NIV iN ARDS

Dr Swapnil Pawar & Dr Jose Chacko

Noninvasive ventilation (NIV) is a well-established treatment for acute respiratory failure, especially in patients with hypercapnia and cardiogenic pulmonary edema.  Conversely, the use of NIV for hypoxemic respiratory failure, including acute respiratory distress syndrome (ARDS), is still controversial.

What’s the pathophysiology of ARDS?

ARDS is characterized by diffuse alveolar damage and pulmonary capillary endothelial injury. The alveolar-capillary barrier is formed by the microvascular endothelium and the epithelial lining of the alveoli. A variety of insults resulting in damage either to the vascular endothelium or to the alveolar epithelium could result in ARDS. Early ARDS results in an increase in the permeability of the alveolar-capillary barrier, leading fluid influx into the alveoli. 

There are two types of alveolar cells, damage to type I cells, leads to the increased entry of fluid into the alveoli and decreased clearance of fluid from the alveolar space. Damage to type II cells results in decreased production of surfactant with resultant decreased compliance and alveolar collapse. Normal repair processes in the lung may lead to the development of fibrosis. Cytokines including tumour necrosis factor, leukotrienes, macrophage inhibitory factor, along with platelet sequestration and activation, also play an important role in the development of ARDS. 

It has been well established that ARDS may be worsened by application positive airway pressure delivered to the lung by mechanical ventilation. This is termed ventilator-associated lung injury (VALI). ARDS is a non-homogeneous disease process. Relatively normal, more compliant alveoli may become overdistended while the diseased alveoli remain closed. The lungs may be subject to barotrauma, volutrauma, and atelectrauma during mechanical ventilation. 

Why do we prefer invasive ventilation over NIV in ARDS? 

There is an increasing concern that continued vigorous spontaneous respiratory efforts may be harmful in the presence of severe lung injury. The adverse impact of using high tidal volumes during invasive mechanical ventilation is well known. The swings in transpulmonary pressure (airway pressure – pleural pressure), representing lung stress, are comparable with spontaneous, controlled, or partially supported breathing when similar tidal volumes are delivered. Thus, high tidal volumes may be equally injurious with spontaneous or assisted breaths compared to controlled breaths. Patients with de novo acute hypoxemic respiratory failure (AHRF) may have a strong respiratory drive, leading to the generation of large spontaneous tidal volumes. 

Besides the adverse effects related to high transpulmonary pressures, the alveolar pressure may drop significantly lower than the end-expiratory pressure during spontaneous breathing. The intravascular pressure within the pulmonary blood vessels decreases proportionally; however, the pleural pressure decreases to a greater extent, thus, increasing the transmural pulmonary vascular pressure. The increase in the transmural pulmonary vascular pressure combined with increased capillary permeability results in leakage of fluid from within the capillaries and may lead to pulmonary edema. These changes may be similar to the pathophysiology of negative-pressure pulmonary edema, characterized by high airway resistance, leading to the precipitous drop in the airway and alveolar pressures. In patients with injured lungs, there may be regional variations in transpulmonary pressures. This may lead to movement of air from the non-dependent to dependent areas of the lung during the early inspiratory phase of spontaneous respiratory efforts (the pendulluft phenomenon). Overstretch injury to the dependent lung may occur due to this phenomenon. The adverse consequences arising from a continued high respiratory drive have been termed patient self-inflicted lung injury (P-SILI).   

Thanks, Dr Chacko. I guess, the potential advantages of NIV in the management of patients with ARDS are mainly related to the avoidance of complications linked to sedation, muscle paralysis, and ventilator-associated complications associated with endotracheal intubation and invasive mechanical ventilation (MV)

What’s the evidence out there in support of or against NIV in ARDS?

The use of non-invasive ventilation (NIV) as the initial modality of ventilatory support has increased several-fold over the years. However, improved outcomes with NIV use in patients with AHRF with no pre-existing cardiopulmonary disease (de novo acute respiratory failure) is less certain. 

From a physiological perspective, there appears to be a strong rationale for the application of NIV in AHRF. Oxygenation may improve with alveolar recruitment, and the work of breathing may improve, thereby ameliorating the subjective feeling of dyspnea. These positive effects may avoid the need for intubation and invasive ventilation, and thus, improve clinical outcomes.

There have been several observational studies, subgroup analysis of ARDS patients among patients with acute hypoxemic respiratory failure, and several meta-analyses with the use of NIV. The results from these studies have been largely inconclusive. 

