope dosing (epinephrine, dobutamine, or milrinone) were adjusted appropriately to maintain mean arterial pressures > 65 mm Hg and arterial saturations > 93%. Most patients were treated with “target temperature management” (TTM) after resuscitation for a goal temperature of 32– 34°C for 24 hours according to the institutional TTM protocol. Neurological status was described with the cerebral performance category (CPC). Good neurological outcome was defined as CPC 1 and CPC 2.9 Results were reported as percentages, means ± standard deviations, and/or medians and interquartile ranges. Categorical data were analyzed using a Pearson’s χ2 test of independence or a Fisher’s exact test when appropriate. Conversely, continuous data were analyzed using a Mann–Whitney U test. Kaplan–Meier survival curves were constructed to examine differences in survival over time between patient groups with heart rhythm previous to ECMO implantation as PEA/asystole, (shockable) ventricular rhythm, or unknown rhythm as well as survival over time for all patients. Subsequently, values between each pair were compared for statistical significance using the log-rank test. All tests were two-tailed, and statistical significance was defined by a p value of < 0.05. IBM SPSS Statistics (IBM Corporation, Armonk, NY) was used for the statistical analyses in this study. ECPR score developed by Park et al.10 was identified. For this score, identification of age, shockable rhythm, CPR time, arterial pulse pressure, and Sequential Organ Failure Assessment (SOFA) score was done. Results In the time period between June 2008 and September 2016, we performed 59 VA ECMOs under mechanical resuscitation for refractory cardiac arrest; characteristics of these patients are summarized in Table 1. The mean patient age was 57 years (range, 19–80 years). Forty patients (67.8%) were male. The mean patient BMI was 32.8 (SD ± 10.7). Twenty patients (47.5%) presented with ventricular tachycardiac rhythm (ventricular tachycardia or ventricular fibrillation) at the time of decision for VA ECMO placement was made. Sixteen patients (27.1%) had PEA, seven patients (11.9%) had an asystolie, and in 8 patients (13.6%) initial rhythm was not reliably documented. Twenty-nine patients (49.2%) had a myocardial infarction that was proven in ECG and/or in coronary angiography
Table 1. Patient Characteristics (n = 59) Variables Age (y), mean ± SD Male sex, n (%) BMI (kg/m2), mean Presenting rhythm, n (%) PEA Asystole Ventricular arrhythmia Unknown Duration of CCPR before ECMO flow or ROSC (min), mean (IQR) CPR < 40 min, n (%) CPR ≥ 40 min, n (%) CPR time not documented, n (%) Location of cardiac arrest, n (%) Emergency room Intensive care unit Operation room Cardiac catheter laboratory Ordinary ward CT scan Other hospital Type of cannulation, n (%) Peripheral Central Reason for cardiogenic shock, n (%) Proven myocardial infarction Suspicion for myocardial infarction Pulmonary embolism Allograft rejection Nonischemic cardiomyopathy Septic shock Other
Value 57.02 (±14.69) 40 (67.8) 32.84 (±10.71) 16 (27.1) 7 (11.9) 28 (47.5) 8 (13.6) 41.47 (±17.99) 15 (25.4) 27 (45.8) 17 (28.8) 4 (6.8) 18 (30.5) 2 (3.4) 16 (27.1) 8 (13.6) 2 (3.4) 9 (15.3) 58 (98.3) 1 (1.7) 29 (49.2) 10 (16.9) 4 (6.8) 5 (8.5) 2 (3.4) 2 (3.4) 7 (11.9)
CCPR, conventional cardiopulmonary resuscitation; ECMO, extracorporeal membrane oxygenation; IQR, interquartile range; PEA, pulseless electrical activity; ROSC, return of spontaneous circulation; SD, standard deviation.
