Hacker News Reader

> There is no reason to believe we've hit a wall and can't invent more antibiotics in the future

That just kicks the can down the road though. Assuming that the use of any antibiotic will eventually lead to a predominance of resistance, we would have to continue to invent new, effective antibiotics indefinitely.

> For example, some countries crazily overprescribe antibiotics, or prescribe broad spectrum ones

I don't think its as simple as over prescription being the problem. My grandmother is pretty I'll these days, she suffers from dementia, is sick frequently, and generally just very warn down. She had a respiratory infection this week and while at the doctor they also found that she has a UTI. The prescribed her two different, non-broad spectrum antibiotics.

I don't think anyone would consider that over prescribing, and she likely would have a very bad outcome without it, but it also seems totally reasonable that her scenario could still lead to antibiotic resistance. Her immune system just doesn't work well anymore, the antibiotics will take out most of the infection and her symptoms will likely go away but I wouldn't expect her body to do the job of cleaning up the rest of the infection, leaving the more resistant strains around in her system.

We don’t have to keep creating new antibiotics indefinitely. There’s always some cost to resistance. Once a drug is no longer present in an environment, non-resistant strains tend to outcompete resistant ones. That is if you just stop using an antibiotic, bacteria will tend to lose resistance to it.

Theoretically once you have enough antibiotics that we can retire drugs with heavy resistance and keep cycling them, you won’t need to keep creating new antibiotics.

There's a lot baked into that theory (well, hypothesis) though. We haven't gone through a cycle like that yet that I'm aware of, moving back to an old antibiotic because the newer one is no longer effective.

Looking just at the evolution of it, there would need to be pressure to actively select against the learned resistance. Maybe it would be lost eventually, but we couldn't rely on that unless something pushed the bacteria away from it rather learning to resist the new antibiotic in addition to the old one.

This has been well researched. Most antibiotic resistance has a fitness cost. Which means the resistance is being actively selected against.

This isn’t always the case. There are some adaptations that don’t have an observable fitness cost, but the majority do. That is, in a lab when you remove the antibiotic, we observed that the number of resistant bacteria drops over time.

We have also observed this in the real world. When we reduce usage of a specific antibiotic. The percentage of resistant bacteria in the wild drops.

The question is how long you’d have to retire an antibiotic and how many different antibiotics you’d need for this strategy to be viable.

Sure, but I think we agree then. What you describe is a hypothesis based on a few important assumptions. For example, that controlled lab studies are predictive analogs for the natural world, that a drop in resistance in such a controlled study will match a real world scenario, and that the rate at which resistance is lost can reasonably be matched by cycling in different drugs.

I'm not saying it isn't possible, maybe it is! Only that its a hypothesis and that the key assumptions are still just that rather than known parameters.

We have more than lab studies, we have real world observations that show the percentage of resistant strains dropping when we reduce the use of a specific antibiotic.

We may not be able to produce new antibiotics fast enough to get to the point where we have enough to cycle them effectively. That’s entirely possible.

But we have very good evidence that there is a point at which we would have enough. I mean at the limit there is a maximum amount of information bacteria can store in their genes, so there certainly exists some maximum number of resistances they can retain.

My point isn’t that we’re going to have enough to do this in x years. My point is that while we don’t know how many different antibiotics we need, we are almost certain that that number isn’t infinite.

Everything that reproduces is in an evolutionary arms race (https://en.wikipedia.org/wiki/Evolutionary_arms_race), so the "kicking the can" objection doesn't really make sense in this context.

There is no alternative to inventing new weapons continuously. We are simply used to winning without much effort since Fleming found penicillin, a naturally occurring fungal compound. Previous successes have limited the economics of continuing to develop new solutions, which is contributes to the uptick in resistance.

This doesn't mean we shouldn't be judicious in the tradeoffs that we choose and try to avoid a purely chemical solution to the arms race. For example, as robotics get better, we should move away from mono-cropping toward combinations of plants that naturally ward off pests. There are better ways to stay ahead than blunt tools.

>There is limited evidence that farming use of antibiotics is what drives resistance in clinical cases.

This is a classic example of asymptotic behaviour analysis.

>That is - we use enormous amounts of antibiotics in farming

You're entirely correct, the data fits an almost exponential curve, the only limiting factor so far has been soil health and costs. the amount of biocides we have to use to keep our productivity will be impossible to match and therefore, even with 0 transference of resistance to human antibiotics, you're gonna see more transference to humans because of farming-biocide resistance. ofcourse, most of this will result in lower crop yield instead of actually selling bad product, this can be slowed down with essentially the same policies and techniques as hospitals employ. But the core problem is the exact same, the extremely quick and deadly takeover of a single strain resistent against our current quo pro due to overuse of a single line of defence.

I agree that there is still a lot we can do to combat fungal disease.

However, your points 1, 2, and 4 unfortunately don't apply well to anti-fungals.

* "There is limited evidence that farming use of [anti-fungals] is what drives resistance in clinical cases."

Certainly not the case with respect to azoles:

-- Rhodes J, et al. Population genomics confirms acquisition of drug-resistant Aspergillus fumigatus infection by humans from the environment. Nat Microbiol. 2022 May;7(5):663-674. doi: 10.1038/s41564-022-01091-2. Epub 2022 Apr 25. Erratum in: Nat Microbiol. 2022 Nov;7(11):1944. doi: 10.1038/s41564-022-01160-6. PMID: 35469019; PMCID: PMC9064804. https://pubmed.ncbi.nlm.nih.gov/35469019/

-- Celia-Sanchez BN, Mangum B, Gómez Londoño LF, Wang C, Shuman B, Brewer MT, Momany M. Pan-azole- and multi-fungicide-resistant Aspergillus fumigatus is widespread in the United States. Appl Environ Microbiol. 2024 Apr 17;90(4):e0178223. doi: 10.1128/aem.01782-23. Epub 2024 Apr 1. PMID: 38557086; PMCID: PMC11022549.

-- Impact of the use of azole fungicides, other than as human medicines, on the development of azole‐resistant Aspergillus spp. https://www.efsa.europa.eu/en/efsajournal/pub/9200

"* There is no reason to believe we've hit a wall and can't invent more antibiotics in the future. "

-- There are very few anti-fungals, because fungi, being eukaryotes, have cell biology so close to ours. The "big three" are 1) azoles (lots of resistance; see above); 2) echinocandins; and 3) amphotericin B (highly toxic to the host)

-- One bright spot is a new antifungal published a few weeks ago in Nature, called Mandimycin: https://www.nature.com/articles/d41586-025-00801-0

* "There are still some antimicrobial treatment techniques we haven't put a lot of resources in. Like, bacteriophage treatments, or cycling antibiotics."

-- Bacteriophage don't work on fungi. Some degree of cycling and combination anti-fungal treatment are already in use.