The Rhythm | Issue 26
Issue 26 | Week of 28 September 2026
Cardiology Intelligence Briefing: the marker is not the outcome
Cardiology depends on markers. The corrected QT interval (QTc), valve gradients and pressure drops across coronary stenoses give us measurable signals before we know whether a patient will develop an arrhythmia, heart failure, myocardial infarction (MI) or die. This week showed both sides: a marker can mislead when overinterpreted, yet help once its link with outcomes is understood.
Two studies moved QTc in opposite directions. In long QT syndrome type 2, lumacaftor shortened QTc by an average of 31 ms in a small, selected cohort, a mechanistic signal that justifies further testing but does not establish clinical benefit.3 In major depressive disorder, some antidepressants lengthened QTc by a few milliseconds on average, yet across 139 randomised trials there was no clear increase in non-suicidal sudden death.2 A favourable change in a marker is not proof of benefit, just as an unfavourable change is not proof of harm.
The imaging studies show markers working when read in context. Flow-based criteria converted most indeterminate stress echocardiograms in low-gradient aortic stenosis into a usable classification.4 In a nationwide English cohort, lower computed tomography-derived fractional flow reserve (FFR-CT) values were associated with MI beyond risk factors and stenosis severity, but mainly at clearly abnormal values; the borderline range chiefly predicted revascularisation, an outcome partly shaped by the test result itself.5
The menopausal hormone therapy study offers a contrast: it measured clinical events directly. In more than one million Danish women, associations with venous thromboembolism (VTE), ischaemic stroke and MI varied by route, dose and duration, while transdermal oestradiol was not associated with an overall increase.1
The lesson is not to distrust markers. The challenge is to be precise about what each marker has been shown to predict, and what it has not. The marker can guide the question. It should not be mistaken for the answer.
