Chapter 29 — Sepsis-Induced Cardiomyopathy

📚 Guideline basis

Surviving Sepsis Campaign 2021; ESICM 2025 recommendations on haemodynamic monitoring in shock; Sanfilippo F, et al. systematic reviews of LV systolic and diastolic dysfunction in sepsis; Parker MM, et al. original description (Ann Intern Med 1984); Guarracino F, et al. on ventriculo-arterial decoupling in septic shock (Crit Care 2013, PMID 23597118).

> ⚠️ Evidence quality

> There is no consensus definition of sepsis-induced cardiomyopathy. Prevalence estimates range from roughly 10% to 70% depending entirely on which definition is applied, and the prognostic significance of the commonest definition (reduced ejection fraction) is contested, with cohorts reporting better, worse, and neutral outcomes. This chapter presents the disagreement rather than resolving it.

Pathophysiology & Mechanisms

Septic myocardial dysfunction is not ischaemic. Coronary blood flow in sepsis is normal or increased, and myocardial necrosis is not the dominant pathology. The mechanisms are metabolic and signalling-level, and they are reversible — which is the single most important clinical fact about the syndrome.

Circulating myocardial depressant mediators. TNF-α and IL-1β depress contractility in isolated myocyte preparations; serum from septic patients depresses contractility in normal myocardium, and the effect is abolished by cytokine neutralisation. Nitric oxide, generated in excess by inducible NOS, contributes through cGMP-mediated reductions in myofilament calcium sensitivity.
Mitochondrial dysfunction and bioenergetic failure. Reduced oxidative phosphorylation with preserved oxygen delivery — "cytopathic hypoxia" — accompanied by mitochondrial structural injury. The myocyte hibernates rather than dies, which is why recovery is the rule in survivors.
Altered calcium handling. Reduced L-type calcium channel current, impaired sarcoplasmic reticulum calcium release and reuptake through ryanodine receptor and SERCA2a dysfunction, and reduced myofilament calcium sensitivity.
β-adrenergic receptor downregulation and uncoupling. Sustained catecholamine exposure, endogenous and exogenous, desensitises the receptor–G-protein–adenylyl cyclase pathway — the mechanism of progressive inotrope resistance.
Microvascular dysfunction and oedema. Endothelial injury, glycocalyx shedding, capillary heterogeneity and myocardial oedema increase diastolic stiffness independently of systolic changes.

The recognition problem

Two properties make sepsis-induced cardiomyopathy systematically under-recognised.

Vasoplegia masks it. Effective arterial elastance falls sharply in distributive shock. Because ejection fraction is arithmetically a coupling ratio (Chapter 10):

Ea / Ees = (1 / EF) - 1

a fall in Ea raises EF at unchanged contractility. A patient with genuinely depressed Ees can present with an EF of 65% and a hyperdynamic-looking ventricle. The dysfunction appears only when noradrenaline restores Ea — which is why "the vasopressor caused the cardiac dysfunction" is a recurring and incorrect bedside inference.

Longitudinal function fails first. Subendocardial longitudinal fibres are the most vulnerable to inflammatory injury and elevated wall stress, while mid-wall circumferential fibres — which dominate ejection fraction — are relatively preserved. GLS, MAPSE and mitral annular s′ are therefore abnormal in a substantial fraction of septic patients with entirely normal ejection fractions.

Clinical Phenotypes

Phenotype	Definition used in the literature	Reported prevalence	Prognostic association
LV systolic dysfunction	EF < 50% (thresholds vary from 45% to 55%)	~20–60% depending on threshold and timing	Contested. Early cohorts (Parker 1984) associated reduced EF with survival, interpreted as adaptive ventricular dilatation; later cohorts report neutral or adverse associations. Meta-analyses have not resolved it
LV diastolic dysfunction	Elevated E/e′, reduced e′	~20–50%	The most consistently adverse association with mortality across systematic reviews
RV dysfunction	Reduced TAPSE, FAC, or RV strain	~30–50%	Associated with worse outcome; frequently overlooked
Subclinical / strain-defined	Abnormal GLS with normal EF	High — abnormal GLS is common with preserved EF	Identifies myocardial involvement reliably; independent prognostic value contested

Two features are characteristic and diagnostically useful:

Non-dilated cavity. Unlike chronic dilated cardiomyopathy, the acutely depressed septic ventricle is usually normal in size. Marked dilatation suggests pre-existing disease.
Reversibility. Function typically recovers over 7–10 days in survivors. A low EF in septic shock is not a chronic diagnosis and should not be labelled as one until repeat imaging after recovery.

Diagnostic Synthesis

There is no diagnostic test. The practical approach is to characterise the physiology rather than to assign a label.

