Michaël Sage1
Wendy See1
Stéphanie Nault1
Christophe Morin1
Christina Michalski2,3
Benoit Chabot4Sofia Marouan5
Pascal M. Lavoie2,3,6
Philippe Micheau7
Jean-Paul Praud1,8
Étienne Fortin-Pellerin1,8*- 1Department of Pharmacology-Physiology, Université de Sherbrooke, Sherbrooke, QC, Canada
- 2BC Children’s Hospital Research Institute, The University of British Columbia, Vancouver, BC, Canada
- 3Department of Experimental Medicine, The University of British Columbia, Vancouver, BC, Canada
- 4Department of Microbiology and Infectiology, Université de Sherbrooke, Sherbrooke, QC, Canada
- 5Department of Pathology, Université de Sherbrooke, Sherbrooke, QC, Canada
- 6Department of Pediatrics, The University of British Columbia, Vancouver, BC, Canada
- 7Department of Mechanical Engineering, Université de Sherbrooke, Sherbrooke, QC, Canada
- 8Department of Pediatrics, Université de Sherbrooke, Sherbrooke, QC, Canada
Animal experiments suggest that total liquid ventilation (TLV) induces less ventilator-induced lung injury (VILI) than conventional mechanical gas ventilation. However, TLV parameters that optimally minimize VILI in newborns remain unknown. Our objective was to compare lung inflammation between low (L-VT) and high (H-VT) liquid tidal volume and evaluate impacts on the weaning process. Sixteen anesthetized and paralyzed newborn lambs were randomized in an L-VT group (initial tidal volume of 10 mL/kg at 10/min) and an H-VT group (initial tidal volume of 20 mL/kg at 5/min). Five unventilated newborn lambs served as controls. After 4 h of TLV in the supine position, the lambs were weaned in the prone position for another 4 h. The levels of respiratory support needed during the 4 h post-TLV were compared. The anterior and posterior lung regions were assessed by a histological score and real-time quantitative PCR for IL1B, IL6, and TNF plus 12 other exploratory VILI-associated genes. All but one lamb were successfully extubated within 2 h post-TLV (72 ± 26 min vs. 63 ± 25 min, p = 0.5) with similar FiO2 at 4 h post-TLV (27 ± 6% vs. 33 ± 7%, p = 0.3) between the L-VT and H-VT lambs. No significant differences were measured in histological inflammation scores between L-VT and H-VT lambs, although lambs in both groups exhibited slightly higher scores than the control lambs. The L-VT group displayed higher IL1B mRNA expression than the H-VT group in both anterior (2.8 ± 1.5-fold increase vs. 1.3 ± 0.4-fold increase, p = 0.02) and posterior lung regions (3.0 ± 1.0-fold change increase vs. 1.1 ± 0.3-fold increase, p = 0.002), respectively. No significant differences were found in IL6 and TNF expression levels. Gene expression changes overall indicated that L-VT was associated with a qualitatively distinct inflammatory gene expression profiles compared to H-VT, which may indicate different clinical effects. In light of these findings, further mechanistic studies are warranted. In conclusion, we found no advantage of lower tidal volume use, which was in fact associated with a slightly unfavorable pattern of inflammatory gene expression.
Introduction
Despite major advances in care, respiratory support remains challenging for many infants hospitalized in the neonatal intensive care unit. Total liquid ventilation has been suggested as an alternative to conventional mechanical ventilation for various conditions affecting the neonate, such as children born with diaphragmatic hernia (Snoek et al., 2014), meconium aspiration syndrome (Avoine et al., 2011) or who are born extremely premature (Shaffer et al., 1999; Davidson and Berkelhamer, 2017). Total liquid ventilation (TLV) uses liquid perfluorochemicals (PFCs), such as perflubron (PFOB, perfluorooctyl bromide), together with a dedicated liquid ventilator. In comparison to conventional gas ventilation, this was shown to ameliorate tidal-volume distribution and to require less positive pressure, thus reducing VILI in animal models (Tooley et al., 1996; Wolfson et al., 2008; Avoine et al., 2011; Pohlmann et al., 2011; Sage et al., 2018b). Perflubron, the PFC used for this study, also has anti-inflammatory properties that could reduce VILI (Thomassen et al., 1997; Woods et al., 2000).
Over the past decade, our team has developed a liquid ventilator prototype that allows for precise control over pressures and flows delivered to the lungs (Robert et al., 2010; Sage et al., 2018a, b). These advances have allowed for tighter control over the end-expiratory lung volume of PFC, mainly through prevention of tracheal collapse occurring when excessive negative pressure is applied during expiration. These repetitive tracheal collapses cause progressive accumulation and trapping of fluid in the lungs (Costantino et al., 2004; Bagnoli et al., 2007). Our most recent prototype, called INOLIVENT, can be used to refine ventilation algorithms and lung protective strategies during TLV (Nadeau et al., 2018; Sage et al., 2018a; Kohlhauer et al., 2019).
