A dementia deep dive
Looking at the topic through the lens of an impact investor
Dementia is a sensitive topic and one of the health conditions that people are most afraid of. It’s also a daunting area for early stage investors to identify what technologies have the potential to create meaningful and lasting change. For example, how do you assess investing in earlier diagnostic technology if there are no apparent effective therapeutics?
We at Eka believe that dementia is an incredibly important area for investment in new diagnostic and treatment technologies given its increasing prevalence due to ageing populations. Therefore, the purpose of this article is threefold: to give the reader a high level foundational understanding of dementia, to explain where we think some of the opportunities may lie and to highlight a few examples of great work being done by startups in this space.
What is dementia?
Dementia is the term given to a syndrome (a group of related symptoms) associated with an ongoing decline of brain functioning. These symptoms include problems with memory loss, mental agility, language and behaviour. The two most common types of dementia are Alzheimer’s disease (AD) and vascular dementia, accounting for ~70% and ~10% of dementia cases respectively.
While the exact aetiology of AD is not yet fully understood, it is thought to be caused by the abnormal build-up of proteins in and around brain cells. One of the proteins involved is called amyloid, deposits of which form plaques around brain cells. The other protein is called tau, deposits of which form tangles within brain cells. Although it's not exactly known as to what causes this process to begin, scientists now understand that it begins many years before symptoms appear. As brain cells become affected, there's also a decrease in chemical messengers (called neurotransmitters) involved in sending messages between brain cells.
Vascular dementia is caused by reduced blood flow to the brain, which damages and eventually kills brain cells. The typical causes of this include a narrowing of the blood vessels inside the brain and/or having had a stroke or multiple mini strokes.
The seven stages of dementia are shown below and can broadly be divided into three main classes: pre-clinical (no symptoms), mild cognitive impairment (early signs) and dementia (clear signs).
There are an estimated 944,000 people living with dementia in the UK and it is most prevalent in people over the age of 65. The risk of AD and other types of dementia increase with age, affecting an estimated 1 in 14 people over the age of 65 and 1 in 6 people over the age of 80.
What are the risk factors?
It’s useful to examine what are the key risk factors that increase the likelihood of dementia to understand how it can be better diagnosed and treated.
Age
Age is the single most significant factor. The likelihood of developing AD doubles every 5 years after the age of 65. But it's not just older people who are at risk of developing AD. Around 1 in 20 people with the condition are under 65. This is called early- or young-onset AD and it can affect people from around the age of 40.
Family History
Inherited genes can contribute to the risk of developing AD, although the actual increase in risk is small. In a few families, AD is caused by the inheritance of a single gene and the risks of the condition being passed on are much higher. APOE4 is the strongest risk factor gene for AD, although inheriting APOE4 does not mean a person will definitely develop the disease.
Cardiovascular Disease
Research shows that several lifestyle factors and conditions associated with cardiovascular disease can increase the risk of AD. These include smoking, obesity, diabetes, high blood pressure and high cholesterol. This risk can be reduced by stopping smoking, eating a healthy, balanced diet, leading an active life (both physically and mentally), losing weight if needed, drinking less alcohol and having regular checks as one gets older.
Other Risk Factors
In addition, the latest research suggests that other factors are also important, although this does not mean these factors are directly responsible for causing dementia. These include hearing loss, having had severe head injuries, untreated depression (though depression can also be one of the symptoms of AD), loneliness or social isolation and/or a sedentary lifestyle.
What is the current diagnosis pathway?
Dementia is one of the health conditions that people are most afraid of. A study by the Alzheimer's Society has shown that more than half of people wait for up to a year before getting help for dementia symptoms, because they feel afraid. While an accurate and early diagnosis of AD can have many benefits, currently only around two thirds of people with dementia have a formal diagnosis of their condition.
Dementia can be difficult to diagnose, especially if symptoms are mild. If the GP has been able to rule out other causes for the symptoms, the patient will be referred to a healthcare professional who specialises in diagnosing dementia. The specialist will usually assess the patient’s mental abilities using tests known as cognitive assessments, which typically involve a series of pen and paper tests and questions, each of which carries a score. It's important to note that test scores may be influenced by a person's level of education. For example, someone who cannot read or write very well may have a lower score, but they may not have dementia. Therefore, while these tests can help doctors work out what's happening, they should ideally never be used by themselves to diagnose dementia.