afterward. In 10 patients (16.9%), myocardial infarction was assumed due to history and/or elevated troponine but could not be proven by EKG signs or coronary angiography. Four of these patients had a very high troponine level (> 16.24), four patients had a significant elevated level (> 0.13), and in one patient the troponine was not determined. In four patients (6.8%) pulmonary embolism was diagnosed as reason for cardiogenic shock. Five patients (8.5%) had an acute allograft rejection after heart transplantation. Two patients (3.4%) had an acute heart failure with history of nonischemic cardiomyopathy. Two patients (3.4%) were diagnosed with a septic shock in the aftermaths. Seven patients (11.9%) had other reason for cardiac arrest or a reliable diagnose could not be accomplished. Eighteen patients (30.5%) were in the intensive care unit during ECPR, and 16 patients (27.1%) were in the cardiac catheter laboratory. Other locations were the emergency department, operation room, ordinary ward, or CT scan. In nine patients (15.3%), resuscitation and ECMO placement were done in an outside hospital where more detailed information of the exact location could not be reliably evaluated. In the majority of patients (98.3%), a peripheral cannulation was done. Table 2 shows the outcome variables. The survival to discharge was 25.4%. In 14 cases (23.7%), the patients passed away during the first day after resuscitation. In five patients (8.5%), ECMO flow could not be established under CPR due
Copyright © American Society of Artificial Internal Organs. Unauthorized reproduction of this article is prohibited.
ECMO AND IMPROVED OUTCOME OF PATIENTS UNDER RESUSCITATION Table 2. Patient Outcome (n = 59)
Variables Survival to discharge Early death (during first day), n (%) Total ECLS days, mean (SD) ICU days, median (IQR), mean (SD) Days on ventilator support, mean (SD) Postdecannulation support, n (%) Left ventricular assist device Total artificial heart AB5000 ventricular assist device Acute kidney injury needed renal replacement therapy, n (%) Initial ECMO flow, mean (SD) ECMO support could not be established, n (%)
Value 15 (25.4) 14 (23.7) 6.00 (±6.75) 13.42 (±20.92) 9.54 (±12.76) 2 (3.4) 1 (1.7) 1 (1.7) 27 (45.8) 3.99 (±1.18) 5 (8.5)
ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit; IQR, interquartile range; SD, standard deviation.
to technical problems with cannulation in the first 75 minutes after cardiac arrest, and these patients were pronounced dead. Mean initial ECMO flow was 3.99 L/min (SD ± 1.18 L/min). The average time on ECLS was 6 days (range, 0–33 days). Two patients (3.4%) were bridged to a left ventricular assist device, and one patient was bridged to a total artificial heart (TAH). One patient was intermittently bridged on “AB5000 Ventricular Assist Device” (Abiomed Inc, Danvers, MA) that could be explanted after the heart function improved. One of the two LVAD patients passed away during the same hospital admission. The other LVAD patient and the TAH patient could be successfully heart transplanted during the course of the following 2 years. Daily plasma hemoglobin (pHb) was taken to detect hemolysis. Nineteen patients (32.2%) had hemolysis during their ECMO run, defined as plasmohemoglobin level > 50 mg/dL.11 There was no significant difference in survival rate between
3
patients with proven hemolysis and ECMO patients without hemolysis. In 36 patients (61.0%), blood culture was taken during ECMO treatment. In three patients (5.1%), bacteremia could be discovered. Other ECMO-related complications were significant groin bleedings leading to blood transfusion in seven patients. Two patients had severe groin bleeding in which surgical intervention was needed. In one patient, a surgical vessel repair was done, the other patient needed a fasciotomy due to compartment syndrome. Limb ischemia occurred in three patients. In one patient, thrombectomy was done. In no patient, amputation was necessary. Twenty-seven patients (45.8%) had to be treated with renal replacement therapy (RRT). Four patients (6.8%) already had indication for RRT before their cardiac arrest. Seven patients (46.7%) of the survived discharged patients went on CRRT during the hospital admission. In three patients (18.8%), RRT-dependent terminal kidney failure (CKD stage 5) remained as diagnose in the following up visits. Of the 16 patients, which survived hospital discharge, all patients had a good neurological outcome defined as CPC score 1 or 2. Figure 1 shows the Kaplan–Meier analysis survival over the first year of the patient group in which ECMO was explanted at least 1 year before data were collected. The 1-year survival of the discharged patients was 73.3%. Twenty-three patients had asystolie or PEA as initial rhythm. The survival rate of these patients was significantly lower (n = 2; 8.7%). In 27 patients, a shockable rhythm (ventricular tachycardia or ventricular fibrillation) was seen as initial rhythm. These patients showed a survival to discharge rate of 40.7% (n = 11). Figure 2 shows the Kaplan–Meier 1-year survival curve separated by initial heart rhythm. Patients with serum lactate level of ≥ 8 mmol/L had a significant lower survival to discharge when compared with patients with lower lactate levels (13.3% vs. 45.0%).