High tidal volumes with conventional gas ventilation have been especially cited as inducing greater VILI (Ciuffini et al., 2018; Farrell et al., 2018). Our ventilation algorithms allow higher respiratory rates and lower tidal volumes to be used, thus maintaining minute ventilation, while avoiding tracheal collapses (up to ∼10 cycles/min compared to 3–6 cycles/min with previously published TLV prototypes) (Robert et al., 2007, 2009, 2010). The higher proportion of liquid tidal volume cycling within the anatomic dead space is, however, of particular concern in the context of low gas diffusion in PFC (Koen et al., 1988; Costantino and Fiore, 2001). To our knowledge, the effect of tidal volume and respiratory rate on lung inflammation during TLV has not been thoroughly addressed. Jiang et al. (2016) showed that a tidal volume of 6 mL/kg was associated with lower transcription and plasma protein levels of IL-6 and IL-8 in piglets, compared to a higher tidal volume of 25 mL/kg during TLV (Jiang et al., 2016). The authors, however, had to use extracorporeal carbon dioxide removal to achieve acceptable blood-gas exchange, thus preventing the need for high minute ventilation and its potential impact on VILI occurrence. Extracorporeal carbon dioxide removal is associated with serious adverse effects and is not currently an option for extremely low birth-weight neonates (Liu et al., 2016). Kohlhauer et al. (2019) recently demonstrated that using low intrapulmonary PFOB volumes, measured as the amount of breathable liquid in the lung at the end expiratory pause (EELV), during a short 30-minute course of TLV was associated with less pulmonary inflammation. The main objective of the present study was to compare lung inflammation induced by low vs. high tidal-volume TLV in a healthy term lamb model without extracorporeal carbon dioxide removal, while maintaining sustainable gas exchanges for a prolonged period. We hypothesize that, like conventional mechanical gas ventilation, high tidal volume (H-VT) will result in greater lung inflammation and impede complete post-TLV weaning compared to low tidal volume (L-VT).
Materials and Methods
Lamb Instrumentation
A total of 21 full-term healthy newborn male lambs aged 3.3 ± 1.2 days were brought to the research facility from a local breeder. The study was approved by the animal research ethics board of the Université de Sherbrooke (protocol #417-17) and performed in accordance with the Canadian Council on Animal Care guidelines.
Sixteen lambs subjected to TLV were premedicated with intramuscular ketamine (10 mg/kg) prior to percutaneous cannulation of the left jugular vein. They were then intubated with a 4.5 mm cuffed endotracheal tube and placed in the supine position. Gas ventilation was initiated (Servo-i, Maquet Critical Care, Solna, Sweden) in the pressure-controlled mode with a respiratory rate of 40 cycles/min and a positive end-expiratory pressure of 5 cmH2O. The inspiratory pressure was manually adjusted to deliver a tidal volume of approximately 7 mL/kg. Hemoglobin oxygen saturation was continuously recorded by pulse oximetry (Radical, Masimo, Irvine, CA, United States) with a probe at the base of the tail. Lambs were maintained under general anesthesia throughout the TLV phase of the experiment using an intravenous infusion of propofol 120 μg/kg/min and ketamine 1 mg/kg/h, together with a 10% dextrose solution at a rate of 5 mL/kg/h. A catheter was inserted into the left carotid artery under sterile conditions to allow for continuous monitoring of systemic arterial pressure and blood-gas measurements. Lambs were then randomized to either the L-VT (high respiratory rate) or H-VT (low respiratory rate) TLV group.
Experimental Protocol
Total Liquid Ventilation
After surgical instrumentation under gas ventilation, the lambs were paralyzed (rocuronium bromide 0.2 mg/kg IV) and disconnected from the gas ventilator for 5 s to allow for lung deflation prior to TLV initiation. The lungs were then filled with 20 mL/kg of pre-oxygenated perfluorooctyl bromide (Exfluor, Round Rock, TX, United States) at 39°C over 18 s, using the INOLIVENT-6 liquid ventilator prototype as previously described (Sage et al., 2018a). Volume-controlled, pressure-limited, and time-cycled TLV was then initiated with a respiratory rate of 10 cycles/min and a tidal volume of 10 mL/kg (L-VT group) or a respiratory rate of 5 cycles/min and a tidal volume of 20 mL/kg (H-VT group). The inspiratory to expiratory time ratio was set at 1:2. The end-expiratory lung volume was progressively increased to 30 mL/kg over the first 10 respiratory cycles to reach a lung volume similar to functional residual capacity (25–30 mL/kg) during conventional mechanical ventilation in this model (Overfield et al., 2001; Wolfson and Shaffer, 2004). Given the complexity of changing respiratory rate with our current liquid ventilator algorithms, respiratory rate was kept constant throughout TLV and tidal volumes were adjusted to maintain targeted blood gases using a permissive hypercapnia approach (pH ≥ 7.20 and PaCO2 from 50 to 65 mmHg (Ryu et al., 2012). Oxygen concentration in the PFC was adjusted in order to maintain pulse oximetry values above 90%. After 4 h of total liquid ventilation, the lambs were placed in the prone position. Neostigmine (0.05 mg/kg) and glycopyrrolate (0.01 mg/kg) were administered to reverse rocuronium effect, and the anesthetic infusion was stopped, allowing the lambs to recover from anesthesia. TLV was halted after an expiration, lambs were disconnected from the liquid ventilator and placed in decline position to favor liquid draining into a bottle by gravity. When no more PFOB was coming out of the endotracheal tube (≈10 s), the gas ventilator was connected.
Weaning From Tidal Liquid Ventilation
Figure 1 shows the standardized weaning protocol used. After TLV, the animals were weaned to conventional gas ventilation in the pressure-controlled mode. The initial FiO2 and ventilatory parameters (peak inspiratory pressure of 15 cmH2O, PEEP of 6 cmH2O, respiratory rate of 40/min) were adjusted throughout the weaning protocol to maintain tidal volume at 6–8 mL/kg, SaO2 > 90%, and PaO2 between 60 and 80 mmHg, PaCO2 between 50 and 65 mmHg, and pH > 7.20. A nasal catheter was inserted into the esophagus to record the electrical activity of the diaphragm (Edi) (Maquet Critical Care, Solna, Sweden) and a nasal mask custom-made for lambs was installed. The lambs were then extubated as soon as they had spontaneous, regular breathing movements with peak inspiratory pressure less than or equal to 12 cmH2O to generate adequate tidal volumes and placed under nasal neurally adjusted ventilatory assist (NAVA). The PEEP, FiO2 and NAVA levels were adjusted to maintain SaO2 above 90% and to minimize the clinically assessed work of breathing. The NAVA level was thereafter decreased until the lambs could be switched to nasal CPAP at 5 cmH2O for at least 15 min. The mask was then removed, and supplemental oxygen was delivered via nasal cannulae to maintain SaO2 >90% and PaO2 between 60 and 80 mmHg. If PaCO2 was above 65 mmHg at any point, the nasal CPAP at 5 cmH2O was reinstalled. If needed, nasal pressure support ventilation and even endotracheal pressure-controlled ventilation could have been instituted again based on the same criteria. The lambs were euthanized 4 h post-TLV with an IV injection of 90 mg/kg pentobarbital.
Figure 1. Standardized weaning protocol after total liquid ventilation. PaCO2: partial pressure of carbon dioxide in arterial blood, PaO2: partial pressure of oxygen in arterial blood, SaO2: hemoglobin saturation in oxygen (pulse oximetry), PEEP: positive end-expiratory pressure, VT: tidal volume. CPAP: continuous positive airway pressure, NAVA: neurally adjusted ventilatory assist.
Control Group
Five lambs were used for histological and RNA transcript analysis without undergoing mechanical ventilation or any other experimental procedure. They were euthanized after peripheral venous cannulation and lung samples were collected.
Histological Score for Lung Inflammation
Samples (2 × 2 × 0.5 cm) from the same region of the peripheral anterior lower lobe and posterior lower lobe of the right lung were collected, fixed, and conserved in 10% formaldehyde. The tissues were then embedded in paraffin, from which 5 μm sections were prepared for hematoxylin and eosin staining. A blinded pathologist examined the slides for the experimental and control groups (SM) using a histological score of lung inflammation previously developed for the newborn lamb (Hillman et al., 2010). Four key components (septation thickness, hemorrhage, inflammatory-cell infiltration, and epithelial sloughing) were assessed, each on a scale of 0 to 2 (total inflammation score of 8).
Gene Transcription Analysis
Samples were collected in the left lung from the same regions as for the histological samples from the right lung and assayed for IL1B, IL6, and TNF gene transcription levels. As the focus of our laboratory will be the use of TLV to prevent bronchopulmonary dysplasia in the years to come, we have oriented the present exploratory analyses as such. We measured the transcription level of 12 other genes selected based on their implication in VILI and bronchopulmonary dysplasia development: AGER, CCL4, CSF2, CXCL1, CXCL8, ICAM1, IL1A, IL1R1, IL33, NFKB1, NFKB2, and TNRRSF1A (Moldoveanu et al., 2008; Huusko et al., 2014). To this end, total RNA extraction was performed on tissue samples preserved in TRIzol (Invitrogen, Carlsbad, CA, United States) at −80°C using TissueLyser (Qiagen, Toronto, ON, Canada). Chloroform was added following the manufacturer’s protocol. The aqueous layer was recovered, then mixed with one volume of 70% ethanol and applied directly to a RNeasy Mini Kit column (Qiagen). DNAse treatment and total RNA recovery were performed as per the manufacturer’s protocol. RNA quality and presence of contaminating genomic DNA were verified as previously described (Brosseau et al., 2010). RNA integrity was assessed with an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, United States). Reverse transcription was performed on 1.1 μg total RNA with Transcriptor reverse transcriptase, Random Hexamer Primers, deoxynucleoside triphosphate kit (Roche Diagnostics, Basel, Switzerland), and 10 units of RNAseOUT (Invitrogen) according to the manufacturer’s protocol in a total volume of 10 μL. Individual forward and reverse primers were resuspended at 20 to 100 μM stock solution in Tris-EDTA buffer (Integrated DNA Technologies, Inc., Coralville, IA, United States). They were thereafter diluted as a primer pair to 1 μM in RNase DNase-free water (Integrated DNA Technologies, Inc.). Quantitative PCR (qPCR) was performed in 384 well plates in a CFX-384 thermocycler (Bio-Rad, Hercules, CA, United States) with 5 μL of 2X iTaq Universal SYBR Green SuperMix (Bio-Rad), 10 ng (3 μL) cDNA, and 200 nM (2 μL) final primer pair solutions. The following cycling conditions were used: 3 min at 95°C, followed by 50 cycles (15 s at 95°C, 30 s at 60°C, and 30 s at 72°C). Relative expression levels were calculated with the qBase framework (Hellemans et al., 2007) and the housekeeping genes glyceraldehyde 3-phosphate dehydrogenase (GAPDH), Beta-2 microglobulin (B2M), ribosomal protein L13a (RPL13A), and TATA-binding protein (TBP) for lamb cDNA. Primer design and validation were evaluated as previously described (Brosseau et al., 2010). In every qPCR run, a no-template control was performed for each primer pair and the results were consistently negative. All primer sequences are available in Supplementary Table S1.
