Heart attacks, strokes and more: thanks to AI, prevention will be ‘omnicomprehensive’ and tailored to the individual

Cholesterol, blood sugar, blood pressure, and a general metabolic check-up. If you like, you can also include a stress ECG. For a healthy person who wants to know their future risk of developing cardiovascular disease, a basic health check-up of this kind is a good place to start (to which the necessary tests can then be added, for example to assess kidney health). The results of the tests should then be interpreted by a doctor, who can offer useful advice on a case-by-case basis to get things back on track. But be careful. In the future, perhaps, health check-ups will be driven by multi-omics. They will harness the potential of Artificial Intelligence to combine a vast amount of information derived from a single blood sample, as well as from the individual’s genetic data and their social environment. In short, it will be possible to create a score that integrates the proteins involved in clotting processes, the inflammatory response and the imperceptible distress of cells. It will also be possible to gather insights into lipid metabolism, providing a comprehensive picture of a person’s actual risks well before any lesions develop. The personal risk score, the result of the combined analysis of hundreds of parameters using deep learning, promises to revolutionise the field of predictive medicine in the future. And the CardiOmicScore – as the test is called – promises to estimate, years before they actually occur, the actual risk of developing six different medical conditions, ranging from heart attack and stroke to heart failure, atrial fibrillation, peripheral arterial disease and venous thromboembolism. This conclusion is based on a study carried out by experts from the LKS Faculty of Medicine at the University of Hong Kong, coordinated by Zhang Quinpeng , and published in *Nature Communications*.

The analysis incorporates a vast amount of complex biological information, integrating data from proteomics (proteins), metabolomics (metabolites) and genomics to calculate individual-specific risk scores. The result? The system is able to analyse nearly 3,000 circulating proteins and 168 circulating metabolites, effectively creating a sort of ‘identity card’ for the invisible at a specific moment in each individual’s biological life. It is important to bear in mind that this also takes into account any genetic characteristics that remain stable throughout a person’s lifetime. Thanks to a specialised ‘deep learning’ system, a sample of data from the UK Biobank was analysed, revealing 2,920 circulating proteins and 168 metabolites measured in blood samples. All these invisible elements can provide a snapshot of a person’s biological state at a given moment, enabling us to go beyond knowledge based solely on an individual’s genetic predisposition.

Cardiovascular diseases, as we know, remain not only the leading cause of death but also of disability. “For too long, our health service has focused almost exclusively on those who are already ill and has never really concentrated on delaying the onset of cardiovascular diseases. – comments Massimo Grimaldi, president of ANMCO (National Association of Hospital Cardiologists) and head of the Cardiology Unit at the “F. Miulli” Hospital in Acquaviva delle Fonti -. “This can only be achieved if we become able to identify those at highest risk and take early action using all the tools at our disposal. Many advanced countries around the world are investing in precisely this area, with excellent results.” The study, referring to Hong Kong, highlights the emphasis that should be placed on primary prevention, because preventing the disease ensures the best possible outcome in terms of both life expectancy and quality of life. ““The research shows how the integration of artificial intelligence, new genetic insights and new analytical capabilities for microparticles such as metabolites and proteins is capable of identifying, much earlier and with high predictive value, those at risk of developing cardiovascular disease,” the expert explains. “Unfortunately, our ability to make opportunistic use of this new knowledge is very limited and cannot be applied to Italia, as we have a different genetic makeup and different environmental factors.”

“It is therefore absolutely essential to collect a large sample of genetic and haematochemical data in Italia, including proteins and metabolites that have been little studied to date, so that we can once again compete in the global race for a life that is not only longer but also healthier; and the privilege of being born in this country will be all the greater,” concludes Grimaldi . Currently in Italia, the assessment of cardiovascular risk is based primarily on a history of atherosclerotic disease and the SCORE 2 risk charts, which combine age with risk factors such as high blood pressure, non-HDL cholesterol, cigarette smoking and diabetes. Early detection of the disease could, however, benefit from simple tests such as a cardiology consultation, an electrocardiogram and the albumin-to-creatinine ratio in urine. The risk scores and the aforementioned tests should ideally form part of a population-based screening programme included within the minimum standards of care. In reality, however, they are very rarely carried out and are not included in effective screening programmes such as that for colorectal cancer. This is paradoxical because more people fall ill and die from cardiovascular diseases than from cancer.”