Reprint requests: Dr. Victor Yu,.Professor of Neona- tology, Monash University, Director of Neonatal Intensive Care, Monash Medical Center, 246, Clayton Road, Clayton Victoria 3168, Australia.
Email: [email protected]

Prevention of respiratory distress syndrome (RDS) is best achieved with prevention or effective treatment of preterm labor, failing which antenatal corticosteroid therapy is the next most efficacious intervention. Postnatal prevention of RDS is the remaining option after the above measures have been taken, and must not be solely relied upon in the place of good obstetric practice. Postnatal prevention of RDS depends on general measures such as prompt treatment of perinatal asphyxia and a high quality of neonatal care. This review focuses' on the use of surfactant therapy as an important component of clinical practice within the neonatal intensive care unit for the management of RDS, in particular its successes and remaining controversies, and surfactant therapy for other indications.

Types of Surfactant

Studies published in the 1970s showed that natural surfactant derived from animal lungs improved lung function and pre- vented respiratory distress in prematurely delivered animals. In 1980, the first successful human study was published, showing that surfactant replacement therapy improved oxygenation, ventilatory requirements, X-ray abnormalities, acidosis and hypotension in.1O pre term infants with RDS(1). Surfactants are of two general classes, natural and synthetic surfactants. Natural surfactants are either heterologous (from minced bovine lung tissue, e.g., Survanta, or extracted from saline lavage fluid of calf lungs, e.g., Infasurf, or from porcine sources, e.g., Curosurf) or homologous (human aminiotic fluid from term pregnancies). Some are. modified by addition of dipalmitoyl phosphotidylcholine, palmitic acid and triglyceride (Survanta) or by removal of neutral lipids by chromato- graphy (Curosurf) to improve the surface properties of the extract. Synthetic surfactants (e.g., Exosurf) contain dipalmitoyl phosphotidylcholine,the main surface- active phospholipid of surfactant, and other components which facilitate surface adsorption. Natural surfactants contain hydrophobic surfactant proteins (SP) Band C which aid in surfactant adsorption and resist surfactant inactivation(2). The main structural properties of these proteins have been clarified which makes it possible to '. design synthetic analogues for use in synthetic surfactants(3). A synthetic peptide (KU) which mimics the aminoacid pattern of SP-B has been found to be effective in treatment of infants with RDS(4).

Clinical Effects of Surfactant Therapy

Since the first randomized clinical trial (RCT) was published on surfactant replacement therapy in 1984, over 50 such trials have been reported. These RCTs have set a new standard for the evaluation of new therapeutic innovations in neonatology. Meta-analyses based on these RCTs have been published(5,6). These showed that by decreasing the need for oxygen and ventilatory support, pneumothorax (PTX) and 1081 mortality rates were reduced. The odds ratio for neonatal mortality was about 0.6; a 30-40% reduction in mortality is expected with surfactant therapy compared to controls. The efficacy of surfactant therapy demonstrated in RCTs has been shown to be translated into effectiveness in routine clinical use. Two observational studies, each involving over 4000-5000 infants weighing < 1500 g in 8 and 14 US centers respectively over a 3-5 year period, reported a 30% reduction in mortality after surfactant was introduced(7,8). These studies estimated that 50-80% of the largest drop in national infant mortality observed in 20 years in the USA (from 9.7 to 9.1 per 1000 births between 1989 and 1990) could be attributed solely to the introduction of surfactant therapy.

Surfactant therapy, by improving cardiorespiratory stability and oxygenation, should theoretically reduce non-pulmonary complications of prematurity such as intra- ventricular hemorrhage (IVH), necrotising enterocolitis and retinopathy of prematurity (RaP). However, published clinical trials have not consistently reported such benefits. Several trials did find a significant improvement in the rate of IVH(9,10) but a multicentre trial in Europe showed an increased risk of periventricular leukoma-Iacia(11). Electroencephalographic data showing periods of electrical silence lasting up to 20 minutes after administration of surfactant was thought, to be suggestive of cerebral ischemia(12) though this was found not to be directly related to alterations in blood gases or systemic circulation(13). A reduction in cerebral blood flow velocity (CBFV) after surfactant administration was interpreted as resulting from a fall in cerebral blood flow(14). A literature review suggested an overall increased risk' for germinal matrix and small IVH with surfactant therapy in infants of 600-750 g but no increased risk, for large IVH or intraparenchymal hemorrhage(15).

