About the study


This is lay information for the ‘All-in-One’ study, funded by the Alzheimer's Society. Aspects of the research plan have been updated to keep up with scientific advances made during the pandemic, to ensure the research remains relevant to the future of the NHS. The study will use blood samples, advanced MRI imaging and cognitive assessment in an All-in-One package to diagnose Alzheimer's disease and aid the prediction of progression.

What is Mild Cognitive Impairment (MCI)?

Between 5 and 20% of people over 60 experience mild cognitive impairment (MCI). MCI is defined as: decline over time in at least one cognitive area (e.g., memory, language, spatial orientation, or forward planning) that is confirmed by pen-and-paper testing. Patients with MCI do not have dementia – they are independent in activities of daily living - but often have the underlying diseases associated with dementia, like Alzheimer disease or damage to the blood vessels supplying the brain cells (neurons). Alzheimer disease is associated with deposits of Amyloid and Tau. Amyloid is a protein which builds up outside neurons before they begin to die but also accumulates with normal ageing. However, Tau builds up inside neurons only during certain disease processes, inevitably leading to cell death.

Does MCI lead to dementia?

People with MCI are at increased risk for developing dementia in the next two years. However, not all patients with MCI will ultimately develop dementia and in those who do, progression of symptoms is often very slow. Dementia will sometimes only develop after 5 or 10 years. This causes huge uncertainty for future planning, often resulting in a feeling of disempowerment for patients, relatives and doctors.

How do we predict who will decline?

There are multiple diagnostic techniques that can provide more certainty. These include scanning by magnetic resonance imaging (MRI) or positron emission tomography (PET) and lumbar puncture to examine cerebrospinal fluid (CSF). Using these techniques, research has shown that there are two specific biological markers that allow better prediction of progression from MCI to dementia within the next few years. The presence of high levels of Tau and altered levels of Amyloid, either on scans or in the CSF, can predict progression quite accurately. However, many barriers exist to the uptake of these diagnostic techniques.

PET scans are expensive (£1600), and not available for Alzheimer’s disease in the NHS. Analysis of CSF requires 1) the clinical skills to safely perform a lumbar puncture, and 2) specialised laboratories for sample processing. CSF analysis is currently being performed at only one or two laboratories nationally. People with MCI are often reluctant to have a lumbar puncture because of its perceived “riskiness” or “painfulness”. Even though that is no longer true, this has led to low uptake. Further, lumbar puncture can be complicated in those on certain medications, for example blood-thinners.

What has changed?

Between 2019 and 2022 there has been a quiet scientific revolution in the analysis of blood for markers of AD and other diseases. Fifteen years of work by investigators like our collaborator, Prof. Henrik Zetterberg, have come to fruition. It is now possible to use an ordinary blood sample to estimate the amount of various proteins in the brain, and therefore the amount of injury to neurons. This helps estimate the risk of progression to dementia in people with MCI. Therefore, for NHS patients there is an urgent clinical need to examine how we can use currently available diagnostic tools: pen-and-paper tests, inexpensive MRI scans and clinical history, alongside these new blood markers, in order to make better predictions for people with the earliest stages of cognitive decline.

What do we want to do?

We aim to test the real-world usefulness and acceptability (to patients and clinicians) of a package of diagnostic tests in people with mild cognitive impairment. Our objective is to prove that such a package can be used across the NHS. We will examine whether it would be cost effective for the NHS to roll this out across memory assessment services nationally, and whether we can even further reduce the MRI scanning duration to make the experience even easier for people with MCI.

The scientific developments in blood-based biomarkers have meant we are now pursuing up-to-date, cost effective tests which avoid the bottlenecks created in the NHS by the COVID19 pandemic. We therefore propose a study using high resolution MRI scanners in Manchester and Cambridge to demonstrate the clinical feasibility of a novel “All-in-One” diagnostic package. This will use combined blood markers, pen and paper tests and MRI scans to deliver standardised, blood-based estimates of brain Amyloid and Tau deposition and neuronal damage in MCI. Previous research studies often excluded patients with significant vascular risk factors, such as hypertension and diabetes, which are common in patients with MCI. These factors are also associated with dementia risk. We will examine whether comprehensive All-in-One assessment, including assessment of neuronal damage due to vascular causes, can deliver the same or better accuracy when predicting who goes on to develop dementia.

Is PET scanning still useful?

PET scans use a radioactive tracer which sticks to harmful proteins in the brain. PET scans still represent the gold standard for the diagnosis of Alzheimer disease in the brain during life. However, in the last 3 years much more research has focused on using PET markers sensitive to Tau, the protein inside brain cells, rather than Amyloid which coats the outside. This is because Tau is 1) more specific to Alzheimer’s disease, 2) does not increase in normal ageing and 3) is a better predictor of cognitive decline. However, PET scans remain extremely expensive, and Tau tracers do not currently have a clinical license in the UK. Tracer supply is precarious and relies on a just-in-time supply chain with expensive tracers, sometimes supplied from abroad. While PET scanning remains an important research tool, over the last 3 years it has become much less likely that it will be used in routine AD diagnosis in the future. The backlog in cancer diagnoses requiring PET scans which has been caused by COVID19 is also likely to make this investigation much less available.

Is MRI scanning still useful?

