Early perfluorodecalin lung distension in infants with congenital diaphragmatic hernia☆☆☆

Materials and methods 

Patient selection and ECMO management 

Six consecutive neonates with CDH who required ECMO support between June 1998 and April 2002 were recruited (Table 1).

Table 1. Patient details

	Patient No.	Weight (kg)	Gestation (wk)	Age at Cannulation (h)	Age at PFC Instillation (h)	Duration of ECMO (d)	CDH Repair (ECMO day)	Duration of PFC (d)	Age at Extubation (d)
	1	2.5	37	20	24	14	8	8	38
	2	3.0	38	11	48	14	7	6	35
	3	3.0	41	12	31	24	6	9	60
	4	3.6	41	127	130	11	3	6	19
	5	2.4	38	16.5	29.5	10	5	7	33
	6	3.6	39	5	10	14	5	10	51

This represented 17% of the total number of infants with CDH treated on the neonatal surgical intensive care unit. All patients were at term (39.2 ± 0.5 weeks) with birth weight 3.13 ± 0.24 kg. Four patients had a left-sided hernia. All were symptomatic within 1 hour of birth. Five were referred for ECMO support within 24 hours having never achieved a sustained period of stability. One patient deteriorated rapidly after an initial period of stability and was referred on day 6. None responded to maximal medical management. This included surfactant and inhaled nitric oxide in all cases and a trial of high-frequency oscillatory ventilation in 3 cases. The inability to adequately oxygenate these patients, despite high ventilation pressures, is indicated by the high oxygenation index (mean, 50.8 ± 8.0). An oxygenation index greater than 40 was associated with 100% mortality rate for CDH in the conventional arm of the UK Neonatal ECMO Trial.5 The first 2 patients were supported with venoarterial (VA) ECMO. Subsequent patients received venovenous (VV) ECMO support through a double-lumen cannula. Local ethics committee approval was obtained for this pilot study. Informed consent was obtained before recruitment.

Results 

Mean initial perfluorodecalin volume required to fill the lungs was 5.3 ± 0.8 mL/kg and mean maximum perfluorodecalin instilled volume was 11.3 ± 0.8 mL/kg (P < .02). In one patient perfluorodecalin instillation was stopped at 9 days because of a fluorothorax leak from the hypoplastic lung on the third postoperative day. As a chest drain already was in place, perfluorodecalin instillation was stopped, and the fluorothorax resolved.

The radiographic area of the lung on the affected side increased significantly (mean percentage increase 272%; P < .02; Fig 1).

View Large Image Download to PowerPoint

Fig. 1. Early and late lung area measured on chest radiograph during perfluorodecalin distension (mean + SEM).

View Large Image Download to PowerPoint

Fig. 2. Fig 2. Appearances of lungs during perfluorodecalin distension in a representative patient.

All patients survived with follow-up of between 3 to 42 months. Four children are no longer oxygen dependent.

Discussion 

ECMO improves survival rate in neonatal respiratory failure,5 but the benefits in CDH are less clear.1 The Extracorporeal Life Support Organisation reports a survival rate to discharge of 54%.6 Of the neonates with CDH who were supported with ECMO between 1991 and 1999 in the United Kingdom, less than 40% survived to 6 months (personal communication). These disappointing results encouraged us to consider other treatment options in an attempt to treat the underlying pulmonary hypoplasia. The duration of ECMO support represents a period with good oxygenation and tissue perfusion independent of ventilation. We utilised this time to introduce a protocol of perfluorodecalin lung distension.

This pilot study results show that perfluorodecalin lung distension results in a gradual increase in the vital capacity of the lungs as measured by volume of liquid instilled and radiographic dimensions. The assessment of lung dimensions described was felt to be reliable because the ECMO cannula used for calibration was of known diameter and in the same plane as the lungs. Measurements made on chest radiographs have been shown to accurately estimate lung volume in children.7

The increase in lung volume may represent nontraumatic recruitment of alveoli or true lung growth. It is possible that both occur. Perfluorocarbon lung distension in an animal model resulted in increased lung volume. Alveolar morphometry showing normal lung architecture3 and increased growth factor gene expression8 suggested that this increase in lung volume represented true lung growth. Histologic analysis was not possible in the current series as all patients survived.

This study differs from that of Fauza at el4 in showing an increased volume of both lungs. Our protocol differs in the timing of liquid distension and the use of perfluorodecalin instead of perflubron (perfluorooctylbromide), the other medical grade PFC. Perfluorodecalin instillation and distension was started as soon as possible after ECMO support was established. In 5 patients, ECMO support was started within 24 hours of birth so the distending force was applied at the earliest possible stage. In all patients, CDH repair was undertaken on ECMO support, and the perfluorodecalin distension protocol was continued after the repair. This ensured that the affected lung could distend further without the constraint of the herniated viscera, particularly the liver.

Perfluorocarbons are an ideal liquid to use for intrapulmonary distension. They are inert,9 distribute uniformly throughout the lung,10 have antiinflammatory properties,11 and are radiopaque, which facilitates measuring lung dimensions on radiographs. There are important differences between perfluorodecalin and perflubron. Perfluorodecalin has a higher kinematic viscosity and lower spreading coefficient, both of which contribute to a more predictable distribution.12 Although both are radiopaque, perfluorodecalin is less radiopaque allowing clear visualisation of other structures such as central lines and ECMO cannulae on the chest radiograph.

To optimise the benefits of perfluorodecalin distension there should be uniform distribution throughout both lungs. Suboptimal distribution can occur if there is insufficient PFC within the bronchial and alveolar tree. A reduction in the amount of PFC over time results from evaporative losses. Changing the perfluorodecalin regularly concealed these losses. This technique also removed mucous plugs, improved distribution, and encouraged more uniform lung distension. A cuffed ET tube prevented leakage of perfluorodecalin and maintained an adequate fill volume. Distribution also was probably improved by allowing the infant to breathe spontaneously.

All 6 patients recruited to this study survived. This exceeded our expectations given the severity of illness of these patients as reflected by their immediate presentation at birth, their failure to respond to maximal medical therapy, high oxygenation index, herniation of the liver, and the size of the ipsilateral lung as demonstrated by the perfluorodecalin. Available evidence suggests that postnatal lung distension results in an improvement in lung volume and lung function and may represent a treatment for pulmonary hypoplasia. A randomised-control trial of early perfluorodecalin lung distension in infants with severe CDH on ECMO support is warranted to clarify the beneficial effects.