Harnessing the Immune System to Fight Cancer | Hacker News Reader

Harnessing the Immune System to Fight Cancer

nytimes.com

171 pointse15ctr0n30 comments

My dad was an immunologist.

By all accounts he was absolutely brilliant. I lost him the day after my 9th birthday in 1982. I grew up with a series of George Washington Cherry Tree kind of stories about him permeating everything.

There's a lot of pressure growing up being the only son of a very accomplished person - having the responsibility of filling shoes that just can't be filled. It got to the point where I started being skeptical. Nature of being a teen/young adult I suppose. In the mid-90's I was having a conversation with my Uncle and he tells me how before he died, my Dad was hell-bent on the Nobel prize. He was going to win it by curing cancer. He told me "the key is in the T-cells. He was always 20 years ahead of his time." I took it with a grain of salt at the time. Another tall tale.

Some years later in 2002 I'm reading an article and lo and behold there's promising research coming around that T-Cell treatment is yielding positive research in cancer. And now I read this article, scroll down a little, and there's a diagram highlighting the T-Cell.

A few years after that I end up working at Stanford and that gave me access to old news articles that aren't up on the web, and a lot of scholarly things that aren't typically in search engine results. I was absolutely floored to discover that the tall tales weren't tall tales at all. They were just stories from people who were honestly proud of him and what he did (even if they didn't understand most of it). I tracked down a few of his colleagues and I'll tell you - it was pretty powerful hearing from some very accomplished people about the work that he did and the kind of man he was.

When you think about it, there has been more progress in medicine in the last 80 years than in the previous 800. It's pretty stinking amazing if you ask me. And most of it has been made by people who's names you'll never hear or read about.

This article brought a lot of memories back to me so I thought I'd share. Miss you dad.

What's your dad's name? Do you have any material we can read?

Steven Bix Kallestad - here's a blurb from early on when he developed a Hepatitis test: https://news.google.com/newspapers?nid=2245&dat=19701221&id=...

Here's an article about his baboons escaping: https://news.google.com/newspapers?nid=1798&dat=19660131&id=... (he rescued them from a research facility, they lived in his bathroom while he was at work, they escaped and lived in the trees outside for a few days)

His public (amex) company getting sold to a private entity shortly after he died: http://www.nytimes.com/1982/06/03/business/company-news-kall... (edited because I originally linked to a consecutive sale)

His products still on the market: http://www.bio-rad.com/evportal/evolutionPortal.portal?_nfpb...

And by the way... his patent on the mouth guard (he was a hockey player and one day decided people shouldn't lose so many teeth so he put this together with a friend) http://patents.justia.com/patent/3943924

There's a whole world of things that he did and obviously there were a lot of other scientists involved, but suffice it to say that he had a hand in many advancements that are taken for granted today like STD tests and treatments, pregnancy tests, etc. He helped develop and improved a lot of equipment as well. A lot of the links I can find right now mention Kallestad products used in research like immunoassays, HGH, nephelometers, etc, but honestly it gets above my comprehension level pretty quickly. I can point to "page 273 of x research article where it mentions a specific product/patent/procedure that the research relied on, but it's not all that interesting if you aren't in the field (I don't think).

My favorite story about him is about an aligator named Zago. My Uncle lived in Florida and decided on a whim to send my dad a live alligator. My dad, being a scientist and used to receiving shipments of animal based serums saw the label on the box and put it in the refrigerator. Days later, my Uncle called and asked him how he liked the new pet. He ran to the refrigerator and it was still alive and kicking. He kept it in the bathtub for a few years until it got too big and donated it to a zoo.

That's a really cool story, thanks for sharing! Your dad has such a large and wide body of work that lives on even today. That's really amazing.

Our lab (http://www.hammerlab.org) solves computational problems in the domain of cancer immunotherapy. Our open source software (https://github.com/hammerlab) is used to create personalized therapeutic vaccines that we are testing in a phase i clinical trial right now.

