I am not an expert on the subject, but isn't it just finding enough solutions to cycle through? As the bacteria become resistant use another solution that works and let the resistant bacteria die out as there becomes no reason to have the resistant traits anymore, thus making them vulnerable to the drug again, or maybe I am just completely wrong.
The problem is that once you have a resistance gene against a given treatment in the bacterial population, it is very unlikely to be completely eliminated, even after you stop the treatment. Unless the gene is actively harmful to the bacteria (some resistance genes are, some are not), there is little selective pressure to get rid of it [1]. It'll be hanging around in plasmids in a tiny fraction of the bacteria, but as soon as you cycle back to that treatment again, the resistant strain will quickly emerge again because they don't have to re-evolve the gene.
You can do the math to calculate, based on the population size, rate of reproduction, rate of mutation, etc., the chance that a gene will still exist in the population after X years when there is negligible selection on it. I don't know how it works out for bacterial pathogens.
[1] Of course there is always the slight selective pressure against anything that enlarges the genome but has no function, since it takes energy to replicate that DNA and/or synthesize that useless protein. But that alone is unlikely to purge the gene from the entire population.
Massive force at a point. If your attack is total enough that you kill everyone in the initial engagement, then the survivors don't get to adapt.
If you made antibiotics something you can't get outside of an isolation ward. If we stopped using them on farm animals, if we made everyone finish their meds (rather than sending them home to maybe do it and maybe not), if we burned the bodies when we failed and sterilised the rooms....
Where's the chance to adapt, where's the chance to spread? You'd be looking at too few generations in a single individual and even if they developed the adaptation, they'd die anyway before it could spread.
I'm not saying that we will, but stranger things have happened. If you only had one antibiotic left in the world that worked, that would not be an unreasonable way to treat its use.
"If you made antibiotics something you can't get outside of an isolation ward. If we stopped using them on farm animals, if we made everyone finish their meds (rather than sending them home to maybe do it and maybe not), if we burned the bodies when we failed and sterilised the rooms...."
Of course all of that is radically better (in the narrow sense of decreasing the rise of resistant bacteria) than what we're doing today, but it's still not any sort of a guarantee. When you treat someone with antibiotics, you're necessarily exposing everything that's in 'em, and you're necessarily not wiping out everything that's in 'em, and bacteria trade genes.
Well, reducing introduction of microbes to the interiors of human beings reduces the need to deploy "antibiotics" in the sense of "sufficiently harmful for bacteria, sufficiently harmless for humans". Not a complete solution, but being better about reducing spread is probably a component - particularly in the surgery portion of all this.
Phages look real promising in this respect but a few problems:
- How do you get a sample to target? And then deliver the treatment?
- Cost for tailor-made treatments?
- Will it arrive in time (before antibiotics run out)?
- Evolving to attack human cells too?
In unrelated news, getting any needed surgeries done in the next few years might not be a bad idea.
If you expressed it in terms of computation, information and entropy, there presumably is a value that could be put on the processing required to outcompete the evolution of a given population of bacteria.
That's an interesting thought. I'm not sure if it's a well defined question though.
Even if you could calculate those things about the bacteria - which would be extremely difficult if not impossible - that doesn't really provide any bounds on how much computation you need to do to stay ahead of the bacteria. The state space of possible ways to kill bacteria and that that the bacterium explore are completely different.
It is just an idea I have been kicking around for a while now. Is basically from wondering if we will ever have the ability to design a targeted antibiotic computationally for individual treatment, which got me thinking on the bacterial evolution vs software thing.
