Hacker News new | past | comments | ask | show | jobs | submit login
In a First, US Doctors Use Crispr Tool to Treat Patient with Genetic Disorder (npr.org)
270 points by mhb 7 days ago | hide | past | web | favorite | 89 comments





That article was from July 2019, there is an update from today:

https://www.npr.org/sections/health-shots/2019/10/10/7667657...

The patient gets to go home today.


Reminds me: this guy used gene therapy to cure himself of lactose intolerance quite a while ago. Not sure how common it actually is among biohackers.

https://www.youtube.com/watch?v=J3FcbFqSoQY


People who do this are usually irresponsible and don't even know whether they've achieved their goal or not.

"""Just to make sure this is really clear because people have been angry about this ever being tested in more humans. This will not be used in humans again until significantly more testing is done and the manufacturing protocol is massively refined. This could mean years before another test is conducted. More tests will need to be done in animal models as well. Maybe we'll do away with the viral aspect all together and completely overhaul the DNA. The reason this video was posted was because I want this to be an open source technology. If you have suggestions for how it can be improved, leave it in the comments."""

I spent 20 years of my life working towards gene therapy and to see idiots do stuff like this (in an incompenent and unconvincing way) is so frustrating.


You didn't actually tell us why it's irresponsible. Is there a real risk of them hurting anyone but themselves?

A neuroscientist (nate1212) on reddit offered the following reasons:

  - He just possibly infected his whole digestive system. Not just small intestine, but stomach 
    as well. Furthermore, AAV can potentially exhibit transcytosis through epithelial layers, 
    suggesting that it's possible the virus infected more than just his digestive system.

  - He did not determine an appropriate dose, and so he likely infected with a HUGE genetic 
    payload. Overexpression with AAV can kill infected cells, which means this man is risking 
    his digestive lining.

  - Neither the promoter nor the encoded protein itself are human, potentially risking 
    (possibly severe) autoimmune reaction.

  - There are few/no long-term studies on effects of AAV integration and expression in humans. 
    There is indeed evidence that AAV increases risk of cancer, almost certainly in a dose- 
    dependent manner.

  - AAVs could integrate randomly into your genome in low frequencies (0.1% to 1%), meaning that they could just by 
    chance disrupt a gene you really need to not get cancer.
The person in the video also suggested "trials for volunteers" (on his reddit post), but immediately backtracked on that after he learned that people cannot legally consent to such experiments.

Good info, but bad formatting. Same quote, more mobile friendly:

- He just possibly infected his whole digestive system. Not just small intestine, but stomach as well. Furthermore, AAV can potentially exhibit transcytosis through epithelial layers, suggesting that it's possible the virus infected more than just his digestive system.

- He did not determine an appropriate dose, and so he likely infected with a HUGE genetic payload. Overexpression with AAV can kill infected cells, which means this man is risking his digestive lining.

- Neither the promoter nor the encoded protein itself are human, potentially risking (possibly severe) autoimmune reaction.

- There are few/no long-term studies on effects of AAV integration and expression in humans. There is indeed evidence that AAV increases risk of cancer, almost certainly in a dose- dependent manner.

- AAVs could integrate randomly into your genome in low frequencies (0.1% to 1%), meaning that they could just by chance disrupt a gene you really need to not get cancer.


None of those answers have anything to do with answering the question of "Is there a real risk of them hurting anyone but themselves?"

Let me rephrase, to make this more explicit: why shouldn't people be allowed to commit probabilistic suicide by non-rigorous experimental protocol, if they want? People do things that could kill them in home machine shops and home chemistry labs all the time, just for fun.

Note that I completely understand why people shouldn't be advertising that they do this, any more than people should be trying to talk other people into building nuclear reactors as a hobby.

I'm more saying that, if the point is just do something for your own sake—with a chance of improving your life, and another chance of killing you—and nobody else is involved or will be involved—then what has anyone else got to do with it?


People shoot up heroin every day and it's pretty bad for your health of course. Now imagine if someone vlogged their heroin habit to the masses, and a handful of those masses were impressionable like minded idiots who also took up H as a result of seeing a video by a supposed internet authority. It's a jackass move, overall.

None of those seem to affect anyone else, so as long as the person experimenting on themselves is of sound mind I don't see the problem.

Not seeming to affect anyone else seems like an insufficient condition. It should be predictably not going to affect anyone else with high reliability - and I strongly suspect this was not the case.

It can also affect people indirectly, which leads to the same reasons we have laws requiring seatbelts or motorcycle helmets.

Not everyone wants to see someone die in a horrible way and have to clean up the mess.


