First UK child to receive gene therapy for fatal genetic disorder is now healthy
livescience.com
livescience.com
> The new gene therapy [...] works by inserting into the body working copies of the genes that are faulty in MLD, thus restoring the ability to break down sulfatides.
How does the new copy of the gene get into every existing cell that needs it? A virus?
In most cases, the original disease-causing stem cells are effectively eliminated by the conditioning regimen (high-dose chemotherapy and/or radiation) used before the (autologous) transplantation, so they do not regenerate and cause further harm to the patient. However, in some cases, residual disease cells may survive the conditioning regimen and persist after transplantation, leading to disease relapse. This is more likely to occur in patients with aggressive or refractory disease or in those who receive a mismatched or haploidentical donor graft. To prevent disease relapse, post-transplantation maintenance therapy, such as immunosuppressive drugs or targeted therapies, may be used to eliminate residual disease cells. Additionally, close monitoring of the patient's blood counts and immune system function can help identify and manage any signs of disease relapse.
This paper seems to have more details on the actual procedure
Patients were treated and monitored according to the schedule described in the appendix (p 17) and as previously reported.16, 17 HSPCs harvested from bone marrow or mobilised peripheral blood (MPB) were transduced with clinical-grade lentiviral vector encoding human ARSA cDNA under the control of the human phosphoglycerate kinase gene promoter.
In general, gene therapy is probably most useful currently for diseases that affect the bone marrow (and the eyes), because these are two tissues where you can precisely deliver the virus. It's still challenging to control where the virus goes and which cells it infects if you just inject it straight into the bloodstream.
I would also have to imagine they would have to do myeloablative chemo or radiation to make sure the fixed cells propagate more than the diseased cell line.
Edit: read the study, they do give them the same chemo used for normal transplants.
(fyi a decent amount of stem cell transplant survivors do not have to take immunosuppressants for life)
https://ashpublications.org/blood/article/104/12/3501/89040/...
>Previous studies have shown that 30% to 70% of all patients surviving beyond 100 days after HCT require treatment for chronic GHVD,2,4-6 often for more than 2 years.
Re: a better system is where you don't need donors:
https://www.cbsnews.com/newyork/news/l-i-woman-dies-after-ma...
One of my friends gave his wife a kidney which allowed her to get off dialysis, so yeah, thanks to anybody who is willing to do this for someone they don't even know.
Hmm...
How does this translate to other places? E.g. can you only match Han in China?
Do Italians and Scots match? Turks and Egyptians?
What matters isn't really the proportion of the subpopulations that are registered, but the absolute numbers.
What's horrific to me is that matching by race is still a thing! These genes are under diversifying selection. That means that they often do not match the surrounding genetic background. That means that racially driven matching is bonkers.
usually a massive dose of radiation, right?
This does not imply that germ line cells were changed and COULD mean the person could still have a high risk of making children with the original gene defect.
> Libmeldy is made using stem cells that are derived from a patient's blood or bone marrow and can give rise to different types of blood cells, according to the European Medicines Agency(opens in new tab) (EMA). These stem cells carry the new, functional genes into the body, where they give rise to white blood cells that travel through the bloodstream.
> In clinical trials, Libmeldy offered clear benefits to infantile and juvenile patients who hadn't yet developed MLD symptoms; these patients were able to break down sulfatides at normal rates and showed typical patterns of motor development, for example. The benefit of the therapy seemed to last several years, but at this point, "it is not yet clear whether it will persist life-long, and extended follow-up is needed," the EMA noted.
> Libmeldy is approved for use in the European Union and U.K., although the U.K.'s drug price watchdog initially rejected the therapy due to its hefty list price of £2.8 million ($3.4 million at today's exchange rates), BBC News(opens in new tab) reported in 2022. The therapy's manufacturer, Orchard Therapeutics, then offered Libmeldy to the NHS at a significant discount.
> The gene therapy has not yet been approved by the U.S. Food and Drug Administration(opens in new tab).First gene therapy for Tay-Sachs disease successfully given to two children - https://news.ycombinator.com/item?id=30408104 - Feb 2022 (77 comments)
“ Mutations in the VHL gene cause von Hippel-Lindau syndrome. The VHL gene is a tumor suppressor gene, which means it keeps cells from growing and dividing too rapidly or in an uncontrolled way. Mutations in this gene prevent production of the VHL protein or lead to the production of an abnormal version of the protein. An altered or missing VHL protein cannot effectively regulate cell survival and division. As a result, cells grow and divide uncontrollably to form the tumors and cysts that are characteristic of von Hippel-Lindau syndrome.”
https://medlineplus.gov/genetics/condition/von-hippel-lindau...
