Noninvasive respiratory support for preterm infants: An alternative to mechanical ventilation

Respiratory distress syndrome

Respiratory distress syndrome (RDS) is a disease of newborn infants, increasing in prevalence with decreasing gestational age. It is characterized by immature lung development and inadequate surfactant production. The lungs of affected infants may not expand normally immediately after birth, do not easily maintain a residual volume, and are at risk of atelectasis. Other factors also contribute to a loss of lung volume, including muscle hypotonia, a compliant chest wall, and slow clearance of fetal lung liquid. Repeated lung expansion, followed by atelectasis during expiration, leads to shearing forces and stretch injury that damage the airway and saccular alveolar epithelium and cause leakage of protein-rich fluid from the pulmonary capillaries. This leakage in turn inhibits any endogenous surfactant present. Damage to the lungs is exacerbated by mechanical ventilation (MV), high oxygen concentrations, and infection.

Apnea of prematurity

The pharyngeal airway of the preterm newborn is very compliant. The cartilaginous components are more flexible, and the fat-laden superficial fascia of the neck that stabilizes the upper airway of term infants is not well developed. The intrathoracic airways, including trachea, bronchi, and small airways, are similarly compliant and prone to collapse during expiration. The breathing patterns of very premature infants are frequently erratic and at times inadequate to maintain oxygenation (see Chapter 8 ). The causes of apnea of prematurity include hypoxia due to a reduced functional residual capacity, particularly in active sleep. Upper airway obstruction, alone or in combination with a central respiratory pause, accompanies most apneic events.

The role of continuous positive airway pressure

CPAP effectively supports the breathing of preterm infants through several mechanisms. It mechanically splints the upper airway, thereby minimizing obstruction and reducing apnea. Distension of the airways reduces resistance to air flow and so diminishes work of breathing. CPAP aids lung expansion and therefore reduces ventilation-perfusion mismatch and improves oxygenation. By preventing repeated alveolar collapse and reexpansion, CPAP reduces protein leak and helps conserve surfactant.

MV via an endotracheal tube (ETT) has been the mainstay of neonatal intensive care almost since its inception. Many lives have been saved by this technique, but its adverse effects are well documented. These include cardiovascular and cerebrovascular instability during intubation; complications of the ETT, including subglottic stenosis and tracheal lesions; infections, both pulmonary and systemic; and acute and chronic lung damage, primarily related to stretch mediated effects of nonhomogeneous tidal volume delivery at the cellular level. By avoiding the local mechanical problems of an ETT as well as those of volutrauma, the use of CPAP has been shown to improve outcomes for preterm infants. ^,

Nasal continuous positive airway pressure interfaces

Several interfaces have been developed for delivering nasal CPAP ( Fig. 6.1 ). Nasal prongs may be short, lying 1 to 2 cm inside the nose, or long, with the tip in the nasopharynx. They may be single or binasal. Both are associated with a leak around the prongs and nasal trauma. Nasal masks have been developed as an alternative interface to ameliorate these problems. An important determinant of effectiveness of nasal CPAP devices is their ability to transmit the pressure to the airways. This ability depends on the resistance to flow of the device, which in turn depends on the length and diameter of the prongs. In an in vitro comparison of popular devices, nasal mask interfaces had lower resistance than short binasal prongs, which in turn had lower resistance than single or long nasopharyngeal prongs.

The highest resistance was seen with the “RAM” nasal cannula, which has been an increasingly popular interface in recent years, particularly in North America. Several bench studies suggest significant limitations in the ability of the RAM nasal cannula to provide ventilator set peak or distending pressure during CPAP or NIPPV support. These studies show that only 60% to 70% of the set pressure is delivered to the proximal airway resulting in only minimal delivery of tidal volume. Two small RCTs have compared the RAM cannula to alternate interfaces for the delivery of NIPPV to preterm infants with RDS. Gokce et al. randomized 126 infants to receive NIPPV via RAM cannula or Hudson prongs. Infants in the RAM cannula group were more likely to require invasive ventilation (32.8% vs. 9.6%) and surfactant therapy (42.1% vs. 19.3%). In a noninferiority trial, Hochwald et al. randomized 166 preterm infants to receive NIPPV via short binasal prongs or mask, or via RAM cannula. In contrast to the advice from most nasal cannula manufacturers, the RAM cannula were fitted snuggly, aiming to fill about 80% of the nares. Surfactant administration was allowed in both groups and was not considered a NIPPV treatment failure. They found no difference in rates of endotracheal intubation but less nasal trauma in infants managed with RAM cannula.

