Part 4: Disorders Specifically Related to Preterm Birth

INTRODUCTION

Neonatal respiratory distress syndrome (RDS) is the most common cause of respiratory failure in preterm infants. Its incidence is inversely related to gestational age and increases to as high as 95% in infants born at 22 to 24 weeks of gestation. Over time, the outcome of infants suffering from RDS has changed dramatically. Forty years ago, approximately 50% of infants with RDS died. However, significant improvements in both prevention and treatment of RDS over the last 4 decades have markedly reduced mortality. As a consequence, the focus of improving outcome has now shifted from mortality to preventing (pulmonary) morbidity following RDS and its treatment.

PATHOPHYSIOLOGY

The main feature of RDS is compromised lung function caused by both structural and biochemical immaturity of the lung.

Structural Immaturity

Lung development during fetal life occurs in different stages. Following organogenesis in the first 2 stages (embryonic and pseudoglandular) of lung development, the canalicular stage, starting at approximately 16 weeks after conception, is the first step in lung differentiation, involving formation of an actual air-blood barrier. Differentiation continues in both the saccular and alveolar stages of lung development starting at 24 and 36 weeks’ postconceptional age, respectively. This means that, at the time of birth, lung development of most preterm infants is still at the saccular stage, resulting in a reduced surface area for gas exchange and limited diffusion capacity due to thickened membranes at the air-blood interface.

Biochemical Immaturity

The hallmark of RDS is a deficiency of pulmonary surfactant, a complex mixture of lipids (90%) and proteins (10%) that is synthesized in alveolar epithelial type II cells. Type II cells are 1 of the 2 epithelial cell types that line the alveolus. The most important function of surfactant is lowering of the alveolar surface tension, the force directed from the wall to the center of the alveolus at the air-liquid interface. This function is mainly attributed to the surfactant phospholipid dipalmitoylphosphatidylcholine and the surfactant hydrophobic proteins B and C. The hydrophilic surfactant proteins A and D play a role in innate host defense. Synthesis and storage of surfactant begins at about 16 weeks’ gestation, and lung homogenates have high concentrations of surfactant by 20 weeks. However, surfactant is not secreted until later, appearing in amniotic fluid at approximately 28 weeks’ gestation, although this may vary greatly among individuals. This explains why some infants with a gestational age of less than 30 weeks do not develop neonatal RDS while other infants, born at a more advanced gestation, do.

The high alveolar surface tension accompanying surfactant deficiency will increase the elastic recoil forces of the lung and decrease compliance of the respiratory system. As a result, preterm infants with RDS need to create large transpulmonary pressures to establish an adequate tidal volume. The absence of surfactant will also compromise lung volume stability, especially at the end of expiration. This makes the lung prone to a low end-expiratory lung volume and atelectasis, partly because the highly compliant chest wall is unable to counteract the increase in elastic recoil forces of the lung. A low end-expiratory lung volume may further compromise lung compliance and increase airway resistance and pulmonary vascular resistance. Due to the fact that the hypoxic pulmonary vasoconstriction response is often not functional in newborns, perfusion of collapsed sacculi-alveoli, also referred to as intrapulmonary right-to-left shunt, increases and results in (severe) hypoxemia.

PATHOLOGY

At postmortem examination, the lungs from infants with RDS are firm and airless. Atelectasis is striking on gross inspection; when the lungs are fixed in inflation, only the airways and a few alveolar ducts are air filled (Fig. 57-1). Diffuse atelectasis and dilated terminal bronchioles and alveolar ducts lined with a homogenous hyaline-staining material characterize the microscopic picture (Fig. 57-2). The hyaline membranes are plasma clots containing fibrin, other plasma constituents, and cellular debris. The small pulmonary arterioles appear constricted. There is congestion of pulmonary capillaries and veins and an increase in pulmonary water with dilation of the lymphatics. Because of these histological features, RDS was initially named hyaline membrane disease.

Interstitial air leaks are common, and collections of air are often seen around small airways and vessels (Fig. 57-3). In some cases, the alveoli and hyaline membranes contain red cells. Electron microscopic examination shows degeneration of epithelial and endothelial cells and rupture of the basement membranes. If death occurs after 3 or 4 days of respiratory distress, the hyaline membranes are fragmented and numerous macrophages appear in the intra-alveolar spaces. The pulmonary interstitium is widened and filled with round cells and fibroblasts. After the first week, there is a proliferation of alveolar epithelial type II cells and capillaries. In severe cases, chronic changes occur, including metaplasia of the bronchiolar epithelium and interstitial fibrosis.

PREDISPOSING FACTORS

As mentioned, the incidence of RDS is inversely related to gestational age. But there are also other factors that increase the risk of neonatal RDS. It is twice as common in males than in females at all gestational ages and is more common in white infants. Delivery by cesarean section, particularly if performed before the onset of spontaneous labor, is an independent risk factor as well. Infants of diabetic mothers are 5 times more likely to develop RDS than infants of nondiabetic mothers with the same gestational age, sex, and mode of delivery. A higher maternal age also predisposes for RDS. Finally, the second-born twin is more likely to be affected, and a family history of RDS increases the risk for any given premature infant.

