Hacker News new | past | comments | ask | show | jobs | submit login
'Living Drug' That Fights Cancer by Harnessing Immune System Clears Key Hurdle (npr.org)
423 points by daegloe on July 13, 2017 | hide | past | favorite | 141 comments

Congratulations! The Chimeric Antigen Receptor (CAR) deployed here is very much unlike the standard 'small molecule' drug that 'disrupts a bad thing', and much more like a rationally engineered tool using the body's very own technologies to overcome a particular limitation. In this case, it gives the patient's own immune system a notion of what the cancer looks like.

If you want to build your own 'living drugs' we've built a digital infrastructure to allow you. Though we just made public our generic protein design software (thanks ShowHN! [1]), we're employing the same underlying digital infrastructure to build, evaluate, and manage CAR designs in high throughput [2]. The drug approved here was painstakingly designed by hand, while we think the technology now exists to permit many more such advances to be created at a much more rapid pace.

[1] https://news.ycombinator.com/item?id=14446679

[2] https://serotiny.bio/notes/applications/car

Design your own 'living' protein drugs here right now: https://serotiny.bio/pinecone/ (and let us know what you think, and how we can make it better!)

3...2..1.. Prion disease!

Seriously though, how would you prevent such automated designing from having unintended side effects?

The software is useful because it allows you to have some idea of the function of each component prior to their assembly. Further, at this point, the software is designed for R&D far far upstream of being put directly used as a therapeutic. Finally, if a design causes some problem, we want to know that - we allow that undesirable property to be discovered in the laboratory, annotated, and communicated so that others know to not use such designs.

Is it actually possible to construct proteins from functional components like we design digital logic and software? Don't the parts interact in highly complex ways that take a lot of compute time to solve?

If you're trying to compute from first principles (atom by atom), yes, exactly - that's a very hard problem. And if you're rearranging entire protein domains rather than point mutations - that's an intractable problem. And it is precisely the problem we are trying to solve - how can you have any idea what the function of a novel protein will be when you are swapping around hundreds of amino acids at a time?

We've taken a different approach. A 'small data' approach, where we manually, and semi-automatedly ingest hard-earned empirical data about each domain, and make it computationally accessible. We give heuristics and logic to a biologist's intuition about 'this protein never seems to work when it's at the N-terminus', rather than trying to compute why it doesn't work by simulating the movements of 100,000 atoms over 10ns. And though many (generally synergistic) interactions won't be predictable in this manner, many will be. We help get you to those designs likely to work as quickly as possible. And the more empirical data ingested, the faster we think we can get you there. Further, even if it's not 100% effective at predicting the single best design, enriching the search space of a large (and expensive) screen is itself valuable.

Life already mixes & matches DNA to create new functional all the time, we're just making that same process explicit and accessible. It turns out the function of any given protein domain is actually reasonably robust to being split up and rearranged. Especially when the goal is 'some function', or 'good enough', and you're not messing with or trying to tune a protein absolutely required for life.

Also a totally new class of bio weapons.

3...2..1.. Prion disease!

Can you explain that?

Prions are 'misfolded' proteins [1] that, in their misfolded form actually cause other proteins to also misfold - a physical viral cascade. A really crazy concept not too dissimilar from a kind of biological 'grey goo' - and in real life actually happens to cause diseases like 'mad cow'. The implication is that blindly engineering a protein might create such a physical virus. I would suggest that the likelihood of accidentally creating such a virus is very much like accidentally creating stuxnet. Possible, but extraordinarily unlikely. And further, would such a thing be made, we'd like to know about it while it's still in the lab and can be contained, learned about, and prevented in the future.

[1] https://en.wikipedia.org/wiki/Prion

I guess OP was alluding to diseases caused by bad protein: https://en.wikipedia.org/wiki/Prion#Prion_diseases_and_their...

A well known example is mad cow disease.

This is fascinating! As someone who knows nothing about this domain whatsoever, could someone point me to some (easy to digest) introductory material? I signed up now, but the terminology is completely alien to me..

Proteins are the machines produced to atomic specification by a genetic blueprint encoded in DNA. The are the tools life uses to accomplish most of its tasks. Proteins are formed like 'beads on a string' - a chemical single-dimensional chain, where twenty different chemical 'amino acids' are mix-and-matched. The amino acids are strung from an 'N' terminus to a 'C' terminus in order to produce nearly every machine all biology uses for every task a cell might need to survive - structural, chemical, metabolic, energetic, sensation, etc. Once produced, the chain of amino acids folds in on itself in 3D, and many 'motifs' or 'domains' have particular functions as sets of tens to hundreds of amino acids. The software, Pinecone, allows you to characterize and then mix-and-match those domains to produce protein machines with novel function. It is essentially a 1D CAD program that outputs 3D atomically defined objects, but compiles down to DNA that nearly any organism can read. It's not all that different from a molecular Kerbal Space Program - a hardware store for protein domains - but one where you can press 'print' and get your DNA delivered.

- Protein Domains: https://www.youtube.com/results?search_query=protein+domains

- How proteins are made: https://www.youtube.com/watch?v=gG7uCskUOrA

- Visualization of protein machines in use: https://www.youtube.com/watch?v=wJyUtbn0O5Y

- Economically, socially & therapeutically valuable Proteins: https://serotiny.bio/notes/proteins/

The basic concepts and terminology are a mix of biochemistry and molecular biology and are describing organic machines that are thousands to hundreds of thousands of atoms, and a couple of nanometers in length on a side.

Is there any plan to open source the software?

Not at the moment. We do love open source and have an interest in seeing what we can make open. But currently we simply don't have the resources to do so.

There is the Synthetic Biology Open Language (SBOL) that is being developed to make this kind of information interoperable. The standard is young, eager and important work that we appreciate and follow. http://sbolstandard.org/

> But currently we simply don't have the resources to do so.

Such as? I am unsure what kind of resources one would need to open source their software.

SBOL certainly seems interesting, thank you for mentioning it and thank you for your reply.

Open sourcing is not a zero sum affair for the creator, I expect licensing agreements would need to be legally cleared and signed off, there would be an expectation of documentation to accompany the release and some kind of communications after release which would call on resources.

The significant thing about CAR-T cell therapy is that it's not very specific to the type of cancer - all cancer cells have damaged DNA that leads to the productions of antigens. Leukemia is the low-hanging fruit because it's easy to inject the T-cells back into the body right where the cancers cells are. It's hard to tell whether you could get enough T-cells to diffuse out of the bloodstream to have an effect on something like prostate cancer. It would be a real breakthrough if you could overcome that hurdle, because then you would have a treatment that works on many different cancers without much modification.

The other 'cool' part about CARs is that 'without much modification' is not near so large a challenge as it is for small molecules. Certainly there are regulatory barriers, but scientifically, the ease with which you could just (genetically) pop on a custom nanobody/binding domain is many many orders of magnitude less than trying to find a small molecule to hit a different target.

In a decade or two, I wouldn't be surprised if CARs were being deployed that were entirely customized to a patient's cancer, where the scaffolding was 'FDA approved', while the binding domain was cultured to specifically react to this patient's cancer, encoded into DNA, printed and designed within the space of a few hours or days. The kind of customizability that comes from a genetic therapy opens a lot of new doors.

Perhaps we'll even be able to generate individualized tissue samples in order to test out side effects prior to treatment


I wonder though if just having these types of cells in the bloodstream could limit or even prevent cancers from metastasizing. That would provide a massive increase in survivability in and of itself.

The downside with adaptive immune system therapy (like CAR-T therapy) is that it doesn't work as well with cancers with fewer mutations that the T cell can detect as anomalous.

My wife has brain cancer and has been reading about all the immunotherapies available. Unfortunately brain cancer tends to have much fewer mutations than say, lung cancer. However there are innate immune system therapies being researched that might be more promising in those cases.

So sorry to hear this. It sounds like you guys are doing the right research. Being in a similar boat, I think that the checkpoint inhibitors look very promising. I try to keep up with the field as best I can. The hard part is you have to find an appropriate clinical trial for whatever you and your doctor think is best... Also, just as an FYI, the NCI is starting up a new precision medicine trial (NCI-MATCH) to target specific pathways instead of tumors... but I'm more of the opinion you have to do combination chemotherapy and modalities unless the drug is super effective. But then again I'm not an oncologist or a doctor.

Very sorry to hear about your wife. I assumed one of the other problems would be the blood brain barrier. Could this therapy overcome that too?

Real, actual question: would it not work to just inject t-cells deep into my taint in that case?

"local' anticancer treatment is achievable but not really effective. the promise in the CAR-T is persistence of the cells and systemic distribution so they can surveil for micro metastases and minimal residual disease, hopefully eliminating it. I'm not sure injecting a ton of cell into ones taint would give the cells the best opportunity to persist and provide systemic prevention.

What I got out of the article was that the approach here seems to be that you create a small number of improved t-cells and inject them. Then these improved t-cells will respond correctly upon contact with cancer cells. At that point they will multiply to meet the required demand. The initial injection would be more like a vaccine than a drug.

The t-cells would multiply? I thought they were created in the bone marrow. Do bloodstream t-cells go through meiosis?

i second that question. if they can carry out needle biopsies of the prostate and other cancer locations, if they can surgically remove cancerous tissue -- why can't they also get these modified cells to the cancerous locations somehow?

A friend of mine is alive today because he was part of one of the early trials. He had been told by his doctor, just before he was accepted into the trial, that he should start putting his affairs in order.

NYTimes also covers the story (https://www.nytimes.com/2017/07/12/health/fda-novartis-leuke...) with more discussion about individual patients.

From the NYT article:

> The panel recommended approving the treatment for B-cell acute lymphoblastic leukemia that has resisted treatment, or relapsed, in children and young adults aged 3 to 25.

Why so young?

> Why so young?

It's a cancer commonly found in children.


> Acute lymphoblastic leukemia is seen in both children and adults; the highest incidence is seen between ages 2 and 5 years. ALL is the most common childhood cancer, constituting about to 30% of cancers before age 15.

If someone over 25 gets it, then this method is not approved for them? That's how I read this.

That's my understanding. Generally, that means they don't have enough data on the non-approved ages from clinical studies yet. Would make sense, given it's considered a childhood cancer - they'd do trials on children first.

I'm guessing there were no patients over 25 in the trial?

Yes, patients older than 25 can still get it as physicians are allowed to prescribe off-label. Whether insurance will pay for it is another question.

i'm surprised it actually is a question. i would have thought there was no question at all and that the answer was just no.

does anyone truly have health insurance with benefits this high?

Insurance pays for off-label therapies all the time and oncology it's very common.

you can get approval on a case by case basis, from what I understand. The doctor or institution will have to send a letter to the FDA beforehand though.

There is no need for physicians to inform the FDA about off-label use. It happens thousands of times per day.

I think when I wrote this comment I forgot what we were talking about. I was thinking about drugs currently being used in clinical trials. My mistake

You are correct about the compassionate use/early access program where patients take drugs prior to FDA approval. All of that needs to be done in concert with the FDA.

A couple of things really surprised me about this article.

First, I'm sure that there is other evidence the FDA is using to determine whether or not to grant this drug FDA approval but a 63 person drug trial seems like an exceedingly small sample size to work with. Perhaps because this disease is so rare they could not put together a larger trial?

Also, it seemed a little bizarre for an FDA panel to receive comments about a decision it is trying to reach from the families of those involved in the drug trial. I suppose there is nothing wrong with that, per se, but shouldn't these types of decisions be reach on the basis of scientific evidence and strive to be devoid of any kind of sentimentality?

Either way, it's always excellent to see new cancer treatments on the market, particularly when they are as groundbreaking as this one.

> a 63 person drug trial seems like an exceedingly small sample size

Sampling is basically decided on 1) recruitability and 2) effect size. If your effect is extremely large, even with a relatively small base size you can demonstrate the benefits of your drug.

Basically very, very large trials are mostly useful to demonstrate benefits that are not that large to begin with.

It's not that small, if the statistics look good. And they really do look better than most everything else on the market.

Also, the approval isn't for patients newly diagnosed yet, it's only for patients that have relapsed or resistant to current therapies. You'll see a larger phase IV trial later, most likely.

Now, off topic from your comment, I'm worried about cost. I know that the R&D for this kind of therapy is exceedingly high, but these therapies need to get cheaper for us to be able to justify using them in a larger population.

> I know that the R&D for this kind of therapy is exceedingly high

Do you have a source for that? I happen to be involved (genetically only - my family) into medical research and everyone I know agrees that costs are vastly inflated:

* Marketing for a new drug is not "research".

* Reformulation of trial-targets is not "research", it's re-shaping of the test settings so you can get $drug to market ASAP.

* When the government/"the people" pay for research (through co-operations with universities), it's not your "R&D cost".

* When you do basic research, it's incredibly easy to claim "10k hours". Ok, but can we please claim those 10k hours only once? Not for every variation of the substance you research again and again?

(I'm a physician who used to work with large pharma companies on trial design and trial innovation.)

I think the cost of R&D in pharma that we hear about is sometimes vastly inflated and sometimes pretty accurate, but definitely oftentimes misapplied (especially by pharma).

On the whole, trials are expensive. Sure, there's the pre-clinical stuff -- basic research that you're alluding to. There's the stuff that is R&D but fails in the pipeline at some point. There's animal studies.

But in most cases, most of the cost comes from human trials (Phase I-III, and mostly Phase III), which can sometimes span dozens of countries and tens of thousands of participants. For some drugs, the Phase III trial(s) account for over 80% of the total R&D cost. Even when they are relatively small trials (as in this CAR-T therapy trial), the administrative and logistical effort to implement something like this is immense. The whole point of these trials is to collect data, and so that data is subject to the highest amount of scrutiny of any data in any medical enterprise. If you come into my clinic for a pre-op before surgery and I measure your blood pressure to be 140/85, give or take a few (oh wait maybe I used the wrong cuff size, lemme try again, oh it's pretty close), that's fine. But if you come into my clinic because you're a participant on a trial for drug X, and I measure it to be 140/85... I better be damn sure that's right (and the study coordinator at my clinic, and the pharma company, and the FDA), even if drug X isn't a blood pressure drug. We've seen cases where certain innocuous data discrepancies trigger central study monitors (study employees) being flown out to remote clinics to manually verify paper records or equipment logs to confirm/reconcile errors. Inaccuracies can cause you to miss things, or patients to be harmed, or can cause a study or a clinic or a hospital to be shut out from performing research again. And of course it can make the difference between a drug approval and failure. It's a lot of resources around the idea of data integrity.

So even small versions of these types of trials can be very expensive.

Now, is it TOO expensive? We often hear about the cost of bringing a drug to market to be around $1B. (Some studies have put it over $2B when you account for failed drugs, etc.) This may or may not be right, but let's not forget, the pharma industry makes more profit than just about any other industry. And the way CMS and insurers agree to pay for medicines... well, pharma has a lot of freedom in pricing (upwards).

You've hit on a lot of things that make the release of some new drugs/devices/therapies less expensive than we are commonly led to believe: reusing previous data, getting new approval for a specific enantiomer of a previously approved racemic drug, making cosmetic updates to existing devices, the tricks go on and on and on. And pharma keeps saying "R&D is so expensive! This drug should definitely be 2X more expensive than the one we're replacing." Planned obsolescence is a pain with your smartphone; it's a lot worse with your insulin pump or even the insulin itself! And one outcome of the high cost of large trials is that pharma does more of these types of un-innovation in some cases, which (due to the patent and regulatory and reimbursement systems) simply give pharma a free pass at making easy money off of our backs.

The high cost of trials also leads to really high prices of "true" innovation, such as in CAR-T trials. This type of therapy (like a lot of new oncology therapy) is highly customized, and no longer simply some chemical compound you make in a factory and then ship all over the world. You have to take a patient's own cells/fluid/materials/etc, process them, and then make modifications (in some cases unique to the patient), and then return the processed product (to the same patient). This is indeed highly costly on top of the cost of performing a lengthy trial, where for certain rare and/or terminal diseases your study endpoints are pretty tough to capture (um, did the patient die? uh, how long do we wait?).

This may or may not be right, but let's not forget, the pharma industry makes more profit than just about any other industry.

This is based on the survivor bias. The successful pharma companies make a lot of money, but the ones that aren't successful lose a ton of money.

And the successful pharma companies don't always stay successful. There are plenty of examples of big pharma companies that are much smaller than they used to be because R&D didn't pan out.

Pharma profits are high because the risk is high.

> most of the cost comes from human trials (Phase I-III, and mostly Phase III), which can sometimes span dozens of countries and tens of thousands of participants

Uhm. Yes. Again, I'd rather have an actual break down of the costs involved here. At least in Germany insanely overpaying doctors for conducting phase III trials replaced the [at least] $3-4k/day budgets pharma companies had to accommodate relevant practitioners at congresses when it was outlawed.

When the head of a public hospital in a western country can legally triple his/her income by conducting a study for 4h every Saturday morning, I file the costs under marketing, not R&D.

It's not that hard to estimate clinical trials costs. Last I checked you were looking at $10-20K per patient per year. It all depends on how much monitoring you are doing.

Now run the math. For a broadly used drug, you might need a trial of 20,000 people over 3 years (if you're looking for a long-term benefit). That's what a lot of the statin drugs did. That alone would be $1.2B at the high end, $600M at the low end.

And you have to run a minimum of two phase 3 trials. Follow on trials are even more.

If it works then it's certainly worth the cost. Also it is likely that you can do donor car-T which will also reduce costs in the future.

Apparently, it's not uncommon for pharma to coach the families that comment to the committee either... And in another orphan drug case with small sample size, that of Sarepta/eteplirsen, in which several evaluating scientists seem confused and the majority were overridden (for approval) by the FDA director in charge. Comments included, “serious irresponsibility” (by selectively publishing data during the whole process) and saying that the drug was a “scientifically elegant placebo”.


This new leukemia treatment is a really interesting, and perhaps the effect size is large enough for the small trial size, but we seem to have some problems in the review process (ignoring data) and incentive structure (coaching parents) for these "orphan drugs".


It's an orphan indication which has special rules to promote development. Otherwise, the thinkingngoes, nobody would develop treatments for uncommon diseases.

The downside of this is a treatment that you can get approved for something rare, but now it is on the market and they can tell doctors it has promise for some common X. Getting approval for treating X would be more expensive because you have to do a larger trial. (you should do a large trial for the rare thing, but since there are not enough people with the rare thing you cannot)

That's almost true, and is a well known part of pharma strategy. For example Botox was initially approved for hyperhydrosis (excessively sweaty palms and armpits) and was used off label for its most well-known application, wrinkles. It was finally approved for wrinkles only recently having been approved for other things along the way.

Any MD can prescribe any (non-opiate) drug for any indication. But the drug companies are prohibited from doing what you say: telling docs about usages they haven't proven in trials. Sometimes the sales rep will drop off a scientific paper (often funded by the drug company) about some other use of the drug -- enforcement in this area is uneven.

If they do want to advertise new uses they need to do new trials. And though existing trial data can also be used, if the dosage regime is different or, say the patient pool tox risks are different (perhaps the new indication is less serious than the orphan one) whole new tox studies may be needed. Which seems fair.

Since the US v. Caronia decision, off-label promotion is more complicated. The judge said restricting off-label promotion violates the 1st amendment, as long as the speech is "truthful"

I'm pretty sure Botox was originally used for Strabismus.

> they can tell doctors it has promise for some common X

That's not authorized, that's called "off label promotion" and it is severly punished when discovered. Also, doctors who use a non-approved drug for another indication can be tracked and investigated as well if needed (conflict of interests). Reimbursment also works only in certain conditions and may not apply if you use the drug outside of its allowed/approved indications.

Slight correction: as noted by others above, off label use while extremely illegal enforcement has been eroded and is patchy. And US law allows any MD (but not other prescribers such as podiatrists or NPs) to prescribe any approved drug for any medical purpose (opioids have recently been somewhat of an exception here).

However, as you point out, insurance companies in practice have a say. Also if the usage is implausible and there's an adverse event I think there could be a malpractice issues so those insurance companies also have an impact.

This is US law I'm talking about BTw.

Isn't 25 supposedly the magic age when more cells die than are created, i.e. when you start 'aging'.

Then again every seven years all the cells in your body are new so is the new you now aging faster as a "cellularly younger" person?

From the article:

   "Scientists use a virus to make the genetic changes in the T cells, raising fears about possible long-term side effects"
Is this a real risk? Is 'using a virus' in this way, still risky at all? or is it just the word 'virus' that makes writers put this line in every article about gene therapy?

{edit: real risk}

It's a standard tool. In fact a virus is not necessary a living device: there's a DNA delivery wrapper (looks like. Little lunar lander with a big spike coming out) and a payload (DNA).

A virus can't reproduce but it can inject its payload into a cell and have the cell's DNA copy/execution capabilities copy the DNA for the virus, making new delivery machines until the cell explodes and emits a cloud of new dlivery agents into the organism.

But if you remove the payload and replace it with your own, that payload presumably won't make more viruses

> It's a standard tool. In fact a virus is not necessary a living device: there's a DNA delivery wrapper (looks like. Little lunar lander with a big spike coming out) and a payload (DNA).