Two recent meta-analyses have compared mask CPAP with prong CPAP. ^, The findings of both are consistent with the benchtop studies showing that the lower resistance of the mask interface delivers more effective CPAP levels to the infant. King et al. noted that while the quality of the evidence included in the review was low, the use of a nasal mask decreased the rate of nasal CPAP failure compared with nasal prongs. Likewise, Jasani and colleagues reported in a pooled analysis that treating nine infants with nasal mask rather than prongs would prevent one intubation. These authors noted that the mask was also associated with a reduction in the rate of nasal injury.

Nasal continuous positive airway pressure devices

There are several ways to deliver CPAP. The most common method is “bubble” CPAP, which uses a continuous gas flow that is directed past, and into, the infant’s nose, and an underwater seal at the distal end of the expiratory limb of the breathing circuit to generate circuit pressure. Tubing is inserted to a specific underwater depth to obtain the desired circuit pressure. The other type of continuous flow CPAP is ventilator-generated CPAP, where the ventilator expiratory assembly employs various methods to maintain a set pressure within the circuit that is applied to the infant.

The alternative to continuous flow CPAP is variable flow CPAP, where alterations to the gas flow result in changes in the delivered pressure. Since its description by Moa and colleagues in 1988, the variable flow nasal CPAP device—Aladdin nasal CPAP Infant Flow System (now called the Arabella; Hamilton Medical AG, Reno, NV), EME Infant Flow Nasal CPAP (CareFusion, San Francisco), or Infant Flow Driver (Electro Medical Equipment Ltd, Brighton, Sussex, UK)—has become widely used around the world. In vitro studies using models of neonatal ventilation have demonstrated less pressure variation and work of breathing with the variable flow device. Pandit and colleagues measured work of breathing in preterm infants with the use of respiratory inductance plethysmography and esophageal pressure monitoring. They demonstrated less work of breathing with variable flow CPAP than with continuous flow CPAP. The same group showed, in a crossover study, that the variable flow device led to better lung recruitment than either nasal cannula or continuous flow CPAP.

There are no large-scale RCTs investigating differences between CPAP devices. Several small studies exist, but there is little convincing evidence to support one form over another. A recent systematic review of bubble CPAP versus other CPAP delivery modes included 19 studies. The authors report that there was no difference in mortality or bronchopulmonary dysplasia (BPD) between modes, but there was less CPAP failure within seven days with bubble CPAP (risk ratio [RR] 0.75, 95% confidence interval [CI] 0.57–0.98). However, more nasal injury was observed with bubble CPAP than other modes (RR 2.04, 95% CI 1.33–3.14).

How much pressure should be used?

The purpose of nasal CPAP is to deliver a set pressure to the airways and lungs. If sufficient pressure is consistently achieved, the choice of device may not be important. A pressure of 5 cm H ₂ O is a traditional starting point. Some neonatal intensive care units (NICUs) hardly vary this pressure and claim good results. There is some evidence from the Cochrane Review of postextubation nasal CPAP that pressures <5 cm H ₂ O are ineffective in this setting. In their landmark publication on CPAP, Gregory and associates used pressures up to 15 mm Hg. A study of infants with mild RDS showed the highest end-expiratory lung volume and tidal volume, the lowest respiratory rate, and the least thoracoabdominal asynchrony occurred at a pressure of 8 cm H ₂ O, when compared with 0, 2, 4, and 6 cm H ₂ O. It is uncertain whether these results apply to more immature infants with more severe lung disease.