Figure 1. Kaplan–Meier survival curve of 59 patients after ECPR.
Copyright © American Society of Artificial Internal Organs. Unauthorized reproduction of this article is prohibited.
4
PABST ET AL.
Figure 2. Kaplan–Meier survival curve separated by initial heart rhythm.
Male gender also showed significant lower survival to discharge rate (17.5% vs. 42.1%). Maximum lactate values of ≥ 8 mmol/L and PEA or asystole as heart rhythm by time of decision were significant risk factors for in-hospital mortality. In this patient cohort, we could not find significance to mortality in age, troponine level, baseline plasma creatinine, BMI, location of CPR, or reasons for cardiogenic shock. Surprisingly, there was no significant difference between groups of shorter and longer CPR-time in the outcome of survival. This was tested for groups more and less than 30 and 60 minutes of CPR-time. ECPR score10 could be identified in 33 patients. The mean ECPR score was 9.06 (minimum 5; maximum 13). Forty percentage of the patients with an ECPR score of > 10 survived when compared with 17.4% to patients with an ECPR score of ≤ 10. Discussion In the last years, ECMO treatment has grown widely as a rescue therapy with several indications like cardiogenic shock, ARDS, and other reasons. Table 3. Patients Survived Discharge (n = 15) Variables Good neurological outcome (CPC ≤ 2), n (%) Renal replacement therapy during admission, n (%) Stayed on RRT after admission, n (%) One year survival, n (%)
Value 15 (100) 7 (46.7) 3 (18.75) 11 (73.3)
CPC, cerebral performance category; RRT, renal replacement therapy.
It was described that VA ECMO placement during CPR could improve the outcome of patients undergoing cardiac arrest.3,4 According to the registry from the Extracorporeal Life Support Organization (ELSO) as of July 2016, 2,885 adult patients have received VA ECMO during ECPR. Survival to discharge rate was 29% in these patients.11 We found a survival rate of 25.4% in our institution and therefore are in the range of the documented ELSO data. Other publications describe a survival to discharge rate from 13% to 54%.12–20 In many publications, inclusion criteria are well selected and exclude patients with asystolie or PEA as initial heart rhythm.15,20 Bellezzo et al.20 defined not only persistent cardiopulmonary arrest as inclusion criteria but also shock (SBP < 70 mm Hg) refractory to standard therapies. Other reasons for high variation of the outcome between studies could be different comorbidities, different percentage of VT/pVT. In our study, all patients were placed on ECMO under ongoing mechanical compression (ECPR) and a high number of patients with asystolie or PEA were included. The survival rate of patients without asystolie or PEA as initial rhythm in our study is 40.7% and is significantly different from the survival to discharge from the patients with asystolie or PEA (8.7%; p = 0.04). ECPR is an invasive procedure, which requires resources including costs for staff and equipment. It should be clearly defined when patients likely would benefit from this procedure and when the resources, risk, and poor outcome does not justify the initiation of VA ECMO during CPR. Which outcome in a certain patient group justifies the initiation of this treatment needs to be discussed. However, knowing predictors for poor outcome might help to simplify the decision in which patient ECPR should be tried. Even if some studies show some
Copyright © American Society of Artificial Internal Organs. Unauthorized reproduction of this article is prohibited.
5
ECMO AND IMPROVED OUTCOME OF PATIENTS UNDER RESUSCITATION Table 4. Survival to Discharge and Potential Predictors for Survival (n = 59)
Variables Baseline plasma creatinine ≥ 2.0 mg/dL Baseline plasma creatinine < 2.0 mg/dL Unknown BMI ≥ 30 kg/m2 BMI < 30 kg/m2 Male Female Age ≥ 75 y Age < 75 y Age ≥ 65 y Age < 65 y Lactate ≥ 7 mmol/L Lactate < 7 mmol/L Unknown lactate Lactate ≥ 8 mmol/L Lactate < 8 mmol/L Unknown lactate Troponine I ≥ 10 ng/ml Troponine I < 10 ng/ml Unknown troponine PEA or asystole Ventricular arrhythmia Unknown rhythm CPR-time ≥ 30 min CPR-time < 30 min CPR-time unknown CPR-time ≥ 60 min CPR-time < 60 min CPR-time unknown AKI needed renal replacement therapy No need for RRT Location cath laboratory No cath laboratory Reason for cardiogenic shock Proven MI No proven MI Pulmonary embolism No pulmonary embolism Allograft rejection No allograft rejection Non ischemic CM No nonischemic CM Septic shock No septic shock ECPR score > 10 ECPR score ≤ 10
n/N (% survival)
p
2/13 (15.4) 10/38 (26.3) 3/8 (37.5) 7/36 (19.4) 8/23 (34.8) 7/40 (17.5) 8/19 (42.1) 0/5 (0.0) 15/54 (27.8) 3/20 (15.0) 12/39 (30.8) 7/36 (19.4) 6/14 (42.9) 2/9 (22.2) 4/30 (13.3) 9/20 (45.0) 2/9 (22.2) 9/27 (33.3) 4/20 (20.0) 2/12 (16.7) 2/23 (8.7) 11/27 (40.7) 2/9 (22.2) 6/32 (18.8) 3/10 (30.0) 6/17 (35.3) 1/9 (11.1) 8/33 (24.2) 6/17 (35.3) 7/27 (25.9) 8/32 (25.0) 6/16 (37.5) 9/43 (20.9)
0.42
10/29 (34.5) 5/30 (16.7) 1/4 (25.0) 14/55 (25.5) 2/5 (40.0) 13/54 (24.1) 0/2 (0.0) 15/57 (26.3) 0/2 (0.0) 15/57 (23.3) 4/10 (40.0) 4/23 (17.4)
0.12
0.19
Table 5. Complications Related to ECMO Treatment (n = 59) Variables
Value
Groin bleeding (nonsevere), n (%) 7 (11.9) Severe vessel bleeding (surgical intervention needed), n (%) 2 (3.4) Limb ischemia without treatment indication, n (%) 2 (3.4) Limb ischemia needed thrombectomy, n (%) 1 (1.7) Positive blood cultures, n (%) 4 (6.8) Hemolysis (plasmohemoglobin > 50 mg/dL), n (%) 19 (32.2)
0.04 0.17 0.19 0.09 0.01 0.31 0.04 0.42 0.37 0.94 0.19
0.98 0.43 1.00 1.00 0.16
BMI, body mass index; CPR, chest compression; PEA, pulseless electrical activity; RRT, renal replacement therapy.
predictors for poor survival rates (for example age ≤ 66 years, shockable arrest rhythm, lactate acid, CPR to ECMO pump-on time between (≤ 45 minutes) and comorbidity10), guidelines that propose in which patient groups ECMO placement during CPR should be initiated are still missing. In our patient cohort, lactate ≥ 8 mmol/L and as mentioned above cardiac rhythm of PEA or asystole were significant predictors for low survival rate. We did not find a significant age boarder, which limited the outcome in our study group. We could show that patients even with PEA and asystole at the time of ECPR could survive. Therefore, a nonshockable rhythm by itself should not be an exclusion criterion for ECMO placement and other criteria should be taken into account and also have to be identified in further studies.
