INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Perfluorochemicals are characterized by a high potential for gas transport, a low surface tension and a high specific weight. The advantageous effects of perfluorocarbon associated gas exchange (PAGE) were described years ago (1, 2). Despite recent advances in liquid ventilation, a standard mode of application of perfluorocarbon (PFC) has not been established yet. Although total liquid ventilation (TLV) with completely liquid-filled lungs is beneficial, the necessity for a liquid filled tube system that contains pumps, heater, and membrane oxygenator to deliver and remove tidal volume aliquots of conditioned perfluorocarbon to the lungs is of great disadvantage. In contrast, partial liquid ventilation (PLV) can be applied using standard ventilators connected with gas-filled standard respirator systems, delivering tidal volumes of oxygen-air mixture to perfluorocarbon filled lungs. The influence on oxygenation, carbon dioxide removal, and lung mechanics has been investigated in several animal studies using different models of lung injury (3). Clinical applications of PLV have been reported in patients with acute respiratory distress syndrome (ARDS), meconium aspiration syndrome, congenital diaphragmatic hernia, and respiratory distress syndrome (RDS) of neonates (8). PLV requires extreme respiratory care, because the ventilatory setting is determined by the perfluorocarbon filled lung (11). Profound expertise is mandatory to perform and maintain filling of the lung with perfluorocarbon to FRC. Disruption of PLV immediately deteriorates gas exchange. Incomplete filling of the lung has been shown to be less effective than filling the lung to FRC volume (13). Severe adverse events affecting gas exchange and pulmonary circulation limit the use of PLV. To enable the wider use of the therapeutic principles of PAGE, less sophisticated application techniques with a lower incidence of adverse effects need to be developed.

Aerosolization of vasodilative agents has proven to be an efficient and safe treatment of gas exchange disturbances caused by ventilation perfusion mismatch (14). Recent studies with aerosolized surfactant have been encouraging (15).

We therefore examined the impact of aerosolized perfluorocarbon on oxygenation, carbon dioxide removal, and lung mechanics in surfactant depleted neonatal piglets with subsequent acute respiratory distress syndrome. To date, no studies on the efficacy of aerosolized perfluorocarbon have been performed.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The study was approved by the Animal Care Committee of the University of Erlangen and the government of Mittelfranken, Germany, and performed according to guidelines of the National Institutes of Health. Twenty newborn piglets weighing 3.5 to 4.3 kg were anesthetized intravenously by an injection of ketamine, midazolam, and fentanyl. An endotracheal tube (4.0 mm in diameter; Mallinckrodt, Hennef, Germany) was placed via tracheotomy and endotracheal pressure was recorded via a 5-Fr. catheter. After paralyzation with vecuronium, a 4-Fr. thermodilution catheter (Arrow International, Erding, Germany) was placed in the pulmonary artery. The left femoral artery was cannulated with a 20-G arterial catheter (Arrow). Arterial blood pressure, central venous pressure, pulmonary artery pressure and body temperature were continuously recorded. Cardiac output was calculated with the CMF 24 Omnicare (Agilent, Böblingen, Germany). Arterial blood gas analysis was performed in 15-min intervals during therapy and in 30-min intervals during the post-therapy period (ABL 330; Radiometer, Copenhagen, Denmark).

The pulmonary data were recorded with a heat wire anemometer (MIM GmbH, Krugzell, Germany), computed with the neonatal respiration monitoring Florian NRM-200 (MIM). To identify lung overdistension, C20/c (20% terminal compliance/compliance) was calculated (16). Intermittent mandatory ventilation (IMV) was performed with a neonatal respirator (Infant Star 950; Mallinckrodt). Respiratory gas was humidified at 39° C (MR 700; Fischer & Paykel, Welzheim, Germany). Breath rate was 50 breaths/min. A peak inspiratory pressure (PIP) of 32 cm H₂O, a positive end-expiratory pressure (PEEP) of 8 cm H₂O, and an FI_O2 of 1 was used. Lung injury was induced by repeated saline lung lavage (17) using 30 ml/kg per side. Lavage was performed in left and right lateral position, respectively. During instrumentation and for the duration of the experiment animals were in the supine position. Lung injury was considered to be stable, when the Pa_O2 constantly remained below 80 mm Hg for 60 min. The animals were randomly assigned to four therapy groups. (1) Aerosol-PFC, (2) FRC-PLV, (3) low-volume (LV)-PLV, (4) Control (IMV).

Respiratory settings remained constant to ensure comparability between the groups. The Aerosol-PFC group received 10 ml/kg/h FC77 (C₈F₁₈ and C₈F₁₆O, density 1.78 g/cm³; 3M, Neuss, Germany) by an aerosolization catheter (oxygen jet aerosol generator; Trudell Medical Inc., Toronto, ON, Canada) (18), (particle diameter, mean: 5.7 µm). The catheter consists of gas and liquid capillaries, converging and terminating at the distal tip of the catheter. The close contact of gas and liquid results in efficient nebulization, with gas flow rates as low as 0.05 L/min.

In the FRC-PLV group, 30 ml/kg FC77 were injected into the lung over a period of 30 min followed by continuous substitution (20 ml/kg/h) with FC77 to compensate FC77 evaporation. The LV-PLV group received 10 ml/kg/h FC77 endotracheally. The control group was ventilated with IMV. Therapy was stopped after 2 h. After an observation period of 6 h, the animals were killed.

Data Analysis and Statistics

Values are expressed as mean ± SEM. After testing for Gaussian distribution, two-way ANOVA was used for comparison between the groups. In case of significance, Bonferroni post-hoc test was applied. A p value of less than 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Arterial Oxygen Tension

Aerosolized perfluorocarbon treatment significantly increased Pa_O2 values in comparison with the untreated control group (p < 0.001) and the LV-PLV group (p < 0.001). The rise in Pa_O2 was slightly slower in Aerosol-PFC-treated animals than in the FRC-PLV group (Figure 1). After discontinuation of PFC-therapy, there was a sustained Pa_O2 increase in the Aerosol-PFC group, but not in the FRC-PLV group. Aerosol-PFC-treated piglets had significantly higher Pa_O2 values than did all the other groups (p < 0.01) 6 h after the end of the therapy.

View larger version (16K): [in this window] [in a new window]   Figure 1.   Mean and SEM of PaO2 obtained before and after induction of lung injury during therapy with Aerosol-PFC, FRC-PLV, LV-PLV, CONTROL, and during the post-treatment observation period in surfactant-depleted neonatal piglets.

Oxygenation Index

Oxygenation index (OI) was calculated {[Pa (cm H₂O) × FI_O2/ Pa_O2 (mm Hg) ] ×100} (19). OI raises with increased mean airway pressure and elevated FI_O2, OI declines with improved Pa_O2. Inversely proportional to the increase of Pa_O2, OI fell from 29.9 ± 3.4 to 17.1 ± 3.8 during the first 30 min of Aerosol-PFC therapy. The fall of OI was even faster and more pronounced in the FRC-PLV group (from 31.3 ± 1.5 to 5.1 ± 0.6), but after 2 h of treatment, the OI of both Aerosol-PFC and FRC-PLV groups was significantly lower than in the LV-PLV and control groups (p < 0.001 each). The effect of Aerosol-PFC treatment on the OI persisted for 6 h after therapy, compared with the control group (p < 0.001), whereas in the FRC-PLV group rapid deterioration was seen after cessation of PFC instillation, with a rise of OI from 7.7 ± 2.5 to 21.2 ± 13.3 within 15 min (Table 1).

Carbon Dioxide Removal

Aerosol-PFC treatment was associated with significantly lower Pa_CO2 values than in untreated control group (p < 0.01) and in animals treated with LV-PLV (p < 0.01) after 60 min of therapy (Figure 2). This effect persisted after discontinuation of aerosol therapy for 6 h (p < 0.01). After 30 min of therapy, the Pa_CO2 levels of the FRC-PLV group were significantly lower than those of the control and LV-PLV groups (p < 0.01). This effect persisted after discontinuation of PFC therapy for 6 h (p < 0.01).

View larger version (16K): [in this window] [in a new window]   Figure 2.   Mean and SEM of PaCO2 before and after induction of lung injury during therapy with Aerosol-PFC, FRC-PLV, LV-PLV, CONTROL, and during the post-treatment observation period in surfactant-depleted neonatal piglets.

Eight hours after the beginning of PFC therapy, the difference between the Aerosol-PFC and the FRC-PLV groups (24.2 ± 1.7 mm Hg versus 35.9 ± 2.8 mm Hg) was not significant.

Ventilatory Efficacy Index

To assess ventilation, independent from respirator settings, ventilatory efficacy index (VEI) was computed: {3800/[PIP-PEEP(cm H₂O)] × respiratory frequency (cpm) × Pa_CO2} (20). The VEI in the Aerosol-PFC treated group was significantly higher than that in the control group (0.116 ± 0.011 versus 0.071 ± 0.011) after 60 min of therapy (p < 0.05). The fastest rise in ventilatory efficacy index occurred in the FRC-PLV group. After 30 min, the VEI in the FRC-PLV group was significantly higher than the VEI of the control group (0.099 ± 0.011 versus 0.071 ± 0.011, p < 0.05). After withdrawal of PFC-application, the VEI in the FRC-PLV group, but not in the Aerosol-PFC group dropped to values comparable with those before the start of therapy (Table 1).

Lung Mechanics

Within 15 min of therapy, dynamic compliance of the respiratory system improved immediately in the Aerosol-PFC and the FRC-PLV groups. High C20/c values indicate a high terminal dynamic compliance and a reduced lung overdistension (16). After 120 min of Aerosol-PFC therapy, the C20/c was significantly (p < 0.01) higher than in the control and LV-PLV groups (Figure 3). Although the FRC-PLV group showed significantly higher C20/c values than the control and the LV-PLV groups did after 2 h of treatment (p < 0.001), the positive effect subsided immediately after the end of the therapy, showing no significant difference when compared with the control group during the following observation period. In contrast, there was a sustained improvement of C20/c in the Aerosol-PFC group post-therapy, which was twofold higher than in the control and the LV-PLV groups (p < 0.01) and significantly higher than in the FRC-PLV group (p < 0.001).

View larger version (16K): [in this window] [in a new window]   Figure 3.   Mean and SEM of the C20/c to estimate the terminal dynamic compliance and lung over distension. High C20/c values indicate a high terminal dynamic compliance and a reduced lung overdistension (16). Obtained before and after induction of lung injury during therapy with Aerosol-PFC, FRC-PLV, LV-PLV, CONTROL, and during the post-treatment observation period in surfactant-depleted neonatal piglets.

Circulatory Measurements

There was no significant difference between the groups with respect to cardiac output, central venous pressure, or body temperature during the experimental observation period.

Safety Measurements

The absence of air leak and the position of the thermodilution catheter were confirmed by chest radiography. Endotracheal pressure was similar in all groups.

During the post-therapy period, a fatal tension pneumothorax occurred in one animal of the LV-PLV group.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Treatment with aerosolized perfluorocarbon resulted in a sustained improvement of oxygenation, ventilation, and lung mechanics, lasting for hours after the end of the intervention in neonatal surfactant depleted piglets with RDS. The application of PFC as aerosol appears to be an essential factor for the effective delivery to ventilated lung areas and subsequently for the observed beneficial effects, as the instillation of equal volumes of non-aerosolized PFC (LV-PLV group) did not show any improvement in gas exchange and lung mechanics.

We did not change positive end-expiratory pressure (PEEP) during the experiment to point out the differences of the therapies. As PEEP improves gas exchange and pulmonary mechanics during partial liquid ventilation (4), it could be supposed that PEEP would influence the effect of PFC-aerosol therapy.

FC77 was used because of its wide use in the literature (12, 21, 22) and personal communication with 3M Germany suggesting the absence of substantially toxic effects. Studies are warranted to evaluate different perfluorocarbons for their suitability for aerosol therapy. Therefore it is necessary to validate the role of vapor pressure of typical PFCs during aerosol therapy.