Exactly. As long as your experiments cannot modify the germline, knock yourself out with CRISPR self-experimentation (as long as you're comfortable with the possible consequences).

Absolutely not. It only takes one idiot to fuck up their body horrifically for the technology to get banned for all the non-idiots.

You can’t ban something I can perform in my garage (not effectively at least) with a few thousand dollars in supplies.

Sure they can. They banned explosives (which I agree with) and you can probably do that with a few hundred dollars.

Sure they’re banned but it’s not particularly hard for amateurs to manufacture after some careful reading, The situation is similar to bans on cryptography IMO.

> The situation is similar to bans on cryptography IMO.

No offense, but a misleading and inaccurate analogy. It's much closer to bans on chemical precursors of drugs like MDMA or LSD. Those chemical precursors have many legitimate uses, but since some people use them to make illegal drugs, normal citizens cannot acquire them easily, or without breaking the law.


I don't see how that's different. Again you can manufacture those precursors after careful reading (that's the fundamental idea behind organic synthesis.)

They can ban the actual effective research and use cases though.

Smoking increases the risk of cancer too, and we let people choose to do that, even though it also increases the risk of cancer for everyone around them as well.

Maybe both of these cancer-causing things are bad.

While I'm sure this is true and there are real reasons why performing such experiments on yourself may be a very bad idea, there are few caveats.

First (which I don't really want to concentrate on) is that's why this is an experiment. People did stupid things (and sometimes died because of them) for entire human history and some hundreds of years later are remembered as "great scientists". Trying to monopolize research and experimentation around academia and "certified professionals" (certified by academia, that is) is a most despicable thing I can imagine. I see that this opinion is basically a point of conflict of two different political views, so there isn't much that can be said to (quickly) persuade someone, who does not feel like this in the first place, but I strongly feel that if people can be "free" in any sense at all, they really should be free to do whatever the fuck they want to themselves. If they wanna die, let them die. If they wanna do drugs, let them do drugs. If they wanna perform stupidly dangerous experiments over themselves, let them do that. If they wanna give somebody else a consent to perform a stupidly dangerous experiment over them — it's their prerogative.

Second is much less philosophical and much more grim. This probably depends on your environment quite a lot (country/place, level of income, connections), so take this as my subjective experience. The problem is that dealing with certified medical professionals performing FDA (or whatever) approved procedures over yourself sometimes feels pretty much as dangerous, if not more. You often have no idea what they are really doing to you, and it's sometimes blatantly obvious they don't really understand it either. Often it is not clear if the treatment (which might be totally safe and correct in some cases!) is adequate to your case, not much of analysis is performed and every little bit you have to do along the way costs quite a lot of money and/or prevents you from living normal life, leaving you quite miserable overall. And the more minor the thing you are trying to fix is, the worse it is. And the worst of it, it is pretty clear that your doctors don't really care. They are relatively safe (legally) if they are working by the book, so they are working by the book. And "the book" often happens to be a result of bureaucratic machine, not common sense. And, after all, they have seen (maybe just 30 minutes ago) people in much worse condition than yourself, so unless you are really dying, what's even the trouble? And all of it even when we ignore the fact, that "the book" and the status quo of the field changes all the time, much of recognized medical research lacks the quality you wish it'd had and accepted views today sometimes are vastly different than 20 years ago. And let's be honest: 20 years of "experience" is nothing, when we are talking about high-stake decision making on the frontiers of the scientific research (and whatever affects your health and well-being always feels like "high-stake").


What you have to say about monopoly of research about certified professionals is apposite but I doubt there were 10 chuckleheads looking to replicate Barry Marshall’s experiment for kicks

Well, that's why Barry Marshall is exactly that kind of guy who will be remembered as a great scientist for doing stupid things. And I doubt there were 10 certified professionals looking to replicate his experiment either.

> AAVs integrate randomly into your genome meaning that they could just by chance disrupt a gene you really need to not get cancer.

I watched the video. I am not a biologist.

I could have sworn at one point he said that it would be inserted at "known, and safe" locations. Not "random"... ?


This is mainly true, however we are still observing random integrations at low frequency such as 0.1% to 1% [0], so if you bring a billion virus particles in, it still poses a non-negligible risk to the agent.

Thank you for your input, I'll word it more accurately.

0: https://www.ncbi.nlm.nih.gov/pubmed/29326962


Very interesting. Thanks for the clarification!

It sets the entire field back by creating a lot of fear and bad press that discourages further research.

This example is from institutional research rather than an indie biohacker, but Jesse Gelsinger's death may have set back the field of gene therapy by many years due to fear (some realistic and some unrealistic). https://www.sciencehistory.org/distillations/the-death-of-je...


