heart disease deaths are a third of what they were in 1950, (thanks to innovations like statins, stents, and bypass surgery.)
I had a look into this a while back. My conclusion was that two big factors in the reduction in heart attack death rates (not numbers) was in large part due to the reduction in smoking rates particularly in older people and the dramatic reduction in the use of toxic trans fats in processed foods and butter substitutes.
The evidence for the life-saving qualities of the 3 items listed was not very strong in the studies I ciykd find. Bear in mind in particular that studies started on or before 2003 and meta-analyses incorporating such studies were conducted under lax rules that allowed all sorts of shenanigans e..g changing the end-points, "run-in periods" etc.
I don''t really want to get into a debate about this but be aware at least that the conclusions in the quote above are controversial.
I think understanding heart diesease both why it went down and what we can do to reduce it further is a whole other topic then cancer and while I'm also not sure about statins I didn't went deeper into that.
When it comes to reduction in smoking reducing heart attacks, we would also expect that it reduces cancer rates.
I have a lot to say about this but I will keep it short. First, I think you're underselling the insight that cancer isn't a single disease (the Atlantic headline was shitty; of course cancer is a disease). This wasn't obvious a priori. The fact that every case of cancer is a unique and horrible snowflake means that we can't expect "a cure for cancer" any more than we can expect "a cure for car trouble". You're right, however, that some things are more likely to go wrong than others, and routine sequencing of tumors from each individual patient can help identify which treatments are most likely to help.
Second, I think there's a link between the decrease in death rate from heart disease and the minimal death rate decline from cancer even with all the increased testing and new treatments. As they say, "something's gonna kill ya," and in my opinion dying from cancer at 65 instead of from a heart attack at 55 is still a win. As a comparator in a disease area where no treatments have really worked to date, see the death rates from Alzheimer's disease.
Third, I'm as bullish on cancer immunotherapy as the next guy, but it turns out that many tumors produce an immunosuppressive environment, where T cells and NK cells just don't do their thing very well. You can immunize against mutated protein fragments presented by the MHC all you want, but in an immunosuppressive microenvironment I still don't think you'll see those sweet sweet CRs.
Finally, even with all the regulatory barriers and misaligned incentives, pharma companies are still working on the best cancer therapy targets we know about. We (I work in pharma, so it's "we") certainly haven't hung up the "Mission Accomplished" banner and moved on. While I expect continued insights about basic cancer biology to come from academic labs that receive public funding, future therapies will continue to arrive primarily from the private sector. The potential pecuniary reward for even incremental increases in cancer survival rate is high enough to keep key players interested.
You're right, however, that some things are more likely to go wrong than others, and routine sequencing of tumors from each individual patient can help identify which treatments are most likely to help.
It can also help with amputating parts of people's body for no useful medical purpose.
Second, I think there's a link between the decrease in death rate from heart disease and the minimal death rate decline from cancer even with all the increased testing and new treatments. As they say, "something's gonna kill ya," and in my opinion dying from cancer at 65 instead of from a heart attack at 55 is still a win. As a comparator in a disease area where no treatments have really worked to date, see the death rates from Alzheimer's disease.
This does suggest that cancer treatments aren't completely worthless. The fact that the US manages compared to other OECD countries relatively low lifespan, high cancer survival rate and high cancer death rate however still points to goodharting to be responsible for a large part of the high cancer survival rate among OECD countries.
Third, I'm as bullish on cancer immunotherapy as the next guy, but it turns out that many tumors produce an immunosuppressive environment, where T cells and NK cells just don't do their thing very well.
There's a difference between natural NK cells and NK cells that gene manipulate to avoid features of the immunosuppressive enviroment. If a NK cells assumes that a cancer cell should have some tumor marker but the cancer cells of a particular patient don't have that particular tumor marker you can alter the NK cells to not care about that tumor marker.
At current technology we can't sequence a tumor and know what we have to change in the NK cells to make them work in the particular immunosuppressive environment of a particular patient for the NK cell to be operational in that enviroment, and then change them. It's still technology that we can build.
Finally, even with all the regulatory barriers and misaligned incentives, pharma companies are still working on the best cancer therapy targets we know about.
One of my main points was that working on cancer therapy targets is not where most of the attention should go but on underlying technology platforms.
