Hacker News new | past | comments | ask | show | jobs | submit login

Many years ago, I remember seeing a BBC documentary about phages that were (and are) used in Russia and former Soviet states in place of antibiotics.

What are phages? "Phages are naturally occurring viruses that kill bacteria. Once they get into bacterial cells the phages' DNA replicates until it kills the host.

Doctors in Georgia, and in other countries that were in the former USSR, have been using this therapy for 90 years. But medics and drug regulatory bodies in most places in the developed world have been reluctant to accept that it works."

From: http://www.bbc.co.uk/news/health-21799534




Too bad they were largely abandoned because antibiotics were cheaper and easier to produce (a situation similar to electric vehicles in the early 1900's, in fact)...


It wasn't just that. The problem with phages is it's impossible to make enough for a therapeutic dose without introducing mutations, so you're never completely sure what you're giving the patient. They're just not as safe as conventional antibiotics.

We'll probably end up using phages because we won't have any other choice. But that's not a step forward for patients.


I'm not sure this analysis is accurate. At any rate, I don't share your concerns.

The big problems with phage are:

   1) Narrow spectrum
   2) Readily cleared by the immune system
   3) Bad at killing bacteria in comparison to antibiotics.
I worked with a a strain of Phi X174 that was optimized for a particular host by serial passages. It did a number on petri dishes, but I'm skeptical that it could have done much to clear wild type e. coli. Antibiotics are good and bad because they indiscriminately kill everything. If phage are every used in a widespread manner, I imagine they'd get deployed in a cocktail to broaden the spectrum.

It's easy to make highly pure phage. Colonies are clonal and form plaques when plated. If you did serial passages to ensure that the phage was near some local fitness maxima, it's unlikely that selective effects would move it away from that maxima. In Phi X174, which admittedly has a short genome (5386bp), I would have been surprised to see any clones with mutations. I was doing site-directed mutagenesis to change single base pairs, so I thought about this a lot. Worrying over a couple of base flips when mass producing phage is simply a hallucinatory fear.

At least one phage is dangerous, but I'm not aware of any other examples. CTX is a temperate phage that makes cholera produce toxin. There is a risk of immune response. The immune system tends to get pissy when you inject foreign matter. I've seen studies that compared oral dosing to intramuscular phage and IM seems to be the way to go for higher blood titers. In one study, serial passages were performed in rabbits. The phage acquired mutations in its capsid proteins and over several generations blood titers increased by something like two orders of magnitude. (I don't recall if they used the same rabbits or naive rabbits.) Makes you wonder if we could recover super fit phage from people receiving phage treatment?

Finally, with regards to safety, phage are currently in use as an anti-microbial again listeria.


Is "highly pure" really pure enough for the FDA? I knew a guy who worked on phages. The problem they ran into is when they submitted their virus to the FDA there was no way they could guarantee the absence of dangerous mutations in a given batch.

EDIT: I should say I didn't follow his work that closely, so it's possible there were efficacy problems as well.


The bar has certainly been met with regards to phage as a food additive: http://www.fda.gov/OHRMS/DOCKETS/98fr/cf0559.pdf

The error rate of an E.coli DNA polymerase in vitro in <9x10^-6. Lambda phage, for example, has a genome that's about 48 kilobases in length. Which gives a mean error rate of 0.432 bases / replication. Ballpark estimate that the phage goes through 30 replications to produce a plaque, so we're talking ~13 mutations (it's a little different than that due to the founder effect, etc).

It isn't like a phage is going to jump Kingdoms and start predating you instead of the preferred host. This simply isn't a realistic risk, while the immunogenicity issues could be.

In terms of my own risk tolerance, I would have been comfortable drinking any of my own samples.


>In terms of my own risk tolerance, I would have been comfortable drinking any of my own samples.

Hmmmm. Maybe so, but would you inject them?


If there was a medical need and they were shown to be efficacious, sure.

I'm currently receiving allergy shots which involves frequent injections of dust mites. Injections are followed by a 30 minute observation period as a caution against anaphylaxis. Staff are on hand to intubate. The frequency of life threatening reactions is around 1 in 10^6.

If phage were ever used clinically, I suspect the protocol would be similar. I mean, heck, some people have life threatening reactions to common antibiotics (I believe sulfa allergies are particularly common), but that doesn't keep them out of use.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: