We will diagnose Alzheimer’s Disease (AD) without only referring to clinical labels – which is helpful, since the clinical presentation can differ greatly even when the underlying pathology is the same. We will be able to establish prognosis not only at the group level but in the individual patient presenting to the memory clinic, so establishing if and when treatment needs to be started. We will have a way of selecting the drug most likely to be helpful in a specific case.

More generally, we will have the capacity to assess drugs in development for their ability to target relevant proteinopathies, and to monitor their efficacy in practice. And not just apply to drugs, since a slowing in loss of hippocampal volume seems able to chart the effects of a nutritional intervention, for example.

All of this will happen in a future made possible by biomarkers.

This was the vision offered by Philip Scheltens (of the Alzheimer Center, VU University Medical Center, Amsterdam, the Netherlands) in his plenary address to the five thousand delegates – from seventy countries – gathered for the opening of the AAIC17 Congress in London. Biomarkers’ best days lie ahead. We ain’t seen nothing yet, Professor Scheltens quipped.

More seriously, he said, biomarkers increase our chances of realizing the dream of one day seeing survivors of AD, and even of being able to prevent the disease. The long delay between the emergence of pathology in the brain – revealed by biomarkers -- and the onset of symptoms offers a wide window of opportunity through which to intervene.

But there are issues with markers, especially in CSF. There is still a need to establish clinical validity, and clinical utility – and there are complex considerations regarding the ethics of disclosing biomarker results to patients, of reimbursement, and of educating physicians in their optimum use.

One day, we will have people who survive AD --  if we select the right patients at the right time and have drugs that do what we want in the brain

We live in exciting times, Professor Scheltens enthused. Among other reasons for being optimistic is the fact that we have new targets in the shape of loss of synaptic function and the physical loss of synapses, which can be shown through the EEG for example. It also seems likely that we will able to subdivide patients based on genetic risk factors such as APO-e4.

A clearer picture of trouble ahead

If you think you’re a good clinician, biomarkers can make you think otherwise. That is another aspect of their impact. In a study from the Netherlands, 18% of patients who had been diagnosed as AD on clinical grounds had that diagnosis changed by evidence that an amyloid PET scan was negative. A positive PET scan in patients who had been diagnosed as non-AD also caused clinicians to change their minds. Even when PET did not provide reasons for a re-think, knowing the results of a scan altered the degree of confidence doctors had in their diagnoses.

There are also grounds for thinking CSF biomarkers – as well as a patient being positive on PET -- can be clinically helpful in predicting which people with mild cognitive impairment (MCI) will progress to dementia and which will not. Abnormal CSF at baseline is associated with greater risk of cognitive decline over five years, while being biomarker-negative is associated (with greater predictive power) with a likelihood of no change in cognition. Moreover, there is some suggestion that the predictive value of biomarkers is additive: in people with MCI who have amyloid pathology, having CSF tau as well is linked to a steeper rate of decline. So it should become possible to select people for intervention – and/or for entry into trials -- based on a profile of biomarkers.

How people will be affected by disclosure of biomarker information remains to be clarified. Preliminary data from a memory clinic cohort suggest – perhaps against our expectations – that anxiety levels do not increase when amyloid data are disclosed. Predictably, levels of uncertainty declines. The Alzheimer’s Disease Biomarkers in Daily Practice (ABIDE) study is aiming to establish whether biomarkers can routinely be used to guide treatment decisions in individual patients.

We have already entered an era in which the diagnosis of AD will be defined by amyloid and tau

Biomarkers first!

We have already entered an era in which the diagnosis of AD will be defined by amyloid and tau pathology rather than clinically. We have known for many years that this is the basis of the disease, but had not abandoned symptomatology. Despite common pathology, phenotypes are different, some being dominated by memory, others by visuospatial difficulties or language. All sufferers have amyloid pathology in the brain, but its presence and location does not necessarily correlate with symptoms.  The distribution of tau, however, does correspond to the brain areas we would expect based on phenotype.

Our understanding of different subytpes of dementia in general is also likely to be helped by biomarkers. Of patients who have dementia with Lewy bodies, for example, around half have abnormal CSF biomarkers indicative of AD; and those who are biomarker-positive progress at a faster rate.