ACACIA-HCM: aficamten met both primary endpoints in non-obstructive hypertrophic cardiomyopathy — but the between-arm differences are 3.0 KCCQ points and 0.67 mL/kg/min, with a 10% versus 1% rate of ejection fraction below 50% (ESC Congress, 28 Aug 2026)

Why it matters (spine). The ESC's own first key takeaway is that "There are currently no approved treatments for symptomatic non-obstructive hypertrophic cardiomyopathy (HCM)," and this trial was positive on both of its pre-specified primary endpoints. That is the news. The discipline this cabinet exists to apply is to print what "positive" measured: a 3.0-point advantage on the KCCQ-CSS symptom questionnaire (that its scale runs 0–100 is background the release does not state) over a placebo arm that itself improved by 8.4, and a 0.67 mL/kg/min advantage in peak oxygen uptake, against a ten-fold higher rate of the heart's ejection fraction dropping below 50% on the drug. All of those figures are in the same press release. A reader who sees only "met dual primary endpoints" and a reader who sees only "10% versus 1%" have each read half the document.

What the ESC document says. Datelined "Munich, Germany – 28 August 2026": "Aficamten treatment was associated with significant improvements in symptom burden and exercise capacity in patients with symptomatic non-obstructive hypertrophic cardiomyopathy, according to results presented in a Hot Line session today at ESC Congress 2026 and published simultaneously in the New England Journal of Medicine." Design: "This was a double-blind, phase III trial conducted at 182 international sites. A total of 517 adults with symptomatic non-obstructive HCM were randomised (1:1) to aficamten or placebo for up to 72 weeks. The dose of aficamten was adjusted based on left ventricular ejection fraction (LVEF). The dual primary endpoints were change in Kansas City Cardiomyopathy Questionnaire Clinical Summary Score (KCCQ-CSS) and change in maximal exercise performance (pVO2) from baseline to week 36. The mean age of participants was 55 years, 54% were women and baseline LVEF was 68%." Results: "At 36 weeks, KCCQ-CSS improved by 11.4 in the aficamten group and 8.4 in the placebo group (p=0.02)," and "pVO2 improved by 0.64 mL/kg/min in the aficamten group and was almost unchanged (−0.03 mL/kg/min) in the placebo group (p=0.003)." Secondary: "Significant improvements with aficamten were also seen in secondary endpoints including New York Heart Association functional class, submaximal exercise performance and NT-proBNP, a marker of heart strain." Safety: "there was an increased incidence of patients having LVEF less than 50% with aficamten vs. placebo (10% vs. 1%) but this was managed with dose adjustment." The principal investigator, Dr Ahmad Masri of Oregon Health & Science University, is quoted: improvements were seen "as early as 12 weeks and there was a pronounced return of symptoms after aficamten treatment was stopped," and "For the first time, we have shown clinically meaningful improvements in symptoms and exercise capacity with a treatment that targets the cause of the disease." The release's own key-takeaway line: "ACACIA-HCM is the first trial to document clinically meaningful improvements in this relatively common form of inherited heart disease." The editors' note states: "ESC Congress 2026 takes place from 28–31 August at Messe München GmbH, Munich, Germany." Funding and disclosures as printed: "The trial was funded by Cytokinetics," and Dr Masri "reports research grants from Attralus, Cytokinetics and Janssen, and consulting fees from" a list of 21 companies including Cytokinetics.

Derived figures, enumerated so a missing one is detectable. Four derivations appear on this card and nowhere else: the between-arm KCCQ-CSS difference of 3.0 points (11.4 − 8.4); the between-arm pVO2 difference of 0.67 mL/kg/min (0.64 − (−0.03)); the ratio of the LVEF-below-50% rates, ten-fold (10 ÷ 1); and the count of 21 companies in the investigator's consulting-fee disclosure, counted from the release's list. Every other figure read from the release — 182, 517, 1:1, 72, 36, 12 (weeks, in the Masri quotation), 50 (the LVEF threshold), 55, 54%, 68%, 11.4, 8.4, 0.02, 0.64, −0.03, 0.003, 10%, 1% — is quoted directly. Five further figures on this card — 518, 160, 19, 53%, 13% — are unverified leads from search summaries, not from the release, and appear only in the correction paragraph below. The release does not state a minimal clinically important difference for either endpoint, confidence intervals, or the per-arm counts, and this card asserts none.

The words "first" and "clinically meaningful" on this card. Both are the ESC's and the investigator's, and appear above only inside quotation marks. Whether a 3.0-point KCCQ-CSS difference is clinically meaningful is a judgement the release asserts and the card does not; the card's contribution is to print the two arms so a reader can form their own.

A correction to the lead, on the record. Figures that circulated in this run's search summaries of the company's releases — 518 patients, 160 sites, 19 countries, and a responder analysis of 53% versus 13% — are not in the ESC document. The ESC release states 517 randomised and 182 international sites and prints no responder analysis. The company's press release and its later "additional results" release were surfaced in search but not opened — their URLs exceed the length the fetch tool accepts — so this card uses the congress's figures only, and the company's are recorded here as unverified leads.

The approval-evidence thread, extended. The 1 September log noted the cabinet holds four evidence shapes under the words "FDA approved." This card is not an approval — the release says nothing about regulatory filing — but it adds a new shape beside those four, under a different heading, of what "positive" can mean: a randomized, double-blind, placebo-controlled trial with dual primary endpoints met on a patient-reported score and an exercise measure, where the active-arm advantage is small in absolute terms and the safety signal is printed beside it. It sits among HKG's cardiovascular entries — semaglutide-cardiovascular-outcomes, olpasiran-lpa-lowering, chmp-obicetrapib-oral-cetp-positive-opinion, merck-enlicitide-oral-pcsk9-fda-approval among them; none of those cards is characterised further here because none was opened for this card.

Money angle. A drug whose investigator describes the trial as "Building on our experience with the cardiac myosin inhibitor, aficamten, in obstructive HCM" now has a positive controlled trial in the non-obstructive form, which the release says is "relatively common"; the drug's regulatory status in the obstructive form is background this card does not source. The commercial consequence depends on a regulatory action the release does not mention and a price this card does not carry. Metacrisis lever: modest and honest — a quality-of-life gain in an inherited disease, carded for the shape of its evidence.

What would prove it wrong (falsifier). Four limbs.

• The facts fail if the ESC release at the URL above does not carry the 28 August 2026 date, the 517 / 182 / 1:1 / 36-week design as quoted, the 11.4 versus 8.4 and 0.64 versus −0.03 results, or the 10% versus 1% safety figure.
• The journal record fails if the NEJM article at DOI 10.1056/NEJMoa2603021 — not opened — reports materially different primary-endpoint figures, or a different randomised count, from the release; the card's numbers are the congress's, and the paper is the authority.
• The significance fails if the effect does not hold: if longer follow-up shows the KCCQ-CSS and pVO2 advantages narrowing, or if the LVEF-below-50% rate translates into heart-failure events or discontinuations that the 36-week readout could not show.
• The regulatory implication — that this trial supports an expanded indication — fails if the sponsor does not file, or if a regulator judges the effect size insufficient; the release itself makes no regulatory claim, and this card carries none as fact.

Provenance (what I actually read). One document, fetched from the main thread on 4 September 2026: the ESC press release at escardio.org/news/press/press-releases/aficamten-trial-brings-positive-news-for-patients-living-with-non-obstructive-hypertrophic-cardiomyopathy/, headlined "Aficamten trial brings positive news for patients living with non-obstructive hypertrophic cardiomyopathy," dated "28 Aug, 2026" on its face and 2026-08-28 in its metadata. Every quotation and every figure above comes from that page, except the five lead figures marked as such — enumerated: the "no approved treatments" key takeaway, the datelined opening sentence, the design paragraph in full, the two results sentences, the secondary-endpoints sentence, the safety sentence, the three Masri quotations (including "Building on our experience"), the "first trial" key-takeaway line, the "relatively common" phrase, the congress-dates sentence from the editors' note, and the funding and disclosure sentences. The NEJM article page returned no readable text when fetched (client-rendered) and its DOI and title are taken from the ESC release and from search-result metadata; nothing on this card comes from the paper. Cytokinetics' releases of 28 and 29 August were not opened. No figure comes from trade coverage. This is a queued promotion from the 1 September log, carded on the primary's own date, which lies outside this run's window.