Comments
Abstract
Rationale: High-flow nasal cannula (HFNC) improves the clinical outcomes of nonintubated patients with acute hypoxemic respiratory failure (AHRF).
Objectives: To assess the effects of HFNC on gas exchange, inspiratory effort, minute ventilation, end-expiratory lung volume, dynamic compliance, and ventilation homogeneity in patients with AHRF.
Methods: This was a prospective randomized crossover study in nonintubated patients with AHRF with PaO2/setFiO2 less than or equal to 300 mm Hg admitted to the intensive care unit. We randomly applied HFNC set at 40 L/min compared with a standard nonocclusive facial mask at the same clinically set FiO2 (20 min/step).
Measurements and Main Results: Toward the end of each phase, we measured arterial blood gases, inspiratory effort, and work of breathing by esophageal pressure swings (ΔPes) and pressure time product, and we estimated changes in lung volumes and ventilation homogeneity by electrical impedance tomography. We enrolled 15 patients aged 60 ± 14 years old with PaO2/setFiO2 130 ± 35 mm Hg. Seven (47%) had bilateral lung infiltrates. Compared with the facial mask, HFNC significantly improved oxygenation (P < 0.001) and lowered respiratory rate (P < 0.01), ΔPes (P < 0.01), and pressure time product (P < 0.001). During HFNC, minute ventilation was reduced (P < 0.001) at constant arterial CO2 tension and pH (P = 0.27 and P = 0.23, respectively); end-expiratory lung volume increased (P < 0.001), and tidal volume did not change (P = 0.44); the ratio of tidal volume to ΔPes (an estimate of dynamic lung compliance) increased (P < 0.05); finally, ventilation distribution was more homogeneous (P < 0.01).
Conclusions: In patients with AHRF, HFNC exerts multiple physiologic effects including less inspiratory effort and improved lung volume and compliance. These benefits might underlie the clinical efficacy of HFNC.
High-flow nasal cannula (HFNC) is a noninvasive form of respiratory support that can reduce reintubation rates and mortality of patients with acute hypoxemic respiratory failure. The physiologic effects potentially underlying these clinical benefits are still largely undefined.
In patients with acute hypoxemic respiratory failure, HFNC improves oxygenation; reduces the patient’s effort; reduces the minute ventilation needed to obtain a physiologic arterial CO2 level; increases the end-expiratory lung volume; and improves dynamic compliance, transpulmonary pressure, and ventilation homogeneity. These beneficial effects might underlie the clinical efficacy of HFNC.
High-flow nasal cannula (HFNC) is a noninvasive respiratory support designed to deliver 30–60 L/min of a heated, humidified mixture of air and oxygen through specifically designed nasal prongs (1). HFNC was first used in preterm infants and pediatric patients (2) and recent large randomized clinical trials have promoted its use in adults with acute hypoxemic respiratory failure (AHRF) (3). These trials demonstrated the potential of HFNC to improve clinical outcomes, such as reintubation rates and 90-day mortality (4–7). However, the physiologic mechanisms underlying the clinical benefits of HFNC are still poorly understood in adult patients with AHRF.
Basic clinical monitoring during HFNC therapy demonstrated rapid improvement of oxygenation and reduction of dyspnea in comparison with a standard facial mask (8, 9). Findings in other populations (10) indicate that advanced respiratory monitoring might demonstrate other specific physiologic effects. A study in pediatric patients with acute bronchiolitis treated with HFNC showed reductions in inspiratory effort, as measured by esophageal pressure swings (11), whereas in another study in patients with chronic obstructive pulmonary disease electrical impedance tomography (EIT) (12) suggested an increase in end-expiratory lung volume. Reduced inspiratory effort and improved lung volume might be particularly important in AHRF, because they may prevent respiratory muscle exhaustion and lower the Vt/end-expiratory lung volume ratio, which are closely related to the need for intubation and the severity in these patients (13, 14).
The present study describes the effects of HFNC on specific advanced physiologic parameters that may be correlated with clinical outcomes of patients with AHRF. Complementing standard clinical monitoring with the use of esophageal pressure and EIT, we assessed the effects of HFNC on gas exchange, patients’ inspiratory effort, minute ventilation (MV), lung volume, dynamic compliance, transpulmonary pressure, and ventilation homogeneity. Our hypothesis was that HFNC significantly improves these key physiologic parameters.
Some of the data reported here have already been presented in the form of an abstract (15).
We enrolled 15 patients with AHRF admitted to the general intensive care unit of San Gerardo Hospital, Monza, Italy. Inclusion criteria were new or acutely worsening respiratory symptoms following a known clinical insult lasting less than a week, and PaO2/setFiO2 less than or equal to 300 mm Hg while receiving additional oxygen by a standard facial mask, as per clinical decision. Exclusion criteria are listed in the online supplement. The ethical committee of San Gerardo Hospital approved the study (reference number: 432_2015bis) and informed consent requirements were met according to local regulations.