On the other hand, complications of pregnancy, such as pregnancy-induced hypertension, chronic maternal hypertension, premature rupture of membranes, intrauterine infection, and subacute placental abruption, all decrease the incidence of RDS. Infants born to mothers addicted to narcotics are also at less risk for developing RDS.

Predicting Lung Maturity and Respiratory Distress Syndrome

Lung fluid moves from the fetal lung into the amniotic cavity. Once surfactant begins to be secreted by the alveolar epithelial type II cells in the fetus, this lung fluid transports secreted surfactant from the alveoli to the amniotic fluid. The concentration of surfactant in amniotic fluid reflects the amount of surfactant available at the alveolar surfaces and thus the potential stability of the respiratory units and risk of RDS. The concentrations of the phospholipids lecithin and sphingomyelin are equal in amniotic fluid in midgestation, but after 34 weeks, there is twice as much lecithin as sphingomyelin. This knowledge led to the widespread use of the lecithin-sphingomyelin (L/S) ratio as an antenatal developmental index of biochemical maturity, to predict which fetuses will develop RDS when delivered.

Several other amniotic fluid tests for assessing lung maturity are now widely available, including the measurement of phosphatidylglycerol, a relatively surfactant-specific phospholipid, and the simple and rapid foam stability or shake test. Although these tests can provide very useful clinical guidance in certain situations, they are all limited by a high false-negative rate secondary to the accumulation of normal intracellular surfactant stores well before amniotic fluid levels of surfactant change. For this reason, amniotic fluid analysis for lung maturation are not routinely used in clinical practice.

Some investigators have also attempted to predict RDS during the first minutes after birth. At this time, gastric contents mainly comprise lung fluid or swallowed amniotic fluid. Several tests can assess the presence or absence of surfactant in the gastric aspirate: lamellar body count, stable microbubble test, and gastric aspirate shake test. Of these, the lamellar body count and the stable microbubble test seem most promising, with reported positive and negative predictive values ranging from 58% to 85% and 75% to 100%, respectively. However, there is currently only 1 randomized trial comparing surfactant treatment based on lamellar body count versus oxygen need, showing no difference in important clinical outcomes and the use of surfactant. The limited amount of prospective evidence is probably the most important reason why gastric content analysis to predict RDS has not found its way into clinical practice.

CLINICAL FEATURES

Some infants with RDS fail to expand their lungs at birth despite vigorous inspiratory efforts, and have respiratory distress from delivery. Others initially inflate their lungs but develop progressive atelectasis and increasingly labored breathing in the first few hours of life. The characteristic clinical features of infants with neonatal respiratory distress are an expiratory grunt, tachypnea, intercostal and sternal retractions, and cyanosis. Grunting respiration is caused by a prolonged expiratory effort against a partially closed glottis. It is usually preceded by a strong inspiratory effort during which the intrathoracic pressure drops well below atmospheric pressure. During the prolonged expiration, intrathoracic pressure is maintained above atmospheric pressure. Infants do not grunt with every breath, and those with severe disease grunt most frequently. By maintaining a positive intrapulmonary pressure during most of the respiratory cycle, grunting probably helps to prevent atelectasis, which is 1 of the hallmarks of RDS.

Due to their decreased lung compliance, infants with RDS need to generate large transpulmonary pressures to establish an adequate tidal volume. The large negative intrathoracic pressures generated as the infant attempts to inflate its lungs cause the soft tissues of the chest cage to retract. These retractions are particularly notable in very small preterm infants as they have highly compliant chest walls. With severe respiratory distress, the lower sternum may be pulled in almost to the vertebral column by the forceful contraction of the diaphragm. Infants breathe primarily with the diaphragm and have very compliant chest walls; as a result they often have paradoxical breathing movements. Thus, as the chest caves in, its circumference becomes smaller while the abdominal circumference increases. As a result, breathing becomes less efficient and tidal volumes become smaller. By increasing their respiratory rate, the infants try to maintain an adequate minute volume as much as possible.

The combination of reduced alveolar ventilation and stability explains that breath sounds are often diminished in intensity and have a harsh, tubular quality. Occasionally, there are fine rales, particularly in those infants born by cesarean section who may have excessive lung liquid.

Cyanosis is an early sign of RDS and, as the disease progresses, may be present even when an infant breathes 100% oxygen. As the lungs become more difficult to ventilate, the work of breathing increases, the infant tires, and arterial carbon dioxide tension rises. At the same time, the hypoxemia and diminished peripheral blood flow cause metabolic acidosis as lactic acid accumulates. With the development of acidosis, potassium leaves the cells and its concentration in serum rises, in some instances to very high levels.

Urine output is usually diminished early in the course of the disease, and the infants may become progressively edematous. Some infants, especially very low–birth-weight infants, have systemic hypotension, peripheral pallor, slow capillary filling, and hypothermia when not treated in time with sufficient respiratory support and exogenous surfactant.

Postnatal Lung Maturation

It takes only 1 to 2 days following birth for an immature lung to mature as it responds to the surge of glucocorticoids and β-adrenergic compounds released by the stress of delivery. Glucocorticoids increase surfactant synthesis, and β-adrenergic stimulation promotes its secretion. At the same time, structural changes occur in the lung. Thin-walled respiratory units develop, and the number of capillaries increases. With these changes, the signs and symptoms of respiratory distress usually subside after 48 to 72 hours of life. Recovery is usually heralded by diuresis. Clinical improvement is accompanied by a rapid fall in pulmonary vascular resistance and a rise in systemic arterial pressure. In some infants, particularly the least mature with birth weight less than 1500 g, this may permit development of a large shunt from the aorta through a patent ductus arteriosus to the pulmonary artery. In these infants, recovery may be interrupted by the development of pulmonary edema.

Radiographic Features of Respiratory Distress Syndrome

The radiographic appearance of the lungs in infants with RDS is characterized by reduced aeration with a diffuse reticulogranular pattern of increased density, usually uniform in distribution (Fig. 57-4) but occasionally more marked in the bases or on 1 side. The densities are due to miliary atelectasis and interstitial edema. Lung volumes are small, and even radiographs taken after a maximal inspiration rarely show the diaphragm to be below the eighth or ninth interspace. The heart is usually normal in size, although it often appears large because of the large thymic shadow and decreased lung volume. The radiographic appearance of the lung can be altered by treatment with nasal continuous positive airway pressure (CPAP) or invasive mechanical ventilation with positive end-expiratory pressure. Later in the course of the disease, pulmonary edema, air leaks, or pulmonary hemorrhage can also affect the radiological appearance.

DIFFERENTIAL DIAGNOSIS

Transient Tachypnea of the Newborn

It may be difficult to distinguish RDS from normal, physiological pulmonary transition after birth. During this transition, fluid needs to be cleared from the lung and an air-filled functional residual capacity need to be created. Preterm infants often show signs of respiratory distress, including grunting and retractions. Furthermore, some infants may be oxygen dependent. In some infants, pulmonary transition may take several hours, a condition also referred to as transient tachypnea of the newborn, or “wet lung.” However, in contrast to RDS, these symptoms are usually not progressive but instead resolve within the first 24 to 48 hours. Furthermore, the chest x-ray shows hyperaeration of the lungs, perihilar streaking and fluid in the interlobar fissures.

Congenital Pneumonia

Infants with congenital pneumonia, especially group B streptococcus, may also present with signs of respiratory distress that can be clinically and radiologically indistinguishable from RDS.

Surfactant Dysfunction

Surfactant function also may be compromised by genetic disorders. For example, surfactant protein B deficiency and ABCA3 mutations can have a striking resemblance to RDS. However, and in contrast to infants with RDS, signs of respiratory distress and cyanosis do not subside over the first 3 days of life and the response to exogenous surfactant treatment is only temporary.

PREVENTION

Neonatal RDS is associated with incomplete development of the lung at the time of birth. Thus, premature delivery should be delayed, where possible. If this cannot be avoided, additional efforts should be made to accelerate lung maturation. Animal studies have shown that antenatal administration of glucocorticoids accelerates both structural and biochemical (surfactant) lung maturation.

In 1972, Liggins and Howie translated these findings to humans when they reported that administration of betamethasone to women in premature labor at least 2 days before delivery significantly reduced the incidence of respiratory distress. Numerous studies following this pivotal publication have confirmed that antenatal glucocorticoids reduce mortality and perinatal morbidities such as RDS, intraventricular hemorrhage, and necrotizing enterocolitis. Treatment consists of 2 doses of 12 mg of betamethasone given intramuscularly 24 hours apart, or 4 doses of 6 mg of dexamethasone given intramuscularly 12 hours apart. The benefits of antenatal steroid typically begin at around 24 hours after initiation of therapy and lasts 7 days. Current guidelines recommend treatment with antenatal glucocorticoids if preterm delivery is imminent between 24 and 34 weeks of gestation. Although observational cohort studies also suggest benefits of antenatal glucocorticoids between 22 and 26 weeks of gestation, the randomized evidence to support this is limited. In addition, a recent multicenter randomized trial found antenatal betamethasone of benefit for late preterm infants born between 34 and 37 weeks of gestation as well.

Repeated courses of antenatal glucocorticoids have been a matter of debate because of possible adverse effects on fetal growth. However, a systematic review of 10 randomized controlled trials clearly showed benefit of a repeated course of antenatal glucocorticoids without evidence for clinical relevant adverse effects on growth or long-term neurodevelopment. As a result, most clinicians have adopted a repeat course of antenatal glucocorticoids but only if preterm delivery is imminent.

