Perfluorochemicals have been explored as respiratory media for more than 20 years in many animal models of respiratory disease and, more recently, in initial phase I and II human trials of liquid ventilation.¹ These chemicals are very stable, have low surface tension, are generally insoluble in water or lipid, and are excellent solvents for respiratory gases. The effectiveness of liquid ventilation is achieved, in part, through recruitment of collapsed gas-exchanging units, increasing lung compliance, and cleansing the lung of pulmonary debris. Sterile perflubron (LiquiVent; Alliance Pharmaceutical Corp, San Diego, CA), is one of very few perfluorochemicals that is produced as a medical-grade drug. The compound has an impressive biocompatibility profile in animals and humans and is the only fluid that is approved by the Federal Food and Drug Administration for testing as a breathing agent. Perflubron has a high solubility for respiratory gases, a positive spreading coefficient on saline, and a low viscosity and is highly radiopaque on plain radiograph or computed tomographic imaging.^8,9

Recent advances in respiratory therapies for the critically ill neonates have expanded the treatment options for infants with severe respiratory failure. In those severely effected infants, extracorporeal life support (ECLS) may result in increased survival. Nevertheless, morbidity remains high, and mortality for select groups of infants who require ECLS may exceed 40%.⁹ The diseases that produce severe respiratory failure that may be refractory to current therapies and ECLS include acute respiratory distress syndrome (ARDS) and congenital diaphragmatic hernia (CDH).^12,13 Therapy for infants receiving ECLS could be optimized through effective recruitment of available lung parenchyma, increasing lung compliance, and removal of lung debris.¹²

Partial liquid ventilation (PLV) is a technique wherein the lung is slowly filled with a perfluorochemical, and ventilation is maintained with conventional gas ventilation of the liquid-filled lung.^3,16 Clinical investigations designed to test the safety and efficacy of PLV with perflubron in humans are now underway. We report the results of one such trial, which used perflubron PLV in critically ill infants receiving ECLS. The objective of this trial was to demonstrate that a period of perflubron PLV improves pulmonary function, without adverse events, in a select group of critically ill infants with CDH or ARDS who are receiving ECLS and have a high likelihood of mortality.

METHODS

1. Have failed conventional support and have been receiving ECLS for respiratory failure;
2. Have failed, or are not candidates for, other means of support (ie, surfactant replacement therapy and high-frequency ventilation);
3. Have demonstrated no improvement with ECLS therapy after 2 days, as indicated by the presence of both of the following: measured lung compliance of less than 0.2 mL/cm H₂O per kilogram and a requirement of at least 70 mL/kg bypass flow to maintain a venous saturation of greater than 75%;
4. Have had a CDH surgically reduced (if present) at least 3 days before entry;
5. Have had no active gas leak from a chest tube; and
6. Were considered critically ill, with a high likelihood of mortality by two physicians trained in neonatal intensive care who are not members of the research team.

Protocol

After this 4-hour period, PLV was continued up to 96 hours at the discretion of the investigator. Weaning from liquid to gas ventilation required the discontinuation of supplemental dosing with perflubron and was initiated by the investigator at any time after the end of 4 hours of PLV. Guidelines for continuation of PLV included the assessed adequacy of the fill (full lung recruitment) and response to PLV. Ventilator changes and changes in ECLS flow rates after the initial 4 hours were made by the investigator and intensive care nursery team to optimize the effects of the perflubron on the lungs (recruiting lung volume) and to promote lung healing. As perflubron evaporated from the lung, supplemental doses were given to maintain a total lung volume of perflubron approximately equal to FRC.

During the first 72 hours after perflubron instillation, blood samples were obtained at 1, 24, 48, and 72 hours to assess perflubron absorption into the blood. Samples of blood (0.1 mL) were placed in 2-mL vials, immediately sealed, and stored at 70°C until analysis. Blood was analyzed using electron capture gas chromatography. For analysis, sealed samples were equilibrated at room temperature, and 10 µL of vapor from each sample was analyzed in a gas chromatograph (model 5890A; Hewlett Packard Corp, Wilmington, DE) equipped with an electron capture detector.

The instrument was calibrated from stock standards prepared by adding 2 µL of perflubron to a capped and sealed 120-mL bottle and allowing sufficient time for complete evaporation (typically 1 hour). From this stock standard, working samples were prepared by transferring 120-µL aliquots to a clean 120-mL bottle already sealed. Samples of working standard (ie, 10 µL per sample) were injected to calibrate the instrument.

Clinical Assessment

RESULTS

Table 1. Patient Demographics [View Table]

The infants tolerated PLV well. All adverse events were considered caused by the underlying illness and unrelated to perflubron. Four of the six infants were weaned from ECLS support to conventional ventilation for at least 3 days. Two are long-term survivors. Both of the infants with CDH died of pulmonary dysfunction (severe pulmonary hypoplasia and persistent pulmonary hypertension). The two infants with ARDS that did not survive died of problems unrelated to lung function (multisystem failure from systemic herpes infection and bowel disease 2.5 months after PLV).

The change in tidal pressure-volume relationship with PLV in a full-term infant with CDH is shown in Fig 1. The infant had hypoplastic lungs, pulmonary hypertension, and consolidation of the pulmonary parenchyma after nearly 3 weeks of ECLS, resulting in a flat loop. This improved with debris removal and volume recruitment with 24 hours of PLV. The infant was removed from ECLS after 72 hours of PLV, and PLV was continued to 96 hours. The increased respiratory compliance was maintained for 48 hours after PLV, but the infant died 3 days later of severe pulmonary hypertension.

The chest radiographs for an infant with ARDS before, during, and after PLV are shown in Fig 2. This infant had failed to clear her persistent pulmonary consolidation despite a prolonged ECLS course (Fig 2, a). Some lung volume was recruited during the initial 4-hour period of PLV (Fig 2, b). Further filling and suctioning resulted in more uniform lung distention by 48 hours of PLV (Fig 2, c) that persisted after ECLS and PLV (Fig 2, d).

The change in pulmonary compliance with PLV in this infant is depicted in Fig 3. As shown, the infant had persistently low pulmonary compliance (and lung consolidation) before PLV. There was a small decrease in pulmonary compliance during the first hour of PLV, which recovered to slightly above pre-PLV values by 4 hours of PLV. A large and sustained increase in pulmonary compliance is not seen until 48 to 72 hours of PLV. This infant was weaned off ECLS while receiving PLV, survived, and is currently well.

Dynamic pulmonary compliance improved in all infants from before PLV to immediately after the end of PLV (mean ± SE, 186% ± 51% increase; range 24% to 344% increase).

Based on gas chromatograms of arterial blood samples, the concentration of perflubron in the serum gradually increased over time after the initiation of PLV. As shown in Table 2, mean ± SE values for perflubron in the blood increased from 0.98 ± 0.64 µg of perflubron/g of blood at 1 hour to 3.81 ± 1.95 µg of perflubron/g of blood at 72 hours after PLV initiation. There were large variations in blood levels of perflubron over time between infants. The greatest perflubron level, however, never exceeded 8.7 µg of perflubron/g of blood.

Table 2. Serum Perfluorochemical Content, µg of Perflubron/g of blood [View Table]

DISCUSSION

PLV has been shown to be very effective in improving lung function and in many different animal models with respiratory failure. Perfluorochemicals are inert liquids generated by replacing the carbon-bound hydrogen atoms on organic compounds with fluorine. The perfluorochemical perflubron (C8F17Br1) carries 50 mL of oxygen and 219 mL of CO₂/dL. It has a relatively low viscosity (1.1 centisokes), low surface tension (18 dyne/cm), and high density (1.93 g/mL). In addition, perflubron is relatively insoluble in lipids and water, evaporates at a rate similar to water, and is biologically inert. Histological evaluation of prematurely delivered animals ventilated with perfluorochemicals and recovered to air respiration demonstrate decreased hyaline membrane formation, reduced injury to the airway epithelium, and distal air spaces, with clearance of alveolar debris.^1,6,7,16,18 Animal studies of liquid ventilation for models of ARDS and CDH have demonstrated improved lung function when compared with gas ventilation.^18,19

Previous human trials have shown the feasibility of PLV in improving lung function in critically ill infants and adults. The improvement in pulmonary compliance with PLV observed in the present study was more gradual when compared with experience in animals and preterm infants with respiratory distress syndrome.^2,5 In the preterm population, the predominant mechanism of action of perflubron PLV is the reduction of surface tension and volume recruitment of collapsed terminal air spaces. In the larger infant with ARDS, the mechanism of action of perflubron PLV is predominantly through volume recruitment of alveoli that have been filled with debris. Subsequent to this relatively slow process of instilling perflubron combined with suctioning of debris, perflubron reaches the alveolar lining and reduces surface tension. Hence, improvement in lung function required hours of PLV in these six infants.

This study suggests that the treatment of patients with ARDS with perflubron PLV is safe and potentially beneficial. This supports the findings of other phase I and II studies of PLV in selected infants, children, and adults with moderate to severe ARDS.² The debris removal, alveolar recruitment, reduction of surface tension, and maintenance of an oxygenated liquid within the alveoli that are the potential acute benefits of PLV therapy may have a direct, favorable impact on the interruption of processes leading to ARDS. In addition, the bacteriostatic and radiopaque qualities of perflubron present added benefits to PLV therapy in patients with this disease process.¹

For infants with CDH, ECLS therapy replaces native cardiopulmonary function and maintains normal gas exchange until pulmonary hypertension, and the difficulties associated with transition from fetal to neonatal cardiopulmonary circulation can resolve (eg, right heart failure and systemic hypoxemia). If the infant's lungs are severely hypoplastic, however, chances of survival are low, because little lung growth can be expected in the 3- to 4-week period usually considered safe for ECLS.¹³ ALthough both infants with CDH in the present protocol died after PLV, both had evidence of severe pulmonary hypoplasia and pulmonary vascular hypertension and had improved lung volume and lung function with perflubron PLV. The diagnosis of the degree of lung hypoplasia and maximal recruitment of available pulmonary parenchyma may better define prognosis, and optimize outcome, in these infants.

In addition to assessing the potential efficacy of PLV, we evaluated the safety of PLV with perflubron in this population. As expected, numerous adverse events were reported in each infant, but none were deemed to be related to perflubron. Small but significant levels of perflubron were detected in blood samples from all tested infants after perflubron PLV initiation. As noted, mean values of perflubron blood concentrations increased gradually in time up to 72 hours, even though patents 2, 3, and 4 were treated with perflubron for 4, 4, and 24 hours, respectively. The large variations in blood levels of perflubron may be related to several issues, including difference in treatment dosing, lung disease, and blood lipid content. Although there are no direct comparative data for uptake of perflubron after direct pulmonary instillation in patients with ECLS, our results are similar to those of other adult animal and human studies, which noted very low concentrations of different perfluorochemical compounds in the blood after short-term intravenous or tracheal instillation of perfluorochemical liquids.²⁴ In addition, pulmonary exposure to perflubron results in blood concentrations that are extremely low compared with the levels associated with intravascular administration of 5 mL/kg perfluorochemical emulsion, which has been used for angioplasty in humans.

The study suggests that PLV with perflubron is safe, improves lung function, and recruits lung volume in critically ill infants receiving ECLS. Although PLV may optimally recruit lung volume in patients with CDH, severe pulmonary hypertension and hypoplasia may only resolve with prolonged or in utero therapy.¹⁹ PLV therapy for infants with ARDS seems to have a great deal of promise. Based on this and other phase I and II trials, studies of PLV on full-term infants with ARDS before ECLS have been initiated.