Liberal transfusion in cancer patients with septic shock: no gain, more clots, trial stopped

Design

Multicentre randomised superiority trial, 18 French centres, stopped early

Population

187 adults with cancer, septic shock, lactate >2 mmol/L and haemoglobin <9 g/dL

Primary outcome

Lactate reduction at 12 hours

Effect

52% vs 50% (OR 1.07, 95% CI 0.60–1.90); thrombosis 13% vs 1%; 28-day mortality 46% vs 53% (NS)

TRANSPORT was a randomised trial in 18 French centres, published in August 2026. Adults with a haematological or solid cancer, septic shock, lactate above 2 mmol/L and haemoglobin below 9 g/dL were assigned to transfusion at a threshold of 9 g/dL (liberal) or 7 g/dL (restrictive) for the first 48 hours of resuscitation. It was planned for 260 patients but stopped at 187 for safety concerns.

The primary outcome, lactate clearance at 12 hours, was 52% with liberal and 50% with restrictive transfusion (OR 1.07, 95% CI 0.60–1.90). Arterial or venous thrombotic events occurred in 13% against 1%. Mortality at 28 days was 46% against 53%, a difference that was not statistically significant.

Cancer patients are often anaemic and are sometimes transfused more freely in shock on the reasoning that oxygen delivery matters more. This trial gives no support for that, and a clear safety signal against it. The authors stop short of preferring either strategy given the early stop and the mortality numbers. In practice, it supports keeping the 7 g/dL threshold used in general septic shock for these patients too, outside specific situations such as active bleeding or myocardial ischaemia.

• Use a haemoglobin threshold of 7 g/dL for transfusion in septic shock, including in patients with cancer.
• Do not transfuse to improve lactate clearance; the liberal threshold did not change it.
• Watch for venous and arterial thrombosis in cancer patients transfused in shock.
• Individualise for active bleeding, acute coronary ischaemia or severe thrombocytopenia with bleeding risk.
• Conserve blood where supply is limited; this trial gives no reason to transfuse more.

Why it matters

It removes the main argument for transfusing anaemic cancer patients more freely in shock.

Don't overread it

The trial stopped early and the mortality difference was not significant in either direction.

The statistics, in plain English

An odds ratio of 1.07 for lactate clearance with an interval of 0.60 to 1.90 means no detectable difference. The 13% against 1% thrombosis rate was statistically significant (p=0.001), but because the trial stopped early, the exact size of that difference is uncertain.

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.