TREATMENT

Assisting the Pulmonary Transition after Birth

At the time of birth, most of the lung is still fluid filled, and the infant clears the fluid by creating large expiratory pressures, also referred to as expiratory braking. Clinically, expiratory braking often presents itself as grunting and retractions. Due to the fact that fluid clearance takes time, most infants are cyanotic at birth and it takes them up to 10 minutes to attain an oxygen saturation (SpO₂) of 90% or more. It is clear that normal physiological pulmonary transition may be difficult to distinguish from RDS. For this reason, the initial treatment should be similar and consist of supporting the infant without causing injury to the vulnerable lungs.

Respiratory Support

If the infant is spontaneously breathing, the expiratory phase of respiration should be supported with CPAP. If spontaneous breathing is absent or inadequate, positive-pressure ventilation should be initiated. There is insufficient evidence to support the use of a prolonged or sustained inflation. The fraction of inspired oxygen should be started between 0.21 and 0.30 and titrated according to the SpO₂ nomograms. If clinical signs of respiratory distress and cyanosis persist after the transitional phase, RDS is very likely and appropriate treatment should be started.

The primary aim in treatment of RDS is to restore lung function and gas exchange while avoiding treatment-induced injury to the vulnerable lungs of preterm infants. As low end-expiratory lung volume is the most prominent feature of RDS, stabilizing end-expiratory lung volume should be the first goal of treatment. In 1971, Gregory and colleagues showed the beneficial effects of CPAP on oxygenation. This improvement in oxygenation has been linked to a higher end-expiratory lung volume (Fig. 57-5). Recent studies have also shown that primary CPAP is the preferred mode of respiratory support after birth, as it reduces the need for invasive mechanical ventilation, surfactant treatment, and the risk of death or bronchopulmonary dysplasia. CPAP is usually administered via the nasal route using either a mask or prongs. Although the optimal level of CPAP still needs to be established, a continuous positive pressure of up to 6 to 10 cmH₂O can be tolerated by the nasal route. Minimal invasive surfactant treatment (see below) and nasal intermittent positive-pressure ventilation might further enhance the success of CPAP in patients with RDS. However, there is insufficient evidence to support the use of humidified high-flow nasal cannula as an alternative primary mode for CPAP in infants with RDS.

In cases in which nasal noninvasive support is not effective in restoring lung function and gas exchange or if severe apnea occurs, intubation and mechanical ventilation should be started using either conventional modes or high-frequency ventilation. Again, maintaining an optimal lung volume should be the primary ventilation strategy when using these modes.

Surfactant Replacement

Surfactant deficiency, 1 of the main features of RDS, can be treated by exogenous surfactant replacement. Administering exogenous surfactant into the lungs will lower the surface tension, improve lung compliance, and stabilize end-expiratory lung volume. Since the mid-1980s, a large number of carefully controlled clinical trials have been published showing that therapy with surfactant is safe, that it clearly reduces mortality from RDS, that it decreases the incidence of air leaks, and that in small infants it decreases the incidence of intracranial hemorrhage.

Surfactant Products

Many different formulations of surfactant preparations are commercially available. Surfactant formulations fall into 3 major classes: (1) animal surfactants purified from animal lungs or bronchoalveolar lavage, (2) synthetic surfactants consisting primarily of phospholipids, and (3) synthetic surfactants that combine phospholipid mixtures with recombinant surfactant proteins or peptides based on surfactant protein structures. A recent systematic review of trials comparing animal-derived surfactant extract to synthetic surfactant consisting only of phospholipids showed that both are effective in the treatment and prevention of RDS. However, the benefits related to the need for ventilator support, the rate of pneumothoraces, and death were more pronounced when using animal-derived products. Although 2 trials of protein-containing synthetic surfactants compared to animal-derived surfactant extract showed no difference in death and chronic lung disease, further well-designed studies of adequate size and power are needed to confirm and refine these findings.

Timing of Surfactant Therapy

There are 2 clearly defined treatment strategies for administration of surfactant: (1) prophylactic therapy, which requires the surfactant preparation to be instilled in the infant’s trachea shortly after birth, preferably in the delivery room; and (2) rescue therapy, which is designed to treat infants with established RDS. The latter can be divided into early (< 2 hours after birth) or delayed (> 2 hours after birth) rescue treatment. It is important to acknowledge that most randomized controlled trials investigating the optimal timing of surfactant were performed in an era when antenatal corticosteroids were still being introduced in clinical care and primary intubation and mechanical ventilation (rather than the less invasive use of CPAP) was the preferred mode of respiratory support after birth. Under these circumstances, it was shown that mortality was lower in infants treated prophylactically compared with selective treatment, and in infants treated early compared with delayed rescue treatment.