If you are an experienced machine learning engineer and would like to help improve the quality of our vaccines, please send me an email at hammer at hammerlab dot org. We are located in NYC.

I can vouch that the hammerlab is legit and doing solid science! I've had great interactions with members of their lab and have been impressed by their work, as well as their committment to open source and open science (preprints, etc).

To piggyback, if you are interested in personalized cancer vaccines, we're also hiring at WashU for a senior scientist position that will be working at least part time on immunotherapy and personalized cancer vaccines. If you're an interdisciplinary person, who knows cancer and can code, drop me a line (c.a.miller at wustl.edu)

Nice work, but as you say, the following: It is common to see these gene sets being referred as pathways and it is for this reason that GSEA is also known as pathway analysis. Will really lead to communication problems with biologist, but you already admit that in the sentence following this one :)

For a very long long time, Pathways are the word used by biologists for defined signalling cascades inside cells, to used the same word for a clustered gene set is quite confusing.

Your approach is very data driven and as it is you are ignoring a vast sea of literature out there that investigates the relation between signalling pathways and target mRNAs by doing research into (among others) transcription complexes directly (i.e. using CHIP [0]) (like they do i.e. here: [1]). I get that for computer scientists this is perhaps the fastest way to results but I feel such research can be done in a smarter way, by teaming up with biologists. Don't you agree?

[0] https://en.wikipedia.org/wiki/Chromatin_immunoprecipitation

[1] http://www.ncbi.nlm.nih.gov/pubmed/24695361

I'm a recent Biology undergrad. Do you have any advice for learning Comp Bio/Bioinformatics? Can it be self-taught, or does one need a PhD?

The programming part can definitely be self-taught, that's basically how virtually everybody actually goes about learning programming successfully.

You already know some biology, so you can probably say more about that side of the equation.

It can definitely be self-taught!

A promising path would be to self-teach as much as possible and then apply for an MS program in bioinformatics or something similar.

Apart from the usual advice for teaching yourself programming, you can practice bioinformatic-specific skills at the wonderful Rosalind Project [0].

[0] http://rosalind.info/about/

I'm a multiple myeloma patient, and started daratumumab a few months ago, which is a monoclonal antibody. It's not a checkpoint inhibitor as described in OP, but rather binds to CD38 which is overexpressed by MM cells, and thereby activates the immune system against those cells.

Certainly activating the body's own defenses against cancer cells seems like an effective approach. Side effects are minimal as the therapy is more targeted, but as the article mentions, we just don't understand enough right now about when they will fail or why they fail.

My own results on dara have been underwhelming. The first month raised our hopes of a complete response, but it totally flat-lined by the second month. By month 3 levels were rising again, and now it seems I've bred a more resistant clone. My next stage of treatment may very well include a CAR-T based therapy, or a checkpoint inhibitor.

One thing the article gets wrong is the cost. Insurance pays up to $10k per infusion, but co-pays are capped at < $10k per year. There is a huge market for these novel agents, and insurance companies are on the hook to pay for FDA approved treatments. From what I can tell this is driving massive R&D investment and a large pipeline of new potential treatments.

Just curious, why the throw away account?

I guess that's one nice thing about chemo that doesn't make you lose your hair, cancer isn't really part of my public identity. Of course my close friends and family know, a few work colleagues, but for now at least that's the extent of it. I guess I have a keenly evolved defense mechanism of just not dwelling on shit I can't handle, and not sharing that particular fact means it doesn't become a part of every conversation. For now anyway, it's just not something I would want to show up on Google tied to my nick.

You shouldn't even have to respond to that question.

I guess I'm getting older, or from a different generation, but a lot of life is personal--right?

I'm shocked at how fast Google scoops up these posts for the world to see forever. There been times where I go back to edit/delete a comment, and figure what's the point; Google set it in stone.

Part of why I love HN: pseudonymous and throwaway accounts allowed + sane community moderation.