I think we have to weigh the risks vs. the potential rewards, and accept "premature death" as a possible cost.

The lack of ability to process out ammonia seems pretty debilitating, so that seems like it was a good candidate.

Jesse Gelsinger is a hero, and should be remembered as such.

That all said, we should glean as much science as possible from the postmortem to understand and prevent whatever misunderstandings resulted in the death, going forward.


He is not responsible for fears of other people. People were afraid of everything throughout the whole history of human science.

No, but he can definitely be responsible for the technology being highly regulated or restricted if his experimentation goes awry .

actually I think the biggest risk is that it will create the impression in people that gene therapy is easy and that a random person on the street can just inject something into them and cure a disease. every self-injecting biohacker I've ever seen basically doesn't even know scientifically speaking whether they had the effect that they thought they did an experiment. contrast this to real gene therapy researchers who have to go through careful clinical trials and do a lot of statistics just to prove that they did what they said they did. the reason that we make them do that is because it's all too easy to fool yourself things like placebo effects and self bias. legitimate researchers go to extensive lengths to demonstrate that the procedure they did did what they said they thought it was going to do and then it didn't have side effects. none of the self biohackers have done that.

> legitimate researchers go to extensive lengths to demonstrate that the procedure they did did what they said they thought it was going to do and then it didn't have side effects. none of the self biohackers have done that.

I trust you (haven't checked biohackers much myself), but as with any field of knowledge/research, the opposition you suggest has nothing to do with credentials and everything to do with intellectual honesty, integrity of the researcher.

I mean, let's take programming since it's HN. Say you compare two programs. You could follow the scientific method (hypothesis -> experiment(s) -> conclusion -> new hypothesis -> etc) like any decent engineer or scientist would... or you could just randomly "feel" which is better "in your opinion" — basically being a charlatan with your own topic.

Unless you're doing marketing, you should be selfishly attracted to the truth — which is really better / faster / etc — because you need the tool for yourself, not to sell it to someone else. I'll never understand charlatanism in research, I won't; but even toy benchmarks for lulz or journaling can be made scientific, easily. It's just a logical framework, which I'd think any self-proclaimed 'reasearcher' knows in and out...

This is isn't to say that you can control all variables in real-world experiments — biological and social research are the hardest in that regard. Sometimes you gotta take a leap because you won't ever be able to "prove" anything beforehand, it's only retrospectively in statistics that you'll know (at the level of metastudies). We've taken many such a bets with public health, family health, our own; whenever the `estimated risks / perceived benefit` ratio makes sense.

I do agree it's wrong to categorize beyond these facts, e.g. credentials of the author of a good idea, or lack of "proof" thereof, have little to no meaning historically — can't stop an idea/solution whose time has come. I mean, Js is the most popular language with Python, let that sink in for a minute. (lol)


Is there a risk in genetic engineering that you're not only hurting yourself but also potentially any children that you have? I don't know the answer to this, but intuitively it seems like a possibility that you could accidentally introduce a genetic trait that won't become apparent for generations.

Don't think so. Traits passed on through eggs and sperm, not muscle, bone etc. If sex cells not affected, then no issue of passing on changes?

I am not a biologist, but do these sorts of therapies discriminate which cells they actually affect?

it depends on the details. if you collect cells from a patient, filter them very carefully so only a particular type of cell is there, treat that in vitro, then reinject.... it should (with some caveats) only affect the cells you treated.

if you inject a therapy into the body (muscle, blood, liver, etc), there is a non-zero chance that some germ cells can be modified. It depends on the vector, the methodology, random chance, etc, etc.


I would guess that people would try and replicate their results and end up hurting themselves, for one.

I don’t understand this question. Being irresponsible and only hurting oneself aren’t mutually exclusive.

Even assuming your premise were true, it just takes one irresponsible idiot to fuck themselves up to result in excessive regulation which slows progress.


Killing yourself harms more people than just yourself, true personal independence is an anarcho-capitalist fantasy. The harms to others include your friends and loved ones being upset that you're dead (obviously), whatever it is you're testing gaining a reputation for killing people, and also it removes your presumably smart brain from the team of people working on the problem. If you merely harm yourself then all of the above apply but to a lesser degree.

If you feel bad about something that does not mean others have a moral obligation to prevent your feeling bad from happening. If there is one thing we are sure about is that we "are" and if someone wants to stop "being" but someone else doesn't think he should because he will be sad then that selfish fuck should respect someone else's decision about himself and do what he can regarding fixing his own feelings. The only time this doesn't apply is with parents leaving kids because they forced their "being" onto them and there's no escaping.

