70 comments

[ 0.20 ms ] story [ 129 ms ] thread
According to the article, Chinese firms are making advances in chimeric antigen receptor cell therapy (Car-T) to treat cancer.
I would love to be able to read this.
The people who spent the time and effort to do the research and write the story would like to get paid. Seems like there is a simple solution here...
(comment deleted)
I count no fewer than eight options on the landing page that would enable you to read this.
This should be the article linked. Mods can you switch, the current one behind a subscription wall?
It's ok to post workarounds in the threads, but we don't change the official URLs to those. Doing that would obscure the article source, among other reasons.

https://news.ycombinator.com/newsfaq.html

https://news.ycombinator.com/item?id=10178989

Hi dang. I really do not understand why it is ok to post pay-walled articles. It makes the frontpage a minefield IMHO. Maybe a short reason can also be stated in the FAQ in addition to that being ok. BTW, is it also ok to post articles behind a registration-wall? I mean, like something shared on FB which can only be viewed by members?
It's explained in the second link there as well as tons of comments I've posted about this (https://hn.algolia.com/?sort=byDate&dateRange=all&type=comme...).

We find the paywalls as annoying as anyone, but as long as there are publicly available workarounds, the articles remain accessible at the price of annoyance. This sucks, of course, but an HN without articles from any such publications would suck worse. Since the core idea of HN is to try to optimize for curiosity, we don't have much of a choice. Hopefully someday the publication problem will get a definitive solution and we won't have to deal with this crap anymore. In the meantime, publicly accessible workarounds are the best of a bad bunch.

(Re registration sites, it's the same principle: if there's a publicly available workaround, it's ok, if not it's not.)

you know i actually think you’re right and something could be done. i literally always open the comments first because something like a quarter of articles here are paywalled and they always have a mirror by the time i read them.

my system works and it seems like people don’t like your sentiment. but i don’t understand why.

Well, I just found my new favorite extension.
Physician here: CAR-T is a potentially amazing technology, but really only works in a small sub-set of the patients for which it is tried. For now
Normal person here: why?
Bad reddit - why aren't physicians normal people?
In this context, pretty clearly, "normal" means "having a layman's understanding of medicine". Physicians had better not be normal (in this sense).
Just took the words out of my mouth, thanks ^^
This far it only works in blood cancers, specifically B cell lymphomas. The approved CAR-Ts target a molecule called CD19 expressed on B cells

It has not worked widely in solid tumors for a variety of reasons, including immune suppression within the tumor microenvironemnt and difficulty finding antigens that are widely expressed on tumors but not also expressed on healthy cells (so you don't kill healthy tissue in addition to the tumor)

Maybe if you tried it as a first-line treatment instead of on relapsed basket cases it'd work better.
yeah, when i was working on CAR-T stuff a few years ago this was the thing we always said.

most of the mouse models that we prove the therapies in (at least to start) are treatment-naive, which makes a difference.

Tox is very severe and the cost of drug plus managing the side effects is massively expensive. So the risk / benefit and health economics are worse in early stage cancer. They'll study it in earlier stage at some point probably
That's not entirely true. It's very effective against blood cancers but less so for solid tumors, where its effect seems to interact strongly with aspects of the tumor microenvironment.
As the father of a kid that died of bone cancer that is par for the course. Other cancers have made HUGE improvements but the hard dense sarcoma tumors still sit in the 1980s in terms survival rates.
I'm very sorry to hear of your loss. I can't begin to imagine.
An unimaginable loss. I am so sorry.
My commiserations
Can anyone explain how you go about re-engineering a cell to fight off cancer?
ELI5: Your immune system fights cancerous cells frequently. There's only a problem when the cancerous cells evolve to avoid this by "hiding" (typically by no longer releasing a "kill me" marker). The therapy basically gives your immune system the tools to identify cancerous cells.
(comment deleted)
T cells kill infected cells by detecting foreign proteins using a lock and key tube mechanism (Key = foreign antigen, Lock = T cell receptor). Normally any given T cell an its off spring have only one specificity of lock (although collectively there are many millions of locks in the body). CAR-T therapy is when they pull out a bunch of T cells for a patient and genetically modify the cell to have a know lock which is specific for a key known to be on the Cancer, then put them back into the patient. So, relatively, there is a massive increase in the frequency of T cell with that specificity in the patient, and hopefully they kill the tumor cells.
Cancer = you, but misregulated. Code for various tissues are turned on at 'incorrect' times, including code for profliferateNow(), ignoreAutoDestruct(), and can start running contradictory code like beNeuronLike() and beMuscleLike() simultaneously.

T-Cells generally sense 'rogue' cells, and command them to auto-destruct. Many cancers are rendered invisible to your immune system because their presentation to the immune system is indistinguishable from a healthy cell in that it doesn't present any feature that is not found elsewhere in the body.

In an engineered Car-T cell, an immune cell is reprogrammed by providing the cell with a DNA blueprint for a synthetically engineered protein called a 'Chimeric Antigen Receptor (CAR)'[1]. That engineered protein has two major functions - sense a particular cell type, and activate the T-cell to kill whatever triggers its sensor. The sensing mechanism is relatively arbitrary from an engineering perspective, so we engineer the sensor to sense a particular cell type which has become cancerous. And the attack trigger is a shortcut of the entire immunological activation pathway - skipping straight to sense->attack, bypassing the natural defenses to prevent autoimmune problems.

Because the sensing mechanism of the synthetic protein is modular, these engineered proteins can swap out various 'sensors' to create new therapies relatively easily. Sense any cell that looks like a [lymphocyte/B-cell] and [kill it]. More sophistocated versions of these synthetic proteins can actually start to incorporate and-gates and other formal logic into their sensing capabilities:

if(sense(neuron && muscle)){ killTarget(); }

(where no neuron-like markers should ever show up with muscle-like markers except in a genetically scrambled cell, i.e. cancer).

[1] https://serotiny.bio/notes/proteins/car19/

> Car-T is at the cutting edge of biology: it involves extracting a patient’s blood cells, re-engineering them in a lab so they can identify and destroy cancer and then re-inserting them into the body.

It would be interesting to see if we could one day make this more optimal by reducing the extra step. Changing the above to:

> Car-T is at the cutting edge of biology: it involves re-engineering a patient's blood cells in a lab so they can identify and destroy cancer in the body.

Re-engineering a patient's blood cells means first having a patient's blood cells. At some point, some cells have to be extracted for the cells to be available.
That would be cool, but also creepy if my blood cells can destroy cancer in my body while they're still in a lab. What if they start destroying cells in other people's bodies?
This is the end-goal. The delivery of synthetic genetic payloads in a live organism is still a bit tricky. Certain therapies do do this by injecting a synthetic 'virus' directly into your bloodstream. However, since it's so easy to extract T-cells, the risk of withdrawing blood, engineering it, & replacing it, it is worth the extra time to guarantee the delivery of the genetic payload is correct and thorough.

As the sophistication of genetic delivery improves, direct delivery will absolutely become more common - and in fact will likely be a required technological step to curing harder-to-reach cancer and genetic abnormalities. That's why the first gene therapies are for the immunologically privileged, and easily accessible sites of the blood and the eye.

> But the most eye-catching research at last year’s conference was presented by a Chinese biotech group that virtually no one had heard of: Nanjing Legend.

I don't mean to belittle this achievement but when will media learn that there are no "unheard" companies in China. Only those which have government backing and those which don't.

While this an amazing story, I think it will be best to wait before pronouncing that there is a boom or not.

Your second and third paragraphs have no relation to each other.

The second paragraph comes from some tabloid about China.

The third one just says “wait and see”

Unlike other medical trials involving chronic illness, cancer trials have a pretty clear indicator of whether the treatment is successful- (most) the patients survive longer than their prognosis.

In this sense, there's not much freedom for the Chinese biofirms to commit fraud during a treatment trial. Everything is verifiable.

Why would falsifying that data be any harder? Don’t you just substitute patient lived for patient had less headaches or whatever?
Its hard to fake that a dead person is still alive. Its easier to say that they reported no headaches.
You can always fudge your trial result. What I am saying is that the line between an effective treatment and an ineffective treatment is very clear- more patients are alive by the end of the trial. Versus, say, a test for a new antidepressant drugs, where being 'happier' is hard to quantify.

Everything else is just record tracking.

"80% of China’s clinical trial data are fraudulent, investigation finds"

ust over 80% of clinical trial data submitted to support new drug registrations in China have been revealed as fraudulent or substandard by the country’s drug regulator.

An investigation of data for 1622 new drugs submitted to China’s State Food and Drug Administration (CFDA) for registration said that 1308 of the applications should be withdrawn because they contained fabricated, flawed, or inadequate data from clinical trials.1

In a damning report released on 9 September the CFDA described the findings of widespread fraud as shocking and vowed to crack down on what it described as a “chaotic” situation in the country’s clinical trials industry.

Much of the blame was attributed to the country’s loosely regulated clinical research organisations, which industry insiders claimed operate in …

https://www.bmj.com/content/355/bmj.i5396

There's a Vice on HBO episode where they covered the Chinese practice of "renting a white guy" [1]. I mention it here because in the segment there was a white expat employed to masquerade as a doctor giving presentations on subjects he had zero qualifications for. They used his foreign appearance and language to lend credence to what was being presented, it's a disturbing scene.

[1] https://www.youtube.com/watch?v=2jGSerN_1sM&t=14m

> Everything is verifiable.

Nah. If there's a will, there's a way.

Trustworthy data is more of a social contract thing than anything else.

You could remove individuals that didn’t pass your trial.
What will happen to american drug cartel now? FOr that matter china would be having its own drug and animal-organ-drug cartel.
I wonder if a similar technique to car-t could be deployed to fight off autoimmune diseases?
Sighhh... apparently they can't just report on a story from China without peppering it with statements about the US -- specifically about how this is a new salvo in some kind of healthcare war with the US. Sometimes stuff just happens when people do things. They're probably not pushing to advance healthcare just to fuck with the US.

Some made-up headlines from world history written from this incredibly narcissistic viewpoint:

Pharaoh to Build Great Pyramid, Challenging Structure-Building Dominance of Future Country "America"

Scientist Marie Curie Discovers "Radioactivity" in Effort to Humiliate USA

In Rare Reversal, Da Vinci's "Mona Lisa" Better Than Similar American Art Works

thanks for making your point with humor, it’s underused on hn. i agree with the sentiment too
An outgrowth of American exceptionalism.
Not to say that they're by any means perfect, but I think what you've pointed out might help explain why many Americans i grew up around see the Chinese government as some sinister and inscrutable entity that is out to silently crush everyone. Also, you don't need to make up fake examples when you can look up anything in the past year about how they're making big strides towards reversing their smog problems and how that's somehow an attack on American industry. Even trump has made these claims ( https://lh6.googleusercontent.com/uDFmIDPzCo4znxgzKlT0_jgv4C... )
Younger me was a news junkie. Iran hostage crisis, USSR was gonna nuke us, Japan was going to buy us, etc. Of course, I believed it all. Over time, I noticed shit just didn’t add up. And no one did a mea culpa.

So I just stopped listening.

I figure all ad supported news media is ghouls profiting from human misery, stoking fear and anxiety to keep people numb, allowing the real outrages to happen unnoticed. (eg Project Censored)

I hope China continues to do well. Ditto everyone. We all do better when we all do better.

That sounds a little bit like survivor bias to me. Nothing bad happened to the US, but in other countries shit went down in the last century, stuff that if you were not informed and aware about, might lead you to... for example go to school in the morning and realise you can't go back home since you're on the other side of a new border.
Been chewing on your reply.

Plenty bad happened in North America. Project Censored is a good starting place, if you want to raise your blood pressure.

I still seek news. Mostly from HN, Democracy Now, Vox, misc. It’s just that now I consciously block any source that I feel is more noise than signal.

Noam Chomsky pointed out that, for the last century, US has been the most secure country, but its people are the most insecure.
Yeah I see what you're talking about, though I would add that there's a whole sub culture in the news about pointing out bad takes from among the fourth estate. It all seems to depend on where you "hang out" so to speak.
This might happen more often in the near future. Whilst medical innovations are booming due to the introduction of cheap options to run machine learning the regulatory units are not keeping up. Getting FDA approvals is a long process but required before they let you enter the US market. EU is easier but soon they will put in place a new regulatory unit that is said to be even more restrictive than the FDA. When that happens it may very well be the Asian countries, China being a good option, India another, that take the lead when it comes to innovation in the medical world.
This would be great if it were true. Unfortunately, it’s a clinical trial from China.
> A few days before Craig Chase was discharged from the Jiangsu Provincial People’s Hospital in Nanjing, his doctor told him something he never expected to hear: his cancer had been cured.

> “His English was not so good, so he used Google Translate. When he said I was cured, I told him it was impossible — there is no cure for multiple myeloma,” he recalls. “But he said, ‘no, you’re definitely cured’. It was unbelievable.”

> Unbelievable, perhaps, but also true. When Mr Chase, now 57, returned home to America after his six weeks of treatment in China to undergo further tests, his doctors could find no trace of multiple myeloma. The blood cancer he had suffered for three years — which had threatened to end his life — was gone.

This is incredible!

If they can figure out how to cure cancer, then how will the pharmaceutical companies make money?

This sounds like a diabolical plan for.. Lex Luthor!

Lex: "Destroy the cure. Instead, make a drug treatment plan, that the patient must take every week, for the rest of their life."