Cardiogenic Shock (NEJM 2026 Review)

Authors, Journal, Affiliations, Type, DOI

• Authors: Holger Thiele (Heart Center Leipzig at Leipzig University; Leipzig Heart Science), Christian Hassager (Department of Cardiology, Rigshospitalet + University of Copenhagen)
• Journal: New England Journal of Medicine, Vol. 394, January 1, 2026 (pp. 62–77)
• Affiliations: Leipzig (Germany) and Copenhagen (Denmark)
• Type: Invited review article (CME credit)
• Note: Holger Thiele is the lead PI of IABP-SHOCK II, ECLS-SHOCK, and CULPRIT-SHOCK trials
• DOI: https://doi.org/10.1056/NEJMra2312086

Overview

This NEJM 2026 review by the lead investigators of major CS RCTs synthesizes current evidence on cardiogenic shock diagnosis, staging, phenotyping, and management. A pivotal epidemiological finding: non-infarction CS now outnumbers AMI-CS in contemporary registries. Mortality ranges by etiology: mixed shock 48% > AMI-CS 41% > new HF-CS 31% > secondary CS 31% > acute-on-chronic HF-CS 25%. Only 5% of all CS patients and 32% of STEMI-CS patients meet DanGer Shock eligibility criteria, constraining applicability of the trial's positive result. A 4-trial IPD meta-analysis of VA-ECMO shows no mortality benefit with consistently higher complication rates. CULPRIT-SHOCK establishes culprit-lesion–only PCI as the revascularization standard in multivessel AMI-CS.

Keywords

Cardiogenic shock, SCAI staging, microaxial flow pump, DanGer Shock, ECLS-SHOCK, CULPRIT-SHOCK, VA-ECMO, IABP, revascularization, norepinephrine, levosimendan, frailty, phenotyping, normotensive cardiogenic shock, mixed shock

Key Takeaways

Definition and Diagnosis

• CS definition (SHARC): cardiac disorder resulting in both clinical and biochemical evidence of sustained tissue hypoperfusion
• Hemodynamic criteria: SBP <90 mmHg for >30 min OR need for inotropes/vasopressors/MCS to maintain SBP ≥90 mmHg; plus evidence of systemic hypoperfusion
• Invasive hemodynamic definition: cardiac index ≤2.2 L/min/m² + high SVR index (>2200 dynes/cm/s)
• Normotensive CS: SBP >90 mmHg despite organ hypoperfusion — distinct and clinically important subset
• Mixed shock: cardiogenic shock with ≥1 additional concurrent cause (e.g., septic or hypovolemic component)
• Initial workup: echocardiogram mandatory (delineates cause and phenotype); PAC for selected patients not responding to initial therapy or in diagnostic uncertainty
• SCAI staging useful for group mortality but imprecise for individual prognosis
• Lactate: established biochemical marker of inadequate perfusion; q1h monitoring recommended until stabilization

Epidemiology and Mortality by Phenotype

• 30-day mortality 40–50%, varying by etiology
• Non-infarction CS now outnumbers AMI-CS (epidemiological shift from improved MI prevention/revascularization)
• Mortality by etiology (n=8,974 SHARC registry):
  • Mixed-cause CS: 48%
  • AMI-CS: 41%
  • New HF-CS (de novo): 31%
  • Secondary causes CS: 31%
  • Acute-on-chronic HF-CS: 25%
• 71% of AMI-CS patients have frailty — major prognostic factor
• Some registries show rising mortality: attributed to aging population + increasing comorbidity burden
• Cause of death: persistent cardiogenic shock (dominant) > arrhythmia > anoxic brain injury > respiratory failure

CS Phenotypes

• LV failure dominant: most common form
• Biventricular failure: RV involvement in 44% of AMI-CS; even more common in non-AMI-CS
• Mechanical complications: papillary muscle rupture, acute mitral regurgitation, ventricular septal defect, free-wall rupture — rare but dismal prognosis; require surgical/percutaneous correction
• Post-cardiac arrest phenotype: cardiac stunning after ROSC; high lactate from no-flow/slow-flow does NOT correlate with cardiac index in comatose patients; 20–30% die from brain injury rather than circulatory failure — critical to distinguish
• SCAI staging limitations: requires ≥2 assessments (dynamic entity); subjective elements limit inter-study comparisons

Systems of Care

• Tertiary-care centers with MCS capability + cardiac ICU + cardiac surgery on-site
• Higher annual CS volume (≥107 cases/year) associated with lower mortality (observational data)
• Multidisciplinary shock teams (intensivists, interventionalists, perfusionists, cardiothoracic surgeons) + regionalized shock-care systems may independently improve outcomes

