List of diseases with a known mechanism but no cure/treatment

[poetinsf]

@poetinsf
We all want treatments, preferably yesterday. But we have to think long term and go for the approach that on average takes the least amount of time. And that’s trying to figure out the core of the problem.

I think the main difference between you and I on this topic is that I'm way more pessimistic about the "basic science". Normally, things tend to converge as science make progress. In ME/CFS, it's been going in circles at best, or more like a random walk. There has been zero convergence.

Given the circumstance, even the attempt at a lotto will be worthwhile if we can approach it safely enough somehow. I don't see it being more expensive than the basic science that, according to @Jonathan Edwards, is wasted 95% of the time and fails to move the needle 1 nm after spending over a $1b.

 

[poetinsf]

The question was how to figure out which one to try.
So which was it? And why?

You could try the likely ones first and then move down the list. In the absence of other indicators, accumulation of anecdotes is one way to assess the likelihood.

Because if you actually tried them in doses adequate to know if they would work and ranges of protocols that you would need to check to know if they worked you would almost certainly kill at least 100 people through adverse reactions.

The safety profile of known compounds are known, aren't they? Just try safe compounds with safe dosage. It's going to be semi random anyway, so no point trying to figure out the right dosage for ME/CFS. I think one concern is making your ME/CFS is worse. Each patient should decide the risk they are willing to take.

 

[poetinsf]

The same reason you don't try to read the British Library's 13 million books in the hope of finding the one you want. It's not a rational way to go about it.

You could though, if you can go through 13 million books somehow. Search engines do that for us these days. Tomorrow, it could be a pharma AI that goes through 13 million compounds for us. For now, we could let 13 million ME/CFS patients try a dozen compounds each and see if any sticks out.

 

[Jonathan Edwards]

You could try the likely ones first and then move down the list.

And what pray is a likely one?

 

[Jonathan Edwards]

Just try safe compounds with safe dosage.

Nearly all drugs are on prescription for the simple reason that none of them are safe. Except ibuprofen maybe. Drugs that have a major impact on illnesses as sever as ME/CFS almost all have a small but significant mortality rate.

 

[Utsikt]

I think the main difference between you and I on this topic is that I'm way more pessimistic about the "basic science".

Given the circumstance, even the attempt at a lotto will be worthwhile if we can approach it safely enough somehow.

I disagree. The difference is not in our perception of how likely ‘basic science’ is to produce a result. It’s in how likely it is that ‘trying promising drugs’ will produce a result.

IMO that chance is essentially zero.

 

[V.R.T.]

I disagree. The difference is not in our perception of how likely ‘basic science’ is to produce a result. It’s in how likely it is that ‘trying promising drugs’ will produce a result.

IMO that chance is essentially zero.

I think the difference is in our perception of how likely basic science is to produce a result. Surely we feel that good science done methodically is likely to produce a result that will lead to treatment. Whereas trying random drugs is much less likely to.

Otherwise why are we all here?

 

[Utsikt]

I think the difference is in our perception of how likely basic science is to produce a result. Surely we feel that good science done methodically is likely to produce a result that will lead to treatment. Whereas trying random drugs is much less likely to.

Otherwise why are we all here?

Thinking through this again, I believe me and @poetinsf disagree on both how likely it is basic science will find a cure and for random experiments to find a cure.

I believe random experiments have a near zero chance of finding a cure. And that basic science has a substantially higher chance.

 

[EndME]

You could though, if you can go through 13 million books somehow. Search engines do that for us these days. Tomorrow, it could be a pharma AI that goes through 13 million compounds for us. For now, we could let 13 million ME/CFS patients try a dozen compounds each and see if any sticks out.