Bellani et al. performed a sub-analysis of the LUNG-SAFE study, a large observational study that evaluated the management of patients with ARDS. NIV failure was strongly correlated with the severity of ARDS. Furthermore, on propensity-matched analysis, among patients with a Pa_O2/Fi_O2 ratio <150, ICU mortality was higher with NIV compared to invasive mechanical ventilation. The magnitude of the decrease in the PaO₂/FiO₂ ratio between days 1 and 2 was an independent predictor of mortality on multivariate analysis. In the FLORALI study that compared high-flow nasal oxygen with NIV or standard oxygen therapy through a non-rebreather face mask in patients with AHRF, 58% of patients failed NIV and required intubation by 28 days. 

Important to note: 1. NIV failure in AHRF carries poor outcomes, including mortality; 2. A combination of sedative and opioid analgesia may increase the rate of NIV failure; 3. An increase in tidal volumes with above a threshold tidal 9.5 ml/kg may be a marker of NIV failure. 

The use of NIV may result in a lower intubation rate in less severe disease;  however, it may lead to reduced survival in the more severe forms of ARDS (P/F < 150). The question remains, how do we correctly identify patients with ARDS who would succeed in an NIV trial? Not an easy question to answer. Besides, how long do you wait for possible stabilization or improvement from NIV? All this would lead us to ask the question, would it be more appropriate, and pragmatic to intubate and ventilate them early as part of a lung-protective strategy rather than continue to press on with NIV? 

That’s the million-dollar question. Avoidance of NIV in these patients is often justified by the association between a failed NIV attempt with a worse prognosis. In these studies, however, it is unclear whether NIV failure  was responsible for the worse prognosis (a causal association, likely by delaying  intubation) or if it was merely a marker of the underlying disease severity, such as  severe sepsis, higher SAPS-II score, and lower arterial partial pressure of oxygen to inspired fraction ratio (PF-ratio).

Despite the debatable recommendation, NIV is regularly used for hypoxemic respiratory failure. Specifically for ARDS  patients, a LUNG-SAFE study showed that NIV was used in 14.4% of patients (436 of 3,022), with 69% of them (300  of 436) being exclusively managed with NIV. These numbers highlight the importance of the topic, on which there is scarce literature.

In a recent study, Patel et al. brought to attention to the importance of the NIV interface in the outcome of ARDS patients. In this single-centre trial, interrupted early for efficacy, 83 ARDS patients requiring NIV by face mask for at least 8 hours were randomized to NIV by helmet or to continue with the face mask. In the helmet group, the intubation rate (the primary endpoint) was less than a third that in the face-mask group (18.2% vs 61.5%). The lower intubation rate was associated with more ventilator-free days and lower mortality.

Can we attribute all of this expressive difference in outcome solely to the interface?

When I look closely at this study, The most common reason for intubation was tachypnea and hypoxemia (83.3% for the face mask vs 37.5% for the helmet). Positive end-expiratory pressure (PEEP), a ventilatory setting used to avert hypoxemia, was set 3 cmH2O higher in the helmet group. Curiously,  pressure support levels were 3 cmH2O higher in the face-mask group leading to comparable total inspiratory pressures (~16 cm H2O) between groups. We know that high IPAP results in higher leak compared to higher PEEP. Also, there was a subjective interpretation of patient tolerance. 

Thus in Patel’s study, subjective decisions regarding patient comfort and excess leaks might have influenced the results.  More stringent criteria for the definition of patient tolerance and acceptable leaks would have been welcome.

Perhaps one of the most important contributions of Patel’s study was to emphasize that protective ventilation is as important during NIV as it is during invasive mechanical ventilation.  This includes the use of low tidal volumes, low plateau pressures, and high PEEP values. Unlike invasive ventilation, the use of  NIV in ARDS currently lacks ventilation protocols directed to avoid the mechanisms of ventilator-induced lung injury. 

In one study, tidal volume was a strong predictor of NIV failure, suggesting that close monitoring of tidal volume is important. In patients with persistently high tidal volumes, either you drop the pressure support or consider early invasive ventilation as a reasonable option to avoid ventilator-induced lung injury.

Another important factor in Patel’s study, besides higher PEEP levels in the helmet group,  driving pressure was also lower, settings again consistent with a lung-protective strategy. Amato et al., analyzing data from previously published trials of protective ventilatory strategies in  ARDS, recently demonstrated that driving pressure (plateau pressure minus PEEP) was the variable most closely related to survival.

A word of caution,  though, in actively breathing patients, driving pressure is usually underestimated because of the unmeasured force exerted by the respiratory muscles.

So in a nutshell, use high PEEP, Low driving pressure strategy and titrate your PS to achieve 4-6 ml/kg tidal volume. However, it’s not as simple as it sounds.

What’s your current practice?

We have a low threshold for invasive ventilation in patients with ARDS unless it is a relatively mild disease. If we try NIV, we re-evaluate every couple of hours and if there is no discernible improvement, we do not delay intubation. We tend to intubate early particularly in patients with pneumonia – particularly, viral pneumonia with bilateral involvement that usually takes several days to improve.