Heart Failure · Guidelines

HFmrEF Is Gone, and the MRA Question Just Got Harder

The 2026 ESC heart failure guideline collapses three ejection fraction phenotypes into two, retires the term GDMT, renames acute heart failure, and issues a Class I recommendation for mineralocorticoid receptor antagonists across the LVEF spectrum that the steroidal trial data do not carry on their own.

Source 2026 ESC Guidelines for the management of heart failurePublished 28 Aug 2026Venue European Heart Journal

The 2026 ESC Guidelines for the management of heart failure were released at ESC Congress 2026 and published simultaneously in the European Heart Journal, replacing the 2021 document and its 2023 focused update.

Four changes matter more than the rest: the ejection fraction categories, the staging framework, the vocabulary used for therapy, and a recommendation on mineralocorticoid receptor blockade that deserves closer reading than its class alone suggests.

Two phenotypes, not three

The category of heart failure with mildly reduced ejection fraction has been eliminated.

HFrEF is now defined by an LVEF below 50% with symptoms or signs of heart failure, and HFpEF by an LVEF of 50% or above with symptoms or signs plus objective evidence of structural or functional cardiac abnormality.

The stated rationale is that patients in the former 41–49% band share pathophysiology with lower ejection fractions and respond to the same therapies, which the accumulated subgroup data support reasonably well.

The practical effect is that a large group of patients who previously attracted a shrug and a Class IIb suggestion now fall inside a phenotype with four Class I drug classes attached.

One caveat deserves stating plainly: the relabelling does not move device thresholds, which remain anchored to their own specific ejection fraction cut-points rather than to the phenotype name.

Where the patients moved. The 41–49% band does not disappear so much as change what it entitles the patient to: the same echo report that produced a Class IIb conversation in 2021 now places the patient inside a phenotype with four foundational drug classes.

Acute becomes decompensated

The guideline replaces acute heart failure with decompensated heart failure, on the reasoning that most presentations reflect gradual failure of compensation rather than sudden onset.

Alongside this, a four-stage framework running from A to D is adopted, spanning risk factors without structural disease, structural disease without symptoms, symptomatic disease, and advanced disease.

The staging language aligns European practice with the American approach and shifts attention toward the two stages where nothing has happened yet.

GDMT is retired for a three-tier scheme

The single term guideline-directed medical therapy has been split into three named tiers.

Foundational medical therapy covers agents with Class I recommendations and convincing morbidity or mortality benefit in broad populations.

Additional medical therapy covers everything recommended at Class IIa or IIb, plus Class I recommendations that apply only to specific subsets or to symptoms.

Guideline-directed interventional therapy covers devices and procedures.

Foundational therapy for HFrEF comprises beta-blockers, an ACE inhibitor, ARNI, or ARB, a mineralocorticoid receptor antagonist, and an SGLT2 inhibitor.

Foundational therapy for HFpEF comprises a mineralocorticoid receptor antagonist and an SGLT2 inhibitor, which is the first time the preserved phenotype has been given a defined foundation at all.

Uptitration of foundational therapy at least every one to two weeks toward target doses carries Class I, at Level C, which is an admission that the interval is consensus rather than trial-derived.

Two foundations of different depth. The preserved phenotype gains a defined foundation for the first time, but it is half the size of the reduced-ejection-fraction one, and its second pillar rests on a narrower evidence base than the label implies.

The MRA recommendation and the trials beneath it

Mineralocorticoid receptor antagonists now carry Class I in chronic heart failure irrespective of ejection fraction, which reads as the single boldest statement in the document.

The steroidal agents do not support that statement on their own.

TOPCAT was neutral overall for spironolactone in preserved ejection fraction, rescued only by a regional subgroup analysis that has been debated ever since.

SPIRIT-HF did not demonstrate benefit for spironolactone in the preserved and mildly reduced range, with enrolment difficulties limiting its interpretation.

The positive trial is FINEARTS-HF, which randomised 6,001 patients with heart failure and an LVEF of 40% or above to the nonsteroidal agent finerenone or placebo.

Over a median 32 months the composite of worsening heart failure events and cardiovascular death fell from 1,283 to 1,083 events, a rate ratio of 0.84 with a confidence interval of 0.74 to 0.95.

The benefit was carried by worsening heart failure events rather than by cardiovascular death, which differed by 8.1% against 8.7% and did not reach significance, and all-cause mortality was unchanged.

Table 1 · What the MRA class recommendation actually rests on

Trial	Agent	Population	Result
TOPCAT	Spironolactone	LVEF ≥ 45%	Neutral overall; regional subgroup contested
SPIRIT-HF	Spironolactone	Preserved and mildly reduced	No demonstrated benefit; enrolment limited
FINEARTS-HF	Finerenone	LVEF ≥ 40%, n = 6,001	Composite RR 0.84 (0.74–0.95); driven by worsening HF events

A class-wide Class I recommendation therefore sits on top of one positive trial of one nonsteroidal molecule, while the agents most clinicians will actually reach for are the two that failed or came close to failing.

Cost sharpens the problem, since generic spironolactone runs around ten dollars a month and finerenone runs roughly seven hundred, which means the recommendation and the prescription will diverge in a large fraction of practices.

Hyperkalaemia is manageable but not trivial, with potassium above 5.5 mmol/L in the finerenone arm at roughly twice the placebo rate, against a reciprocal reduction in hypokalaemia.

The defensible reading is that the Class I recommendation is a statement about mineralocorticoid receptor blockade as a mechanism, not an assertion that the three available agents are interchangeable across the ejection fraction range.

SGLT2 inhibition becomes universal

SGLT2 inhibitors now carry Class I in preserved ejection fraction as well as reduced, extending a recommendation that was previously confined to one end of the spectrum.

