Can antibiotics save each other?

Please read the original article at: https://doi.org/10.1038/s41467-024-49122-2

Disclaimer: Below is an interpretation of the article thus do not necessarily represent the views of the authors.

Can antibiotics save each other?

Maybe.

Collateral sensitivity (CS) is when resistance to an antibiotic makes a bacteria more vulnerable to another antibiotic. We could take advantage of this by making drug combinations using these CS-pairs. However, most known CS are caused by mutations that can't be transferred between bacteria via plasmids. In fact, the scariest antimicrobial resistance genes (ARGs), are on plasmids. Among them is blaOXA-48, which inactivate carbapenems, drugs we use when nothing else works.

This study by Herencias et al. discovered a CS link in E. coli between blaOXA-48 and susceptibility to azithromycin and colistin. Azithromycin, a macrolide, is often used for many infections, and colistin is another drug-of-last-resort. They showed that this link is causal and broadly consistent among penicillinases and important E. coli strains, though the strength of the effect is variably weaker.

So give carbapenems and azithromycin or colistin together, right?

It's never that simple, sadly.

If a bacteria already has macrolide or colistin resistance, does CS weaken the resistance? Against susceptible bacteria, it could reduce drug dosages and side-effect risk. It could slow resistance development too, since a pathogen has to be resistant to both to survive. But, if the CS effect is too small to reduce resistance significantly, it won't make untreatable patients treatable.

In the case of colistin, increasing sensitivity to a drug-of-last-resort using another drug-of-last-resort is counter-productive. Misuse would accelerate resistance to both critical antibiotics and it can be even more prohibitively expensive. Especially when the greatest burden is in less affluent countries.

These are why, I think, the most important finding was that other penicillinases have a similar CS effect as blaOXA-48.

ARGs against penicillins are widespread, so using them in combination with azithromycin or colistin may not work in patients. If the mechanism of how this CS effect is figured out, the same effect could be produced using small molecules, or plasmids containing CS-inducing genes could be spread.