Like, it is very, very small thing. How do you even measure it, let alone replace it?
Measuring is an absolutely fascinating procedure. The predominant method is called sequencing by synthesis. Illumina is the company that basically rules the market for this technology, they have some decent explanatory videos if you’re interested.
The first steps are to isolate DNA, then break it into small bits, replicate those small pieces thousands of times, and then anchor these to a glass slide. Then flow a solution over the slide with the individual DNA bases (ATCG) that have been modified to fluoresce when illuminated with a certain wavelength of light. These fluorescently labeled nucleotides get assembled one by one into strands that match the small fragments from the DNA sample. Each time the fluorescent DNA bases are added a laser is shined on the glass slide and a very sensitive microscope camera reads how many strands shone at the frequency that corresponds to A T C and G. This tells you which base just got added to which strand. Repeat a couple hundred times and you have a huge number of sequences of ATCG’s corresponding to each short strand of DNA in the sample.
The problem now is assembling these short sequences into a full genome. Basically, the puzzle is finding the arrangement of short sequences that maximizes the overlap between them and their agreement with a reference genome. It’s a pretty interesting problem that (I believe) is classically solved with dynamic programming, although I don’t know the most recent methods.
I really don’t think it’s well enough appreciated how integral computation is to DNA sequencing. Signal processing in the camera, assembling the genome fragments, determining whether a mutation exists, and all the subsequent analysis to give biological meaning to a sequence.
Edit: editing is a whole other, extremely complicated procedure. In fact, direct editing isn’t really possible now. Instead we can basically just delete and insert.
This is the kind of content I visit this site for.
This formulation has the added benefit of not being easily answered with a flippant remark.
Don't take the chance, people. Edit your DNA with notepad.
We have a set of tools that let you determine biological structures (sometimes entirely, sometimes only partially).
We also have a set of tools to modify genes. Some viruses already do this, so we use them. Were also using CRISPR which is based on the machinery of bacteria.
So you use the modification tools to change genes and the analytical tools to prove you did it correctly.
Of course the level of complexity is way higher since your dealing with a living organism and the modification tools (and analytical tools) aren't perfect, so there is a ton of trial and error to do to correctly.
> Roughly six months out from treatment, "Teddi is a happy and healthy toddler showing no signs of the devastating disease she was born with," the NHS statement reads.
and later
> In clinical trials, Libmeldy offered clear benefits to infantile and juvenile patients who hadn't yet developed MLD symptoms; these patients were able to break down sulfatides at normal rates and showed typical patterns of motor development, for example. The benefit of the therapy seemed to last several years, but at this point, "it is not yet clear whether it will persist life-long, and extended follow-up is needed," the EMA noted.
Which is much more nuanced.
With genetic diseases often the difference between healthy and not is a single enzyme is missing.
If you can prove the enzyme level is restored and all the downstream effects match healthy individuals, youve basically restored them to what healthy people are.
Durability of the change is the big issue. Some gene therapies wane over time.
And of course if damage had occurred before the gene therapy, that may never go away.
Of course, it shouldn't have to be either/or. Unfortunately, the institution (NHS) has been driven into the ground and we're all paying taxes for practically nothing in return.
> The National Centre for Pharmacoeconomics (NCPE) in Ireland recommends "that atidarsagene autotemcel not be considered for reimbursement unless cost effectiveness can be improved relative to existing treatment."
Wow… instead of a lifesaving cure they recommend the treatment of the symptoms until the kid dies because it’s cheaper.
Somewhere in the middle is where most people disagree.
Twenty years ago I had Hodgkins Lymphoma. Obviously the NHS fixed that. This was doubtless quite expensive, it needed a team of experts, expensive chemicals and radiotherapy machinery. But, if they don't they're out a young productive software engineer because Hodgkins kills people, even assuming you're so callous your response is to just immediately put me down like a dog so as to reduce the cost of palliative care or diverting resources from relatives - Whereas since they fixed it that's maybe 30-40 years of economic productivity and ensuing tax income they're getting.