We still use MRI scans to examine the structure of the brain [see figure 1]. This remains important because of the high rate of cognitive problems caused by cerebrovascular disease which is only evident on MRI. MRI can provide us with measures of the volume of different segments of the brain, and how well the brain areas are connected, a technique called “diffusion imaging”. MRI scanners are widely available, which means that for the foreseeable future MRI scans will be a routine part of clinical diagnosis, especially at the earliest stages of disease. One of our aims is to use the high-quality information we will get from blood markers to develop new techniques in MRI scanning and image analysis. This will help make the best use of this already-existing, widespread, (and relatively cheap) technology. More recent advances in MRI technology like Arterial Spin Labelling (ASL) can tell us about the brain’s blood flow, and may be able to mimic some of the aspects of FDG PET scans. Arterial Spin Labelling [see figure 2] tracks the water in the blood as it circulates in the brain by briefly magnetising it first. An image is then captured after a few microseconds. This image is subtracted from a background image to measure brain blood flow.

What can blood tests tell us?

The blood markers we will be testing include the same two proteins we look for on PET scans (Beta-Amyloid and Tau) as well as markers of inflammation (Glial Fibrillary Acid Protein GFAP) and markers of damage to neurons (Neurofilament light-chain NFL). Taken as a group these markers may be able to tell us whether Alzheimer’s disease is present in the brain even if cognitive impairment is still very mild [see figure 4]. When used with MRI scanning and pen-and-paper tests, they may be able to tell us how much damage has been done by Alzheimer’s disease or cerebrovascular disease, and therefore what the prognosis is. During the development of blood tests, they are often perfected on blood samples from people with few comorbid illnesses, who are - on average - very well educated and well-off, with often very little diversity. We need to make sure that these tests are fit for purpose for everyone treated by the NHS, and that a diagnostic package like this can be properly interpreted by doctors and patients. We need to ensure it is acceptable to patients, including how best to disclose the results. Most importantly, we need to ensure it is an accurate predictor of decline within the next two years.

Methods and techniques

We will scan and sample 100 people with MCI using fast, high resolution MRI scanners in Manchester and Cambridge. Then, we will check every 6 months for two years whether they have developed dementia. To do this we will use two short scales: the Clinical Dementia Rating Scale (CDR) and the Addenbrooke’s Cognitive Evaluation Scale (ACE-III) [see figure 3]. The CDR takes 20 minutes with the patient and 20 minutes with an informant. The ACE-III takes 20 minutes with the patient. Overall assessment time would be about an hour or even less.

Participation in the study will be offered to NHS patients (age 50 or above) who are referred to memory assessment services. Once we confirm there is a cognitive problem in at least one area (e.g. memory or language) using the pen-and-paper tests and rule out other sources of cognitive problems (severe depression or alcohol for example), people can be included. Many of these people would be referred for an MRI scan anyway in the course of the memory clinic work, so the scan itself does not represent much extra effort for participants. The only addition to routine assessment is blood sampling. People who already have dementia will not be eligible. We will also not include people in whom early Alzheimer’s disease has already been ruled out, who cannot undergo MRI scanning, who have other known significant disorders of the brain, or those with terminal cancer. During the first study year we will conduct MRI scanning, pen-and-paper assessment, and blood sampling on 50 people in Manchester and 50 in Cambridge.

How long will people be followed up for?

Participants will be asked to have clinical follow-up assessments with repeated CDR and ACE-III every 6 months for two years to determine whether there has been progression to dementia. This will be the main outcome criterion to assess the accuracy of the new test. At any time, they will be free to leave or to join other studies. There will be clinical trials in Manchester and Cambridge which could reduce their risk of dementia. We expect that up to 20% of people to either drop out of the study or join a clinical trial and we have calculated that 80 people at follow-up will still be enough for good results.

Additional secondary questions will be addressed together with collaborators:

1. What are the relative merits of various MRI imaging markers, including Arterial Spin Labelling, and techniques to assess neuronal damage, shrinkage of the hippocampus and other brain regions, vascular lesions, or changes in blood flow?

2. Can we compare the cost of this 1hr diagnostic package to the cumulative cost of a wait-and-see approach in MCI?

3. Do brain networks decay in early disease based on MR data collected in Manchester (a potential future marker of neuronal damage), and how is this reflected in blood markers?

How will we recruit?

The study will be registered with the NIHR Clinical Research Network (Prof. O’Brien is the National Speciality Lead for Dementias) and Join Dementia Research. Recruitment will be from NHS services in the Greater Manchester Mental Health Foundation Trust, including brainHealth Manchester, our new specialist MCI clinic, and Cambridge & Peterborough NHS Foundation Trust.

Dissemination and implementation

How will we tell people about our results?

Naturally, we aim to publish the results of our research at scientific and clinical meetings and in scientific journals. We will advertise the results of the study on social media and through the Alzheimer Society and University websites, with press-releases to the national media & through patient information sites.

Will other people be able to use this technology?

If this is a sensitive, specific and cost-effective package for the diagnosis of underlying Alzheimer’s disease in people with MCI, it will help to make the case that a network of laboratories should be established capable of carrying out commercial, clinical versions of these tests, and that in combination with a routine MRI scan, people can have state-of-the-art diagnosis within everyday NHS practice.

What about national policy?

We will discuss results with NICE and commissioners. Both Prof. O’Brien and Prof. Burns are national opinion leaders in dementia. We will also see whether participants understand better after the consent and scanning procedure that Alzheimer disease is a process that precedes dementia, and this may help national policymaking. We will engage in the ongoing public debate to promote a better understanding of the biological causes, multiple risk factors and the clinical course of cognitive impairment and dementia.