It concluded that the risk of IVH was related to the management of surfactant delivery. A learning curve operates in the use of surfactant in regards to anticipatory ventilator management to minimize cardio-respiratory instability following surfactant administration. With careful management of oxygenation,1 and ventilation as surfactant therapy becomes more familiar and customary  only small and transient perturbations occur with cerebral oxyhemoglobin concentration and cerebral blood volume measured by near infrared spectro-scopy(16).

Although a greater proportion of preterm infants are surviving with surfactant therapy, the risk of ROP among the survivors was lower than control infants(17-19). There was also no difference in their prevalence of neurodevelopmental disability, growth retardation and late respiratory or allergic disease(20-39). Indeed, 9 of the 10 double-blind RCT follow-up studies published in 1995(30-39) reported lower proportions and absolute numbers of infants with disability with significant gains in developmental stores in the surfactant-treated group. No adverse long- term effect on pulinonary function was identified as the two groups were found to have similar functional' residual capacity (FRC), tidal volume, compliance and time constant of the respiratory system(26-27). One study showed that surfactant treatment was associated with beneficial long-term effects on resistive airflow properties as reflected by a lower resistive work of breathing, improved expiratory reserves and reduced evidence of airflow obstruction(28).

Surfactant therapy has become routine in neonatal intensive care units except in some developing countries. A number of questions which have been raised on how to optimize its use. These include the method of administration, timing of therapy, amount to give and number of doses, type of surfactant, reason for non- responsiveness, cost implications, role of adjunctive therapies, and surfactant for indications other than RDS.

Method of Administration

It is generally recommended that the surfactant is delivered as a bolus directly into the lungs through an endotracheal tube over a IS-minute period, either via a catehter inserted into the endotracheal tube in 2-4 aliquots after disconnecting the infant from the ventilator, or into the side-port on the endotracheal tube adaptor without the need of removing the infant from the ventilator. Alternatively, surfactant can be delivered via a catheter inserted into the endotracheal tube via a side-port in the ventilator circuit adjacent to the endo-tracheal adaptor without. the need of removing the infant from the ventilator. Infants who remained connected to the ventilator during surfactant instillation have been shown in a RCT to experience less oxygen desaturation compared to those who were disconnected(40). The effect of maintaining a positive end-expiratory pressure during surfactant instillation has also been shown to result in more homogeneous distribution of surfactant within the lungs(41).

Several protocols have been advocated to vary the position of the infants during administration to enable uniform distribution of sufactant to the alveoli, but their efficacy has not been tested. Different protocols of positioning were found in a RCT not to affect the infants' clinical outcomes(40), and a study using an animal model showed that keeping the chest in the horizontal may result in the most even distribution of the surfactant in the two lungs(42). This simple practice may eliminate possible hypoxemia and cardiovascular disturbances associated with the handling of the infants using current positioning techniques. Rapid instillation over S minutes is not recommended as it results in an increase in CBFV and pCOz and requires closer monitoring of blood gases to maintain adequate ventilation compared to a slower IS-minute bolus(43,44). A transient increase in the peak inspiratory pressure is believed to overcome acute airway obstruction and minimize cardiorespiratory disturbances. Although surfactant has been given slowly via a syringe pump over 30 minutes or longer(45), this is no longer recommended as a slow infusion has been shown to result in very ununiformed surfactant. distribution and poorer response than bolus treament(46).

Ultrasonic nebulized surfactant treatment has been studied in preterm lambs(47). Compared with instilled surfactant, nebulized surfactant has a lower efficiency of deposition in the non-compliant lung, and a less homogeneous distribution. The potential for nebulized surfactant therapy may therefore be limited. Surfactant has been administered intra-amniotically to four human fetuses in utero with immature amniotic fluid indices(48). After birth, all the infants had an uneventful clinical course without respiratory distress. This was the only report on the possibility of using surfactant antenatally in the prevention of RDS.