Just a slight correction. Not all viruses look like what you describe. [0] Others, like the HIV, are circular. [1]

[0] https://en.wikipedia.org/wiki/Virus#Structure

[1] https://en.wikipedia.org/wiki/Structure_and_genome_of_HIV

A real risk compared to, say, dying from cancer that you already have? What, are they worried it might give you cancer in 10 years? If the cancer you have now is going to kill you, that still sounds like a good deal.

(This is not hypothetical. My wife's grandfather had radiation treatment for cancer. 20 years later, he got cancer again, plausibly from the radiation. That's still a net win of 20 years.)

The virus is inactivated HIV. The risks about long-term side effects pertain more to gene editing than the use of a virus as a delivery mechanism.

Which are risks similar in nature to the risk of getting cancer after one has been infected with any other virus that integrates into one's genome.

There is am ecosystem of many types viruses in all of us, healthy or not. Part of that ecosystem is HERVs or Human Endogenous Retroviruses (PERVs are the same in pigs--im not kidding). Understanding how they do or don't cause disease is an active area or research.

Not only are viruses a cool (if somewhat nefarious) DNA-delivery system (though, HIV being a retrovirus probably means that it delivered RNA instead of DNA), but it can be targeted. HIV viral particles attack human immune cells, which just happen to be the general cell type that needs to be treated in this case! Using something like hepatitis or HPV probably would do the trick here since those viruses wouldn't have the keys to unlock the doors to get inside of the T cells.

A virus is a machine for replacing DNA in a cell with other DNA. Seems like the right tool for the job.

Not to be flippant, but isn't a "virus that cures cancer" exactly how the Will Smith movie I Am Legend starts?

ok, sure, the pessimistic among us can accentuate the negative aspects of this medical therapy, but it did solve the NYC subway crisis. and don't forget the wildlife habitat expansion.

these things aren't all bad.


Virus retroviral internal machinery has been used for ages.

> "Scientists use a virus to make the genetic changes in the T cells, raising fears about possible long-term side effects"

Isn't that the plot of Resident Evil/Biohazard? ;)

"While Novartis will not estimate the price it will ultimately put on the treatment, some industry analysts project it will cost $500,000 per infusion."

Meanwhile, the latest version of the US Senate's healthcare bill includes the so-called Cruz Amendment[1], which would allow insurance companies to offer health insurance plans without essential health benefits, which would allow lifetime caps on insurance[2], which could mean that your six year old with recurring leukemia gets pulled off their treatment when they're halfway through. Not because you did anything wrong, per se, but because maybe your employer refuses to spring for health care plans with more than an $x dollar cap. Or you never anticipated something so horrific and catastrophic happening to your family.

[1] https://www.nytimes.com/2017/07/13/us/politics/senate-republ...

[2] https://www.brookings.edu/2017/05/02/allowing-states-to-defi...

You could always pay the alternate price of kidnapping a Novartis executive's child until the treatment is completed, and then surrendering yourself to a lengthy prison term afterward. I can't pay $500k out of my own pocket, but I can buy a roll of duct tape and rationalize the hell out of an ethically tricky situation.

That's one of the classic posers to gauge someone's level of ethical maturity, isn't it? Your child is dying, and you cannot afford the treatment. Would you steal it instead of buying it? Why?

That sounds horrible.

Although, I do wonder what effects there would be if a cancer cure-all were discovered.

Since nearly everyone would need the treatment at some point, it wouldn't really be insurance anymore; more like a mortgage.

We're already at this stage - we have many life-extending (there's no life saving, only extending it for smaller or larger amounts) treatments and procedures, and the amount is growing. For pretty much every patient who dies currently we could extend their life a bit more (not they'd always want that, mind you) if we put in more resources in that patient.

It's tautologically clear that it's not possible to do everything for everyone, i.e. a community 100% composed of doctors and nurses wouldn't be able to provide all the possible life-extending things (especially late in life/close to death) to everyone of themselves. So one way or another we need a process to decide where we stop, i.e. what life-extending things will not be provided to which people.

Of course, there's a major practical difference between in a process that takes/costs one day of labor and extends life expectancy by a year, and a process that takes/costs a year of labor and extends life expectancy by a day - but there's no conceptual difference, and we have options all along that scale to find where the tradeoff starts/stops making sense.

> Another big concern is the cost. While Novartis will not estimate the price it will ultimately put on the treatment, some industry analysts project it will cost $500,000 per infusion.

Welp, guess my insurance premiums aren't stabilizing anytime soon.

There won't be that many people who need this drug, and supposedly it's a one-time cure that isn't a life-long drug treatment. I'm a layman when it comes to biotech, but my hope is that longterm the prices come down for this and other drugs that use a similar technique to treat diseases. I'm happy researchers are focusing on cures rather than "living-with" a disease.

Also, shoutout to Boston (and Cambridge) which I don't think gets enough love on HN. #1 in the entire world for biotech, #2 for software, #1 for higher ed, #1 for healthcare, you can actually afford to live here, solid public transportation, amazing skiing / hiking / beaches, NYC in 4 hours. If you are tired of SF think about Boston

You can afford to live there? I was looking at 1 br apartments in Cambridge and it was pretty darn close to SF Bay. Interviewers I was talking to said they commuted in 45 minutes just so they could own a house.

