Old Antibiotics Fight Back as ICU Bloodstream Infections Turn Resistant

In intensive care units around the world, the bacteria that reach the bloodstream are increasingly the kind that modern medicine struggles to kill. A new retrospective study from a single Turkish hospital adds to a growing body of evidence that carbapenem resistance has become the dominant feature of Gram-negative bloodstream infections in critically ill patients, and it asks a question that many clinicians in resource-limited settings face daily: can the older, more toxic antibiotics amikacin and colistin still be relied upon when the newest drugs are out of reach?

The study, published in BMC Infectious Diseases by Sinan Mermer of the Department of Infectious Diseases and Clinical Microbiology and Ertuğrul Çağlayan of the Department of Microbiology at Medical Point Hospital, İzmir University of Economics, examined bloodstream infections caused by Gram-negative bacilli in ICU patients. The researchers focused on two things: how common carbapenem resistance had become among these isolates, and how susceptible the resistant organisms remained to amikacin, an aminoglycoside, and colistin, a polymyxin that many microbiologists regard as a last-resort antibiotic.

The headline finding is stark. Among Gram-negative bloodstream isolates from the ICU, 85.4 percent were carbapenem-resistant. That figure means that in this setting, the great majority of Gram-negative bacteria recovered from the blood had already lost susceptibility to carbapenems, the beta-lactam antibiotics that are often considered the backbone of empirical therapy for serious infections in critically ill patients. Two species dominated the resistant population: Acinetobacter baumannii, with 94.5 percent of isolates resistant, and Klebsiella pneumoniae, with 92.6 percent resistant. Together these two organisms accounted for more than 80 percent of the carbapenem-resistant isolates in the study.

Both organisms are familiar adversaries. Acinetobacter baumannii is a resilient environmental survivor that thrives on hospital surfaces and medical equipment, and it has an extraordinary capacity to accumulate resistance determinants. Klebsiella pneumoniae, a normal inhabitant of the human gut, has become one of the most successful carriers of carbapenemase enzymes worldwide, spreading through hospitals and long-term care facilities with alarming efficiency. When either organism enters the bloodstream, mortality is high, and the window in which effective therapy changes outcomes is narrow, often measured in hours rather than days.

The technical explanation for carbapenem resistance in these organisms lies largely in enzymes. Carbapenemases come in several molecular classes. The Klebsiella pneumoniae carbapenemase, or KPC, is a class A serine beta-lactamase that hydrolyzes carbapenems and most other beta-lactams. Metallo-beta-lactamases, such as NDM enzymes, belong to class B and use a zinc ion in their active site, which makes them intrinsically resistant to inhibition by conventional beta-lactamase inhibitors. Oxacillinases of the OXA family, particularly OXA-23 and OXA-58, are the dominant carbapenemases in Acinetobacter baumannii. Because these enzymes are often encoded on plasmids or integrated into mobile genetic elements, they can move between strains and species, which is why resistance in one ICU isolate can foreshadow resistance in the next patient’s infection.

Against this backdrop, the study’s second question becomes clinically urgent. Amikacin is an aminoglycoside that binds to the bacterial 30S ribosomal subunit, causing misreading of mRNA and halting protein synthesis. It was introduced in the 1970s and was largely sidelined as safer beta-lactams proliferated, but it has retained activity against many carbapenem-resistant organisms because the resistance mechanisms that defeat beta-lactams do not affect ribosomal binding. Colistin works differently: it is a cationic polypeptide that disrupts the outer membrane of Gram-negative bacteria by binding to lipopolysaccharide, displacing calcium and magnesium ions that stabilize the membrane. Its clinical revival over the past two decades has been driven almost entirely by necessity, since it often remains active against organisms resistant to everything else, despite nephrotoxicity and the awkward pharmacokinetics that stem from its development in an era before modern dosing studies.

In the new study, amikacin showed moderate activity, particularly against Klebsiella pneumoniae, while colistin retained some in vitro activity against the resistant isolates. The authors are careful, however, to attach a significant caveat: methodological limitations in susceptibility testing mean these findings should be interpreted with caution. That caution is well grounded. Colistin susceptibility testing is notoriously difficult. The polymyxin molecules adhere to plastic surfaces and diffuse poorly in agar, which makes disk diffusion unreliable, and broth microdilution, the reference method recommended by both EUCAST and CLSI, is technically demanding and not universally available in routine laboratories. Gradient methods and automated systems can misclassify isolates, and the clinical breakpoints themselves have been revised in recent years as pharmacokinetic and pharmacodynamic data accumulated. A laboratory report of colistin susceptibility can therefore be a less certain guide to treatment than clinicians might assume.

The study’s context matters as much as its numbers. In many regions, novel beta-lactam and beta-lactamase inhibitor combinations, such as ceftazidime-avibactam, meropenem-vaborbactam, and imipenem-relebactam, are either unavailable or unaffordable, and agents active against metallo-beta-lactamase producers remain scarce. Where access to these drugs is restricted, clinicians fall back on the older arsenal: colistin-based combination regimens, aminoglycosides, tigecycline, and sometimes high-dose carbapenems in organisms with only moderately elevated minimum inhibitory concentrations. The authors conclude that in settings with a high incidence of carbapenem-resistant Gram-negative bacilli and restricted access to novel agents, amikacin and colistin may be considered in selected cases, guided by local susceptibility data.

That emphasis on local data is the study’s most actionable message. Resistance rates for these organisms vary enormously between hospitals, regions, and even units within the same institution, driven by local clonal spread and differing antibiotic use. A susceptibility rate measured in one ICU cannot be safely extrapolated elsewhere, which is why the authors argue that early appropriate antimicrobial therapy should be guided by local resistance patterns rather than by global guidelines alone. For a clinician facing a septic ICU patient, the difference between an 85 percent resistance rate and a 40 percent one changes the entire calculus of empirical therapy, from the choice of initial regimen to the urgency of escalation once culture results return.

The findings also carry a broader warning for antimicrobial stewardship. Reliance on colistin and aminoglycosides is not a stable solution. Colistin resistance in Klebsiella and other Gram-negatives, typically mediated by modifications to lipopolysaccharide that reduce the drug’s binding target, has been documented worldwide and can emerge during therapy. Aminoglycoside resistance arises through modifying enzymes and ribosomal mutations and can be selected rapidly under drug pressure. Every course of a last-resort antibiotic consumes a finite resource, and the study’s authors implicitly frame their results accordingly: these older drugs retain value, but that value depends on surveillance, susceptibility testing of adequate quality, and stewardship programs that prevent indiscriminate use. As carbapenem resistance in ICU bloodstream infections approaches saturation levels, the margin between effective therapy and therapeutic failure narrows to the precision of the laboratory and the speed of the clinician, and studies like this one document exactly how thin that margin has become.

Subject of Research: Carbapenem resistance and susceptibility to amikacin and colistin among Gram-negative bloodstream isolates in ICU patients

Article Title: Do amikacin and colistin retain activity against carbapenem-resistant gram-negative bloodstream isolates in a single-center ICU?

Article References: Mermer, S., & Çağlayan, E. (2026). Do amikacin and colistin retain activity against carbapenem-resistant gram-negative bloodstream isolates in a single-center ICU?. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-13988-y

Image Credits: AI Generated

DOI: 10.1186/s12879-026-13988-y

Keywords: carbapenem-resistant Gram-negative bacilli, bloodstream infection, intensive care unit, amikacin, colistin, antimicrobial resistance, Acinetobacter baumannii, Klebsiella pneumoniae, antimicrobial stewardship, susceptibility testing, EUCAST, surveillance