I used to work in the phage field and apart from regulation (phages are near impossible to get approved by the FDA), the big problem is the phages people have isolated have a very narrow specificity - most target only individual strains of a specific bacterial species (i.e. they have a very narrow host range). This means you need to have hundred or thousands of different phages to treat even a single bacterial species.
The really interesting thing is that there are phages that have a broad host range (multi-species or even multi-genera), but due to the nature of how phages are typically isolated they are not often found by researchers. Phages, like other living organisms, show differences in ecological adaption with some being specialists (leopards) while others are generalists (rats). The specialists grow faster in their specific hosts than the generalists do so in the normal isolation process which uses a single host bacteria strain (plaque plate assay) and so they appear first. The generalists grow slightly slower and so are typically missed by the scientists.
In my lab we developed a different assay that favoured the isolation of generalist phages and we were really successful in finding broad host range phages (many were multi-genera in host range). I no longer work in this area and it is something I miss as it is such an interesting topic :(
In my lab we developed a different assay that favoured the isolation of generalist phages and we were really successful in finding broad host range phages (many were multi-genera in host range).
Did your group publish anything on this approach? If so, could you share a link to one or some of the papers?
Yes we did, but we didn’t make a huge deal about it in the papers (it is just in the methods sections). If you want to find my phage papers just have a look at my profile on ResearchGate [1].
The basic idea is pretty straightforward. Rather than plating a filtered sample directly onto a overlay plate of a single bacterial strain, we pre-enriched for phages in a mixed culture of multiple bacterial strains (we would sometime use up to 50 different strains/species in the one enrichment flask). What this does is give the broad host range phages a competitive advantage since they can reproduce in multiple hosts and outgrow the narrow host range phages. When you plate out onto a specific strain for isolation of the phages you end up with mostly broad host range phages.
The idea is so obvious I am surprised it is not more widely used in the field. It seems to be the norm to just use a single bacterial strain for the phage isolation. I think most people just assume all phages are narrow host range.
Do you think it is feasible to allow some kind of voluntary FDA opt-out for individuals, in extremely limited cases?
I find myself wondering what I would want to do if I were in a similar situation: long-term, treatment-resistant illness, pain and inevitable death. At this point in my life, I think I would be okay saying "okay nothing is working. I volunteer to be an experimental subject - and if I die, please absolve anyone working to cure me from punishment."
Interesting fact: a doctor can prescribe you any drug for any reason, regardless of what the drug was meant for, subject to a couple restrictions. So, if you are in such a terminal situation, and some studies have shown drug X, normally used to treat some other disease, could help people with your disease, the doctor can prescribe it to you even if the drug hasn't been FDA approved for that purpose. This is intentional: it allows doctors to use cutting-edge research and don't have to wait years for FDA approval.
There are restrictions, of course. First, if a doctor start prescribing opiates to every patient s/he'll probably get a call from the DEA :). Second, insurance companies may refuse to pay for experimental treatments. And third is potential malpractice liability, but making sure the patient has informed consent can limit that. And this all only applies to drugs that are already FDA approved, just not for your specific use case.
I should point out the only reason I know this is from an article about some drug companies that have been accused of encouraging doctors to use their drugs for non-FDA approved purposes to boost sales. So I may be missing some key details here.
There is also an orphan drug classification which has a reduced approval process which is intended for very rare cases or cases where clinical trials are simply impossible to orchestrate due to the life expectancy at the clinical stage.
This is a good example of a downside of regulation: regulations embed assumptions that favor certain outcomes. Those assumptions might have made sense when the regulations were written, but are very slow to adapt to changes in the industry.
The FDA drug approval process was basically designed for things like antibiotics: Single molecular compound, high purity, standardized dosage and application, and run trials with many many patients.
Treatments that don't fall under that umbrella (like phages) are very difficult to develop and bring to market, even if they work very well, just because of the impedance mismatch.
Phages are basically individually tailored to the patient, require a lot of work in administering, and difficult to test under the FDA's guidelines. Many people have unnecessarily died because treatments that could be are lifesaving aren't brought to market due to too much rigidity in the system.
To be fair, the FDA has been active in coming up with ways for things like phages to be brought to market safely, but it's a very slow, uphill battle.
FWIW, Silicon Valley investor Jim O'Neill, who might well be the next head of the FDA, said this in 2014 about that:
“We should reform FDA so there is approving drugs after their sponsors have demonstrated safety -- and let people start using them, at their own risk, but not much risk of safety,” O’Neill said in a speech at an August 2014 conference called Rejuvenation Biotechnology. “Let’s prove efficacy after they’ve been legalized.”
That makes a ton sense, except that one wonders if any manufacturer will ever bother proving efficacy. Worse, they may instead double-down on marketing campaigns. Doctors and other professionals, working scientists or not, are susceptible to marketing like anybody else.