DeepMind beating Big Pharma at protein folding prediction suggests relatively little investment in the basic technology. It's quite plausible that you actually need working protein folding prediction to do what I pointed towards above for modifying the NK cells to work in more immunosuppressive environments.
While I expect continued insights about basic cancer biology to come from academic labs that receive public funding, future therapies will continue to arrive primarily from the private sector.
I'm completely okay with the private sector developing therapies whenever they think that they can develop a working therapy and bring it profitably to market. My post was more about how the large chunk of government and non-profit money should be spent.
If would also be okay with simply redirected the cancer research budget elsewhere (like anti-aging) and leaving the problem completely to the private sector but there's political desire to use public funds to do something about aging.
Thanks for engaging! I think there's a real debate to be had about how public research money is spent. I put a higher expected value on continuing to fund basic cancer research than I think you do. I also am more bullish on doing working at the object level (going after specific targets) relative to the meta level (technology platforms). Maybe this is myopia on my part, working as I do in the pharmaceutical industry, but I have also spent a fair amount of time thinking about the problem.
DeepMind beating Big Pharma at protein folding prediction suggests relatively little investment in the basic technology.
I actually think DeepMind is plausibly the only entity in the world who could have made AlphaFold when they did. The sheer amount of compute involved puts it out of the reach of nearly everyone else, plus pharma companies would have found it hard to hire away the caliber of ML talent DeepMind attracts. There's a case to be made that this is a nearly-ideal outcome for the pharma industry: the problem was cracked, publicly, by a company with little to no interest in making medicines. My prediction is that DeepMind either licenses the technology to pharma companies or contracts with them on specific targets (if the compute requirement is prohibitive for licensing). That seems to satisfy the incentives of DeepMind (this should be a significant money-maker, plus good publicity if and when their structures help lead to new drugs) and pharma (get structures for important targets that we can't get other ways).
Solving protein folding doesn't only give you the ability to know how existing proteins fold. It also gives you the ability to design new proteins.
If you take a problem like the one of NK cells and cancer immunity against them there will be cases where the machinery that NK cells have by human nature won't be enough and you need to design new proteins for them to properly recognize the cancer cells.
You won't get there by just licensing some protein folding technology from Google.
The sheer amount of compute involved puts it out of the reach of nearly everyone else, plus pharma companies would have found it hard to hire away the caliber of ML talent DeepMind attracts.
Compute and programmers are something you can hire. If the big pharma companies would be functional, it would be appropriate for each of them to spend a billion per year on AI.
A few years ago I spoke a few times with someone doing new business development at Pfizer. According to his perspective Pfizer is too bureaucratic to effective develop software.
It might be that new biotech companies that can actually manage to employ programmers in a productive way without stiffling them with too much bureaucracy will win out.
While not in the cancer area Proteon Pharmaceuticals would be a company where bioinformatics is at the core of their phage therapy product. As they grow and slowly push out most of the antibiotics from the market they have a need for more IT investment to better simulate how changes in phages and hosts interact and which mutations have which effects.
It might be that after they have a really good system to simulate phages, they go and reapply their knowledge to NK cells.
A company like Moderna or BioNTech that focus on cancer vaccines might also make major IT investments to optimize target selection and fully simulate the cancer cells to know which antigens they actually present.
Solving protein folding doesn't only give you the ability to know how existing proteins fold. It also gives you the ability to design new proteins.
I don't agree with this claim. AlphaFold gives you the ability to calculate how a given amino acid sequence is likely to fold. That is very different from being able to predict an amino acid sequence that performs a specific function or even has a given shape. Small modifications of known shapes or functionalities would be tractable using AlphaFold's technology, but there are other ways to get that, for example directed evolution. Search in the space of amino acid sequences is possible in principle but even with several orders of magnitude increase in compute the size of the search space still seems intractable to me.
If the big pharma companies would be functional, it would be appropriate for each of them to spend a billion per year on AI.
Isn't this significantly more than DeepMind spends? I realize increased competition for ML talent would drive up salaries but I just can't see that kind of budget allocation happening for something that pharma companies don't consider to be core to their business.
Thanks again for engaging. It's been fun to see how someone in the Silicon Valley mindset looks at the biopharma landscape.
AlphaFold gives you the ability to calculate how a given amino acid sequence is likely to fold. That is very different from being able to predict an amino acid sequence that performs a specific function or even has a given shape. Small modifications of known shapes or functionalities would be tractable using AlphaFold's technology, but there are other ways to get that, for example directed evolution.
This is basically why you need more then just licensing the technology from AlphaFold. You actually need to employ a bunch of programmers for a few years. It's a hard problem but it's solveable.
Isn't this significantly more than DeepMind spends? I realize increased competition for ML talent would drive up salaries but I just can't see that kind of budget allocation happening for something that pharma companies don't consider to be core to their business.
Tesla is today worth more then the other car makers combined. The legacy car makers largely understood to late that driverless cars are going to be a core part of their business.
It might be that currently the idea of some big pharma companies is to wait for the equivalent of Cruise and then buy it.
Besides understanding what goes on in a cell with computer models and protein design, predicting phase 3 trial outcomes better based on phase 1 data would be another task where a lot of data can be gathered and analysed with machine learning. There's an incredible amount of profit in prediction phase 3 trial outcomes before doing the trial.
Thanks for this article. I was diagnosed with metastatic prostate cancer last year. With the caveat that you've obviously spent more time than me on understanding the general conditions of cancers and that I'm in Europe, I'd still like to give my impressions. I think way too much money is spent on cancer drug research compared to fundamental research understanding the human body. Doctors and patients are way too eager to spend a lot of money for small amounts of improvement in overall survival. I think you are too optimistic about immunotherapy. I was offered to participate in a trial and looked into it and for PCa the record is abysmal. The side effects are also significant. I decided to decline the trial (which did feel a bit selfish.)
Bear in mind a lot of studies are for me-too drugs i.e. slight variants of existing drugs that have the tremendous advantage of being patentable, even if they are no better. Such trials provide little benefit to humanity.
As a fellow member of the reluctant brotherhood I have seen many friends enter trials only to suffer greatly with no, or even a negative, effect on survival. (Sometimes, I suspect, people will have treatment because it allows them to avoid facing The Horrible Truth*).
*That they are indeed mortal.
When it comes to talking about the prospect of immunotherapy what's possible with today's technology and what's possible with better technology are not the same. The approaches we have today with NK cells are relatively unsophisticated compared to what's possible.
Suggestion: before theorizing why "survival rates" diverge from not-dying rates, explain the difference between the two metrics.
Epistemic status: This isn’t medical advice. To the extent that it’s advice it’s health policy advice. I’m no domain expert. If you are faced with the prospect of cancer, consult with multiple experts.
Introduction
Richard Nixon declared the war against cancer in 1971. Beau Biden, the son of Joe Biden, died in 2015 due to brain cancer. Having their child die before them is one of the worst experiences a parent can have. Joe Biden, then the vice-president, decided to start the Vice President's Cancer Moonshot. On the campaign trail on his run for presidency he declared in 2019: “I promise you if I'm elected president, you're going to see the single most important thing that changes America: We're gonna cure cancer.” Joe Biden essentially declared the second war against cancer.
With both Nixon and Biden wanting to fight wars on cancer the issue is essentially bipartisan. Biden has personal reasons for fighting the war that distinguish him from other recent presidents but they are not party-political. I applaud the ideal of fighting wars against cancer instead of burning resources to fight expensive wars against human people.
While I applaud the principle, the first war against cancer failed. In this article I want to lay out the problems with the policies that came with the last war on cancer and make a case of how we can approach health policy in a better way.
The difference between the cancer survival rate and the cancer death rate
One mainstream view of the war on cancer is that it was partly a success. Vincent DeVita writes in The 'War on Cancer' and its impact:
To a layperson that claim might seem impressive but as Sarah Constantin writes in Is cancer progress stagnating:
There are three possible explanations of why those two numbers diverge.
The longer lifespans thesis
While it's true that we increased lifespan and a higher lifespan increases the likelihood of getting cancer you would expect the same with heart disease. Given that we see a lot of success at cutting heart disease deaths but not cancer deaths, this thesis doesn't explain what's going on and why we aren't getting more progress in cancer which is the area at which we throw the most research dollars.
The toxic environment thesis
The first explanation is the toxic environment thesis. According to it we have a lot more cancer causing substances in our environment and as a result even though we are better at curing cancer, we have a similar amount of cancer deaths. In the last decades we drastically reduced air pollution, removed cancer causing substances such as asbestos and reduced smoking rates. We have regulations that try to remove cancer causing substances from the market. Given those efforts, we should expect a less toxic cancer causing environment.
If we have a more toxic environment, there's a significant unresolved policy failure. Scenarios such as microplastics causing as much cancer as the substances we removed, are scary and underexplored. There are strong lobbying interests against seriously studying products for safety issues that are currently not under our radar and the narrow focus of US cancer policy doesn't fight against them to see whether we have a lot of unknown toxic substances in our current environment.
The goodharting thesis
There's an easy and reliable way to significantly increase the cancer survival rate. If you double the amount of people that are diagnosed with cancer and the healthy people you diagnosed with cancer don't die due to cancer you massively increase your relative cancer survival rate.
One of the insights at the time the war against cancer started is that it's much easier to treat a small cancer in its early stages than it is to treat a larger cancer after time passed. Out of this insight, the idea that a good way to fight cancer is to increase the amount of early cancer diagnoses arose. Public health campaigns that taught the populations about early signs of cancer got started and especially in the US expensive medical imaging was promoted to catch cancer in its early stages.
This resulted in the US having the fifth place at highest cancer survival rates while at the same time having the fifth highest cancer death rate and only the 40th place at life expectancy. Cancer treatment that happens when it wasn't needed is very costly. Some women lost their breasts without gaining anything in return.
After passing the affordable care act, a taskforce of the Obama administration decided that they believe in the goodharting thesis enough to cut back on testing in 2009. In addition to reducing the amount of women who needlessly lose their breasts this policy was also a way to save healthcare costs and critics argued that it was a health care policy that reduced treatment quality to save costs.
Initially, the American Cancer Society spoke against it and it took them till 2015 to come around to recommend the same policies. The new language was still very broad and they never really owned up to the fact that the recommended policies that made their members a lot of money that needlessly took women's breasts for a long time. Without the American Cancer Society owning up to their problematic past, it's not an organization to listen to when we want better cancer policy.
Trump administrations reduction of regulations
Given the high cost of drug approval the Trump administration decided to create a right-to-try law for patients with life-threatening diseases to bypass the FDA's application process for "compassionate use" of experimental drugs. The American Cancer Society was again at the forefront of fighting the new policies.
What's cancer?
Given that we don't want to fall into the goodharting trap again, it's worth exploring what cancer happens to be.
Cells gather mutations as they divide and are exposed to external stressors. They have non-perfect repair mechanisms. When the genes for the repair mechanisms mutate there are a lot more mutations. Some mutations lead to cells constantly dividing even when they are in a situation where a normal cell wouldn't divide. Some mutations make the cell ignore signals to self-destruct. Some mutations lead to the cell producing telomerase to escape the hay flick limit that limits how often a cell can divide.
Mutations happen all the time and in the normal case the immune system catches the mutated cell and eliminates it. The immune system can eliminate cancer cells because they have a different cell surface then regular cells. When proteins get broken down inside a cell the cell presents substrings of the proteins on its cell wall via a process called antigen presentation. When genes mutate there are some substrings in the mutated proteins that don't appear in the other cells in the organism. Genes that are normally only expressed in fetal development can mutate to be expressed in adult humans and then the existence of the corresponding proteins is a signal for the cell being cancerous. Cancer cells can mutate to shut down antigen presentation and therefore don't show their mutated proteins. In that process they however also stop presenting the antigens that a normal cell presents which provides a different avenue for their recognition.
The immune system can fail to do its job either broadly in the body or in a specific location and problematic cells replicate in a way that results in cancer. Sometimes the immune system starts effectively fighting the cancer after it's already visible on imaging methods. Sometimes the genes for telomerase don't get expressed and while a cell cluster mutates in a visible way it stops growing when it goes against the hay flick limit.
Currently, we don't have a good idea to what extent the overall mutation rate due to environmental stressors, global immune system failure or local immune system failure in a specific part of the body is the driving force for cancer.
Our goal should either be to find out which people actually need treatment to survive their cancer or find treatments that have no harmful side-effects for early stage cancer so that it doesn't matter if we treat it even when it would go away on it's own. For those that would die without treatment it's okay to have treatments with serious side-effects, and we need better treatments for late stage cancers as well.
How much does cancer matter?
Suzanne Wu argued that given that curing all cancer would only extend lifespan by three years money would be better spent fighting aging then fighting cancer. While we want to extend lifespan a lot longer than three years, curing cancer would allow us to use other therapies more aggressively that we currently don't use because they have the risk of causing cancer. Anti-aging therapies to regrow parts of the body come inherently with cancer risks and we would get further with them if we wouldn't need to worry about cancer.
Growth hormone increases the speed in which cancer grows and for anti-aging therapies there’s a good chance that we will want to inject growth hormone.
Is Cancer a Disease?
The Atlantic wrote an article titled Cancer Isn't a Disease. That headline comes out of asking a person working in biotech “What is a common and/or annoying misconception about your vocation?” The slogan is that cancer isn’t one disease but a cancer is a collection of diseases.
The background of this statement is that different cancers indeed react differently to many treatments. Contrary to the reality of pharma companies doing many clinical trials of cancer drugs the person they ask asserts: “That fact alone—that cancer is a collection of diseases—dissuades Pharma from attacking it, with the absence of blockbuster potential. It’s becoming reminiscent of antibiotics, albeit for somewhat different reasons.”
If you ask yourself why Pharma develops drugs that are targeted at individual cancers a large part of the answer is the Orphan Drug Act of 1983. Under the Orphan Drug Act drugs, vaccines, and diagnostic agents would qualify for orphan status if they were intended to treat a disease affecting less than 200,000 American citizens. Orphan status inturn reduced the regulatory barriers for bringing drugs to market. Given that regulatory barriers constitute a major part of the cost of developing drugs, this encourages Pharma to develop drugs for orphan diseases instead of more general solutions.
Orphan drugs also make it easier to charge higher prices EvaluatePharma® estimates based on an analysis based on the top 100 drugs in the US in 2018an that the mean cost per patient per year of an orphan drug was $150,854 versus $33,654 for a non-orphan drug. The report predicts:
Pipeline orphan drugs account for over a third of total R&D pipeline sales through to 2024, with the annual growth rate from sales forecast to be 163% compared to 146% for non-orphan R&D products.
As a society we prefer if drug companies develop cheaper drugs that help more patients but set up our regulatory environment to encourage expensive drugs that help fewer people.
Orphan drug status also provides a few other advantages to drug companies that I won’t list here, see the EvaluatePharma® report for more information.
Effective use of research money
Within the NIH the National Cancer Institute had a budget of $6.9 billion in 2020. In contrast, the human genome project spent $5,1 billion in 2020 dollars between 1990 and 2003 to accelerate DNA sequencing technology and uncover the human genome. While the knowledge about the human genome that they published wasn't very useful, the technology that came out of the human genome project that allowed for cheap DNA sequencing to be developed turned out to be very useful.
Even if we only look at cancer the ability to sequence the genome of a cancer and thus get data about the mutations of the cancer of a particular cancer is plausible worth more then all the cancer research that the NCI funded in that timeframe. Sequencing is a basic building block for effective immunotherapy which is one of the most promising technologies to tackle cancer in the coming years as I will discuss later.
New technology often allows a research task to be done for a tenth of the price in a decade. A high percentage of public research dollars should go into technology development.
Cloud labs
When researchers use their equipment in their own lab, their priority isn't to improve their research technology but to make scientific findings in the domain of their grant to publish papers. If the scientists would instead concentrate on doing their science and outsource the execution of their experiments to cloud lab companies, the cloud lab companies could focus on bringing the cost of the experiments.
Besides allowing the cloud lab itself to optimize their technology, cloud labs also help with researcher productivity. EmeraldCloudLab for example claims that scientists who use their platform increase the amount of samples per year from 2,220 to 7,064, take 1 year to publish a paper instead of 1.96 years, reduce cost per paper from $146.3 K to $107.6 K and reduce time to first publication quality data from 3 months to 24 hours.
Cloud labs seem to be currently held back by requiring researchers to think differently about the way their lab works and having Phd students do less cheap manual work. Grant giving should earmark for a large portion of grants that aren't about building new technological capability part of the grant to pay for cloud lab costs.
Theoretic research
Researchers seek large research budgets, universities seek professors that are likely to bring in large research budgets. Running expensive experiments comes with more research costs than theoretical research. As a result of this dynamic we don’t have professors who research cancer completely on the theoretical level and integrate the large amount of information we have into coherent theories on a basic level. We should give out a type of grant that’s focused on theoretical research without experiments.
If we would have more theoretical researchers instead of just researchers who focus on experiments we would have understood the goodharting problem of cancer testing earlier. We don’t need to spend as much on theoretical research as we spend on experimental research but giving 1% of total cancer research money to theoretical research where the involved researchers aren’t engaging in experiments would be great.
Treatment perspectives for cancer
While we should be open to a lot of different treatment modalities I will discuss approaches here where I’m confident that executing them well will improve cancer care.
Cancer Immunotherapy
In the last decade cancer immunotherapy appeared on the scene. The idea of cancer immunotherapy is to help the body fight the cancer more effectively.
As we describe above, many cancer cells present antigens about their mutations on their cell wands. If we get the body to build antibodies against those antigens, the immune system uses those antigens to detect and fight cancer cells. In the beginning there was the hope that targeting proteins that normally don’t get expressed in adults is enough to attack the cancer. Clinical trials suggest that it isn’t. Fortunately, we can use gene sequencing to learn about all the mutations in a cancer. With the help of computer models we can determine which of those mutations will be displayed as antigen on the cell wall and vaccinate patients against those mutations that get displayed.
The benefit of this method is that it puts little stress on the patient, so it matters less when we use the method with a patient that doesn’t need treatment. For patients with more advanced cancers we can combine this method with other methods.
Multiple technology platforms might be usable for cancer vaccines. We could use traditional adjuvants, we could also use mRNA vaccines. Given that the technology is advanced enough that multiple companies are doing clinical trials, the field does not need non-commercial research money.
When it comes to cancer cells that remove antigen presentation mechanisms, natural killer cell based therapies already lead to approved cancer treatments. Like cancer vaccines the treatments have little toxicity. At this stage there are still many challenges that need research to optimize treatment. Just like cancer vaccines that are individualized to individual patients are better, natural kill cell based therapies that are individualized to target the cancer of a specific person are likely more effective and there are many open research questions that need to be solved, so the field needs funding.
We need technology to determine which antigens on the surface of cells of a particular cancer are missing.
We need technology to effectively grow natural killer cells in the lab that are specialized to be sensitive to particular missing antigens and not attack when antigens that are generally missing in a particular patient. While we are at it, we have to study whether we can increase in-vivo persistence.
Nutrient optimization
It’s likely that the blood nutrient content of cancer patients frequently deviates from optimal levels, given that cancer is taxing the organism a lot. While in most cases nutrient optimization won’t be enough to cure cancer alone it can easily be used in combination with other therapies to improve treatment success.
Needed technology
Blood testing
After Theranos failing the appetite to invest into a new generation of general blood testing technology is currently low. Having better and cheaper blood testing technology would allow us to take less blood from patients to get information about what goes on in the organism of cancer beyond cancer markers.
Understanding better what goes on in the whole body when it suffers from cancer helps us to progress science.
Therapeutically, understanding more variables of a patient gives us more points to intervene.
3D open source anatomical model creation
Human anatomy is a neglected research field. Important aspects of human anatomy such as the lymphatic system existing in the brain have only been found in 2015. Operating cancer with our current understanding of anatomy produces needless damage that we could avoid if we would understand human anatomy better.
Understanding anatomy better and what differences among healthy humans are normal will allow us to understand abnormal anatomy to detect cancer when it happens. Cancers produce stress on the organism by pressuring other parts of the body. A better understanding of anatomy is needed to understand the effects better. Sometimes anatomy will create a microenvironment that has effects on the cancer. Surgery comes with side effects and those can be reduced with better understanding of the underlying anatomy.
Instead of just gathering a 3D atlas of cancers, 3D models should include larger parts of the body. Instead of just having the 3D models as raw data we need open-source tools that turn the raw data into models with which both researchers and clinicians can better interact.
While commercial providers might produce 3D models out of raw data, having the models based on open-source software is essential for researchers who study aspects of the models and need to adapt them for their research questions.
We already scan the whole body of cancer patients to discover possible metastases. Better software would allow us to get more out of the data that we already gather.
Open Research questions:
What roles do transposons play in cancer?
With next-generation sequencing that sequences the DNA 100 base pairs at a time, it’s not possible to see when a transposon that’s 6000 bases in length gets duplicated. Third-generation sequencing brings us the ability to sequence 10000 base pairs at a time so that we can see how often a transposon is replicated.