Isn't causing other people pain a pretty basic thing that you aren't supposed to do? If I went around delivering electric shocks to my friends, the best advice would be for me to stop, not for my friends to achieve a level of zen so high that pain didn't bother them.

> Isn't causing other people pain a pretty basic thing that you aren't supposed to do?

Uh, obviously not? Seriously, this is absolutely self-inconsistent way to see things. No, it is not a "pretty basic thing" and the world isn't a friendly place, deal with it. People always cause pain to the others whether you like it or not. If I've been dating a girl for several years and still love her, but she does not love me anymore, she is causing me pain by breaking up with me. Quite likely more so, than a dear friend's death and pretty much for sure more that, say, a robbery. So what, she isn't supposed to break up with me now or what?


[flagged]


flagged for politics

no reason to post about hate crimes on a thread about crispr


> Making someones opinion or feelings criminal is insanity

You are right, but "hate crimes" are not just bad thoughts, they are bad thoughts brought through to action. It's one thing to rationally defend yourself from an attack and unintentionally kill the attacker, it's entirely another thing if you act on your toxic worldview based on bullshit and someone ends up harmed or worse. It's just too bad we don't apply this thinking to things like... oh, I don't know, religions that actively try to prevent their congregations from getting immunized.


flagged for politics

no reason to post about hate crimes on a thread about crispr


"Hate crimes" in some countries also comprise expression of hate. Hate crimes can be a useful designation for research and social intervention, but prosecuting them more harshly is literally policing thought as a secondary offense and makes the population malleable to further authoritarianism. It's a bailey type of argument because criticizing it means you must be sympathetic to racists.

We punish "thought as a secondary offense" all the time. eg: involuntary manslaughter vs voluntary manslaughter vs third degree murder vs second degree murder vs first degree murder.

Or more pointedly - terrorism is a different charge from murder, the difference being the "thought as a secondary offense" of intending to cause a political effect with your act of murder.

Hate crimes are analogous to terrorism. They are an offense not just against an individual, but an entire class of individuals. They are intended to send a message (even if it's as simple as "I don't think you should be present in my society") and cause fear and intimidation. And as such it's perfectly fine to punish that differently from plain old "killed you for the paper in your wallet".

Why do you feel it necessary to bring up how much you hate crime laws in a thread about CRISPR?


flagged for politics

no reason to post about hate crimes on a thread about crispr


Well, he can eat pizza now. What use your 20 years of experience were to him anyways?

It's a narrative that he can eat pizza. I don't think he really cured anything.

Amusingly, none of my research in gene therapy is relevant today. It was predicated on an already obsolete concept- designing transcription factors that can bind to a unique site in the genome to guide integration. CRISPR (which is what the article is about) made my approaches technically less viable, because they require extensive simulation, modelling, and experiment, while with a good CRISPR system, you just need an appropriate guide RNA.

Of course the really hard problem isn't the integration technique, it's knowing what to integrate.


As someone else mentioned, he likely hasn't really cured himself of anything. Regardless, he could have just taken lactase and called it a day

A frequently successful alternative therapy is to eat some yogurt once in a while.

I've never had terrible symptoms (merely unpleasant), but I am lactose intolerant, and the tiny bit of lactose that is present in mozzarella (~½ gram in a serving) doesn't bother me any.


There are some pretty big methodological issues here, and some of the comments make good cases that this didn't actually do anything.

> cure himself of lactose intolerance

Which is funny, because lactose intolerance is actually the norm, before a bunch of ancient Europeans decided that they'd figure out how to live off cows and there was selection pressure to extend lactose tolerance out of infancy where it normally remains


You keep using this word "normally"...

Are you saying that natural selection is not normal if it was only a few thousand years ago?


"Normally" simply means majority of humanity is lactose intolerant.

interesting .. didn't know that it was already possible

He could've also just, y'know, stopped consuming garbage lactose.

How can one tell the difference between good and "garbage" lactose, then?

Simple: All lactose is garbage.

Does anyone know of any companies hiring software engineers in this field? It's basically my dream job to somehow combine my interest in DNA/Genes with my skillset (programming)

In the bay area:

We are! (Serotiny): https://serotiny.bio

Synthego: https://www.synthego.com/

Mammoth: https://mammoth.bio/

Zymergen: https://www.zymergen.com/

Boston:

Arbor: https://arbor.bio/

Asimov (kind of): https://www.asimov.io/

Ginkgo: https://www.ginkgobioworks.com/

Also just happy to chat about the (very interesting, vibrant, if still tiny) intersection of software & synbio.