To rule out other possible causes of your symptoms and look for possible signs of damage caused by AD, the specialist may recommend having a brain scan. This could be a CT scan or MRI scan. It may take several appointments and tests over many months before a diagnosis of AD can be confirmed, although often it may be diagnosed more quickly than this. It is important to note that there is no definitive test for diagnosing AD or any of the other common causes of dementia. Findings from a variety of sources and tests must be pooled before a diagnosis can be made and the process can be complex and time consuming. Even then, uncertainty may still remain, and the diagnosis is often conveyed as “possible” or “probable”. Despite this uncertainty, a diagnosis is accurate around 90% of the time.
Clinical trials and drug development
The world is unfortunately still lacking in a drug that can have a dramatic effect on slowing or curing AD. There are many reasons behind this, one being that the underlying scientific mechanisms of the disease are not yet fully understood and there is a large heterogeneity of people with dementia and pathologies responsible. Despite this, one would still expect that Big Pharma would be rushing to develop drugs given the scale of the problem and the revenue opportunity it represents. Unfortunately, this is where costly clinical trials come in.
Compared to other therapeutic areas, AD clinical trials are complicated, expensive, and slow. As we’ve seen above, it’s hard to accurately diagnose AD in a timely and cost effective manner. Screening procedures include neurocognitive tests, MRIs, PET scans and CSF tests, all of which are expensive and time-consuming, especially for asymptomatic patients. The screen-failure rate of a clinical trial refers to the percent of subjects who undergo screening but do not meet the enrolment criteria of a trial and is a key driver of costs for clinical trials. The chart below shows how the screen failure rates for AD are higher than that of other therapeutic areas and that this increases with the earlier the progression of the disease.
The result of these high screen failure rates is that AD disease modifying therapy (DMT) trials tend to take a lot longer than trials in other therapeutic areas due to the difficulty of recruiting each patient. Furthermore, many of the trials are about trying to slow the progression of the disease and therefore the trial may require years and not months to prove this. Not only are the trials longer but they are also more expensive. Patient-screening costs account for 50–70% of total per-patient costs for AD trials. For instance, each cognitive test can cost $600, an MRI $2,400 and each amyloid pathology test can cost anywhere from $1,250 (for a CSF test) to $8,000 (for a PET scan). Screen failures at each step mean more tests for more patients and more costly trials for sponsors. This high cost and lengthy nature of AD clinical trials are likely to be a major reason that we are still lacking effective AD therapeutics.
The economic impact
Finally, while it may seem quite utilitarian to look at the economic impact of such a devastating disease, it’s useful to assess it through this capitalist lens to gain an understanding of the size of the problem and whom it affects.
The high economic burden of dementia is spread across three main areas: healthcare, social care and informal care. In the UK, healthcare costs mainly relate to the NHS and are due to hospitalisation of people living with dementia. Social care costs relate to services such as nursing homes, home care and respite care. Informal care costs relate to the family providing unpaid care for people living with dementia.
A recent study in 2019 estimated that the total cost of dementia to the UK economy is £34.7bn, which is ~1% of GDP. This cost is forecasted to rise to £94.1bn by 2040 due to an ageing population. The cost of social care in the UK was estimated to be £15.7bn and it was found that families are providing £13.9bn in unpaid care. It’s interesting to note that social and informal care make up around 85% of the costs of dementia, which sit outside the clinical setting.
Where are the opportunities?
Having examined the status quo, we will now look at some of the areas in which we see opportunities for technology to create more positive outcomes for people suffering from dementia in the future.
Earlier and more effective diagnosis
Current diagnostic methods for dementia/Alzheimer’s are expensive, complex, intimidating and long winded. This creates stress for the patient, financial burden for the healthcare system and results in only two thirds of patients having a formal diagnosis. There is also evidence that medicines aimed at slowing the onset of dementia are more effective if given early in its development. Earlier diagnosis is also important from the aspect of allowing the patient and family to get their affairs in order and organise community support. This can ensure the patient lives independently for longer and reduces the emotional and financial cost of care.
Novoic is one such company in the UK tackling this problem. They analyse how people speak using AI to detect subtle cognitive impairment and signs of AD neuropathology. They’ve already demonstrated impressive initial results and are now being rolled out to a massive trial involving 20,000 patients. Other startups outside of the UK looking to detect early signs of mild cognitive impairment or dementia using speech analysis include Langaware, ki:elements, Winterlight Labs and Canary Speech.