At enrollment, we collected the patients’ main demographics and clinical data. An esophageal balloon catheter was placed in the esophagus, as demonstrated by the appearance of cardiac artifacts and appropriate negative swings of pressure tracings during inspiration (16, 17). Esophageal pressure waveforms were continuously recorded by a dedicated data acquisition system throughout the study. An EIT-dedicated belt was placed around each patient’s chest and connected to a commercial EIT monitor. During the whole study, EIT data were registered at 20 Hz and stored for offline analysis by dedicated software (18).
Each patient was entered in the two study phases with the same set FiO2 for 20 minutes in computer-generated random order: standard nonocclusive oxygen facial mask with gas flow set at 12 L/min, and HFNC with gas flow 40 L/min.
Set FiO2 during both phases was selected clinically by the attending physician before enrollment to achieve peripheral saturation between 90 and 95% on pulse oximetry during standard oxygen facial mask breathing. Set FiO2 during both study phases was measured by a dedicated system (AIRVO 2; Fisher and Paykel Healthcare, Auckland, New Zealand) connected to the standard facial mask or the nasal cannula. This system can deliver airflows between 2 and 60 L/min with FiO2 between 0.21 and 1.0 by connection to a wall supply. FiO2 is continuously measured at the gas outlet of the system. In summary, in keeping with previous studies (12, 19, 20), we set the gas delivery system at 12 L/min during the facial mask phase and at 40 L/min during HFNC with the same measured set FiO2, obtained by modifying the additional oxygen wall supply. However, we could not avoid or verify lower tracheal and alveolar FiO2 during the oxygen facial mask phase because of entrainment of room air.
At the end of each phase, we collected arterial blood gas analysis data, respiratory rates (RR), and hemodynamics. From the esophageal pressure waveforms recorded during the last 3–5 minutes of each phase we measured the following (21):
| 1. | The average pressure time product over a minute (PTPmin), as a measure of the metabolic work of breathing per minute (see online supplement for detailed methods); | ||||
| 2. | The per-breath PTP, as a measure of the metabolic work of breathing per single breath (see online supplement for detailed methods); | ||||
| 3. | The esophageal pressure swings during inspiration (ΔPes) as a measurement of the patient’s inspiratory effort; | ||||
| 4. | The dynamic end-expiratory transpulmonary pressure (PL,ee), calculated as the difference between airway pressure (assumed to be 0 cm H2O with the facial mask and 2.5 cm H2O [22] during HFNC) and the absolute Pes measured at the end of expiration (zero flow); | ||||
| 5. | The dynamic end-inspiratory transpulmonary pressure (PL,ei), calculated as the difference between airway pressure (assumed to be 0 cm H2O with the facial mask and 2.5 cm H2O during HFNC) and the absolute Pes measured at the end of inspiration (zero flow); | ||||
| 6. | The driving transpulmonary pressure (ΔPL), calculated as (PL,ei−PL,ee). | ||||
During the last minutes of each phase, we measured the following EIT parameters:
| 1. | The average global Vt and those distending nondependent and dependent lung regions (Vtglob, Vtnon-dep, and Vtdep, respectively). | ||||
| 2. | The MV. | ||||
| 3. | Corrected MV (MVcorr), defined as MV multiplied by the ratio of the patient’s PaCO2 to 40 mm Hg (23) (with lower values indicating improved CO2 clearance, reduced CO2 production, or both);. | ||||
| 4. | Global and regional changes in end-expiratory lung impedance (corresponding to end-expiratory lung volume) during the HFNC phase (ΔEELIglob, ΔEELInon-dep, and ΔEELIdep, respectively) (18). | ||||
| 5. | Global inhomogeneity index, a measure of inhomogeneous distribution of tidal ventilation (24). The global inhomogeneity index gave a reliable and interpatient comparable synthetic assessment of inhomogeneous distribution of the Vt in the lungs (24). The risk of additional lung injury seems to increase linearly with inhomogeneous distribution of lung densities (25), which is correlated with ventilation inhomogeneity (26). Thus, any change in this index might be clinically significant. However, to our knowledge, no prospective clinical validation of specific thresholds has been done yet. | ||||
| 6. | The global and regional peak inspiratory and expiratory airflows (PIFglob, PIFnon-dep, and PIFdep; PEFglob, PEFnon-dep, and PEFdep, respectively) (27). | ||||
| 7. | From the EIT-derived airflow tracings, we also measured the inspiratory (Ti) and expiratory (Te) times, and the total cycle time (Ttot) (28). | ||||
Please note that all the lung volumes (e.g., Vt) mentioned above were not measured but rather we assumed that the changes in chest electrical impedance reflect changes in lung volumes. We chose the sample size on the basis of previous studies (9–14, 16–20, 24). For the sake of clarity, EIT measures (e.g., Vtglob) during HFNC were transformed from arbitrary units of impedance change to the percentage change from their baseline values during the oxygen facial mask phase. Normally distributed variables are expressed as mean ± SD and were analyzed by a paired Student’s t test. Nonnormally distributed variables are expressed as medians (interquartile range) and were compared by Wilcoxon signed rank test. Correlations were analyzed by Pearson coefficient. A level of P less than 0.05 (two-tailed) was considered statistically significant.
