Background

Patients with interstitial lung disease and usual interstitial pneumonia pattern (ILD-UIP) may experience severe acute hypoxic respiratory failure (AHRF) during acute exacerbations (AE-ILD-UIP) [1], requiring invasive respiratory support (MV) [2]; nevertheless, the mortality following MV exceeds 80% [3]. Patho-physiologically, AE-ILD resembles acute respiratory distress syndrome (ARDS) with diffuse alveolar damage (DAD), superimposed on a background of lung fibrosis [4].

In ARDS, lung-protective MV strategies contributed to mitigate ventilatory induced lung injury (VILI), thus decreasing mortality [5, 6]. Talmor and coworkers showed that an esophageal pressure (Pes)-guided positive end-expiratory pressure (PEEP) titration to obtain a positive end-expiratory transpulmonary pressure (PL,EE) is useful to recruit dependent lung regions, improve lung mechanics and minimize atelectrauma in these patients [7].

Retrospective data suggest that patients with AE-ILD are particularly susceptible to stress and strain, and hence at higher risk of VILI [8]. Thus, it seems straightforward to use lung-protective ventilatory strategies in these patients. However, little is known on PL,EE in patients with AE-ILD-UIP and even less on the potential impact of lung-protective strategies aimed at maintaining positive PL,EE.

We studied the impact of different PEEP settings (zero PEEP [ZEEP], PEEPLOW and PEEPTITRATED to obtain positive PL,EE) in patients with AE-ILD-UIP, compared it with matched primary ARDS controls. We hypothesized that the impact of PEEP titration on partitioned respiratory mechanics could be different between the groups.

Methods

Study setting and population

The study (ClinicalTrial.gov ID NCT05098717) was carried out at the Respiratory Intensive Care Unit (RICU) of the University Hospital of Modena (Italy) in accordance with the Ethics Committee “Area Vasta Emilia Nord” approval (registered protocol number 327/2022). Informed consent to divulgate data was obtained from participants or their relatives, as appropriate.

Patients with AE-ILD-UIP developing AHRF and consecutively admitted to the RICU (August 1st, 2016, to July 1st, 2022) were eligible for enrollment. Inclusion criteria were age > 18 years; established diagnosis of ILD with a UIP pattern on a high-resolution computed tomography scan; invasive MV in volume-controlled mode. Patients suffering from chronic obstructive pulmonary disease, neuromuscular disease and chest wall deformities were excluded. AE-ILD-UIP were then matched 1:1 by body mass index, PaO2/FIO2 and acute physiology and chronic health evaluation (APACHE) II score at admission, to a group of patients with primary ARDS under MV extracted from our dataset over the same period.

Study procedures and aim

According to our institutional protocol, patients with AE-ILD-UIP or ARDS requiring MV were submitted to a partitioned respiratory mechanics measurements within 24 h from admission during three different lung-protective strategies including low VT (6 ml/Kg/PBW) and three consecutive PEEP levels, i.e., 0 cmH2O (ZEEP), 4–8 cmH2O (PEEPLOW), and Pes-guided titration to obtain positive PL,EE (PEEPTITRATED). At each phase, PEEP level was maintained for 30 min before recording all respiratory parameters and arterial blood sampling (see details in Additional file 1: Supplement [9, 10]).

The aim was to report measures of partitioned respiratory mechanics under lung-protective MV at different PEEP levels in patients with AE-ILD-UIP compared with ARDS.

Data collection and analysis plan

Demographics, clinical characteristics, available pulmonary functions tests within 12 months before AE-ILD and partitioned respiratory mechanics were collected.

Data were displayed as median and interquartile range for continuous variables and numbers and percentages for dichotomous variables. Group comparison was built using a one-to-one propensity score matching procedure with the nearest-neighbor method without replacement (caliper = 0.2). Comparison between continuous variables was performed with Wilcoxon and Wilcoxon signed-rank tests. Dichotomous variables were compared using the χ2 test. Kruskal–Wallis was used to test as an interaction for whether the change in respiratory mechanics and physiological variables according to PEEP settings was different between groups. Statistics was performed using SPSS version 25.0 with PSMATCHING3 R Extension command (IBM Corp., Armonk, NY, USA) and GraphPad Prism version 8.0 (GraphPad Software, Inc., La Jolla, Ca, USA) unless otherwise indicated.

Results

Respiratory mechanics of AE-ILD-UIP

Over the study period a total of 21 patients with AE-ILD-UIP underwent MV. Of these, 10 patients were analyzed according to inclusion criteria (see Additional file 1: Supplement). All of them died while on MV.

Respiratory mechanics of AE-ILD-UIP at different PEEP levels are reported in Table 1, while changes in respiratory mechanics at different levels are shown in Fig. 1 (and Additional file 1: eFigure 2, Supplement). At ZEEP the median lung elastance (EL) was 44.4 cmH2O/L, transpulmonary driving pressure (ΔPL) was 21.1 cmH2O, PL,EE was − 4.3 cmH2O and end-inspiratory transpulmonary pressure (PL,EI) was 16.7 cmH2O (Table 1). During the PEEPLOW phase PL,EE remained below 0 cmH2O (Table 1), median EL and PL,EI did not change (Fig. 1, panel A and E) while ΔPL significantly decreased from baseline (p = 0.018,). During the PEEPTITRATED phase PL,EE was 0.8 cmH2O (Table 1) and EL significantly increased as compared to both ZEEP and PEEPLOW (p = 0.04 and p < 0.0001, respectively, Fig. 1, panel A), while PL,EI and ΔPL were higher as compared to PEEPLOW (p < 0.001 and p = 0.003, respectively, Fig. 1, panel E and G).

Table 1 Blood gas analyses and partitioned respiratory mechanics of the AE-ILD-UIP and the ARDS population at different PEEP levels. Data are presented as median value and IQR
Full size table
Fig. 1
figure 1

Measured individual values of EL, PL,EE, PL,EI and ΔPL the matched study groups at ZEEP, PEEPLOW and PEEPTITRATED phase. When testing as an interaction for whether the change in physiological variables at different PEEP levels was different between AE-ILD-UIP and ARDS (dotted p-values line), statistical difference was found for EL (p < 0.001, panel A and B), PL,EI (p < 0.001, panel E and F, p < 0.00) and ΔPL (< 0.001, panel G and H). EL, lung elastance; PL,EI, end-inspiratory transpulmonary pressure; PL,EE, end-expiratory transpulmonary pressure; PL, transpulmonary driving pressure; ZEEP, zero positive end-expiratory pressure; PEEP, positive end-expiratory pressure; AE-ILD-UIP, acute exacerbation of interstitial lung disease with usual interstitial pneumonia pattern; ARDS, acute respiratory distress syndrome

Full size image

AE-ILD-UIP as compared with historical, matched, ARDS controls

AE-ILD-UIP and matched ARDS groups were similar for SAPS II score (Additional file 1: eTable 1, Supplement). Lung infection was the cause for developing ARDS in all patients.

PEEPTITRATED, but not PEEPLOW setting, resulted in higher PEEP in ARDS () compared with AE-ILD-UIP (14 VS 12 cmH2O, p < 0.001). At ZEEP, ARDS patients had lower EL (17.9 cmH2O/L, p ≤ 0.0001), PL,EI (4.4 cmH2O, p < 0.0001), and ΔPL (9.3 cmH2O, p < 0.0001) compared with AE-ILD-UIP. During the PEEPLOW and the PEEPTITRATED phases, ARDS patients still had lower EL (14.6 cmH2O/L, p < 0.0001 and 15.2 cmH2O/L, p < 0.0001 respectively), PL,EI (10.5 cmH2O, p < 0.0001 and 16.9 cmH2O, p = 0.001 respectively), and ΔPL (12.3 cmH2O, p = 0.02 and 13.9 cmH2O, p = 0.0001 respectively) as compared to AE-ILD-UIP.

Figure 1 shows that during the PEEP trial EL, PL,EI and ΔPL were different in AE-ILD-UIP and ARDS patients. EL remained unchanged at PEEPLOW and worsened at PEEPTITRATED in AE-ILD-UIP, whereas it did not change in patients with ARDS (Fig. 1, panel A and B).

Discussion

With this study, we report for the first time that AE-ILD-UIP patients under lung-protective MV strategy respond favorably in terms of respiratory mechanics to low PEEP levels, whereas respond unfavorably (and rather uniformly) to a Pes-guided PEEP strategy to obtain positive PL,EE. The mechanical behavior of AE-ILD-UIP was different from that of matched “pulmonary” ARDS controls.

In our AE-ILD-UIP patients a low PEEP strategy resulted in reduction of ΔPL indirectly suggesting alveolar recruitment, probably occurring in the areas of DAD superimposed to UIP. Indeed, despite we did not measure alveolar recruitment, we assume that in AE-ILD-UIP patients the lung regions spared by fibrosis but affected by DAD were likely de-recruited at ZEEP. Thus, it seems that the low PEEP strategy could be wise in patients with AE-UIP-ILD at least in terms of lung mechanics. These results are novel and referred to a cohort of patients rarely studied in the intensive care context. A previous study by Nava et al. assessed the respiratory mechanics during MV in seven patients with end-stage idiopathic pulmonary fibrosis [11] and reported values of lung elastance (46.1 cmH2O/L) similar to those found in our work. However, in that study lung mechanics were only measured at ZEEP.

Tailored PEEP titration in the context of lung-protective ventilation is still under debate [12]. During controlled MV, PL,EE may be negative at ZEEP, indicating that the dependent lung regions are compressed [13]. This condition predisposes to tidal alveolar collapse and re-opening, resulting in high local shear forces that enhance VILI (atelectrauma) [14]. Negative PL,EE is common in ARDS patients ventilated with lower PEEP levels in supine position and this largely explains the beneficial physiological effects of PEEP titration to achieve a positive PL,EE reported in preclinical and clinical studies [15, 16]. Notwithstanding, the EPVent-2 trial showed that these positive physiological effects have to deal with the potential PEEP-induced lung injury caused by overdistension in the non-dependent lung regions [17]. We hypothesized the “Talmor” PEEP titration protocol in AE-ILD-UIP could lead to beneficial physiological effects also in patients with AE-ILD-UIP; however we found a rather sharp increase in EL and ΔPL in all of them (Fig. 1). It is tempting to speculate that Pes-guided PEEP titration resulted in squishing among the patchy fibrotic tissue of the non-fibrotic lung regions (so called “squishy ball lung” phenomenon) [18] and that this effect invalidated the potential benefits of alveolar recruitment in the dependent lung regions.

Our study suffers from limitations: the small sample, the lack of quantitative analysis of hyper-inflated lung tissue [19] during PEEP titration and no end-expiratory lung volume assessment [20] allow only preliminary pathophysiological insights. Moreover, we did not assess the role of fluid balance as confounding factor. Finally, a selection bias should be acknowledged, as patients with AE-ILD are not usually placed on MV given the poor prognosis.

Conclusions

In AE-ILD-UIP mechanically ventilated patients, low PEEP strategy may improve respiratory mechanics and, at difference with primary ARDS, PEEP titrated to obtain a positive PL,EE significantly worsened lung mechanics. This paves the way to larger studies to clarify the best physiological response to PEEP in these patients. However, we feel that our findings could have practical implications when managing patients with AE-ILD-UIP under MV, suggesting that low PEEP strategy may be preferable to prevent lung injury.