Atrial fibrillation after PCI: the first fortnight is where dropping aspirin costs you

Design

individual patient data network meta-analysis of randomised trials, Cox models stratified by trial, with 14-day landmark analysis

Population

10634 patients with atrial fibrillation undergoing percutaneous coronary intervention across six trials

Primary outcome

composite of cardiovascular death, myocardial infarction or stroke; and TIMI major bleeding

Effect

no difference in efficacy at 1 year (DOAC plus P2Y12 inhibitor HR 1.16, 95% CI 0.97 to 1.41 vs VKA plus DAPT); TIMI major bleeding 0.48 (0.38 to 0.65); intracranial haemorrhage 0.21 (0.07 to 0.64)

An individual patient data network meta-analysis pooled six randomised trials and 10634 patients with atrial fibrillation undergoing percutaneous coronary intervention, comparing four strategies: a direct oral anticoagulant with a P2Y12 inhibitor, a vitamin K antagonist with single antiplatelet therapy, a vitamin K antagonist with dual antiplatelet therapy, and a DOAC with dual antiplatelet therapy.

At one year, the composite of cardiovascular death, myocardial infarction or stroke did not differ between any of them, with vitamin K antagonist plus dual antiplatelet therapy as reference: DOAC plus P2Y12 inhibitor 1.16 (95% CI 0.97 to 1.41), vitamin K antagonist plus single antiplatelet 1.14 (0.85 to 1.54), DOAC plus dual antiplatelet 0.99 (0.75 to 1.30). No prespecified subgroup, including bleeding and thrombotic risk, showed an interaction. But a 14-day landmark analysis found more myocardial infarction and definite or probable stent thrombosis in the two strategies that had already dropped to a single antiplatelet - and the protocols recommended that transition at a median of 1 day and 3 days respectively.

On bleeding the picture is clearer. A DOAC with a P2Y12 inhibitor halved TIMI major bleeding compared with vitamin K antagonist plus dual antiplatelet therapy (0.48, 0.38 to 0.65), beat vitamin K antagonist plus single antiplatelet (0.62, 0.40 to 0.97), and reduced intracranial haemorrhage to about a fifth (0.21, 0.07 to 0.64). The practical reading is that the anticoagulant choice is settled in favour of a DOAC, and the open question is how fast to come off the second antiplatelet - with these data arguing against doing it on day one.

• Use a DOAC rather than a vitamin K antagonist: bleeding is halved and intracranial haemorrhage cut to about a fifth.
• Do not drop the second antiplatelet on the day of the procedure - the early excess of stent thrombosis sits in the first 14 days.
• Keep the duration of triple therapy short but deliberate, and write the stop date rather than leaving it to the next clinic.
• Bleeding and thrombotic risk scores did not identify subgroups with different treatment effects here.
• For a high thrombotic-risk lesion, the case for a fortnight of dual antiplatelet cover alongside the DOAC is stronger than for a straightforward one.

The statistics, in plain English

Individual patient data is what makes a network meta-analysis of this kind worth more than the sum of the trials: it allows a landmark analysis at 14 days that none of the original trials was designed to report. The efficacy comparisons are all null with intervals that include meaningful harm - DOAC plus P2Y12 inhibitor at 1.16 with an upper bound of 1.41 does not exclude a real increase in ischaemic events, which is exactly what the landmark analysis then localises to the early period. The bleeding reductions are large and precise by contrast. Reading only the one-year efficacy row would have missed the timing signal entirely, which is the point of the paper.

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.