Data Analysis
All RNA levels were normalized to a group of housekeeping-gene expression. RNA transcription results were presented in log2 (fold change between L-VT and H-VT lambs). Principal component analysis and hierarchical clustering were performed, as exploratory analyses, with R software1. Statistical analyses were performed with SPSS 19 (IBM, Armonk, NY, United States). Gene expression data in each group were reported as mean ± SD (unless specified otherwise), and the groups compared with a 2-sided non-paired t-test. Histological-score results were reported as median (Q1; Q3) and compared with the Kruskal–Wallis test followed by the Mann–Whitney U test. A value of p < 0.05 was deemed to be statistically significant.
Results
The lambs from all groups were similar in age and weight (Table 1). All were successfully weaned to spontaneous breathing at 4 h post-TLV, except for one lamb in the H-VT group that remained intubated because of anesthesia-related weak respiratory efforts; the lamb’s extubation time was not considered in the analysis.
Table 1. Lamb characteristics at baseline and blood gases.
Table 2 provides the ventilatory parameters. They were identical in the H-VT and L-VT groups at all time points, except for tidal volume and respiratory rate. Tidal volume nevertheless needed to be raised in the L-VT group from 10 mL/kg (at TLV initiation) to 14 ± 2 mL/kg within the first 2 h of TLV to achieve the targeted PaCO2/pH values; the tidal volume remained stable thereafter. Minute ventilation in the L-VT group was thus significantly higher than in the H-VT group in order to achieve similar PaCO2 and pH values, as expected. The end-expiratory lung volume of PFC was closely monitored and kept minimal in both groups, while avoiding end-expiratory tracheal collapse (35 ± 11 mL/kg).
Table 2. Ventilatory parameters.
Lambs in the L-VT group were extubated after 72 ± 26 min, while those in the H-VT group were extubated after 63 ± 25 min (p = 0.5). At 2 h post-TLV, four lambs in the L-VT group and five lambs in the H-VT group were under nasal NAVA. At 4 h post-TLV, one lamb in the L-VT and two lambs in the H-VT group remained under NAVA. The majority (seven in the L-VT, five in the H-VT group) of the lambs were only on supplemental oxygen. FiO2 requirements at the end of the experiment were similar in the L-VT and H-VT groups at 27 ± 6 and 33 ± 7%, respectively, (p = 0.3). Blood gases were similar in both groups at all time points (Table 1).
Histological Inflammation Score
Both TLV groups had significantly higher inflammation scores compared to the control non-ventilated group, mainly due to increased septation thickness (Figure 2). The histological score also tended to indicate higher inflammation in the L-VT group compared to the H-VT group in the anterior lung region, although this difference did not reach statistical significance (Figure 2).
Figure 2. Inflammation histological score in lambs subjected to TLV with L-VT (n = 8), H-VT (n = 8), and CTL (n = 5). The total score out of eight represents the sum of the four categories scored over two (septation thickness, hemorrhage, cell infiltration, and epithelial sloughing). A: anterior region of the lung; P: posterior lung region. A higher score indicates more inflammation. Each point represents a lamb and each line represents the median and the first and third quartiles (Q1; Q3). *: statistically significant difference (p < 0.05).
Inflammatory-Gene Expression
Lung tissue IL1B mRNA was significantly more expressed in the L-VT group than the H-VT group in both the anterior and posterior lung regions (Figures 3A,B). The expression of both IL6 and TNF transcripts was greater in the L-VT group, but the difference was not significant (Figures 3A,B). When compared to the control lambs, the L-VT lambs exhibited a significantly higher expression of IL1B (p = 0.01), IL6 (p = 0.006), and TNF (p = 0.04) in the anterior region, while IL1B and IL6 were significantly upregulated in the posterior region (p < 0.001; p = 0.03, respectively) (Figures 3C,D). On the other hand, only TNF was upregulated in the posterior region (p = 0.002) in the H-VT group compared to the control lambs (Figure 3D).
Figure 3. Expression of the IL1B, IL6 and TNF genes illustrated by log2 fold changes in L-VT relative to H-VT lambs in the anterior (A) and posterior (B) lung regions, and fold changes in gene expression of L-VT, H-VT and unventilated lambs relative to housekeeping genes using qBase relative quantification framework method [glyceraldehyde-3-phosphate dehydrogenase (GAPDH), beta-2-microglobulin (B2M), ribosomal protein L13a (RPL13A), and TATA-binding protein (TBP)] in the anterior (C) and posterior (D) lung regions. Each dot represents a lamb, with mean ± standard deviation. *p < 0.05. See legend of Table 1 for other abbreviations.
A principal component analysis showed clear separation between unventilated, H-VT and L-VT ventilated lamb groups, with consistent gene expression profiles between the anterior and posterior lung regions (Figure 4A). Unsupervised clustering of individual gene expression revealed four main findings: (1) pronounced inflammatory transcriptional changes in both groups of liquid ventilated lungs compared to unventilated lambs; (2) three main gene expression clusters, with an increased pro-inflammatory gene expression profile in L-VT lungs, but not in the H-VT or unventilated lambs; (3) a distinct transcriptional response in lung tissues from L-VT compared to H-VT ventilated lambs; and (4) greater ventilation-associated changes in the posterior, compared to the anterior lung regions (Figure 4B). Altogether, our gene expression analysis indicates that H-VT induces qualitatively distinct transcriptional changes compared to L-VT.
Figure 4. Inflammatory gene expression analyses. (A) Principal component analysis of gene expression data in both anterior (Light colored dots) and posterior (Dark colored dots) lung samples from unventilated, L-VT (TLVLow) and H-VT (TLVHigh) lambs. Distance between dots approximates gene-expression pattern differences among experimental conditions/lung regions; the closer the dots, the closer the expression levels for each linearly uncorrelated gene group (i.e., principal component). (B) Heatmap and unsupervised clustering of selected gene expression in the anterior and posterior lung regions in L-VT, H-VT or unventilated lambs (CTL). Colors indicate expression z-score after normalization on housekeeping genes.
Discussion
Given TLV’s potential to reduce VILI in neonate chronic lung disease, our study provides important information about how to optimize its potential using different levels of tidal volume. The main finding of our study is that TLV with low tidal volume was associated with more inflammatory-gene expression than TLV using high tidal volume. In addition, our results show the possibility to consistently achieve weaning from TLV to spontaneous breathing within 2 h post-TLV.
The overall transcription profile suggests that a TLV strategy using an H-VT may induce less lung inflammation than L-VT. Interestingly, observations were consistent between the anterior and posterior regions of the lungs implying that our results are no simply due to a sampling bias. In addition to IL1B, genes encoding for important proteins of the inflammatory cascade, such as CXCL8, TNF, CSF2, and NFKB1, were also more expressed in the L-VT group. Nevertheless, IL33 was upregulated in the H-VT group compared to the L-VT group. This cytokine can be pro-inflammatory or anti-inflammatory depending on its environment (Zhao and Zhao, 2015). In this case, it remains unclear whether the level of IL33 transcription induced less or more lung inflammation in the H-VT group. Overall, although heterogeneous, the gene-expression profile we found in the L-VT lambs was more consistent with VILI literature data than the profile observed in the H-VT lambs. Moreover, the gene expression profile in H-VT lambs suggests that higher tidal volume may be more lung protective compared to L-VT in TLV. It is interesting to note opposing inflammatory cytokine gene expression profiles between the L-Vt and H-Vt experimental groups. Indeed, instead of having only up regulation in one group and no change in the other, we had genes that were overexpressed in one group while the same genes were down regulated in the other groups. This effect was reproducible in two different regions and therefore is unlikely to represent a measure artifact. This intriguing finding suggests that different tidal volumes trigger qualitatively distinct inflammatory responses in the lungs. These findings definitely require a closer look in a future study, in order to understand the mechanisms involved in these qualitative differences and their clinical implication in terms of choosing the right lung protective ventilatory strategy.
Our results contradict our hypothesis favoring the use of L-VT to prevent lung inflammation. Importantly, L-VT lambs required 50% higher minute ventilation–i.e., a higher respiratory rate–than H-VT lambs in order to maintain PaCO2 and pH values within the targeted limits. This was likely due to the known low gas diffusion velocity in perfluorochemicals compared to a gaseous medium (Costantino and Fiore, 2001). This effect could be exacerbated in the ovine model, known for its larger anatomic dead space (Albertine, 2013). It should be noted that our strategy aimed at strictly maintaining the end-expiratory liquid volume at the lowest possible level might have mitigated lung over-distension/volutrauma, a crucial VILI mechanism during TLV, as shown by Kohlhauer et al. (2019). The benefits of limiting the end-inspiratory lung volume might, however, be offset by the disadvantages of the higher respiratory rate required in L-VT lambs. Indeed, a lower respiratory rate has been associated with less alveolar shear stress and lung inflammation during conventional gas ventilation (Hotchkiss et al., 2000; Rich et al., 2000; Vaporidi et al., 2008; Chen et al., 2015).
Very few authors have demonstrated the possibility of achieving complete weaning a neonatal animal after TLV (Jackson et al., 1994; Stavis et al., 1998). Jackson et al. (1994) were able to wean two near-term healthy non-human primates after 5 h of conventional mechanical ventilation following 3 h of TLV, while Stavis et al. (1998) reported weaning three full-term lambs from 4 h of TLV to spontaneous breathing after 16 h. The present study confirms that it is possible to consistently wean healthy full-term lambs from 4 h of TLV to spontaneous breathing within 2 h.
Although the L-VT lambs exhibited a more inflammatory profile, no clinically relevant difference was observed during the weaning process, including oxygen requirement, respiratory rate, and arterial-blood gases. Most animals presented mild respiratory distress during the first 30 to 90 min following extubation. Most required oxygen, and some required continuous positive airway pressure for 4 h after TLV. PaCO2 levels and respiratory rate were, however, normal at that point. As anticipated, residual liquid PFC likely affects respiratory mechanics and gas exchange during the first few hours following TLV, while it is being cleared from the lung through evaporation and coughing (Bendel-Stenzel et al., 1998). This is consistent with a past study showing that recovery from TLV in near-term non-human primates was associated with the need for supplemental oxygen in the first 24 to 72 h post-TLV (Jackson et al., 1994). A study is currently underway in our lab to assess respiratory mechanics during weaning after TLV.
Limitations of the Study
One limitation of our study is its small sample size, which limits power and thus, our ability to demonstrate further differences between the groups. In addition, we decided to proceed to weaning from TLV in order to identify clinically relevant differences post-TLV between the two groups as well as to assess the feasibility of such weaning. The transition from TLV back to gas ventilation might come at the cost of inducing inflammation. Since lambs from both groups were, however, consistently weaned according to the same standardized protocol, we do not believe this introduced a bias in our study. The only animal that could not be extubated because of profound sedation was in the H-VT group; this would have affected the results in favor of the L-VT group, contrarily to our results.
As with all animal models, anatomic and gene expression differences with the human newborn are possible. Clinical trials will be needed before this technology can be used as a standard treatment in the neonatal intensive care units. Gene expression levels were assessed in this study. The authors acknowledge that protein levels as well as functional respiratory assessment after a longer course of TLV would be beneficial in confirming our findings. Nevertheless, this study represents a first step and brings new knowledge in delineating the optimal TLV strategy in newborns.
Conclusion
Our results suggest that a TLV strategy using high tidal volume and low respiratory rate is associated with reduced lung inflammation compared to low tidal volume and high respiratory rate. TLV modality did not, however, affect weaning to spontaneous breathing, which was achieved within a 4-hour time window. The ability to consistently wean neonatal lambs is a major milestone in TLV research and paves the way for an eventual clinical trial.
Data Availability Statement
The original contributions presented in the study are publicly available. This data can be found here: https://rnomics.med.usherbrooke.ca/palace?purl=/pcrreactiongroup/list/475.
Ethics Statement
The animal study was reviewed and approved by Animal Research Ethics Board of the Université de Sherbrooke (protocol #417-17).
Author Contributions
ÉF-P, PM, J-PP, BC, and MS contributed to the original idea and design of the study. MS, WS, SN, and ÉF-P contributed to the experiments. MS, WS, SN, CMi, ÉF-P, BC, SM, PL, and J-PP contributed to the data collection, analysis, and interpretation. MS, WS, CMo, CMi, SN, SM, PL, BC, J-PP, and ÉF-P all contributed to the manuscript.
Funding
This work was funded by a grant from the Respiratory Health Network of the Fonds de recherche du Québec – Santé, the Centre Hospitalier Universitaire de Sherbrooke, and the Faculté de Médecine et des Sciences de la Santé de l’Université de Sherbrooke.
Conflict of Interest
PM and J-PP are co-inventors of patents related to the ventilator prototype used in this study (Apparatus for conducting total liquid ventilation with control of residual volume and ventilation cycle profile, US 7,726,311 B2, EP 1 424 090 B1, CA 2,451,261 and Indirect measurement in a total liquid ventilation system, PCT/CA2014/205548 US 2016/0271348 A1).
The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Acknowledgments
The authors would like to thank the RNA transcript platform of the Université de Sherbrooke for the design of the primers and the quantification of ARN transcript. The experiments were performed at the Université de Sherbrooke animal facility.
Supplementary Material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphys.2020.00603/full#supplementary-material
Abbreviations
CPAP, continuous positive airway pressure; CTL, control lambs without ventilation; EELV, end-expiratory liquid volume; EEPP, end-expiratory pause pressure; EIPP, end-inspiratory pause pressure; FgO2, fraction of oxygen bubbled in PFOB; FiO2, fraction of inspired oxygen; HCO3–, arterial concentration of bicarbonate ions; H-V T, total liquid ventilation with high tidal volume; IL1B, interleukin-1 gene; IL6, interleukin-6 gene; IL8, interleukin-8 gene; L-V T, total liquid ventilation with low tidal volume; NAVA, neurally adjusted ventilatory assist; PaCO2, partial pressure of carbon dioxide in arterial blood; PaO2, partial pressure of oxygen in arterial blood; PEEP, positive end-expiratory pressure; PFC, perfluorochemical; PFOB, perfluorooctyl bromide; PIP, peak inspiratory pressure; SaO2, hemoglobin saturation in oxygen (pulse oximetry); TLV, total liquid ventilation; TNF, tumor necrosis factor- α gene; VE, minute ventilation; VILI, ventilator-induced lung injury; VT, tidal volume.
Footnotes
References
Albertine, K. H. (2013). Progress in understanding the pathogenesis of BPD using the baboon and sheep models. Semin. Perinatol. 37, 60–68. doi: 10.1053/j.semperi.2013.01.001
Avoine, O., Bosse, D., Beaudry, B., Beaulieu, A., Albadine, R., Praud, J. P., et al. (2011). Total liquid ventilation efficacy in an ovine model of severe meconium aspiration syndrome. Crit. Care Med. 39, 1097–1103. doi: 10.1097/CCM.0b013e31820ead1a
Bagnoli, P., Tredici, S., Seetharamaiah, R., Brant, D. O., Hewell, L. A., Johnson, K., et al. (2007). Effect of repeated induced airway collapse during total liquid ventilation. Asaio J. 53, 549–555. doi: 10.1097/MAT.0b013e318148449d
Bendel-Stenzel, E. M., Mrozek, J. D., Bing, D. R., Meyers, P. A., Connett, J. E., and Mammel, M. C. (1998). Dynamics of spontaneous breathing during patient-triggered partial liquid ventilation. Pediatr. Pulmonol. 26, 319–325. doi: 10.1002/(sici)1099-0496(199811)26:5<319::aid-ppul3>3.0.co;2-v
Brosseau, J. P., Lucier, J. F., Lapointe, E., Durand, M., Gendron, D., Gervais-Bird, J., et al. (2010). High-throughput quantification of splicing isoforms. Rna 16, 442–449. doi: 10.1261/rna.1877010
Chen, Z. L., Song, Y. L., Hu, Z. Y., Zhang, S., and Chen, Y. Z. (2015). An estimation of mechanical stress on alveolar walls during repetitive alveolar reopening and closure. J. Appl. Physiol. 119, 190–201. doi: 10.1152/japplphysiol.00112.2015
Ciuffini, F., Robertson, C. F., and Tingay, D. G. (2018). How best to capture the respiratory consequences of prematurity? Eur. Respir. Rev. 27:170108. doi: 10.1183/16000617.0108-2017
Costantino, M. L., Bagnoli, P., Dini, G., Fiore, G. B., Soncini, M., Corno, C., et al. (2004). A numerical and experimental study of compliance and collapsibility of preterm lamb tracheae. J. Biomech. 37, 1837–1847. doi: 10.1016/j.jbiomech.2004.02.035
Costantino, M. L., and Fiore, G. B. (2001). A model of neonatal tidal liquid ventilation mechanics. Med. Eng. Phys. 23, 457–471.
Davidson, L. M., and Berkelhamer, S. K. (2017). Bronchopulmonary dysplasia: chronic lung disease of infancy and long-term pulmonary outcomes. J. Clin. Med. 6:E4. doi: 10.3390/jcm6010004
Farrell, O., Perkins, E. J., Black, D., Miedema, M., Paul, J. D., Pereira-Fantini, P. M., et al. (2018). Volume guaranteed? Accuracy of a volume-targeted ventilation mode in infants. Arch. Dis. Child Fetal. Neonatal. Ed. 103, F120–F125. doi: 10.1136/archdischild-2017-312640
Hellemans, J., Mortier, G., De Paepe, A., Speleman, F., and Vandesompele, J. (2007). qBase relative quantification framework and software for management and automated analysis of real-time quantitative PCR data. Genome Biol. 8:R19. doi: 10.1186/gb-2007-8-2-r19
Hillman, N. H., Kallapur, S. G., Pillow, J. J., Moss, T. J., Polglase, G. R., Nitsos, I., et al. (2010). Airway injury from initiating ventilation in preterm sheep. Pediatr. Res. 67, 60–65. doi: 10.1203/PDR.0b013e3181c1b09e
Hotchkiss, JR Jr., Blanch, L., Murias, G., Adams, A. B., Olson, D. A., Wangensteen, O. D., et al. (2000). Effects of decreased respiratory frequency on ventilator-induced lung injury. Am. J. Respir. Crit. Care Med. 161(2 Pt 1), 463–468. doi: 10.1164/ajrccm.161.2.9811008
Huusko, J. M., Karjalainen, M. K., Mahlman, M., Haataja, R., Kari, M. A., Andersson, S., et al. (2014). A study of genes encoding cytokines (IL6, IL10, TNF), cytokine receptors (IL6R, IL6ST), and glucocorticoid receptor (NR3C1) and susceptibility to bronchopulmonary dysplasia. BMC Med. Genet. 15:120. doi: 10.1186/s12881-014-0120-7
Jackson, J. C., Standaert, T. A., Truog, W. E., and Hodson, W. A. (1994). Full-tidal liquid ventilation with perfluorocarbon for prevention of lung injury in newborn non-human primates. Artif. Cells Blood Substit. Immobil. Biotechnol. 22, 1121–1132. doi: 10.3109/10731199409138807
Jiang, L., Feng, H., Chen, X., Liang, K., and Ni, C. (2016). Low tidal volume reduces lung inflammation induced by liquid ventilation in piglets with severe lung injury. Artif. Organ. 41, 440–445. doi: 10.1111/aor.12784
Koen, P. A., Wolfson, M. R., and Shaffer, T. H. (1988). Fluorocarbon ventilation: maximal expiratory flows and CO2 elimination. Pediatr. Res. 24, 291–296. doi: 10.1203/00006450-198809000-00003
Kohlhauer, M., Boissady, E., Lidouren, F., de Rochefort, L., Nadeau, M., Rambaud, J., et al. (2019). A new paradigm for lung-conservative total liquid ventilation. EBioMedicine 52:102365. doi: 10.1016/j.ebiom.2019.08.026
Liu, Z., Duarte, R. V., Bayliss, S., Bramley, G., and Cummins, C. (2016). Adverse effects of extracorporeal carbon dioxide removal (ECCO2R) for acute respiratory failure: a systematic review protocol. Syst. Rev. 5, 98–98. doi: 10.1186/s13643-016-0270-0
Moldoveanu, B., Otmishi, P., Jani, P., Walker, J., Sarmiento, X., Guardiola, J., et al. (2008). Inflammatory mechanisms in the lung. J. Inflammati Res. 2, 1–11.