Prophylactic versus rescue therapy

Surfactant can be given as prophylactic therapy (at or within 30 minutes of birth to those infants at risk of developing RDS, if possible even before the infant has breathed or received positive pressure ventilation) or as rescue therapy (given only when the diagnosis of severe RDS could be accurately made, usually at 3-6 hours after birth). The advantages of prophylactic over rescue therapy are that it may facilitate initial lung aeration and re- sorption of lung liquid, improve distribution of surfactant administered, and reduce barotrauma and thus leakage of inhibitor proteins. Eight RCTs involving over 2000 infants have -been published on"prophylac- tic versus rescue therapy(49-56). Three meta-analyses of 4 to 7 of these trials revealed a lower incidence of PTX and mortality with prophylactic therapy(57-59). These findings are relevant for infants born extremely preterm, as the majority of these RCTs exclusively 'enrolled infants < 30 weeks gestation. One of these RCTs also reported a reduced risk of severe IVH and ROP(55). However, the risk of chronic lung disease (CLD), or either CLD or death, does not differ according to treatment strategy. No neurodevelopmental advantage has been found on long-term follow-up with prophylactic therapy(60). The disadvantages of prophylactic therapy are that it destabilises the infant during initial resuscitation and that it may be an unnecessary treatment in some infants (thus leading to increased cost, increased risk of side effects and unnecessary endotracheal intubation). A prophylaxis policy would require nearly twice the number of infants treated(58) but result in a 39% reduction in neonatal mortality and an estimated saving of seven more infants for every 100 born < 32 weeks(59). Although there are advantages to prophylactic therapy among those born < 30-32 weeks gestation, the practical disadvantages result in a reluctance to routinely intubate and administer surfactant to all infants born of a certain gestation or birthweight.

It has been recommended that all infants < 32 weeks gestation should be treated with surfactant as soon as they are intubated(59). This would mean surfactant given at birth for such infant who need endotracheal intubation for resuscitation at delivery, or given as Soon as such infants require intubation for ventilation subsequently. Prophylactic therapy at birth has also been advocated for the most immature infants < 28 weeks or < 1000 g who had received no antenatal corticosteroid therapy(53). Delivery room administration of surfactant should be carried out only by someone experienced in neonatal resuscitation and surfactant administration. Other- wise, surfactant therapy should be given as soon as feasible when clinical signs of RDS develop, using the need for endotracheal intubation for mechanical ventilation as the criterion for surfactant administration(61). A protocol of early treatment (less than 2 hours of age in infants at high risk of RDS) compared to late treatment (at about 3 hours when RDS has established) has been shown in a RCT to result in significant 6% reduction in the risk of death or CLD(62). Delaying treatment until the infant has established RDS reduces the efficacy of surfactant therapy. The study showed that early therapy to an estimated 32 infants, when compared with therapy of established RDS, would prevent one infant from death and one infant from developing CLD, but would entail the additional use of surfactant in 8 infants.

In infants with moderate to severe RDS treated early with nasal continuous positive airway pressure (CP AP), administration of a faster acting natural surfactant during a short endotracheal intubation has been shown to reduce the need for subsequent mechanical ventilation (43% vs 85%)(62). Even when a synthetic surfactant was used with CPAP, it has been found to result in a reduction in the. duration of oxygen therapy and days of hospitalization(63). These studies demonstrated the benefits of surfactant therapy in infants treated with nasal CP AP rather than intubated for mechanical ventilation.

High Versus Low Dose

Preterm infants with RDS have a surfactant pool of 5 mg/kg. The surfactant dose is based empirically on animal experiments(64). In term newborn animals, the surfactant pool is about 100 mg/kg which is the usual dose used in surfactant therapy (100mg/kg in 3-5 ml/kg saline). One study using Surfacten from Japan showed that a single dose of 120 mg/kg, compared to 60 mg/kg, resulted in a shorter duration of oxygen and ventilatory therapy and a lower incidence of IVH and CLD(65). An- other study using another bovine surfactant, Alveofact, showed that 100 mg/kg is better than 50mg/kg, with significant improved oxygenation and a lower incidence of pulmonary interstitial emphysema(66). The instillation of surfactant down the en- dotracheal tube is generally well tolerated, especially with the first dose soon after birth and the infant is very sick and requires high peak inspiratory pressures which drives the surfactant rapidly into the lungs. Paradoxically, subsequent doses are less well tolerated because of airway obstruction, since the infants would have been stabilized on lower peak inspiratory pressures by then.

Single Versus Multiple Doses

Only about 20-30% of the surfactant dose can be recovered from the air spaces after 24 hours of ventilation(67). Most of it is found in the lung tissue, being recycled to maintain endogenous metabolic pathways with a turnover time of about 13 hours. Exogenous surfactant does not inhibit the synthesis and secretion of endogenous surfactant. It is of benefit to the surfactant deficient lung by providing it with phospholipids as substrate for the recycling pathways. The improvement after a single dose of surfactant is unsustained because its function can be inhibited by proteins in the small airways and al- veoli(68,69). Multiple doses can be used to overcome this functional inactivation. Three RCTs have shown that multiple doses, compared to a single dose of surfactant, resulted in a decrease in oxygen and ventilatory requirements, PTX and mortality(70-73). Clinical experience suggested that infants treated with a third dose were more preterm and had more neonatal complications such as patent ductus arteriosus (PDA)(73). A RCT has shown that a treatment regime which permits up to four doses has no additional benefit compared to a two-dose regime(61). Although a multiple dose strategy has been shown to be more effective, the optimal timing and indications of re-treatment remain uncertain.

