I think the major problem being made is people assuming Alzheimer's is one disease. Look at the history of cancer research for a comparison - it started out with all the researchers thinking cancer was a single disease and each researcher had their own pet theories as to the cause. We now know that each cancer is an unique swam of related cancers and the best treatment is found by identifying and exploiting the particular mutational weaknesses the swam displays.
You might wonder why this matters - the problem is if Alzheimer's is more than one disease then it is very hard to develop effective treatments as any new treatment will only work on a sub-population of patients. This makes it very hard to prove to the FDA that your treatment should be licensed.
Yes, disease classification is a rapidly moving target.
First off, AD is one of several currently classified neurodegenerative diseases. Others include frontotemporal dementia, "Parkinson's Plus" syndromes such as Lewy Body Dementia, Multiple System Atrophy and Supranuclear Palsy.
Some people classify vascular dementia(s) as a set of separate entities such as multi-infarct dementia, cerebrovascular small vessel disease and so on. It is however becoming increasingly clear that vascular factors affect most (if not all neurodegenerative diseases).
These classifications are based mostly on clinical evaluation which is tremendously unreliable. Bloodwork, CSF, genetic studies and imaging contribute to the clinical diagnosis (mostly neuroimaging - most notably MRI and PET scanning).
All the aforementioned diseases overlap to some extent and have patient subgroups within each of them, sometimes based on clinic feature (such as tremor-dominant parkinson's) and some on imaging features (such as limbic dominant AD).
Newer imaging methods are moving us towards dividing patients up into different phenotypes based on for example, patterns of tau deposition. Increasing genomic knowledge will likely also have a huge impact in this regard.
Yes until we get the sub-populations sorted it is going to be very hard to get anywhere with AD. Combine that with poor animal models, late diagnosis, and the need for large and very long human trials it is not surprising almost all pharma companies are avoiding AD.
From my understanding only one one major (Eli Lilly) now has a serious AD program - all the others have effectively walked away after losing tens of billions [1]. I have to say that Biogen’s antibody [2] is looking interesting. If I had to invest my own money in AD I would spend it on better early detection/sub-population classification and then on developing monoclonal antibodies targeting each sub-population.
Thank you for this. Really interesting blog, hadn't seen it before.
I can vouch for the fact that at least two other majors continue to invest heavily in AD diagnostics and/or therapeutics. You're right though that all have thrown a ton of money at this without much ROI yet (apart from those producing diagnostic radio-pharmaceuticals).
You might wonder why this matters - the problem is if Alzheimer's is more than one disease then it is very hard to develop effective treatments as any new treatment will only work on a sub-population of patients. This makes it very hard to prove to the FDA that your treatment should be licensed.