Extracorporeal cardiopulmonary resuscitation for in- and out-of-hospital cardiac arrest: The race against time

Abstract

Objectives

Extracorporeal cardiopulmonary resuscitation (ECPR) is increasingly used due to its beneficial outcomes and results compared to conventional CPR. Cardiac arrests can be categorized depending on location: in-hospital cardiac arrest (IHCA) and out-of-hospital cardiac arrest (OHCA). Despite this distinction, studies comparing the two are scarce, especially in comparing outcomes after ECPR. This study compared patient characteristics, cardiac arrest characteristics, and outcomes.

Methods

Between 2016 and 2022, patients who underwent ECPR for cardiac arrest at our institution were retrospectively analyzed, depending on the arrest location: IHCA and OHCA. We compared periprocedural characteristics and used multinomial regression analysis to indicate parameters contributing to a favorable outcome.

Results

A total of n = 157 patients (100%) were analyzed (OHCA = 91; IHCA = 66). Upon admission, OHCA patients were younger (53.2 ± 12.4 vs. 59.2 ± 12.6 years) and predominantly male (91.1% vs. 66.7%, p=<0.001). The low-flow time was significantly shorter in IHCA patients (41.1 ± 27.4 mins) compared to OHCA (63.6 ± 25.1 mins). Despite this significant difference, in-hospital mortality was not significantly different in both groups (IHCA = 72.7% vs. OHCA = 76.9%, p = 0.31). Both groups' survival-to-discharge factors were CPR duration, low flow time, and lactate values upon ECMO initiation.

Conclusion

Survival-to-discharge for ECPR in IHCA and OHCA was around 25%, and there was no statistically significant difference between the two cohorts. Factors predicting survival were lower lactate levels before cannulation and lower low-flow time. As such, OHCA patients seem to tolerate longer low-flow times and thus metabolic impairments compared to IHCA patients and may be considered for ECMO cannulation on a broader time span than IHCA.

Introduction

There are about 350,000 estimated out-of-hospital cardiac arrests (OHCA) in the European Union annually,¹ and cardiac arrest (CA) is considered the third leading cause of death in Europe.2, 3 The reported survival rates range from 15%-34% for in-hospital cardiac arrest (IHCA) and, on average, 8% for OHCA treated by EMS.² As such, considerable efforts have been made to understand the causes and pathophysiology of CA to increase survival rates, which have not been significantly improved for decades.

A recently introduced technique that shows promising improvements in survival rates is extracorporeal membrane oxygenation (ECMO) during CA (extracorporeal cardiopulmonary resuscitation = ECPR).4, 5, 6, 7 While two randomized controlled trials (RCT) showed survival rates of ≥ 30% for ECPR in OHCA, high-quality data for ECPR in IHCA is lacking.

Based on the availability of these RCTs for OHCA, beneficial cannulation criteria for ECPR in OHCA become more evident (e.g., initial-shockable rhythm, CPR-time < 60 mins, high-quality CPR). Nevertheless, it is unclear if these inclusion criteria may also be utilized for IHCA because IHCA patients are a very different patient population.8, 9

Is there a need to establish distinct inclusion criteria for IHCA and OHCA? There is some evidence that OHCA incidents typically involve sudden, unexpected, and catastrophic events without prior warning. The affected individuals are usually younger, have fewer comorbidities, and may endure longer periods of low-flow circulation with more severe physiological derangements before Extracorporeal Membrane Oxygenation (ECMO) is initiated.10, 11, 12 However, OHCA encompasses a diverse range of pathologies, some of which may respond well to ECMO, while others, such as intracranial hemorrhage or occult malignancy, may not.13, 14 Additionally, obtaining information on past medical history for candidacy assessment is often challenging in OHCA. In contrast, IHCA cases exhibit a higher prevalence of comorbidities, and the underlying pathology is more likely to be known, aiding in the identification of suitable candidates. Moreover, the hospital-based location facilitates shorter low-flow times, potentially allowing for broader inclusion criteria, including non-shockable rhythms, where the underlying pathology may be more susceptible to low-flow intervals.

As ECPR is a relatively new technique, there is a dearth of studies comparing the outcomes of IHCA vs. OHCA with ECPR. As such, if essential differences in arrest characteristics and outcomes warrant consideration, this knowledge gap might be problematic.

