Deferring PCI before TAVR cut bleeding, but complex disease did worse

Design

Post hoc substudy of the randomised PRO-TAVI trial

Population

466 patients with significant coronary artery disease undergoing TAVR, stratified by SYNTAX score

Primary outcome

Composite of death, myocardial infarction, stroke and major bleeding at 1 year

Effect

No difference overall; SYNTAX above 17: ischaemic composite 30.9% vs 14.8% (adjusted HR 2.43, 95% CI 1.04 to 5.64)

This substudy of the PRO-TAVI trial included 466 patients with significant coronary artery disease who were randomised to deferral of PCI or PCI before transcatheter aortic valve replacement. The authors graded disease extent with the SYNTAX score and compared outcomes by quartile at one year.

The primary composite of death, myocardial infarction, stroke and major bleeding did not differ between deferral and PCI across SYNTAX categories. Bleeding was less frequent with deferral in every category. In patients with a score above 17, the ischaemic composite of death, infarction and stroke occurred more often with deferral (30.9% vs 14.8%; adjusted HR 2.43, 95% CI 1.04 to 5.64), driven mainly by death and stroke.

The subgroup was defined after the trial and the interval is wide, so this generates a hypothesis. It argues for a closer look at lesion complexity when the heart team decides, not for a change in practice.

• Deferral of PCI before TAVR gave less bleeding across all SYNTAX score groups.
• With a SYNTAX score above 17, ischaemic events were 30.9% with deferral vs 14.8% with PCI.
• The complex-disease signal comes from a post hoc subgroup with a wide interval (HR 2.43, 1.04 to 5.64).
• Document the SYNTAX score in the heart team discussion.
• Consider individualising, and do not change a pathway on this analysis alone.

Why it matters

It challenges the idea that the safety of deferral is uniform across all coronary anatomy.

Don't overread it

This is a post hoc subgroup analysis of one trial and the authors themselves say it needs confirmation.

The statistics, in plain English

A subgroup finding that was not planned in advance is weaker than the headline result, because slicing a trial many ways produces chance differences. The lower end of the confidence interval (1.04) sits just above 1.0, so the result is borderline even before this caveat.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.