A study from Buzzella and colleagues suggests that CPAP pressures of 8 cm H ₂ O compared to 5 cm H ₂ O may be more effective at preventing extubation failure in extremely low birth weight infants. Additionally, the higher pressure seems to be well tolerated from a cardiovascular standpoint. Comparisons of cardiac output, cerebral circulation, and venous return suggest the higher pressure does not compromise cerebral circulation or venous return but may reduce pulmonary shunting across the ductus arteriosus. The choice of CPAP pressure should be tailored to the infant’s clinical condition. A baby with RDS, relatively stiff lungs, a high fraction of inspired oxygen (FiO ₂ ), and a chest X-ray showing opaque lungs may need a higher pressure to support lung volume than a baby with a low FiO ₂ being treated for apneic episodes. If CPAP is to be effective in infants with very low lung compliance, the pressure may need to be increased to 8 to 10 cm H ₂ O in increments of 1 cm H ₂ O and the effect observed. It is important to note, however, that high distending pressures, if used in a baby with compliant lungs, can interfere with pulmonary blood flow and cause overdistention, leading to carbon dioxide retention.

The optimal CPAP pressure is likely to vary between infants and over time as an individual infant’s pulmonary mechanics change. In the absence of evidence-based guidelines, we typically use CPAP pressures in the range 5 to 8 cm H ₂ O, adjusting them on the basis of FiO ₂ and clinical progress.

Nasal CPAP for babies with RDS or at risk of developing RDS

The focus of studies on the use of nasal CPAP in infants who have, or are at risk for, RDS has changed over the decades. From a clinical perspective, studies conducted before the availability of surfactant are of limited relevance in the modern neonatal intensive care era. In general, presurfactant trials comparing the risk and benefit of prophylactic CPAP versus no continuous distending pressure (i.e., oxygen by hood or standard nasal cannula therapy) showed that CPAP was associated with lower risk of treatment failure (RR 0.64, 95% CI 0.50–0.82), lower use of ventilatory assistance (RR 0.72, 95% CI 0.54–0.96), and lower overall mortality (RR 0.53, 95% CI 0.34–0.83). However, CPAP was associated with increased risk of pneumothorax (typical RR 2.48, 95% CI 1.16–5.30).

CPAP in the “surfactant era”

Surfactant is the most comprehensively evaluated treatment in neonatology. Initial randomized trials of surfactant were done nearly 30 years ago, when early CPAP was not commonly used for very preterm infants, antenatal corticosteroids were given to only 10% of eligible mothers, and rates of neonatal mortality and morbidity were much higher than at present. Surfactant therapy was provided via an ETT, and whether given prophylactically or as treatment, surfactant reduced mortality and the combined outcome of death or chronic lung disease. ^, Surfactant therapy appeared more beneficial when given early in the course of RDS. It became common practice for all very preterm infants to be intubated in the delivery room for surfactant administration. In the past few years, a number of randomized trials have evaluated “less invasive” approaches to surfactant administration in conjunction with CPAP support.

Is CPAP a suitable alternative to routine intubation of very preterm infants at birth?

More than a decade ago, several groups reported their experience following policy change from early intubation to early nasal CPAP, ^, describing lower mortality and morbidity in the CPAP-treated group. Since then, several RCTs have compared intubation in the delivery room with early nasal CPAP. The “COIN” trial randomly assigned 610 breathing infants at 25 to 29 weeks’ gestation to CPAP or intubation and ventilation if they manifested signs of respiratory distress at 5 minutes after birth. Surfactant was administered to intubated infants at the discretion of the treating clinician. There were no important differences in the rates of death or BPD between the groups. CPAP halved the intubation rate, decreased the risk of the combined outcome of death or the need for oxygen therapy at 28 days, and resulted in fewer days of MV. However, the CPAP group had a higher rate of pneumothorax (9% vs. 3% in the intubated group).

The “SUPPORT” trial enrolled 1316 infants between 24 and 27 weeks’ gestation who were randomly allocated before birth to immediate CPAP or intubation in the delivery room. Intubated infants were also treated with surfactant within 1 hour after birth. The rates of death or BPD did not differ significantly between the groups after adjustment for gestational age, center, and familial clustering. Although more than 80% of infants in the CPAP group ultimately required intubation and MV, only about one-third were intubated in the delivery room. Overall, infants randomly assigned to CPAP were less likely to be intubated, less likely to receive postnatal corticosteroids, had fewer days of MV, and were more likely to be alive without MV by day 7.