Discussing certain things under full name would almost be irresponsible for some people as long as Google, Facebook and various TLAs in both the East and the West has this landgrab going.

I assumed he didn't want people to know, but I was curious to know his response just the same.

Over the years, I have gotten to know a couple MM patients who have had great results from the double stem cell transplant approach. Happy to put you in touch with some of them if you want to explore that option.

We are getting better at it already! Perhaps it is more difficult with blood cancers such as yours but for solid tumors one can roughly say (in linked article the focus is on colon cancer): the more mutations, the more proteins that deviate from your normal proteins, the better the chance the immune system has at recognizing a tumor as foreign. Also clues (as to if someone will respond to immuno therapy) are in how many and what type of immune cells are inside the tumor (as opposed to only in the surrounding healthy tissue). This can be seen in the standard pathologist slides that are made from biopsies.

That said, there are still some responders among those who show unfavorable signatures for immune therapy, so there is work to be done.

[0] https://www.researchgate.net/publication/298427742_Integrati...

One of the heroes of the article explained how his team discovered the first drug to prolong survival in advanced melanoma patients:

The discoveries that led to the drug, Dr. Allison said, came entirely from years of basic research in immunology — experiments in test tubes and mice — and not from the clinical or “translational” science, aimed at moving rapidly into humans, that is so heavily favored now by institutions that pay for studies.

“None of this came from cancer research, none,” Dr. Allison said, adding that without support for basic research, “progress, if any, will be incremental, not a big leap.”

I keep getting hints that there is some fight behind the scenes in US biology over how much money should be devoted to exploratory vs translational research. Comments like these pretend as if this wasn't a difficult trade off, that we need a huge shift in one direction, and that this direction is so obviously better we don't even need to discuss the relative amounts involved.

Aren't we doing that already? Just the rhetoric has changed: we just dress up all our basic science as being about cancer.

The problem with current immunotherapies is that they can trigger a wide range of auto-immune side effects: skin, lungs, gastrointestinal, thyroid, pituitary, liver, and more. Each dose is like spinning a roulette wheel and hoping it lands on "cancer cell" instead of a useful organ.

These side effects can often be dampened with high dose steroids, but they can persist for a very long time. After all, that's why immunotherapy is so exciting, it creates a durable response via immune memory.

In my opinion the next big step will be injecting drugs like ipilimumab (Yervoy) plus adjuvants directly into solid tumours and malignant fluids. This would lower the total dose, reducing the chance of systemic side effects, while still delivering a high concentration where it is needed most: where dendritic cells carrying tumour antigens meet T cells in the nearest lymph node. The subsequent immune response will still provide a systemic benefit. For example, a recent melanoma study saw an abscopal (distant) response in 10 out of 12 patients with intra-tumoral treatment with ipilimumab + IL-2, with no serious distant side effects [1]. (Yes this is a very small study, but bear in mind response rates to systemic ipilimumab are normally more like 15%.)

Unfortunately I wouldn't expect a company like Bristol-Myers Squibb to fund these kind of studies. The current model for pricing is by the mg, and intra-tumoral treatments use about 100 times less product than systemic treatments.

Other drugs like nivolumab/pembrolizumab would still need to be delivered systemically because they operate on the front-line where T cells meet cancer cells.

The step after intra-tumoral treatment will be generating and amplifying the immune response in a test tube instead of in the body. This is also very promising. [2]

[1] http://www.ncbi.nlm.nih.gov/pubmed/27391442

[2] http://www.ncbi.nlm.nih.gov/pubmed/26894495

My dad was telling me how interventional radiologists have been doing chemo embolization, where they deliver the chemotherapy directly at the tumor site/organ. I don't see why they couldn't use ipilimumab and similar drugs to do the same with immunotherapy.

Wim Hof (aka "the Iceman") has developed a fascinating method which enables anybody to consciously manage one's immune system, no drugs involved (scientific studies are available).

