A second look at declined donor lungs, and two ESC readouts land in NEJM

1. A two-surgeon review system for declined donor lungs

Journal · Published: JTCVS, 1 September 2026 — genuinely new, 1 day old Gil-Barturen, Yamamoto, Pessoa, Wakeam, Yeung, Pierre, de Perrot, Yasufuku, Waddell et al.

A single high-volume centre implemented a system in which donor lungs declined by the primary on-call surgeon are reassessed by a second surgeon. Retrospective review of lung transplant recipients, August 2022 – September 2024, comparing recipients of initially-accepted organs with those accepted only after second review.

• 428 transplants in two years; 34 (8%) used donors accepted after second review
• Of 880 initially declined offers, 57 were accepted on review and 34 transplanted
• EVLP use similar (38.2% vs 32.9%, p=0.666); donor and recipient characteristics comparable
• PGD grade 3 at 72 h: 11.8% vs 12.9%
• Postoperative ECMO 5.9% vs 5.3%; time to extubation 70 vs 52 h; ICU stay 6 vs 6 days; hospital stay 26 vs 23 days; overall survival similar
• Transplant volume increased by 8.6%

Interpretation. An unusually clean example of a process intervention with a real denominator behind it. The finding that matters is not the 8.6% but the 880 declined offers it came out of: a single centre turned down that many organs in two years, and a second look converted 6.5% of them into transplants with indistinguishable early outcomes.

That reframes donor decline as partly a judgement-variance problem rather than purely an organ-quality one. The intervention costs nothing but a phone call and a colleague’s time, and it is immediately portable to any centre with more than one transplant surgeon on staff.

Two honest limits. This is retrospective and single-centre, so the second reviewers were selecting, not randomising — the 34 transplanted were the ones a second surgeon was confident about, which is precisely the point but also means the comparison is between two groups both judged acceptable by someone. And 34 patients cannot exclude a modest excess of PGD or late graft dysfunction; longer follow-up matters more here than the early numbers.

Read the paper · PMID 42680010

2. SWITCH-SWEDEHEART — prasugrel versus ticagrelor after ACS

Journal · Published: NEJM, 29 August 2026 — 4 days old, from the ESC Congress batch Omerovic, Koul, Andersson, Fröbert et al.

Registry-based, open-label, stepped-wedge, cluster-randomised trial in Sweden. Seven regions in three clusters switched their default P2Y12 inhibitor from ticagrelor to prasugrel in randomised sequence across four 9-month periods, 2021–2024. Outcomes ascertained from national registries; intention-to-treat with a mixed model adjusted for cluster and calendar time, giving a policy-level estimate.

• 17,095 patients — 9,444 under default ticagrelor, 7,651 under default prasugrel
• Mean age 69.8 ± 11.5; 36.8% aged ≥75; 27.8% women; 39.4% STEMI, 43.7% NSTEMI, 16.8% unstable angina
• Primary composite (death, MI, or stroke at 1 year): 11.8% vs 11.1%, adjusted OR 0.90 (95% CI 0.77–1.06) — not significant
• Major bleeding: 4.4% vs 4.2%, adjusted OR 0.80 (95% CI 0.64–0.99)

Interpretation. The design is the interesting part. A registry-based stepped-wedge cluster randomisation of a default prescribing policy answers the question clinicians actually face — which drug should the department default to — rather than the question trials usually answer, which is how a selected patient fares on an assigned drug. The 17,095 unselected patients, over a third of them 75 or older, are a population ISAR-REACT 5 never enrolled.

On results: no ischaemic advantage for prasugrel, and a bleeding odds ratio of 0.80 whose upper bound touches 0.99 — nominally favouring prasugrel, but a secondary endpoint sitting right on the boundary, in an open-label trial, is a signal to watch rather than to act on. The practical reading is that this is a genuine equipoise result: departments defaulting to either drug have no strong reason to switch.

Note this contrasts with ISAR-REACT 5, which found prasugrel superior for ischaemic events. The difference is likely population and adherence under real-world defaults rather than a contradiction — which is itself the argument for doing trials this way.

Read the paper · PMID 42670968

3. TRIC-I-HF — transcatheter tricuspid repair in heart failure

Journal · Published: NEJM, 30 August 2026 — 3 days old, ESC Congress batch Hausleiter, Stocker, Geisler, Lurz, Rottbauer, Thiele et al.

Patients with symptomatic severe tricuspid regurgitation and increased risk of future heart-failure events, randomised 2:1 to transcatheter tricuspid repair plus medical therapy or medical therapy alone. First primary endpoint: hierarchical composite of death, heart-failure hospitalisation and quality-of-life improvement at 1 year, assessed by win ratio, with a second primary endpoint (death or HF hospitalisation through 3 years) tested if the first was significant.

• 360 patients — 237 repair, 123 medical therapy
• Mean age 80.3 ± 6.4; 56.4% women
• Win ratio 2.42 (95% CI 1.76–3.33) for the first primary endpoint
• Also lower risk of death or HF hospitalisation through 3 years

The deposited abstract truncates at the p-value for the win ratio; the confidence interval is well clear of 1.

Interpretation. A positive tricuspid trial, which the field has been waiting for. TRILUMINATE showed quality-of-life benefit without moving death or hospitalisation, leaving tricuspid intervention in the awkward position of being demonstrably symptom-improving and unproven on hard endpoints. TRIC-I-HF selected patients at increased risk of heart-failure events and reports benefit on the three-year composite — enriching for events is exactly what TRILUMINATE arguably failed to do.

Read the win ratio carefully, though. A hierarchical composite ranks death first, then hospitalisation, then quality of life — so a win ratio of 2.42 does not mean 2.42-fold fewer deaths; much of the win can accrue at the quality-of-life tier. The three-year hard-endpoint result is the more meaningful claim, and worth checking the absolute numbers on when the full text is accessible.

For perioperative relevance: mean age 80, 56% women, and a population being selected for intervention rather than medical management means more octogenarians presenting for structural procedures with right-heart failure physiology — a distinct and unforgiving anaesthetic problem.

Read the paper · PMID 42670975