Step	Measurement	Purpose
1	LVOT VTI and stroke volume index	Establishes whether output is actually low. A low EF with SVI 45 mL/m² is not a perfusion problem
2	Ea = 0.9 × SBP / SV	Quantifies the arterial limb; a low Ea identifies vasoplegia as the dominant lesion
3	EF, plus a longitudinal index (MAPSE, s′, GLS)	Detects involvement masked by vasoplegia
4	Diastolic assessment — e′, E/e′, TR velocity	The phenotype most consistently linked to mortality; note the ASE 2025 algorithm exclusions (Chapter 8)
5	RV assessment — RV:LV, TAPSE, TAPSE/PASP	Frequently abnormal and frequently missed
6	Repeat after vasopressor titration	Restoring Ea unmasks reduced Ees; this is diagnostic, not iatrogenic
7	Repeat at 7–10 days	Distinguishes reversible septic dysfunction from unmasked chronic disease

Differential

Alternative	Distinguishing features
Pre-existing dilated cardiomyopathy	Dilated cavity, chronic remodelling, prior imaging, no recovery
Acute coronary syndrome	Regional wall motion abnormality in a coronary distribution, ECG and troponin trajectory
Stress (Takotsubo) cardiomyopathy	Apical ballooning or a variant pattern crossing coronary territories; may complicate sepsis (Chapter 38)
Myocarditis	Often younger patient, marked troponin rise, may be fulminant
Dynamic LVOT obstruction	Hyperdynamic ventricle, dagger-shaped LVOT envelope, worsens with inotropes (Chapter 28)
Acute cor pulmonale	RV dilatation with septal shift; the LV is underfilled, not failing (Chapter 32)

Troponin elevation is near-universal in septic shock and does not establish an ischaemic aetiology.

Therapeutic Logic

The central principle: treat the measured haemodynamic deficit, not the ejection fraction.

Situation	Action
Low Ea, normal/high SVI, hypotension	Vasopressor. This is arterial failure; inotropes worsen it
Low SVI with hypoperfusion after adequate MAP and preload	Inotrope — dobutamine is first-line in most protocols; titrate to flow and perfusion markers, not to EF
Low EF with normal SVI and clearing lactate	No inotrope. Observe and repeat
Progressive inotrope requirement with falling response	Consider β-receptor downregulation; consider mechanical support rather than dose escalation
Diastolic dysfunction with congestion	Restrict fluid; rate and rhythm control; decongest (Chapters 12, 13)
RV dysfunction	Vasopressor for coronary perfusion, reduce PVR, avoid volume (Chapter 9)
Refractory	VA-ECMO — septic cardiomyopathy is one of the few settings where outcomes with mechanical support are favourable, precisely because the lesion is reversible

Agents with a contested evidence base

Agent	Rationale	Evidence
Dobutamine	β₁ agonism increases contractility	Standard of care by consensus; no mortality benefit demonstrated; increases myocardial oxygen demand and arrhythmia
Levosimendan	Calcium sensitiser, avoids increasing calcium load; theoretically ideal for the calcium-desensitisation mechanism	LeoPARDS was negative — no reduction in organ dysfunction or mortality, with more supraventricular arrhythmia. Not recommended for routine use
Esmolol	Reduces catecholamine toxicity, improves diastolic filling by slowing rate	A single-centre randomised trial (Morelli, 2013) reported striking mortality benefit; not replicated at scale; remains investigational
VA-ECMO	Bridges reversible dysfunction	Observational series report favourable survival in refractory septic cardiomyopathy relative to other cardiogenic shock aetiologies

🛑 Critical pitfall: Starting an inotrope because the ejection fraction is low. If stroke volume index and lactate are acceptable, the number does not require treatment. Inotropes in septic shock increase arrhythmia and myocardial oxygen demand for no demonstrated benefit.

🛑 Critical pitfall: Stopping or reducing noradrenaline because the ejection fraction fell after it was started. Restoring arterial elastance unmasked pre-existing myocardial depression. Returning the patient to hypotension worsens coronary and systemic perfusion.

🛑 Critical pitfall: Labelling a patient with "dilated cardiomyopathy, EF 25%" on the basis of a study performed in septic shock. Most recover within 7–10 days. The diagnosis requires repeat imaging after resolution.

💡 Clinical pearl: Diastolic dysfunction is the phenotype most consistently linked to mortality in the systematic review literature, and it is the one least often reported. Measure e′ and E/e′ in every septic shock study.
💡 Clinical pearl: A normal EF with MAPSE 6 mm and s′ 4 cm/s is an abnormal ventricle. Report the longitudinal indices explicitly.
💡 Clinical pearl: Because septic cardiomyopathy is reversible, refractory cases are among the better candidates for temporary mechanical support. Escalate earlier than instinct suggests in the young patient with a clearly reversible septic insult.

References

Parker MM, Shelhamer JH, Bacharach SL, et al. Profound but reversible myocardial depression in patients with septic shock. Ann Intern Med 1984;100:483–90.
Evans L, Rhodes A, Alhazzani W, et al. Surviving Sepsis Campaign 2021. Crit Care Med 2021;49:e1063–143.
Sanfilippo F, Corredor C, Fletcher N, et al. Diastolic dysfunction and mortality in septic patients: a systematic review and meta-analysis. Intensive Care Med 2015;41:1004–13.
Guarracino F, Ferro B, Morelli A, et al. Ventriculoarterial decoupling in human septic shock. Crit Care 2013;17:R213. PMID 23597118.
Gordon AC, Perkins GD, Singer M, et al. Levosimendan for the prevention of acute organ dysfunction in sepsis (LeoPARDS). N Engl J Med 2016;375:1638–48.
Morelli A, Ertmer C, Westphal M, et al. Effect of heart rate control with esmolol on hemodynamic and clinical outcomes in patients with septic shock. JAMA 2013;310:1683–91.