A Vermont Oxford Network trial compared three approaches to initial respiratory management in 648 very preterm infants born at 26 to 29 completed weeks’ gestation: (1) prophylactic surfactant followed by a period of MV, (2) prophylactic surfactant with rapid extubation to CPAP, or (3) initial management with CPAP and selective surfactant treatment. The primary outcome was the incidence of death or BPD at 36 weeks’ postmenstrual age. Both prophylactic surfactant with rapid extubation to CPAP and initial management with CPAP reduced the RR of death or BPD compared with prophylactic surfactant followed by a period of MV. About half the infants managed with initial CPAP avoided MV and surfactant. Mortality and other adverse outcomes were similar between the groups.

A meta-analysis including these RCTs and a fourth trial found a significant reduction in the combined outcome of death or BPD, at 36 weeks’ corrected gestation for infants treated with early CPAP: RR 0.91 (95% CI 0.84–0.99); number needed to treat: 25 infants. The trials show that nasal CPAP can be safely used from birth in very preterm infants and that nearly half of such infants may not need ventilation or surfactant treatment. Current evidence supports the use of early nasal CPAP as the initial respiratory mode for most extremely preterm infants in an effort to prevent intubation/MV and subsequent BPD.

Is CPAP with early intubation for surfactant and brief mechanical ventilation better than CPAP alone?

A systematic review has investigated whether early, brief intubation for surfactant administration (the Intubation, SURfactant, Extubation, or ‘INSURE’ procedure) followed by extubation to nasal CPAP was better than nasal CPAP and selective intubation, surfactant, and continued MV. Meta-analysis of the six studies identified showed that intubation, ventilation, and early surfactant therapy followed by extubation to nasal CPAP ventilation was associated with lower rates of later MV (RR 0.67, 95% CI 0.57–0.79), air leak syndromes (RR 0.52, 95% CI 0.28–0.96), and supplemental oxygen requirement at 28 days of age (RR 0.51, 95% CI 0.26–0.99). There was no difference in supplemental oxygen requirement at 36 weeks’ postmenstrual age. The early surfactant group received about 60% more surfactant than the selective intubation group. Stratified analysis of FiO ₂ at study entry suggested that a lower treatment threshold (FiO ₂ <0.45) reduced air leaks and oxygen requirement at 28 days of age compared with a higher threshold (FiO ₂ ≥0.45).

Further studies have been published since this review. The Colombian Neonatal Network enrolled infants of 27 to 31 weeks’ gestation, who were receiving oxygen and had increased work of breathing, at 15 to 60 minutes after birth. The infants were treated with bubble nasal CPAP 6 cm H ₂ O and then randomly allocated to either nasal CPAP plus surfactant (n = 141) or nasal CPAP alone (n = 137). The primary outcome was the need for MV started because either the FiO ₂ was > 0.75 or the partial pressure of carbon dioxide (PaCO ₂ ) was > 65 mm Hg. The nasal CPAP plus surfactant group received two doses of surfactant (Survanta ^TM , Abbott Nutrition, Abbott Park, IL, USA), 2 minutes apart, and were then extubated, if possible, to nasal CPAP 6 cm H ₂ O. The need for MV was significantly lower in the nasal CPAP plus surfactant group (26% vs. 39%, RR 0.69 [95% CI 0.49–0.97]), although all babies who had received surfactant had been temporarily intubated and ventilated. Mortality, BPD, duration of MV, and oxygen therapy did not differ between the groups. There were fewer air leaks in the group receiving surfactant.

The “CURPAP” trial randomly assigned 208 infants born at 25 to 28 weeks’ gestation who were not intubated within 30 minutes of birth to either (1) intubation, surfactant (Curosurf ^TM , Cornerstone Therapeutics, Inc., Cary NC, USA) and extubation within an hour (if possible) to nasal CPAP or (2) nasal CPAP with early selective surfactant. Infants were intubated for MV when the FiO ₂ exceeded 0.40, the infant had four episodes of apnea per hour, or for PaCO ₂ >65 mm Hg. There were no significant differences between the groups in the need for MV at 5 days of life, death or BPD, or the rate of pneumothorax.

These trials show that outcomes in infants stabilized in the delivery room with either nasal CPAP or prophylactic surfactant and extubation to nasal CPAP appear to be similar to those in infants managed with surfactant followed by MV.

