2 ms·

One of the sort of strange things about gene therapy is how cheap a custom solution really is. You can arbitrarily genetically modify a plant or bacteria for a

One of the sort of strange things about gene therapy is how cheap a custom solution really is. You can arbitrarily genetically modify a plant or bacteria for a few hundred dollars to the level of arbitrarily picking the letters of a gene sequence (or you know, doing some research and doing something real) and getting the gene modifying tools manufactured very very cheaply. If you have a slightly different form of a much more common ailment, there probably is good hope that eventually patient specific gene therapy for you will be something of a routine. You may be taking extra risks because of the lack of available research, but I'm sure any patient can become well informed enough to make the choice for themselves. I'm not saying it's commonplace or anything yet, but the tools are there or almost there to fix a slightly incorrect single gene issue (like a single point deletion or letter swap, etc). Like just look https://www.idtdna.com/pages/products/crispr-genome-editing/alt-r-crispr-cas9-system https://www.idtdna.com/pages/products/crispr-genome-editing/... obviously this is for research and not in vivo human treatment which takes quite a bit more care and specifics to target whatever tissues / organs / whatever is broken in you... but the reagents for doing custom sequence edits are priced for research in the hundreds of dollars. sure once an established research program finished and got approved they'd spend tens or hundreds of millions and charge each patient millions but it's not insane to imagine a biohacker curing you in their garage for $5,000 while only maybe killing you Given that, I think there's a lot of hope that custom gene therapy will become commonplace and relatively affordable.

rubidium 1y ago
You’ve taken a very small slice of the total workflow to develop cas9 crispr edits. Yes some components are “not crazy expensive”, but that’s not therapeutic production levels. And no sequencing. And no bioinformatics. And then zero tissue targeting/ delivery or safety studies. So $5k is a massive stretch. I do think costs are coming down. But don’t want to oversell current state of “garage hacking” either.
- colechristensen 1y ago
  Sequencing all of the interesting parts of the human genome is $300-$400 and a few weeks wait. The bioinformatics to find a simple mutation and design a crispr fix or one of the other gene therapy methods is about 1 grad student week. I am talking about garage hacking and there sure are a bunch of risks and chances of ineffective treatments, but researchers used to take a lot more risks and I can see folks with degenerative, terminal, or otherwise very severe conditions being happy to say fuck it let's try something. And with an appetite for risk and a fuck it attitude, you can try something for less than the cost of a used toyota. And as mainstream treatments develop, custom gene editing as a fully safe therapy option is going to become common.
mandevil 1y ago
We are not close to that. Gene therapies are still very much hit or miss in RCT's, if some biohacker offers you something for 5k you are definitely getting scammed. The first gene therapy approved for in vivo treatment by the FDA, Luxterna, was approved in 2017. It treated RPE65 associated Retinitis Pigmentosa (at an original list price of 425,000/eye). Just a few months ago, J&J's LUMEOS trial revealed that basically the same process but targeting a different gene (RPGR) failed to meet its primary endpoints in the Phase III clinical trial (only 22 of 55 patients treated showed improvement on at least two measures, and there was no statistically significant improvement on the main measure at all). At this point, we can't even reliably take a process that fixed one gene and apply it to another gene to treat the same disease, that's how far away we are from "some dude in a garage." Will we get there? Maybe, but the human body is far more complicated than software, and analogies based on how software work mislead more than they help.
- colechristensen 1y ago
  If you were a grad student with RP, you could replicate the LUMEOS study drug for $400 + your time for a garage level safety self treatment. I can certainly see someone being willing to take that risk themselves for a 40% chance of improvement to reverse or prevent their own blindness. There are many guardrails for safety which are quite expensive but the actual materials are absolutely not expensive at all.
  - mandevil 1y ago
    The experience of patients who received Luxterna was that the single most important difference in how large the improvement was and how many complications they had was the quality of the surgeon who delivered the actual therapy. The best surgeons saw much better results and far fewer complications, and these surgeons are way beyond the capability of some grad student in a lab. The best surgeons were way better than other licensed, trained surgeons! That these sorts of non-gene factors dominate the outcomes- and costs- of gene therapy is the point that the original article is making. The articles example for this is Casgevy, a gene therapy that can mostly cure Sickle Cell. The problem is in order for that gene therapy to work it takes surgeons, full transplant teams, super-chillers, chemotherapy, and full hospitals to deliver it and for the patient to recover in while being carefully monitored. This is what drives the cost, and so it doesn't matter how cheap the actual gene sequencing or editing is, the rest of the costs dominate. This is Amdahl's law, but for costs- the costs are dominated by the non-gene editing part of the process and so that determines the improvement in cost you will see.