Besides, the international guidelines for the treatment of CA are almost identical in IHCA and OHCA, irrespective of the location of arrest.15, 16

In this study, we investigated baseline and CA characteristics in IHCA and OHCA with ECPR, exploring factors that may explain differences in outcomes and arrest characteristics with respect to the heterogeneity of both populations.

Patients and methods

This is a retrospective single-center study from January 2016 until December 2020.

Ethical statement

The local Ethics Committees of the University of Cologne (No. 20–1262) approved the study. Individual patient consent was waived due to the retrospective design of the study. The study conformed to the principles outlined in the Declaration of Helsinki.

Institutional ECPR protocol

Our institution's ECPR protocol for OHCA has been described before. 17, 18, 19 For OHCA with ongoing CPR, physicians and surgeons are involved in further treatment. ECMO (Maquet, Rastatt, Germany) implantations are performed in the catheterization laboratory, intensive care unit (ICU), operating room, or emergency department. After a general clinical examination and initial echocardiographic evaluation, the left or right groin is sterilized and prepared for ECMO implantation. Additional concomitant blood gas analysis focuses on assessing the initial parameters of oxygenation and metabolism. The decision to cannulate is based on participating physicians' judgment, following institutional criteria for OHCA:

ECPR inclusion criteria OHCA

• -No terminal illness
• -Age ≤ 75 years
• -Observed cardiac arrest
• -No-flow-time ≤ 5 mins
• -Low flow time (time under CPR until the commencement of ECMO) ≤ 90 mins
• -refractory shockable rhythm or non-shockable in case of cardiac tamponade or pulmonary embolism

ECPR exclusion criteria OHCA

• -arterial pH < 6.6
• -non-shockable rhythm > 20 mins after examination for cardiac tamponade or pulmonary embolism
• -hemoglobin < 8 g/dL

In the case of IHCA, the cannulation decision was mainly based on case-to-case decisions. Patients were solely cannulated in the ICU and the OR.

ECMO cannulas are implanted via femoral vessels using Seldinger's technique. Extracorporeal circulation is established by using a Rotaflow centrifugal pump (Maquet, Rastatt, Germany) with a Quadrox membrane oxygenator (Maquet, Rastatt, Germany) and the PLS2050 circuit system (Maquet, Rastatt, Germany).

To avoid thromboembolic events during ECMO circulation, our anticoagulatory protocol includes the intravenous infusion of unfractionated heparin, with a target-activated clotting time (ACT) of 140 to 160 seconds and PTT between 60 and 80 seconds.

Neuroprognostication protocol

Our institutional neuro prognostication protocol includes a post-ECPR whole-body CT scan repeated after 72 hours. Furthermore, lab tests include neuron-specific enolase test (NSE) and long-term EEG.

Variables of interest

The analysis was performed using a retrospectively maintained institutional patient database. The variables evaluated included: patient demographics (age, height, weight, sex), patients' status before ECMO support, CPR data, pre-implantation parameters in blood-gas analysis, the cause of out-of-hospital cardiac arrest (OHCA) or in-hospital cardiac arrest (IHCA), outcome data on ICU (blood transfusions, in-hospital mortality, discharge, hospital, and ICU stay). The low-flow time was defined as the time between the beginning of chest compressions until the commencement of extracorporeal circulation; no-flow time was defined as the time between CA and the beginning of CPR.

Endpoints of the study

Our analysis's primary endpoint was survival to discharge.

Secondary endpoints were: length of ICU- and in-hospital stay, length of ECMO support, brain death, and the feasibility of ECMO weaning and withdrawal.

Statistical methods

Statistical analysis was performed using Statistical Package for Social Sciences, version 27.0 (SPSS IBM, Chicago, Illinois) and has been motioned before.20, 21, 22, 23 All data were presented as continuous or categorical variables. Categorical data were expressed as total numbers and percentages. Continuous data were evaluated for normality using a one-sample Kolmogorov-Smirnov test and were expressed as the mean ± standard deviation (SD) is normally distributed or median (interquartile range) in cases of non-normally distributed continuous variables. Univariate analysis was performed using either Student́s test or Mann-Whitney U test for normally and non-normally distributed continuous variables, respectively. Pearson's χ² or Fisheŕs exact tests were used to compare categorical data depending on the minimum expected count in each cross-tab. P values < 0.05 were considered statistically significant.

