Fifth Universal Definition of Myocardial Infarction (2026)

Authors, Journal, Affiliations, Type, DOI

• Authors: Mills NL, Newby LK, Zaman S, et al; ESC/ACC/AHA/WHF Task Force for the Universal Definition of Myocardial Infarction
• Journal: Circulation 2026;154 (co-published identically in European Heart Journal, JACC and Global Heart)
• Affiliations: Joint European Society of Cardiology / American College of Cardiology / American Heart Association / World Heart Federation Task Force; endorsed by EACTS and STS, with SCAI affirmation of value
• Type: Expert consensus document
• DOI: 10.1161/CIR.0000000000001477
• Task force: 17 members from 12 countries across 5 continents; 18 independent peer reviewers plus a patient panel
• Grading: none — the document issues no COR/LOE or Class designations. Its normative language is a four-tier diagnostic ladder applied identically in Boxes 3, 4, 5 and 7: considered → likely → confirmed → presumed

Overview

The Fifth Universal Definition abolishes the numerical type 1–5 classification that has organised MI diagnosis since 2007 and replaces it with a three-part clinical taxonomy built on pathophysiology: primary MI (spontaneous, from any acute coronary pathology), secondary MI (oxygen supply–demand imbalance caused by another acute condition), and procedure-related MI (a complication of any percutaneous or surgical cardiac procedure within 30 days). The redesign is explicitly a trade in diagnostic operating characteristics — primary criteria prioritise sensitivity to avoid missing an acute coronary pathology, while secondary and procedure-related criteria prioritise specificity so the label carries treatment implications. Two structural moves drive it: all non-atherosclerotic acute coronary pathologies (SCAD, embolism, vasospasm) and all late restenosis, stent thrombosis or graft failure beyond 30 days move out of "type 2" and into primary MI; and biomarker multiples of the URL (5× for PCI, 10× for CABG) are abandoned as diagnostic thresholds in favour of imaging-confirmed coronary complication and/or new regional wall motion abnormality or loss of viable myocardium. Alongside this the document introduces sex-specific 99th percentile URLs as the sole definition of myocardial injury, accelerated diagnostic pathways, objective criteria for unrecognised and sudden-death MI, a low-resource presumptive pathway, and ICD-11 stem and sixth-digit codes aligned to the new taxonomy.

Keywords

Acute myocardial injury · cardiac troponin · chronic myocardial injury · coronary and cardiac imaging · high-sensitivity cardiac troponin · ICD-11 · myocardial infarction · MINOCA · primary myocardial infarction · procedure-related myocardial infarction · recurrent myocardial infarction · secondary myocardial infarction · Takotsubo syndrome · unrecognized myocardial infarction · unstable angina

Key Takeaways

What Changed from the Fourth Universal Definition

• The rationale for abandoning type 2 was that grouping non-atherosclerotic coronary causes with supply–demand imbalance "has limited its adoption"; the diagnosis was often made on symptoms alone without imaging and "had no direct treatment implications for many patients."
• The rationale for abandoning types 4a/5 was that the 5× and 10× URL thresholds were arbitrary: in a large prospective international study 97.5% of cardiac-surgery patients had troponin above 10× URL, and thresholds of >218× URL (95% CI 40–318) were needed to identify peri-procedural mortality risk.

Myocardial Injury — Acute and Chronic

• Acute myocardial injury (Box 1): a rise and/or fall in cardiac troponin I or T with at least one value above the sex-specific 99th percentile URL. It may be ischaemic or non-ischaemic; ischaemic injury is the hallmark of MI and is classified as MI only when the MI criteria are also met.
• Non-ischaemic mechanisms include inflammation (myocarditis, transplant rejection, cytokine-mediated sepsis), haemodynamic stress (tachy- or bradyarrhythmia, acute HF, acute PE, malignant hypertension), physiological stress (strenuous exercise), catecholamine stress (Takotsubo, subarachnoid haemorrhage, stroke, seizures, phaeochromocytoma), toxicity (anthracyclines, ErbB-targeted therapies, TKIs, checkpoint inhibitors, radiation) and trauma (cardioversion, PCI, structural intervention, ablation, device implantation, cardiac surgery, contusion).
• Troponin elevation does not equal necrosis — Takotsubo and myocarditis can recover fully, and athletes after exertion likely have reversible release.
• Chronic myocardial injury (Box 2): considered when two or more cTnI or cTnT values are above the sex-specific 99th percentile in a stable clinical setting; confirmed when a condition associated with cardiac remodelling is identified; excluded when analytical interference or reduced clearance explains it. Chronic injury is up to five times more frequent with cTnT than cTnI.
• A change of less than 20% between troponin measurements suggests chronic rather than progressive injury, though the document concedes the criterion "is simply based on the analytical imprecision of cardiac troponin assays."
• Renal handling: troponin is cleared renally at normal concentrations and hepatically at high concentrations; halving eGFR doubles basal troponin.

Primary Myocardial Infarction

• Likely (Box 3): acute myocardial injury plus at least one of — symptoms consistent with acute myocardial ischaemia; new or presumed-new ischaemic ECG changes; development of pathological Q waves.
• Confirmed with at least one imaging feature: (a) an acute coronary pathology — atherothrombosis, SCAD, vasospasm, embolism, or restenosis, stent thrombosis or bypass graft failure more than 30 days from revascularisation; (b) new or presumed-new regional wall motion abnormality and/or absence of viable myocardium in an ischaemic pattern.
• Where angiography and imaging are unavailable, a final diagnosis may rest on biomarkers plus ECG plus clinical criteria, coded BA41.09/BA41.19 (aetiology unknown).
• Angiographic frequencies: SCAD 0.1%–4.0%; coronary embolism 0.2%–13%; epicardial vasospasm 1%–20%; in-stent restenosis 1.0%–6.5%; stent thrombosis 0.6%–0.8%.
• SCAD should be considered in women under 50 and in pregnancy or the post-partum period.

Secondary Myocardial Infarction

• Considered (Box 4): acute myocardial injury with an alternative acute condition causing supply–demand mismatch, plus at least one of symptoms, new ischaemic ECG changes or pathological Q waves. Likely with known CAD or strong clinical suspicion from the extent of ECG change or troponin elevation.
• Confirmed by either obstructive CAD — ≥70% stenosis in an epicardial vessel by angiography, or ≥50% stenosis that is flow-limiting on physiological assessment — WITHOUT an acute coronary pathology, or new/presumed-new regional wall motion abnormality or absence of viable myocardium in an ischaemic pattern.
• The commonest trigger is tachyarrhythmia; triggers frequently coexist, so no universal threshold can be set for any one of them.
• If imaging reveals an acute coronary pathology, reclassify as primary MI.
• Diagnostic yield of systematic angiography: in a prospective cohort of prior "type 2 MI" patients, 1 in 20 had atherothrombosis — that is, primary MI.

Procedure-Related Myocardial Infarction

• Window is 30 days, explicitly "an arbitrary period" chosen to align with post-procedure complication studies; it replaces the Fourth UDMI's 48-hour window for types 4a and 5.
• Considered (Box 5): suspected coronary complication within 30 days, acute myocardial injury, and symptoms or new ischaemic ECG changes. Confirmed by an angiographic coronary complication (side branch occlusion, iatrogenic dissection, no/slow flow, recoil, stent thrombosis, graft failure, coronary impingement, embolism), imaging showing new regional wall motion abnormality or loss of viability in the territory of the artery in which the complication arose, new pathological Q waves, or unexplained sudden deterioration in a sedated patient.
• Both features are required when the complication arises during the index procedure or when the procedure is being performed in the management of acute MI; either suffices after an initially uncomplicated procedure.
• Biomarker magnitudes survive only as "likely" triggers — a continuing rise above pre-procedural levels at 6 h (>5× URL for intervention) or 24 h (>35× URL for surgery). These are not diagnostic and should prompt imaging.
• Peak timing: surgery 3–12 h; a later peak at 12–24 h suggests a procedure-related complication. Events beyond 30 days are primary MI, not procedure-related.

Myocardial Infarction After Sudden Death

• Type 3 is removed. Where MI is the likely cause of death, the clinical classification is applied from the setting or post-mortem findings.
• Primary MI is more likely with age >40, prior CAD, preceding chest pain, and documented VF without known cardiomyopathy or channelopathy.
• No ECG and no autopsy → unspecified MI (BA41.Z); incomplete evaluation → MI of uncertain aetiology.
• Post-mortem CT-guided biopsy with CMR has >95% sensitivity and specificity for acute MI, but protocols are unstandardised and availability is limited.

