Evaluation of Physical Capacity and Energy Requirements in Patients With Hypoxemic Acute Respiratory Failure

Multimodal Evaluation of Physical Capacity and Energy Requirements in Patients With Hypoxemic Acute Respiratory Failure Under Non Invasive Ventilation Support

NIVAR01 (Multimodal evaluation of physical capacity and energy requirements in patients with hypoxemic acute respiratory failure and community acquired pneumonia under non invasive ventilation support) aims to comprehensively assess energy requirements and physical capacity in patients with acute hypoxemic respiratory failure (AHRF) due to community-acquired pneumonia (CAP) undergoing non-invasive ventilation (NIV).

Through a stepwise, multimodal approach integrating physiological measurements, imaging, and molecular biomarkers, the project seeks to identify patient subphenotypes that can guide targeted interventions, improve outcomes, and reduce the need for invasive mechanical ventilation. The project is structured in five progressive phases, each building on the findings of the previous one to refine and personalize care strategies.

The first phase of the project aims to assess the correlation between BIA (bioelectrical impedance analysis) and IC (indirect calorimetry) in sedated patients, in oder to extrapolate this correlation in patients under NIV in the next phase.

The second phase uses this correlation to validate the use of indirect calorimetry (IC) in patients under non-invasive ventilation (NIV) by comparing it with data obtained from patients under invasive mechanical ventilation, where IC is better established and technically more accurate. This comparative approach will serve to assess the reliability and feasibility of IC in NIV settings, incorporating correction factors for air leaks and comparing with BIA data, laying the groundwork for its integration with other physiological and biochemical measurements in subsequent phases.

The third phase expands to a multimodal prospective observational cohort study integrating IC, BIA, ultrasound, biomarkers of mitochondrial and endothelial dysfunction, and innate immune system paralysis. These data will be used for patient phenotyping through advanced machine learning. The fourth and fifth phases will develop and test personalized nutritional and motor interventions (e.g., phrenic nerve stimulation, tailored physiotherapy) in clinical trials based on identified patient subtypes.

Previous studies (Georges et al., Siirala et al., Singer 2024) have demonstrated the feasibility of using IC in NIV patients, although limited by methodological constraints. This project brings novelty by incorporating BIA-derived metabolic rate estimates and segmental analysis, offering new insights into body composition, fluid balance, and muscle integrity, including diaphragm function.

Additionally, biomarkers such as cf-mtDNA, FGF-21, GDF-15, HSP60/10, and cytokine profiles (TNFα, IL-6, IL-10, CRP), as well as endothelial markers (e-Selectin, sICAM-1, vCAM-1, Syndecan-1), will be analyzed. Immune paralysis will be studied via monocyte HLA-DR expression and LPS-stimulated cytokine release.

By leveraging multimodal data integration and gender-specific analysis, NIVAR 01 aims to optimize prediction of NIV failure beyond current tools such as the HACOR index and enable individualized patient management.

Non-invasive ventilation (NIV) is widely used in patients with acute hypoxemic respiratory failure (AHRF), particularly those with community-acquired pneumonia (CAP), as a strategy to avoid intubation and invasive mechanical ventilation. However, predicting NIV failure remains a significant clinical challenge, and the existing tools, such as the HACOR index, are limited in accuracy and do not take into account crucial aspects like energy expenditure, muscle integrity, or other patient's physiological parameters.

NIVAR01 addresses this gap through a comprehensive, stepwise research plan aimed at understanding and anticipating NIV failure by integrating metabolic, physiological, imaging, and biomolecular data. The ultimate goal is to define patient subphenotypes that enable the personalization of therapy and improve clinical outcomes.

Objectives

The main objective of NIVAR 01 is to develop and validate a multimodal strategy to predict NIV failure in patients with AHRF and CAP, enabling the design of personalized therapeutic interventions. Specifically, the project aims to:

• Validate the use of indirect calorimetry during NIV and its correlation with BIA.
• Identify physiological and molecular phenotypes associated with NIV failure.
• Develop and test targeted nutritional and functional interventions based on patient subtypes.

Planning

The project is structured in five sequential phases, each building upon the results of the previous one. The approach moves from methodological validation to biological characterization and finally to clinical intervention. It will integrate sex-specific analysis into predictive models and intervention strategies.

- Phase 1: Correlation between BIA and IC in sedated patients

The project begins by evaluating the correlation between metabolic rate data obtained from BIA and from indirect calorimetry in sedated, invasively ventilated ICU patients. Since IC is a robust validated method in this setting, it will serve as the reference method. The correlation obtained will validate the accuracy of BIA and support its use in future phases.

- Phase 2: Validation of IC in patients under NIV

Using the reference from Phase 1, IC will be adapted for use in patients under NIV. This includes addressing the challenges posed by air leaks through integration of correction factors provided by ventilator device and its monitoring software, and potentially the development of specific circuit adaptors if required. The metabolic data collected via IC will be compared with BIA estimates to confirm feasibility and reliability in the NIV setting.

- Phase 3: Multimodal cohort study and patient phenotyping

In a prospective observational study, a cohort of patients with AHRF and CAP under NIV will undergo multimodal evaluation. This includes IC, BIA (with segmental analysis of the different body compartments both at the anatomic and physiological level), muscle and diaphragm ultrasound, and blood sampling for biomarkers of mitochondrial dysfunction (cf-mtDNA, FGF-21, GDF-15, HSP60/10), endothelial injury (e-Selectin, sICAM-1, vCAM-1, Syndecan-1), oxidative stress, and immune paralysis (HLA-DR expression, TNFα response to LPS). Data will be analyzed using machine learning to identify patient subphenotypes.

- Phase 4: Design of personalized interventions

Based on the phenotypes identified, targeted interventions will be developed. These include nutritional strategies informed by metabolic requirements and muscle function assessments, as well as physiotherapy protocols that may involve early mobilization and respiratory muscle stimulation (e.g., phrenic nerve). Strategies will be adapted to sex-specific physiological patterns and the patient's compartmental body composition.

- Phase 5: Randomized controlled trials

In the final phase, the proposed interventions will be tested in randomized clinical trials (RCTs). Patients will be stratified by phenotype and randomized to receive standard care or personalized therapy. Primary outcomes will include NIV failure (need for intubation), ICU/hospital length of stay, mortality, and functional recovery.

Expected Outcomes

NIVAR 01 is expected to generate:

• A validated protocol for using IC during NIV, including leak correction methods.
• Evidence supporting the use of BIA as a practical bedside tool to estimate metabolic rate and assess muscle and fluid status.
• Identification of subphenotypes among AHRF patients under NIV, based on multimodal data integration.
• Personalized intervention protocols tailored to patient profiles, with demonstrated clinical benefit in RCTs.
• New predictive models that outperform traditional tools like the HACOR index.
• Practical contributions to gender-informed critical care.

This integrated model will help clinicians anticipate NIV failure earlier, adapt treatment proactively, and improve patient outcomes through individualized care strategies.

Conclusion

NIVAR01 aims to develop a new standard in the management of patients with acute respiratory failure under non-invasive ventilation. By combining validated physiological measurements, molecular biomarkers, and clinical decision tools, the project introduces a precision medicine approach to intensive care. Through systematic validation, rigorous phenotyping, and targeted interventions, NIVAR 01 aims to reduce NIV failure, shorten ICU stays, and improve survival and recovery in this patient population.

Inclusion Criteria:

• Moderate-to-severe Acute Hypoxemic Respiratory Failure (PaO2 /FiO2 < 200mmHg) due to respiratory infection who underwent NIV support for at least 24h.

Exclusion Criteria:

• Presence of electrical implants, wounds or skin damage at the designated electrode sites or the inability to perform BIA.
• Patients who met criteria for immediate intubation upon ICU admission were also excluded.
• Other circumstances or clinical conditions at the discretion of the investigator that precluded the participation of the patient in the study were also considered (i.e., limitation of therapeutic effort and poor prognosis).