There is no point for a pharma AI to go through 13 million compounds because there is no indication on what is going wrong in ME/CFS. You could just as well ask your neighbors cat to type something on a keyboard and then take a compound that has a similar name to what was spelled. If you want your pharma AI to be able to tell you something meaningful you will have to feed it something meaningful. This will inevitably require basic research. If you now let 13 million ME/CFS patients try a dozen compounds each, you're just going to get the kind of useless anecdotal data we already have (assuming you're only trying out safe compounds as you're suggesting and the other scenario is obviously a lot worse).

 

[Creekside]

But what is your criteria for deciding what to spend money on and test? What makes cumin more likely to work than any other of the drugs above?

Since I haven't found any fancy drugs that seem likely, I select test treatments that are cheap, convenient, and safe. Containing untested potentially strong chemicals is also a positive factor. The carom seeds are a good example. As a side benefit, trying new foods might reveal something I'll want to include in my meals. The carom seeds fail that, since I didn't like the first taste, but $1.69 isn't a big loss. I'll still try a larger amount to check for an effect on ME.

My point of testing available foodstuffs is that it's (relatively) cheap and safe and convenient, and trying new foods might provide other benefits (a more interesting diet). If all the PWME around the world tested a few dozen foods they never tried before (and the more they differ from previous food families, the better), someone might find something helpful. Unfortunately, the foods that PWME haven't yet encountered are most likely to be found in more primitive societies, where PWME are less likely to know about ME. There might be a small plant or fungus that is an effective treatment for ME ... but it's limited to a few square km deep in the Amazon rainforest, which has had a human pass through once every few decades.

As for cumin, I didn't choose it as a possible treatment for ME. I simply noticed that I didn't have my expected ME response to a meal of curry. Experimentation verified the reliability of the effect, and further experimentation identified the cumin as the active ingredient. Larger doses provided temporary full remission, for at least a few days. Years later, I noticed that PEM from some mandatory exertion didn't occur, and my journal said that I'd had cumin the day before. Experimenting verified that cumin was effectively blocking my PEM. Further experiments with other herbs ruled out the constituents other than cuminaldehyde. I didn't manage to find any perilla, which also has a significant cuminaldehyde content, before the PEM blocking became permanent (cured of PEM).

So, it's not just trying new foods, it's paying attention to changes in ME symptoms, and then figuring out what might have caused the change.

 

[Creekside]

If you now let 13 million ME/CFS patients try a dozen compounds each, you're just going to get the kind of useless anecdotal data we already have (assuming you're only trying out safe compounds as you're suggesting and the other scenario is obviously a lot worse).

If 13 millions PWME try a dozen compounds each, just by choosing a food available locally that they haven't tried before, there's a chance that someone would find something that could lead to an understanding of ME. If a research group had done extensive testing on me, after exertion with PEM blocked or not blocked by cumin, maybe they'd have found out the mechanism of PEM.

 

[EndME]

If 13 millions PWME try a dozen compounds each, just by choosing a food available locally that they haven't tried before, there's a chance that someone would find something that could lead to an understanding of ME. If a research group had done extensive testing on me, after exertion with PEM blocked or not blocked by cumin, maybe they'd have found out the mechanism of PEM.

Yes, but many are already trying dozens of different compounds each day simply by eating food and drinking. In fact most of us have probably tried tens of thousands different compounds already. I've pretty much eaten all the local foods and non-local foods available to me. Now you can argue about dosage, but there is no indication in a random experiment what the right dosage should be! After all your own experiment would suggest that cumin doesn't work, precisely because if you ask multiple people to try cumin you wouldn't get any useful effects.

 

[EndME]

If a research group had done extensive testing on me, after exertion with PEM blocked or not blocked by cumin, maybe they'd have found out the mechanism of PEM.

Probably not because nobody knows what to test for in PEM. If we can't differentiate between healthy controls and pwME there's currently no indication anybody would have been able to see a "before and after" effect in you either. Having a single person to test would reduce some other noise but only having a single person also introduces a lot of noise!

 

[Kitty]

I believe random experiments have a near zero chance of finding a cure. And that basic science has a substantially higher chance.

Yup, and I'd add that it's true a lot of studies have been done, but most of them were rubbish. The work needs to be based on detailed clinical knowledge, epidemiology, clear thinking, and not making sh*t up. Too many failed that test.

Not much of the basic science has happened yet. Some of the datasets we'll need to work from have been gathered so recently they're still being analysed, and even they can only tell us a limited amount about the epidemiology and natural history.

Imagine it's another illness, one that none of us has any personal stake in. If the clinicians and researchers and patients are still struggling even to characterise and describe it, how much progress would we expect the research to have been made?

 

[Creekside]

I believe random experiments have a near zero chance of finding a cure. And that basic science has a substantially higher chance.

It isn't one or the other. There are a small number of PWME who have treatments that work reliably for them, for at least one symptom. We don't know what to test for for PEM, but using whatever methods are available on a patient with a reliable switch of states (PEM or PEM blocked, for example) is more likely to reveal something than testing some patients without that clear difference in states. Maybe the difference in a factor isn't noticeable in a single pair of tests, but maybe it will show up if you do 20 tests spread out over time. Maybe IL-77 is consistently higher in PEM than in PEM-blocked.

 

[Creekside]

patients are still struggling even to characterise and describe it

So true. Two days ago a eye specialist asked me what ME was ... and despite years of reading descriptions here, I still struggled to explain it.

 

[Utsikt]

It isn't one or the other.

I didn’t say it was one or the other. I talked about relative probabilities.

We don't know what to test for for PEM, but using whatever methods are available on a patient with a reliable switch of states (PEM or PEM blocked, for example) is more likely to reveal something than testing some patients without that clear difference in states.

Or those patients could have an inherently different underlying mechanism, so whatever you uncover might just apply to them?

 

[poetinsf]

And what pray is a likely one?

How about things that current crop of hypotheses predicts or implicates? Then move on to ones that accumulated anecdotes. In the absence of any indicator, resort to random ones. I'm thinking there may be a way to systematically approach it, collectively. Screen drugs that are safe enough to try, make them available to willing patients in random fashion and collect results. (No, I'm not a medical professional!)

Nearly all drugs are on prescription for the simple reason that none of them are safe.

That would depend on how unsafe they are. The side effects of those drugs are what we are looking for, so not being safe to some extent is a prerequisite, I'd think.

Except ibuprofen maybe. Drugs that have a major impact on illnesses as sever as ME/CFS almost all have a small but significant mortality rate.

Not sure if severe side effects are a prerequisite. If some OTC or supplement is effective for some people (anecdotes, yes), I don't see a reason why non-lethal prescription drugs can't be effective.

 

[poetinsf]

I think the difference is in our perception of how likely basic science is to produce a result. Surely we feel that good science done methodically is likely to produce a result that will lead to treatment. Whereas trying random drugs is much less likely to.

Otherwise why are we all here?

Passive observations and RCT are not the only way to do the science. You can figure out by effectuating the change, learn how the system behaves and then moving on to the next (semi) random effectuating till get to the solution in Bayesian way. That is often how we fix computer bugs, and how science progresses in other fields. Medical science is obviously more complicated and is bound by Hippocrates' Oath. But that doesn't mean that we need to exclude debugging method even when opportunity is available.

Drugs is one way to effectuate that change. Try different drugs, see how ME/CFS responds and that could give us information even if that doesn't bring cure.

 

[poetinsf]

This will inevitably require basic research.

True. I don't think anybody is saying that we should stop basic research. But the "basic research" is not the only way to solve the problem or do the science.

If you now let 13 million ME/CFS patients try a dozen compounds each, you're just going to get the kind of useless anecdotal data we already have (assuming you're only trying out safe compounds as you're suggesting and the other scenario is obviously a lot worse).

Do we know that for sure? We should at least go through hundreds of thousands of compounds out there, screen ones safe enough, and see what has not been tried.