High frequency oscillatory ventilation and prone positioning in a porcine model of lavage-induced acute lung injury

Background This animal study was conducted to assess the combined effects of high frequency oscillatory ventilation (HFOV) and prone positioning on pulmonary gas exchange and hemodynamics. Methods Saline lung lavage was performed in 14 healthy pigs (54 ± 3.1 kg, mean ± SD) until the arterial oxygen partial pressure (PaO2) decreased to 55 ± 7 mmHg. The animals were ventilated in the pressure controlled mode (PCV) with a positive endexpiratory pressure (PEEP) of 5 cmH2O and a tidal volume (VT) of 6 ml/kg body weight. After a stabilisation period of 60 minutes, the animals were randomly assigned to 2 groups. Group 1: HFOV in supine position; group 2: HFOV in prone position. After evaluation of prone positioning in group 2, the mean airway pressure (Pmean) was increased by 3 cmH2O from 16 to 34 cmH2O every 20 minutes in both groups accompanied by measurements of respiratory and hemodynamic variables. Finally all animals were ventilated supine with PCV, PEEP = 5 cm H2O, VT = 6 ml/kg. Results Combination of HFOV with prone positioning improves oxygenation and results in normalisation of cardiac output and considerable reduction of pulmonary shunt fraction at a significant (p < 0.05) lower Pmean than HFOV and supine positioning. Conclusion If ventilator induced lung injury is ameliorated by a lower Pmean, a combined treatment approach using HFOV and prone positioning might result in further lung protection.


kg), sufficient PEEP-level and limitation of the plateau inspiratory pressure to 35 cm H 2 O [1-3].
Although it is known, that the degree of hypoxemia is inconclusive to predict mortality [4], the early response of the PaO 2 /FIO 2 -ratio to therapeutic interventions might be an indicator for an increased survival rate in ARDS [5,6]. This calls for the most rapid amelioration of hypoxemia with a mono-or multimodal organ protective treatment approach.
High frequency oscillatory ventilation (HFOV) with its constant mean airway pressure (P mean ) with superimposed small tidal volumes and active in-and expiration at a high respiratory frequency might be the ideal lung protective ventilatory strategy [7]. In a multicenter randomized controlled trial investigating the effectiveness of HFOV, the significant early improvement of the PaO 2 / FIO 2 -ratio in the HFOV-group was associated with a tendency towards a reduced 30-day mortality compared with the conventional ventilation group. The PaO 2 /FIO 2 -ratio was the most significant predictor of survival independent of the selected ventilator strategy [8].
Prone positioning was shown to increase the PaO 2 in 70-80% of patients with ARDS and to improve alveolar ventilation without influencing the 28-day mortality [9,10]. If PaCO 2 -reduction was achievable with prone positioning, 28-day mortality in ARDS patients was significantly reduced [11]. Prone positioning and application of PEEP were shown to have an additive effect on oxygenation [12]. However, prone positioning is a potentially dangerous manoeuvre with acute and long term complications such as tracheal tube dislocation, and pressure sores [13].
Combination of different treatments are used in desperation for salvage therapy in patients with ARDS [14]. A recently published study in 39 medical ARDS-patients randomized to conventional lung protective ventilation and HFOV showed comparable increases of the PaO 2 / FIO 2 -ratio after prone positioning. An additive effect of prone positioning and HFOV could not be demonstrated [15].
The objective of our study was to evaluate the effects of prone positioning on gas-exchange, hemodynamics and respiratory parameters in HFOV-ventilated pigs with severe lavage induced acute lung injury [16]. We hypothesized, that during HFOV oxygenation can be improved at a lower P mean with the animals positioned prone than supine.

Animals
The study was conducted in accordance with the National Institutes of Health guidelines for ethical animal research and was approved by the Laboratory Animal Care and Use Committee of the District of Unterfranken, Germany.
The experiments were performed in 14 healthy pigs, Pietrain breed, all negative for the malignant hyperthermia gene. The animals were 14 to 18 weeks old, with a mean (± SD) body weight of 54 ± 3.1 kg.

Experimental preparation
The animals were fasted for 24 hours without limiting water access. Prior to instrumentation the animals were sedated with intramuscular ketamin (10 mg/kg), xylazine hydrochloride (1 mg/kg) and atropine (25 µg/kg) and placed supine on an operating table armed with a heating pad to provide core temperature stability (37.3 ± 0.5°C). Anesthesia was induced with an intravenous bolus of sodium thiopental (5 mg/kg) using an auricular vein. The animals' trachea was orally intubated with a cuffed 8.0mm ID Edgar tracheal tube (Rueschelit ® , Ruesch, Kernen, Germany) providing an additional lumen embedded in the tubes inner wall for tracheal pressure monitoring. Anesthesia and complete muscle relaxation were maintained with continuous intravenous infusion of ketamin (2 mg/kg/h), midazolam (0.5 mg/kg/h), fentanyl (0.01 mg/kg/h) and vecuronium (0.1 mg/kg/h).
The animals were mechanically ventilated with a Servo ® 900C ventilator (Siemens-Elema AB, Solna, Sweden) using pressure controlled ventilation (PCV) with a PEEP of 5 cmH 2 O, an inspiratory to expiratory ratio (I:E) of 1:1 and a fraction of inspired oxygen (FIO 2 ) of 1.0. A V T of 6 ml/kg and a respiratory rate (RR) of 25-30 breath/min were applied resulting in normocapnia.
2 gm Cefotiam was administered intravenously. After systemic heparinization (300 U/kg Liquemin ® , Roche, Reinach, Switzerland) arterial and central venous access were established transcutanuously using ultrasound guidance (SonoSite 180 Plus ® , SonoSite Inc., Botell, WA, USA). Activated clotting time (ACT II ® , Medtronic, Minneapolis, MN, USA) was measured hourly and maintained between 150 and 200 seconds throughout the experiment with heparin bolus injections as needed. The left carotid artery was cannulated with a 20-gauge catheter (Vygon, Ecouen, France). The right internal jugular vein was cannulated with a 9 French introducer sheath (Arrow, Reading, PA, USA) and a 7,5 French flow directed thermodilution pulmonary artery catheter (831F75, Edwards Lifescience, Irvine, CA, USA) was advanced into the pulmonary artery under transduced pressure guidance.

Hemodynamic, ventilatory and blood gas measurements
For hemodynamic monitoring pressure transducers referenced to atmospheric pressure at the mid-thoracic level (Combitrans ® , Braun, Melsungen, Germany) and a modular monitor system (Servomed ® , Hellige, Freiburg i. Br., Germany) were used. Mean arterial pressure (MAP), mean pulmonary artery pressure (MPAP), central venous pressure (CVP) and pulmonary artery occlusion pressure (PCWP) were transduced. Heart rate (HR) was traced by the electrocardiogram.
Trifold injections of 10-ml aliquots of ice cold saline into the right atrium at random phases of different respiratory cycles were used for pulmonary artery catheter-based cardiac output (CO)-measurements (Explorer ® , Edwards Lifescience, Irvine, CA, USA).
For tracheal pressure monitoring air filled pressure transducers (Combitrans ® , Braun, Melsungen, Germany) referenced to atmospheric pressure were used [18]. Temperature was measured by thermistor in the pulmonary artery.

Experimental procedure Lung injury
After instrumentation the animals were stabilized for 30 min in the supine position and mechanically ventilated with PCV (V T = 6 ml/kg, I:E = 1:1, FIO 2 = 1.0, PEEP = 5 cmH 2 O). RR was adjusted to achieve normocapnia. Baseline measurements were obtained.
Lung injury was induced by bilateral pulmonary lavages with 30 ml/kg isotonic saline (38°C) and repeated every 10 minutes until PaO 2 decreased to 40-60 mmHg and was stable for 60 minutes with unchanged ventilator parameters. During induction of lung injury all lungs were ventilated with PCV, FIO 2 = 1.0, PEEP = 5 cmH 2 O, V T = 6 ml/kg, RR = 40/min. Post injury measurements were obtained.

Positioning
Prone positioning was performed with supportive rolls under shoulders and pelvis providing a free abdomen in order to minimized increases in intra-abdominal pressure.

Study protocol
The FIO 2 (1.0) remained unchanged throughout the experiment. A 20-min equilibration period was given for each modification following the study protocol. After time point T 0 the standard ventilator was replaced by an oscillatory ventilator (Sensormedics 3100B, Viasys, Conshohocken, PA, USA) without changes in P mean . The animals were randomly assigned to two groups (n = 7 each): A 20-min period was given for equilibration between each modification and followed by measurements of hemodynamics, blood gases and respiratory parameters. The following modifications were performed after completion of measurements terminating the previous 20-min period ( Figure 1): (1) T baseline : 30 min after instrumentation.

: Hemodynamic and metabolic data at baseline (T baseline ), after injury (T 0 ), after starting HFOV (T 20 ), after randomisation (T 40 ), during P mean -augmentation (T 60 -T 160 ), at end of experiment (T 180 )
Group T baseline  T 0  T 20  T 40  T 60  T 80  T 100  T 120  T 140  T 160  T   At the end of the experiment the animals were killed using an intravenous overdose of sodium thiopental and T 61 (Intervet, Unterschleissheim, Germany).

Statistical analysis
Values are reported as mean ± SD. Statistical analyses were performed with Statistica for Windows, version 5.1 (StatSoft, Tulsa, OK, USA). Two-way analysis of variance (ANOVA) for repeated measurements with factors mode and time were used for data analysis. Student-Newman-Keuls' post hoc test was used for comparison of significant ANOVA results within and between the groups. Data of the first measurement set (T baseline ) were only compared with data of the second measurement set (T 0 ). P values less than 0.05 were considered significant.

Results
Detailed data regarding hemodynamics, blood gases and respiratory parameters are presented in table 1. PaO 2 -, OIand CO-changes during the experimental period are displayed in figures 2, 3, 4.

Pulmonary gas exchange
Oxygenation improved significantly with rising P mean in both groups. At T 80 and T 100 PaO 2 was significantly higher in the prone positioned animals. A significantly higher PaO 2 compared to the preceeding time point was detected at T 80 in the prone positioned animals and at T 120 in the animals positioned supine. Significant improvement of OI occurred immediately after prone positioning (T 40 ) lasting until T 100 . SvO 2 was significantly higher from T 40 onwards in the HFOV-prone group if compared to T 0 without detectable significant differences between the groups. All animals remained hypercapnic with a PaCO 2 greater 70 mmHg resulting in a pH of less than 7.23 throughout the experiment in both groups.

Respiratory parameters
PIP increased significantly in all groups with rising P mean without differences between the groups.

Hemodynamics
MAP remained stable in both groups. CVP and PCWP started to rise significantly in the HFOV group from T 100 and T 140 respectively if compared to T 0 . MPAP was increased significantly in both groups from T 140 if compared to T 0 . From T 40 to T 140 CO and HR were significantly lower and continuously falling in the HFOV-prone group. At T 160 and T 180 no differences between the groups could be detected regarding CO and HR. Qs/Qt was significantly lower from T 40 to T 120 in the HFOV-prone group without differences between the groups from T 140 onwards.

Discussion
We evaluated the effects of prone positioning combination of HFOV and prone positioning in an adult animal model of ARDS. The major findings of our study are: 1) HFOV and prone positioning improves oxygenation at a lower P mean than HFOV and supine positioning. 2) HFOV and prone positioning result in significant reduction of pulmonary shunt fraction and normalisation of cardiac output at a lower P mean than HFOV and supine positioning. 3) Hypercapnia was neither ameliorated by HFOV nor combination of HFOV with prone positioning.
Since therapeutic alternatives are lacking and the underlying concepts sound reasonable, multimodal therapeutic approaches are commonly used for salvage therapy in patients with ARDS [19]. Apart from subsets of patients in other HFOV trails, the combined use of HFOV and prone positioning is described in one case report and was investigated systematically in a prospective randomized study including 39 medical patients [15,20]. Papazian et al.
Oxygenation Index (OI) throughout the study protocol Figure 3 Oxygenation Index (OI) throughout the study protocol Data are mean ± standard deviation. # p < 0.05 vs. T 0 ; * p < 0.05 vs. T T -20 ; § p < 0.05 HFOV vs. HFOV prone found the prone position combined with HFOV and PCV superior to HFOV and supine positioning in terms of oxygenation, but failed to demonstrate additive effects. However, the inflammatory mediators were elevated during HFOV-prone but not during HFOV-supine. The authors themselves put these results into perspective, since a control group was lacking and a time dependent natural change in the concentration of inflammatory mediators could not be excluded. It is a limitation of our study, that we did not investigate a control group ventilated in a conventional lung protective mode and positioned prone in order to detect additive effects of the two treatment modalities. Papazian et al. stressed the difficulties associated with bronchoalveolar lavages in ARDS patients in terms of patient safety and feasibility. This calls for long term experiments with large animals comparing conventional lung protective ventilation and HFOV with and without prone positioning looking not only at gas exchange and respiratory mechanics but also at histology and inflammatory mediators.
Current concepts to ameliorate the detrimental effects of VILI focus on reduction of volutrauma, barotrauma, atelectrauma and biotrauma [21]. It was shown in a small animal model, that HFOV had the same effect on oxygenation and pulmonary compliance than a conventional lung protective ventilatory approach but also reduced the systemic inflammatory response [22][23][24]. However, tracheal tube size, respiratory frequency and pressure amplitude are markedly different in small animals resulting in non-comparable changes of pulmonary mechanics and oscillatory pressure transmission. Therefore, experiments in large animals should be performed before HFOV is assessed systematically in adult patients with ARDS. Aiming to simulate a life-threatening clinical scenario, we induced an acute lung injury with severe hypoxemia and hypercapnia.
It was striking, that 19 ± 2 lavages with 30 ml/kg isotonic saline were needed to reach the targeted PaO 2 -value, suggesting a lung protective effect of the low-tidal-volume approach during ARDS-induction even on a low PEEPlevel (5 cmH 2 O). Intrinsic PEEP was measured during PCV by means of an endexpiratory occlusion maneuver for five seconds after every third lavage and was always less than 1 cm H 2 O. In two studies using sheep with a body weight of 30 kg, 4 lavages were needed to achieve a PaO 2 of less than 120 mmHg. These animals were ventilated with a V T of 12 ml/kg in a volume controlled mode [25,26]. Stability of the experimentally induced ARDS was proven two-fold: 1) A stabilisation period of 60 min. with unchanged ventilatory parameters was kept between the last pulmonary lavage and T 0 . 2) After T 160 , HFOV and prone positioning were discontinued, all animals were Cardiac output (CO) throughout the study protocol Although the combined application of HFOV and prone positioning improved oxygenation, normalized cardiac output and significantly reduced pulmonary shunt fraction at a lower P mean than HFOV alone, hypercapnia was not influenced in our experiment. This is consistent with clinical results, since normocapnia was not achievable with HFOV alone in many adults with ARDS [8,27].

Conclusion
In this saline lavage induced porcine model of ARDS, we showed in a clinically relevant scenario, that the combination of HFOV and prone positioning improved oxygenation at a lower P mean than HFOV combined with supine positioning. In addition, reduction of the pulmonary shunt fraction and normalisation of the cardiac output was achieved at lower airway pressures. The ventilator pressure amplitude is a major determinant of VILI. HFOV might be a step towards further lung protection, since sufficient oxygenation can be restored or maintained with a significant reduction of the ventilator pressure amplitude when compared to standard respirator modes. However, HFOV failed to be a major component in ARDS treatment algorithms in adult patients. Having in view a long history of failed multimodal treatment approaches in ARDS research, we now conclude from our results that a combination of HFOV and prone positioning seems promising and should be further investigated systematically and compared to conventional lung protective ventilation. Long term trials in large animals and aquisition of histologic and immunologic data clearly seem justified.