Educational information only. Nothing on this site is medical advice, and no dose mentioned here is a recommendation. Speak to a prescriber who knows your history.

Colistin (colistimethate sodium)

polymyxin E, colistimethate sodium, colistin sulfate, CMS, colistin methanesulfonate

Colistin, also called polymyxin E, is a cyclic lipopeptide antibiotic discovered in 1949 that kills Gram-negative bacteria by disrupting their outer membrane. It is almost always given as colistimethate sodium, an inactive prodrug that converts to colistin in the body. Like polymyxin B it was sidelined for decades over toxicity concerns and returned as a last-resort treatment for highly resistant organisms.

High-quality evidence Antimicrobial & antibiotic Reviewed 2026-09-04

Mechanism

Colistin works by the same route as polymyxin B. Its five cationic diaminobutyric acid side chains bind the anionic phosphate groups of lipid A in Gram-negative lipopolysaccharide, displacing the stabilising calcium and magnesium bridges. Hydrophobic insertion of the fatty acyl tail and the D-Leu6-Leu7 region then destabilises the outer membrane, permitting self-promoted uptake and subsequent inner membrane disruption, leakage of cytoplasmic contents and death. Additional contributions from hydroxyl radical generation and inhibition of the type II NADH-quinone oxidoreductase have been proposed but remain secondary and less well established.

The pharmacology is dominated by the prodrug problem. Colistimethate sodium is not itself antibacterial. After administration it hydrolyses slowly and incompletely to colistin, while a large fraction is cleared unchanged by the kidney before conversion occurs. Early exposure to active drug is therefore low and highly variable between patients, which is the reason regulators now insist on a loading dose in critically ill patients and on expressing doses in international units to reduce the risk of dosing errors between differently labelled products. Resistance mirrors that of polymyxin B: lipid A modification with 4-amino-4-deoxy-L-arabinose or phosphoethanolamine via PmrAB and PhoPQ signalling, and horizontally transmissible mcr genes.

What the research shows

The most important modern trial is AIDA (Paul 2018), an open-label randomised controlled trial in 406 patients with severe infection caused by carbapenem-resistant Gram-negative bacteria, predominantly Acinetobacter baumannii. Colistin monotherapy was compared with colistin plus meropenem. There was no significant difference in clinical failure between arms, and no benefit from combination therapy in the Acinetobacter subgroup, which formed the majority of the population. Mortality was high in both groups, a reminder of how sick this population is.

The OVERCOME trial subsequently addressed the same question with a double-blind design in 423 randomised patients with pneumonia or bloodstream infection, again mostly Acinetobacter baumannii, and reached a consistent conclusion: adding meropenem to colistin did not improve mortality (43% against 37%, p=0.17) or clinical outcomes. Taken together these two trials substantially settled a question that had been argued for a decade on the basis of laboratory synergy data.

The EMA Article 31 referral, completed in October 2014, is a regulatory rather than a research output but is arguably as consequential. Reviewing the pharmacokinetic, efficacy and safety data, the CHMP concluded that dosing should be expressed in international units, that a conversion table was required in product information, that critically ill patients should receive a loading dose, and that intravenous colistimethate should be reserved for serious infections in patients with limited options and combined with another suitable antibiotic where possible.

A 2025 secondary analysis of the AIDA cohort (Nutman and colleagues) examined acquisition and molecular characterisation of carbapenem-resistant Enterobacterales during colistin monotherapy versus combination therapy, addressing the ecological consequences of each strategy.

Evidence assessment

High-quality evidence

Confirmed as strong on audit, with a caveat about what the strength consists of. Colistimethate holds current approved labelling in the United States, the United Kingdom and the European Union, including a nebulised indication for chronic Pseudomonas aeruginosa infection in cystic fibrosis, and unlike polymyxin B it has genuine randomised trial data: AIDA (PMID 29456043, 406 patients) and OVERCOME (PMID 37538951, 423 patients) both verified in this pass, along with the published AIDA protocol and a 2025 secondary microbiological analysis. What those trials established, however, is a null result: adding meropenem does not help. Mortality was high in both arms of both trials. The tier reflects the quality and quantity of the evidence, not a favourable efficacy signal.

Tiers are applied consistently across the library and re-checked when new trials read out. Read the grading method.

Key studies

Colistin alone versus colistin plus meropenem for treatment of severe infections caused by carbapenem-resistant Gram-negative bacteria: an open-label, randomised controlled trial Preclinical only

Paul M, Daikos GL, Durante-Mangoni E, Yahav D, Carmeli Y, et al. · The Lancet Infectious Diseases · 2018

Multicentre open-label randomised controlled trial; 406 patients with severe carbapenem-resistant Gram-negative infection

Combination therapy was not superior to colistin monotherapy. Adding meropenem did not reduce clinical failure overall or in the Acinetobacter baumannii subgroup that made up most of the cohort. Mortality was high in both arms.

Multicentre open-label randomised controlled trial to compare colistin alone with colistin plus meropenem for the treatment of severe infections caused by carbapenem-resistant Gram-negative infections (AIDA): a study protocol Preclinical only

Dickstein Y, Leibovici L, Yahav D, Eliakim-Raz N, Daikos GL, et al. · BMJ Open · 2016

Published trial protocol

Sets out the design, power calculation and pre-specified analyses for AIDA, including the definition of clinical failure and the loading-dose colistin regimen used, which is useful for judging how far the null result can be trusted.

Colistin Monotherapy versus Combination Therapy for Carbapenem-Resistant Organisms Preclinical only

Kaye KS, Marchaim D, Thamlikitkul V, Carmeli Y, Chiu CH, et al. · NEJM Evidence · 2023

Multicentre double-blind randomised controlled trial (OVERCOME); 423 randomised patients with pneumonia or bloodstream infection caused by extensively drug-resistant organisms, predominantly Acinetobacter baumannii

Adding meropenem to colistin did not improve mortality (43% versus 37%, p=0.17) or clinical outcomes in carbapenem-resistant Gram-negative infection, independently confirming the AIDA result under blinded conditions.

Polymyxin-containing medicines: Article 31 referral, European Medicines Agency review of colistin and colistimethate sodium Preclinical only

2014

Regulatory safety and efficacy review by the EMA Committee for Medicinal Products for Human Use

Concluded that doses must be expressed in international units with a conversion table in product information, that critically ill patients require a loading dose to reach effective concentrations promptly, and that intravenous colistimethate should be reserved for serious infections with limited alternatives, combined with another agent where possible.

Carbapenem-resistant Enterobacterales (CRE) acquisition and molecular characterization following colistin monotherapy and colistin-meropenem combination therapy: findings from the AIDA randomized trial Preclinical only

Nutman A, Temkin E, Lellouche J, Amar Ben Dalak M, Kaplan E, et al.; AIDA Study Group · Antimicrobial Resistance and Infection Control · 2025

Secondary microbiological analysis of the AIDA randomised trial cohort (volume 14, article 133)

Examined acquisition and molecular characterisation of carbapenem-resistant Enterobacterales during and after each treatment strategy, addressing whether combination therapy carries an additional ecological cost.

Safety

Nephrotoxicity is dose-related and common, driven by proximal tubular accumulation of colistin. Neurotoxicity presents as perioral and peripheral paraesthesiae, dizziness, ataxia, weakness and visual disturbance. Neuromuscular blockade with respiratory arrest is rare but potentially fatal, and the risk rises with concurrent anaesthetic agents, aminoglycosides or neuromuscular disease. Inhaled colistimethate causes cough, sore throat and bronchoconstriction, and there have been fatalities where colistimethate was reconstituted and stored before nebulisation, allowing premature hydrolysis to colistin. Hypersensitivity occurs. Renal function requires close monitoring, and dosing must be adjusted for renal impairment.

Regulatory status

Status summary. Regulation changes-verify against the current regulator position before relying on this.
JurisdictionStatus
United KingdomColistimethate sodium is a licensed prescription-only medicine, available as a powder for injection or infusion and as nebulised and dry-powder inhaled formulations for chronic Pseudomonas aeruginosa lung infection in cystic fibrosis. Following the EMA Article 31 review completed in October 2014, UK and EU product information expresses doses in international units, includes a conversion table, and recommends a loading dose in critically ill patients. Intravenous use is reserved for serious infections where treatment options are limited, preferably with a second active agent.
United StatesColistimethate sodium for injection is an approved prescription medicine, a legacy registration for serious Gram-negative infections. Labelling carries warnings for nephrotoxicity, neurotoxicity and neuromuscular blockade. Dosing conventions differ from those used in Europe, which has been a recognised source of medication error.
WADA (sport)Not prohibited. Colistin and colistimethate do not appear on the WADA Prohibited List in or out of competition.

Questions

Colistin sulfate is markedly more toxic when given by injection. Sulfomethylating the five free amine groups produces colistimethate, which is far better tolerated intravenously because the masked positive charges reduce direct membrane interaction with human tissue. The trade-off is that colistimethate has no antibacterial activity of its own and must hydrolyse back to colistin in the body, a process that is slow, incomplete and variable. This is why regulators now require a loading dose in critically ill patients.

On the best available evidence, no. Laboratory synergy studies suggested it should, and this drove widespread combination use for years. Two randomised trials, AIDA in 2018 (406 patients, open-label) and OVERCOME in 2023 (423 patients, double-blind), both found that adding meropenem to colistin did not improve clinical failure or mortality in carbapenem-resistant infection. This is a good example of laboratory synergy failing to translate into clinical benefit.

mcr-1 was the first plasmid-borne colistin resistance gene identified, reported in 2015. It encodes a phosphoethanolamine transferase that modifies lipid A and reduces its negative charge, so colistin binds less well. Chromosomal colistin resistance had been known for a long time, but because mcr genes sit on plasmids they can move horizontally between bacteria and between species. That converts colistin resistance from a strain-specific problem into a transmissible one, which is why colistin use in agriculture has been curtailed in many countries.