This should interest everyone who's dealing with immune system overreactions, such as any auto-immune disease, allergies, etc. and persistent inflammation issues (also touching depression).

For an easy intro, check out the Vice documentary: http://www.vice.com/en_uk/video/iceman

Joe Rogan has a long interview with Wim Hof: https://www.youtube.com/watch?v=Np0jGp6442A

A more recent interview: http://www.richroll.com/podcast/the-iceman-wim-hof/

For the biological explanations: http://www.icemanwimhof.com/files/2016wimhofmethod-revealed....

One of the scientific studies: http://www.icemanwimhof.com/files/pnas.pdf

Wim Hof's 2 Websites: http://www.wimhofmethod.com and http://www.icemanwimhof.com

If you don't have the USD 200.-- for the 10 week program, you'll find it on your favorite torrent site.

We're a Start-up building software for this new emerging cancer treatment. Looking for strong engineering talent. At Vitruvian Networks (http://vineti.com)

The future of cancer is targeting, in one way or another. It has been interesting to watch immunotherapy win out as the dominant approach to targeting; there are other options, such as nanoparticle platforms in which targeting and kill mechanisms could be assembled as needed, or oncolytic viruses, and so forth. All of those are still under development, just with nowhere near as much backing as, collectively, goes to forms of immunotherapy.

Still, most immunotherapies packaged and deployed to trials are more of the same in terms of being very specific to specific types of cancer. There are hundreds of types of cancer, depending on how you want to count. It is much better to be able to build a therapy with few side-effects, given that chemotherapy has a very unpleasant effect on health and life span over the long term even when successful, but the extreme specificity is still problematic from the strategic point of view. The cancer community has shown that they can't produce new therapies for all cancers at a fast enough rate to completely control cancer any time soon. The strategy needs to shift to producing therapies that can target many types of cancer.

Once class of immunotherapy shows particular promise on this count, and that's the chimeric antigen receptor approaches that have produced such good results for leukemia. It should considerably reduce the cost of building new therapies for many types of cancer once a couple of types of cancer have been successfully treated that way and the basic platform is defined. We'll see how that goes over the next fifteen years or so. All in all, immunotherapies look to be the replacement for chemotherapy / radiotherapy as the coming generation of cancer therapies, and people will keep their hair, and will get treated as outpatients in clinics, and cancer survivorship will increase, and it is likely that they won't lose a decade of life expectancy as a result of treatment.

It is worth looking beyond that, however, as the technical basis for the generation of therapies after immunotherapy is actually being worked on today. For my money the thing that comes after immunotherapy will involve blocking telomere lengthening, as that is a - possibly the only - basis for a universal cancer therapy. All cancers need to lengthen telomeres, cannot evolve their way around that need, and building one therapy to control every type of cancer seems a much better path forward than the present strategy of seeking ever more specific biochemistry to interfere with. A fair number of people are working on the telomerase side of that equation, and not enough people are working on the alternative lengthening of telomeres (ALT) side of that equation, but this will all come to a useful conclusion somewhere in the late 2020s, I'd imagine.

I thought telomeres lengthening was for curing old age? Latest issue of Science News is on age extension; I'm a layman.

Cancer genesis relies on a sequence of genetic mutations occurring and has many defining characteristics. The telomere lengthening he's discussing is a mechanism by which cancer cells make themselves "immortal".

So the goal is to lengthen the telomeres if healthy cells and shorten them for cancerous cells. Will be interesting to see how it works out for Elizabeth Parrish:

http://arstechnica.com/science/2016/04/ceo-tests-crazy-genet...

These treatments are truly awesome.

The pharmaceutical companies now making them.... not so sure.

It makes me generally very angry that nearly all the research has been conducted using charitable donations and a large volume of open source software.

then a few pharmaceutical executives get to charge either the insurance company or in many cases in the US the patients family - hundreds of thousands of dollars a month to provide access to treatment.

Reply on news.ycombinator.com