Why haven't biologists cured cancer?

This idea of "curing cancer" as if cancer was a monolithic disease really needs to die.

Cancer is just a catch-all term for somatic cell mutations that become unmanageable. There are as many ways to get "cancer" as there are ways to mutate cells into these sort of uncontrollable states. And there are cures for multiple cancers today (Imantinib, Herceptin, etc). There actually have been enormous recent strides in drugs like PD-1 inhibitors that do have a broad spectrum affect against cancers, but I think outside of this one particular example, the real math problem is how do you expect to contain a mutational state of a cell with 10000x ways to reach that state with a single, or even handful of drugs?

The little anecdote about Peter Thiel about shitty physicists going into biology reminded me of a chapter from Richard Feynman's book, "Surely Your Joking Mr. Feynman".

So it turns out Richard Feynman, after working in Los Alamos on the first nuclear bomb, and before going on to win a Nobel in Physics, actually did a stint as a biologist at CalTech. In fact he was doing some of the earliest work on ribosomes... so fundamental that he could have been the first person to discover that most all ribosomes were functionally equivalent:

"It would have been a fantastic and vital discovery if I had been a good biologist. But I wasn't a good biologist. We had a good idea, a good experiment, the right equipment, but I screwed it up: I gave her infected ribosomes ­­ the grossest possible error that you could make in an experiment like that. My ribosomes had been in the refrigerator for almost a month, and had become contaminated with some other living things. Had I prepared those ribosomes promptly over again and given them to her in a serious and careful way, with everything under control, that experiment would have worked, and we would have been the first to demonstrate the uniformity of life: the machinery of making proteins, the ribosomes, is the same in every creature. We were there at the right place, we were doing the right things, but I was doing things as an amateur -- stupid and sloppy."

So one of the world's best physicists almost became one of the world's most prominent molecular biologists, but actually fucked up the experiment for practical reasons.

I think this really underscores how not every problem is fundamentally a math problem... In the real world, especially with biology there's a shit ton of practical tedious things to deal with that can hamstring even the most brilliant experiments.

Contrary to the political priors of VCs, I think the real answers are pretty mundane:

1. Funding. Drugs have a low probability of success and a long lag time. Investors think in discount rates. A high-risk venture like biology is less appealing than an advertising-based tech platform with zero marginal costs.

2. Costs. Biology uses a LOT of proprietary instruments, kits, and chemical reagents. A lot. It also needs a lot of manual labor that would be difficult to roboticize.

3. Time. Biological experiments operate on biological timescales. Code takes seconds to run. Cell cultures take a day to grow. Even fancy new multiplexed sequencing assays take a while. You have the library prep time, the sequencing, and the downstream analysis. Its a long process. Now imagine waiting years and years to see if a drug in clinical trial prevents Alzheimer's.

4. Complexity. How do you make an equation for a giant network of weakly-interacting parts? Biology is a very "data-driven" field for this reason. The introduction of new microscopy, chemical conjugation techniques, and high-throughput assays has only made things worse. I genuinely hope some black box AI will be able to help us make sense of this mess and cure cancer. But medicine is full of interventions and incomplete prior histories, which will make naive association models hard to use.

This article is a good illustration of what's wrong with the current predominant approach in biology — it's all about genes and proteins, it's all way too reductionist to yield any meaningful higher-level outcomes. It's like trying to understand and manipulate an alien technology equivalent to a computer running a React web app, except you're only ever looking at and manipulating individual transistors in the CPU.

As Michael Levin said it, "reductionism is aptly named, it reduces what you can do". Do check out his research btw, it's some seriously impressive stuff. According to him, cancer, in particular, is simply a result of a group of cells electrically disconnecting from the surrounding tissue. He was able to force them to reconnect in one of his experiments, curing the tumor without killing anything.

As someone who has been involved in and seen how cancer research is done, it is no surprise that progress is slow.

The "skilled", "famous" and "influential" scientists in cancer research spend almost all their time fighting for grants, undermining each other, jockeying for influence, exaggerating their own importance, publishing phony papers etc.

Underpaid novices, trainees, grad students, and postdocs with little prior experience do the actual work in the lab.

It is much, much worse than a reasonable person would imagine!

This reminds me of the company Loyal - a startup aiming at medical approaches to longevity by first treating pets/dogs. It's easier (I assume) from a regulation/moral standpoint to test therapies on animals instead of humans. I wonder if developing therapies for animal cancers first could be profitable, to bring in investment and prove a theory, and then expand to humans in a similar way.

Lovely article. The part that resonates most with me is towards the end: What some software developers would refer to as "friction". Medical tests are high friction for reasons the author hinted at. I'm optimistic some of the cfDNA or fragmentome cancer screenings will be useful for catching cancer early. If you try to dig on the topics, you'll find that the projects (GRAIL etc) are proud their methods and datasets are proprietary. The ones that play the political and business game best are getting funding for trials.

It seems there is a steady stream of progress in biology and its medical tie-ins, but it is slow and cumbersome. Unrelated: It seems like a lot of the techniques in biology are discovered partly by chance (ie dedicated work in an area that bears fruit, perhaps not in an expected form). Ways to leverage various bacteria, phages, proteins etc to provide insight, or perform a manipulation. The molecular biology techniques we have available seem like a tiny fraction of what we could discover. And in a way, it seems as much engineering or tool-making as science.

This lecture[0] by Robert Austin, a physicist who has done extensive work in biomedical including cancer-related research, proposes a fundamentally different theory for how cancer develops and progresses. Namely, the mutational theory of cancer is wrong, and also the now conventional treatments we have for treating cancer patients, that is chemotherapy and radiotherapy, are often worse than the disease (both modalities promote metastasis of cancer as they evolve and migrate away from the source of the toxic stressors). Highly recommended to anyone who has an interest in the subject.

[0]: https://www.youtube.com/watch?v=Q7iWB6xbkwQ

Cancer is among the stickiest of medical wickets, for several reasons:

* Cancer is not a disease, it's a family of diseases, with a wide spectrum of causes, symptoms, and treatments.

* Cancer is, fundamentally, a mutiny of your own cells against the rest of your body. Because cancer is fundamentally a part of you, it can shanghai some of the resources your body normally uses for itself, like blood vessel infrastructure. It can also masquerade as a healthy part of you, for example to your immune system. Hell, part of your immune system can become cancer (as in leukemia)!

* Cancer metastasizes, which means that even if you get rid of the main tumor, a few of those rebellious cancer cells can use your circulatory or lymphatic systems to start new rebel colonies elsewhere in your body, which means you could be fucked all over again. Treating this form of cancer can be an arduous, painful, sickness-inducing process of chemo- and/or radiotherapy.

We are closer than ever to "a cure for cancer" with things like immunotherapy treatments which can program your immune system to target the specific kind of cancer you have. Unfortunately, most of these need to be tailored per patient, which procedure is neither cheap nor routine. And it still doesn't always work for all cancers!

What haven’t software engineers fixed all the bugs?

So Peter Thiel is an idiot.

Biology especially human biology is a lot harder simply because experiments have bigger ethical issues.

Even if you have desperate cancer patients there are limits how and what you can test.

And we have others fields where all these brilliant people according to him are working and all the money goes and the problems are also not solved.

Even the improvements in 5yr survival could be the fact that detection has improved. Cancer detected much earlier then before naturally improving the 5yr numbers.

Thanks for sharing a great article. Overall, I think the path to develop a drug is very multi-dimensional, cost and time intensive and very slow. Some other elements to add to the article:

- The preclinical drug discovery phase was not mentioned much. This phase involves discovering/designing/creating an actual drug against a potential biological mechanism/target uncovered by biology/genomics. Although this is generally seen as a more tractable element of the drug development path, it can still be very difficult and requires many years of additional research. Even so, some targets are well known to have great potential in disease, but it has been very difficult to generate a selective and potent drug against it. This phase typically does involve more "theoretically defined" research (although it is still messy) with chemists, biophysicists and pharmacologists, which fit more into the article's mention of the "mathematician". Yet, in line with the article, these often-talented people cannot always create a suitable drug for a given target. Providing further evidence that it is not "just" a lack of mathematically talented individuals.

- The reasons for drug failures in the clinic can be more numerous than alluded to in the article. Some drugs are very toxic, not only because of off-target side effects, but potentially also due to on-target side effects. In deadly cancers, this might not be so much of a problem, but it can still be a limiting element when we combine drugs to limit the surfacing of drug resistant cancer cells.

