Double cycling with breath-stacking during partial support ventilation in ARDS: Just a feature of natural variability?

Abstract

Background

Double cycling with breath-stacking (DC/BS) during controlled mechanical ventilation is considered potentially injurious, reflecting a high respiratory drive. During partial ventilatory support, its occurrence might be attributable to physiological variability of breathing patterns, reflecting the response of the mode without carrying specific risks.

Methods

This secondary analysis of a crossover study evaluated DC/BS events in hypoxemic patients resuming spontaneous breathing in cross-over under neurally adjusted ventilatory assist (NAVA), proportional assist ventilation (PAV +), and pressure support ventilation (PSV). DC/BS was defined as two inspiratory cycles with incomplete exhalation. Measurements included electrical impedance signal, airway pressure, esophageal and gastric pressures, and flow. Breathing variability, dynamic compliance (C_Ldyn), and end-expiratory lung impedance (EELI) were analyzed.

Results

Twenty patients under assisted breathing, with a median of 9 [5–14] days on mechanical ventilation, were included. DC/BS was attributed to either a single (42%) or two apparent consecutive inspiratory efforts (58%). The median [IQR] incidence of DC/BS was low: 0.6 [0.1–2.6] % in NAVA, 0.0 [0.0–0.4] % in PAV + , and 0.1 [0.0–0.4] % in PSV (p = 0.06). DC/BS events were associated with patient’s coefficient of variability for tidal volume (p = 0.014) and respiratory rate (p = 0.011). DC/BS breaths exhibited higher tidal volume, muscular pressure and regional stretch compared to regular breaths. Post-DC/BS cycles frequently exhibited improved EELI and C_Ldyn, with no evidence of expiratory muscle activation in 63% of cases.

Conclusions

DC/BS events during partial ventilatory support were infrequent and linked to breathing variability. Their frequency and physiological effects on lung compliance and EELI resemble spontaneous sighs and may not be considered a priori as harmful.

Introduction

Double cycling (DC) is a patient-ventilator asynchrony typically associated with a sustained inspiratory effort that persists beyond the ventilator’s inspiratory time, causing the inspiratory valve to open twice consecutively [1,2,3]. Due to incomplete exhalation, the volume of the second cycle is added to the first cycle, causing breath-stacking (BS), a phenomenon traditionally associated with an increased risk of lung injury [4, 5].

The combined phenomenon DC/BS has been mainly described during assist-control ventilation (ACV) associated with high respiratory drive [2]. Indeed, a short and fixed inspiratory time set during ACV in the presence of a high drive and effort are the factors suggested to induce double cycling.

The mechanisms underlying DC/BS during partial ventilatory support as well as their impact on the lungs remain poorly understood. When patients are switched to assisted breathing with no mandatory breaths, respiratory variability increases as it should physiologically do, but the ventilator working mechanism may not necessarily adapt well to intermittent abrupt changes in effort as it naturally exists during sighs, which are inherent markers of this variability [6].

We hypothesized that DC/BS events observed during partial support ventilation may not be necessarily harmful (e.g. requiring sedation) but may primarily reflect natural variability in breathing patterns. Consequently, the incidence of DC/BS in partial support ventilation should be associated with the coefficient of variation of tidal volume (V_T) and respiratory rate (RR).

We used data from a randomized crossover physiological study in patients with hypoxemic respiratory failure switched to assisted ventilation and assessed during neurally adjusted ventilatory assist (NAVA), proportional assist ventilation (PAV+), and pressure support ventilation (PSV) in cross-over [7] to assess the incidence of DC/BS events in each mode and their relationship with breathing variability. Additionally, we evaluated the characteristics and consequences of DC/BS on regional lung stretch, end-expiratory lung impedance (EELI), and dynamic lung compliance (C_Ldyn).

Methods

This secondary analysis is based on a short-term crossover study (ethical approval number N.027/2016) involving hypoxemic patients resuming spontaneous breathing. After a period of controlled mechanical ventilation exceeding 72 h, patients were randomized to receive partial ventilatory support using NAVA, PAV + , or PSV, each for 20-min intervals in a randomized cross-over fashion [7]. The patients were hemodynamically stable under moderate-light sedation, used individualized levels of end-expiratory pressure (PEEP) (defined as the crossing point of the overdistension and collapse using electrical impedance tomography (EIT)), and received similar mean V_T during all modes [7]. Airway pressure (Paw), esophageal (Pes) and gastric (Pga) pressures, transpulmonary pressure (P_L) and flow, were measured. P_L was calculated as the difference between Paw and Pes.

Three experts, blinded to the esophageal pressure swings, reviewed the continuous tracings of V_T, Paw and flow from each patient to identify double cycling (defined as two consecutive ventilatory cycles separated by an expiratory time less than half the mean insufflation time) and diagnosed breath-stacking when the second insufflation occurred before complete exhalation of the previous (defined as an expired tidal volume lower than half of the inspired tidal volume) [4]. The incidence of DC/BS per patient was calculated as the number of events per 100 ventilatory cycles. Coefficient of variation (CV = [standard deviation / mean] * 100) of V_T and RR were used to estimate the variability of breathing pattern during the time of observation in each mode. We analyzed muscular pressure (P_MUS), V_T and regional inflation (tidal impedance change in arbitrary units “ΔZ_AU”) in DC/BS cycles and in the 5 previous regular cycles (control). P_MUS was calculated as the difference between Pes and the chest wall elastic recoil pressure (Pcw) during inspiration [8]. Chest wall elastance was measured under controlled ventilation, before the study. P_MUS in DC/BS cycles considered the maximum Pes swing and total inspired tidal volume. EELI changes in arbitrary units (ΔEELI_AU) and dynamic lung compliance changes (ΔC_Ldyn) were obtained by comparing 5 cycles before and after the DC/BS. ΔEELI_AU was transformed in volume in milliliter (ΔEELI_ml), by multiplying ΔEELI_AU by the ratio of V_T/ΔZ_AU [9]. C_Ldyn was calculated as V_T divided by the maximal transpulmonary pressure swing.

Friedman test followed by Dunn's post-hoc test was performed to compare the DC/BS incidences, and CV-V_T and CV-RR, within NAVA, PAV + and PSV. Linear mixed models with patients as random intercepts, adjusted by ventilatory mode, were performed to associate breathing variability with DC/BS incidence as dependent variable. Linear mixed models were also used to evaluate the association between DC/BS and respiratory variables. We looked at the consequence of DC/BS on ΔEELI_ml and ΔC_Ldyn, considering the expiratory muscle activity when Pga was available. Analyses were performed in Stata v17.0 and GraphPad Prism v10.

Results

The 20 patients had a median and interquartile range (IQR) of 9 [5–14] days on mechanical ventilation with a PaO₂:FiO₂ ratio 275 ± 46 mmHg. PEEP was 10 [7,8,9,10,11,12] cmH₂O, V_T was 7.8 [7.4–8.9], 7.9 [7.2–8.8], and 7.7 [7.2–9.3] ml/kg of predicted body weight in NAVA, PAV + and PSV respectively (p = 0.859); RR was 24.9 ± 7, 24.9 ± 7, and 22.3 ± 6/min in NAVA, PAV + and PSV respectively (p = 0.060).

DC/BS was not frequent; the median [IQR] incidence was 0.6 [0.1–2.6] % in NAVA, 0.0 [0.0–0.4] % in PAV + and 0.1 [0.0–0.4] % in PSV (p = 0.06); i.e. approximately 1–2 cycles with DC/BS every 5 min across all modes. Representative cases are shown in Fig. 1. DC/BS was observed in 15 (75%), 9 (45%) and 10 (50%) of patients in NAVA, PAV + and PSV respectively; 5 patients showed DC/BS with all three modes.

Fig. 1

Representative tracings with breath-stacking cycles in NAVA, PAV + and PSV Fig. 1 illustrates three representative cases with respiratory variables and EIT-derived signals over time showing regular and breath-stacking cycles during NAVA (patient #5), PAV + (patient #18) and PSV (patient #14). On each panel, from top to bottom: tidal volume, airway pressure (Paw), flow, esophageal pressure (Pes) and global impedance change (∆Z), in arbitrary units, A.U. Breath-stacking events are indicated under gray boxes. Please notice that breath-stacking cycles have two consecutive efforts

The median [IQR] of CV-V_T was 26.9 [18.1–45.6] % in NAVA, 16.5 [11.1–26.6] % in PAV + and 17.0 [11.9–20.3] % in PSV (p = 0.0001 for NAVA vs PSV); CV-RR was 19.3 [11.4–25,4] % in NAVA, 10.7 [8.4–13.3] % in PAV + and 12.8 [8.7–17.4] % in PSV (p = 0.0007 for NAVA vs PSV). Independently of the ventilatory mode and excluding the DC/BS cycles, the incidence of DC/BS was positively associated with increased variability in V_T (p = 0.014) and RR (p = 0.011). For every 10% increase in CV-V_T or CV-RR, breath-stacking incidence increased by around 0.5%.