Nadeau, M., Denaclara, J.-Y., Tissier, R., Walti, H., and Micheau, P. (2018). Patient-specific optimal cooling power command for hypothermia induction by liquid ventilation. Control Eng. Pract. 77, 109–117. doi: 10.1016/j.conengprac.2018.05.007
Overfield, D. M., Bennett, S. H., Goetzman, B. W., Milstein, J. M., and Moon-Grady, A. J. (2001). Hemodynamic effects of positive end-expiratory pressure during partial liquid ventilation in newborn lambs. J. Pediatr. Surg. 36, 1327–1332. doi: 10.1053/jpsu.2001.26360
Pohlmann, J. R., Brant, D. O., Daul, M. A., Reoma, J. L., Kim, A. C., Osterholzer, K. R., et al. (2011). Total liquid ventilation provides superior respiratory support to conventional mechanical ventilation in a large animal model of severe respiratory failure. ASAIO J. 57, 1–8. doi: 10.1097/MAT.0b013e3182018a9f
Rich, P. B., Reickert, C. A., Sawada, S., Awad, S. S., Lynch, W. R., Johnson, K. J., et al. (2000). Effect of rate and inspiratory flow on ventilator-induced lung injury. J. Trauma 49, 903–911. doi: 10.1097/00005373-200011000-00019
Robert, R., Micheau, P., Avoine, O., Beaudry, B., Beaulieu, A., and Walti, H. (2010). A regulator for pressure-controlled total-liquid ventilation. IEEE Trans. Biomed. Eng. 57, 2267–2276. doi: 10.1109/tbme.2009.2031096
Robert, R., Micheau, P., and Walti, H. (2007). A supervisor for volume-controlled tidal liquid ventilator using independent piston pumps. Biomed. Signal Process. Control 2, 267–274. doi: 10.1016/j.bspc.2007.07.010
Robert, R., Micheau, P., and Walti, H. (2009). Optimal expiratory volume profile in tidal liquid ventilation under steady state conditions, based on a symmetrical lung model. Asaio J. 55, 63–72. doi: 10.1097/MAT.0b013e3181911821
Ryu, J., Haddad, G., and Carlo, W. A. (2012). Clinical effectiveness and safety of permissive hypercapnia. Clin. Perinatol. 39, 603–612. doi: 10.1016/j.clp.2012.06.001
Sage, M., Nadeau, M., Forand-Choiniere, C., Mousseau, J., Vandamme, J., Berger, C., et al. (2018a). Assessing the impacts of total liquid ventilation on left ventricular diastolic function in a model of neonatal respiratory distress syndrome. PLoS One 13:e0191885. doi: 10.1371/journal.pone.0191885
Sage, M., Stowe, S., Adler, A., Forand-Choiniere, C., Nadeau, M., Berger, C., et al. (2018b). Perflubron distribution during transition from gas to total liquid ventilation. Front. Physiol. 9:1723. doi: 10.3389/fphys.2018.01723
Shaffer, T. H., Wolfson, M. R., and Greenspan, J. S. (1999). Liquid ventilation: current status. Pediatr. Rev. 20, e134–e142. doi: 10.1542/pir.20-12-e134
Snoek, K. G., Houmes, R. J., and Tibboel, D. (2014). Liquid ventilation in congenital diaphragmatic hernia: back on stage? Pediatr. Crit. Care Med. 15, 914–915. doi: 10.1097/pcc.0000000000000284
Stavis, R. L., Cox, C. A., Wolfson, M. R., Cullen, A. B., Roache, R., Hipp, S., et al. (1998). Tidal liquid ventilation (TLV) transition to spontaneous breathing(SB): 24 hour follow-up of physiologic and radiographic correlates 1748. Pediatr. Res. 43, (Suppl. 4), 298–298. doi: 10.1203/00006450-199804001-01770
Thomassen, M. J., Buhrow, L. T., and Wiedemann, H. P. (1997). Perflubron decreases inflammatory cytokine production by human alveolar macrophages. Crit. Care Med. 25, 2045–2047. doi: 10.1097/00003246-199712000-00023
Tooley, R., Hirschl, R. B., Parent, A., and Bartlett, R. H. (1996). Total liquid ventilation with perfluorocarbons increases pulmonary end-expiratory volume and compliance in the setting of lung atelectasis. Crit. Care Med. 24, 268–273. doi: 10.1097/00003246-199602000-00015
Vaporidi, K., Voloudakis, G., Priniannakis, G., Kondili, E., Koutsopoulos, A., Tsatsanis, C., et al. (2008). Effects of respiratory rate on ventilator-induced lung injury at a constant PaCO2 in a mouse model of normal lung. Crit. Care Med. 36, 1277–1283. doi: 10.1097/CCM.0b013e318169f30e
Wolfson, M. R., Hirschl, R. B., Jackson, J. C., Gauvin, F., Foley, D. S., Lamm, W. J., et al. (2008). Multicenter comparative study of conventional mechanical gas ventilation to tidal liquid ventilation in oleic acid injured sheep. Asaio J. 54, 256–269. doi: 10.1097/MAT.0b013e318168fef0
Wolfson, M. R., and Shaffer, T. H. (2004). Liquid ventilation: an adjunct for respiratory management. Paediatr. Anaesth. 14, 15–23. doi: 10.1046/j.1460-9592.2003.01206.x
Woods, C. M., Neslund, G., Kornbrust, E., and Flaim, S. F. (2000). Perflubron attenuates neutrophil adhesion to activated endothelial cells in vitro. Am. J. Physiol. Lung. Cell Mol. Physiol. 278, L1008–L1017.
Keywords: liquid ventilation, ventilator-induced lung injury, tidal volume, newborn lamb, transcriptome
Citation: Sage M, See W, Nault S, Morin C, Michalski C, Chabot B, Marouan S, Lavoie PM, Micheau P, Praud J-P and Fortin-Pellerin É (2020) Effect of Low Versus High Tidal-Volume Total Liquid Ventilation on Pulmonary Inflammation. Front. Physiol. 11:603. doi: 10.3389/fphys.2020.00603
Received: 06 March 2020; Accepted: 14 May 2020;
Published: 18 June 2020.
Edited by:
Yu Ru Kou, National Yang-Ming University, TaiwanReviewed by:
Noah H. Hillman, Saint Louis University, United StatesBeth J. Allison, Hudson Institute of Medical Research, Australia
Copyright © 2020 Sage, See, Nault, Morin, Michalski, Chabot, Marouan, Lavoie, Micheau, Praud and Fortin-Pellerin. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Étienne Fortin-Pellerin, etienne.fortin-pellerin@usherbrooke.ca