More recent studies comparing invasive mechanical ventilation combined with prophylactic surfactant treatment to primary nasal CPAP and, if indicated, rescue surfactant treatment have shown that prophylactic treatment is associated with an increased risk of death or bronchopulmonary dysplasia (BPD). For this reason, prophylactic surfactant treatment is no longer recommended when using primary nasal CPAP at birth.

Modes of Surfactant Delivery

Historically, surfactant was administered via an endotracheal tube during invasive mechanical ventilation. As the latter has been identified as a major risk factor for ventilator-induced lung injury and subsequent BPD, alternative, less invasive strategies of surfactant administration have been investigated. During the Intubate SURfactant and Extubate (INSURE) protocol, infants are intubated and surfactant is administered followed by immediate extubation to nasal CPAP. Compared with traditional surfactant treatment INSURE reduces the need for mechanical ventilation, but the effect on other clinical outcomes such as mortality and BPD at 36 weeks postmenstrual age is limited. With the aim of avoiding intubation altogether, some investigators explored the feasibility of administering surfactant during spontaneous breathing on nasal CPAP. Several studies have now shown that this route of administration, often referred to as less-invasive surfactant application (LISA) or minimally invasive surfactant therapy (MIST), reduces the need for invasive mechanical ventilation and death or BPD at 36 weeks postmenstrual age. Other modes of surfactant delivery, including the use of a laryngeal mask or nebulized surfactant administration, still need further evaluation.

General Support of Infants with RDS

The infant should be cared for in a warm, neutral thermal environment. Initially, fluid intake should be restricted to 60 to 80 mL/kg and thereafter titrated on diuresis and serum sodium concentration. If the arterial pressure is low in the early course of the disease and if peripheral circulation is inadequate, as judged by poor capillary filling, the circulating volume may be increased with normal saline. Infusion of dopamine (5–10 μg/kg/min) may help maintain the circulation and avoid excessive fluid administration, especially in the very low–birth-weight infant. Given that the clinical presentation and radiographic findings of RDS and neonatal pneumonia—particularly secondary to group B streptococcus—may initially be indistinguishable, there should be a low threshold for starting antibiotics in infants presenting with respiratory distress after birth.

COMPLICATIONS AND OUTCOMES

As a result of antenatal steroids, exogenous surfactant and gentler (noninvasive) modes of respiratory support, the majority of preterm infants nowadays survive the phase of early respiratory failure due to RDS. The most common short-term complication of RDS is pulmonary air leak. In the longer term, BPD can develop, especially if infants need a longer period of invasive mechanical ventilation. BPD itself results in prolonged need for supplemental oxygen, a higher incidence of respiratory illness with wheezing in the first years of life, and an increased risk of adverse neurodevelopmental outcome.

SUGGESTED READINGS

INTRODUCTION

The ductus arteriosus (DA) is a vital component of the fetal circulation, when the placenta is the source of oxygen to the fetus. The DA provides a conduit for blood to bypass the high resistance pulmonary vascular bed and shunt toward the descending aorta and low-resistance placental circulation. At birth, the lungs are inflated and resistance in the pulmonary vascular bed decreases as the lungs become the source of oxygenation, rendering the DA no longer necessary. Over the first few days of life, the DA undergoes active constriction and occlusion; failure of this process results in a patent ductus arteriosus (PDA). In the majority (> 95%) of term infants and those born > 28 weeks of gestation, closure of the ductus occurs within hours. In contrast, PDAs occur in 70% of infants born prior to 28 weeks of gestation, with an incidence that is inversely proportional to gestational age (GA) at birth.

BIOLOGICAL FACTORS

Closure of the DA involves a number of cellular and molecular processes that occur over 2 phases: (1) “functional” closure within hours after birth, characterized by ductal smooth muscle constriction; (2) “anatomic” occlusion over the next several days, characterized by ductal remodeling in response to hypoxia/ischemia of smooth muscle cells, resulting in permanent closure. Premature birth interrupts the normal maturation of ductal contractile mechanisms, increasing the likelihood of a PDA.

Pathophysiology and Functional Closure of the Da

In the fetus, most of the cardiac output from the right ventricle bypasses the lungs and flows from the main pulmonary artery (MPA) into the descending aorta via the DA. A number of factors regulate the magnitude of ductal shunting, including its size (diameter, length), pressure differences between the aorta and pulmonary artery, and systemic and pulmonary vascular resistances (SVR and PVR, respectively). If the diameter of the PDA is small, the diminutive cross-sectional opening offers a high resistance to flow; thus, shunting is minimal despite potentially large pressure differences. On the other hand, with a large ductal diameter, pressures tend to be similar, and the magnitude of the ductal shunting is determined primarily by a balance between SVR and PVR. The right-to-left ductal shunting in utero is due to the high PVR, which in combination with a low SVR, results in a low pulmonary blood flow during fetal life.

At birth, ventilation of the lungs triggers a large decrease in PVR, which coincides with an increase in SVR following removal of the placental circulation by clamping of the umbilical cord. These events result in a marked increase in pulmonary blood flow, and as PVR falls below SVR, there may be left-to-right ductal shunting through the DA from the aorta and into the pulmonary circulation (see also Chapter 45).

