A study published in the journal "Circulation: Arrhythmia and Electrophysiology" has found that progressive changes in the electrical signals recorded by implantable automatic defibrillators, sent periodically and remotely to the hospital, make it possible to dynamically estimate the risk of ventricular fibrillation, considered the most lethal cardiac arrhythmia.
The work, coordinated by the Centro Nacional de Investigaciones Cardiovasculares Carlos III (CNIC), has designed an analysis model of the heart’s electrical signal capable of detecting the periods in which people carrying an implanted defibrillator have a higher probability of suffering a severe arrhythmia.
As the CNIC explains, thousands of patients with heart failure carry an implantable automatic defibrillator. These devices identify and treat potentially fatal arrhythmias, sometimes by delivering an electric shock, in people at risk of sudden death, including those who suffer from heart failure. Although these devices are decisive for saving lives, their shocks can have adverse effects, which is why identifying periods of maximum arrhythmic vulnerability remains a pending clinical challenge.
The research proposes a shift in the current approach: rather than limiting itself to intervening once the arrhythmia has already been triggered, the defibrillator could help anticipate its onset. The scientists started from a clear premise: when heart failure progresses, intracardiac pressure increases, the walls of the heart become distended, and the way in which the electrical impulse propagates is altered.
This hypothesis was evaluated in experimental models of heart failure with an implantable defibrillator. The data showed that, as the disease progressed, the local electrical signal decreased in intensity and ventricular activation slowed down. At the end of the follow-up, it was observed that these changes were associated with a significantly higher risk of ventricular fibrillation compared with the control group without heart pathology.
With this information, the team trained a neural network, an artificial intelligence system oriented to pattern recognition, using records of cardiac electrical activity sent remotely from 5,089 patients treated in 50 hospitals in Spain. After that, the model was tested on an independent cohort made up of 2,040 patients from seven national centers carrying devices from another manufacturer.
The algorithm monitors the evolution of the electrical signal and premature beats (extrasystoles), and generates a risk score between 0 and 100 that is updated with each new remote transmission. In the patients who finally developed ventricular fibrillation, that score was increasing in the months prior to the episode, while in the rest it remained stable.
The tool offered a predictive capacity for arrhythmic events greater than that obtained through assessment based on classic clinical and demographic parameters, such as age or medical history. "Unlike other algorithms, our model is based on the pathophysiology of heart failure, so its predictions have a clinical meaning that the physician can interpret," David Filgueiras, a cardiologist at Hospital Clínico San Carlos de Madrid, head of the Advanced Development in Mechanisms and Arrhythmia Therapies group at CNIC, and coordinator of the study, has explained.
On the other hand, Jesús Diz, a researcher at CNIC and first signatory of the work together with José María Lillo, from the Fundación Interhospitalaria para la Investigación Cardiovascular (FIC), emphasizes that "the tool uses information that defibrillators already send to the hospital, so it could be incorporated into remote monitoring platforms without the need to implant another device, which could help bring forward check-ups or adjust treatment during periods of greatest risk".
Toward a more preventive and personalized cardiology
The authors consider that this finding drives a more preventive cardiology, adapted to each patient. In the coming years, cardiologists could receive automatic alerts when the electrical profile of a specific patient shows a relevant increase in the risk of serious arrhythmia, which would facilitate more individualized care and earlier therapeutic decisions.
As they emphasize, the importance of the work goes beyond technological development. Researchers have shown that the information needed to anticipate certain potentially fatal arrhythmias is already available in the millions of remote transmissions that implanted defibrillators generate every day. "The key is to interpret it correctly," they conclude.