Now, obviously not every such intervention is literally profitable in a financial sense, but I chose my case because even boiled down to simply money this actually can make sense anyway. If you agree it made sense to fix me (and I don't see why you wouldn't) then beyond that it's only about value judgements.
You can't have it both ways and preventing people from making money means, not surprisingly, they will be very hesitant to spend money researching those areas.
Even if it were true, turning a discovery into a distributable, replicable, intervention is the opposite of a trivial task.
And yes, our lovely regulatory environment makes it enormously expensive to study drugs and treatment and to get them approved. Who knows how many lives have been lost by this.
When people talk about for example university research driving technology, I think the practicalities of the real world … aka business aka living in 4dimensional reality … are inaccessible to them. They elide all these pesky details which represent like 98% of what it takes to actually do anything.
"Hey, that's a nice life-saving treatment you spent $500M developing. It would be a shame if we just took it from you."
Expect tarrifs, sanctions and trade secrets so you won't be able to reproduce future treatments.
You might not like the utilitarian approach but this is how the UK measures effectiveness.
Ireland could simply let a local company violate the drug patents. A country like the UK could impose conditions such as profit caps on pharmaceutical companies who base their work on publicly funded research. We could reduce the length of drug patents. There are many, many options besides “role over and let pharma companies charge $4 million/treatment when it costs them $1 million.”
So it does seem like they've already discounted it heavily.
The humanitarian approach is to save that one live AND improve the other 100 as well. We can afford to do so, because those expensive cases are rare.
3,100 cases a year in all of Europe, and 3,600 cases a year in the entire US. That's pretty damn rare really. And it's a genetic condition, which means we should be working to eradicate the problem entirely by screening parents and gene editing ideally making it both rare and increasingly rare from here out.
If you're spending $4 million on this you're either raising taxes again (and government expenditure is already roughly 50% of gdp in most developed European countries, cant go much higher without destroying the thing thats making the taxes) or you're not rehabilitating 5 homeless people, educating 5 special needs kids and discovering a potential new treatment for 1 rare disease.
Let’s suppose a drug company is setting their price to maximize revenue.
Suppose they make the following projections:
They determine that if the treatment price is $10 million above the actual cost of providing the treatment, 5 people will buy it. $50 million total profit.
If the price is set so the profit per treatment is $1 million, 100 people will use it. $100 million total profit.
If the profit per treatment is $100k, 800 people will use it. $80 million total profit.
If the company isn’t factoring in the value of a life saved, they will pick the $1 million price point. If ethicists then just run with that price, they may come to the conclusion that the treatment isn’t cost-effective. However, they are relying on data that’s an output of a process with conflicting values, and that pollutes the result of their calculation. Garbage in, garbage out. The 700 people who didn’t get treated lose out for a pretty bad reason.
We could imagine a policy where the drug companies are mandated to maximize lives saved when setting prices. One might argue that companies won’t develop as many drugs if profits can’t be maximized. We could adjust the policy to subsidize companies for income lost when setting lower prices. Ie. if the drug company picks the $100k price point to save 700 more lives, the government gives them $20M compensation so they can profit like they would have at the $1M price point. That way society spends the same amount of money on this drug, but more lives get saved. I’m sure there’s a lot of challenges in designing a program like that, but the opportunity to save lives makes it seem worthwhile to attempt.
No policy decision can change that.
I suspect we can somehow find enough pennies in the couch cushions to get those kids a therapy, especially if it's curative.
People are dying due to resource rationing and it occasionally makes the news.
Healthcare is expensive and a large proportion of taxes go to it.
If you buy 1600 doses of it at a time they're going to sharpen the pencil for you.
Inflation caused by excess money supply (NB: not inflation caused by excess money velocity) cannot survive taxation. Taxation destroys money supply.
It’s no more (or less) complicated than any of the thousands of other fiscal decisions a government makes every year.
There’s a fallacy alluded to in the statement “it’s a very complicated problem” - the allusion is to one of inaction being a safer default choice but nothing about inaction in the face of a complex problem guarantees against making things worse
By qualifying with “complex” problem i mean a problem that’s as complex as choosing what to fund in a national budget. A problem for which no answer can be proven correct. A problem which can only be addressed by taking a considered view on risk. A problem with unresolvable uncertainties.
