The Rhythm | Issue 16

Issue 16 | Week of 20 July 2026

A quiet week on the conference calendar, but not a quiet one in the journals. Five papers this week span antithrombotic strategy, a novel mechanistic approach in hypertrophic cardiomyopathy (HCM), population-level prevention, and the accuracy of lipid measurement.

Cardiology Intelligence Briefing: precision over reflex

This week's evidence kept returning to the same question: not simply whether to treat, but how intensively, for how long, and in whom. Two large antithrombotic trials reached seemingly opposite conclusions in very different clinical settings.^1,2

In patients with multivessel coronary disease who had remained event-free for a year after stenting, continuing dual antiplatelet therapy for a second year reduced cardiovascular death, myocardial infarction, or stroke without a statistically significant increase in clinically relevant or major bleeding.¹ In atrial fibrillation after coronary intervention, shortening dual antithrombotic therapy to one month before continuing anticoagulation alone was non-inferior to twelve months of combination therapy for death or thromboembolic events, while substantially reducing bleeding.² The apparent contradiction is the lesson: the right duration depends on whether ischaemic or haemorrhagic risk dominates.

Elsewhere, TEMPEST offered a novel mechanistic direction in HCM. The copper-modulating agent trientine produced a modest reduction in left ventricular mass through a pathway distinct from the myosin inhibitors that have dominated recent HCM drug development.³ It remains a surrogate-endpoint signal from a relatively small trial, but it opens a different therapeutic avenue worth watching.

At the population level, almost nine in ten US ischaemic heart disease (IHD) deaths in 2023 were estimated to be attributable to modifiable risk factors, while the pace of improvement in mortality has slowed markedly since 2010.⁴ Blood pressure, diet, and LDL cholesterol remain major contributors, but rising obesity and dysglycaemia increasingly threaten to offset gains made elsewhere.

Finally, a machine-learning simplification of the Martin-Hopkins LDL cholesterol equation retained essentially the same accuracy while making the method easier for laboratories to implement.⁵ Improving an equation used across millions of lipid panels is unglamorous work, but small gains in accuracy can have an unusually large reach.

The throughline is precision over reflex. How much treatment, for how long, and for whom matters as much as the treatment itself.