This is the least controversial change in the document, and it converts the drug class into the one agent appropriate for essentially every heart failure patient without a contraindication.

The obesity phenotype gets a drug

Semaglutide and tirzepatide receive Class IIa in preserved ejection fraction with obesity, formalising a phenotype that had no targeted therapy at all two years ago.

The SUMMIT trial randomised 731 patients with a mean BMI of 38 kg/m², reporting cardiovascular death or worsening heart failure in 9.9% on tirzepatide against 15.3% on placebo, a hazard ratio of 0.62.

Symptom benefit was substantial at a 6.9-point improvement in the Kansas City Cardiomyopathy Questionnaire summary score at 52 weeks.

Separately, GLP-1 receptor agonists are recommended at Class IIa in type 2 diabetes with additional cardiovascular risk factors to reduce incident heart failure or cardiovascular death.

Devices, advanced disease, and the smaller upgrades

Early referral to a specialist centre for transplantation or mechanical circulatory support assessment now carries Class I, which converts a judgement call into an obligation with a date attached.

Durable mechanical circulatory support and transcatheter mitral valve repair both move up, reflecting device iteration and trial accumulation rather than any single result.

Cardiac resynchronisation therapy should be considered in preference to right ventricular pacing in reduced ejection fraction with high-degree AV block, at Class IIa.

Digoxin rises to Class IIa for symptomatic patients with LVEF at or below 40% despite optimal foundational therapy, a modest rehabilitation of an old drug.

Vericiguat holds at Class IIb for symptomatic HFrEF below 45% despite optimal foundational therapy.

Table 2 · Named agents, manufacturers, and cost reality

Agent	Brand · Manufacturer	Position in 2026	Approximate US cash cost
Finerenone	Kerendia · Bayer (OTC: BAYRY)	Nonsteroidal MRA carrying the preserved-EF evidence	Roughly $690–800 monthly — GoodRx
Spironolactone	Generic, multiple	Steroidal MRA; the affordable option with the weaker preserved-EF data	Around $10 monthly
Dapagliflozin	Farxiga · AstraZeneca (NASDAQ: AZN)	Foundational across both phenotypes	Commonly $250–450 monthly — GoodRx
Empagliflozin	Jardiance · Boehringer Ingelheim and Eli Lilly (NYSE: LLY)	Foundational across both phenotypes	Around $250 with a discount card — GoodRx
Sacubitril/valsartan	Entresto · Novartis (NYSE: NVS)	Foundational RAAS option in HFrEF	Wide spread by pharmacy; generic now listed — GoodRx
Tirzepatide	Zepbound · Eli Lilly (NYSE: LLY)	Class IIa in preserved EF with obesity	Self-pay around $299 monthly — GoodRx
Semaglutide	Wegovy · Novo Nordisk (NYSE: NVO)	Class IIa in preserved EF with obesity	Self-pay tiers via NovoCare

Pricing moves frequently and every figure above should be treated as approximate rather than quoted to a patient.

Case scenario

A 68-year-old man is referred with exertional dyspnoea, a BMI of 36 kg/m², type 2 diabetes, and an echocardiogram reporting an LVEF of 44% with grade 2 diastolic dysfunction and an indexed left atrial volume of 40 mL/m².

Under the 2021 framework he is labelled HFmrEF, receives a diuretic, and attracts a Class IIb conversation about whether the neurohormonal agents are worth starting.

Under the 2026 framework he is HFrEF, and all four foundational classes are indicated from the first visit rather than considered.

The plan is an SGLT2 inhibitor and a low-dose beta-blocker at visit one, sacubitril/valsartan and a mineralocorticoid receptor antagonist added over the following month, with review every one to two weeks and potassium and creatinine checked at each step.

His obesity is treated as a target rather than a comorbidity, and because his ejection fraction sits below 50% the incretin recommendation applies to his diabetes risk profile rather than to a preserved-EF indication, which is a distinction worth documenting so the next clinician understands the reasoning.

What changes on Monday

Rewrite the echo report interpretation macro, because an LVEF of 44% now names a phenotype with four indicated drug classes rather than a diagnostic grey zone.

Replace the phrase GDMT in local order sets and letters with the foundational, additional, and interventional tiers, since the whole point of the new vocabulary is that it tells the reader whether a drug is a cornerstone or an add-on.

Decide institutional policy on which mineralocorticoid receptor antagonist the Class I recommendation means at each ejection fraction, because the answer is not the same drug across the range and the cost difference is seventy-fold.

Book the uptitration visits at the time of discharge rather than leaving them to a routine follow-up interval that will not meet a one-to-two-week cadence.

Treat a Class I referral to an advanced heart failure centre as a referral, not as a topic for the next clinic letter.

References

1. 2026 ESC Guidelines for the management of heart failure. European Heart Journal, August 2026. Oxford Academic
2. ESC clinical practice guidelines hub — heart failure, with slide set and patient versions. escardio.org
3. Major changes made to the ESC guidelines on heart failure. ESC press release, August 2026. escardio.org
4. FINEARTS-HF: finerenone in heart failure with LVEF ≥ 40%. ESC press release. escardio.org
5. SUMMIT: tirzepatide in HFpEF with obesity. TCTMD. tctmd.com
6. Beyond steroidal MRAs: rethinking mineralocorticoid receptor blockade in heart failure. American College of Cardiology. acc.org

Further viewing

• 2026 ESC Guidelines for the management of heart failure
• Optimisation of heart failure therapy — what's new from ESC
• 2026 ACC expert consensus decision pathway on HFpEF

Educational content for clinicians; not a substitute for the full guideline text or for individual clinical judgement. Recommendation classes and evidence levels are summarised from the published document and contemporaneous reporting, and the primary text should be consulted before implementation. Pricing is approximate, varies by pharmacy and coverage, and changes frequently. The case scenario is fictional and contains no patient identifiers.