At birth, the dramatic, rapid rise in arterial oxygen content upon ventilation of the lungs triggers a cascade of events in which constriction of the DA would normally occur. With increasing gestational maturation, the amount of ductal constriction in response to increasing partial pressure of oxygen (pO₂) is greater, and the level of pO₂ required for initiation and maintenance of this process decreases.

The potential cellular mechanisms responsible for ductal constriction are summarized in Figure 58-1, and include the inhibition of voltage-dependent potassium channels by production of reactive oxygen species, the endothelin-1 constrictor pathway, and a fall in circulating prostaglandin E₂ (PGE₂) levels following clamping of the umbilical cord. PGE₂ is the most important prostanoid to regulate ductal patency in the fetus and neonate, and is produced by the placenta and ductus. In addition, the marked increase in pulmonary blood flow at birth promotes lung catabolism of circulating PGE₂, promoting ductal closure.

The mechanisms responsible for ductal constriction may be developmentally regulated. The formation of alveoli is believed to begin at approximately 28 weeks’ gestation. Studies in preterm animal models have demonstrated an immaturity of both potassium and calcium channels. These impairments and other physiological consequences of preterm birth lead to ineffective oxygen-mediated constriction of the DA. The reduced pulmonary blood flow and metabolism among preterm infants contributes to increased circulating concentrations of PGE₂ and persistent ductal patency. Additionally, the sensitivity to PGE₂ is greater in preterm neonates than in more mature infants. Excessive fluid intake (> 170 mL/kg/day) has also been associated with increased risk of continued patency of the DA.

Anatomic Closure of the Da

Successful “functional” closure of the ductus results locally in ischemic hypoxia of the ductal wall. This results in cell death and production of hypoxia-mediated growth factors, which play a role in vascular remodeling and permanent ductal closure. Intimal thickening of the DA is also required for permanent ductal closure. Insufficient ductal smooth muscle constriction prevents the histological changes (intimal thickening, fragmentation of the internal elastic lamina) necessary for permanent closure of the DA.

Even minor alterations in the hypoxia-dependent mechanism may prevent anatomic DA closure. This potentially explains the proclivity of the ductus to “reopen” after echocardiographic evidence of its closure. The localized region of hypoxia within the ductal wall is promoted by the initial constriction of the ductus, which leads to cell death and ductal remodeling. It has been reported that, in DAs with no clinical signs of ductal patency after pharmacological therapy, those with residual luminal blood flow had a higher rate of reopening than did those with no luminal flow. The residual blood flow may provide sufficient oxygen to the luminal cells and tissues, preventing hypoxia-promoted closure. Mechanisms promoting ductal relaxation may then eventually lead to “reopening” of the ductus.

Diagnosis and Evaluation

Clinical Assessment

The clinical features associated with shunting through the PDA depend on the magnitude of the ductal shunt and the compensatory ability of the immature myocardium to handle the additional volume load. The cardiac function of preterm infants (< 37 weeks) is more vulnerable to preload and afterload changes than those in their more mature counterparts.

The murmur associated with a PDA is commonly heard best at the left sternal border in the second and third intercostal spaces. While older children may display the classic continuous, machinery murmur, very premature infants more commonly have a high frequency systolic murmur with a “rocky” quality. With increasing left-sided volume loading, a prominent second heart sound, mid-diastolic flow rumble at the apex, and a third heart sound may be present. As the left-to-right ductal shunt increases, the precordium becomes hyperactive, pulse pressure widens, and the peripheral pulses become more prominent. If the shunt becomes sufficiently large, evidence of left ventricular failure, including tachycardia, tachypnea, and rales on auscultation of the lung, may develop.

Usefulness of Clinical Assessment

Recent investigators have challenged the isolated use of clinical signs, with evidence of poor consistency and limited accuracy. During the first few days of life, preterm infants with large PDAs may have no clinical signs on physical examination due to a higher PVR, lower ductal shunt velocity, and compensatory increases in pulmonary lymph flow clearing the respiratory interstitum of excess fluid and protein. Thus, in most cases, even in the presence of a large PDA, pulmonary mechanics may not be adversely affected during the first 72 hours after birth. Following the expected drop in PVR over the first week of life, the murmur becomes apparent. After this time, the sensitivity and specificity of a murmur to diagnose a PDA are 79% and 94%, respectively. The left-to-right ductal shunt leads to tachycardia and increased stroke volume, which manifest clinically as hyperactive precordium, increased pulse volume, and wide pulse pressure. While these physical exam findings, alongside evidence of pulmonary overcirculation and left heart dilation on a chest radiograph, increase clinical suspicion for a PDA, the accurate diagnosis of ductal patency requires Doppler echocardiography.

Echocardiography

Echocardiography is the gold standard for PDA diagnosis. Two-dimensional echocardiography with Doppler flow studies provide accurate assessment of (1) ductal size (diameter), (2) pattern of ductal shunting, and (3) volume of ductal shunting. These echocardiographic indices can be used to define a hemodynamically significant PDA (HSPDA), discussed below (Figs. 58-2, 58-3, 58-4, 58-5, 58-6).

PDA Size

Ductal size is measured at the site of maximal constriction in end systole (Fig. 58-3). Traditionally, a ductal diameter > 1.5 mm is used define a HSPDA. However, this dimension cannot alone determine the need for PDA treatment or predict the likelihood of spontaneous closure. Alternatively, others have proposed indexing ductal diameter-to-body weight (in mm per kg) or left pulmonary artery (LPA) diameter (PDA:LPA ratio). The likelihood of spontaneous ductal closure during the first 3 days of life when there is a low PDA:LPA ratio (< 0.5) is 3-fold higher than if the ratio is large (≥ 1) or moderate (≥ 0.5–1). The relative benefits and limitations of different indices of PDA size remain incompletely characterized.

Flow Pattern

The primary direction and velocity of ductal shunting can be approximated using color Doppler; however, accurate determination of shunt velocity and directionality requires pulsed Doppler interrogation (Figs. 58-4, 58-5, 58-6). The pattern of Doppler shunting reflects the relative pressures (aortic and main pulmonary artery) at each end of the ductus in systole and diastole. For instance, when aortic pressure exceeds the pressure in the MPA, the shunt is left to right (positive tracing), but when MPA pressure exceeds aortic pressure, the shunt is right to left (negative tracing).

Shunt Volume

The accurate measurement of the volume of ductal shunting requires cardiac catheterization, but the invasive nature of the procedure limits the feasibility of this in fragile neonates. A number of echocardiographic indices have been proposed to assess the volume of ductal shunting indirectly, including left atrial–to–aortic root ratios, left ventricular size, left ventricular output, and diastolic flow in the descending aorta. Retrograde diastolic blood flow in the descending aorta may represent the echocardiographic marker that best reflects the circulatory consequences of an HSPDA. The normal diastolic flow in the postductal descending aorta is low-velocity forward flow but as ductal shunting increases, the diastolic flow becomes progressively less and may become retrograde (Fig. 58-6).

DEFINING A HEMODYNAMICALLY SIGNIFICANT PDA

Rather than an all-or-none approach in the determination and treatment of PDAs, efforts to define HSPDA according to the severity of echocardiographic criteria may provide a more selective approach to care and help to identify the subset of infants that are most likely to benefit from treatment (Table 58-1). Targeted use of PDA treatment in those with an HSPDA may enable healthcare providers to minimize risk and yield the greatest benefits.

An HSPDA has been defined based on ductal size or left heart dimensions (ratio of left atrium to aorta), and those echocardiographic indices are correlated to the likelihood of subsequent ductal closure or to a range of short or longer-term clinical outcomes. For instance, neonates with ductal diameters > 1.5 mm had the highest odds of an adverse outcome, whereas infants with PDA diameters ≤ 1.5 mm had outcomes similar to those with no PDA. Moreover, HSPDAs are less likely to close spontaneously than are small PDAs. However, marked heterogeneity in the criteria to define an HSPDA and the threshold for diagnosis or treatment of an HSPDA, limit equitable comparisons among studies. Despite widespread use of the term, consensus on the definition of HSPDA is lacking. More importantly, no studies have shown that treating, or not treating, an HSPDA improves longer-term outcomes. Recently, PDA scoring systems that take into account both clinical and echocardiographic data have been proposed and may provide more robust models for outcome prediction and allow further refinements in PDA management.

RISK FACTORS AND COMORBIDITIES ASSOCIATED WITH PDA

A ductal left-to-right shunt increases the pulmonary blood flow, which may eventually lead to pulmonary edema and decreased lung compliance. These changes in lung mechanics often require positive pressure and higher oxygen concentrations to maintain respiratory stability, but over time these contribute to the development of bronchopulmonary dysplasia (BPD).

Among preterm infants, persistent left-to-right shunting not only decreases lung function but also may lead to cardiac dysfunction, due to left-sided volume overload. These factors, coupled with a ductal steal phenomenon, may worsen systemic perfusion and contribute to a number of PDA-related morbidities. Although no causal link has been established, a PDA is considered by some to be a precursor to a number of morbidities, including BPD, neurological abnormalities, and necrotizing enterocolitis (NEC). Infants with a PDA have an 8-fold higher mortality than age-matched controls with a closed ductus.

Neurological Morbidities

Immediately after birth, preterm infants may lack cerebral autoregulation, placing them at risk for decreased cerebral perfusion in the setting of hypotension. This renders premature neonates vulnerable to neurological morbidities, including intraventricular hemorrhage (IVH). Because IVH typically occurs shortly after birth, prophylactic PDA treatment as a strategy to prevent IVH has gained considerable interest. Prophylactic indomethacin reduces IVH, but longer-term neurodevelopmental benefits at a corrected age of 18 months have not been observed.