Silent or Unrecognised Myocardial Infarction

• Q waves alone are poorly specific; confirm with imaging, preferably CMR with late gadolinium enhancement.
• Prevalence: ~1 in 5 people over 70 when CMR is applied systematically, and 3% of asymptomatic middle-aged individuals (mean age 50).
• "Silent" is often a misnomer — many patients recall symptoms retrospectively, and reduced pain sensitivity is demonstrable on cold pressor testing.

Unstable Angina

• Considered when there is new rest or minimal-exertion ischaemia (CCS class 3 or 4) or sudden intensification after MI has been excluded by serial troponin; confirmed if an acute coronary pathology is found.
• The diagnosis has become progressively rarer with hs-cTn, and many historical "UA" patients were reclassified as MI when stored samples were retested.
• Unstable angina remains part of the ACS spectrum, with its own ICD-11 code (BA40).

Myocardial Injury With Non-Obstructive Coronary Arteries (MINOCA)

• Redefined: MINOCA now means "myocardial injury with non-obstructive coronary arteries" in patients with possible MI and no stenosis ≥50% on angiography. It is a working diagnosis, not a final one.
• Most such patients ultimately have a non-coronary cardiac cause (myocarditis, Takotsubo, cardiomyopathy) or a non-cardiac cause such as pulmonary embolism; if MI is later confirmed it is classified as primary, secondary or procedure-related.
• Intravascular imaging may reveal plaque erosion or rupture, mural thrombus without rupture, or SCAD; provocative testing may reveal spasm or microvascular dysfunction.
• CMR ideally within 2 weeks; MI is the final diagnosis in only 22%–27%, and the alternative diagnosis changes management.

Symptoms

• Chest discomfort is the most common symptom with similarly high sensitivity in women and men but low specificity; severity does not predict MI.
• "Chest discomfort" is preferred over "chest pain", and "typical"/"atypical" are discouraged because "atypical" labels are applied more often to women and contribute to delayed diagnosis.
• Women with ACS have higher odds of interscapular pain, nausea and vomiting, and dyspnoea, and lower odds of chest pain and diaphoresis. Absent chest pain is more common in older people and in diabetes, and is associated with delayed diagnosis and higher in-hospital death.

The Electrocardiogram

• STEMI working diagnosis: new ST elevation at the J-point in ≥2 contiguous leads — ≥2.5 mm in men <40 y, ≥2 mm in men ≥40 y, ≥1.5 mm in women regardless of age in V2–V3, and/or ≥1 mm in all other leads — absent LVH or LBBB.
• Posterior MI: ST depression ≥1 mm in V1–V3, especially with a dominant R in V1–V2; confirm with ST elevation in V7–V9. Right ventricular MI: ST elevation in V3R–V6R, especially with aVR elevation.
• ST-segment depression: new horizontal or down-sloping ≥0.5 mm at the J-point in ≥2 contiguous leads. T-wave inversion: new or dynamic ≥1 mm in ≥2 contiguous leads.
• Pathological Q wave: duration ≥40 ms and/or depth ≥25% of the R wave in the same lead — the classical definition retained because it correlates better with transmural MI on CMR than the alternatives proposed in earlier UDMI versions.
• Sgarbossa (LBBB or ventricular pacing): concordant ST elevation ≥1 mm; ST depression ≥1 mm in V1–V3; discordant ST elevation ≥5 mm. Modified Sgarbossa replaces the absolute discordant threshold with an ST/S-wave ratio ≤0.25.
• Up to 1 in 4 patients managed as NSTEMI have acute occlusion of the culprit artery without classical ST elevation.
• In secondary MI, ECG changes are more often global; marked changes that do not resolve with treatment of the trigger should prompt reconsideration of primary MI.