Less invasive or minimally invasive surfactant administration

Over the last decade, various techniques of administering surfactant to spontaneously breathing preterm infants on noninvasive respiratory support without using an ETT, known as less invasive surfactant administration (LISA) or minimally invasive surfactant therapy (MIST), have been described and studied ( Fig. 6.2 ). RCTs of LISA are heterogeneous, with differences in method of delivering the intervention, the mode of noninvasive respiratory support used, and the study population. Pooled analyses of these RCTs found that LISA reduced rates of MV, the composite outcome of death or BPD at 36 weeks, and BPD in survivors.

The recent Cochrane Review included 2164 preterm infants in 16 studies and compared LISA techniques with endotracheal intubation and surfactant. The management of infants in the control group varied between studies: either early extubation after surfactant, delayed extubation after surfactant, or continuation of nasal CPAP and rescue surfactant administration only when prespecified criteria were met. Pooled analysis showed that LISA decreased the risk of the composite outcome of death or BPD: RR 0.59 (95% CI 0.48–0.73); the need for intubation within 72 hours: RR 0.63 (95% CI 0.54–0.74); severe intraventricular hemorrhage: RR 0.63 (95% CI 0.42–0.96); death during first hospitalization: RR 0.63 (95% CI 0.47–0.84); and BPD among survivors: RR 0.57 (95% CI 0.45–0.74). There was no significant difference in risk of air leak requiring drainage. Only two of the included RCTs have reported longer-term outcomes at 2 years, ^, and more evidence of longer-term effects is required. Herting et al. found no statistically significant differences in weight, length, or neurodevelopmental outcome. However, Mehler and colleagues reported lower incidences of a psychomotor development index <70 in infants receiving LISA and a lower incidence of mental developmental index <70 in the subgroup of more mature infants born 25 to 26 weeks’ gestation who received LISA.

Since that Cochrane Review, Dargaville and colleagues have completed the largest RCT of the “MIST” technique to date: the “OptiMIST-A” trial. The trial enrolled preterm infants 25 to 28 weeks’ gestation supported with nasal CPAP and receiving an FiO ₂ ≥ 0.30 within 6 hours of birth. Infants were randomized to either receive surfactant (200 mg/kg of poractant alfa ) via a 16-gauge vascular catheter under direct vision of the vocal cords or to continue CPAP without routine surfactant. Nasal CPAP was continued thereafter in both groups unless specified intubation criteria were met. OptiMIST-A was unique in that the control group underwent a sham procedure so that caregivers and outcome assessors were blinded to treatment allocation. The planned sample size was 606 infants. Ultimately, the trial, conducted at 33 centers around the world, enrolled 485 preterm infants.

In the OptiMIST-A trial, the primary outcome of death or BPD occurred in 43.6% of the MIST group and 49.6% of the control group: RR 0.87 (95% CI 0.74–1.03), P = 0.10. The overall incidence of death before 36 weeks’ postmenstrual age did not differ significantly between the groups. However, there was an interaction between death and gestational age: the more immature subgroup of infants born 25 to 26 weeks’ gestational age who received LISA seemed to have a higher rate of death. BPD was lower in the MIST group: RR 0.83 (95% CI 0.70–0.98). The incidence of other adverse events was similar between groups. There were some additional apparent benefits of MIST, including reductions in the incidence of intubation within 72 hours of birth and at any time, pneumothorax requiring drainage, patent ductus arteriosus requiring medical therapy, and the need for oxygen at home after discharge. The requirement for surfactant therapy via ETT was halved after treatment with MIST (32.8% vs. 68.4%), and MIST reduced the duration of MV (1 vs. 4 days), CPAP (17 vs. 22 days), and all forms of mechanical respiratory support (40 vs. 45 days).

The results of the OptiMIST-A trial, combined with evidence from earlier meta-analyses of RCTs, suggest important benefits from LISA/MIST. However, some caution is warranted in the most immature infants born 25 to 26 weeks’ gestation given the signal for higher mortality. Overall, the evidence suggests that LISA/MIST may be preferable to routine endotracheal intubation for surfactant therapy in infants born ≥25 weeks’ gestation. Of course, this procedure still requires skill with laryngoscopy. These findings are now being incorporated into clinical practice.