I used to intern for a company called FlowJo that does DNA flow cytometry data analysis, which helps power projects like these. They are hiring a few engineers right now -> https://jobs.bd.com/search-jobs/Ashland%2C%20OR/159/4/625200...

Here's their homepage https://www.flowjo.com/about/company


We make a variant caller, which is used in the pipeline for this kind of work, you might find it an interesting rabbit hole. https://magnolia.sh

Yes. Mine. pa11@sanger.ac.uk.

For a second I was really pleased to see a 'decline and go to plain text site' button on the cookie mask, but it goes to a page with only links to 'top news stories' etc. and not the actual article.

What's the point?


Back in 2017, 60 minutes covered a group that was cured of sickle cell disease with gene therapy. How does this approach differ when they already showed success? https://www.cbsnews.com/news/could-gene-therapy-cure-sickle-...

This study uses a different vector and targets a different gene.

The NIH study uses a modified HIV to introduce the complete gene for the defective protein.

https://www.nhlbi.nih.gov/news/2019/sickle-cell-patients-rec...

This study uses CRISPR/Cas9 to increase expression of a gene that the patients already have.


Just skimming through the article you posted, it says trials with a similar treatment resulted in one patient dying from a massive immune response and two other got cancer. Sounds like it relies on a virus to deliver the DNA, but I didn't read too carefully, so maybe CRISPR will turn out to have fewer side effects or a greater success rate. I'm pretty sure one of the up-sides to CRISPR is compared to other gene-editing techniques it's relatively cheap, which is always a plus.

The death and cancers mentioned over there were from gene therapy trials in general, not the specific study in the link.

FYI a little misleading that it's posted now, since the linked article came out in July.

Sarah Cannon, where this trial took place, purchased Boston based Genospace ~18 months back. Curious of Genospace's role in this and if they are able to add value to any other Sarah Cannon trials.

It would be interesting to see the legal release this patient had to sign to get this done in the US.

From what I recall CRISPR is always used with another technology allowing more precision. CRISPR is too broad so it needs something with more precision to assist.

I skimmed the article and did a quick search but I can't recall the name of the second method.


Are you thinking of eg CRISPR Cas9 and or Cpf1?

https://en.wikipedia.org/wiki/CRISPR#Cas9


Assuming they use CRISPR outside the body, on blood, then infuse the blood back in...

Are there any credible risks that something unexpected happens?


That isn't what they are doing.

They are extracting bone morrow and editing that and then reintroducing it. It's basically a bone marrow transplant using the patient's edited stem cells.

http://www.crisprtx.com/programs/hemoglobinopathies

Words like "credible" and "unexpected" sort of make it impossible to answer the question. The doctors probably expect some editing errors and things like that. They are also doing a really small study to see what happens.


Words like "credible" and "unexpected" sort of make it impossible to answer the question

I guess I could reword. Is there more than a remote chance the patient dies because of an error related to the gene editing?


To know that with any confidence, you'd have to do this for a lot of people and look at the stats. We don't currently have the technology to answer a question like that convincingly for n=1 studies.

They don't know for certain, hence a small pilot study. They of course believe that it will work and that there is a limited chance of such things.

Yup: "During a routine data run, they unexpectedly found foreign, non-bovine DNA that had bound itself to the animal’s genetic sequence during the edits—specifically, genes from the lab material"

https://newfoodeconomy.org/fda-gene-edited-cattle-antibiotic...


What's the probability that the sequencing sample was contaminated and that the sequence result doesn't represent the animal in question? Has that been ruled out?

Obviously such integration errors occur in reality. Though, in therapy I'd expect the rates to be quite low. It'd be interesting to find out how often it happens in these applications.


Integration errors aren’t the main concern with treating patients - it’s the off target effects. In other words, changes to genes you didn’t expect. You don’t want to fix one problem, only to cause something else.

IIRC, the reason why the integration errors were such a big deal is because they made it to the germline of the animal and could be transferred to offspring. This would have been extensively validated given the amount of press and regulatory pressure.


superheros?

Well, the Milkybar Kid is one of the few superheroes whose origin story has not been done to death in the endless reboots that afflicts the genre, so that could work as a good story…

She is literally GMO.

So can we stop it with the anti-GMO-at-all-costs worldview?


Editing human DNA will intensify the debate compared to editing crop plant DNA, I am sure.

It's a little easier in this case since it's not a heritable change (only bone marrow DNA was edited), but as others have pointed out elsewhere, heritable changes would be a much bigger deal.


I'm generally pro GMO food, but that's not a sound argument at all.



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

Search: