Hacker News new | comments | show | ask | jobs | submit login
A Battle Plan for a War on Rare Diseases: Conversation with Matt Might (nytimes.com)
61 points by onuralp 8 days ago | hide | past | web | favorite | 20 comments






Matt Might (the subject of this article) gave a jaw-dropping talk a year ago, named "Winning the War on Error - Solving the Halting Problem and Curing Cancer" [1]. At least you can't blame him for the lack of ambition! But the best part is he's really getting it done. His security research led to developing very successful static analysis tools that exposed security bugs in real code. And he did manage to find a treatment for his own son's rare genetic disorder.

[1] https://www.youtube.com/watch?v=MdmQUlD7P40


Off topic, but he also gave a great talk on parsing with derivatives https://www.youtube.com/watch?v=ZzsK8Am6dKU which was used by the clojure team https://paperswelove.org/2016/video/david-nolen-parsing-with...

Rare family diseases tend to be genetic diseases with a relatively simple genetic cause that can be found by comparing family members.

This article is about treating rare diseases, but there is a simpler way - having identified the genes for the disease, we could screen for them, and then advise people carrying them to reproduce only via selective IVF. That way, the next generation of that family can be free of carriers, eliminating the disease in the long term even if a treatment isn't found. (This is not an either or with finding a treatment)

The above approach is much cheaper, effective, but raises important moral questions we need to think hard about. At what point does a disease become bad enough that it needs to be excluded like this? Slippery slope, etc.


Perspective from a US physician:

Pre-implantation genetic diagnosis (or PGD, what you described as 'selective IVF') is not cheap. It costs upwards of US$100,000 and no commercial or government insurance will cover it.

Genetic screening is also not cheap. An exome (essentially looks at all ~22,000 known exons, or the genes that directly contribute to protein formation) costs US$6,000 - $8,000 and most insurances will fight tooth and nail not to have to cover it even when it is ordered by a geneticist. You might think "but there are much cheaper genetic tests besides an exome" which is true, but remember we are talking about rare diseases. Exome testing is still imperfect but it is essentially the only realistic option if your goal is to rule out as many known genetic diseases as possible, or if you know there is a high pretest probability of a genetic disease but less exhaustive testing was negative or inconclusive.

Actually, current PGD doesn't use exome testing but much more selective testing often based on the parents. One technique is FISH which can essentially only tell you if there are any "large" genetic abnormalities, like (micro-) duplications or deletions, which are large relative to, say, a point mutation. As an example, FISH testing would not reveal cystic fibrosis. The other is PCR, which is currently the technological limit for PGD testing because of the small number of cells and minimal amount of genetic material available for testing. There are several PCR assays available for specific conditions in the setting of PGD, which is where the parents fit in. We will probably see more and more assays becoming available, but it is unlikely that anything resembling an exome will be feasible for PGD in the foreseeable future.

Disclaimer: I am not a geneticist or reproductive specialist. If you can find sources by physicians in these fields, I would defer to them.


Matt Might makes the argument that there are a lot more rare diseases than many people realize. About every instance of cancer is effectively a rare disease (even though we agree to classify them according to a taxonomy), because each is caused by some kind of unique gene mutation which requires a specialized treatment, just like genetic disorders.

Furthermore, once a cancer is in the body, it can keep mutating, and one can have several (or many) slightly different kinds of cancer at the same time, which is one of the primary reasons it is so hard to treat them. So by creating tools to treat rare diseases we are getting closer to curing any and every kind of cancer.

And you can't prevent cancer by screening, we all become susceptible at later stages of our lives.


See here: http://www.counsyl.com/

But, about 30 genomic mutations are spontaneous every generation. I have one that causes brittle bone disease. So, just knowing the parents' genomic profile and screening out with IVF will not be sufficient.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3322360/


> he’s begun to create a road map for other families facing rare diseases — as much as 10 percent of the population, or 30 million Americans.

10% of our families are effected by "rare" diseases.

September is Pediatric Cancer Awareness Month: 100% of Pediatric Cancer is classified as "rare." It is used to give so little in research and most money is used to help the families financially. 4% of the US Cancer Grant is allocated to Pediatric Cancer research.

  Lost a sister at 15 to brain cancer and a son at 12 to bone cancer. The worst thing to do is to call it rare. How many thousands didn't start school this year? How many thousands of families have lost

Well it is rare, though. More kids are killed by automobiles than die of pediatric cancer.

That doesn't make it any easier for those families dealing with pediatric cancer. But it is rare, and if your goal is saving the most lives with a limited amount of money, there is lower-hanging fruit.


I really appreciate level-headed moralism like this. Many of our problems today are because we get painted into corners by people making emotional arguments and winning because the other side is unwilling to appear uncaring in the implied false dichotomy.

Forget that. How is that a win. I make a statement and then he can say only a few thousands kids die so tough. Sure I'm emotional but I can give some good reasons why Pediatric Cancer Research needs more funding.

Pediatric Cancer went over 20 years without a single chemo? How is that reasonable? https://www.theatlantic.com/health/archive/2013/01/our-dispr...

1) Pediatric Cancer Treatments are beneficial to adult cancers. They have less side effects and are usually more effective due to the fact that children's bodies and cells are all growing. Most adult treatments are too harsh and attack all the cells.

2) Treat a 60 year old smoker for lung cancer and you gain maybe 10 years. Treat a 10 year old and you have 60+ years saved. Most adults get cancer because of poor life choices.

3) Save your own life as a full grown adult or save a child. MOST of us would choice to save the kid. https://phys.org/news/2014-08-driverless-car-child-life.html


Americans spend $12 billion on bottled water every year, for example.

There’s enough money in the economy for people to pursue curing rare diseases.

I wish we could spend more of our $15 trillion economy on research.


Disregarding the morality issue, Matt Might makes the argument that there are a lot more rare diseases than many people realize. About every instance of cancer is effectively a rare disease (even though we agree to classify them according to a taxonomy), because each is caused by some kind of unique gene mutation which requires a specialized treatment, just like genetic disorders.

Furthermore, once a cancer is in the body, it can keep mutating, and one can have several (or many) slightly different kinds of cancer at the same time, which is one of the primary reasons it is so hard to treat them. So by creating tools to treat rare diseases we are getting closer to curing any and every kind of cancer.


These rare diseases are interesting in their utility. Previously one would think ignore them budgetwise - not out of callousness but because of compassion for others affected by more common diseases. However their oddball effects also boost understanding of otherwise obscure effects. Sonic hedgehog was named for the protein shape - and then it turned out defects in the linked gene can lead to horrifying birth defects and well SHH is a less frivolous term for it.

"Sonic hedgehog was named for the protein shape"

FWIW, Shh was named for the phenotype in Drosophila. It makes their bristles spiky.

I can't find a pic of Shh off the top of my head, but here's a gene with a similar phenotype in flies: https://twitter.com/kirkmmaxey/status/577886323912015872


I appreciate the intensity and urgency of this attempt, but surely there is a way to frame the work without resorting to militaristic metaphors. Surely there are other forms of expressing an effort than “the war on ...”?

If there’s one thing worth waging war on, I think it’s all these diseases since they have a “common enemy” position against mankind as a whole.

I worked with Matt at the US Digital Service. Amazing stand up guy!

Dr. Matt Might is one of the few people that I remember from high school [0]. He's the only one from that school whom I've communicated with post-graduation. [edit: one other, actually...]

[0] https://news.ycombinator.com/item?id=11366288

I think it important to put medicine's lack of progress in adequately treating many conditions in context, and only consider genetic factors as one consideration in the web of causality. Too often doctors make a diagnosis and treat the diagnosis. Much more work needs to be made in respecting the complexity of even the simplest of diagnoses.

Consider the simple ear infection. Sometimes children have regular ear infections, but these usually go away as the person grows up. I started having chronic ear infections in college, well after the usual age that one experiences chronic ear infections. The prescribed antibiotics took care of the problem for a few weeks before the infection returned. After a year of this I figured out how to take care of my ears myself with non-prescription interventions, and after maybe 6 years I figured out how to treat the causes behind my ears' poor drainage.

The GQ story "Sperm Count Zero" was recently discussed here: https://news.ycombinator.com/item?id=17942744

GQ didn't really drill down into the causes behind the male population's shrinking sperm counts, because the experts themselves aren't at all sure about the causes behind the symptom. While the causes of an individual man's low sperm count are unique in every case, it is a combination of factors that causes our society's shrinking sperm counts. The causes that I suspect include evolutionary-inappropriate modifications to our diet, mass exposure to endocrine disruptors (xeno-estrogens), etc. The Mayo Clinic's page about this condition [1] shows there's a few known simple causes for a man's low sperm count, but there is no effort to put society's condition in context.

[1] https://www.mayoclinic.org/diseases-conditions/low-sperm-cou...

Women are having similar responses to the causes behind men's shrinking sperm counts. Because women are supposed to release one egg a month rather than millions, their condition gets a label (PCOS/Endometriosis/anovulation/etc) and standardized treatments. Women are often subjected to palliative treatments for their diagnosis -- frequently this is some form of birth control. While some women find their condition to be helped with these prescriptions, medicine has failed to 'cure' their problem. They've just shut down the symptoms enough to make the patient's life a little more tolerable.

Sometimes women think they have "polycystic ovaries" [PCOS] because their mother had symptoms of polycystic ovaries, and think their own condition "must" be genetic. But habits get passed down too, and "stress" is a major factor in many women's conditions. Poverty is not genetic, but it usually gets passed down with all the other habits we learn from our parents.

Treating a woman's diagnosis of PCOS with birth control is using hormone analogues to suppress the normal functioning of the reproductive system. The use of these drugs for PCOS is a 'palliative' rather than 'curative' treatment.

There are certainly 'rare diseases' with genetic causes, but an effort also needs to be made to clear obsolete and harmful treatments from medicine's tool chest. Sometimes "war" is appropriate, but war can also be prevented with good diplomacy.


By definition obsolete treatments get removed only when something better comes along or it is proven lack of treatment leads to better outcomes. The least aggressive forms of prostate cancer defy conventional wisdom and are left untreated - putting aside quality of life treating it is more likely to kill the patient who would probably die of something else before it becomes a problem.

The place of doctors is to treat with today's proven means - experimenting is usually left to researchers unless there is little to lose. Medicine is a conservative field in the roll out. Given how many "advances" of newly discovered aspects don't work quite as expected like megadosing antioxidants it is pretty prudent.




Applications are open for YC Winter 2019

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

Search: