Temocillin matched carbapenems for resistant blood infections in a trial

Temocillin, a narrow-spectrum β-lactam antibiotic that fell out of use for decades, worked as well as carbapenems for treating bloodstream infections caused by drug-resistant gut bacteria in a Spanish randomised trial [s1]. The result, from the 328-patient ASTARTÉ trial published in The Lancet, matters because it gives hospitals a way to treat these infections without reaching for a carbapenem — the broad-spectrum class whose overuse drives the next, harder-to-treat tier of resistance [s1].

The bacteria in question are third-generation cephalosporin-resistant Enterobacterales, the family that includes Escherichia coli and Klebsiella pneumoniae. When these organisms enter the blood, clinicians have leaned on carbapenems as the reliable option, and that habit has a cost: carbapenem-resistant Enterobacterales sit in the "critical" tier of the World Health Organization's 2024 bacterial priority pathogens list, the group for which new treatments are most urgently needed [s2]. Every course of a carbapenem given for an infection that a narrower drug could clear adds to the selection pressure producing those critical-priority organisms.

What the trial did

ASTARTÉ was an investigator-initiated, phase 3, open-label, non-inferiority trial run at 29 Spanish hospitals [s1]. Adults with a single-organism bloodstream infection caused by a cephalosporin-resistant Enterobacterales that was also susceptible to both study drugs were randomly assigned to intravenous temocillin (2 g every 8 hours) or a carbapenem — meropenem (1 g every 8 hours) or, where appropriate, ertapenem (1 g per day) [s1].

Between December 2020 and November 2024, 334 patients were enrolled, of whom 328 formed the analysis population: 163 assigned temocillin and 165 assigned carbapenems [s1]. This was an older, comorbid group, with a median age of 72 years [s1].

The primary endpoint was clinical success by day 28 — a composite requiring cure, no need to stop or switch the study drug for an adverse event or failure, no recurrence, and survival [s1]. Clinical success occurred in 120 of 163 patients on temocillin (74%) and 121 of 165 on carbapenems (73%), a difference of 0.3 percentage points [s1]. Because the lower bound of the confidence interval around that difference (95% CI −7.7 to infinity) stayed above the prespecified non-inferiority margin of 10 percentage points, temocillin was declared non-inferior (non-inferiority p=0.017) [s1]. Serious adverse events were reported in 31 temocillin patients (19%) and 35 carbapenem patients (24%) [s1].

How to read a non-inferiority result

"Non-inferior" is not "better," and the design decides what the number means. A non-inferiority trial asks whether a new option is not unacceptably worse than the standard by a margin agreed in advance — here, 10 percentage points [s1]. Temocillin cleared that bar comfortably, with an observed difference near zero, so the trial supports treating it as a genuine alternative rather than as a proven equal down to the last point [s1].

Two design features bound the conclusion. The trial was open-label, so patients and treating doctors knew which drug was given, which can influence subjective decisions such as whether to switch therapy — a component of the primary endpoint [s1]. And enrolment required the infecting organism to be susceptible to temocillin in the laboratory; the trial speaks to targeted therapy once susceptibility is confirmed, not to empirical treatment started before culture results are back [s1].

Why a "neglected" antibiotic matters here

Temocillin's appeal is precisely that it is narrow. It is stable against many of the enzymes that render Enterobacterales resistant to cephalosporins, yet it lacks the broad reach that makes carbapenems so disruptive to a patient's microbiome and to hospital ecology [s1]. Using a carbapenem-sparing drug for infections a narrower agent can treat is the core logic of antimicrobial stewardship, and it is the mechanism by which the critical-priority resistance the WHO tracks is meant to be slowed [s1][s2]. ASTARTÉ provides the kind of head-to-head evidence that has been missing for this specific, common scenario. Related coverage has examined why antibiotics do nothing for colds and how that misuse feeds resistance and how surveillance networks track resistant organisms in practice.

What to watch

The trial was conducted in one country's hospital system, in patients whose isolates were confirmed susceptible, so uptake elsewhere will depend on local resistance patterns and on laboratories being able to test temocillin susceptibility reliably [s1]. The larger question is whether stewardship programmes and guideline committees now fold temocillin into recommended pathways for cephalosporin-resistant bloodstream infections — the step that would turn a single trial's result into fewer carbapenem prescriptions [s1][s2].

This article describes research and regulatory context and is not medical advice. Antibiotic choice for a bloodstream infection is a decision for treating clinicians based on culture and susceptibility results.

Sources

• Temocillin versus carbapenems for bacteraemia due to third-generation cephalosporin-resistant Enterobacterales in Spain (ASTARTÉ): a multicentre, phase 3, open-label, non-inferiority, randomised clinical trial — The Lancet, 9 July 2026
• WHO bacterial priority pathogens list, 2024 — World Health Organization, 17 May 2024