- As mentioned by others, cancer cells are cells from our own body, and they utilize our body's functions in excessive or highly altered manners to grow. However, blocking these functions selectively in cancer cells can be difficult (especially if non-genetically) as these functions often are still present in all other cells.

- Cancer metastasizes, cancer cells can spread across the body and generate new tumors elsewhere in the body. This can be almost anywhere, and it can be very difficult to detect early metastases in a patient. Hence, stopping treatment too early, even though the doctors might not see any cancer cells and the treatment has strong side effects, means you could redevelop cancer. Moreover, some metastatic sites might be in locations that are hard to reach for a given drug, hence they might not be fully targeted by a given drug.

- Full-blown cancers typically do not develop solely because of a single driving genetic alteration. Instead, a series of 2~5 genetic/biological alterations from a potential pool of dozens of genetic factors in combination leads to an aggressive tumor. Note though that it can be true that a single genetic alteration is dominant and drives a large part of cancer growth. Even in a single cancer type (e.g. colon cancer) the combination of 2-5 alterations leading to aggressive cancer can be different. Moreover, even within a single patient, different metastatic sites might evolve on their own and acquire different combinations of these driving factors. Hence, to truly treat some cancers targeting multiple drivers would be ideal, and each patient might require a relatively unique approach.

- Cancer cells are genetically unstable and can rapidly alter their genetic makeup. The DNA of normal human cells consists of two sets of 23 chromosomes that are well-organized and add up to ~3 billion DNA base pairs (the code of life). Cancers show very variable chromosome numbers and some advanced cancers can have more than 100 chromosomes. Moreover, cancer chromosomes can be heavily altered, where pieces of other chromosomes integrate into others, translocate, bridge, reconnect. It can be a total soup of >10 billion DNA base pairs. Moreover, these changes are different for each cancer, so every cancer patient will be more or less unique. This genetic instability also allows cancer cells to rapidly mutate and adapt/develop resistance to a given drug treatment.

I don't have anything to add, but still wanted to say thanks for sharing. One of my friends passed away earlier this year due to cancer and one of the questions I've kept wondering about has been about our slow progress in curing cancer. (Which admittedly stems from my own utter ignorance on the subject!)

Edit: I've thought of something worth adding. I've been told that certain datasets related to biology and chemistry are normally paywalled. I wonder if restricted access to information due to copyright is also hampering the field's ability to iterate more quickly. It seems like a field that's serious about rapid iteration could make as much data and information available to the public to encourage increased collaboration and propagation of knowledge. Is Nature still considered the most prestigious journal?

That we haven't cured cancer is a problem of culture. And I'm not talking about that culture that we put in paintings and films, but about the more fundamental one that tells everybody what's the proper way to live a life. That base level of culture that tells people they should have kids and buy a car and a house (though not necessarily in that order), and go to a grill party from time to time. That part of culture that tells people is better to die of cancer than risk a leak of something dangerous from a laboratory, or letting the ayatolahs produce biological weapons.

In Vinge's novel "Rainbow's End" there is a particular fragment about how they had automated biological research, with robots at big connected research factories making all the experiments. Such factories would go a long way to shorten the "long cycles" the article mentions, and they perhaps would democratize bio-sciences research. I don't think there are any technological obstacles to building those factories; robots have been employed at factories for a long time and there many robots already used in expensive biological labs. But note the word "expensive". In cultural terms, "expensive" means "big commitment that has to be more important than anything else that can be bought with that money.". This is the cultural part I refer to. As a society, we discuss many things, but biological research is very, very low in that awareness list. In Europe, for example, putting spyware on people's phones is way higher in the agenda. And thus, the public would never approve a 20% (or any other number) of the government budget going towards biological research, unless they had material evidence that that money would indeed cure cancer. For the same reason, they wouldn't use use 20% of their savings to invest on biotech. But if, by the right means, those perceptions could be changed, we would see a revolution in biology and really substantial breakthroughs in life-expectancy.

Kinda reminds me of that software law, where any hardware gains are negated by bloated software. Any medical discoveries are negated by deteriorating diet.

$$$$$$$$$$

There have been many cures that get buried due to not being able to make money from them