Cardiac Troponin

• High-sensitivity assay definition: coefficient of variation ≤10% at the sex-specific 99th percentile URL, and detection of troponin in >50% of healthy females and males in the reference population.
• Female 99th percentile is below the uniform URL and male above; for cTnT the female URL is half the male URL in a large global healthy reference population.
• Age: 1 in 3 patients over 75 with possible MI have elevated troponin, versus 1 in 20 under 50. No assay has regulatory approval for age-adjusted thresholds.
• Analytical interference: haemolysis and biotin lower measured troponin, fibrin micro-clots raise it, and macrotroponin complexes cause false positives with a half-life of several weeks. Macrotroponin may account for as many as 1 in 2 elevated cTnI values in asymptomatic populations.
• Serial testing is unavoidable above very low thresholds: sensitivity of the 99th percentile at presentation is ~90% overall and as low as 70% within 3 h of symptom onset. Around 1 in 6 STEMI patients have troponin below the URL in the first hours because the infarct territory is not perfused.
• Serial change cannot separate primary from secondary MI — initial values differ by less than two-fold and peaks by up to five-fold, with overlapping absolute and relative deltas.
• No single delta applies to all presentations and assays. The three determinants are time from symptom onset, the interval between samples, and the baseline concentration; absolute deltas are assay-specific and non-transferable.

Accelerated Diagnostic Pathways

• Three families: single-sample rule-out in two risk tiers; 0/1 h or 0/2 h serial sampling with multiple thresholds and deltas across three tiers (rule-out, observe, rule-in); and conventional pathways combining a clinical risk score with the 99th percentile at 0 and 3 h.
• Pathway thresholds are applied uniformly, not sex-specifically, and are for risk stratification only — the final diagnosis still requires a rise or fall with at least one value above the sex-specific 99th percentile.
• Box 6 adds that early presenters may not yet show a rise, recurrent ischaemic symptoms require serial testing before MI is excluded, and analytical interference should be suspected when values are clinically incongruent or vary without pattern.
• Assay availability: in a survey of 663 laboratories across 76 countries, troponin was available in 98% of hospitals and high-sensitivity assays in 92% of those. Where no hs assay exists, use conventional troponin rather than CK-MB; where no troponin exists at all, CK-MB mass assay.

Coronary and Cardiac Imaging

• Invasive angiography is the gold standard, but ~14% of NSTEMI patients lack a clearly identifiable culprit lesion; in older non-Q-wave MI series more than a third had none.
• Intravascular imaging (IVUS/OCT) identifies plaque rupture, erosion, intraplaque cavity, mural thrombus, intramural haematoma, isolated thrombus, in-stent restenosis, stent thrombosis and intimal bumping from spasm. Invasive functional testing with acetylcholine or ergonovine identifies epicardial and microvascular spasm; FFR/RFR/iFR detects flow-limiting stenosis in suspected secondary MI.
• Echocardiography distinguishes global from regional dysfunction — the key discriminator between secondary and primary MI — and identifies mimics including pericarditis, severe AS, aortic dissection and massive PE. Not all MI produces a regional wall motion abnormality.
• CMR uses cine, T2, T2*, first-pass perfusion and LGE, and models injury in four stages: oedema only; necrosis plus oedema; plus microvascular obstruction; plus intramyocardial haemorrhage. Yield is highest within 2 weeks in MINOCA, while oedema, MVO and haemorrhage can persist up to 6 months in large primary MI before becoming LGE fibrosis.
• Coronary CT angiography is not routine for intermediate- or high-risk suspected MI, but is useful in low-risk NSTEMI, for plaque phenotyping to separate primary from secondary MI, for dissection and for graft patency.
• Nuclear imaging is limited to transmural rather than subendocardial infarcts by spatial resolution.

