Response-tailored antibiotics for stable left-sided endocarditis: two weeks shorter, safe on the composite, more relapse

Design

International, randomised, open-label trial

Population

508 stable adults with left-sided IE (S. aureus, E. faecalis, streptococci)

Primary outcome

Days alive without antibiotics at 6 months; death, surgery or embolism

Effect

183 vs 169 days (difference 13); safety 8.2% vs 10.7%; relapse 5.1% vs 1.6%

POET II was an international open-label trial of 508 stable adults with left-sided infective endocarditis caused by S. aureus, Enterococcus faecalis or streptococci. All had received at least 2 to 4 weeks of antibiotics and met clinical stabilisation criteria before randomisation to stop treatment, or to continue to a standard 4 to 6 weeks.

Tailored therapy gave a median 183 against 169 antibiotic-free days over 6 months (difference 13 days). Death, unplanned cardiac surgery or symptomatic embolism occurred in 8.2% against 10.7%, meeting non-inferiority. But relapse of bacteraemia or endocarditis was 5.1% against 1.6%.

The standard 4 to 6 weeks is mostly consensus, and this trial shows it can be shortened in selected stable patients without more deaths or emboli. The relapse signal is the price, and it is not trivial. A shorter course is reasonable when stabilisation criteria are clearly met and follow-up with repeat cultures is assured; it is not a blanket change.

• Consider stopping at 2 to 4 weeks only in stable patients meeting clear response criteria
• Confirm cleared blood cultures, falling inflammatory markers and no uncontrolled focus first
• Arrange repeat blood cultures and review after stopping
• Tell patients to return promptly with fever
• Discuss the relapse risk openly and record the decision

Why it matters

Six weeks of antibiotics is consensus, not evidence, and this trial shows where it can safely be shortened.

Don't overread it

Relapse was a secondary endpoint and more frequent with shorter therapy; open-label design and selected stable patients limit how widely this applies.

The statistics, in plain English

Non-inferiority means the safety composite was not worse by more than 7.5 percentage points. The interval for the difference (−7.7 to 2.7) sits within that margin. But relapse, 13 against 4 patients, was significantly higher (p = 0.04). About one extra relapse for every 29 patients treated with the shorter strategy is the trade to weigh.

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.