Additional details on the method for this study are provided in the online supplement.
Patients’ main characteristics are reported in Table 1. Patients were 60 ± 14 years old and six (40%) were women. At enrollment, all patients had PaO2/setFiO2 less than 200 mm Hg, with three (20%) less than 100 mm Hg. Seven patients (47%) had bilateral infiltrates on chest radiograph.
| Patient | Sex | Age (yr) | SAPS II at ICU Admission | Number of Organs Dysfunction | Etiology of Acute Respiratory Failure | PaO2/setFiO2 (mm Hg) | Bilateral Infiltrates on Chest Radiograph |
|---|---|---|---|---|---|---|---|
| 1 | M | 60 | 40 | 2 | Primary, infectious | 119 | No |
| 2 | M | 55 | 36 | 2 | Primary, infectious | 134 | Yes |
| 3 | M | 53 | 34 | 3 | Primary, infectious | 193 | No |
| 4 | F | 43 | 26 | 1 | Primary, infectious | 97 | No |
| 5 | F | 66 | 56 | 2 | Primary, infectious | 121 | Yes |
| 6 | M | 68 | 43 | 1 | Primary, infectious | 107 | Yes |
| 7 | M | 47 | 33 | 2 | Extrapulmonary, noninfectious | 114 | No |
| 8 | F | 56 | 26 | 1 | Primary, infectious | 117 | No |
| 9 | F | 47 | 42 | 1 | Primary, infectious | 158 | Yes |
| 10 | F | 78 | 43 | 1 | Primary, infectious | 146 | No |
| 11 | M | 70 | 44 | 2 | Extrapulmonary, noninfectious | 171 | No |
| 12 | M | 49 | 51 | 1 | Primary, infectious | 68 | Yes |
| 13 | M | 95 | 26 | 1 | Primary, infectious | 83 | Yes |
| 14 | M | 47 | 35 | 1 | Primary, infectious | 144 | Yes |
| 15 | F | 74 | 48 | 1 | Primary, infectious | 180 | No |
| Total or mean ± SD | 6 F/9 M | 60 ± 14 | 38 ± 9 | 1 ± 2 | 13 primary/2 extrapulmonary; 13 infectious/2 noninfectious | 130 ± 35 | 7 yes/8 no |
During HFNC, ΔPes were significantly lower than with the standard nonocclusive oxygen facial mask (P < 0.01) (Table 2, Figure 1A), indicating that patients had less inspiratory effort. Interestingly, the Vt/ΔPes ratio (i.e., an estimate of the dynamic lung compliance) was significantly higher during HFNC (P < 0.05), possibly indicating external “ventilation support” by the mandatory flow of HFNC during inspiration, improved lung mechanics, or both. PTP and PTPmin were both significantly lower with HFNC (P < 0.05 and P < 0.001), suggesting lighter metabolic work of breathing per breath and per minute (Table 2, Figure 1B).
| Variable | Oxygen Facial Mask | HFNC | P Value* |
|---|---|---|---|
| ΔPes, cm H2O | 9.9 ± 4.2 | 8.0 ± 3.4 | <0.01 |
| PTP, cm H2O × s | 9.5 (5.7 to 12.1) | 7.4 (4.1 to 9.4) | <0.01 |
| PTPmin, cm H2O × s/min | 216.3 ± 100.5 | 154.8 ± 84.8 | <0.001 |
| PL,ee, cm H2O | −10.1 ± 5.0 | −7.5 ± 5.2 | <0.001 |
| PL,ei, cm H2O | −3.6 ± 4.9 | −2.6 ± 4.5 | 0.16 |
| ΔPL, cm H2O | 5.7 ± 3.4 | 4.3 ± 2.9 | 0.08 |
| RR, bpm | 24 (20 to 27) | 22 (17 to 24) | <0.01 |
| Vt (change from facial mask), % | — | −5 ± 32 | 0.44 |
| Vtnon-dep (change from facial mask), % | — | 3 ± 49 | 0.59 |
| Vtdep (change from facial mask), % | — | −5 ± 33 | 0.54 |
| Minute ventilation (change from facial mask), % | — | −19 ± 16 | <0.001 |
| Corrected minute ventilation (change from facial mask), % | — | −18 ± 15 | <0.001 |
| Set FiO2 | 0.60 (0.50 to 0.75) | 0.60 (0.50 to 0.75) | 1.00 |
| PaO2, mm Hg | 72 (68 to 75) | 98 (78 to 131) | <0.001 |
| PaO2/setFiO2, mm Hg | 130 ± 35 | 184 ± 53 | <0.001 |
| PaCO2, mm Hg | 40.7 ± 5.7 | 41.1 ± 5.9 | 0.27 |
| pH | 7.45 ± 0.02 | 7.44 ± 0.03 | 0.23 |
| SBP, mm Hg | 141 ± 25 | 137 ± 27 | <0.05 |
| MAP, mm Hg | 90 ± 15 | 88 ± 16 | 0.11 |
| CVP, mm Hg | 4.6 ± 5.2 | 5.8 ± 4.7 | <0.05 |
| HR, bpm | 85 ± 9 | 84 ± 9 | 0.44 |
Figure 1. Reduction of inspiratory effort (A) and metabolic work of breathing (B) during treatment with high-flow nasal cannula (HFNC). In patients with acute hypoxemic respiratory failure, HFNC delivered at 40 L/min reduced the negative swings of esophageal pressure (a measure of patients’ inspiratory effort) and the esophageal pressure-time product (a measure of patients’ per breath metabolic work of breathing) in comparison with a standard nonocclusive facial mask at 12 L/min and the same set FiO2. Horizontal bars represent mean values. ΔPes = inspiratory esophageal pressure swing; PTP = pressure-time product.
[More] [Minimize]MV was significantly lower during HFNC (P < 0.001) than with the standard nonocclusive oxygen facial mask (Table 2). This was caused by changes in RR or Vt, which were inversely related (ΔRR × ΔVt: R2 = 0.74, P < 0.001). On average, RR decreased (P < 0.01), whereas Vt did not differ in the two phases, at either the global or regional level. Despite the decrease in MV, HFNC significantly improved oxygenation (P < 0.001) (Table 2, Figure 2A), whereas PaCO2 and pH did not change. MVcorr dropped significantly during HFNC (P < 0.001) (Table 2, Figure 2B).
Figure 2. High-flow nasal cannula (HFNC) improves oxygenation (A) and reduces minute ventilation (B) without affecting arterial CO2 tension. In comparison with the standard nonocclusive facial mask, in patients with acute hypoxemic respiration, HFNC increased PaO2 with no change in set FiO2. During HFNC, the minute ventilation (MV) needed to obtain normal arterial CO2 tension (i.e., corrected MV = MV * PaCO2/40 mm Hg) decreased, as if CO2 removal was more effective, CO2 production was reduced, or both. Horizontal bars represent mean values. MVcorr = corrected minute ventilation.
[More] [Minimize]There was a significant correlation between the reductions in PTP and the changes of MVcorr during HFNC (ΔPTP × ΔMVcorr: R2 = 0.46, P < 0.01), possibly indicating that enhanced CO2 clearance by washout of the upper airways reduced the inspiratory work of breathing, that HFNC lowered CO2 production, reducing the ventilation needs, or both. The relative reduction of PTPmin passing from the oxygen facial mask to HFNC was significantly larger than the reduction of MVcorr (32 ± 12% vs. 18 ± 15%; P < 0.05). The relative reductions of PTPmin and RR passing from the mask to HFNC were not correlated (P = 0.147, data not shown).
Lung volume, as measured by ΔEELI, significantly increased during HFNC, globally and in the dependent and nondependent lung regions (P ≤ 0.01 for all) (Table 3, Figure 3A). The increase in global gas content in the lungs was 51 ± 57% of the baseline Vt. This suggests the generation of positive end-expiratory pressure by HFNC that might have improved oxygenation and, in the presence of unchanged Vt, reduced regional lung strain. Similarly, PL,ee and PL,ei increased during HFNC (P < 0.05) (Table 2) and became less negative, possibly indicating a lower tendency to alveolar collapse (21). Driving transpulmonary pressure fell during HFNC, although not significantly (P = 0.08) (Table 2).
| Variable | Oxygen Facial Mask | High-Flow Nasal Cannula | P Value* |
|---|---|---|---|
| ΔEELIglob (change from facial mask), % of baseline Vt | — | 51 ± 57 | <0.001 |
| ΔEELInon-dep (change from facial mask), % of baseline Vt | — | 29 ± 36 | ≤0.001 |
| ΔEELIdep (change from facial mask), % of baseline Vt | — | 26 ± 33 | ≤0.01 |
| GI index | 0.50 (0.49 to 0.57) | 0.47 (0.43 to 0.60) | <0.01 |
| PIFglob (change from facial mask), % | — | −15 ± 23 | 0.07 |
| PEFglob (change from facial mask), % | — | −27 ± 22 | ≤0.001 |
| PIFnon-dep (change from facial mask), % | — | −11 ± 29 | 0.29 |
| PIFdep (change from facial mask), % | — | −20 ± 19 | <0.01 |
| PEFnon-dep (change from facial mask), % | — | −19 ± 32 | 0.07 |
| PEFdep (change from facial mask), % | — | −34 ± 18 | <0.001 |
| Ti, s | 1.2 ± 0.2 | 1.2 ± 0.3 | 0.84 |
| Te, s | 1.3 ± 0.2 | 1.5 ± 0.6 | <0.05 |
| Ti/Ttot | 0.5 ± 0.0 | 0.4 ± 0.0 | <0.05 |
Figure 3. High-flow nasal cannula (HFNC) increases lung volume (A) and raises the ratio of Vt to inspiratory effort (B). In comparison with a standard nonocclusive low-flow facial mask, in patients with acute hypoxemic respiratory failure HFNC increased the lung volume at end-expiration as measured by electrical impedance tomography (see text and the online supplement for details), suggesting a positive end-expiratory pressure effect. HFNC also gave a higher ratio of Vt to negative inspiratory pressure during inspiration (ΔPes), suggesting improvement in the mechanical properties of the respiratory system, external “ventilation support,” or both. Average values were significantly different even though in five patients Vt/ΔPes actually decreased. Horizontal bars represent mean values. ΔEELI = change of end-expiratory lung impedance.
[More] [Minimize]The global inhomogeneity ventilation index fell slightly but significantly during HFNC (P < 0.01) (Table 3), indicating more homogeneous distribution of ventilation throughout the lungs, which might correspond to better distribution of lung densities (26).
Patients’ PEF decreased significantly overall, by the reduction of PEF from the dependent lung regions (Table 3) and this might be regarded as an indirect sign of improvement of lung compliance in this region. PIF was reduced during HFNC, although not significantly (P = 0.07) and this too might have contributed to improving oxygenation by giving higher alveolar FiO2. Finally, the Ti/Ttot ratio was lower during HFNC (P < 0.05) (Table 3), which in presence of lower inspiratory effort and unchanged maximal inspiratory pressure suggests a lower tension-time index of the inspiratory muscles (28).
There was a significant correlation between ΔPes and PTP during HFNC and patients’ baseline PaCO2 (R2 = 0.433, P < 0.01 and R2 = 0.275, P < 0.05, respectively) (Figure 4; see Figure E1 in the online supplement). Baseline PaO2 was not correlated with lowering of either of these variables (P = 0.42 and P = 0.35, data not shown).
Figure 4. Baseline arterial CO2 tension is correlated with changes in inspiratory effort with the high-flow nasal cannula (HFNC). PaCO2 measured during the baseline standard oxygen facial mask phase (x-axis) was correlated with the decrease in patient’s effort induced by the switch to HFNC. Thus, HFNC might be particularly effective in unloading the respiratory muscles of patients with acute hypoxemic respiratory failure with more impaired CO2 clearance (i.e., those with higher dead space fraction). The solid line represents the regression line. ΔPes = inspiratory esophageal pressure swing.
[More] [Minimize]The present study shows that in patients with AHRF, HFNC improves several key physiologic parameters including oxygenation, inspiratory effort, MV, RR and lung volume, dynamic lung compliance, transpulmonary pressure, and homogeneity.
Esophageal pressure swings and PTP are validated, commonly used measures of patients’ inspiratory effort and metabolic work of breathing, respectively (21). Previous studies described the reduction of work of breathing by HFNC in pediatric populations (11, 29). We found this was also true in adult patients with AHRF, because measures of inspiratory effort and metabolic work of breathing significantly decreased during HFNC therapy. Our data, the physiologic background, and previous publications suggest that many factors contribute to reducing respiratory workload during HFNC. We observed better oxygenation, which reduced the patients’ hypoxic drive. Higher arterial oxygenation during HFNC might have simply mirrored higher alveolar FiO2: set FiO2 might in fact have been significantly higher than tracheal and alveolar FiO2 with the standard facial mask, because of entrainment of room air, whereas the higher external flow coupled with lower inspiratory airflow during HFNC might have permitted minimal differences between set and alveolar FiO2 (8). Improved lung volume, indicating positive end-expiratory pressure, might have contributed to the increase in the PaO2/setFiO2 ratio and the lower hypoxic drive too (30). During HFNC, the MV needed to obtain normal arterial CO2 tension was lower than with the facial mask and this might have followed enhanced CO2 clearance by washout of upper airways (1, 8), leading to lower ventilation needs and work of breathing. However, lower CO2 production from the respiratory muscles (8), linked to the decrease in MV, may also have helped reduce the inspiratory effort and ventilation needs. The possible improvement in dynamic lung compliance might have established more favorable working conditions during inspiration and unloading of the inspiratory muscles.
Among all these factors, the relation between the reduction in MVcorr and changes in PTP suggests that better CO2 clearance might have had a key role in reducing patients’ work of breathing. However, this decrease was significantly larger than the reduction in MVcorr, possibly indicating that multiple mechanisms work to reduce the workload. As a sign to clinicians, in this study the reduced inspiratory effort during HFNC positively correlated with higher baseline arterial CO2 tension, suggesting HFNC was more effective in the presence of more severely impaired CO2 clearance (e.g., in patients with AHRF with higher dead space fraction) (31). The reduction of RR by HFNC was not related with that of PTPmin. Previous findings linked this reduction of the RR to clinical success (32) but our result suggests that mechanisms besides reduced work of breathing underlie the clinical benefits of HFNC. The causal link between acute reductions of respiratory workload, long-term prevention of respiratory decompensation, reduced need of intubation, and improved survival seems reasonable but has yet to be proved.