And if you print money and inflate then the price of goods rises. So sure, you might only pay a bit more for the treatment but now fruits are more expensive and your population gets slightly more unhealthy.
>> >> Inflation caused by excess money supply (NB: not inflation caused by excess money velocity) cannot survive taxation. Taxation destroys money supply.
It depends - in the specific scenario of an incredibly expensive drug for small numbers of patients (and other problems with similar dynamics) then tax might not be the best solution. But in other cases yes.
In this particular case the returns would go to successful investors - not historically a money user class prone to causing inflation (quite the opposite). In this case you may find that you don’t need any intervention - they might take their winnings and go speculate on more investments (great!). It may be that conditions are such that they don’t presently fancy investing, perhaps you could offer them inflation tracking risk free gilts/tbills to park in for now. However you would need to be ready for their choice to go on an inflationary non productive consumption spree. In which case - directed taxation should be applied.
I think a big problem is you don't currently realise just how many different worthy causes are currently underfunded. At a guess you could probably spend a couple of trillion a year in the UK before you've exhausted most of the nice things that could be done. There are sooooo many good things to do and only so many person hours (which is what money pays for mostly) to do them.
If you did scrap the defence budget then the money shouldn't go on treating these kids but instead should be invested in the surgery backlogs, ambulance wait times and the justice system trial backlog.
Look at the Blue Cross Blue Shield plan for Michigan. It covers a number of gene therapies at a cost of $2.2M for one of them.
Not all European countries pay for it.
https://www.bcbsm.com/amslibs/content/dam/public/employers/d...
The economics would never work out unless premiums were sky high or the insurance is mandatory or across most of your population.
It only works if premiums are spread across a large population so a $2.5M therapy becomes a $1 increase in premium for 2.5M people.
Best of the both worlds.
So, depending on where you are, private healthcare means the government pays their share and you pay the rest through an insurance or directly. Again depending on where you are the insured is often provided by your employer, so if you happen to get ill you can have your treatment in a luxury.
In the UK I earn a decent wage (well more than 80% of people in the UK), and last year I paid less than £3k / $3.5k for healthcare which covered me, my wife, my 2 children. It also covers my 70+ year old mother and my 90+ year old grandmother who don't earn enough to pay income tax.
80% of people pay far less than I do.
What type of health insurance can I buy in the US for $290 a month to cover a family with no co-pays.
>>you have to fund the super expensive good for nothing public one anyway.
My father had battled cancer in Polish oncology hospitals and has received nothing but top tier treatment wherever he went. What's more, NFZ completely funded the treatment with then-experimental Glivec for him, which was billed at around 200k zł/year(in US, a year of Glivec treatment is around $100k). In the course of 8 years of cancer he had to stop working due to the ilness, and yet was still fully covered and received the same treatment as always.
So no, it's not "expensive but good for nothing". It's not expensive compared to private treatment, and it's extremely good if you have a major ilness. I agree that routine treatment and appointments are in shambles, but that's universally true pretty much everywhere in Europe right now, the waiting times in countries with public healthcare systems have inflated into infinity due to a combination of factors. But even in Poland, I wouldn't dare to not pay into the system - it's too good not to pay for it(and I do, in fact, pay for it).
> but that's universally true pretty much everywhere in Europe right now, the waiting times in countries with public healthcare systems have inflated into infinity due to a combination of factor
I have heard that's not the case in Germany and the private appointments are almost not a thing. I could be wrong though.
https://www.icalculator.info/poland.html
Says that people pay 1.25% in health insurance.
Poland in general pays 6% of its GDP on health care. The US spends 3 times that.
Rules for life insurers are different and in some circumstances they may deny coverage.
If given more money then that value can go up, but the UK economy is in trouble enough as it is with the state taking half of GDP.
That's the same sort of ethical calculus that goes into deciding the distribution of any life-saving resources.
We can always spend more (resources/social capital) to save more lives/years of life. How do you decide where that line gets drawn?
Most medical systems decide it by looking at the cost of treatment (Dollars, organs, risks), and quantifying years of quality life gained. If you'd like that bar raised, increase the tax rate/insurance costs.
'Save every life at any cost' is not compatible with a world where you have decide whether your budget goes on a low-ROI, or a high-ROI intervention. 'First come-first-serve' is not compatible with a world where you are optimizing for overall positive outcomes.
