How much money costs a breakthrough for PFS? Is a genetic study the way to go?

Key points:

  • A genetic study is very expensive
  • WGS is preferable
  • Good genetic studies takes often more than 1000 participants and can easily cost more than a million dollars
  • Unlikely that there is a strong genetic influence; unlikely that it is possible to genetically design a PFS mouse model
  • The money needed to solve PFS is much higher than anticipated. Only grants and viral campaigns can reach that number.
  • The problem with grants and how to solve it
  • The problem with data noise and how to solve it
  • Will a genetic study bring the breakthrough for PFS?
  • Genetics: Going for grants rather than for private money
  • A possible alternative to genetics
  • Aptamer-based proteomic analysis of the CSF
  • Last words

Before we start:
The following opinions are my own. Contact me to discuss your thoughts or if you have corrections.

How much money does it need to solve a complex disease like PFS:

If you have a look at other diseases that share a symptom overlap and you see how much money they raised, you get an idea of how much money could be needed for PFS. But as a note of caution, it can always be that we have a lucky early shot and an early study leads to a lottery win. For example, currently, 2 independent researchers are currently looking at 2 different compounds that they have hope for, could affect PFS positively. No need to get excited though. There is a high likelihood that these compounds have little to no effects, but you never know.

Example ALS:

ALS raised 115 million dollars with their extraordinary ice-bucket challenge in 2014. A great deal of the money went into genetic studies that yielded at least 2 new genes for ALS, but didn’t bring any huge breakthrough — ALS remains a deadly disease.

Example ME/CFS:

I couldn’t find exact numbers, but it looks like they raised at least 100 million in total over the years for research both from grants and private donations. ME/CFS still remains a devastating disease with a reduced life expectancy and little to no treatment options.

Example Parkinson’s:

One of my favourite foundations is the Michael Fox Foundation (you know, the young guy from back to the future who got Parkinson’s). He raised more than 1 billion dollars with his foundation and improved the progress of Parkinson’s research a lot. Still, the disease remains deadly with only symptomatic treatment options.

So taken together I estimate that we should aim to raise at least 100 million dollars in the next years if we realistically want to solve PFS. This number seems high, but it is not impossible. Of course, it needs private donations, but the vast deal has to come from outside the community like national/international research grants and/or really successful campaigns. As I said the ice-bucket campaign made it possible to raise 115 million dollars for ALS in 2014 alone. For such a campaign to actually work, it would need the community to unite and get active. Just a 10–20 people working on this wouldn’t be enough. That means the community has to awake from its “sleeping beauty” sleep ASAP. I talked to national and international funding agencies. There are hundreds of millions of dollars for rare disease research. But when I talked to the funding agencies they refused to recognise PFS as “rare disease” as it is caused by a medication. That means we are stuck in a governmental funding gap and there are little to no grants for us. I am currently pulling together the leaders of other post-drug syndromes. I hope to form an alliance that can do lobby work and put pressure on the national and international funding agencies to release specific grants for post-drug syndrome research. The problem here is that these agencies are very bureaucratic, it can take years for them to finally release a grant, therefore we have a lot of urgencies to increase pressure now and not in a year.

Problems with an early genetic approach:

I think if we had sufficient money — I mean some millions — a genetic study would be good, but at the current stage with very very limited resources, I see this very critical.

a) Whole-genome sequencing (WGS) vs Whole-exome sequencing (WES)

I talked to a researcher at the New York Genome Center. She said she favours whole-genome sequencing (WGS) over whole-exome sequencing (WES). The problem is that WGS is so much more expensive, but you capture also a lot more data that might go missing if you just do a WES.

b) Sample size, data quality:

She said it’s very difficult to estimate the numbers of patients that are needed to do a good study. This depends on the homogeneity (very similar patients) of the patient cohort (group) and also on how strong the genetic impact is. What we can say for sure is that PFS is not inherited. That means there is no dominant or recessive pattern. This means genetics can be an influence, but not the cause.

If genetics were the cause of PFS, PFS would occur naturally in rare cases just like it does in ALS, Alzheimer’s, Parkinson’s, MS but it doesn’t.

c) Costs:

She stated for weak associations you need usually a thousand or even more samples to achieve significance. Currently, we are calculating here with 100 PFS patients for a 300k study. So there is a good chance that the study either would cost 10x more or that the outcome would not be significant.

d) Homogeneity:

A very big problem with PFS is that the symptoms and symptom severity vary a lot from person to person. That means the underlying cause might not always the same (or it is a spectrum disease which I find likely). In this case, you would need a lot more patients, probably thousands to achieve significance. The researcher from the genomic centre recommended to make WGS in a small number of “very interesting” cases and “look if you find something”. That means f.e. looking at very extreme cases, or cases that got sick from just one pill. A caution here: I found that many who said they got sick from a single pill actually had been taken finasteride or Accutane before somewhen in their lives and they got sick at re-administration. She. recommended arrays to look for specific stuff. I am not an expert in genetics so, not exactly sure what kind of arrays she was talking about, but she mentioned Ilunima arrays.

e) Who has PFS and who has not = data noise:

Big, big unsolved and unrecognized problem: Who has really PFS and who just believes he has PFS but is actually suffering from something else?! We don’t know how big the percentage of patients is who actually don’t have PFS, but similar problems but from another cause. These people mess up the data. No one can say for sure if he has PFS as there is no data. Sometimes I am pointing out at this point that only 56% of the Melcangi patient cohort showed methylation in the SRD5A2 gene. What if the other 46% don’t even have PFS, but something else?! The same problem will occur in a genetic study, as we cannot say for sure who has PFS and who has not this might mess up the data and lower the significance. What we need is a very strong good and at best somehow validated questionnaire-based test that filters out the ones who are most likely to have PFS from the ones who are least likely to suffer from PFS. Only people with a high score should be allowed to participate in such a study not endanger the data and study results. I proposed and improvement of Axo’s very good survey as such a screening tool. But it’s up to them to give their consent on making adjustments and using it for this kind of purpose.

f) Will genetics really bring the breakthrough?

I heard that once we know the underlying genetics, we could create a real PFS mouse model that leads to an understanding of PFS.

I would be very careful here:

  • This might work for ALS, Parkinson’s and Alzheimer’s as these diseases naturally occur there must be a genetic factor to it. In PFS it’s different! I haven’t seen a single PFS case that naturally occurred. That means the strongest factor is not the genetic factor, but the drug Finasteride itself. That makes me very sceptical that it will be possible to create a genetically designed PFS mouse model.
  • Another thing I heard is: PFS is that rare, it must be genetical. That’s not true. There might be other rare constellations such as the individual microbiome composition, inherited epigenetics, stress, ROS, androgen levels before taking finasteride etc. Genetics are not the only factor for rare diseases.
  • Also, I want to point out again towards the ALS ice-bucket challenge that raised 115 million dollars. There are now more than 30 known genes, as far as I know, the investments into the genetic studies from that money did only yield 2 new genes. 7 years have passed since then and ALS is still poorly understood and there is no good treatment option. The guy who started the campaign in 2014 just passed away.

Going for grants not private money for a genetic study

I am afraid a genetic study would take 300k without an outcome that possibly won’t make much progress. Our community is very tiny and our resources are more than limited. I would rather like to see 10–20x 30K in vitro studies that yield 10–20 insights than a 300k study that doesn’t reach significance or comes up with a slight association. I think therefore it is too early to make a genetic study in a community that is lacking strong financial resources.

I think the idea of an untargeted approach like PH follows it is good, especially now, at the early days of PFS research.
Public grants if at all are usually released for very ambitious big studies. I know a very small association which received a 9M grant from Horizon2020 to build a biobank and do related studies.
We won’t get grants for small 20–100k studies such as small basic molecular biological research on in-vitro and mouse models. But it will be easier to get a 1M grant for a genetic study. Therefore I would be in favour of not wasting vast amounts of money for a genetic study but wait for a good grant opportunity to make this possible. With the 300K, we can do 3–30 smaller studies that yield other important molecular biological results that otherwise would have had problems getting funding.

Possible Alternatives to a genetic study

Looking at the cerebrospinal fluid again:

Personally, I would love to see a broad investigation in the cerebrospinal fluid again. From past results from the Melcangi study, we know there is something going awfully wrong in the brain of PFS patients, together with symptoms like loss of libido, loss of motivation, cognitive deficits, brain fog and insomnia it makes it very likely that there is much more to figure out there.

Melcangi’s rat model doesn’t reflect PFS in my opinion, but still, it is currently the closest model that we have to PFS. I found it very interesting that the one rat study yielded neuroinflammation in rat brain after withdrawal. That fits very well to the brain fog, head pressure, the cognitive dysfunction, the depression and insomnia in many patients. Neuroinflammation always means the involvement of the immune system. Some people have speculated on changes in the Th1/Th2 system. The least what we could do in another study that looks at the cerebrospinal fluid is to look at a full panel of different cytokines to see which immune pathways are activated, so we can see, what’s the involvement of the immune system here. Recent studies have done similar things for ME/CFS, Parkinson’s, ALS, MS and Alzheimer’s. So I think neuroimmunology deserves a closer look.

Adapter-based proteomic study with SomaScan

What I personally would find the best study would be an aptamer-based proteomic study of the cerebrospinal fluid. SomaLogic a company for proteomic analysis of body fluids has a new technology that is called SomaScan and that was used for several studies published in the journal Nature.

What is intriguing about their approach is that they can look at 4000 different markers at the same time for costs of 500 dollars per patient. Talking to SomaLogic they estimate that a study under 100k — including healthy controls — is feasible. With this approach, it would be possible to look at which pathways are up and which ones are downregulated. So this is not looking at the genes, but directly at the gene products. That means both genetic and epigenetic changes would leave their markers in the signatures, making it possible to have a closer look at certain genetics and epigenetics for a way lower cost. Also, the insights into the activated and downregulated pathways would give us a pretty good understanding of what is going on in PFS. The aptamer analysis is also possible for blood that means for an additional 100k one could even correlate the pathways of CFS and plasma and maybe even find a diagnostic marker for PFS (witch more than 4000 possible endpoints that’s not unlikely IMO.

Last words:
So my opinion:
Let’s go for genetics later, and if we do so let’s organise grant money. Let’s have a look at possible changed markers of the disease first. It will help us figure out what’s wrong and how we can fix it.

Physician & President of the PFS Research Association e.V.