First human Crispr gene-editing trial gets early approval
wired.co.uk
wired.co.uk
This is misleading, right? Yes, it's true that Stage I trials are designed (and in particular, statistically powered) to test safety rather than effectiveness. But that doesn't mean they are making gene edits with no therapeutic purpose. That would be highly unethical. The gene edits are definitely intended to be an "actual treatment". It's just that the number of subjects is so low there isn't any hope of measuring the effect.
They'll be especially careful with the CRISPR drugs due to the disasters that Gene Therapy trials have previously undergone. Most notably, TGN1412 was given at a dosage about 500x smaller than the safe dosage in animals but caused almost immediate cytokine storms that nearly killed all 6 participants in the trial.
OK, good point. But even then there needs to be a reasonable belief that either (1) the likelihood of clinical effectiveness is nonzero, and is still worth trying compared to the smaller chance of bad effects, or (2) the same patient will be able to receive a clinically effective dose in the future based on the safety information gained testing the smaller, clinically ineffective dose on them.
The thing that puzzles me most is that if we do it by creating some cells and injecting them, how do we guarantee that these cells replace their existing unmodified counterparts?
This should also answer your second question: You target some cells higher up in the hirachy like CMPs or even MPP and HSC (stem cells). Other tissue, other stem cell type. And ultimately obviously some lineage commited (cells aimed to become a certain cell type) iPS cells (which are stem cells generated from normal body cells). Thats all possible, the only thing really missing is good understand of epigenetics (only a question of time) and some way for reliable epigenetic programming/modification. (Sorry if thats too short too complicated but maybe it guides you in the right direction)
Edit: And third question: I am not entirely sure why you would need to replace all the existing cells. Some additional functionality might be sufficent. Where this is not the case, you usually go for irradition (and full replacement).
>This should also answer your second question: You target some cells higher up in the hirachy like CMPs or even MPP and HSC (stem cells). Other tissue, other stem cell type. And ultimately obviously some lineage commited (cells aimed to become a certain cell type) iPS cells (which are stem cells generated from normal body cells). Thats all possible, the only thing really missing is good understand of epigenetics (only a question of time) and some way for reliable epigenetic programming/modification. (Sorry if thats too short too complicated but maybe it guides you in the right direction)
If I understand you correctly, we can replace the cells in charge of producing other cells (stem cells) to make the changes lasting, but we are not yet certain how the changes will play with the rest of their environment (epigenetics).
>I am not entirely sure why you would need to replace all the existing cells. Some additional functionality might be sufficent. Where this is not the case, you usually go for irradition (and full replacement).
To make an example, let's say I'm a secret ginger carrying the mutation that makes me far more susceptible to skin cancer[1]. And let's say we know what exactly to change in DNA to fix my problem. Now, I don't know how the skin producing machinery works, but I guess replacing it in one go isn't an option. So, if we go with changing cells higher up in skin production, can we guarantee that in due time they will replace all (or most of) the skin with modified cells? As a bonus question, since we limit ourselves to targeting only particular cell types (skin in this case) that means the rest of the body is carrying unmodified DNA, hence, those changes are not hereditary, right?
[1] http://www.telegraph.co.uk/science/2016/07/12/one-quarter-of...
Not quite, we are not yet 100% certain what these changes are (or whether we even know all epigenetic marks). We know that epigenetics determines what cell type a particular cell is, but we do not know yet what all of these are. Slight analogy: What we have with the genome is the compiled code of the program, the sum over all epigenetic marks somewhat is the state at run time.
>To make an example, let's say I'm a secret ginger carrying the mutation that makes me far more susceptible to skin cancer[1]. And let's say we know what exactly to change in DNA to fix my problem. Now, I don't know how the skin producing machinery works, but I guess replacing it in one go isn't an option. So, if we go with changing cells higher up in skin production, can we guarantee that in due time they will replace all (or most of) the skin with modified cells? As a bonus question, since we limit ourselves to targeting only particular cell types (skin in this case) that means the rest of the body is carrying unmodified DNA, hence, those changes are not hereditary, right?
That is going to be hard, as you basically want to replace a whole organ with another. The (idealistic) idea there is rather to take a virus that performs the modification in the body in all cells. But that is still somewhat in the future as you have to make sure the virus does not spread to other persons, the changes are lasting etc.
Note: What we did not cover here is the whole interplay with the immune system. It may well happen that some B or T cells recognize the modified cells as foreign (at least if the change is in a protein) and fight it, so you'd rather have the modified cells as nonessential. In the study here this is not that important as you still have other T cells and you'd be happy enough if they (the modified cells) fought the cancer for some time and then disappear.
Now you're talking. That's a brilliant analogy, I'm stealing it for the next time I explain CS to biologists, thanks!
>The (idealistic) idea there is rather to take a virus that performs the modification in the body in all cells. But that is still somewhat in the future as you have to make sure the virus does not spread to other persons, the changes are lasting etc.
Indeed, I forgot about viruses. To sum up, to get to changing adult DNA we need a) more knowledge of the "runtime"/epigenetics to know what to change, b) ways to handle viruses if we go with them. That sounds reassuring, actually, at least we might not get stuck in Gattaca like future for too long.