The unspoken assumption is that it's affordable if you have roommates (or a solid dual-income household). I have a place in Providence, RI and commute to Cambridge, MA by train (38 minutes to South Station via Amtrak, double that via MBTA). It's a shorter commute than when I rented in Brookline. Because of all of the Boston-area colleges, students set the standard. The housing stock is higher quality (and lower price) once you escape Boston/Cambridge, where students will live in any sort of derelict apartment, and use loans to pay inflated prices for the privilege. Boston has many things going for it, but the housing market is a real problem. Don't get me started on "broker fees"...

it's affordable if you have roommates

This is the story in NY too. It definitely does not qualify as "affordable" for anyone with a family.

>commuted in 45 minutes

I live in and work in Boston, and my commute is 45 minutes. I'm planning to move further out to buy a house, and my commute will probably still be roughly 45 minutes.

Unless you live within walking distance of your office or feel like you have a safe enough bike route to the office, 45 minutes is a fairly standard commute here.

I live in Waltham and while I don't work in Boston (work from home) it only takes me 15-25 minutes (depending on where in the city) to get into Boston.

The Boston area has awesome suburbs that aren't really suburbs like Newton, Arlington, Waltham, Brighton, Melrose (where incidentally Martin Fowler lives) etc. I know everyone wants to live in the city cause its cool but these mini cities are pretty darn cool as well.

From most of these areas you can get into the city in 20 minutes by car. 40 minutes if there is bad traffic.

And of course there is public transportation.

Compare this to most of the metropolises in America: Atlanta, Dallas, Houston, etc that have massive traffic and suburbs that are truly suburbs (boring as can be).

Compare this to most of the metropolises in America: Atlanta, Dallas, Houston, etc that have massive traffic and suburbs that are truly suburbs (boring as can be).

Yet those boring metropolises have something key: affordable housing. Which is why the world is looking a lot more like them: http://www.newyorker.com/magazine/2017/07/10/americas-future.... While most MA suburbs prohibit most development, thus causing prices to skyrocket: https://www.bostonglobe.com/business/2017/05/16/somerville-h...

I generally agree but often those affordable houses are massive cheaply made McMansions (particularly in Texas).

Living in New England I'm often envious and this is probably coming from jealously but really family's of 5 don't need 6000 square feet, a pool, a 3 car garage, and 2 acres of land. You can always live like the Europeans and most of the rest of the world does and there are condos sub 300,000 in the MA suburbs.

Sure you can, but would you want to? 6000 square feet means I have a place for my hobbies, with the tools all setup so I can do something for 10 minutes if that is all the time I have for them. I like my pool, it is a lot easier to go swimming when the pool is just out your door as opposed to having to go someplace else. I like having room in my garage for my collectable car. I like having my garden close enough that I can get fresh greens on a whim.

The point is different life choices enable different lifestyles. I've seen wood workers in Europe produce beautiful things in a tiny apartment. They prove it can be done, but they don't have the space for a table saw and as a result everything they do takes longer. Even if hand tools are the goal, in the large shop you turn around and your saw bench is there, while the European has to get it out of the closet. When done the American sweeps and is done while the European counterpart has to put the bench away as well. When I'm working I want the American style shop.

Your choices may be different, that is your choice. Do not try to force other people to like your choice.

Funny... Weren't these studies done in Philly, not Boston? And isn't the Novartis lab based out of Switzerland? I missed the part about Boston in this article...

You lost me at "amazing skiing" ;-)

This won't affect your insurance - the patient population is too small.

Now, if a similar treatment gets approved for, say, prostate cancer... that's going to affect your insurance.

Eh, probably not for prostate cancer, very treatable if caught early, and a lot of people have "watchful waiting" as it does less harm to leave it in for a while than take it out.

Now, if there was a gene therapy for breast, colon or liver cancer, then it would be a problem.

Yes, well, that's all true for prostate cancer, but if we had a drug that could cure it people would want that over surgery or "watchful waiting".

It depends ... the treatment may be wanted, but unless it has a much higher cure rate than the current regimen, no insurer will offer it, and no single payer system will either.

In the UK we have NICE (National Institute for Clinical Excellence) who make the decision over what drugs can be used on the NHS (and most private insurers use the same gradings), if it doesn't offer a clear benefit over the existing treatment in both effectiveness and then cost it won't be approved.

> the patient population is too small.

I think that cancer is one of the great filters. At some point a species will develop an unreliable way to treat cancer (chemotherapy); at which point cancer becomes a genetic problem as natural selection is displaced by medicine. Over many generations, this will kill a species unless a reliable cure is found - we seem to have that cure. Our cure is symptomatic (cancer being the symptom) but the same technology could eventually be used for designer genetics.

Essentially, more and more people are going to rely on this treatment until we are able to completely eradicate hereditary cancer. Probably not in our lifetimes.

What displacement of natural selection? Hereditary cancers are the exact type of cancers that haven't been affected by natural selection since they generally are post-reproduction age.

Furthermore, eradicating hereditary cancer (and many other diseases) is technologically possible already (screening of embryos, genetic testing of parents) but it's socially and politically complex; we as a society currently don't want to e.g. enforce restrictions on procreation just to prevent these diseases.

This discussion already happened below.

What is the selective pressure /for/ cancer-predisposing genes? And it's not as if most cancers act as a reproductive filter -- the majority of them emerge well after the onset of reproductive ability, drastically reducing, if not eliminating, their selective effect.

> What is the selective pressure /for/ cancer-predisposing genes?

Evolution does not and never has selected /for/ things. It is a process of elimination. What this means is that, since the advent of chemotherapy, we have eliminated selection /against/ certain types of cancer.