Do you have any alternative guiding principle for where money in the healthcare system should be spent?
It's worth mentioning that the study in Ireland indicates that the treatment extends life by 14.49 QALYs (average "Total Life-years" moved from 8.92 to 22.74), which is a long way from a cure. If this is truly a cure, and the treated population lives a full life (life expectancy in Ireland is current 82 years, not 23), then this treatment will become cost effective without any change in the treatment or it's cost. The posted article agrees that this is the big open question.
> The benefit of the therapy seemed to last several years, but at this point, "it is not yet clear whether it will persist life-long, and extended follow-up is needed," the EMA noted.
Even if the estimate is accurate, there is a massive qualitative difference between slowly dying horribly for 'x' years and living a normal life for 'y' years. You can't just subtract 'x' from 'y' and come up with a delta and compute based on that.
Reminds me of the studies that showed that Tamiflu is ineffective because it only reduced the duration of influenza symptoms by 1/2 a day on average. Yes, that's true, I've taken it myself and the symptoms continued for about the normal period of time. But it reduces the severity massively. It's like a light switch. The most severe flu suddenly turns into the mildest of mild cold-like symptoms in a matter of hours. But... that's hard to measure objectively, so it is not recommended because the according to a metric that doesn't matter it doesn't work.
The statistics and metrics in medical papers are woeful, which is why it's commonly accepted that 1/2 of all medical research is false.
Ok, this probably needs some extraordinary evidence …
> which is why it's commonly accepted that 1/2 of all medical research is false.
The irony
The difference in measured in QALYs - quality-adjusted life years. The study attempts to take that into account.
> They can't possibly know that the drug extends life to "22.74" years, because it has only been approved for use for the last 3 years! This is like asking for 30 years of Kubernetes experience on a job application.
Yes, exactly my point. The long-term effects of this treatment are what will make it worthwhile or not. Trying to extrapolate a few years to a full life does not make that an easy exercise.
If only there was a measure that was adjusted for the quality of the year. Some sort of quality-adjusted life year.
Yes, it's not a perfect measure (how could any quantitative measure of a qualitative thing be perfect?).
Compromises always have to be made because the world does not have infinite resources. Unfortunate, but true.
One way to allocate resources between all our competing needs is by using money.
You can come up with other systems of allocation, however all systems will be unfair and arbitrary in surprising ways. There is always a need to choose between ugly choice A and ugly choice B, because our resources are constrained.
You have X amount of dollars. You can spend Y percent on healthcare. How do you spend Y to have the greatest benefit? Save 1 patients with a $300,000 gene therapy, or save 1,000 patients with effective treatment for their diabetes.
Until we have unlimited resources, that trade off will always be made.
Holy shit. My god, I'm just thinking of the parents.
"Sorry, we can only save one of your kids."
"You can't give Nala the treatment?"
"Well, we could. We won't, though."
I'm sure there are good reasons, but Jesus, that's harsh.
Does that mean clincal trials didn't show a benefit for patients who had already developed symptoms? Does it mean it wasn't studied in patients who had already developed symptoms?
I don't think it's unreasonable that the NHS won't spend large amounts of other people's money on a treatment that hasn't been studied or wasn't found to be effective.
It likely wasn't studied.
It's a pretty new treatment, so the vendors start with clinical trials for use cases where the clearest/greatest impact is expected, and maybe later on move to related use cases.
In the case of the child who already shows neurological symptoms, the treatment can help stop the progression of the disease, but most likely not reverse the already existing damage. However, there are already treatments approved to help stop the progression of the disease, including enzyme replacement therapy (provide the enzyme in charge of degrading the sustrate through the bloodstream every so often).
Some would argue that if treatments with similar outcomes already exist and are approved, it could be unethical to test this experimental treatment on a patient (I disagree).
King Cnut demonstrated it with the tide.
Just shockingly amazing!
'Only' the mass extinction of species part isn't fixable without clarketech.
Geoengineering can stop global warming if necessary, but bringing back a destroyed ecosystem is effectively impossible.
I wouldn't make any bets based on Musk's posturing about Mars.
In general if Elon says it'll be done in two years time you can be confident it will take at least 3 years. Average of 4, but possibly more than 10