Every Cancer Patient Is One in a Billion | Hacker News Reader

Every Cancer Patient Is One in a Billion

wsj.com

119 pointsmighty-fine64 comments

Yep, if we really want to cure cancer, we need to go back into the tissue archives and study every single cancer that's ever been diagnosed, at the molecular level, to start getting our heads around what really happened to all those people. 10 million people a year. Every year. For a the last 6-7 decades of chemotherapy.

Meanwhile, there's this perverse medicolegal incentive to retain the excised tissue samples for the minimum necessary period of time, which is 10 years for US laboratories accredited by the College of American Pathologists.

You heard it right, the lawyers won. We spend medical administration time filling out medical administrative paperwork to spend medical care dollars burning most of the cure to cancer. Every year.

hmm...does this happen in other countries?

Interestingly, the CAP has seen very strong uptake internationally. For people in charge of nation-state level assets like major medical research centers, the scientific rigor the CAP has learned to impose on laboratory medicine over the last 50 years sells itself.

I asked the director of a local tissue archive here in Australia. Their specimens go back decades. The regulatory minimum is 10 years for many tissues (more for paediatrics, and indefinitely for clinically significant rarities in molecular genetics) but in practice they hold longer than they have to.

Yes, there are some really valuable studies out of Australia showing that protein signatures can be retrieved by immunohistochemistry back 80+ years.

What would be the value in finding a cure for cancer for a patient who is already dead if that patients particular mutation is unlikely to be found in another person?

Patterns. There's very rarely one mutation. It's all about patterns. The problem, and what makes this a potentially tractable problem in the age of machine learning, is that we can now catalog an arbitrarily large number of patterns and associate any given new cancer with some number of known patterns.

As has already been stated you can use the prior tissue to better individualize current treatment. For example I'm a really rare patient, most people in my situation are already dead by this point in their treatment. Chemotherapy worked so well most of my primary tumor was gone within days and they haven't seen it since my 6th treatment. They still think it's there but it's generally too small to be seen. They didn't save the samples from my cancer to figure out why it worked so welln or why I've been able to handle aggressive treatment for far longer than average with minimal side effects. No hair loss, minimal nerve damage, etc. If that information could help even one other person, I would want it out there. I've contributed to every study I can be a part of and consented to every bit of sharing of my generic material I could, I only wish I could do more. Nearly everyone with my same diagnosis at the same time as me is dead. Speaking just for myself, being able to contribute to keeping other people from ever going through this means the world to me.

What kind of cancer did/do you have and at what stage did you begin treatment?

This is truly astonishing, by the way.

I'm stage 4 Stomach cancer, started there because it's super aggressive, the OS for 5 years at stage 4 is 5% at last reading, based on my current treatment outcomes I've got at least another year in me based on available treatments as I'm still first line and the cancer is still responding to chemo. I'm aiming for 10 years, because go big or go home right? It was all through my periodontal cavity, but apparently aside from the primary tumor my cancer cells are too small to be seen on imaging, PET or CT. I'm apparently supposed to be dead already, and yet here I am kicking around. I get "you don't look like you have cancer" a lot, because I'm on the bigger side. It was the best damned weight loss program I was ever on though, lost 60lbs before my first chemo. It took me a while to realize just how stupendously lucky I was. Every single day I wake up terrified that it could be my last and every single night I go to bed terrified I might die in my sleep. Every treatment has a risk of death, or get this, causing new cancers. I'm heading into my 16th treatment of 5FU and just recently dropped the Oxaliplatin, which causes permanent nerve damage in most people, but it's way better than death, which it can also cause if you have a bad reaction. I've been planning out a Youtube channel for younger people like me who get cancer, and I suppose older people would appreciate it too. I've missed very few days of work over this whole thing and it's been mostly an inconvenience. I've got mostly good days, been way more chipper since getting diagnosed oddly enough, I have some bad days, and nights I straight up cry in the shower so my wife doesn't see. I think the best thing about getting cancer is that I have an excuse to make people in a cancer ward laugh with my dark humor and no one will say anything because we all need to laugh, it's the only time that constant nagging in your head about how close you are to death seems to go away. I suspect that's kind of like soldiers in long-term dangerous deployments, there's a lot of waiting and everything is normal and you get into a routine, and then without warning you're in danger and everyone's moving and serious, and there's no time for jokes because everyone's just trying to focus on staying alive. But again, I'm one of the lucky ones. I'm still lucky enough to have the privilege to worry about this. I'm still lucky enough to wake up and see the amazing person my child is becoming. I've been lucky enough to be able to finally push myself into doing the charity work I've always wanted. I want everyone to have the opportunity to feel this lucky.

