High-flow nasal cannula (HFNC) has recently emerged as a crucial therapeutic support strategy for hypoxemic patients [1]. The effects of HFNC on the airways and lung mechanics have been extensively described ^[2][3][2,3]. In brief, through the generation of a humidified and heated flow of medical gas, whose Fraction of Inhaled Oxygen (FiO₂) can be easily regulated and monitored, the device is able to:

• humidify the epithelium of the airways, thus improving ciliary motion and mucus clearance [4];
• wash out the dead space, removing excess carbon dioxide (CO₂), thus improving hypercapnia [5];
• reduce the work of breathing [6];
• generate a small end-expiratory positive pressure (PEEP), which helps wash out CO₂ and prevents the airways from collapsing during the expiratory phase [7].

In addition, in comparison with Non-Invasive Ventilation (NIV) and Conventional Oxygen Therapy (COT), HFNC is often perceived as more comfortable by patients [4]. Therefore, HFNC has been increasingly used in clinical practice for several different respiratory conditions, from bronchiolitis in children [8] to acute respiratory failure (ARF) in adults ^[9][10][9,10], with a crucial role during the COVID-19 pandemic in patients with different severities of hypoxemia ^{[11][12][13][14]}[11,12,13,14]. Moreover, HFNC has been applied in the setting of chronic respiratory failure (CRF) secondary to lung diseases, such as interstitial lung diseases (ILDs) and chronic obstructive pulmonary disease (COPD). There is also evidence of a significant beneficial effect of long-term domiciliary treatment with HFNC in patients with bronchiectasis ^[15][16][15,16]. HFNC has been used in mild hypercapnia secondary to COPD with promising results [17]. In particular, one metanalysis of three studies involving stable hypercapnic COPD patients demonstrated that long-term domiciliary treatment with HFNC reduced the exacerbation rate and improved the patients’ Quality of Life (QoL) compared with long-term COT, albeit without any significant change in the arterial oxygen (O₂) and CO₂ partial pressures (PaO₂ and PaCO₂, respectively) [18].

According to the 2022 European Respiratory Society practical guidelines [19], HFNC should be preferred over COT in ARF and over NIV in hypoxemic ARF; HFNC is equivalent to NIV in post-operative patients at high risk of pulmonary complications and is considered a second-line treatment, after trialing for NIV, in hypercapnic COPD patients.

Finally, recent evidence has shown an emerging role for HFNC in the setting of exercise training (ET) and pulmonary rehabilitation (PR) as a tool for improving endurance time and dyspnea scores in respiratory patients [20].

2. HFNC during ET and PR in ILDs

With fewer results in comparison with COPD, some research groups investigated the role of HFNC in the setting of ILDs. Among the 32 patients evaluated by Chihara et al. ^[21][34], there were 15 subjects with interstitial pulmonary fibrosis (46.9%). As described above, patients were randomly assigned to the interventional group, which was characterized by the presence of HFNC at 50 L/min and FiO₂ 100% during training, or to the control group, which received O₂ at the flow of 6 L/min through nasal cannula. All patients underwent an incremental-load exercise test and three constant-load exercise tests (with daily LTOT, HFNC at 50 L/min FiO₂ 100%, and OT at 6 L/min) before and after rehabilitation. The protocol was structured on exercises in five sessions per week on the same cycle ergometer with an initial setting at 80% of the maximum workload achieved at incremental-load testing. After 4 weeks, the interventional group presented overall with a significant improvement in the 6MWT in terms of meters walked, with a mean difference of 55.2 ± 69.6 m versus −0.5 ± 87.3 m in the control group. However, only in five out of nine (55.6%) pulmonary fibrosis patients assigned to the HFNC group an increase of ≥30 m in the 6MWD was observed, versus a proportion of four out of six patients (66.67%) in the control group. In both groups, the duration of the constant-load exercise test increased at 4 weeks in comparison to baseline. Harada et al. ^[22][39] designed a randomized crossover trial that investigated the effect of applying HFNC in 24 patients during exercise testing in comparison to VM. All patients had been diagnosed with idiopathic pulmonary fibrosis (IPF) according to the 2018 ATS/ERS/JRS/ALAT guideline ^[23][40] and manifested exertional dyspnea and desaturation (SpO₂ < 90% at the 6MWT) at baseline. All the patients enrolled underwent an incremental load exercise test and a baseline constant-load exercise test at 80% peak work. Afterward, they underwent the same test another time with HFNC and another with VM, in a random order, thus forming two groups (Arm A, with HFNC test done prior to VM, and Arm B, the opposite). The primary outcome was endurance time in the constant-load exercise test, while the secondary outcomes were SpO₂, dyspnea, leg fatigue, heart rate, and comfort. The study met the primary outcome because the HFNC group had an increased endurance time than the VM one; the study also demonstrated that, when on HFNC support, patients showed a higher peripheral saturation and reduced leg fatigue. However, dyspnea, maximum heart rate, and comfort at 80% peak work rate were not affected by the application of HFNC. Suzuki et al. ^[24][41] investigated how HFNC could alter the response to ET in patients diagnosed with fibrotic ILDs in comparison with VM. The study was a randomized crossover one and enrolled 20 patients who had only mild desaturation at the basal constant-work-rate endurance test (SpO₂ ≥ 88%). All patients underwent a symptom-limited incremental exercise test; then, patients were randomly assigned to group A, where a high-intensity constant-work-rate endurance test was performed first with VM, and, on the following day, on HFNC; and group B, where the tests were performed in the reverse order. The primary outcome was endurance time, while the secondary outcomes were SpO₂, HR, dyspnea assessed with the BORG scale, comfort, and adverse events. The study did not show any statistically significant differences between the two groups in all the outcomes. Neither the subgroup analysis of the “good responders” (defined as an improvement >100 s or 33% in endurance time in comparison to the baseline test) nor the stratification according to the presence of pulmonary hypertension demonstrated any superiority of HFNC on VM. These discouraging results, however, should be interpreted with caution, because the sample included patients with diverse kinds of ILDs with different underlying pathological mechanisms and systemic involvement. To conclude, it can be stated that, in the setting of ILDs, the evidence supporting HFNC therapy for ET is sparse and inconsistent, given the limited number of studies and the high heterogeneity of the patients involved. However, one study ^[21][34] has demonstrated an improvement in the performance in the 6MWT, although not clinically significant, while dyspnea was substantially unaffected in all the three studies included in the current review ^[21][22][24][34,39,41]. This could be related to the fact that dyspnea in ILDs can be controlled by supplementing oxygen at higher FiO₂ without a significant risk of hypercapnia, which, on the contrary, is an important limitation in COPD patients.

3. HFNC and ET in Patients with Primary or Secondary Lung Cancer

One study by Hui et al. ^[25][42] evaluated the role of the supplementation of O₂ through HFNC or COT in 45 non-hypoxemic patients diagnosed with cancer and primary or secondary lung involvement. Patients were randomly assigned to one of the four investigational groups (HFNC + FiO₂ 100%, HFNC without O₂ enrichment, COT, and low flow air at 2 L/min). Patients in the HFNC + FiO₂ 100% group improved dyspnea and leg discomfort at peak exercise as well as endurance time in constant-work-rate exercise testing. The study demonstrated a significant reduction in dyspnea in non-hypoxemic patients and an improvement in endurance time, but such intriguing results need to be confirmed by future larger trials. The studies discussed above have been summarized in Table 1.

Study	Subjects	Design	Intervention	Performance Tests	Main Findings
Chihara et al. (2022) [21][34]	15 IPF patients with CRF	Single-center RCT/crossover trial	HFNC at 50 L/min and FiO2 100% vs. Oxygen at 6 L/min through nasal cannula during PR	6MWT, Constant-load exercise testing at 80% maximal capacity	↑ 6MWD, ↔ Endurance time, ↔ Dyspnea, ↔ HR, ↔ BP, ↔ RR
Harada et al. (2022) [22][39]	24 IPF patients with exertional dyspnea and desaturation	Randomized crossover trial	HFNC vs. VM at the same FiO2	Constant-load exercise testing at 80% maximal capacity	↑ Endurance Time, ↑ SpO2, ↓ Leg Fatigue, ↔ Dyspnea, ↔ HR, ↔ Comfort
Suzuki et al. (2020) [24][41]	20 fibrotic ILDs patients with mild exertional desaturation	Randomized crossover trial	HFNC vs. VM at the same FiO2	Constant-load exercise testing	↔ Endurance Time, ↔ SpO2, ↔ Dyspnea, ↔ HR, ↔ Comfort
Hui et al. (2021) [25][42]	45 non-hypoxemic patients with primary or secondary Lung Cancer	Single-center RCT	HFNC with FiO2 100% vs. HFNC without O2 vs. COT vs. Low-Flow Air at 2 L/min	Constant-load exercise testing	↑ Endurance Time +, ↓ Dyspnea +