Backward multivariate logistic regression analysis by odds ratio (OR) and 95% confidence interval (CI) was conducted for factors significant in the univariate analysis, assuming they were independent factors for hospital survival. Factors which remained significant in multivariate logistic regression analysis. Were included in a a propensity score-matched analysis with 0.2 standard deviation of the logit of the propensity score to reduce confounding factors.

Median overall survival was assessed by Kaplan-Meier curves and compared by log-rank test. P values < 0.05 were considered statistically significant.

Results

A total of 157 patients who underwent ECPR for out-of-hospital cardiac arrest (OHCA, n = 91) and intra-hospital cardiac arrest (IHCA, n = 66) at our institution were retrospectively analyzed (Fig. 1).

Fig. 1.

Kaplan-Meier survival curves for the OHCA and IHCA cohort.

Baseline characteristics

The baseline characteristics are presented in Table 1.

Table 1.

Baseline characteristics before ECMO-cannulation.

	Total cohort n = 157	IHCA n = 66	OHCA n = 91	p-value
Age (years)	55.7 ± 12.8	59.2 ± 12.6	53.2 ± 12.4	0.711
Male gender (%)	80.3	66.7	91.1	<0.001
Body-Mass Index	27.9 ± 6.4	28.0 ± 2.6	27.4 ± 4.9	0.09
Low-Flow Time (min)	56.08 ± 28.0	41.1 ± 27.4	63.6 ± 25.1	<0.001
Lactate (mmol/L)	12.9 ± 8.77	11.1 ± 6.4	14.4 ± 10.6	0.71
pH	6.92 ± 0.36	7.1 ± 0.2	6.8 ± 0.8	0.008
pO2 (mmHg)	93.1 ± 60.7	108.4 ± 117.4	70.0 ± 36.2	0.08
pCO2 (mmHg)	60.2 ± 27.7	56.6 ± 26.6	64.9 ± 25.0	0.30
Hb prior (g/dL)	11.2 ± 3.2	10.9 ± 2.8	12.19 ± 3.6	0.04
CK (U/L)	3376 ± 5936	2169 ± 3540	4380 ± 6600	0.02
CK-MB %	44.5 ± 156	31.8 ± 76.8	49.06 ± 162.1	0.50
Charleston-Comorbidity Score		4.1 ± 2.5
Presenting rhythm
shockable (%)	59.5	43.8	65.6	0.04
PEA (%)	24.3	40.6	17.7	<0.001
Asystole (%)	16.2	15.6	16.5	0.76

The mean age of the whole cohort was 55.7 ± 12.8 years. Overall, there were significantly fewer male patients in the IHCA group compared to the OHCA group (66.7% vs. 91.1%, p=<0.001).

The low-flow time was significantly shorter in the IHCA arm (41.1 ± 27.4 min vs. 63.6 ± 25.1 min, p=<0.001). Upon ECMO-cannulation, initial blood gas analysis showed significant differences in acquired pH-values between the groups (IHCA: 7.1 ± 0.2 vs. 6.8 ± 0.8, p = 0.008), there were no statistical differences between the groups in initial lactate, pO2 or pCO2.

The initial acquired electrocardiogram showed significantly fewer shockable rhythms in the IHCA group (43.8%) compared to the OHCA group (65.5%), p = 0.04. Whereas pulseless electrical activity (PEA) was significantly more often seen in the IHCA group (40.6%) compared to the OHCA group (17.7%), p=<0.001.

Outcome data

ICU outcome data are presented in Table 2.

Table 2.

ICU-stay outcome data.

	Total cohort n = 157	IHCA n = 66	OHCA n = 91	p-value
Limb ischemia (%)	26.0	25.1	26.8	0.84
Active bleeding (%)	63.6	87.0	89.2	0.45
PRBC	12.2 ± 13.9	14.9 ± 18.6	10.4 ± 10.6	0.10
GI-complications	40.1	40.1	40.0	0.99
ECMO-explant (%)	38.9	53.0	28.1	0.05
ECMO duration (%)	3.61 ± 2.49	4.0 ± 2.1	3.8 ± 2.8	0.89
CVVH (%)	32.2	40.7	24.04	0.04
In-hospital mortality (%)	75.2	72.7	76.9	0.31
ICU stay (days)	8.2 ± 13.8	10.1 ± 6.98	6.8 ± 14.2	0.16
Hospital stay (days)	11.4 ± 19.5	15.6 ± 7.7	6.7 ± 16.2	0.36

PRBC = packed red blood cells; GI = gastro-intestinal, CVVH = continuous veno-venous hemofiltration

The most common complication in the cohort was the need for continuous venovenous hemofiltration (CVVH) in 32.2%. The CVVH was significantly more often used in IHCA (40.7%) compared to OHCA (24.04), p = 0.04. Limb ischemia was present in 26% of the whole cohort. In-hospital mortality was 75.2% in the whole cohort, with no statistically significant difference between the groups (Fig. 1). ICU stay was 8.2 ± 13.8 days, and ECMO duration was 3.61 ± 2.49 days in the cohort. We assessed overall survival in the two groups statistically using the Kaplan-Meier method (Fig. 1).

Subgroup analysis survivors versus non-survivors, the whole cohort

We conducted a subgroup analysis with a multinomial regression analysis to investigate factors associated with a favorable outcome, which are depicted in Table 3.

Table 3.

Baseline characteristics before ECMO-cannulation.

	Total cohort n = 157	Survived n = 39	In-hospital death n = 118	p-value	Multinominal P-value
Age (years)	55.7 ± 12.8	53.8 ± 11.6	56.3 ± 13.1	0.28
Age > 70 years (%)	39	5.5	14.4	0.009	0.14
Male gender (%)	80.3	82.1	79.7	0.82
IHCA (%)	42.0	46.2	40.7	0.33
OHCA (%)	58.0	53.8	59.3	0.33
Witnessed arrest (%)	82.8	94.6	88.8	0.24
No-flow time (min)	3.25 ± 4.9	1.5 ± 2.3	3.8 ± 5.4	0.03	0.14
Low-flow time (min)	56.08 ± 28.0	43.6 ± 22.4	60.6 ± 28.1	<0.001	0.005
Lactate (mmol/L)	12.9 ± 8.77	8.6 ± 5.8	13.9 ± 9.6	0.003	0.002
pH	6.92 ± 0.36	7.0 ± 0.2	6.9 ± 0.8	0.15
pO2 (mmHg)	93.1 ± 60.7	99.5 ± 31.4	65.5 ± 36.2	0.05	0.12
pCO2 (mmHg)	60.2 ± 27.7	61.6 ± 29.6	54.6 ± 13.0	0.06
Hb (g/dL)	11.2 ± 3.2	10.9 ± 2.8	12.19 ± 3.6	0.14

There were significant differences between survivors and non-survivors in baseline characteristics. As such, survivors showed a significantly shorter no-flow time (1.5 ± 2.3 min vs. 3.8 ± 5.4 min, p = 0.03), a significantly lower low-flow-time (43.6 ± 22.4 min vs. 60.6 ± 28.1 min, p=<0.001) as well as significantly lower lactate levels upon cannulation (8.6 ± 5.8 mmol/l vs. 13.9 ± 9.6 mmol/l, p = 0.003). Additionally, survivors showed higher pO2 levels (99.5 ± 31.4 mmHg vs. 65.5 ± 36.2 mmHg, p = 0.05) and were less likely to depict asystole as their initial acquired rhythm (8.3% vs. 24.1%, p = 0.01). After multinominal regression analysis, factors significantly influencing survival were low-flow-time and initial lactate upon ECMO cannulation.

Subgroup analysis survivors versus non-survivors, OHCA cohort

The subgroup analysis for the OHCA cohort is shown in Table 4.

Table 4.

Baseline characteristics before ECMO-cannulation OHCA.

	Total cohort n = 91	Survived n = 21	In-hospital death n = 70	p-value	Multinominal P-value
Age (years)	53.2 ± 12.4	53.8 ± 11.6	56.3 ± 13.1	0.26
Age > 70 years (%)	7.7	0	10.0	/	/
Male gender (%)	91.1	82.1	79.7	0.65
witnessed arrest (%)	86.9	94.7	84.6	0.23
No-flow time (min)	4.3 ± 5.2	1.5 ± 2.3	3.8 ± 5.4	0.04	0.15
Low-flow time (min)	63.6 ± 25.1	54.6 ± 16.8	68.0 ± 26.5	0.03	0.05
shockable rhythm (%)	64.6	68.4	63.3	0.91
Lactate (mmol/L)	14.4 ± 10.6	11.6 ± 5.1	15.5 ± 10.6	0.003	0.04
pH	6.8 ± 0.8	6.9 ± 0.1	6.7 ± 0.9	0.26
pO2 (mmHg)	70.0 ± 36.2	71.2 ± 31.4	67.8 ± 39.2	0.05	0.13
pCO2 (mmHg)	64.9 ± 25.0	61.3 ± 12.6	65.6 ± 27.0	0.14
Hb prior (g/dL)	12.19 ± 3.6	13.7 ± 2.8	11.8 ± 3.6	0.11

There were significant differences between survivors and non-survivors in the baseline characteristics. As such, survivors showed a significantly shorter no-flow time (1.5 ± 2.3 min vs. 3.8 ± 5.4 min, p = 0.03), a significantly lower low-flow-time (54.6 ± 16.8 min vs. 68 ± 26.5, p = 0.03) as well as significantly lower lactate levels upon cannulation (11.6 ± 5.1 mmol/l vs. 15.5 ± 10.6 mmol/l, p = 0.003). Additionally, survivors showed higher pO2 levels (71.2 ± 31.4 mmHg vs. 67.8 ± 39.2 mmHg, p = 0.05). After multinominal regression analysis, factors significantly influencing survival were low-flow-time and initial lactate upon ECMO cannulation.

Subgroup analysis survivors versus non-survivors, IHCA cohort

The subgroup analysis for the OHCA cohort is shown in Table 5.

Table 5.

Baseline characteristics before ECMO-cannulation IHCA.

	Total cohort n = 66	Survived n = 18	In-hospital death n = 48	p-value	Multinominal P-value
Age (years)	59.2 ± 12.6	55.6 ± 13.1	60.6 ± 12.1	0.09
Age > 70 years (%)	18.2	11.1	20.8	0.03	0.36
Male gender (%)	66.7	72.2	64.6	0.31
witnessed arrest (%)	95	94.4	95.2	0.95
No-flow time (min)	0.0	0.0	0.0	/	/
Low-flow time (min)	63.6 ± 25.1	23.5 ± 15.2	45.0 ± 28.6	0.03	0.02
shockable rhythm (%)	33.3	60	37	0.05
Lactate (mmol/L)	14.4 ± 10.6	5.3 ± 4.8	12.5 ± 6.0	0.002	0.002
pH	7.1 ± 0.2	7.2 ± 0.2	7.1 ± 0.2	0.15
pO2 (mmHg)	108.4 ± 117.4	116.8 ± 26.4	64.8 ± 28.2	0.05	0.13
pCO2 (mmHg)	56.6 ± 26.6	61.3 ± 12.6	65.6 ± 27.0	0.53
Hb (g/dL)	10.9 ± 2.8	11.7 ± 2.8	10.6 ± 2.6	0.46

There were significant differences between survivors and non-survivors in the baseline characteristics. There were significantly more patients over 70 years in non-survivors compared to survivors (20.8% vs. 11.1%, p = 0.003). Additionally, survivors showed a significantly shorter low-flow time (23.5 ± 15.2 min vs. 45 ± 28.6, p = 0.03) as well as significantly lower lactate levels upon cannulation (5.3 ± 4.8 mmol/l vs. 12.5 ± 6.0 mmol/l, p = 0.002). Additionally, survivors showed higher pO2 levels (116.8 ± 26.4 mmHg vs. 64.8 ± 28.2 mmHg, p = 0.05). After multinominal regression analysis, factors significantly influencing survival were low-flow-time and initial lactate upon ECMO cannulation.

Survival in relation to low-flow-time

The percentage of survivors and non-survivors concerning low-flow time (<30 mins, <60 mins, and < 120 mins) is depicted in Fig. 2.

Fig. 2.

Survival related to low flow time in the IHCA and OHCA cohort. All analysed patients have been divided based on their low-flow times in three distinct time intervals: <30 mins of low-flow time, >30–<60 mins and > 60–<120 mins. Survival rates in all 3 times intervals are displaced.

In the IHCA group, the majority of deaths occurred in a time interval between ≥ 30 mins and ≤ 60 mins of flow-flow time (n = 22; 46.7%). No patient survived after ≥ 60 mins of low-flow time and ≤ 120 mins.

In the OHCA group, the majority of deaths occurred in a time interval between ≥ 30 mins and ≤ 60 mins of flow-flow time (n = 34; 48.3%). After ≥ 60 mins of low-flow time and ≤ 120 mins, n = 4, 20% of patients survived, whereas 40% of deaths occurred.

Discussion

In this retrospective analysis comparing CA patients who received ECPR, we showed that:

• (1)Survival to discharge was 24.8% for the whole cohort; there was no statistically significant difference between survival in IHCA (27.3%) versus OHCA (23.1%).
• (2)Factors associated with survival were low-flow time and lactate upon ECMO cannulation, regardless of where CA occurred.
• (3)Patients with OHCA seem to tolerate longer low-flow times compared to IHCA and thus lower pH levels than IHCA patients.

In this analysis, we showed that survival-to-discharge with using ECMO, regardless of whether the arrest occurred inside or outside the hospital, was nearly 25%. This is a comparably large improvement compared to universally published data, stating a worldwide survival after OHCA of below 9%.²⁴ Data analyzing survival after OHCA treated with ECPR are heterogeneous. There have been three RCTs investigating outcomes for ECPR after OHCA.5, 6, 7 The reported survival rates of these RCTs for patients with OHCA treated with ECPR are between 20–43%. There is a dearth of studies comparing IHCA to OHCA patients, especially in ECPR. Our reported outcome data fit along with the most recent highest-quality publications. Additionally, we only deemed patients eligible for ECPR who had ongoing CPR and had not achieved return of spontaneous circulation ROSC. This is an essential fact because ECPR is defined as the commencement of ECMO in a period of up to twenty minutes from CA to ROSC; as such, the outcomes of patients who have already achieved ROSC will be superior compared to the ones with ongoing CPR.

Besides the RCT for ECPR in OHCA, no RCTs investigated the effect of intra-arrest ECMO in IHCA. One of the largest retrospective analyses by Chen et al. investigated the survival in 135 patients after IHCA treated with ECPR.²⁵ In a post-hoc propensity-score matched analysis, survival to hospital discharge was 23.7% in the ECPR group and 10.6% in the conventional CPR group.²⁶ In a retrospective analysis, survival after two years was investigated for ECPR and conventional CPR for IHCA. Shin et al. showed 20% survival in ECPR patients compared to 5% in the conventional CPR group.²⁷ Concerning our reported survival for IHCA, those numbers fit into already published registries. However, there is much room for bias in investigating ECPR for IHCA, mainly because patients may have a plethora of different clinical conditions that may have led to CA in the hospital. As such, most patients treated in the IHCA arm of our study are post-cardiac surgery. As such, in an attempt by Tonna et al. to design a risk score for ECPR in IHCA, post-cardiac surgery was shown to be beneficial in terms of overall survival.⁸

Investigating time under CPR (low-flow time) as a major determinant of survival

In general, survival after 45 min of CPR is grim28, 29. However, based on available published data from RCT, utilizing ECMO in CA offers a survival advantage when applied in an experienced center.⁴

ECPR is a time-critical intervention. Regardless of the location of CA, whether inside or outside of the hospital, low flow is considered the primary determinant of neurologically favorable outcomes.³⁰ It is defined as the time from initiation of resuscitation to initiation of extracorporeal membrane oxygenation. Previous analysis showed a lower low-flow time was associated with favorable outcomes for OHCA patients.31, 32, 33 However, low-flow durations in OHCA patients are considerably longer, compared to IHCA, due to the required time for transportation to the hospital and resuscitation in the field. In a pooled analysis of two RCTs for ECPR in OHCA, Belohlavek et al. showed that the positive effect of utilizing ECPR was the most prominent in patients with low-flow times over 45 mins and shockable rhytms.⁴

Additionally, Bartos et al. showed a neurologically favorable survival of 100% of patients after OHCA when cannulated for ECMO within 20–29 minutes of arrest, compared to standard CPR.²⁸

For IHCA patients, the outcomes were expected to be superior to OHCA because of the rapid availability of advanced life support and ECMO. In this regard, only a few studies are available investigating the effect of low-flow time in IHCA, and outcomes were not sufficiently investigated.10, 12, 33, 34 Nevertheless, it has been shown that survival-to-discharge in these studies was between 31.7–43.9% with a median low flow time of 27.0–30.0 minutes. In 2023, Ohbe et al. investigated 9,844 patients who received ECPR in IHCA in Japan.³⁵ They showed that the estimated survival-to-discharge rate dropped sharply by nearly 20% after the first 35 minutes of low-flow time and that the chance of survival to discharge was below 8% after 90 minutes of CPR for IHCA. This is in line with our results, showing 0% survival between 60–120 mins of CPR for IHCA. However, this effect could also be heavily influenced by the fact that IHCA patients showed significantly more non-shockable rhythms upon the first analysis in our cohort. Furthermore, initial-shockable rhythm has been reported to be a critical factor influencing survival in ECPR patients.⁴

Overall, one could conclude that a later initiation of ECPR may worsen the risk of hypoxic brain injury and multiple organ failures.

However, there seems to be a significant discrepancy in tolerable low-flow times between IHCA and OHCA cohorts.

This ultimately leads to the question if IHCA is sufficiently different from OHCA. As such, OHCA involves a population of patients who suffer a sudden unexpected CA. This population has been reported to be younger and with fewer comorbidities, although data on baseline difference is heterogenic.10, 11, 36 Also, OHCA includes mostly patients with CA due to myocardial infarction (MI), a condition which has shown to benefit from ECPR and early revascularization.37, 38 These differences in comorbidities and overall health may partly explain the already reported differences in the OHCA patients' ability to sustain longer low-flow times with favorable outcomes compared to IHCA.³⁹

Additionally, it was previously shown that there is a relation between the duration of arrest and rising levels of lactate, and thus lower pH was an independent predictor of mortality.28, 40, 41

As such, there is now evidence to support that the optimal window for initializing ECPR on OHCA is anywhere between 20 and 60–70 minutes of low-flow time, whereas, in IHCA, the cutoff point where a favorable outcome may no longer be reached might be under 40 minutes.

Further investigations, especially in ECPR for IHCA, are needed to better understand the ECPRs effect in IHCA.

Conclusion

Survival-to-discharge for ECPR in IHCA and OHCA was around 25%, and there was no statistically significant difference between the two cohorts. Furthermore, factors predicting survival were lower lactate levels before cannulation and lower low-flow time. As such, OHCA patients seem to tolerate longer low-flow times and thus metabolic impairments compared to IHCA patients and may be considered for ECMO cannulation on a broader time span than IHCA.

Limitations

The present study has some limitations. The definition of ECPR in our study is purely based on ongoing CPR, whereas in the literature, it may include percutaneous extracorporeal membrane oxygenation after the sustained return of spontaneous circulation (after 20 consecutive minutes following cardiac arrest) or chest compression after the establishment of percutaneous extracorporeal membrane oxygenation. Furthermore, we were unable to report long-term survival outcomes. Additionally, as stated above, our IHCA cohort included mainly patients after cardiac surgery, which might add to better outcomes in general. Furthermore, the results have to be interpreted cautiously since it was a retrospective single-center analysis.

Source of funding

No funding was provided for this work.

Data availability

Raw data were generated at the University Hospital of Cologne, Department of Cardiac Surgery. Derived data supporting the findings of this study are available from the corresponding author Gaisendrees C. on request.

CRediT authorship contribution statement

Christopher Gaisendrees: Conceptualization. Georg Schlachtenberger: Data curation. Lynn Müller: Resources, Data curation. Deborah Jaeger: Conceptualization. Ilija Djordjevic: Project administration. Ihor Krasivskyi: Methodology, Investigation. Ahmed Elederia: Resources. Sebastian Walter: Visualization. Mattias Vollmer: Resources, Conceptualization. Carolyn Weber: Investigation. Maximilian Luehr: Visualization, Resources. Thorsten Wahlers: Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.