Wow. You are incredible.

Eh won't go that far myself. I'm here because I've gotten a lot of help along the way when I've really needed it. I just try to give back in whatever way I can with the time I have left. I have good days and bad days still. It's just about not letting the bad days be more than the good days.

These rare mutations are actually do come up in other patients and are worth studying to work out a molecular mechanism that explains the cancer. Mutations that drive cancer usually affect the same couple of pathways, and different mutations could bring about the same phenotype.

I was unaware of this 10 year rule. Here at UH Cancer Center I know there are teams with samples in the freezer from decades ago. Any links where I can learn more?

You can keep samples as long as you want. The CAP requires accredited labs to keep them 10 years. Attorneys simply want a record retention policy, which creates a natural tension many lab directors respond to by getting rid of the blocks after 10 years. See (1) for the checklist, the specific line item is ANP.12500. In theory things can get kicked to a biorepository, but the standards for a CAP-accredited biorepository are quite high. Running a biorepository that's not accredited raises the specter of questions like "What are you doing with this material, doctor?", so a lot of folks would rather burn our hope for the future than have any possible source of questions about what they're doing.

(1) http://elss.cap.org/elss/ShowProperty?nodePath=/UCMCON/Contr...

I see. I guess on the bright side collecting new data is becoming exponentially cheaper. For example, this year the world will collect more heart rate data than it did than all previous years in human history combined. So even if we lost all heart rate data from decades past, we generate that quite quickly now. You certainly lose a lot of longitudinal value, but perhaps that won’t turn out to be too critical? (Or perhaps it will)

I think there is a problem even on a higher level. I am not aware for example that the detailed history of cancer patients, is recorded centrally for later data analysis. Which treatment worked, which did not? Even airlines now share data about flights to help engineer around the world to find remedies for problems with small incidence levels. I wonder why that is so complicated in the case of cancer

Researchers can usually request that data from the hospital records (needs to be specifically requested due to HIPAA i believe) and some of it is already in public databases like tcga.

Ironically, living in Germany makes me scared that I won't be able to pay for a therapy by someone like Dr. Nagourney as our governmental health insurance providers are very eccentric when it comes to paying for treatments.

That combined with my high tax-rate and only slightly above average salary means I won't be able to save enough money to afford it to send a loved one to him when it becomes necessary. That is crazy scary to me.

A programmer from close to my area had to publicly raise funds (~350k iirc) to send his 2 year old son to the US for a surgery that saved his life, but which no hospital in Germany dared to perform.[1]

[1] https://www.facebook.com/hilf.david/

I'm a developer who has been living for 4+ years in Germany. Does that mean that it is a good idea to perhaps change to private insurance?

I thought that when I get old here, as I pay all those fucking taxes, I would get a good treatment for whatever diseases I may have until it's my last day.

That sucks. I feel ripped off.

I think the OP is confusing a few things here. Public health insurance will always pay for whatever treatment you need. Age is only a factor in very few cases. What matters is that the doctor or hospital is also in Germany. The system allows you a huge amount of freedom in choosing who to get care from. But the system doesn't cover getting care from foreign facilities. Private health insurance policies have the same restriction.

> What matters is that the doctor or hospital is also in Germany.

Given that ~90% of cancer research is funded in the United States [0], that might work for established therapies but anything on the cutting edge might be hard to get?

According to this article from MGH, cutting edge immunotherapies cost ~$1M per patient: http://protomag.com/articles/solid-tumor-barrier

And this one demonstrates the delays in their availability in Europe. Sucks big time for the people currently needing treatment: https://www.bloomberg.com/news/articles/2018-09-12/novartis-...

[0] #62 Peter Attia M.D. podcast with Keith Flaherty MGH

That's precisely my point.

I don't get much freedom when it comes to international healthcare and won't be able to save enough to pay out of pocket/take a loan for something as dramatic as this. The only insurance answer to this is a Dread Disease insurance which pays out once specific illnesses appear, but the amount and illnesses are completely dependent on your financials aswell, so you're still worse off in this case compared to simply earning more in the US.

It's great that I don't have to fear going to the hospital for normal stuff, but everything extraordinary is very scary as our public healthcare system is simply crumbling, getting appointments for specialists can take months, hospitals are overcrowded (my mother was taken by an ambulance to a hospital due to a heart condition, only to be ferried further as all beds were full that night) and doctors and nurses are working themselves to the bone or simply leave the country outright.

I don’t know about the German insurance, but for example the Swiss insurance makes a specific exception: in case there is no adequate treatment or expertise available within Swiss borders, the insurance must cover a foreign treatment.

Sadly, as shown in my linked case, that isn't the case here. The boy would most likely have died if his parents didn't publicly raise the funds themselves and pursued the therapy in the US.

Well the private insurance is indeed a problem if the insured cohort gets old, i.e., all the clients that signed up for a certain insurance product in around the same time frame, and when it is cheapest: mostly at age 25-40. After a while, the clients get older and the product becomes more expensive for the clients, and won’t attract new clients anymore. Then, all remaining clients become old „together“, hence the old aged themselves need to share ever raising costs.

Although there tends to be great molecular diversity in each person's cancer, there are still avenues of genetic commonality that can be pursued. For example, the tumor suppression gene TP53 is mutated in a high percentage of cancer cases. If this mutation can be reliably repaired, our ability to treat a number of cancers could be significantly improved. There are recent results suggesting that this may indeed soon be the case (for example, treatment with APR-246).

Many cancers are heterogeneous. Once a malignant community forms the established organized population deteriorates into chaotic competition.

Wait, are we back to talking about how to organize programming teams again?

Well, you do hear humanity likened to a cancer fairly often. Look at what we do to other species, and we're even trying to metastasize to other planets. Lucky for us that this universe doesn't go in for it's millennial checkups, huh?

If cosmic things had millennial checkups, they'd be cosmic-scale in their hypochondria. That would be like me getting a checkup every second.

I wanted to find out the unit of time of "every billion years" - first thought was "billennial".

Apparently, a millenium is otherwise known as a kiloannum; a megaannum is a million years, and a gigaannum is a billion.

So, a reasonable universe should be getting gigaannual checkups.

This is why Valve hires at its current rate?

The reason disease, viruses, technologies, and team organisation are likened to Darwinian information transfer systems is because that's what they are.

The book "Tripping over the truth" is a great history of the failure of the genetic mutation theory of cancer to cure cancer. It argues that it's a dead end and we should go back to exploring Otto Warburg's metabolic theories.

Mutations impact pathways. Studying genetics and epigentics is fine. Metabolism is another angle but the brain itself consumes a lot of energy... it’s not just dividing cells. If you target metabolism you run a huge risk of killing the patient before the cancer does.

Warburg's metabolic theory is an important part of current cancer care and many prospective efforts. However, tumor metabolism is an effect, not a cause. The simple-as-possible-but-no-simpler model of cancer is that mutations occur, breaking the blueprint code for cellular processes. In most cases this leads to cell death. In rare cases, the mutant survives and either fails to interpret a death signal or over-interprets a growth signal, or the growth switch is constitutionally activated. This leads to more mutations and some number of daughter lineages also survive. This all stems from very straight forward underlying rates of mutation predicted by the laws of thermodynamics, requiring no other input. There are plenty of example problems available to work through on Rosalind.info.

That is, cancer is a disease of information. And this is extremely compatible with the genomic evidence in every cancer we've encountered, many, many times over. Viruses you say? They insert into the genome and interfere with normal cell processes. UV radiation. Chemical exposures, oxidative stress, etc, etc. Those are intermediary chains of causation. Interesting for specific examples (wear gloves, wear sunscreen, be careful about sex, etc).

One of the blueprints that tends to fall apart is the complex aerobic respiratory pathway, including the Kreb's cycle, etc. But there are other, usually latent energy pathways, like glycolysis, that are available. It is the loss of aerobic respiration that unmasks the glycolysis pathway and all its associated inefficiencies and metabolic stress effects.

I have that book, but regret buying it. Frankly, it's lightweight for a HN audience.

The original technical book "Cancer as a Metabolic Disease" by Thomas N. Seyfried and published by Wiley seems pretty accessible to me.

I'm not advocating for or against the theory. It seems though a lot of time and money has been invested into the genetic mutation approach, without the expected returns in actual health outcomes, so that there is least a case to answer.

Would open sourcing all cancer research and allowing anyone to contribute and share in the open source community... would that assist in potentially finding a cure/treatment for various cancers or would the profit motive be required for these research to be successful?

There are a few case studies on that

Protein folding and grid computing networks are based on altruism and enthusiast novelty interest “wow I can use my spare computational power on this global network of other enthusiasts to cure cancer”, you know, novel in 2004.

As soon as some enthusiasts introduced cryptocurrency to these networks, see Curecoin and Gridcoin, users of those quickly became the most significant contributors to those networks (until those particular crypto currencies’ economic models fell apart). It wasn’t “spare computational resources” it was capital investment for a yield in those digital assets. I believe there are communities still experimenting with this regarding protein folding and looking for mutations iteratively.

So a profit motive will always beat out a non-market based approach, which isn’t really news

> So a profit motive will always beat out a non-market based approach, which isn’t really news

Call me a dreamer, but I wonder if that would remain true in a UBI-style environment?

if there is a scarce useful resource being used and acquired then yes?

I think people would be tempted to try more and crazier things to earn more money because the risk of failure is lessened in that scenario.

how would any random person test their cures? Joe Schmoe thinks fasting for a week and then consuming 100 ml/kg of deionized Jagermeister will stop leukemia dead in its tracks. So which dying cancer patient does he get to test on?

If its NIH funded its automatically open sourced 1 year after publication.

https://mobile.twitter.com/AmandaHaddock/status/922493664856...

http://archive.is/VESVr

Paywall :/

> Since there are some 1,000 cancer-related genes and each cancer requires up to three distinct gene alterations to succeed, every cancer patient is literally one in a billion.

Errors in his math:

- not every combination of any of the 3 genes (1000999998 = "1 billion" ordered combinations) causes cancer

- the order doesn't matter

- the occurrence rates of each mutation are not the same and don't appear in the same rate among various populations

Thus, the number of kinds is much smaller than 1 billion, and many people do have the same type genetically.

Holistic practitioners should all be studying machine learning.

They were right about the problem of not having individualized medicine, way off with the solution. Now they can win with data

If we really want to cure cancer we should stop polluting our air, water and food. And even if we can't stop we need to find ways to disincentivize making products that require our air and water to be polluted. For food:

https://www.basicknowledge101.com/subjects/verticalfarming.h...

I applaud your motivation, and I agree with your suggestion, but that's not where most cancer comes from. Most cancer arises out of a random failure of the mechanisms underlying apoptosis.

Random mutation rates are linked to rates of cell division. There is an error every x number of times a cell divides. The faster cells divide (say, because of inflammation caused by cigarette smoke), the greater the risk of a random mutation that will disable the normal mechanisms of cell death.

Environmental mutagens and carcinogens play a part by damaging DNA or increasing cell division rates (asbestos, for example, which destroys immune cells come to engulf the invader, leading to a cycle of cell damage) but they are only a tiny part of the picture.

Cancer is caused by math, in a very real sense. Mutation rates * cell division rates = risk of cancer. You can absolutely get cancer without any adverse environmental exposure, which is why some children get cancer very early on--genetic factors increase risk.

The universe is chaotic. There is often no reason for what happens to an individual. It's just random chance. Man's search for meaning is often futile.

We better also stop polluting the sun so we can cure skin cancer.