> the majority of them emerge well after the onset of reproductive ability

Correct, we have never had an impact on that and likely won't for a very long time (we don't know enough about our genome to do so). As-per this study, we have been curing cancer in children (as we absolutely should) with very little knowledge as to how/why they developed cancer so young. Will their children also be at risk for leukemia? What about their millions of descendants? We don't know. If treatments like this work out then we might not have to care.

Actually evolution may well be selection against cancer in older people. (may is key - we don't know for sure) Enough Young girls of breeding age have a small tendency to breed with old men. This increases our lifespan by ensuring those who have whatever genetics it takes to live older have more children. It is a small effect, but it is something.

But cancer isn't primarily a genetic problem; it's a consequence of the way our bodies work at the cellular level. We all accumulate genetic damage over time from a variety of sources, and we'll all get cancer eventually if something else doesn't kill us first.

That is the story for most people today but is not the entire story. If your mother had breast cancer then you are at a higher risk - if not for genetics that would be magic. That is why I specifically said "hereditary cancer."

Yep, other hereditary cancer syndromes too, such as Von-Hippel Lindau, which is a mutation in a tumour suppressor gene.

how much is getting cured of cancer at a very young age worth? dying vs. living another 80-90 years?

less than a Tesla? more than a house?

interesting questions arise around immuno-oncology.

should Apple be the most profitable company - or someone that literally cures cancer?

> how much is getting cured of cancer at a very young age worth? dying vs. living another 80-90 years? > less than a Tesla? more than a house?

It's worth so much that a person shouldn't be required to pay for it. Like a right.

>> It's worth so much that a person shouldn't be required to pay for it. Like a right.

It can't be a right. That would mean someone has an obligation to provide it. I would say it's worth so much that a monopoly on it should not be allowed.

> It can't be a right. That would mean someone has an obligation to provide it.

By that logic, voting can't be a right, because someone has to register voters, run the polling booths, count the votes, etc.

If the health care system was mostly staffed by unpaid volunteers, I might agree. As it stands health care is 17% of US GDP (10% global) and growing.

Society has an obligation to provide it, just as it does drinking water.

I get a water bill every month...

Also, I know folks on well water that live outside of the city, society does not provide their drinking water.

Also, There is Flint MI.

YOU get a bill. Society pays for public drinking fountains, homeless shelters, section 8 housing, etc... where water is provided for the less fortunate at a scaled societal contribution (or free).

Flint MI is an absolute crisis and the government has stepped in, the state of Michigan sued the city, and there have been over a dozen criminal indictments. Multiple governments have stepped in to repair the obligation of clean drinking water, and to punish those who failed the citizens.

You got it backwards. There is no obligation to provide drinking water. In fact most places where the city puts it in they force people to use the provided water and NOT put in wells. Then they send a bill.

This is bonkers logic. It's wrong to coerce someone into providing the cure, but it's right to coerce someone out of forming a monopoly?

An IP-based monopoly can't exist unless the society enforces the concept of that IP, e.g. coerces anyone to not make the treatment without a valid licence.

The monopoly isn't made by the company, it's a conditional and time-limited offer given by the society to promote the creation of such things.

If you show up in a random hospital with an injury or disease, they are already obligated to provide care, it's just you're directly responsible for the bill.

The individual shouldn't be paying.

In the US it actually doesn't work this way. Hospitals are only obligated to provide care for people whose lives are in immediate danger. Once the patient is stabilized, the hospital is free to kick them out. There are a handful of exceptions and many hospitals will provide care out of charity, but don't expect to get a half-million dollar cancer treatment unless you have health insurance or really good credit.

Whether or not an individual pays for it has nothing to do with the pricing the manufacturer sets. In fact, setting prices becomes even more of a concern when these things are being funded by tax dollars as the government is going to need a way of determining what's reasonable.

Anglo-Saxons (in general, not all, but in particular Americans) are so very weird when it comes to the "free market".

IMO (dirty socialist/communist that I am), healthcare is not part of the free market. The industry providing healthcare should be strictly regulated and no benefits allowed.

Also, what's worth healthcare compared to the US DOD spending?

Quite a lot actually. US defense spending in 2014 was $614 billion, 3.5% of GDP. US healthcare spending in 2014 was estimated by the world bank as 17.1% of GDP.

That's total spending, but even in terms of public money, the US is at 8.3% of GDP: larger than Belgium, the UK, Switzerland, Finland, Canada and many others.

This would certainly be possible but would require a major change in gov't research funding, including many extra billions for drug R&D, "productionizing" research, and large scale clinical trials; since a "no benefits allowed" industry wouldn't be investing to create drugs like this one.

>The industry providing healthcare should be strictly regulated and no benefits allowed.

Then how would anyone be interested in launching such a company or working in that industry ?


Some people believe other people to be fundamentally good and not interested only in personal enrichment (financial enrichment I mean here).

But some of you make me doubt that.

But why are some people quantum physicists researchers ? Why are there mathematicians and biologists ? They don't earn much.

Your line of thinking is so stupid and self-centered, it really makes me want to use swear words.

Public research is rather different from launching a company. It's not exactly the same hassle nor honors. There's some difference between understanding the world and selling stuff. Research rarely has products in mind, at least it's not its primary goal, unlike companies. They also don't have clients, which is a central thing for companies (and not the most enjoyable one).

And guess the reason why many researchers end up in private companies ?

Thanks for the insult, I thought this community was supposed to be mature.

