Case Study 25.1 — The Drug That Came Back

Colistin's abandonment and return, and what a returning drug tells you about the state of a field

The situation

Some drugs are withdrawn because they stop working. Some are withdrawn because they were never as good as they looked. Colistin was withdrawn for neither reason. It worked, it kept working, and clinicians stopped using it anyway — because something better arrived and colistin's toxicity was no longer worth accepting.

Then, decades later, they started using it again. Nothing about the molecule had changed. Nothing about its toxicity had changed. What changed was the alternative.

This case study asks you to reason about a drug whose value moved without its properties moving, because that pattern is one of the most under-taught ideas in pharmacology and one of the most useful for reading claims about any therapy.

Part 1 — The molecule and its first career

Colistin, also called polymyxin E, is a cyclic lipopeptide produced by a soil bacterium and described in the late 1940s. It is a peptide by composition, though not a ribosomally synthesized one: it is assembled by bacterial enzymes and contains non-standard amino acids and a fatty acid tail. Like the host defense peptides of §25.2, it is cationic and amphipathic, and it kills by the mechanism of §25.3.

Its target is the Gram-negative outer membrane. The positively charged peptide binds the negatively charged lipid A of lipopolysaccharide, displaces the divalent cations that bridge and stabilize adjacent LPS molecules, and permeabilizes the outer membrane — then reaches and disrupts the inner membrane as well. It is one of the few agents that can breach the Gram-negative outer membrane, which is a formidable permeability barrier and the reason many antibiotics that work beautifully on Gram-positives are useless against Gram-negatives.

Colistin entered clinical use around 1960. It was effective. It was also nephrotoxic and neurotoxic — kidney injury and neurological effects including neuromuscular blockade were recognized early and were not rare.

Through the 1970s and 1980s, newer agents — aminoglycosides, then expanded-spectrum beta-lactams — offered comparable Gram-negative coverage with better tolerability. Colistin's systemic use fell away almost entirely in high-income settings. It survived in topical preparations and in a few niche applications, but as an intravenous drug it was, for practical purposes, retired.

Consider what that decision was. It was not a judgment that colistin did not work. It was a comparison, and colistin lost the comparison on tolerability while winning nothing on efficacy that the alternatives did not also offer.

Part 2 — What changed

Beginning in the 1990s and accelerating through the 2000s, three organism groups became progressively harder to treat: carbapenem-resistant Enterobacterales including Klebsiella pneumoniae, multidrug-resistant Acinetobacter baumannii, and resistant Pseudomonas aeruginosa.

Carbapenems had been the reliable last resort for serious Gram-negative infection. As carbapenem-resistant isolates spread, clinicians in some settings began encountering infections against which the modern armamentarium had nothing to offer.

They went back to the shelf.

Colistin returned to systemic use as a last-line agent against multidrug-resistant Gram-negative organisms. It is administered as an inactive prodrug converted to the active peptide in the body, it is used under specialist supervision with monitoring of renal function, and it retains every one of the toxicities that caused its abandonment.

The comparison had changed, and only the comparison. In 1980, the alternative to colistin was an aminoglycoside. In 2015, the alternative in some cases was an untreated infection with high mortality.

Part 3 — And then resistance

In 2015, surveillance of isolates from food animals, retail meat, and hospitalized patients in China identified a plasmid-borne gene, mcr-1, encoding an enzyme that adds phosphoethanolamine to lipid A. The modification reduces the outer membrane's negative charge and weakens the electrostatic attraction colistin depends on.

Chromosomal colistin resistance had been described before. The significance of mcr was that it sat on a plasmid, which meant it could move horizontally between bacteria — including between species — rather than being confined to a single lineage. Within roughly a year of the report, mcr-1 had been identified in isolates from many countries across several continents.

This did not end colistin's usefulness, and it has not. But it is the observation that most directly falsifies the strong version of the AMP resistance argument, and it did so on the most clinically consequential membrane-active peptide in use.

Part 4 — The rating

The chapter rates colistin as follows:

Claim: Colistin is an effective treatment option for infections caused by multidrug-resistant Gram-negative organisms when other agents are unavailable or inactive. Rating: ✅ Strong clinical evidence Reason: Approved, in guideline-directed clinical use for decades, with well-characterized activity against organisms including carbapenem-resistant Enterobacterales, Acinetobacter, and Pseudomonas. What would change it: Widespread dissemination of transferable resistance to the point of routine clinical failure; or head-to-head trials showing newer agents are superior for the same organisms, which would narrow rather than eliminate the claim.

A ✅ on a drug with documented nephrotoxicity and neurotoxicity is not a slip. It is the system working as designed. ✅ means the evidence supports the claim as stated. It does not mean the drug is pleasant, well tolerated, or anyone's first choice.

Discussion questions

1. Colistin's properties did not change between 1980 and 2015, but its place in therapy reversed completely. Articulate precisely what did change, and express the change as a comparison rather than as a property of the drug. Then name one other area of medicine where the same reasoning governs treatment choice.

2. The chapter rates colistin ✅ while describing serious organ toxicity in the same section. Explain to someone who has never used this rating system why that is coherent. Then write the sentence you would add to a consumer-facing summary to prevent ✅ from being read as "safe."

3. Suppose a new, well-tolerated agent is approved with reliable activity against carbapenem-resistant Enterobacterales. Using the chapter's rules for what would change a rating, state exactly how colistin's ✅ should be revised — and be careful about whether the claim is eliminated or merely narrowed.

4. mcr demonstrated that resistance to a cationic membrane-active peptide can be plasmid-borne and internationally mobile. Does this falsify the argument in §25.3 that membrane disruption makes resistance harder to evolve? Distinguish between the claim it does falsify and the claim it does not, and state which claim the field is entitled to make going forward.

5. Colistin's revival is described in the chapter as "a striking and slightly grim demonstration" and "a report on the state of §25.1." Make the case that reviving an abandoned drug is a success of clinical pharmacology. Then make the case that it is a failure of the drug development system. Which argument is stronger, and what evidence would settle it?

6. A hospital newsletter describes colistin as "a natural peptide antibiotic — nature's own answer to superbugs." Identify every misleading implication in that phrase, using specific sections of Chapter 25. Then rewrite the sentence so that it is accurate and no longer than the original.