Cancer has always existed regardless of pollution. It's cell mutation, which is part of evolution. You can't fix it by getting rid of pollution. You fix it by making your own immune system recognize it again so that it can not the cell properly. Either that or restore the proper cell death instruction. Cancer deaths have been going down for years, firstly due to less smoking and less salted meat in our diets and now due to better understanding. It's slow, but it's steady and every oncologist in the world is waiting for the hockey stick moment.

There was an interesting study here [1].

They researched the prevalence of cancer in ancient societies using a variety of methods - remains, reporting, etc. The goal was to determine whether the relative lack of cancer in ancient societies was driven by some sort of external factor such as inability to detect cancer in remains, poor reporting, lower life expectancies, etc or rather it was simply because cancer was simply much less common. The evidence available tends to point very much towards the latter which, in turn, suggests that we have created environmental conditions which are driving cancer. The paper references two papers which have pointed to general pollution as a major potential driver there.

Death rates are not a relevant measurement on this topic either since they're heavily confounded by increase in survival rates. You need to look at incident rates, which continue to increase worldwide. Even lung cancer is now on the rise among females.

[1] - https://www.nature.com/articles/nrc2914

How does this paper account for the increase in life expectancy? Many types of cancer appear in people of older age and reaching that age was not nearly as common in ancient societies.

Two ways. By looking at the causes of death of those that did live to old age as well as by considering cancers (such as bone) that tend to occur disproportionately in young people.

I think the study is biased towards cancers that are visible in the remains. If you died from leukemia, would that be evident based on your skeleton?

I do not know the answer for leukemia but this is another issue that was considered in the paper.

If people are put off by the paywall: https://sci-hub.tw/#about

Literally the only thing they seem to say for sure in that paper is that they expect to see bone evidence they don't see. While I don't doubt the earnestness of the paper's authors, their primary study seems to be in Egyptology first. They practically gush about how the ancient Egyptian medicines are still in use today...but then only list ones based on papyrus. They don't account for the fact that our sedentary lifestyle is shown to increase cancer rates, they barely explore different cultures, with only a passing reference to Greece, who has numerous treatments, including excision to remove cancer, and admit that the ancient Egyptians didn't even classify benign tumors differently than no-benign. They don't account for the increased consumption of meat in a western diet. They don't account for massively higher numbers of humans.

Based on their paper alone I can think of at least 5 other hypothesis that they didn't even cover.

1. Periods of fasting and lack of food decrease cancer. 2. Less sedentary lifestyles decrease cancer risks(this has actually been shown to have some promise, with HIIT workouts showing improved response rates and decreased reccurance) 3. Decreased infant mortality caused an increase in susceptibility to cancer because weaker physical specimens are surviving. 4. Firearms. Cancer rates have increased with the output of firearms. 5. Decreased prayer to the Egyptian gods. As these gods declined cancer rates increased.

Some of these are obviously less serious than others, however it doesn't excuse the plausible possibilities. These aren't oncologists, their paleontologists, also known as historians who guess about things based on what they dig out of the ground.

https://www.nytimes.com/2015/08/14/world/asia/study-links-po...

https://dceg.cancer.gov/research/what-we-study/drinking-wate...

https://www.theguardian.com/us-news/2019/may/01/california-t...

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2782753/

To sum up - cancer is driven by our environment, not just random mutations. As commented by others - we are doing a lot more to our environment and food sources to drive cancer (and other health problems) than random chance.

This link is for fun: https://www.youtube.com/watch?v=wVnMBGXVVUI

Something like 60% of cancers are due to genetics - you inherit a faulty gene or a spontaneous mutation gives you one.

When it comes to environmental causes, tobacco smoke and diet make up the vast majority. Pollution barely registers.

Genetics are important, but when people from one population move to a different region they tend to get the cancers of the new population.

Prostate cancer and breast cancer rates are low in Osaka, but in Japanese people living in Hawaii the rates match the Hawaiian population.

Japanese people living in Japan have high rates of stomach cancer, but those rates drop for Japanese people living in Hawaii.

Yes, the difference in diets and smoking and drinking habits will do that. It's still not pollution in the examples you gave.

Reply on news.ycombinator.com