So we'd need to see other mechanisms, regulatory or otherwise[1], that would counter that effect. FDA approval is a powerful motivator for 1) companies to find actually effective therapies and 2) as an extremely efficient signaling mechanism to the market regarding efficacy. But it's sub-optimal for many reasons, including erring too heavily on the side of safety. It's just not obvious what would replace it.
Don't forget, snake oil thrives in the free market.
[1] Medicare reimbursements predicated on FDA efficacy approval? I'm not sure insurance companies can be expected to do the right[2] thing. It might be cheaper to reimburse a snake oil treatment that allows the patient to die than to pay a premium for an effective treatment. And while presumably a good doctor would insist on the effective treatment, at the margins the insurance company could expect to profit from allowing their patients to choose the snake oil.
[2] Where "right" means a course of action that maximizes long-term social wealth (i.e. preferring effective treatment and therefore subsidizing research), rather than simply maximizing the short-term utility of patients and profits (i.e. allowing the patients the salve of snake oil, with reimbursements, in lieu of effective treatment).
The problem with that is that insurance companies won't pay for a drug without demonstrable efficacy. A great example is Exondys 51 for DMD. They got FDA approval with very limited data on efficacy. Anthem said they won't pay for it.[1]
Now that's something interesting. I found it funny that the article didn't seem to even briefly question whether the drug approval process could be changed to make it simpler to get things like phages to patients.
If we're getting lots of political change, this sounds like something that could benefit.
To be fair, Phages are also vastly worse than antibiotics in they need to be extremely targeted. So, it's really not clear they are worth investing a lot of time or money developing until we have zero other options.
Antibiotics are extremely expensive to develop (on the order of billions of dollars), and patients regularly die because 1) the antibiotics were ineffective against the particular bacteria in their body or 2) the antibiotics wiped out their normal flora, and give them intractable cases of C. difficile and the like.
> zero other options
Those patients literally had zero other options than just dying.
Phages should be another weapon in our arsenal. You're right that they might not end up working out, but right now there is chronic underinvestment in research to find out whether work, because nobody is going to heavily fund research for therapies that will never pass the regulatory hurdle. It seems like bad risk management to back ourselves into a corner -- the whole point of research is to find out whether other things work. Defense in depth isn't just for software security.
The problem is they are not free. If we trade 1 new Antibiotic and get X Phage treatments how large an X do you need to break even? Next is it cheaper to find the next Antibiotic or get those X Phage treatments adopted? Current estimates suggest Antibiotics are still a vastly better investment.
I mean we have plenty of ridiculously expensive treatments out there already. It's not clear creating such things are a good investment vs. reducing traffic fatalities etc.
Well, the only way to answer that question is to do some more research on phages, and I guarantee you, the funding for antibiotics research is more than 100x that for phage therapies, in large part because of regulation.
Phage therapies are actually used regularly in Georgia, Russia, and other parts of the former Soviet Union, in part because they couldn't get access to antibiotics during WW2 and right afterwards.
You can do the math as to whether we've passed the point of diminishing marginal returns on antibiotics.
Look for actual numbers, the USSR for example did quite a bit of Phage research and then gave it up. So, no it's not 100x in favor of Antibiotics, which get vastly less funding than you might think.
Yes, before computers and labs-on-chips, growing them is difficult and time-consuming and maybe even dangerous, making them worse than antibiotics of the time. But now with MRSA spreading everywhere and chips that could measure purity, we should be able to make them cheaper and safer with fewer side effects.
Do bacteria develop resistance to phages as quickly as they might an antibiotic? A longer usable lifetime may change the balance of costs considerably.
Fun fact about the T4 bacteriophage: the dissection of its protein structure led to important refinements of SDS-PAGE, which is a protein analysis method that is still in wide use today. For that, the article that describes its analysis is one of the most widely cited in history[1]. (Also, yes, this is shamelessly a link to a tool I've developed to examine the importance of biomedical research papers.)
If you just handed out phage freely, as you do abx, bacteria will "eventually" evolve resistance to phage....at a 2-3 orders slower rate, but it will eventually happen.
Of course, the solution is easy, you don't develop a phage therapy like you would an antibiotic, but rather like a constantly updating flu vaccine.
The 'therapy' might not be handing out pre-grown phages, but literally doing the labwork to isolate and grow phages right in the hospital for each patient.
This was how phage therapy was originally developed, and it's still being done that in some surprising places like Georgia [1]. They have phage 'banks' for common bacterial species, and it's not effective, they end up taking bacterial samples from each patient and isolating more phages.
My impression is that virus evolution is much slower than bacterial evolution. They can't reproduce sexually, so there is no gene recombination to spread beneficial mutations. This difference is partly why vaccines are done for viruses and not so much for bacteria.
Influenza is an RNA virus. So is HIV and the coronavirus. And of all RNA viruses, influenza and HIV are among those that mutate the fastest. The influenza virus mutates so fast that 99 percent of the 100,000 to 1 million new viruses that burst out of a cell in the reproduction process are too defective to infect another cell and reproduce again. But that still leaves between 1,000 and 10,000 viruses that can infect another cell.
This is very interesting, however, is it answering the right question? HIV and flu vaccine are specifically called out as having unusually high mutation rates. Apparently the difference is between RNA viruses and DNA viruses - the RNA ones have no "double checking" to repair errors, so have much higher mutation rates. The rates quoted here [1] put DNA viruses as similar to humans. I don't know if these phages are RNA or DNA ones (Wikipedia says there are phages of both types and while the T4 virus they have a picture of in the article appears to be DNA-based, they probably just chose it for its striking appearance).
More importantly - is mutation rate what we want to consider? Bacteria exchange DNA with each other through horizontal gene transfer and Wikipedia quotes that as the main means of transmitting antibacterial resistance, for example [2], even across species (to the extent that bacterial species are even well-defined). I'm not sure how to make the comparison with that to viral mutation rates. This [3] source gives the following interesting, but imprecise, quote:
> This antagonistic coevolution results from the constant emergence of countermeasures by which bacteria resist phages, while phages, mutating at a more rapid rate than their prey, find means to overcome this resistance.
The paper it cites [4], definitely shows phage evolution alongside bacteria evolving counter-measures. So I think that's enough to answer the original question posed, though the question of "which is evolving faster" is harder to answer and probably isn't very well-posed.
That all said, I know nothing about this field and would be happy for some better insight.
More importantly - is mutation rate what we want to consider?
I think it is quite important when considering this particular interaction. Horizontal gene transfer will help a population of bacteria share a successful strategy, rampant mutation will help the virus try more strategies.
The article mentions that the particular infection reached a balance with the phage overnight, so it is likely to be a quite an issue. But the population of bacteria resistant to the phage had become susceptible to an antibiotic (for reasons related to the resistance). So it can still be a valuable tool, even if resistance happens quickly.
Broadly speaking, most viruses have no mechanism to control mutations other than becoming inactive/nonfunctional. Some bacteria are also like this, in that they can tolerate a high mutation rate without going extinct, while other bacteria have rudimentary control mechanisms that slow their mutation rate in favor of higher survivability and specialization.
Phage therapy is very promising but it's labor intensive and specific to each particular infection. That also poses a monetization problem for pharmaceutical companies who prefer to create a pill that can be universally applied to many patients, which is perhaps why it is not more widespread.
But, it is currently practiced in Poland and Russia.
> A virus from a pond just 40 miles from his house apparently gave him a new lease on life—and gave a new meaning to the term “local medicine.”
That is, are phages for a specific bacterium more likely to be found local to the person infected? Would they have been even more likely to find the right phage if they had collected samples from the patient's neighborhood, family and friends? Nearby people who weren't infected?
The short answer is no. Bacteria like pseudomonas don't really have geographical ranges because they are not just human pathogens- they are prolific in temperate soils throughout the world. The same species can be found throughout the world and spread from place to place with great ease, and there will be hundreds of pseudomonas phages capable of infecting any given strain. I am fairly confident that you could isolate an infectious phage for this pseudomonas strain from many temperate soils or mucky ponds anywhere in the world. This is not universally true for all bacteria, but is mostly true for bacteria like pseudomonas.
I'm not a doctor or research biologist, but I have the same question. My default assumption is that it depends on whether this is a locally evolved bacterium. If it's a common strain found everywhere then the phage is probably pretty common too. If this is some unusual strain of Pseudomonas that's localized to his area then a phage that coevolved locally with that specific strain might be the most effective fighting it. Either way, it's a probability situation as there's no guarantee a particular phage will fight a particular bacterium until it's tried.
> Tbilisi draws patients from around the world who are suffering from untreatable UTIs, acne, cystic fibrosis, and intestinal infections.
I've heard of a case where someone with a horrible Staph leg infection was desperate so flew to Tbilisi, apparently it worked for them.
Wonder if they can monetize that better. Say set up something in Mexico right across the border, so people don't have to fly half way across the world...
I wonder if the study of these phages can give us better treatments than just gambling with cocktails of the phages themselves. Like understanding their point of infection and how we can exploit that to deliver some chemical to kill the hostile bacteria more effectively.
I remember when i was young whenever we had a cut that had got a bit red my grandmothers solution was to go for a swim in the ocean, maybe we were phage hunting.
The really interesting thing is that there are phages that have a broad host range (multi-species or even multi-genera), but due to the nature of how phages are typically isolated they are not often found by researchers. Phages, like other living organisms, show differences in ecological adaption with some being specialists (leopards) while others are generalists (rats). The specialists grow faster in their specific hosts than the generalists do so in the normal isolation process which uses a single host bacteria strain (plaque plate assay) and so they appear first. The generalists grow slightly slower and so are typically missed by the scientists.
In my lab we developed a different assay that favoured the isolation of generalist phages and we were really successful in finding broad host range phages (many were multi-genera in host range). I no longer work in this area and it is something I miss as it is such an interesting topic :(