Another part of our findings suggests new hypotheses on the effects of HFNC on the physiologic determinants of ventilation-induced lung injury. We noted a decrease in driving transpulmonary pressure swings along inspiration. Additionally, absolute estimates of PL,ei and PL,ee were higher during the HFNC phase (as expected from the additional positive end-expiratory pressure effect), possibly resulting in a smaller tendency to lung collapse. As higher driving transpulmonary pressure and derecruitment (18, 21) potentially aggravate lung injury, HFNC might hypothetically lower this risk. Increased end-expiratory lung volume induced by HFNC with unchanged Vt might have reduced lung strain, which seems linearly correlated with the severity of ventilation-induced lung injury (14). Finally, the slight but significant decrease in the inhomogeneity of ventilation distribution indirectly suggests there may be fewer or smaller regional areas of alveolar collapse (26), potentially reducing the risk of focal multiplication of the alveolar wall tension and additional injury (25). In summary, our findings suggest that HFNC might affect key determinants of ventilation-induced lung injury, such as lung stress, strain, and inhomogeneity. There is lively debate on how to minimize the detrimental effects of spontaneous breathing on preexisting lung injury (33–35) and the hypotheses on the potential role of HFNC in improving physiologic determinants of ventilation-induced lung injury may well merit further scrutiny.
Our study has several limitations. First, EIT imaging covers only part of the lungs (approximately 50%); it cannot detect an increase of lung volume along the vertical axis unless an abdominal belt is used, and most of the validation studies of EIT compared with other techniques were conducted in different settings (i.e., intubated patients or animals). However, previous studies have showed linear correlations between EIT measurements of lung volume changes and those obtained by other validated methods (36, 37), even in cases of significant changes in intrathoracic pressure (38). Moreover, the randomized crossover design of this study (with each patient compared with herself or himself in the different phases) might have made the comparison of EIT measures more accurate. Second, the study phases were short; we aimed for the shortest time needed to equilibrate lung volumes and gas exchange given the difficulties of stable and reliable advanced respiratory monitoring in awake patients with AHRF. Third, we assessed only objective rather than subjective measures of dyspnea and did not investigate patients’ comfort; however, subjective indexes have already been extensively reported (1, 8).
Fourth, we assessed PTP by analyzing only the esophageal pressure tracings rather than the traditional PTPes that includes the chest wall static recoil pressure-time curve (39) because we lacked direct measure of chest wall compliance and absolute Vt. However, in the two study phases, chest wall compliance was most probably unchanged and the Vt measured by EIT did not vary, so changes in PTP should have been linearly correlated with changes in PTPes. Moreover, we calculated PTPes by using standard formula for chest wall compliance (21) and by calculation of absolute Vt through an arbitrary milliliter/impedance unit conversion factor derived from the lower positive end-expiratory pressure phase of a previous study (38) and we found tight correlation with PTP (see the additional results section in the online supplement). Fifth, the population was small and this might have explained the lack of significance in some of the correlations. Finally, we measured only set FiO2 and not that delivered to the lower airways, which might have differed, especially during the facial mask phase because of entrainment of room air. Thus, we cannot know how much the difference between set and delivered FiO2 could have influenced the study findings.
HFNC exerts various specific physiologic effects in patients with AHRF including improved gas exchange, lower RR and effort, improved lung volume, dynamic compliance, transpulmonary pressures, and homogeneity. All these physiologic benefits might positively affect the clinical outcome of patients with AHRF.
| 1. | Papazian L, Corley A, Hess D, Fraser JF, Frat JP, Guitton C, Jaber S, Maggiore SM, Nava S, Rello J, et al. Use of high-flow nasal cannula oxygenation in ICU adults: a narrative review. Intensive Care Med 2016;42:1336–1349. |
| 2. | Wilkinson D, Andersen C, O’Donnell CP, De Paoli AG, Manley BJ. High flow nasal cannula for respiratory support in preterm infants. Cochrane Database Syst Rev 2016;2:CD006405. |
| 3. | Nishimura M. High-flow nasal cannula oxygen therapy in adults: physiological benefits, indication, clinical benefits, and adverse effects. Respir Care 2016;61:529–541. |