Biologists often don’t care for what they can’t see and transposons just move into our view.The fact that while transposons regularly replicate within DNA the transposon count of our species stays constant. There needs to be a mechanism of how increased transposons reduce the fitness of individuals. The most plausible mechanism is that they regularly cut the DNA and induce mutations. Given that cancer happens downstream from DNA mutations, cancer might be one way that individuals with a transposon count that’s too high get wiped out.
PGBD5 that codes for a transposase is expressed in a majority of pediatric solid tumors while it’s possible that PGBD5 only gets expressed after the cancer grows a bit, it's also possible that PGBD5 produces mutations that create the majority of pediatric solid tumors.
If that’s the case, we have to check whether PGBD5 is needed or whether we can vaccinate against PGBD5 to let the immune system kill cells that express it long before cancer gets developed. We have to rethink which substances are cancerogenous based on how they interact with transposons.
The main approach shouldn’t be to think about how we can shut down DNA repair in cells affected by PGBD5 but how we can generally prevent PGBD5 from bringing cells into a cancerous state. Shutting down repair is not a strategy that’s likely to be a permanent solution as it just means that another cell that has problems with PGBD5 is going to mutate into a cancerous state. We should focus applied research in a way that actually has a chance of creating major progress.
How does the microenvironment around cancer affect cancer formation?
We don’t understand well how the immune system sometimes fails at detecting cancer and fighting it. It’s plausible that there are conditions under which the immune system works less well in certain parts of the body.
Drug approval
Prediction-based medicine (PBM) for compassionate use
The current state of affairs where doctors can give parents unwarranted hope when they sell treatments to the patients under the compassionate use clause sets bad incentives. To the extent that drugs get used without a company producing studies for them, that also removes the incentives to fund studies to investigate the merits of treatments.
We need a new mechanism to deal with patients whom we give treatments under compassionate use. I propose to use the mechanism of Prediction-based medicine (PBM). In PBM a doctor has to tell the patients about the likely outcomes of a treatment and submit his prediction to a central authority. That central authority then publishes aggregated data about the quality of the predictions of individual doctors and hospitals.
In the case of cancer, I propose telling the patient about the 1-year, 3-year and 5-year survival rate when he uses the treatment.
Under Prediction-based medicine doctors are incentivised to give patients drugs when they have justified belief that the drug will help the patients without an expensive approval process. Pharma companies in turn are incentivised to run studies that allow doctors that care about being a doctor with good prediction accuracy to make good predictions.
A side-effect of this system is that we can identify the best doctors at knowing whether a drug will help a patient in the absence of a formal drug approval, how we should evolve our standards for formal drug approval so that we approve drugs that work with a minimum of bureaucracy.
Pharma companies can also hire the best doctors at estimating the usefulness of drugs to guide them at making decisions about which clinical trials to run.
Drug approval denationalization
Currently, the incentives of the FDA are more about not approving drugs that pose risk of criticism. Drug approval denationalization is about creating competition between drug approval agencies of developed countries. In it we take a list of countries in whose systems we trust and declare that approval by any of those countries is enough to bring the drug to market.
End the Orphan drug act
The idea that a drug that’s taken by 300,000 patients needs different evidence then a drug that’s taken by 100,000 is flawed. In both cases patients deserve drugs that work and that don’t put them at risk. To the extent we need to allow usage of drugs where approval is too expensive, Prediction-based medicine is a better system.
Conclusion
The US discourse goodharted on the cancer survival rate which is a bad metric, instead the success of cancer policy should be measured by reducing death due to cancer. Spending money on tool building is often higher return then bringing another substance that might produce a small effect that changes little in the big picture to market.
While reducing the barriers to bringing new drugs to market, we still need to know which drugs work and experimenting with Prediction-based Medicine for drug use for compassionate use is a good way to get started. The Orphan drug act sets bad incentives as we want cheap drugs that help a lot of people instead of expensive drugs that help few people.
General platforms like cancer immunotherapy that can be adapted to different types of cancers are more promising than narrow drugs that only work in very specific kinds of cancer.
While I consider the areas towards which I point to be important, research should always be open for new approaches and not be too much committed to old strategy given that the nature of science means, that the domain under investigation is uncertain.