Preventive Cardiology · Guidelines

Cardiac Rehab Finally Gets a Prescription, Not a Referral Slip

The first dedicated ESC guideline replaces a decade of “refer to rehab” with dated start windows, a minimum session count, and an exercise intensity anchored to the ventilatory threshold rather than a percentage of peak.

Source 2026 ESC Guidelines on Cardiac RehabilitationPublished 28 Aug 2026Venue EHJ / EJPCDOI 10.1093/eurheartj/ehag099

The 2026 ESC Guidelines on Cardiac Rehabilitation are the first standalone European recommendations on a therapy that has carried a Class I indication inside other documents for more than fifteen years without ever receiving its own.

The document was released at ESC Congress 2026 and published simultaneously in the European Heart Journal and the European Journal of Preventive Cardiology, with endorsement from the European Society of Physical and Rehabilitation Medicine and WONCA.

It runs to 131 recommendations across 30 recommendation tables and roughly 675 references, of which 58 are Class I, 50 Class IIa, 22 Class IIb, and exactly one Class III.

Only six recommendations reach Level A evidence, a distribution that says more about the funding of non-pharmacological trials than about the therapy itself.

14 dTarget start after acute coronary syndrome

24Minimum ambulatory sessions, with no observed ceiling

1 : 7European programme slots per eligible coronary patient

8.4×Enrolment gain from automatic referral plus personal contact

The referral window is now a number

Previous documents recommended rehabilitation “early” after an event, a word that in practice meant whenever a slot opened.

The guideline replaces it with explicit windows: within 14 days after acute coronary syndrome and no later than 30 days, and within 28 days after coronary artery bypass grafting and no later than 42 days.

Patients hospitalised with decompensated heart failure should start once decongestion and haemodynamic stability are achieved, rather than waiting for an arbitrary post-discharge interval.

A dose is specified as well, with a minimum of 24 ambulatory sessions and a dose–response relationship in which each additional session is associated with fewer major adverse cardiovascular events and no identified plateau.

Table 1 · Initiation windows and programme dose

Index event	Target start	Outer limit	Class
Acute coronary syndrome	≤ 14 days	30 days	I
CABG / cardiac surgery	≤ 28 days	42 days	I
Decompensated heart failure	After decongestion	—	I
Programme dose	≥ 24 sessions	3–4 months median	I

Intensity moves off percent-of-peak

The most consequential technical change is the abandonment of percentage-of-peak-effort targets as the default method of prescribing aerobic intensity.

Percent-of-peak methods assume that a given fraction of VO₂peak or peak heart rate corresponds to the same metabolic strain in every patient, an assumption that fails badly in deconditioned, beta-blocked, and heart failure populations where the same 70% can land anywhere from light to near-maximal.

The guideline instead anchors intensity domains to the first and second ventilatory thresholds, with moderate-domain training below VT1, heavy-domain training between VT1 and VT2, and severe-domain work above VT2.

Because cardiopulmonary exercise testing is not universally available, two prediction equations are provided that estimate threshold heart rates from resting and peak heart rate, removing gas exchange analysis as a hard prerequisite for individualised prescription.

Why the anchor matters. Two patients with identical peak heart rates receive the identical percent-of-peak prescription (top), yet their thresholds sit in different places: the same target that lands mid-heavy domain for Patient A sits above VT2 for Patient B, who trains in the severe domain every session and tolerates the programme poorly.

Delivery mode is graded, not assumed equivalent

The guideline separates three delivery formats and refuses to call them interchangeable.

Supervised centre-based rehabilitation retains the strongest recommendation, and remains the only format with demonstrated reductions in hospitalisation and mortality in heart failure.

Cardiac telerehabilitation and hybrid models are recommended at Class IIa on the basis of preserved gains in functional capacity and quality of life, positioned as the answer to capacity and geography rather than as a superior product.

Unsupervised home programmes without digital support fall to Class IIb, reflecting thin comparative data rather than demonstrated harm.

Safety data are reassuring across supervised formats, with serious adverse events reported at approximately one per 53,770 patient-hours in supervised telerehabilitation and one per 11,333 patient-hours for centre-based high-intensity interval training.

Table 2 · Delivery models

Model	Class	Evidence position
Centre-based, supervised	I	Only format with mortality and hospitalisation benefit in heart failure
Telerehabilitation with remote supervision	IIa	Preserves functional capacity and quality-of-life gains; expands access
Hybrid (centre → home transition)	IIa	Practical bridge for working patients and long travel distances
Home-based without digital support	IIb	Limited comparative evidence

The indication list widens well past the coronary patient

Rehabilitation is recommended at Class I after acute and chronic coronary syndromes, after valve surgery, in adult congenital heart disease, and in recipients of cardiac implantable electronic devices.

Heart failure with preserved ejection fraction receives a Class I recommendation for the first time in a dedicated rehabilitation document, closing a gap that had left the largest and fastest-growing heart failure phenotype without formal exercise guidance.

Atrial fibrillation following catheter ablation and cancer therapy–related cardiovascular toxicity are both recommended at Class IIa, the latter formalising cardio-oncology rehabilitation as a distinct programme rather than an improvised referral.

The single Class III recommendation in the entire document is a prohibition on exclusion: patients are not to be denied rehabilitation on the grounds of frailty or multimorbidity, both of which carry their own Class I recommendations for tailored programmes.

Core components beyond the treadmill

Structured exercise is one of ten core components, and the guideline treats the other nine as recommendations rather than aspirations.

Optimisation of guideline-directed pharmacotherapy during the supervised period is Class I, positioning rehabilitation as the titration window that outpatient clinic follow-up rarely delivers.

Assessment of patient-reported outcome measures covering physical and mental health carries Class I, Level A, one of only six Level A recommendations in the document.

Smoking cessation programmes of at least one month with adjunctive pharmacotherapy are Class I, Level A, and GLP-1 receptor agonists enter as a Class IIa nutritional and weight-management adjunct.

Psychological screening with access to cognitive behavioural therapy is Class I, addressing the cardiac anxiety and avoidance behaviour that drive much of the functional disability after an event.

Two components are genuinely new to a cardiology guideline: routine screening for sexual dysfunction with explicit counselling, and structured counselling on environmental cardiovascular risk including air pollution, traffic noise, artificial light at night, and temperature extremes.

Table 3 · Pharmacological adjuncts named within core components

Agent	Brand · Manufacturer	Role in programme	Cost note
Varenicline	Chantix · Pfizer (NYSE: PFE); generics widely available	First-line cessation pharmacotherapy alongside a ≥1-month programme	Generic cash pricing roughly $25–40 per month with a discount card; see GoodRx
Nicotine replacement	Multiple, largely OTC	Combination patch plus short-acting formulation	Low; OTC pricing varies by retailer
Semaglutide	Wegovy · Novo Nordisk (NYSE: NVO)	Class IIa weight-management adjunct within nutritional counselling	Self-pay tiers published by NovoCare; insurance coverage highly variable
Sildenafil	Viagra · Pfizer (NYSE: PFE); generics widely available	Considered safe in stable coronary disease off nitrates, within sexual health counselling	Generic pricing low; varies by pharmacy

Pricing changes frequently and figures above should be treated as approximate rather than quoted to patients.

The bottleneck is capacity, not evidence

Europe offers roughly one rehabilitation slot for every seven patients with coronary disease, a shortfall in the range of 3.4 million places annually, and participation across countries ranges from 9% to 50% of eligible patients.

Approximately 40% of European programmes operate without financial support from social security systems, which converts a Class I therapy into an out-of-pocket purchase.

The intervention with the largest measured effect on enrolment is administrative rather than clinical: automatic referral combined with a personal contact from programme staff increases participation 8.4-fold compared with routine referral.

Where the guideline intervenes. The recommendations act at two points the referring cardiologist controls — the referral mechanism and the timing — before delivery mode, which is selected by risk, geography, and patient preference rather than by default.

The American gap

No equivalent United States guideline exists, and domestic guidance remains distributed across scientific statements on core components and on home-based delivery rather than a graded, standalone document.

Utilisation is correspondingly poor, with roughly one in four eligible Medicare beneficiaries participating, only about a quarter of those starting within three weeks of the index event, and completion of the full 36-session benefit in a minority of enrollees.

Participation falls further among women, patients over 85, and Hispanic beneficiaries, and is lowest across the Southeast and Appalachia.

The practical value of the European document for American practice is that it supplies defensible numbers — a start date, a session floor, an intensity anchor — that can be written into an order set and audited.

Case scenario

A 58-year-old woman presents with anterior STEMI, undergoes primary PCI of the LAD with a drug-eluting stent, and is discharged on day three with an ejection fraction of 45% and a 20-pack-year smoking history.

She lives 45 minutes from the nearest programme, works full time, and declines the standard three-day-per-week centre-based schedule at the bedside referral conversation.

Under the new framework she is auto-referred with a follow-up call from programme staff within 48 hours, and enrolled on day 11 in a hybrid model: two supervised centre visits weekly for four weeks, then remotely supervised sessions with heart rate telemetry.

Submaximal testing places her VT1 at 104 bpm and VT2 at 128 bpm, so her moderate-domain prescription is set at 95–104 bpm rather than the 70%-of-peak target of 118 bpm that would have placed every session in the heavy domain.

She screens positive for elevated cardiac anxiety on the programme's patient-reported outcome battery, is referred for a brief course of cognitive behavioural therapy, starts varenicline with a 12-week cessation plan, and completes 26 sessions over 14 weeks.

What the guideline concedes

The evidence gaps are stated rather than glossed, and they are substantial.

Prehabilitation before planned procedures, cost-effectiveness across delivery settings, optimal programme design in complex multimorbidity, and the incremental value of wearables and consumer applications are all flagged as unresolved.

Long-term comparative outcome data for telerehabilitation against centre-based delivery remain thin, particularly for rare adverse events, which is the honest reason centre-based care holds the stronger class.

Cultural and ethnic determinants of adherence, and predictors of maintenance after programme completion, are identified as priorities without current answers.

Practical takeaways

Referral should be automatic and event-triggered rather than discretionary, with a human contact attached.

The order should carry a date, not a phrase, targeting 14 days after acute coronary syndrome and 28 days after surgery.

Intensity should be prescribed from threshold-anchored heart rates, using the prediction equations where cardiopulmonary exercise testing is unavailable.

Frailty, advanced age, preserved ejection fraction heart failure, device implantation, and prior cancer therapy are indications for a modified programme, not reasons to withhold one.

Twenty-four sessions is the floor, more is better, and completion rather than enrolment is the metric worth auditing.

