A liberal transfusion target in septic shock with cancer brought no lactate benefit and more thrombosis

Design

Multicentre randomised superiority trial, stopped early for safety

Population

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

Primary outcome

Lactate normalisation or 30% fall at 12 hours

Effect

52% vs 50%; OR 1.07 (0.60–1.90); thrombotic events 13% vs 1%

TRANSPORT, published in Intensive Care Medicine in August 2026, randomised adults with a haematological or solid cancer, septic shock, lactate above 2 mmol/L and haemoglobin below 9 g/dL at 18 French centres. Transfusion was guided by a threshold of 9 g/dL or 7 g/dL for the first 48 hours.

The trial stopped early for safety after 187 of a planned 260 patients. Lactate improvement at 12 hours was similar (52% vs 50%; OR 1.07, 95% CI 0.60–1.90). Arterial or venous thrombotic events were more frequent with liberal transfusion (13% vs 1%). Mortality at 28 days was 46% and 53%, not significantly different.

Early stopping makes all estimates imprecise, and the authors decline to recommend either strategy firmly. But the trial removes the rationale that cancer patients with septic shock need a higher haemoglobin to clear lactate, and it adds a thrombotic signal to the case against liberal transfusion.

• Do not assume a higher haemoglobin target will improve lactate clearance in cancer patients with septic shock.
• Watch for thrombosis when transfusing to higher targets in patients with cancer.
• Use a restrictive threshold as the default unless there is a specific indication, such as active bleeding or myocardial ischaemia.
• Mortality did not differ, so neither target can claim a survival advantage from this trial.

Why it matters

It challenges the instinct to transfuse anaemic cancer patients more liberally in shock.

Don't overread it

A trial stopped early with 187 patients gives imprecise estimates in both directions.

The statistics, in plain English

An odds ratio of 1.07 with an interval from 0.60 to 1.90 means no detectable difference in lactate response. The thrombosis difference was statistically significant, but a trial stopped early can exaggerate effects, so the true size 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.