World Alzheimer's Day

Is Alzheimer's scary? Thus the diagnosis: from biomarkers of resilience to language analysis

Research on understanding and managing this neurodegenerative disease opens up to protective factors and screening with artificial intelligence

by Francesca Cerati

21 September 2024

Costi, cure e solitudine: così l’Alzheimer grava su pazienti e famiglie

4' min read

4' min read

The World Alzheimer's Day plays a key role in bringing the global community together to address the challenges posed by the disease and dementias in general, raising awareness, reducing stigma and raising hope for a future where these conditions are better understood and managed.

There is still a lot of work to be done on the understanding and management of this neurodegenerative disease - which affects about 5% of people over 60 and in Italy is estimated at about 500,000 patients (Iss data) with an imposing organisational and economic impact -. Not least because the number of people suffering from dementia is set to increase due to the ageing of the population.

Although Alzheimer's research is very vital, the goal of a cure in the true sense of the word is still a long way off, despite the fact that in the last three years, three monoclonal antibodies have been approved by the American Fda, but have not been approved by the European Medicines Agency, the Ema, because they are considered ineffective, very costly and have several, even serious, side effects.

Costi, cure e solitudine: così l’Alzheimer grava su pazienti e famiglie

Drugs

'The biological efficacy is there, but it is modest. The deposition of beta-amyloid disappears, but the disease proceeds anyway, so let's say that the biological and clinical aspects do not go hand in hand,' explains Paolo Maria Rossini, head of the department of neuroscience at the Irccs San Raffaele in Rome. 'Moreover, these therapies are very expensive, costing tens of thousands of euros. They are given intravenously and therefore require administration involving hospital facilities. Every 2-3 months an MRI should be done to see if there are any side effects, such as oedema or microhaemorrhages, and if there are, the therapy should be stopped'.

'Ema's extreme caution,' comments Camillo Marra, president of SinDem (SinDem's autonomous association for dementia), professor of Neurology and Neuro-Psychology at the Catholic University of the Sacred Heart and director of the Memory Clinic Unit of the Fondazione Policlinico Gemelli, 'is probably linked to the uncertainty surrounding the 'profiling' of patients who are candidates for treatment. These drugs are indeed expensive and not without undesirable effects; it would therefore be desirable to administer them only to patients who really need them and can respond to these therapies. In short, it would be better to define when to use them, how to use them and for how long'. But the information to answer these questions is still lacking.

Biomarkers of disease

'At the moment, we only have disease biomarkers,' Marra continues, 'But the positivity of beta-amyloid or phospho-tau biomarkers and the presence of neuro-inflammation markers only tells part of the story, because this does not give us information about the speed and time in which that particular patient will reach the development of the first cognitive symptoms of Alzheimer's. An eventuality that might occur for some subjects after a year, for others after 20-30 years, for others even never, thus making medication useless'. In practice, it is not enough to define only the biological diagnostic criteria of the disease, but it is necessary to identify the protective factors, resilience, cognitive and cerebral reserve factors, which ensure that the disease in certain individuals will perhaps never manifest itself.

Resilience Tests

From this perspective, the view on predictive testing takes on a different connotation when considering resilience (protection) biomarkers, as opposed to disease biomarkers.

'The resilience factors that allow those with the altered biomarker not to fall ill should be investigated much more,' Rossini resumes. 'Alzheimer's and other dementias are diseases that work in the dark for years or decades before symptoms become apparent, in a continuous 'ping-pong' between risk factors and resilience factors. If resilience factors were studied in depth, new therapeutic avenues could probably be found that enhance these biological factors that are present but not sufficiently expressed in those who become ill'.

'Until now,' says Marra, 'we have defined patients on the basis of pathology and not their complexity and protective factors. We need to improve their 'profiling', to understand who has the highest risk of progressing (and how fast) to dementia. Only in this way will we be able to personalise treatment and identify potential 'responders' to therapies; this could narrow down the candidates for treatment with anti-amyloid antibodies to 25-50,000 people in Italy'.

In short, we still lack the operational tools to identify a person with Alzheimer's at an early stage, the speed of progression of their disease, and to predict whether or not they will respond to therapies that, including those that will come, are very expensive. This is why healthcare systems must equip themselves with tools to select people at high risk.

'The Interceptor project,' recalls Rossini, 'which ended recently and whose data we are currently processing, goes in this direction. And Italy will be the first country in the world to have an organisational model developed in 19 centres from Piedmont to Sicily. The aim of Interceptor is to develop a panel of biomarkers that are sustainable in terms of costs and health organisation, which - together with neuropsychological tests - will make it possible to diagnose Alzheimer's disease at an early stage.

Language analysis with Ai

There is also a European study on artificial intelligence and dementia (Ai-Mind), financed by the European Commission with around €14 million, in which Italy has no less than four operational units. And from this point of view, language analysis is proving to be much more effective than memory. We are talking about computerised linguistic analysis, which measures the number of words produced per minute, the number of errors in language production from the point of view of names, which studies the frequency of use and the typicality of the words produced.

'Some cognitive tests,' Marra points out, 'could help us predict who, among the many people suffering from a minimal cognitive disorder, will become ill in a year's time and who in 10 years' time, giving us an indication of the speed of progression. There are many studies under way in this field and we too are moving in this direction: one day it could allow us to diagnose a disease simply by having the patient tell us a story, and in this artificial intelligence will make a great contribution, and in the future it could be used as a means of screening, quickly and effectively, even over the phone'.