Special Situations

• Recurrent MI: troponin often stays elevated for weeks, so serial measurement is key; symptoms differ from the index event in ~10% of males and ~16% of females; CK-MB "adds cost without providing diagnostic clarity." Recurrent MI is classified by mechanism — there is no separate recurrent-MI code.
• Heart failure: up to 90% of acute HF patients have cTnT above the URL, through ischaemia, supply–demand mismatch, apoptosis and autophagy, infiltrative and inflammatory injury, and neurohormones; preload-induced stretch is a proposed non-ischaemic mechanism.
• Chronic kidney disease: troponin is three times more likely elevated — cTnT 70% versus 24%, cTnI 47% versus 16%. The final diagnosis is primary MI more often (24% versus 11%), yet half of CKD patients with elevated troponin have injury without infarction. The frequency of elevated troponin rises six-fold from 10% at eGFR ≥90 to 66% at eGFR <30 mL/min/1.73 m², and CKD patients are on average 20 years older. Contrast nephropathy risk is "often overstated."
• Critical illness: 1 in 4 critically ill patients with known cardiovascular disease had unrecognised MI; almost 40% have myocardial injury plus ischaemic ECG change; LV dysfunction occurs in ~25% of ICU patients, half non-cardiac and mostly reversible.
• Non-cardiac surgery: most events occur within 2 days and up to two-thirds have no ischaemic symptoms. With active surveillance, PMI occurs in ~1 in 8 high-risk patients and primary MI in 1 in 20 of those with post-operative injury. MINS carries an eight-fold higher adjusted 30-day mortality. Up to 40% of patients with CVD or risk factors have baseline elevated troponin, so a pre-operative value is essential.
• Structural cardiac intervention: MI complicates up to 5.6% of TAVI (pooled 1.1% by previous definitions), with leaflet-related ostial obstruction in 0.7% and valve embolism in 1%. New ischaemic-pattern LGE appears in as many as 18% of systematically CMR-imaged TAVI patients, typically small, multi-territorial and unrelated to CAD.
• Cardio-oncology: anthracyclines and checkpoint inhibitors raise baseline troponin and cause myocarditis; atherothrombosis remains the commonest cause of primary MI, but vasospasm and embolism are more common in cancer.

Classification, Coding and Public Health

• The Fourth UDMI with hs-cTn increased diagnoses of type 1 and type 2 MI by 11% and 22%, but ICD-10 misalignment meant this never reached morbidity and mortality data — in two European countries fewer than 20% of patients meeting Fourth-UDMI type 2 criteria received an ICD-10 MI code.
• ICD-11: BA41.0 acute STEMI; BA41.1 acute NSTEMI. Sixth-digit stems for primary MI: _.1 atherothrombosis; _.2 SCAD; _.3 coronary embolism; _.4 coronary vasospasm; _.5 restenosis >30 days; _.6 stent thrombosis >30 days; .7 graft failure >30 days; .8 undetermined aetiology after angiography; .9 unknown aetiology without coronary imaging. **.A** secondary MI (with the alternative condition's code); **.B** procedure-related percutaneous within 30 days; **.C** procedure-related open cardiac surgery within 30 days. BA40 unstable angina; BA41.Z unspecified MI; BA50 unrecognised MI.
• The Sixth-digit structure is what allows routine data to study SCAD, vasospasm, embolism, restenosis, stent thrombosis and graft failure incidence for the first time.

Low-Resource Settings

• 2021: 31.9 million new cases of ischaemic heart disease globally, 31.8% more than 2010.
• Access varies starkly — in Uganda 43% of facilities had troponin and 55% had ECG; in China the proportion of hospitals able to measure troponin rose from 25% to 84% between 2001 and 2011.
• Box 7 — presumed MI: symptoms consistent with acute myocardial ischaemia plus at least one of new ischaemic ECG changes or development of pathological Q waves.
• Proposed solutions: affordable point-of-care troponin, telemedicine ECG interpretation, and alignment with the UN 2023 universal health coverage commitments for non-communicable diseases by 2030.

Research Agenda

• Box 8 priorities include implementation and incidence studies of the new framework, the impact of secondary-MI criteria on diagnosis frequency, outcomes by coronary aetiology in primary MI, prognosis after procedure-related MI, AI for ECG and biomarker interpretation, sex-specific thresholds, URLs incorporating age, sex and renal function, and cost-effectiveness of point-of-care hs-cTn.
• The document notes that trials have enrolled presumed primary MI without angiographic confirmation, diluting treatment effect — the reason objective criteria are proposed.

