Patients with pulmonary embolism who are haemodynamically stable but have right ventricular strain and raised cardiac biomarkers sit in an uncomfortable middle ground. They are too well for systemic thrombolysis, which carries a real bleeding cost, and some of them deteriorate anyway. PRAGUE-26 randomised 558 such patients — stable, simplified Pulmonary Embolism Severity Index of 1 or more, right ventricular dysfunction plus raised troponin or natriuretic peptide — to catheter-directed alteplase with anticoagulation, or anticoagulation alone. Median age was 64 and 40.9% were women.
The seven-day composite of death from any cause, recurrent pulmonary embolism, or cardiorespiratory decompensation or collapse occurred in 2 patients (0.7%) with thrombolysis and 19 (6.8%) with anticoagulation alone: relative risk 0.10 (95% CI 0.02 to 0.44, p<0.001). Decompensation drove the difference. Bleeding did not differ — clinically relevant bleeding 4.6% against 5.0%, major bleeding 1.4% against 2.2% — but intracranial haemorrhage occurred in 2 patients on thrombolysis and none on anticoagulation.
That is a large effect on a soft-ish endpoint with a small but non-zero catastrophic risk on the other side of the ledger. Cardiorespiratory decompensation is a real event, not a surrogate, but it is also the component most sensitive to how an unblinded clinician responds to a deteriorating patient. Two intracranial bleeds in 280 patients is 0.7% — low, and permanent when it happens.
What this changes in practice depends heavily on where you work. In a centre with an interventional service and a pulmonary embolism response team, this strengthens the case for intervening rather than watching. In a hospital without one, the message is about triage: identify these patients deliberately, and decide early whether transfer is feasible, rather than discovering the question at 3 am when the patient crashes. Across much of India that decision has to be made against real transfer times and cost, which makes recognising the risk category early the part that is actually within your control.
- Identify intermediate-high-risk PE deliberately: stable, sPESI at least 1, right ventricular dysfunction AND raised troponin or natriuretic peptide.
- Where catheter-directed thrombolysis is available, discuss it early rather than after deterioration.
- Where it is not, decide on transfer at the point of diagnosis, not at the point of collapse.
- Weigh the roughly 0.7% intracranial haemorrhage risk explicitly in the consent conversation.
- Anticoagulation alone remains correct for low-risk and most intermediate-low-risk PE.
The statistics, in plain English
A relative risk of 0.10 means a 90% reduction, but the interval of 0.02 to 0.44 is wide because the whole result rests on 21 events. The absolute figures are the ones to carry: 6.8% against 0.7%, so about 16 patients treated to prevent one event. Set against that, two intracranial haemorrhages in 280 treated patients — an event the trial was far too small to quantify properly. The composite endpoint was dominated by cardiorespiratory decompensation rather than death, and in an open-label trial the decision to call a patient 'decompensated' is made by clinicians who know the allocation.
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.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for top clinical updates, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free