Questionable is still after how many minutes of CPR outcome is futile and when attempt of ECPR should be avoided or withdrawn. Kelly et al.21 described two cases with good neurological outcome after 176 minutes and 97 minutes cardiopulmonary resuscitation before extracorporeal rescue. The CPR time in our study cohort was up to 75 minutes. Surprisingly, we did not find a significant correlation between CPR time and survival rate. Although ECPR can increase survival rates after cardiac arrest, it could also result in poor postresuscitation neurological status, and the data for this are still poor. Studies reported high serum lactic acid levels before ECMO not only as predictors for poor survival rate but also for poor neurological outcome.22 In our data, all survivors of discharge had a good neurological outcome with CPC of 1 or 2. It is still unclear which role acute kidney injury might play in terms of survival rate. Oliguria seen during the first 24 hours after ECMO may be an independent predictor of mortality.23 In our patient cohort, there was no significant higher mortality for patient needed CRRT during admission. The reason for the cardiac arrest also did not show any differences in outcome. Some studies mention that the outcome of ECPR in adult patients for septic shock is controversial.24,25 In our cohort, we had two patients in whom septic shock was diagnosed as reason for cardiac arrest in the aftermath. Both patients did not survive hospital discharge. Patients undergoing ECMO support through femoral vessels, especially under CPR are prone to complications.26–28 The incidence of complications in our cohort was comparable with complications reported in other publications.26,28,29 Most complications include femoral artery injuries and were manageable. Severe complication was limb ischemia needed fasciectomy and severe groin bleeding needing immediate intervention, but no amputation. Hemolysis during extracorporeal membrane oxygenation is a known complication.30 Daily pHb levels were measured. Elevated pHb could be seen in 19 patients (32.2%), which had no significant influence on survival to discharge rate. Experience of the provider (especially in terms of cannulation) could influence a successful placement of the cannulas and therefore increase the outcome.29 Technical challenge during CPR is particularly to place the arterial cannula because due to low flow the vessel is often small and under ongoing CPR difficult to detect. In this study, in five patients (8.5%) ECMO support could not be established. Most of our patients who were placed on ECMO in the cardiac catheter laboratory already had femoral arterial access that were used for placing the ECMO cannula. Anyway, there was no significant difference in mortality between patients who were cannulated in the cardiac catheter laboratory and patients who were placed somewhere else on ECMO.
Copyright © American Society of Artificial Internal Organs. Unauthorized reproduction of this article is prohibited.
6
PABST ET AL.
Our study has some limitations. Nine patients were cannulated in other hospitals and were transported for further care to our institution. Values for variables like first detected rhythm on CPR, baseline creatinine, troponine were sometimes difficult to evaluate in these patients. Chest compression time was well documented in 42 patients. In 17 patients, no reliable CPR time was documented. Male gender was a significant predictor for poor survival in our cohort. Reasons for this remain unclear, and it might be reevaluated in cohorts of bigger studies. Park et al.10 developed a score to monitor survival to discharge in patients after ECPR. The score ranges from 3 to 15 and includes the age of the patient, shockable arrest rhythm, CPR time, post-ECMO arterial pulse pressure, and the SOFA score. When the ECPR score was > 10, the sensitivity and specificity for prediction of survival to discharge in Parks study were 89.6% and 75.0%, respectively. The ECPR score could be identified in 33 patients in this cohort. Ten patients had an ECPR score of > 10. Forty percentage of these patients survived to discharge when compared with 17.4% of the patient with an ECPR score less than 10. Although there was a remarkable difference in survival between these two groups, our patients did not reach the significant difference in outcome expected by ECPR score. One reason might be the smaller cohort in our study. Anyhow, the ECPR score has already shown in a bigger cohort that it can be a useful tool for clinical providers.10 Although this study did not find any significant difference in the survival outcome between groups with longer and shorter CPR time, we believe that shortening time to establish a sufficient blood circle is important in patients with cardiac arrest. Logistic structures might improve survival rate by shortening time of CPR. Several case series could even show that prehospital placement of ECMO is successfully possible27,31,32 and might be considered as a challenge for future ECMO programs. The results of this study do not deviate much from the published ELSO data. Slightly lower survival to discharge results can be explained mainly with the higher number of patients having PEA/asystole while ECMO placement. Demographic differences in matters of age do not seem to play an important role in our study since we did not see a clear cutoff point for age. In conclusion, we showed that ECLS after cardiac arrest could be successful even in a nonhighly selected group of patients. In patients with high serum lactate levels, PEA or asystole presenting as initial heart rhythm, ECPR should be carefully considered since these were predictors for poor outcome in our study. References 1. Stiell IG, Wells GA, Field B, et al; Ontario Prehospital Advanced Life Support Study Group: Advanced cardiac life support in outof-hospital cardiac arrest. N Engl J Med 351: 647–656, 2004. 2. Olasveengen TM, Sunde K, Brunborg C, Thowsen J, Steen PA, Wik L: Intravenous drug administration during out-of-hospital cardiac arrest: A randomized trial. JAMA 302: 2222–2229, 2009. 3. Chen YS, Lin JW, Yu HY, et al: Cardiopulmonary resuscitation with assisted extracorporeal life-support versus conventional cardiopulmonary resuscitation in adults with in-hospital cardiac arrest: An observational study and propensity analysis. Lancet 372: 554–561, 2008. 4. Blumenstein J, Leick J, Liebetrau C et al. Extracorporeal life support in cardiovascular patients with observed refractory in-hospital
cardiac arrest is associated with favourable short and long-term outcomes: A propensity-matched analysis. Eur Heart J Acute Cardiovasc Care 2015. 5. Ortega-Deballon I, Hornby L, Shemie SD, Bhanji F, Guadagno E: Extracorporeal resuscitation for refractory out-of-hospital cardiac arrest in adults: A systematic review of international practices and outcomes. Resuscitation 101: 12–20, 2016. 6. Ouweneel DM, Schotborgh JV, Limpens J, et al: Extracorporeal life support during cardiac arrest and cardiogenic shock: A systematic review and meta-analysis. Intensive Care Med 42: 1922–1934, 2016. 7. Kim SJ, Kim HJ, Lee HY, Ahn HS, Lee SW: Comparing extracorporeal cardiopulmonary resuscitation with conventional cardiopulmonary resuscitation: A meta-analysis. Resuscitation 103: 106–116, 2016. 8. Riggs KR, Becker LB, Sugarman J: Ethics in the use of extracorporeal cardiopulmonary resuscitation in adults. Resuscitation 91: 73–75, 2015. 9. Cummins RO, Chamberlain DA, Abramson NS, et al: Recommended guidelines for uniform reporting of data from out-of-hospital cardiac arrest: the Utstein Style. A statement for health professionals from a task force of the American Heart Association, the European Resuscitation Council, the Heart and Stroke Foundation of Canada, and the Australian Resuscitation Council. Circulation 84: 960–975, 1991. 10. Park SB, Yang JH, Park TK, et al: Developing a risk prediction model for survival to discharge in cardiac arrest patients who undergo extracorporeal membrane oxygenation. Int J Cardiol 177: 1031–1035, 2014. 11. Registry from the Extracorporeal Life Support Organization (ELSO). 2017. Available at elso.org. 12. Johnson NJ, Acker M, Hsu CH, et al: Extracorporeal life support as rescue strategy for out-of-hospital and emergency department cardiac arrest. Resuscitation 85: 1527–1532, 2014. 13. de Chambrun MP, Bréchot N, Lebreton G, et al: Venoarterial extracorporeal membrane oxygenation for refractory cardiogenic shock post-cardiac arrest. Intensive Care Med 42: 1999–2007, 2016. 14. Mazzeffi MA, Sanchez PG, Herr D, et al: Outcomes of extracorporeal cardiopulmonary resuscitation for refractory cardiac arrest in adult cardiac surgery patients. J Thorac Cardiovasc Surg 152: 1133–1139, 2016. 15. Stub D, Bernard S, Pellegrino V, et al: Refractory cardiac arrest treated with mechanical CPR, hypothermia, ECMO and early reperfusion (the CHEER trial). Resuscitation 86: 88–94, 2015. 16. Maekawa K, Tanno K, Hase M, Mori K, Asai Y: Extracorporeal cardiopulmonary resuscitation for patients with out-of-hospital cardiac arrest of cardiac origin: A propensity-matched study and predictor analysis. Crit Care Med 41: 1186–1196, 2013. 17. Kagawa E, Inoue I, Kawagoe T, et al: Assessment of outcomes and differences between in- and out-of-hospital cardiac arrest patients treated with cardiopulmonary resuscitation using extracorporeal life support. Resuscitation 81: 968–973, 2010. 18. Mégarbane B, Leprince P, Deye N, et al: Emergency feasibility in medical intensive care unit of extracorporeal life support for refractory cardiac arrest. Intensive Care Med 33: 758–764, 2007. 19. Sakamoto T, Morimura N, Nagao K, et al; SAVE-J Study Group: Extracorporeal cardiopulmonary resuscitation versus conventional cardiopulmonary resuscitation in adults with outof-hospital cardiac arrest: A prospective observational study. Resuscitation 85: 762–768, 2014. 20. Bellezzo JM, Shinar Z, Davis DP, et al: Emergency physician initiated extracorporeal cardiopulmonary resuscitation. Resuscitation 83: 966–970, 2012. 21. Kelly RB, Porter PA, Meier AH, Myers JL, Thomas NJ: Duration of cardiopulmonary resuscitation before extracorporeal rescue: How long is not long enough? ASAIO J 51: 665–667, 2005. 22. Ryu JA, Cho YH, Sung K, et al: Predictors of neurological outcomes after successful extracorporeal cardiopulmonary resuscitation. BMC Anesthesiol 15: 26, 2015. 23. Lee JJ, Han SJ, Kim HS, et al: Out-of-hospital cardiac arrest patients treated with cardiopulmonary resuscitation using extracorporeal membrane oxygenation: Focus on survival rate and
Copyright © American Society of Artificial Internal Organs. Unauthorized reproduction of this article is prohibited.
ECMO AND IMPROVED OUTCOME OF PATIENTS UNDER RESUSCITATION
neurologic outcome. Scand J Trauma Resusc Emerg Med 24: 74, 2016. 24. Bréchot N, Luyt CE, Schmidt M, et al: Venoarterial extracorporeal membrane oxygenation support for refractory cardiovascular dysfunction during severe bacterial septic shock. Crit Care Med 41: 1616–1626, 2013. 25. Cheng A, Sun HY, Lee CW, et al: Survival of septic adults compared with nonseptic adults receiving extracorporeal membrane oxygenation for cardiopulmonary failure: A propensity-matched analysis. J Crit Care 28: 532.e1–532.10, 2013. 26. Bisdas T, Beutel G, Warnecke G, et al: Vascular complications in patients undergoing femoral cannulation for extracorporeal membrane oxygenation support. Ann Thorac Surg 92: 626– 631, 2011. 27. Lamhaut L, Jouffroy R, Soldan M, et al: Safety and feasibility of prehospital extra corporeal life support implementation by non-surgeons for out-of-hospital refractory cardiac arrest. Resuscitation 84: 1525–1529, 2013.
7
28. Aziz F, Brehm CE, El-Banyosy A, Han DC, Atnip RG, Reed AB: Arterial complications in patients undergoing extracorporeal membrane oxygenation via femoral cannulation. Ann Vasc Surg 28: 178–183, 2014. 29. Rupprecht L, Lunz D, Philipp A, Lubnow M, Schmid C: Pitfalls in percutaneous ECMO cannulation. Heart Lung Vessel 7: 320– 326, 2015. 30. Palanzo DA, El-Banayosy A, Stephenson E, Brehm C, Kunselman A, Pae WE: Comparison of hemolysis between CentriMag and RotaFlow rotary blood pumps during extracorporeal membrane oxygenation. Artif Organs 37: E162–E166, 2013. 31. Grunau B, Scheuermeyer FX, Stub D, et al: Potential candidates for a structured Canadian ECPR Program for out-of-hospital cardiac arrest. CJEM 18: 453–460, 2016. 32. Kippnich M, Lotz C, Kredel M, et al: [Venoarterial extracorporeal membrane oxygenation for out-of-hospital cardiac arrest. Case series of prehospital and in-hospital therapies]. Anaesthesist 64: 580–585, 2015.
Copyright © American Society of Artificial Internal Organs. Unauthorized reproduction of this article is prohibited.