Synthetic Versus Natural Surfactant

The difference between natural and synthetic surfactants appears to be that the synthetic surfactants have a delayed and modest effect 12-18 hours after administration, while oxygenation and lung function improves sooner and to a greater extent after naturalsurfactants(74). Lung function studies showed that the improvement in oxygenation after surfactant instillation is associated with an increase in FRC(74-76). However, even with the use of natural surfactants, dynamic lung compliance does not improve immediately(77-80) but early improvement in static lung compliance was found to correlate with increased oxygenation and recruitment of lung volume(75,81). In contrast, improvement in lung compliance and FRC occurs much later with the use of synthetic surfactants, usually following the clinical improvement in gas exchange(82-84), although pulmonary artery pressure falls by an hour after administration(44,85). It is now generally accepted that surfactant therapy improves aeration of the lung by stabilizing gas exchange units already being ventilated in addition to recruiting new units(74,86). There is probably an improvement in the efficiency in gas mixing and an increase in effective pulmonary blood flow to account for the improvement in oxygenation(74,86). Meta-analyses of surfactant RCTs(5,6) have shown that both synthetic and natural surfactants, whether given as prophylactic or rescue therapy, are effective in reducing the PTX and mortality: rate. Rescue therapy using synthetic surfactant and prophylactic therapy using natural surfactant appeared also to be effective in decreasing the incidence of CLD. Meta-analysis also suggested that the incidence of PDA with synthetic surfactant increased with prophylactic therapy and decreased with rescue therapy but it was not affected by natural surfactant. Although the problem of PDA with surfactant therapy remains controversial, it has been shown that surfactant reduces pulmonary vascular resistance over hour after administration, resulting in a decrease in pulmonary artery pressure(85) and an increase in the velocity of the left-to-right shunting through the PDA(87). In the presence of a large PDA after surfactant therapy, it has been shown that the heart of the preterm infant can mount a compensatory increase of cardiac output sufficient to maintain cerebral blood flow, but postductal organ blood flow cannot be maintained(88). A large left-to-right ductal shunt may lead to hemorrhagic pulmonary edema or "pulmonary hemorrhage". A retrospective study showed that pulmonary hemorrhage was associated with the presence of a clinical detectable PDA be- fore or at the time of the hemorrhage(89). Clinical pulmonary hemorrhage has been reported to increase significantly from 1% to 2% with surfactant therapy though the incidence of pulmonary hemorrhage at necropsy was reported to be similar in treated and control infants(90). Prophylactic use of prostaglandin inhibitors after surfactant therapy within 4 hours after birth has been shown to reduce the incidence of a hemodynamically significant PDA(91).
Nine RCTs involving over 3000 infants, of which three are large multicentre trials(92-94), compared the efficacy of a synthetic surfactant (Exosurf) and natural surfactant (Survanta and Infasurf). Meta-analysis of these studies(58,95) demonstrated with the use of natural surfactant, the oxygen requirement and mean airway pressure are lower for at least.3 days after treatment, the risks of PTX, ROP, neonatal mortality and the combined endpoint of death or CLD are lessened, and fewer doses are required. The use of natural surfactant instead of synthetic one was estimated to prevent one PTX for every 14 infants treated and to prevent one death for every 42 infants treated(96}.

There is some evidence that natural surfactants differ in their effects and are not equivalent. RCTs have shown that Curosurf and Infasurf, compared with Survanta, resulted in more rapid improvement in oxygenation and mean airway pressures up to 48 hours after treatment and a longer duration of effect, with a non-significant trend towards less PTX and large IVH(97,87). No significant long term benefits have yet been established. It is possible that the different surfactant preparations have different physical properties and physiological effects under in vitro and in vivo conditions(99). In the Curosurf study, its beneficial effects over Survanta might also have been explained by the fact the different initial dose of phospholipids used, namely 200 mg/kg for Curosurf and 100 mg/kg for Survanta.