rubidium 1y ago

You’ve taken a very small slice of the total workflow to develop cas9 crispr edits. Yes some components are “not crazy expensive”, but that’s not therapeutic production levels. And no sequencing. And no bioinformatics. And then zero tissue targeting/ delivery or safety studies. So $5k is a massive stretch. I do think costs are coming down. But don’t want to oversell current state of “garage hacking” either.

colechristensen 1y ago
Sequencing all of the interesting parts of the human genome is $300-$400 and a few weeks wait. The bioinformatics to find a simple mutation and design a crispr fix or one of the other gene therapy methods is about 1 grad student week. I am talking about garage hacking and there sure are a bunch of risks and chances of ineffective treatments, but researchers used to take a lot more risks and I can see folks with degenerative, terminal, or otherwise very severe conditions being happy to say fuck it let's try something. And with an appetite for risk and a fuck it attitude, you can try something for less than the cost of a used toyota. And as mainstream treatments develop, custom gene editing as a fully safe therapy option is going to become common.

mandevil 1y ago

We are not close to that. Gene therapies are still very much hit or miss in RCT's, if some biohacker offers you something for 5k you are definitely getting scammed. The first gene therapy approved for in vivo treatment by the FDA, Luxterna, was approved in 2017. It treated RPE65 associated Retinitis Pigmentosa (at an original list price of 425,000/eye). Just a few months ago, J&J's LUMEOS trial revealed that basically the same process but targeting a different gene (RPGR) failed to meet its primary endpoints in the Phase III clinical trial (only 22 of 55 patients treated showed improvement on at least two measures, and there was no statistically significant improvement on the main measure at all). At this point, we can't even reliably take a process that fixed one gene and apply it to another gene to treat the same disease, that's how far away we are from "some dude in a garage." Will we get there? Maybe, but the human body is far more complicated than software, and analogies based on how software work mislead more than they help.

colechristensen 1y ago
If you were a grad student with RP, you could replicate the LUMEOS study drug for $400 + your time for a garage level safety self treatment. I can certainly see someone being willing to take that risk themselves for a 40% chance of improvement to reverse or prevent their own blindness. There are many guardrails for safety which are quite expensive but the actual materials are absolutely not expensive at all.
- mandevil 1y ago
  The experience of patients who received Luxterna was that the single most important difference in how large the improvement was and how many complications they had was the quality of the surgeon who delivered the actual therapy. The best surgeons saw much better results and far fewer complications, and these surgeons are way beyond the capability of some grad student in a lab. The best surgeons were way better than other licensed, trained surgeons! That these sorts of non-gene factors dominate the outcomes- and costs- of gene therapy is the point that the original article is making. The articles example for this is Casgevy, a gene therapy that can mostly cure Sickle Cell. The problem is in order for that gene therapy to work it takes surgeons, full transplant teams, super-chillers, chemotherapy, and full hospitals to deliver it and for the patient to recover in while being carefully monitored. This is what drives the cost, and so it doesn't matter how cheap the actual gene sequencing or editing is, the rest of the costs dominate. This is Amdahl's law, but for costs- the costs are dominated by the non-gene editing part of the process and so that determines the improvement in cost you will see.