References

1. 2026 ESC Guidelines on cardiac rehabilitation. European Heart Journal, 28 August 2026. doi:10.1093/eurheartj/ehag099
2. 2026 ESC Guidelines on cardiac rehabilitation. European Journal of Preventive Cardiology. doi:10.1093/eurjpc/zwag417
3. ESC Clinical Practice Guidelines hub — cardiac rehabilitation, including slide set and patient versions. escardio.org
4. ESC introduces first clinical guideline for cardiac rehabilitation. TCTMD, September 2026. tctmd.com
5. ESC releases new guidelines for cardiac rehab, CKD and HF. American College of Cardiology, 29 August 2026. acc.org
6. Cardiac rehabilitation availability and delivery in Europe. European Journal of Preventive Cardiology 2019;26(11):1131–46. Oxford Academic

Further viewing

• How to find ventilatory thresholds VT1 and VT2 — CPET interpretation
• Interpretation of cardiopulmonary exercise tests, part 1
• 2026 ESC Guidelines for the management of heart failure — companion release

Educational content for clinicians; not a substitute for the full guideline text or for individual clinical judgement. Recommendation classes and evidence levels are summarised from the published document and secondary analyses available at the time of writing, and the primary text should be consulted before implementation. Pricing is approximate and changes frequently. The case scenario is fictional and contains no patient identifiers.

TARGET-CTCA: When MI Has Already Been Ruled Out, Does Coronary CT Angiography Add Value

At the European Society of Cardiology (ESC) Congress 2026, investigators presented the TARGET-CTCA Trial, examining whether Coronary CT Angiography (CCTA) improves outcomes in patients presenting with acute chest pain after myocardial infarction (MI) has already been excluded. The results challenge a long-standing assumption that identifying more coronary artery disease automatically translates into better patient outcomes. 

Why This Study Matters

Modern emergency departments increasingly rely on high-sensitivity troponin testing to rapidly evaluate patients with suspected acute coronary syndrome (ACS). Many patients ultimately have MI ruled out but remain at intermediate cardiovascular risk due to modest troponin elevations or clinical risk factors.

Current U.S. chest pain guidelines support selective use of CCTA in such patients. However, whether imaging-guided management reduces future cardiovascular events has remained uncertain.

TARGET-CTCA Trial Design

The study enrolled 3,170 patients presenting to emergency departments across 14 hospitals in the United Kingdom.

Patient Characteristics

• Median age: 61 years
• Women: 30.2%
• Prior MI: 17.5%
• History of angina: 15.1%
• Nearly 90% presented with chest discomfort

Patients were randomized to either:

1. CCTA-guided management
2. Standard care

The median follow-up period was approximately three years. 

What Did CCTA Reveal?

Among patients undergoing Coronary CT Angiography:

• 42.5% had non-obstructive coronary artery disease
• 22.6% had obstructive coronary artery disease
• More patients received statin therapy
• More patients received antiplatelet therapy
• Rates of percutaneous coronary intervention (PCI) were modestly higher

These findings demonstrated that CCTA successfully identified previously undiagnosed atherosclerotic coronary artery disease and altered medical management. 

The Primary Outcome

The primary endpoint was a composite of:

• Myocardial Infarction
• Death from a Cardiac Cause

After approximately three years:

Group	Primary Event Rate
CCTA	7.1%
Standard Care	7.3%

Despite identifying more coronary disease and increasing preventive treatments, CCTA did not significantly reduce myocardial infarction or cardiac death. 

Why Didn't More Imaging Improve Outcomes?

Lead investigator Dr. Nicholas Mills noted that elevated high-sensitivity troponin is not specific for coronary artery disease alone.

Troponin elevations may result from:

• Cardiomyopathy
• Heart Failure
• Chronic Kidney Disease
• Structural Heart Disease

The findings suggest that future adverse events in this population may be driven by factors beyond obstructive coronary disease, limiting the impact of imaging-guided interventions.

Clinical Implications

The TARGET-CTCA Trial provides several important takeaways:

✅ CCTA effectively detects coronary artery disease

✅ Imaging increases use of preventive medical therapy

✅ Routine CCTA after MI exclusion does not reduce future MI or cardiac death

✅ No subgroup demonstrated a clear clinical benefit

✅ Troponin elevation alone should not automatically trigger coronary CTA

For practicing cardiologists, these results support a more selective and individualized approach to imaging rather than routine CT evaluation of every intermediate-risk chest pain patient. 

What Comes Next?

Attention now turns to the ongoing FAST-CCTA Trial, which may help identify specific patient populations that benefit from CT-based evaluation.

Researchers are also exploring:

• Artificial intelligence–based plaque analysis
• Advanced plaque characterization
• Coronary inflammation assessment
• Enhanced risk stratification tools

These technologies may ultimately improve identification of high-risk patients beyond traditional stenosis assessment.

Bottom Line

The TARGET-CTCA Trial demonstrated that while Coronary CT Angiography identifies significant coronary artery disease and leads to more aggressive preventive therapy, it does not reduce rates of myocardial infarction or cardiac death after MI has already been ruled out.

For cardiologists managing patients with acute chest pain, the study reinforces an important principle: better diagnosis does not always translate into better outcomes, and careful patient selection remains essential when considering advanced cardiac imaging.

Reference

Lee KK, Wereski R, Lowe D, et al. Targeted Use of Computed Tomographic Coronary Angiography in Acute Chest Pain. New England Journal of Medicine. 2026.

ESC · ACC · AHA · WHF  |  28 August 2026

The Numbers Are Gone

What the Fifth Universal Definition of MI Actually Changes

Bishnu Subedi, MD, FACC

Types 1 through 5 are retired. Three named categories replace them, troponin thresholds go sex-specific, and "Type 2 MI" finally has a bar to clear.

For nineteen years we have been writing numbers in charts. Type 1. Type 2. Type 4a. On 28 August 2026, the European Society of Cardiology, the American College of Cardiology, the American Heart Association and the World Heart Federation retired them.

The Fifth Universal Definition of Myocardial Infarction replaces the numerical classification with three named categories based on mechanism: primary, secondary and procedure-related myocardial infarction. It also mandates sex-specific troponin thresholds, strips the biomarker multiples out of the periprocedural criteria, and demotes MINOCA from a diagnosis to a working diagnosis.

Presented at ESC Congress 2026 and published simultaneously in the European Heart Journal, JACC, Circulation and Global Heart. The task force was co-chaired by Professor Nicholas Mills for the ESC and Professor Kristin Newby for the ACC/AHA. Mills was blunt about the motivation: the previous definition's numerical categories "were not always easy to apply in clinical practice." Newby framed the payoff at the bedside — clinicians can now discuss the cause of an MI with a patient in terms the patient can actually follow.

This is not a relabelling exercise. Four things genuinely change.

1. Three names, and three things that crossed categories

The translation is mostly one-to-one — Type 1 becomes primary MI, Type 2 becomes secondary MI, Types 4a and 5 merge into procedure-related MI. The interesting part is what moved.

Late stent thrombosis, in-stent restenosis and late graft failure are now primary MI. Under the numbered system these were permanently procedural — Type 4b and 4c — however many years after the index PCI they occurred. The 2026 framework treats a coronary event beyond the 30-day procedural window as de novo disease, which is what it pathophysiologically is. Thrombus on a stented segment at fourteen months behaves like thrombus on a plaque.

Type 3 is gone as a category. It described a circumstance — cardiac death before biomarkers could be drawn — rather than a mechanism, which is why it always sat awkwardly beside the others. It is replaced by explicit criteria for diagnosing MI following sudden death, after which the case is named for whatever mechanism is established.

The translation, at a glance

You used to write	You now write
Type 1	Primary MI
Type 2	Secondary MI — but see below
Type 3	No longer a category
Type 4a	Procedure-related MI
Type 4b, within 30 days	Procedure-related MI
Type 4b, beyond 30 days	Primary MI
Type 4c	Primary MI
Type 5	Procedure-related MI
MINOCA	A working diagnosis, then reclassify

2. Secondary MI finally has a bar to clear

This is the change most likely to alter your hospital's numbers, and the one most worth reading carefully.

Under the 2026 framework, secondary MI requires more than an imbalance and a rising troponin. It requires either obstructive coronary disease without acute coronary pathology, or a new regional wall motion abnormality.

Consider what that excludes. A septic patient in fast AF with a troponin that rises and falls, no prior coronary imaging and no echocardiogram, does not have secondary MI. That patient has acute myocardial injury — which is not a lesser diagnosis, carries a serious prognosis of its own, and is simply the accurate term.

Watch for this

"Type 2 MI" had drifted into being a way of noting a raised troponin in a sick patient rather than making a diagnosis. The new criteria stop that. Expect your secondary MI counts to fall and your acute myocardial injury counts to rise — and expect somebody to misread that as a clinical improvement.

3. Sex-specific troponin thresholds become the standard

Healthy women have lower circulating troponin than men, principally because left ventricular mass is lower. A single overall 99th percentile therefore sits above the true female limit — and women with genuine myocardial injury get reported as normal.

The magnitude is not subtle. On a widely used hs-cTnI platform the published sex-specific limits are roughly 16 ng/L for women against 34 ng/L for men, with an overall limit of 26 ng/L. Every woman falling between 16 and 26 ng/L has been sitting in a diagnostic blind spot. The High-STEACS programme showed a decade ago that closing it roughly doubles the proportion of women diagnosed with MI, and that those newly identified women carry real risk while being less likely than men to reach angiography or secondary prevention.

Two caveats worth stating plainly. First, correcting the threshold identifies injury; the ischaemia gate is unchanged, and these women still need the same diagnostic reasoning as anyone else. Second, this is a laboratory information system project, not a memo — the reference limit, the result flag, the chest pain pathway, the decision support and the audit queries all have to ship in the same release, or you will produce results flagged abnormal that your pathway still treats as normal.

4. Procedure-related MI: one criterion set, no biomarker multiple

The 5× and 10× thresholds are gone. PCI and cardiac surgery are now assessed by identical criteria within a 30-day window, and the diagnosis turns on a demonstrated coronary complication — dissection, side-branch occlusion, no-reflow, acute closure, a graft problem — rather than on how high the troponin climbed.

Anyone who has adjudicated a revascularisation trial will recognise why this matters. Holding the surgical arm to a numerically higher bar than the percutaneous arm made every composite endpoint containing periprocedural MI internally inconsistent. That structural problem is now fixed.

The corollary for clinical work: troponin release after uncomplicated cardiac surgery is procedural myocardial injury. Expected, prognostically meaningful, worth recording — and not an infarction. When you are asked to see a post-procedure patient with a rising troponin, the first question is no longer "how high?" but "what did the operator or surgeon actually see?"

What to do on Monday

1. Call your laboratory. Ask whether the troponin report applies a sex-specific 99th percentile and whether it says so on the result. If not, every "normal" troponin in a woman in that band is being mishandled by your pathway right now. Until it changes, know your assay's female limit and interpret manually.
2. Stop writing bare "NSTEMI". Write the category and the mechanism as a clause: "Non-ST-elevation primary myocardial infarction due to plaque rupture, mid-LAD." Coders code the diagnosis line, not your narrative.
3. Use "acute myocardial injury" out loud. In handover, on the ward round, in the note. If it is harder to say than "MI", people will say "MI".
4. Get the echo. Under the new criteria, a new regional wall motion abnormality is often the deciding evidence between secondary MI and acute myocardial injury.
5. Do not let a patient leave with "MINOCA" as the final diagnosis. Either you can name the mechanism, or the discharge summary says which investigation is outstanding and who is arranging it.
6. Warn your audit team. Every time series that crosses this transition has a break in it. Map your historical codes to the new categories before anyone draws a trend line.

The part that is easy to miss

Two smaller changes deserve more attention than they will get.

The document provides explicit guidance for settings without access to coronary and cardiac imaging — the first universal definition to acknowledge that most of the world's infarctions occur where the diagnostic pathway it describes cannot be performed. And the proposed ICD-11 alignment means the three categories can finally be recorded distinctly and compared across health systems, with better surveillance of mechanisms that have been chronically under-captured. The societies name spontaneous coronary artery dissection as the example — a predominantly female condition that a mechanism-blind coding system was never going to count properly.

The honest caveat

Everything above reflects the societies' release material and the first wave of reporting. The primary document is the authority on its own wording, and the exact criteria — particularly the supporting evidence required for secondary MI and the specifics of the procedure-related definition — should be read in full before anyone rewrites a protocol. Assay-specific numbers quoted here are illustrations of magnitude, not cut-points to adopt.

But the direction is unambiguous, and it is the right one. A number could be written by a clinician who had noticed a raised troponin and gone no further. A name is a claim about what happened — and the 2026 criteria now ask you to have evidence for it.

---

Sources
Fifth Universal Definition of Myocardial Infarction (2026), J Am Coll Cardiol, doi:10.1016/j.jacc.2026.07.025 · ESC guideline page · ESC Congress 2026 press materials · Thygesen K, et al. Fourth Universal Definition of Myocardial Infarction (2018).

Educational commentary for a professional audience. Not a clinical guideline and not a substitute for the primary document.

ESC Congress 2026 · Munich · Day 2

One Win, Three Reality Checks

Saturday, 29 August 2026 — what mattered for practising cardiologists

The day's clear winner was STAREE, which finally gives us randomised evidence for statins in primary prevention after age 70. Among 9,971 community-dwelling adults free of vascular disease, diabetes and dementia, atorvastatin 40 mg cut major cardiovascular events by roughly a third (6.0% vs 8.3%; HR 0.70). The catch: the co-primary endpoint of disability-free survival was unchanged, and muscle, hepatic and dysglycaemic adverse events were more common. So the drug prevents infarcts and strokes in older people — it does not, on this evidence, buy them independent years.

The other three Hot Lines all landed negative — and each was a widely held assumption.

In LIBREXIA-ACS (14,194 patients), the oral factor XIa inhibitor milvexian added to antiplatelet therapy after acute coronary syndrome did nothing to ischaemic risk (5.4% vs 5.1%; HR 1.05), though it did keep its clean bleeding signature. Meanwhile AMUNDSEN showed that giving evolocumab before primary PCI doubles the proportion reaching LDL-C goal at one year (82% vs 40%) — but with no significant difference in death or unplanned cardiovascular hospitalisation in the intention-to-treat analysis. Surrogate met, outcome not yet.

Most provocative for anyone running a chest-pain pathway: PRESC1SE-MI, a stepped-wedge trial across 67,624 emergency presentations in 11 countries, found the ESC 0/1-hour high-sensitivity troponin algorithm to be reassuringly safe (30-day death or type 1 MI 1.1% vs 1.2%, non-inferior) yet completely neutral on emergency department length of stay — a median of 309 minutes in both arms. Faster biochemistry does not equal faster discharge; the bottleneck is the workflow around the assay, not the assay.

Key message

Day 2 rewarded an old, cheap drug in an undertreated age group — and reminded us that a better biomarker, a cleaner anticoagulant, or a lower LDL-C does not automatically become a better outcome.

Learn more

• STAREE — atorvastatin, cardiovascular events and disability-free survival in older adults (NEJM)
• STAREE trial summary (American College of Cardiology)
• LIBREXIA-ACS — milvexian after recent acute coronary syndrome
• AMUNDSEN — evolocumab before PCI for acute myocardial infarction (JAMA)
• PRESC1SE-MI — ESC 0/1-hour algorithm and emergency department length of stay
• Full ESC Congress 2026 Hot Line programme

Trial data as presented at ESC Congress 2026, Munich, 29 August 2026. For educational use; verify against the primary publications before changing practice.

🫀 ESC Congress 2026 — Day 1 Roundup (Hot Line 1)

Munich, Germany — Friday, August 28, 2026. ESC Congress 2026 opened with Hot Line 1, spanning amyloid cardiomyopathy, nonobstructive HCM, AF anticoagulation thresholds, endocarditis antibiotic duration, and CMR-guided ICD selection. Key data and clinical takeaways from all five presentations are summarized below.

---

🧬 CARDIO-TTRansform: Eplontersen Misses Primary Endpoint in ATTR-CM

The largest-ever ATTR-CM trial randomized 1,432 patients with wild-type or hereditary transthyretin amyloid cardiomyopathy (mean age 72; 57% on a baseline TTR stabilizer) to eplontersen 45 mg SC every 4 weeks or placebo. The composite primary endpoint of CV mortality plus recurrent CV events through week 140 was not met (rate ratio 0.89, 95% CI 0.73–1.09, P = 0.277), despite robust on-target TTR suppression.

A prespecified subgroup of patients not on a baseline stabilizer (monotherapy) showed a nominally significant reduction in events, while those already on a stabilizer derived no incremental benefit — raising questions about background therapy saturation or trial timing relative to disease stage. Safety was consistent with prior eplontersen data.

Presenters framed this as a signal for further exploration in monotherapy or earlier-stage populations rather than a closed door for TTR silencing in ATTR-CM.

Coverage:
CARDIO-TTRansform: Eplontersen Misses Primary Endpoint in ATTR-CM

💊 ACACIA-HCM: Aficamten Positive in Nonobstructive HCM

ACACIA-HCM randomized 517 symptomatic nonobstructive HCM patients (mean LVEF 68%, mean KCCQ-CSS 65.7) to aficamten (titrated 5–20 mg) or placebo. At week 36, aficamten produced a greater KCCQ-CSS improvement (11.4 vs 8.4 points, P = 0.02, widening to 7.0 points by week 72), higher peak VO2 (+0.64 vs −0.03 mL/kg/min, P = 0.003), and more ≥1-class NYHA improvement (41.9% vs 27.8%).

A composite responder analysis (symptoms, exercise capacity, LAVI, septal e′, NT-proBNP) showed a "clinical response" (≥3 domains) in 53% vs 13% (P < 0.001). Reversible LVEF reduction <50% occurred in 10.5% vs 0.8%; 2.7% dropped below 40% and paused dosing, with two permanent discontinuations. LVEF normalized after washout.

With no approved pharmacotherapy currently available for this phenotype, the benefit-risk profile is viewed as sufficient to support regulatory filing, contingent on LVEF monitoring similar to a REMS program.

Full coverage:
Aficamten Wins in Nonobstructive Hypertrophic Cardiomyopathy: ACACIA-HCM

🩸 SINGLE-AF: First RCT Evidence for DOACs at Intermediate CHA2DS2-VASc Risk

Current Class IIa recommendations for anticoagulation at intermediate stroke risk (CHA2DS2-VASc 1 in men, 2 in women) rest on conflicting observational data. SINGLE-AF randomized 1,803 such patients (mean age 60.4; 71.8% paroxysmal AF) to DOAC (apixaban or rivaroxaban) versus no routine anticoagulation.

At 24 months, the composite of stroke, systemic embolism, ISTH major bleeding, or CV death occurred in 0.5% (DOAC) vs 1.5% (control) (HR 0.31, 95% CI 0.10–0.94), driven mainly by fewer strokes (3 vs 10 events). Serious adverse events were similar between arms (8.9% vs 9.3%).

Given the small absolute event count, this is regarded as hypothesis-generating rather than practice-changing; CKD was an exclusion criterion, and generalizability outside the Korean cohort is uncertain. Larger, more heterogeneous confirmatory trials are needed before any guideline revision.

Full coverage:
DOACs Benefit AF Patients With Intermediate Stroke Risk: SINGLE-AF

💉 POET II: Response-Tailored Antibiotic Duration Noninferior in Left-Sided IE

Building on POET I's oral step-down strategy, POET II randomized 508 clinically stabilized patients with streptococcal, S. aureus, or E. faecalis left-sided infective endocarditis to response-tailored discontinuation (median 26 days total therapy) versus standard 4–6 week duration (median 41 days).

The primary safety endpoint (death, unplanned surgery, or symptomatic embolism at 6 months) met noninferiority: 8.2% (tailored) vs 10.7% (standard), P < 0.001. Days alive without antibiotics favored the tailored arm (183 vs 169 days, P < 0.001 for superiority). Relapse was more frequent with tailored therapy (~5% vs 1.6%, P = 0.04), driven largely by E. faecalis (6.4% relapse rate); most relapses were uncomplicated and managed medically.

Discussants endorsed a shift toward personalized, stabilization-criteria-driven duration, with caution advised for enterococcal endocarditis and implementation limited to high-volume endocarditis centers with rigorous stabilization protocols.

Full coverage:
POET II Supports Shorter Antibiotic Course in Infective Endocarditis

🧲 CMR GUIDE: Scar Burden Doesn't Clearly Extend ICD Benefit to LVEF 36–50%

Current primary-prevention ICD criteria are anchored to LVEF ≤35%, leaving patients with mild-to-moderate systolic dysfunction unaddressed despite known arrhythmic risk conferred by myocardial scar. CMR GUIDE randomized 353 patients with ischemic or nonischemic cardiomyopathy, LVEF 36–50%, and LGE-positive scar on CMR to ICD versus implantable loop recorder.

The composite primary endpoint (SCD plus hemodynamically significant VA) was not met (7.8% vs 9.2%; HR 0.76, 95% CI 0.37–1.58). SCD as a standalone endpoint was reduced 72% with ICD (1.7% vs 5.8%; HR 0.26), and a significant age interaction (P = 0.01) showed benefit confined to patients <70 years (HR 0.28) with a signal toward harm in those ≥70 (HR 2.33). NNT over the 6.3-year median follow-up was 24.

Discussants characterized this as absence of evidence rather than evidence of absence, given underpowering from lower-than-expected event rates. No guideline change is anticipated pending a properly powered, age-stratified trial.

Full coverage:
Scarring Doesn't Tip the Scale Towards ICDs in Cardiomyopathy With LVEF of 36–50%

---

📚 References

CARDIO-TTRansform (Maurer MS, et al.):
Rationale and Design of CARDIO-TTRansform. Circ Heart Fail. 2026;doi:10.1161/CIRCHEARTFAILURE.126.014205 (outcomes publication pending; design paper linked)

ACACIA-HCM (Masri A, et al.):
Aficamten for symptomatic nonobstructive hypertrophic cardiomyopathy. N Engl J Med. 2026;doi:10.1056/NEJMoa2603021

SINGLE-AF (Kim D, et al.):
Anticoagulation for atrial fibrillation with intermediate stroke risk. N Engl J Med. 2026;doi:10.1056/NEJMoa2607978

POET II (Bundgaard H, et al.):
Response-tailored or standard-duration antibiotic treatment for infective endocarditis. N Engl J Med. 2026;doi:10.1056/NEJMoa2607887

CMR GUIDE (Selvanayagam JB, et al.):
Cardiovascular magnetic resonance to guide defibrillator implantation for LVEF of 36% to 50%. JAMA. 2026.

Source reporting: TCTMD, HCPLive, ESC Congress 2026 (Munich). Summarized and paraphrased from original coverage; consult linked primary publications for full methodology and data.