Intensive Care Unit Management

• Blood glucose control; lactate q1h until stabilization; adequate oxygen delivery; thromboprophylaxis; stress-ulcer prophylaxis; early enteral feeding after stabilization
• All vasoactive medications must be given intravenously
• Lung-protective ventilation: tidal volume 6–8 mL/kg predicted body weight if mechanically ventilated
• Non-invasive ventilation (CPAP): option to avoid intubation in borderline situations; caution if RV failure dominates
• Renal replacement therapy: initiate for uremia, otherwise-untreatable volume overload, metabolic acidosis, or refractory hyperkalemia — conventional indications only
  • Earlier RRT (without conventional indications) has NO effect on outcome (RCT evidence)
  • RRT usage rates across major trials: IABP-SHOCK II 18%, CULPRIT-SHOCK 14%, DanGer Shock 11%
  • Optimal timing of RRT initiation remains unclear

Hemodynamic Monitoring

• Echocardiography / POCUS: first-line for cause identification and phenotyping
• No consensus on invasive hemodynamic monitoring method
• PAC: guidelines suggest using early for patients not responding to initial therapy OR in diagnostic/therapeutic uncertainty (e.g., mixed shock)
• Multiple hemodynamic profiles defined (LV, RV, biventricular) — prognosis affected by RV/biventricular failure
• No RCT data yet demonstrating outcome benefit from PAC in CS; PACCS trial ongoing

Fluid Management

• Crystalloids only if central hypovolemia without congestion AND hemodynamics improve after leg-raise test
• IV loop diuretics for volume overload/pulmonary edema
• Fluid management must be tailored to LV-dominant vs RV-dominant failure (different physiological needs)
• Avoid hypovolemia

Pharmacological Management

Inotropes / Inodilators:

• No superiority established for any inotrope (Cochrane analysis)
• Dobutamine (primary first-line in LV failure) vs milrinone: no outcome difference (DOREMI trial)
• Levosimendan: Ca²⁺ sensitizer → positive inotropy + afterload reduction; did NOT facilitate VA-ECMO weaning in a recent RCT; ongoing trials for other CS endpoints
• First-line inotrope choice lacks clear consensus; selection varies widely in practice

Vasopressors:

• Dopamine vs norepinephrine (n=1,679, diverse shock): dopamine associated with substantially more arrhythmic events; no mortality difference
• Epinephrine vs norepinephrine: similar cardiac index but epinephrine → unfavorable effects on HR and lactic acidosis (metabolic derangement)
• Norepinephrine: preferred vasoconstrictor for low BP/insufficient tissue perfusion pressure
• Vasopressin in CS: data lacking
• MAP target: >65 mmHg considered adequate; ongoing trial testing MAP 55 vs 65 mmHg in AMI-CS

Temporary Mechanical Circulatory Support

IABP:

• Enhances coronary perfusion during diastole; reduces afterload during systole
• IABP-SHOCK II: did NOT improve cardiac index or survival vs medical therapy in AMI-CS; no benefit at 30 days or 6-year follow-up
• Altshock-2 (HF-CS; n=101): early IABP did NOT improve survival or bridge to heart replacement therapy at 60 days
• Still widely used for ease of insertion, cost, and favorable adverse-event profile
• Recommended for mechanical complications in European guidelines (but not US guidelines)
• Routine use in AMI-CS: not recommended

VA-ECMO:

• Provides up to 6 L/min flow; full respiratory and circulatory support
• ECLS-SHOCK (AMI-CS, planned early revascularization): 30-day mortality 47.8% ECMO vs 49.0% control (RR 0.98; 95% CI 0.80–1.19; P=0.81); similar at 1-year follow-up
  • Substantial cardiac-arrest phenotype enrollment may explain neutral result
• 4-trial IPD meta-analysis (VA-ECMO in AMI-CS): no mortality benefit; consistently higher complication rates across all devices
• Potential harm from VA-ECMO: retrograde aortic blood flow increases LV afterload ("Windkessel effect")
• LV unloading strategies to mitigate VA-ECMO afterload:
  • Microaxial flow pump or IABP added to VA-ECMO: observational data suggest lower mortality
  • Transseptal LA cannula vs VA-ECMO alone (recent RCT): no effect on mortality
  • Additional RCTs ongoing
• Routine VA-ECMO in CS: not recommended (US guidelines)

Microaxial Flow Pump (Impella):

• Peak flow ~4.3 L/min; treats CS with predominant LV dysfunction
• Large propensity-matched studies (>100,000 patients): consistently no survival benefit + higher complication rates
• DanGer Shock (n=360; STEMI-CS; no hypoxic brain injury risk; randomized within 24h at experienced centers):
  • Microaxial flow pump vs standard care
  • HR 0.74 (95% CI 0.55–0.99; P=0.04) at 180 days — first positive tMCS RCT
  • Survival benefit sustained up to 10 years (proportional-hazards ratios suggesting lasting effect)
  • More bleeding, limb ischemia, and renal replacement therapy in pump group
• Patient selection limitation: only 5% of all CS patients and 32% of STEMI-CS patients meet DanGer Shock eligibility criteria
• IPD meta-analysis (9 trials; n=1,059; 6-month follow-up): no overall mortality benefit; benefit only in LV-dominant, low hypoxic brain injury risk subgroup (HR 0.77; 95% CI 0.61–0.97; P=0.024); complications higher across all devices
• Two additional microaxial flow pump trials in AMI-CS started; one suspended after DanGer Shock publication
• US guidelines: Class IIa indication for selected patients
• HF-CS: MCS only if chance of myocardial recovery OR eligible for durable LVAD/heart transplant

TandemHeart (LA-to-femoral arterial device):

• Rarely used clinically; 2 small trials in AMI-CS; no conclusive clinical outcome evidence

General MCS reflections:

• 50–60% of AMI-CS patients survive without any MCS device — device complications in these patients can be lethal
• Device selection influenced by efficacy, institutional experience, complications, reimbursement, and expert opinion
• Circumstances where MCS unlikely to help: severe shock + older age + frailty; anoxic brain injury
• Further RCTs with phenotype-specific enrollment are urgently needed

Revascularization in AMI-CS

Culprit vs multivessel PCI — CULPRIT-SHOCK:

• 70–80% of AMI-CS patients have multivessel CAD
• CULPRIT-SHOCK: culprit-only PCI vs immediate multivessel PCI
  • Death + RRT composite: 45.9% vs 55.4% (RR 0.83; 95% CI 0.71–0.96; P=0.01), driven by mortality reduction
  • Most surviving culprit-only patients underwent staged revascularization after stabilization
  • Current preferred strategy: culprit-lesion–only PCI with staged revascularization after clinical stabilization
• If not amenable to PCI: CABG may be considered

SHOCK trial:

• Early revascularization: no 30-day mortality reduction; but reduced mortality up to 6 years
• Delay in revascularization → worse outcomes (multiple registries)
• Door-to-balloon time reduction emphasized

Pharmacoinvasive approach:

• Observational data: STEMI-CS with prolonged interhospital transport → fibrinolysis + subsequent PCI may benefit vs primary PCI alone; no increase in major bleeding

Future Perspectives

• Advanced phenotyping critical for trial design (heterogeneity explains neutral results)
• Ethical challenges in CS trials (acuity, informed consent)
• International shock research networks needed to power phenotype-specific trials
• MAP target optimization (55 vs 65 mmHg trial ongoing)
• LV unloading strategies for VA-ECMO need RCT evidence
• Inotrope selection trials needed (especially levosimendan in CS beyond VA-ECMO weaning)

Limitations of the document

• Review article: synthesizes existing evidence rather than generating new data
• Many recommendations based on expert opinion or observational data rather than RCTs
• DanGer Shock benefit may not be generalizable given strict eligibility (5% real-world applicability)
• Non-AMI CS (HF-CS) largely without RCT evidence; management extrapolated
• Post-cardiac arrest CS phenotype has unique pathophysiology (brain injury, sedation, vasodilation) that confounds MCS trial interpretation
• Most evidence from European/Western centres; may not reflect global practice

Key Concepts Mentioned

• concepts/Cardiogenic-Shock — updated with mortality phenotype data, frailty statistics, VA-ECMO IPD meta-analysis, 10-year DanGer Shock follow-up, eligibility limitation (5%/32%), CULPRIT-SHOCK
• concepts/SCAI-Shock-Classification — SCAI stages with specific lactate thresholds per stage; limitation noted: imprecise for individual prognosis
• concepts/Temporary-Mechanical-Circulatory-Support — updated with Altshock-2 (HF-CS IABP negative), VA-ECMO retrograde afterload harm, LV venting RCT (transseptal, negative), 4-trial VA-ECMO IPD meta-analysis (negative), 10-year DanGer Shock follow-up, real-world eligibility fractions, levosimendan VA-ECMO weaning failure

Key Entities Mentioned

• entities/DanGer-Shock-Trial — HR 0.74 at 180 days; 10-year benefit sustained; 5%/32% real-world eligibility
• entities/ECLS-SHOCK-Trial — 47.8% vs 49.0% at 30 days; similar at 1 year; 4-trial IPD meta-analysis confirms no benefit
• entities/CULPRIT-SHOCK-Trial — culprit-only PCI (45.9%) vs multivessel (55.4%) composite; current revascularization standard
• entities/Altshock-2-Trial — IABP in HF-CS; n=101; no benefit for survival or bridge to heart replacement therapy at 60 days
• entities/IABP-SHOCK-II-Trial — confirmed negative; no hemodynamic or survival benefit in AMI-CS

Wiki Pages Updated

• Created wiki/sources/cardiogenic-shock-nejm-2026.md (this file)
• Updated wiki/concepts/Cardiogenic-Shock.md
• Updated wiki/concepts/Temporary-Mechanical-Circulatory-Support.md
• Updated wiki/concepts/SCAI-Shock-Classification.md
• Updated wiki/wikiindex.md
• Updated wiki/sourceindex.md