Limitations of the document

• No grading system. As a consensus document it issues no Class, COR or LOE designations, so the strength of evidence behind each criterion is not individually rated.
• The 30-day procedure window is conceded to be arbitrary, chosen for alignment with prior studies.
• The 5× and 35× URL figures survive as "likely" triggers despite the document's own criticism of URL multiples; their sensitivity and negative predictive value for deciding who does not need imaging is stated to be unknown.
• No age-adjusted 99th percentile URLs are approved by any regulator; the document defers this to research.
• Chronic myocardial injury lacks a validated quantitative definition beyond two values above the URL in a stable setting, and the 20% change heuristic is acknowledged to reflect analytical imprecision rather than biology.
• Secondary MI criteria rest largely on single-cohort evidence — one study on reduced diagnosis frequency and risk reclassification — and no trial has tested outcomes under the new criteria.
• Procedure-related MI harmonisation across percutaneous and surgical procedures is explicitly unvalidated.
• Routine pre- and post-procedural troponin remains contested, with EACTS and some guidelines opposed and EAPCI supportive, so the criteria may be applied to populations in which troponin is never measured.
• Imaging is central to confirmation yet often unavailable, CMR is limited early after surgery, echocardiographic windows are poor post-operatively, and not all MI produces a regional wall motion abnormality — leaving a large clinical-diagnosis-only residual category.
• ICD-11 codes were still under WHO review at publication, so the coding architecture is provisional.
• Several quantitative claims are drawn from single studies without meta-analytic synthesis, including the post-TAVI 18% LGE figure and the post-operative 1-in-8 PMI figure.
• No data exist on secondary-MI patients in cardiac rehabilitation, the frequency of macrotroponin in possible MI is unknown, and the reliability of ECG interpretation in critical illness is unquantified.

Key Concepts Mentioned

• entities/Acute-Coronary-Syndrome — primary, secondary and procedure-related definitions
• concepts/Acute-Myocardial-Injury — Box 1; rise and fall above the sex-specific 99th percentile
• concepts/Chronic-Myocardial-Injury — Box 2; two values in a stable setting; 20% heuristic
• concepts/High-Sensitivity-Cardiac-Troponin — assay definition, 99th percentile, interference, kinetics
• concepts/MINOCA — redefined as myocardial injury with non-obstructive coronary arteries
• concepts/Accelerated-Diagnostic-Pathways — single-sample, 0/1 h and 0/2 h families
• concepts/Perioperative-Myocardial-Injury — PMI versus MINS
• concepts/Unstable-Angina — retained with ICD-11 BA40
• concepts/Unrecognized-Myocardial-Infarction — Q waves plus CMR/LGE
• concepts/Coronary-Vasospasm — reclassified from type 2 to primary MI
• entities/Acute-Coronary-Syndrome — the retained STEMI/NSTEMI axis
• concepts/Myocardial-Viability — absence of viable myocardium as a confirmatory criterion
• concepts/Recurrent-Myocardial-Infarction
• concepts/Myocardial-Injury-in-Critical-Illness

Key Entities Mentioned

• entities/Acute-Coronary-Syndrome — the umbrella retained for unstable angina plus primary MI
• entities/Takotsubo — non-ischaemic injury with troponin rise, regional wall motion abnormality and no LGE; a leading MINOCA mimic
• entities/Chronic-Kidney-Disease — troponin three times more often elevated; six-fold eGFR gradient
• entities/Heart-Failure — up to 90% elevated cTnT in acute HF
• entities/Pericarditis — the principal ECG confounder after cardiac surgery
• entities/Atrial-Fibrillation — tachyarrhythmia as the commonest secondary-MI trigger
• Spontaneous coronary artery dissection — 0.1%–4.0% of primary MI; women under 50 and pregnancy
• Coronary embolism — 0.2%–13% of primary MI
• concepts/Aortic-Stenosis — TAVI-related procedure-related MI

Wiki Pages Updated

• wiki/sources/5thMI-AHA-2026.md — created
• wiki/concepts/Myocardial-Infarction.md — created as the owner page for the three-setting classification
• wiki/entities/Acute-Coronary-Syndrome.md — type 1–5 framework replaced with the three-setting classification
• wiki/concepts/MINOCA.md — redefined as myocardial injury with non-obstructive coronary arteries
• wiki/concepts/OMI-NOMI-Paradigm.md — retained STEMI/NSTEMI axis added as a contradiction
• wiki/concepts/Coronary-Vasospasm.md — reclassified from type 2 to primary MI
• wiki/concepts/Perioperative-Cardiovascular-Assessment.md — PMI and MINS definitions aligned to the Fifth UDMI
• wiki/concepts/Chronic-Kidney-Disease.md — troponin figures added
• wiki/sourceindex.md — new row
• wiki/wikiindex.md — new row