| 4. | Maggiore SM, Idone FA, Vaschetto R, Festa R, Cataldo A, Antonicelli F, Montini L, De Gaetano A, Navalesi P, Antonelli M. Nasal high-flow versus Venturi mask oxygen therapy after extubation: effects on oxygenation, comfort, and clinical outcome. Am J Respir Crit Care Med 2014;190:282–288. |
| 5. | Frat JP, Thille AW, Mercat A, Girault C, Ragot S, Perbet S, Prat G, Boulain T, Morawiec E, Cottereau A, et al.; FLORALI Study Group; REVA Network. High-flow oxygen through nasal cannula in acute hypoxemic respiratory failure. N Engl J Med 2015;372:2185–2196. |
| 6. | Stéphan F, Barrucand B, Petit P, Rézaiguia-Delclaux S, Médard A, Delannoy B, Cosserant B, Flicoteaux G, Imbert A, Pilorge C, et al.; BiPOP Study Group. High-flow nasal oxygen vs noninvasive positive airway pressure in hypoxemic patients after cardiothoracic surgery: a randomized clinical trial. JAMA 2015;313:2331–2339. |
| 7. | Hernández G, Vaquero C, González P, Subira C, Frutos-Vivar F, Rialp G, Laborda C, Colinas L, Cuena R, Fernández R. Effect of postextubation high-flow nasal cannula vs conventional oxygen therapy on reintubation in low-risk patients: a randomized clinical trial. JAMA 2016;315:1354–1361. |
| 8. | Spoletini G, Alotaibi M, Blasi F, Hill NS. Heated humidified high-flow nasal oxygen in adults: mechanisms of action and clinical implications. Chest 2015;148:253–261. |
| 9. | Lenglet H, Sztrymf B, Leroy C, Brun P, Dreyfuss D, Ricard JD. Humidified high flow nasal oxygen during respiratory failure in the emergency department: feasibility and efficacy. Respir Care 2012;57:1873–1878. |
| 10. | Bräunlich J, Beyer D, Mai D, Hammerschmidt S, Seyfarth HJ, Wirtz H. Effects of nasal high flow on ventilation in volunteers, COPD and idiopathic pulmonary fibrosis patients. Respiration 2013;85:319–325. |
| 11. | Pham TM, O’Malley L, Mayfield S, Martin S, Schibler A. The effect of high flow nasal cannula therapy on the work of breathing in infants with bronchiolitis. Pediatr Pulmonol 2015;50:713–720. |
| 12. | Fraser JF, Spooner AJ, Dunster KR, Anstey CM, Corley A. Nasal high flow oxygen therapy in patients with COPD reduces respiratory rate and tissue carbon dioxide while increasing tidal and end-expiratory lung volumes: a randomised crossover trial. Thorax 2016;71:759–761. |
| 13. | Babcock MA, Pegelow DF, Harms CA, Dempsey JA. Effects of respiratory muscle unloading on exercise-induced diaphragm fatigue. J Appl Physiol (1985) 2002;93:201–206. |
| 14. | Bellani G, Guerra L, Musch G, Zanella A, Patroniti N, Mauri T, Messa C, Pesenti A. Lung regional metabolic activity and gas volume changes induced by tidal ventilation in patients with acute lung injury. Am J Respir Crit Care Med 2011;183:1193–1199. |
| 15. | Mauri T, Eronia N, Grasselli G, Tagliabue P, Rona R, Suriano G, Bellani G, Pesenti A. Mechanisms underlying relief of dyspnea by high flow nasal cannula in acute respiratory failure patients. Am J Respir Crit Care Med 2015;191:A1181. |
| 16. | Mojoli F, Chiumello D, Pozzi M, Algieri I, Bianzina S, Luoni S, Volta CA, Braschi A, Brochard L. Esophageal pressure measurements under different conditions of intrathoracic pressure: an in vitro study of second generation balloon catheters. Minerva Anestesiol 2015;81:855–864. |
| 17. | Mojoli F, Iotti GA, Torriglia F, Pozzi M, Volta CA, Bianzina S, Braschi A, Brochard L. In vivo calibration of esophageal pressure in the mechanically ventilated patient makes measurements reliable. Crit Care 2016;20:98. |
| 18. | Mauri T, Eronia N, Abbruzzese C, Marcolin R, Coppadoro A, Spadaro S, Patroniti N, Bellani G, Pesenti A. Effects of sigh on regional lung strain and ventilation heterogeneity in acute respiratory failure patients undergoing assisted mechanical ventilation. Crit Care Med 2015;43:1823–1831. |
| 19. | Vargas F, Saint-Leger M, Boyer A, Bui NH, Hilbert G. Physiologic effects of high-flow nasal cannula oxygen in critical care subjects. Respir Care 2015;60:1369–1376. |
| 20. | Schwabbauer N, Berg B, Blumenstock G, Haap M, Hetzel J, Riessen R. Nasal high-flow oxygen therapy in patients with hypoxic respiratory failure: effect on functional and subjective respiratory parameters compared to conventional oxygen therapy and non-invasive ventilation (NIV). BMC Anesthesiol 2014;14:66. |
| 21. | Mauri T, Yoshida T, Bellani G, Goligher EC, Carteaux G, Rittayamai N, Mojoli F, Chiumello D, Piquilloud L, Grasso S, et al.; PLeUral pressure working Group (PLUG—Acute Respiratory Failure section of the European Society of Intensive Care Medicine). Esophageal and transpulmonary pressure in the clinical setting: meaning, usefulness and perspectives. Intensive Care Med 2016;42:1360–1373. |
| 22. | Parke RL, McGuinness SP. Pressures delivered by nasal high flow oxygen during all phases of the respiratory cycle. Respir Care 2013;58:1621–1624. |
| 23. | Wexler HR, Lok P. A simple formula for adjusting arterial carbon dioxide tension. Can Anaesth Soc J 1981;28:370–372. |
| 24. | Zhao Z, Pulletz S, Frerichs I, Müller-Lisse U, Möller K. The EIT-based global inhomogeneity index is highly correlated with regional lung opening in patients with acute respiratory distress syndrome. BMC Res Notes 2014;7:82. |
| 25. | Cressoni M, Cadringher P, Chiurazzi C, Amini M, Gallazzi E, Marino A, Brioni M, Carlesso E, Chiumello D, Quintel M, et al. Lung inhomogeneity in patients with acute respiratory distress syndrome. Am J Respir Crit Care Med 2014;189:149–158. |
| 26. | Mertens M, Tabuchi A, Meissner S, Krueger A, Schirrmann K, Kertzscher U, Pries AR, Slutsky AS, Koch E, Kuebler WM. Alveolar dynamics in acute lung injury: heterogeneous distension rather than cyclic opening and collapse. Crit Care Med 2009;37:2604–2611. |
| 27. | Bodenstein M, Boehme S, Bierschock S, Vogt A, David M, Markstaller K. Determination of respiratory gas flow by electrical impedance tomography in an animal model of mechanical ventilation. BMC Pulm Med 2014;14:73. |
| 28. | Barnard PA, Levine S. Critique on application of diaphragmatic time-tension index to spontaneously breathing humans. J Appl Physiol (1985) 1986;60:1067–1072. |
| 29. | Saslow JG, Aghai ZH, Nakhla TA, Hart JJ, Lawrysh R, Stahl GE, Pyon KH. Work of breathing using high-flow nasal cannula in preterm infants. J Perinatol 2006;26:476–480. |
| 30. | Pesenti A, Rossi N, Calori A, Foti G, Rossi GP. Effects of short-term oxygenation changes on acute lung injury patients undergoing pressure support ventilation. Chest 1993;103:1185–1189. |
| 31. | Nuckton TJ, Alonso JA, Kallet RH, Daniel BM, Pittet JF, Eisner MD, Matthay MA. Pulmonary dead-space fraction as a risk factor for death in the acute respiratory distress syndrome. N Engl J Med 2002;346:1281–1286. |
| 32. | Sztrymf B, Messika J, Bertrand F, Hurel D, Leon R, Dreyfuss D, Ricard JD. Beneficial effects of humidified high flow nasal oxygen in critical care patients: a prospective pilot study. Intensive Care Med 2011;37:1780–1786. |
| 33. | Yoshida T, Roldan R, Beraldo MA, Torsani V, Gomes S, De Santis RR, Costa EL, Tucci MR, Lima RG, Kavanagh BP, et al. Spontaneous effort during mechanical ventilation: maximal injury with less positive end-expiratory pressure. Crit Care Med 2016;44:e678–e688. |
| 34. | Brochard L, Slutsky A, Pesenti A. Mechanical ventilation to minimize progression of lung injury in acute respiratory failure. Am J Respir Crit Care Med 2017;195:438–442. |
| 35. | Hraiech S, Yoshida T, Papazian L. Balancing neuromuscular blockade versus preserved muscle activity. Curr Opin Crit Care 2015;21:26–33. |
| 36. | van der Burg PS, Miedema M, de Jongh FH, Frerichs I, van Kaam AH. Cross-sectional changes in lung volume measured by electrical impedance tomography are representative for the whole lung in ventilated preterm infants. Crit Care Med 2014;42:1524–1530. |
| 37. | Grivans C, Lundin S, Stenqvist O, Lindgren S. Positive end-expiratory pressure-induced changes in end-expiratory lung volume measured by spirometry and electric impedance tomography. Acta Anaesthesiol Scand 2011;55:1068–1077. |
| 38. | Mauri T, Eronia N, Turrini C, Battistini M, Grasselli G, Rona R, Volta CA, Bellani G, Pesenti A. Bedside assessment of the effects of positive end-expiratory pressure on lung inflation and recruitment by the helium dilution technique and electrical impedance tomography. Intensive Care Med 2016;42:1576–1587. |
| 39. | Bellani G, Mauri T, Coppadoro A, Grasselli G, Patroniti N, Spadaro S, Sala V, Foti G, Pesenti A. Estimation of patient’s inspiratory effort from the electrical activity of the diaphragm. Crit Care Med 2013;41:1483–1491. |
Supported in part by institutional funding of the Department of Medicine, University of Milan-Bicocca, Monza, Italy. Fisher and Paykel Healthcare, Auckland, New Zealand, provided the device and disposables to deliver high-flow nasal cannula therapy free of charge. The supporting company had no role in the conception, design and conduct of the study, data analysis, and writing of the manuscript.
Author Contributions: Substantial contributions to the conception or design of the work, T.M., N.E., G.G., G.B., and A.P. Acquisition, analysis, or interpretation of data for the work, all authors. Drafting the work or revising it critically for important intellectual content, T.M., C.T., G.G., C.A.V., G.B., and A.P. Final approval of the version submitted for publication, all authors. Accountability for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved, T.M. and A.P.
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org
Originally Published in Press as DOI: 10.1164/rccm.201605-0916OC on December 20, 2016
Author disclosures are available with the text of this article at www.atsjournals.org.
