15. Extracorporeal Cardiopulmonary Resuscitation (ECPR)

Quick Recap

Cross-cutting protocol — companion to Post-Cardiac Arrest Syndrome (Neurology System) and Vasopressor & Inotrope Selection & Titration protocols. Addresses a genuinely discordant, actively evolving evidence base across three major RCTs (ARREST, Prague OHCA, INCEPTION) with different intention-to-treat results — substantially explained, though not fully resolved, by system logistics and time-to-cannulation factors rather than the underlying physiological rationale being disproven.

1. Definition

Extracorporeal Cardiopulmonary Resuscitation (ECPR): the use of veno-arterial extracorporeal membrane oxygenation (VA-ECMO) initiated during ongoing cardiopulmonary resuscitation for cardiac arrest refractory to conventional advanced cardiac life support (ACLS) — aims to restore systemic perfusion during arrest, enabling diagnosis and treatment of reversible causes (notably acute coronary occlusion) while "buying time" for definitive treatment or spontaneous cardiac recovery.

Refractory cardiac arrest: cardiac arrest persisting despite conventional ACLS for a defined duration (commonly >15–20 minutes, with specific thresholds varying by trial/protocol) without return of spontaneous circulation — conventional CPR is generally understood to be minimally effective beyond approximately 20–30 minutes, which is the physiological rationale for considering an alternative circulatory support strategy in appropriately selected patients.

Low-flow time: the duration of inadequate systemic perfusion during the resuscitation process, from arrest onset to restoration of adequate flow (either via ROSC or ECMO initiation) — a critical, time-dependent variable that the current evidence base identifies as one of the central determinants of ECPR success, explaining much of the divergence between trials (Section 11).

2. Pathophysiology

Conventional CPR generates only a fraction of normal cardiac output (typically 10–30% of baseline), providing marginal, time-limited perfusion to vital organs — as resuscitation duration extends, the cumulative hypoxic-ischemic burden to the brain and other organs progressively worsens, and the probability of meaningful neurological recovery even with eventual ROSC declines sharply. ECPR's physiological rationale is to restore near-normal systemic perfusion via mechanical circulatory support, interrupting this progressive ischemic injury while a reversible underlying cause (most commonly acute coronary occlusion, amenable to percutaneous intervention) is identified and treated.

Why this is fundamentally a time-dependent, systems-dependent intervention, not merely a device efficacy question: ECPR's potential benefit is critically dependent on minimizing low-flow time — the interval between arrest onset and restoration of adequate perfusion (via ECMO flow) — meaning the intervention's real-world effectiveness is inseparable from the speed and reliability of the surrounding system: prehospital recognition and transport logistics, in-hospital cannulation team readiness, and institutional experience/case volume. This system-dependency is the central lens through which the discordant trial results (Section 11) should be understood.

3. Immediate Stabilization (ABCDE) — ECPR Candidacy and Activation

Not a standalone acute stabilization scenario in the traditional sense; ECPR activation is itself a time-critical decision embedded within ongoing resuscitation:

Checklist:

4. Focused History

• Time of arrest onset and witnessed status (essential for total downtime/low-flow time calculation)
• Bystander CPR provision and quality
• Initial rhythm (shockable [VT/VF] vs. non-shockable [PEA/asystole] — a consistent factor across trials associated with differential benefit, Section 22)
• Presumed etiology (cardiac vs. non-cardiac)
• Relevant comorbidities affecting candidacy and prognosis

5. Comprehensive System-wise Examination

During active resuscitation, examination is necessarily limited to standard ACLS assessment (rhythm, pulse checks, quality of compressions/perfusion); post-cannulation assessment shifts to standard post-cardiac-arrest and ECMO-specific monitoring (cross-reference Post-Cardiac Arrest Syndrome protocol, Neurology System).

POCUS integration: bedside echocardiography during resuscitation can help identify a reversible mechanical cause (tamponade, massive PE with RV strain) and assess for spontaneous cardiac activity, informing both the ECPR decision and cannulation planning.

6. Syndrome Identification — Reframed as Candidacy Classification

• Witnessed, shockable-rhythm refractory arrest, short anticipated low-flow time, high-volume/experienced ECPR system: the population with the clearest, most consistent evidence of benefit across trials (Section 11)
• Non-shockable rhythm, unwitnessed arrest, or anticipated prolonged low-flow time: substantially less favorable evidence base; institutional criteria vary considerably in how (or whether) these patients are considered candidates
• Low-volume or logistically less mature ECPR system: the INCEPTION trial's findings (Section 11) specifically caution that system maturity/experience meaningfully modifies expected outcomes — candidacy assessment should account for the actual capability of the treating system, not only the patient's individual clinical characteristics

7. Differential Diagnosis — Reversible Causes to Actively Identify Once Perfusion Is Restored

Once ECMO flow is established, the clinical priority shifts to identifying and treating the underlying reversible cause — most commonly acute coronary occlusion (warranting urgent coronary angiography/PCI), but also pulmonary embolism, tamponade, and other classically reversible arrest etiologies (cross-reference relevant Cardiovascular System protocols).

