New Research Challenges Noninvasive Support, Favors Protective Ventilation for ARDS: Luca Menga, MD

Preclinical data presented at the European Respiratory Society Congress 2025 explored how methods of assisted breathing may or may not contribute to lung injury. Luca Menga, MD, a critical care physician at University of Toronto, presented, “Effects of Spontaneous Breathing, Positive Airway Pressure, and pRoteCtive Mechanical Ventilation in a Preclinical Model of Acute Lung Injury. The SPARC Study,” and here he discusses findings to date from a model of acute respiratory distress syndrome (ARDS). Overall, these findings seems to suggest that protective mechanical ventilation may be superior to continuous positive air pressure and volume control ventilation.

This transcript was lightly edited for clarity; captions were autogenerated.

Transcript

Can you share an overview of the SPARC study and key findings from your late-breaking abstract?

In the late-breaking abstract that we presented, we compared the effects of 3 strategies, which are unassisted continuous breathing, which we called the ceiling; continuous positive air pressure, which we call CPAP; and volume control ventilation in a large animal model of acute respiratory distress syndrome, or acute lung injury. We decided to this study because all the patients with acute lung injury, at a certain point, they breathe spontaneously, most of them before getting intubated, and we wondered, can this be injurious or not? Because there are evidences in the literature showing that this may be injurious, but there are also strategies that can allow patients to breathe spontaneously in a safe way without needing and attracting intubation. For example, CPAP.

Why did your research target a muscular inspiratory pressure of 15-25 cm H₂O, and how did you manage potential variability in sedation?

We recently did a systematic review and meta-analysis that's going to be published in Critical Care soon, where we studied what are the inspiratory efforts of patients with ARDS breathing spontaneously, with and without interfaces such as NIV [noninvasive ventilation] or CPAP. What we found is that, on average, in the studies published until now, the muscular inspiratory pressure, it's around 15 cm of water, and we wanted to capture a model which would be representative of the most severe patients, of the patients who are the sickest ones that we can possibly treat. How to manage this? We changed the sedations and also that space to obtain a strong inspiratory effort. We didn't decrease the sedation to make the animal uncomfortable, of course, so we just stopped at a certain point with the sedation. If that wasn't enough to increase the inspiratory effort to where we wanted it, we would just increase that space, obtaining a strong effort while maintaining animal comfort.

Did you observe any specific histological or biomarker evidence of lung injury?

These are only preliminary data, of course. Histology Is something that we have looked at for 4 animals per group, and we haven't presented it in the abstract just because it wasn't ready yet. We found some very strong trends in favor of protective mechanical ventilation compared to the other groups, with the CPAP groups looking kind of right between completely unassisted and completely protective ventilation. More in detail, I will say that we found stronger markers of histological injury in the region around the pleura, which is the region where we think that the inspiratory effort is transmitted to, compared to the peribronchial area, which is the more conventionally affected area in cases of lung injury.

What additional evidence is needed to more definitively support your findings and their potential translation to clinical practice in ARDS management?

For sure, we will need to finish enrolling all the animals and see if this data confirms in the whole sample size, and also histology for the whole sample size. Then we also need to take with a grain of salt everything that's concerning any more research, because they may be as close as we can get to the actual patients' phenotypes, but they're not exactly the same. The model is obviously exaggerated in the sense that it's very short, while with patients we have a slower building up of symptoms, both in sense of lung injury, for example, but also in terms of contraindication for controlled ventilation. For example, in our model, you wouldn't see typically the ICU complications that you may see after 1 week in the ICU.