It’s also worth noting that there is emerging literary evidence that the onset of dementia can be prevented or delayed if detected at the mild cognitive impairment stage and lifestyle changes are enacted. A recent Lancet paper suggested that 40% of dementia cases could be prevented or delayed by targeting 12 risk factors throughout life. However, the evidence on this is still light and conflicting as it’s hard to conduct large clinical trials on this without an effective methodology of diagnosing mild cognitive impairment, which again adds to the importance of developing these diagnostic technologies.
Faster and cheaper clinical trials
The development of effective therapeutics for dementia is hampered by the slow speed and costly nature of running a clinical trial in this space. Early, effective and cheap diagnosis of AD has the potential to radically alter this. However, clinical trials in general are still too slow and expensive to run, with issues around site selection, recruitment, retention, data collection and monitoring to name but a few. In a data-driven digital world, it feels like there has to be a way to speed up these clunky trials to make them faster, cheaper and more effective.
Lindus Health is a UK startup that’s aiming to become the next-gen CRO for clinical trials, delivering the full end-to-end solution from study design and recruitment through to delivery and other services. They’ve done over 60 clinical trials to date and claim to be 3x faster than incumbents. AQ Trials is another startup that’s building the operating system for allowing sites to run clinical trials in a smarter way, managing everything from one platform. Stitch Health is a platform that aims to improve the patient experience and gather their feedback to reduce dropout. Finally, Huma is developing a platform for decentralised clinical trials, which uses digital technologies to monitor patients outside of the clinical setting, increasing access and improving patient reporting.
Personalised holistic disease management
There is a controversial book called The End of Alzheimer’s by Dr. Dale Bredesen which argues that AD is not a single disease but rather a collection of different processes that can lead to cognitive decline. The book outlines a personalised approach to tackling the root causes of these different processes by identifying and tackling the patient’s specific conditions through lifestyle changes, supplements and medication. However, there have been papers rebuking many claims within the book and the efficacy of the so-called Bredesen protocol given the lack of clinical evidence.
Personalisation is a core tenet of Eka’s healthcare thesis and regardless of the veracity of the above book, we think a personalised approach is important in dementia care given the large heterogeneity of affected people and the various stages at which the disease can present. Mindstep is a UK startup that exemplifies this by providing brain training programmes that are tailored to you as an individual. Five Lives is another startup based in France that has developed an app that has content and a brain game platform designed to detect and reduce the chances of getting dementia by managing five lifestyle areas linked to the disease: diet, mental stimulation, physical activity, sleep and mood. They recently raise €3.7m, with 20% of the round being allocated to collecting research to clinically prove that the app can prevent dementia.
We’d also be excited to see how emerging technologies such as causal AI can be used to understand and quantify the effects of multiple personalised interventions on a specific outcome and whether this could be used to accelerate the world of clinical trials.
AI for drug discovery
Given the lack of effective therapeutics for dementia, it would feel remiss of us not to mention the plethora of AI for drug discovery companies that have been founded in the UK over the last decade. Each of these has the potential to discover a drug that could slow down or cure dementia.
Isomorphic Labs is one of these companies that we’re most excited about given that it has Demis Hassibis at the helm and all the learnings from Deepmind and AlphaFold to build upon. Exscientia was the first major UK startup that combined AI with drug discovery and it has recently advanced a drug candidate for Alzheimer’s into human testing. AbsoluteAI is another recent entrant to this field that is designing a platform to find toxicology issues faster during the drug development phase so that these drugs can be screened out earlier in the development pathway.
Conclusion: There is hope
“The best minds of my generation are thinking about how to make people click ads”
Jeff Hammerbacher, early employee of Facebook
I stumbled across this quote when reading a talent manifesto that the founder of Novoic had written to persuade people to join their company and work on something meaningful. The quote is from 2011 and I like to think that much has changed in the decade since, with some of the best minds of this generation being drawn to work on meaningful problems. This is evidenced when you look at the teams being built at companies mentioned above.
I’m hopeful that the combination of this mission-driven top talent and the opportunities above will produce promising solutions to help tackle dementia. If you are working on such a solution, please do get in touch with us via our website or on LinkedIn, we’d love to learn more and see how we can help.