When a manufacturer sets the price, they want to recoup the costs and then make a profit on each unit sold.

When a government funds the research, the marginal cost for each unit provided is time and materials, with no profit.

Government funds projects every day that are waste, spending tens, even hundreds of millions of dollars. We can continue to pour money down the F-35 drain but not provide healthcare? Make America Intellectually Honest Again

I still think it might be best if government funds the research and then allows anyone with the capability to use it. In other words, no patents. There would be competition and the price would come down. The big problem I have with this suggestion is that I don't believe in governments ability to do anything cost effectively - even funding research - but I might be wrong.

Dean Baker, an economist, has written here and there about alternative ways to pay for drug research:

"There are clearly better and worse ways to structure a system of government financed research. For example, the Free Market Drug Act, a bill recently introduced in the U.S. Congress, called for establishing a set of competing government corporations that would be evaluated at periodic intervals (e.g. 10 years) for the quality of their work. The worst performers would be put out of business with new ones created to take their place."

More here:



Considering the number of poorly performing gov't programs, I'm sure the "quality of their work" would be based on things other that actual output. Nepotism and buying votes are probably more likely.

> how much is getting cured of cancer at a very young age worth? dying vs. living another 80-90 years?

A tough question, made tougher by the side effects of doing so. $500k per infusion pays for quite a few mostly-healthy low-income uninsured kids to get basic pediatric care. The US healthcare system is already priced too high for a lot of people - I pay $2,002.25/month for my family.

> should Apple be the most profitable company - or someone that literally cures cancer?

Should healthcare really be a for-profit endeavour?

on the latter:

is it more ethical to profit from curing a disease or from causing it?

Pfizer vs. Red Bull.

and if you can't become a billionaire from curing things, how will this sector attract the same kind of talent/genius like other industries?

Biotech vs. Quants vs. SnapChat

The US federal government has basically standardized, in terms of public health and safety measures, on treating doing those interventions that will save lives if they will save lives for less than $7.5 million on average (it's been going up over time). This is for stuff like lead abatement, public safety commercials, pollution controls, etc.

That doesn't apply to medicine but it looks like this treatment is well worth it compared to the money the government spends, e.g., cleaning up lead paint. And since the number I gave is for an average life when saving a child we should be willing to spend more.

"should Apple be the most profitable company - or someone that literally cures cancer?"

Nobody needs an Apple product. Life is a basic need.

They deserve to be profitable but in a way that doesn't require bleeding dry the cancer patients themselves.

getting downvoted for asking questions.

HN sometimes really can fuck right off.

Does anyone know: what are the failure rates like for the gene editing technology being used for this? Thinking like a software engineer, are there transposition errors (GATC --> GTAC) , atomicity issues (GATC --> GA)? Mutations afterward?

In this particular case the therapy is being applied to cells that have been extracted from a patient. That allows a reasonable error rate where errors can be filtered out, and success verified before the cells are reimplanted. This is a strategically nice intermediate before having to run a therapy on a living organism. The technology in use here is not quite the same kind of 'DNA editing' as is found with tools like CRISPR, but rather a much more therapeutically mature (if technologically blunt) form of viral 'insertion of a block of code'.

With respect to actual code fidelity, errors in DNA come from a number of different sources. Every time DNA is copied (a cell divides) there is an inherent fidelity rate of the copy (on the order of a single mistake per billion writes). The payload here is on the order of a few thousand base pairs so copies should have a very high fidelity.

In this case a viral protein is 'inserting' its DNA randomly into the genome of the target cells. Imagine inserting a library of code randomly into a codebase. Certainly not ideal, and an issue that CRISPR technologies promise to help improve. However, given that the therapy is only being applied to immune cells that are only running the 'immune' section of the human codebase, and no progeny of those cells will ever have to become a brain or skin or run any of the other programs, the chance that the inserted DNA disrupts the 'immunological' code in the codebase is relatively small. And if there is disruption to some cells' genomes those cells could be screened out if they really distort something they should not.

With respect to DNA generally, common errors arise from undesirable but common chemical modifications to the code itself. The DNA can become damaged (by reactive oxygen, UV light, and through other chemical reactions), and while there are significant systems to repair that damage, oftentimes since there is only a single backup (DNA is 'double-stranded'), it's often impossible for that machinery to determine whether the error is on strand1 or strand2, so 50/50 chance of 'repairing' into the error.

Is this the same CAR-T treatment that Juno Therapeutics tried and scrapped[0] after 5 trial patients died after receiving the treatment?

[0]: http://www.xconomy.com/seattle/2017/03/01/after-trial-deaths...

No. From the article:

> In the past, a handful of patients who were getting similar treatments developed by other companies died from serious brain swelling. Although those sorts of complications did occur in some patients receiving CTL019, the patients recovered and there were no fatalities, the company says.

  it will cost $500,000 per infusion

Making the choice between being indebted for your entire life and just dying is surprisingly hard for me.

Isn't this how vaccines work?

Vaccines work by previewing a foreign object, that if seen again, the immune system will attack. The big difference in treating cancer is that there is nothing 'foreign' about cancer - it arises from one's own cells. Cancer is an unproductive or even malicious mashup of material already found in the human body, very much unlike a viral or bacterial infection. There are no natural differences in kind that can be detected by an immune system that would not otherwise attack healthy tissue as well.

In this case there is essentially a synthetic sensor designed and provided to the immune system that is precisely tuned in the lab to detect the (very subtle) differences between a cancer cell and a healthy cell.

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