We analyzed 50 DC/BS cycles with reliable esophageal and gastric signals (25 in NAVA, 10 in PAV + and 15 in PSV). DC/BS was associated with either a single or two apparent consecutive inspiratory efforts in 42% and 58% of the cases (Fig. 1); one inspiratory effort was more frequent in NAVA (17 of 25), and two apparent efforts in PAV + (9 of 10) and PSV (12 of 15).

Compared with regular cycles, DC/BS was associated with slightly higher P_MUS (15.1 ± 5.4 cmH₂O vs 12.2 ± 4.2, p < 0.001), V_T (550 ± 208 ml vs 459 ± 116, p = 0.001) and regional stretch (non-dependent region: 11.5 ± 6.6 A.U. vs 9.5 ± 5.2 A.U., p = 0.004; dependent region: 15.9 ± 8.6 A.U. vs 12.7 ± 5.1 A.U., p = 0.002). These differences were mainly explained by the events displaying two apparent efforts (Fig. 2A and B).

Fig. 2

Comparisons between breath-stacking and regular cycles, and changes of end-expiratory lung impedance and dynamic lung compliance before-and-after breath-stacking cycles. Individual values and mean (solid line) of [A] tidal volume (VT) and [B] muscular pressure (P_MUS), in regular cycles (blue empty circles), double cycling with breath-stacking cycles (DC/BS) with one inspiratory effort (1-Eff, black empty circles), and DC/BS with two inspiratory efforts (2-Eff, dark gray empty circles). The p-values correspond to linear mixed models adjusted by modes, comparing regular breaths with DC/BS with one and two efforts. Similar findings were observed in regional stretch at dependent and non-dependent regions. Panels C and D show individual values of delta end-expiratory lung impedance (ΔEELI_ml) and delta dynamic lung compliance (ΔC_Ldyn) between the regular cycles before-and-after breath-stacking cycle, depending on the behavior of expiratory muscle activity after DC/BS. The horizontal lines represent the means of ΔEELI_ml and ΔC_Ldyn (blue without expiratory activity and black with expiratory activity). The p-values correspond to linear mixed models adjusted by modes. The associations between expiratory muscle activation and ΔEELI or ΔC_Ldyn are independent of the number of inspiratory efforts

In terms of expiratory muscle activity following each DC/BS event, we could measure a Pga rise during expiration on 27 tracings (9 in NAVA, 7 PAV + and 11 in PSV). The majority of DC/BS cycles (n = 17, 63%) were not followed by any visible expiratory activation, and were associated with an improvement in C_Ldyn comparing the five breaths after DC/BS to the five breaths before. In 10 cases (37%), DC/BS cycles were followed by a Pga rise more than twice that of the previous cycles and accompanied by a negative drop in both EELI and C_Ldyn (Fig. 2C and D).

Discussion

Our findings indicate that DC/BS events were rare and associated with greater variability of V_T and RR during partial support ventilation. Interestingly, we observed that DC/BS cycles were often generated by two apparent consecutive, closely spaced inspiratory efforts.

In terms of the characteristics and physiological effects of DC/BS, these asynchronous cycles were characterized by higher P_MUS, increased V_T and regional lung stretch. Notably, in the majority of cases, DC/BS were followed by an improvement in dynamic lung compliance and EELI. The combination of their positive effects on the respiratory system, their low incidence, and their association with natural variability raise doubts about their potential harm.

Our findings align with prior studies demonstrating greater variability in breathing patterns during NAVA compared to PSV [10, 11]. The relatively low frequency of the combined phenomenon of DC and BS in our patients may explain why we observed only a trend toward a higher incidence in NAVA. Importantly, variability in V_T and RR appears to be a strong predictor of DC/BS occurrence, irrespective of the ventilatory mode.

Classical physiological studies demonstrate that sighs typically occur during the early expiratory phase of a normal tidal volume, a phenomenon that is part of natural breathing variability [6, 12]. Sighs may occur approximately at the frequency found for the DC/BS observed in this study and do not reflect a sustained high respiratory drive [6]. Sighs may assist in re-expanding atelectatic airspaces and stimulating surfactant secretion [6, 13]. In our study, the majority of case patients showed increased end-expiratory lung volume and lung compliance following a DC/BS, particularly when expiratory effort was absent.

Our findings should be interpreted with caution due to the small sample size, the secondary nature of the analysis, and the short duration of spontaneous ventilation.

In conclusion, DC/BS during partial support modes may reflect natural breathing variability and may not be considered a priori as harmful. Their low frequency and their effect on lung compliance and lung volume suggest a physiological role similar to sighs.