Heart Failure Diagnostic Performance of an Expirogram Analysis Algorithm Evaluating 4 Biomarkers (SenseIR-IC)

Evaluation of the Diagnostic Performance of an Algorithm for Analyzing Expirograms of 4 Biomarkers in Exhaled Air in Patients With Heart Failure

Telemonitoring is a key clinical issue in heart failure (HF). Bedside measurement systems using handheld devices provide "digital biomarkers" useful for remote monitoring. A recent systematic review and meta-analysis showed that teleconsultations and telemonitoring at home improved the prognosis of HF patients compared with usual care. Biomarkers contained in exhaled air could constitute "digital biomarkers" in HF, as measurement is non-invasive, and 4 different species have shown their potential interest: NO, CO, acetone and isoprene. The assessment of these species in the exhaled air to remains an issue in the perspective of non-invasive biomarkers in HF. Indeed, it requires selective sensors with low limit of detection. In addition, these sensors should be miniaturizable. Quartz-enhanced photoacoustic spectroscopy (QEPAS) are sensors that are suitable in this context. Last, the measured concentration should be informative and directly related to the HF. However, the concentration each of these biomarkers is not homogeneous during the expiration as it reflects the different lung compartments. While the end-expiratory concentration constitutes a sample of the alveolar concentration (AC) that reflects the blood concentration (BC) of one specie, the relationship between alveolar and blood concentrations is complex as exchanges that take place within these different compartments. Thus, measuring the concentration of a specie in exhaled air during a complete exhalation (or "expirogram") depends not only on the BC of the specie, but also on changes in lung function. Because both BC and changes in lung function depend on the severity of the HF, obtaining a full expirogram each specie should provide valid diagnosis information in HF. The mathematical modelization of real-time QEPAS sensors based expirograms together with lung function parameters (volume, flow) and lung compartment identification (capnography i.e. exhaled CO2 concentration) could provide valid algorithms with a an acceptable diagnosis performance in HF.

In the context of chronic diseases and heart failure (HF) in particular, telemedicine and telemonitoring have emerged as major clinical challenges today. The development of point-of-care measurement systems using hand-held devices provides "digital biomarkers" that are a key element in remote monitoring. Biomarkers contained in exhaled air could constitute "digital biomarkers", as the measurement of molecules in exhaled air is non-invasive. Currently, 6 exhaled biomarkers are validated by the US FDA and used in a clinical context. In heart failure, 4 different species have been shown to be of potential diagnostic or prognostic interest: NO, CO, acetone and isoprene. However, while the concentration of these species in alveolar air (CA) reflects their blood concentration (Cs), the relationship is more complex, having to take into account the different compartments of the bronchial tree and the exchanges that take place within these different compartments. Thus, measuring the concentration of a species in exhaled air during a complete exhalation (or "expirogram") using a real-time measurement, turns out to be dependent not only on the systemic concentration of the species, but also on changes in lung function.Thus, obtaining an expirogram not only makes it possible to specify the measurement of the endogenous source of the species, but also provides information on changes in pulmonary function, directly induced by heart failure, and which have a well-recognized prognostic value. → The combination of different candidate exhaled biomarkers in IC, during a real-time measurement of forced expiration, using selective, sensitive and miniaturizable sensors would provide diagnostic, prognostic and patient outcome information in heart failure. Quartz-enhanced photoacoustic spectroscopy (QEPAS) is a suitable method for remote monitoring of heart failure patients. It enables the creation of sensors characterized by good selectivity and low detection thresholds. What's more, real-time analysis is possible, and the sensors are potentially miniaturizable. These sensors are therefore capable of providing expirograms for different species (rather than simply measuring CA at the end of expiration). Complex signals of this type can be analyzed using mathematical modeling and artificial intelligence techniques such as "deep neural networks". These mathematical modelling methods have been used to model pulmonary, neurological or cardiac function parameters. As part of a translational research project in collaboration with Dr. A. Vicet (MCF- UM, Institut d'Electronique et des Systèmes) and Prof. N. Molinari (CHRU Montpellier, Institut Desbrest d'Épidémiologie et de Santé Publique), the research team is currently developing sensors for various exhaled biomarkers using the QEPAS method, which have been coupled with synchronous quantification of volumes, flows and lung compartments. These sensors are currently undergoing analytical validation (in the laboratory). The first expirograms have been obtained and modelled using a spline regression dimension reduction method.

The device will be tested to measure biomarkers in a group of 55 Heart Failure patients and 20 healthy subjects.

Stable or relapsing heart failure patients aged between 25 and 85, diagnosed by a cardiologist and requiring respiratory and cardiac investigations at the Montpellier University Hospital.

Recovery of heart failure data (left ventricular ejection fraction LVEF on cardiac echography (in mL), VO²max during, maximal cardio-pulmonary exercise test, plethysmographic spirometry, carbon monoxide transfer, NT-ProBNP assay,), clinical examination and finally expirogram with the device measuring the 4 exhaled biomarkers (CO/NO/Acetone/Isoprene).

Diagnostic performance (area under curve of ROC analysis) for the diagnosis of chronic heart failure of exhaled CO

Diagnostic performance (area under curve of ROC analysis) for the diagnosis of chronic heart failure of exhaled NO

Diagnostic performance (area under curve of ROC analysis) for the diagnosis of chronic heart failure of exhaled acetone

Diagnostic performance (area under curve of ROC analysis) for the diagnosis of chronic heart failure of exhaled isoprene

Inclusion Criteria: Age from 25 to 85, Stable or worsening heart failure, diagnosed by a cardiologist, Patients requiring respiratory function tests and cardiac explorations as part of a work-up at Montpellier University Hospital (pre-cardiac transplant, pre-rehabilitation, pre-surgery, etc....). Exclusion Criteria: Subject in a period of relative exclusion in relation to another protocol, or for whom the maximum annual indemnity of €6000 has been reached. Adult protected by law (guardianship, curatorship or safeguard of justice) Subject participating in another interventional research protocol Subject not affiliated to a social security scheme or not benefiting from such a scheme. Pregnant or breast-feeding woman, patient unable to give consent, protected adult, vulnerable persons (art.L. 1121-6, L. 1121-7, L. 1211-8, L. 1211-9 of the french legislation). Subject deprived of liberty by judicial or administrative decision Subject having refused to give consent Healthy Volunteers : Age from 25 to 85, Volunteers with no previous cardiorespiratory history or treatment.