Positive effects on oxygenation of PLV and TLV with voluminous fillings of the lungs are well known (23), whereas the application of PFC as aerosol is a new approach. Liquid perfluorocarbon follows gravity to dependent areas and perfluorocarbon-assisted gas exchange is most effective when a lung volume near to FRC volume is filled with PFC. Changes in filling volume result in deterioration of pulmonary gas exchange. This could be confirmed by our data, as the FRC-PLV group showed only temporary improvement in gas exchange, whereas the effect in the Aerosol-PFC group was maintained after discontinuation of treatment. In contrast to PLV or TLV, the application of PFC as aerosol requires a minimum of additional technical effort: inhalation catheter, infusion pump and an oxygen flow. Vaporisation of PFC, resulting also in a significant improvement of gas exchange and lung compliance 2 h after a 30-min therapy in adult sheep (26), requires an anesthesia machine with two serial vaporizers especially modified for use with PFC. PFC-vapor therapy was performed with perfluorohexane (C6F14). Its high vapor pressure (177 mm Hg at 20° C and about 375 mm Hg at 37° C) and relatively low boiling point (57° C) allow the vapor to be generated in an anesthesia machine. Vapor administration with PFCs like FC77 (3M) (vapor pressure, 85 mm Hg at 37° C; boiling point, 97° C) or LiquiVent (Alliance) (vapor pressure, 11 mm Hg at 37° C; boiling point, 147° C) is not effective at body temperature. Data concerning the improvement of lung mechanics using liquid ventilation are controversial in the literature (5, 17, 23, 27). Our data show a significant but temporary improvement of C20/c in the FRC-PLV group and a persistent improvement in the Aerosol-PFC group. As a recent study using PLV with very small amounts of PFC (3 ml/kg) in combination with high frequency oscillation ventilation (HFOV) showed augmented compliance, it seems that PFC distribution on the lung surface plays an important role (28). Several factors may contribute to the effectiveness of treatment with aerosolized PFC. The oxygen carrier effect of aerosol particles, generated by the oxygen jet of the inhalation catheter to the alveolar space, might contribute to the increase in oxygenation. The overwhelming effects on gas exchange and lung mechanics, especially regarding its persistence after the end of inhalation in aerosol-PFC treated animals in our study, might be caused by coating the bronchial and alveolar epithelium with PFC. This liquid layer might lead to an effective reduction of alveolar surface tension, yielding to an increase of lung compliance. Surface tension of PFC is very low compared with that of water (e.g., FC77 15 versus 72 dyne/cm at 25° C); therefore, PFC could act like surfactant. Both aerosol-PFC and FRC-PLV first might work as artificial surfactant. We speculate, that in this 2-h PLV intervention, removal effects of the newly produced surfactant caused by PLV might contribute to the difference between aerosol treatment and PLV at FRC volume, whereas aerosol-PFC therapy might not interfere with the synthesis of natural surfactant. The extent of improvement of gas exchange and pulmonary compliance suggests an effective volume recruitment in even non-dependent, atelectatic lung areas because of a more homogenous intrapulmonary delivery of aerosolized PFC. In contrast to PLV, where liquid PFC follows gravity and performs liquid levels in the lung (FRC-PLV group), we suppose a more homogenous distribution of aerosolized PFC in the lung because of the transport of aerosol particles with the inspired air. We confirmed the absence of liquid levels by roentgenoscopy in the aerosol and the LV-PFC group. Further studies are required to prove the pulmonary distribution of aerosolized PFC. To date, we have no data proving different effectiveness of PFC-aerosol therapy in prone, supine, or changing position, as it has been discussed for liquid ventilation. Further studies focusing on lung injury have shown that the early inflammatory response is reduced in animals treated with aerosolized PFC (K. von der Hardt and colleagues, submitted). Improving alveolar gas exchange without pressure-induced alveolar overdistension might reduce barotrauma and volutrauma, leading to a reduction of the inflammatory cascade of ARDS and decreasing the perpetuation of the disease (29).

Although the number of animals was too small to answer all safety questions raised by the new mode of PFC application, our data did not reveal any adverse effect associated with this method.

In summary, our study has suggested that aerosolization represents a new, efficient, safe, and easy-to-handle way of perfluorocarbon application in surfactant-depleted piglets.