8. Severity/Risk Assessment

Low-flow time: the single most consistently identified determinant of outcome across this literature — median downtime/low-flow times differed substantially across the major trials and appear to substantially explain their divergent results (Section 11).

Initial rhythm: shockable rhythms (VT/VF) consistently associated with better outcomes than non-shockable rhythms across the pooled analysis and individual trials.

Center volume/system experience: an ELSO registry analysis specifically found survival after ECPR was associated with modifiable post-resuscitation practices and center volume — reinforcing that outcomes are not purely a function of patient selection but also of institutional capability and experience.

9. Investigations

Immediate bedside: continuous rhythm/perfusion monitoring during resuscitation; POCUS for reversible mechanical causes and cardiac activity assessment

Post-cannulation: urgent coronary angiography in most protocols given the high prevalence of acute coronary occlusion as the underlying cause; standard post-cardiac-arrest workup (cross-reference Post-Cardiac Arrest Syndrome protocol) once perfusion is restored

10. Point-of-Care Ultrasound

Central to both identifying reversible mechanical causes of arrest (tamponade, massive PE) during the resuscitation itself and to guiding/confirming cannulation (vascular access site assessment, cannula positioning).

11. Evidence-Based Management — The Discordant Trial Evidence, Explained in Detail

Three RCTs, Three Different Intention-to-Treat Results

• ARREST trial (Yannopoulos et al., Lancet 2020, single-center, Minnesota, n=30, VT/VF only): the first RCT in this field — stopped early for superiority; suggested ECPR improved survival and neurological outcomes compared to conventional CPR. Survival was substantially higher with ECPR than the subsequent trials would later reproduce, reflecting this being a small, single-center, highly experienced-system trial
• Prague OHCA trial (Belohlavek et al., JAMA 2022, single-center, Prague, n=256/264, any rhythm, intra-arrest transport + ECPR-based approach vs. conventional CPR): the trial's primary intention-to-treat analysis did not reach statistical significance for the primary outcome (180-day survival with favorable neurological outcome: 31.5% ECPR-based approach vs. 22.0% standard care, OR 1.63, 95% CI 0.93–2.85, p=0.09) — the trial was actually terminated by the DSMB for futility, yet the point estimate numerically favored the ECPR-based approach throughout
• INCEPTION trial (Suverein et al., NEJM 2023, multicenter, Netherlands, n=160, shockable rhythm only): no significant difference in favorable neurological outcome (20% vs. 16%) — notably, this trial had considerably longer time intervals (hospital arrival to cannulation initiation: 16 minutes; cannulation initiation to ECMO flow: 20 minutes) than ARREST and Prague OHCA, where near-immediate cannulation was achieved, and included lower-volume centers with less mature, less standardized ECPR systems and caseload

Reconciling the Discordance — As-Treated and Pooled Analyses Point Toward System/Timing Factors

• As-treated secondary analysis of Prague OHCA: among patients who did not achieve prehospital ROSC (the population actually eligible for ECPR), 180-day survival was 23.9% with ECPR vs. 1.2% with conventional ACLS alone (log-rank p<0.001); after adjustment, ECPR was associated with substantially reduced risk of 180-day death (HR 0.21, 95% CI 0.14–0.31, p<0.001) — a striking, clinically important difference when analyzed by actual treatment received rather than intention-to-treat, though this as-treated framing carries its own well-recognized bias risks (patients who survived long enough to actually receive ECPR may differ systematically from those who did not)
• Pooled individual patient data analysis of ARREST and Prague OHCA (2023): combining both trials' individual patient data, an early invasive ECPR-based strategy significantly improved 180-day favorable neurological survival compared to standard ACLS — 32.4% (ECPR) vs. 19.7% (control) across all rhythms (absolute difference 12.7%, 95% CI 2.6–22.7, p=0.015), with an even larger effect in the shockable-rhythm-only subgroup: 47.1% vs. 28.3% (absolute difference 18.8%, 95% CI 7.6–29.4, p=0.01), corresponding to a number needed to treat of 5 in the shockable-rhythm population — a genuinely favorable, clinically meaningful NNT when restricted to this well-organized-system, shockable-rhythm population
• INCEPTION's own as-treated analysis: showed only a numerical, non-statistically-significant benefit for ECPR, though the authors noted patient numbers for this specific analysis were low, limiting the strength of any conclusion from it
• Explicit expert commentary on the discordance: analyses of low volume per center, absence of standardization, and delays in ECPR delivery in INCEPTION specifically have been proposed as contributing to its more muted effect, given that ECPR is, by definition, a time-dependent intervention where system logistics substantially determine achievable outcomes

Practical Synthesis

The current evidence base, taken as a whole, supports ECPR as a potentially effective rescue strategy specifically within well-organized, high-volume, low-latency systems of care, particularly for patients with witnessed arrest, shockable initial rhythm, and short achievable low-flow time. The trials' divergent intention-to-treat results should not be interpreted as ECPR having been definitively disproven as an intervention — rather, they collectively demonstrate that ECPR's effectiveness is substantially system-dependent, and that the intervention's real-world benefit is likely concentrated in centers capable of achieving the rapid cannulation times seen in ARREST and Prague OHCA, rather than being a uniform property of the device/technique itself regardless of delivery system. Current international guidelines (ILCOR, incorporated into AHA/ERC ALS guidelines) reflect this nuance with only a weak recommendation, explicitly reserving ECPR consideration for selected patients with a suspected reversible cause of arrest, in whom conventional measures are failing, within institutions capable of rapid, standardized delivery — not as a universal rescue strategy applicable regardless of system capability.

Practical Implementation Considerations

• Institutions considering an ECPR program should recognize that outcomes are substantially determined by system logistics (prehospital recognition/transport protocols, in-hospital cannulation team readiness, achievable time-to-flow), not solely by the decision to acquire ECMO capability itself
• A prospective observational implementation study (Stockholm program) illustrates the real-world logistics challenge: of 95 patients meeting ECPR criteria, only 9% achieved ECMO initiation within the predefined 60-minute threshold, with 36% ultimately receiving ECMO and 25% overall survival to discharge — concretely demonstrating the gap between eligibility and achievable rapid cannulation even in a dedicated program, and reinforcing why low-flow-time optimization is a genuine, ongoing systems challenge rather than a solved logistics problem

12. Organ Support

ECPR is itself the organ (circulatory) support intervention this protocol addresses; post-cannulation management interacts directly with Vasopressor & Inotrope Selection & Titration and Post-Cardiac Arrest Syndrome protocols for the comprehensive post-arrest, ECMO-supported management that follows successful cannulation.

13. Disease-Specific Therapy — Not Applicable

This is a device/systems-of-care intervention protocol rather than a pharmacotherapy one; standard ACLS pharmacotherapy continues throughout the resuscitation and cannulation process per standard ALS guidelines.

14. Consultation Matrix

15. Monitoring Framework

• During resuscitation/cannulation: continuous rhythm, compression quality, and time-tracking (arrest onset to ECMO flow) given the low-flow-time dependency central to this evidence base
• Post-cannulation: standard ECMO and post-cardiac-arrest monitoring (cross-reference Post-Cardiac Arrest Syndrome protocol)
• Institutional/systems level: tracking of time-to-cannulation and time-to-flow metrics as ongoing quality indicators, given their established relationship to outcome

16. ICU Bundle Checklist

17. Complications

Early:

• Vascular access complications (bleeding, limb ischemia) — documented in real-world implementation data (e.g., 2/95 patients with severe access-site bleeding requiring intervention in the Stockholm program)
• Failure to achieve adequate flow within a clinically meaningful time window
• Ongoing hypoxic-ischemic brain injury if low-flow time is prolonged despite ECPR activation

Late:

• Neurological injury proportional to total low-flow time, regardless of eventual perfusion restoration
• Standard VA-ECMO complications (cross-reference relevant ECMO-specific content where available in this library)

Prevention: minimizing low-flow time via rapid, standardized activation and cannulation protocols; appropriate patient selection per institutional, evidence-informed criteria

Rescue: standard vascular complication management; escalation to relevant specialist teams (vascular surgery for access complications, interventional cardiology for the underlying cause)

18. Escalation & De-escalation

Escalation: refractory arrest meeting institutional criteria → rapid ECPR activation per Section 3 checklist, recognizing the intervention's time-dependency.

De-escalation: cross-reference Post-Cardiac Arrest Syndrome and Goals of Care & Palliative Care Integration protocols for the post-cannulation trajectory, including neurological prognostication and, where outcomes are poor, appropriate transition to comfort-focused care per that protocol's evidence-based communication framework.

19. ICU Discharge Criteria — Not Directly Applicable

Cross-reference Post-Cardiac Arrest Syndrome and ICU Discharge Criteria & Step-Down protocols for the relevant discharge framework once a patient survives the acute ECPR/post-arrest period.

20. Documentation & Medicolegal Checklist

• Time of arrest onset, witnessed status, and bystander CPR documented
• Institutional ECPR eligibility criteria assessment documented
• Time-to-cannulation and time-to-ECMO-flow explicitly documented, given their established prognostic significance
• Reversible cause investigation and treatment documented
• Rationale for ECPR activation or non-activation documented, particularly in borderline candidacy cases

21. Key Guidelines

• ILCOR/AHA/ERC Advanced Life Support guidelines: incorporate ECPR as a weak recommendation, for selected patients suspected to have a reversible cause of arrest in whom conventional measures are failing — reflecting the genuinely low-to-moderate certainty of the underlying evidence base and the systems-dependency detailed in Section 11

22. Landmark Trials

23. Controversies

• The core, still-live controversy: whether ECPR's benefit (clearly demonstrated in ARREST and the pooled ARREST/Prague-OHCA analysis) fails to materialize in INCEPTION because of genuine system/logistics limitations (the leading current explanation) versus a more fundamental limitation of the intervention's generalizability that experienced-center enthusiasts may be underweighting. Both explanations are actively debated in recent literature (e.g., the explicitly titled "No one steps twice into the same river" commentary on this exact tension), and this protocol presents the systems-dependency explanation as the leading, evidence-supported interpretation without treating it as fully settled.
• As-treated vs. intention-to-treat analysis interpretation: the striking as-treated Prague OHCA finding (23.9% vs. 1.2% survival) is dramatically more favorable than the trial's own primary ITT result — this discrepancy itself is instructive about the limits of as-treated analyses in this specific context (patients who survive long enough to actually receive ECPR are inherently a different population than all patients randomized to the ECPR-intent arm), and should be interpreted with appropriate caution rather than as simply "the real effect size."
• Resource allocation and system investment questions: given the clear system-dependency of outcomes, a genuine policy-level question exists regarding whether resources are better invested in establishing a small number of very-high-volume, rapid-response ECPR centers versus broader but less mature program distribution — this is an active health-systems planning question beyond the scope of purely clinical evidence, but directly informed by the clinical evidence detailed above.
• Non-shockable rhythm and unwitnessed arrest populations: evidence for these populations remains considerably less developed than for the witnessed, shockable-rhythm population that has been the focus of the major trials — institutional criteria vary, and this represents a genuine area of ongoing uncertainty rather than an extension of the shockable-rhythm evidence base.

24. References

1. Yannopoulos D, Bartos J, Raveendran G, et al. Advanced reperfusion strategies for patients with out-of-hospital cardiac arrest and refractory ventricular fibrillation (ARREST): a phase 2, single centre, open-label, randomised controlled trial. Lancet. 2020;396(10265):1807-1816.
2. Belohlavek J, Smalcova J, Rob D, et al; Prague OHCA Study Group. Effect of intra-arrest transport, extracorporeal cardiopulmonary resuscitation, and immediate invasive assessment and treatment on functional neurologic outcome in refractory out-of-hospital cardiac arrest: a randomized clinical trial. JAMA. 2022;327(8):737-747.
3. Suverein MM, Delnoij TSR, Lorusso R, et al. Early extracorporeal CPR for refractory out-of-hospital cardiac arrest. N Engl J Med. 2023;388:299-309.
4. Belohlavek J, Yannopoulos D, Smalcova J, et al. Extracorporeal cardiopulmonary resuscitation for refractory out-of-hospital cardiac arrest: a pooled individual patient data analysis. EClinicalMedicine. 2023;59:101988.
5. Rob D, et al. Extracorporeal versus conventional cardiopulmonary resuscitation for refractory out-of-hospital cardiac arrest: a secondary analysis of the Prague OHCA trial. Crit Care. 2022.
6. Rob D, et al. Long-term follow-up of refractory out-of-hospital cardiac arrest treated by hyperinvasive approach after Prague OHCA trial. Crit Care. 2024;28:165.
7. Scquizzato T, et al. Extracorporeal CPR after the INCEPTION trial: no one steps twice into the same river. Artif Organs. 2023.
8. Implementation of an extracorporeal resuscitation (ECPR) program for out-of-hospital cardiac arrest in Stockholm, Sweden: feasibility, safety, and outcome. 2023.
9. The Washington Manual of Critical Care, 4th ed. 2025 — VV/VA ECMO and cardiac arrest chapters.
10. ICU Protocols: A Step-wise Approach, 2nd ed. — VV ECMO chapter.