Reasons for Non-Responsiveness

Up to 20% of infants believed to have RDS have little or no response to surfactant therapy. The reasons are multifactorial: (a) extremely preterm infants also have structural lung immaturity, (b) some may have other diseases such as pneumonia or pulmonary hypoplasia, (c) perinatal asphyxia is associated with a poor response, (d) pulmonary edema from lung damage or fluid overload results in protein inhibition and inactivation of surfactant, (e) pulmonary edema from left-to-right shunting through the PDA has the same effect, and if) there may be maldistribution of surfactant in the lungs, though this is not believed to be a common cause of a poor response(100). Clinical studies have shown that a poor response to surfactant is associated with severe RDS, perinatal asphyxia, infection, early PDA, pulmonary interstitial emphysema, PTX and excessive fluid and colloid intake(100-104). Extreme prematurity is not a factor for poor response as surfactant is known to, improve survival without increasing the proportion of disabled survivors in infants < 750 g born at. 23-24 weeks(105,106). It has been shown that infants with a poor immediate response to surfactant therapy has a higher mortality and PTX rates than the good responders(107-108).

Cost Implication

Surfactant is expensive. Nevertheless, surfactant therapy has been shown to result in a 25% reduction in daily hospital charges, a 52% reduction in daily ancillary charges (laboratory, x-ray, respiratory therapy) and, as a result of improved survival, a 22% reduction in hospital charges per survivor(109). It was reported that the cost of care with surfactant therapy declined by 10% among infants who survived and 30% among those who died(8). the lower cost of care. found among those who died shows that the availability of surfactant did not prolong dying for those in whom neonatal intensive care was futile. However, the improved survival rate also increases the average hospital stay, resulting in a cost per extra survivor of 14,000 pounds Sterling(110). Nevertheless, the latter study showed that when the cost per quality adjusted life year was calculated for surfactant therapy, it is one-half that for renal transplantation and one-tenth for that of coronary bypass surgery and hemodialysis.

The cost-effectiveness of surfactant therapy depends on the price of surfactant and whether it is used in large or small preterm infants(111). In the larger infants, surfactant therapy reduces medical costs because it shortens the duration of assisted ventilation and decreases the complications of RDS. In the smaller infants, their increased survival rate prolongs the need for neonatal intensive care among the additional survivors and increases overall medical costs. The savings from the use of surfactant in the larger infants should more than pay for the higher costs incurred in using surfactant in the smaller infants. Therefore, surfactant saves lives, reduces morbidity, and overall also saves money.

Combination Therapy

Non-randomized studies of combined antenatal corticosteroid therapy and surfactant therapy versus surfactant therapy alone have shown that combined therapy resulted in a significant reduction in the incidence of RDS, airleak, PDA, IVH and mortality(112-115). Although postnatal surfactant therapy by itself does not reduce the risk of IVH, antenatal corticosteriod therapy has been shown to have a protective effect on the brain unrelated to enhanced lung maturation(116). Data from animal experiments also indicate that antenatal corticosteroids and postnatal surfactant therapy have synergistic beneficial effects on neonatal lung function(117). Such combined therapy is estimated to result in 125 extra survivors out of 1000 infants born below 30 weeks gestation(118). The study also showed that the cost implications of combined therapy are that 7-11% extra intensive care beds are required. Further- more, 12-24% more special care beds are required after each type of intervention. Nevertheless, the combination of antenatal corticosteroid therapy and postnatal surfactant therapy is the most cost-effective because it produces the greatest number of survivors and the lowest number of intensive and special care days in hospital.

One study in preterm lambs has shown that surfactant therapy and subsequent high-frequency oscillatory ventilation (HFOV) leads to better oxygenation and alveolar expansion at comparable mean airway pressures compared to surfactant therapy followed by conventional mechanical ventilation(119). Another study in preterm monkeys found that the use of surfactant with HFOV, compared to either surfactant or HFOV alone, reduced alveolar proteinaceous edema(120). This reduction in lung injury may be beneficial in lowering the incidence of severity of CLD. The findings of the two studies await confirmation by RCTs in human preterm neonates.

Surfactant for Other Indications

Surfactant deficiency may be present in term infants with respiratory failure due to meconium aspiration syndrome (MAS), pneumonia and the adult type of acute RDS. Meconium inhibits the surface lowering properties of surfactant(121) and surfactant improves lung function in animals with experimentally induced MAS(122). When surfactant was 'given to a series of infants with MAS, it was reported that their oxygenation improved significantly within 2 hours, although 40% showed little or no response(123). The only RCT reported that surfactant-treated infants had improved oxygenation, less time on ventilation, fewer PTX, and less likely to reach the criteria for extracorporeal membrane oxygen- ation(124). Infants with MAS have been treated with tracheal bronchial lavage, using 15 ml/kg of natural surfactant saline suspension at a phospholipid concentration of 5 mg/ml(125). Large RCTs have been proposed to determine whether such a treatment has a role in the management of MAS. Several infants with respiratory failure from bacterial sepsis have also been treated with surfactant which resulted in a reduction in oxygen requirement and mean airway pressure, without apparent side effects(126).