Supportive Therapy

Irrespective of the treatment approach, all neonates with PDA should receive supportive care to minimize symptoms related to ductal shunting.

Positive Pressure Ventilation

Positive pressure ventilation (positive end-expiratory pressure, or PEEP) is intended to improve gas exchange and minimize the adverse effects of ductal shunting with pulmonary edema. However, the impact of prolonged positive pressure respiratory ventilation on feeding behavior and neurodevelopment remains poorly understood.

Fluid Restriction

Fluid restriction is widely recommended as a measure to limit the physiologic consequences of excessive volume load on the pulmonary circulation. Randomized studies and meta-analyses have demonstrated an association between higher fluid intake and PDA, and in turn between PDA and both congestive heart failure and NEC. Fluid restriction has been shown to reduce the risks of PDA, without increasing the risk of adverse consequences.

Diuretics

Diuretic therapy is intended to optimize lung compliance and oxygenation. However, furosemide (a loop diuretic) stimulates renal synthesis of PGE₂, and has been shown to increase the rate of ductal patency twofold when compared to chlorothiazide. In view of the evidence that furosemide may increase the prevalence of PDA, its routine use is not recommended.

Treatment of the Patent Ductus Arteriosus

There exists consensus that a PDA should be closed by 2 years of age, to avoid the risks of pulmonary hypertension or bacterial endocarditis, but disagreement remains about the thresholds and strategies for PDA closure among extremely premature infants earlier in life. Some advocate closing PDAs aggressively, based on observational data showing improved short-term outcomes following ductal closure. However, evidence for improved outcomes following treatment is not strong. Although associated with higher mortality and morbidity, PDA may not be causative. The treatment of PDA in preterm infants can be conservative, pharmacological (indomethacin, ibuprofen, acetaminophen), or nonpharmacological (catheter-based closure or surgical ligation), but the optimal approach has not been determined through prospective studies with long-term follow-up (Fig. 58-7).

Conservative Management

The conservative approach to the management of PDA through supportive therapy is aimed at reducing symptoms from the ductal shunt and providing the opportunity for spontaneous closure, thereby avoiding the risks of potentially unnecessary interventions. To date, the only randomized controlled trial (RCT) comparing treatment of an HSPDA versus conservative management was conducted over 30 years ago. In that study, the early surgical ligation group had lower rates of adverse outcomes than did the nonintervention group; however, the applicability of these findings in the setting of contemporary neonatal treatment (eg, antenatal corticosteroids, surfactant therapy) is controversial.

The rate of spontaneous ductal closure within the first 4 days of life is greater than 95% in neonates with birth weights > 1500 grams. In a single-center, prospective study of 122 infants with average birth weight of less than 800 grams, spontaneous permanent ductal closure occurred in 42 (34%), whereas the vast majority (80/122) had a persistent ductal shunt. Factors that predicted spontaneous closure included greater GA and absence of respiratory distress syndrome.

The natural history of the PDA in premature infants remains poorly characterized, though persistence beyond the first few days of life can still be associated with later closure. Spontaneous closure rates of around one-third have been observed by day 8 of life in infants with birth weights < 1000 grams, while closure by 165 days has been reported in approximately three-quarters of those born at 24 to 27 weeks of gestation. In a single-center, retrospective, observational study of infants with birth weights < 1500 grams, 21 infants were discharged home with an open ductus. Of those infants, 18 (86%) PDAs spontaneously closed at a mean post menstrual age of 48 weeks (51 days of life); none of the infants had evidence of heart failure secondary to prolonged PDA exposure.

Despite insufficient data on the risks associated with prolonged exposure to PDAs, recent evidence suggests a marked change among US providers away from traditional drug therapy and surgical ligation toward a conservative treatment approach. While conservative treatment may be a useful adjunct to ductal closure and may reduce unnecessary medical therapy or surgical ligation in an appreciable number of infants, the question remains as to what to do when PDAs fail to close following a period of conservative treatment, particularly if there are clinical and echocardiographic signs of an HSPDA or a need for continued respiratory support at a corrected age of 36 weeks.

Pharmacological Therapy

Among infants with a PDA warranting closure, 3 primary pharmacological options are available: indomethacin, ibuprofen, and acetaminophen (paracetamol) (Table 58-2). Studies have consistently shown that cyclooxygenase (COX) inhibitors, including indomethacin and ibuprofen, are more effective than placebo in closing the PDA in preterm infants. While indomethacin and ibuprofen are the only 2 US Federal Drug Administration–approved drugs for closure of a ductus in preterm babies, recent published experiences suggest that acetaminophen also may have a role. Data are not clear on which infants will respond to drug therapy. Some studies report no differences between responders and nonresponders, while other studies have suggested that failure is related to maturational factors (eg, birth weight, GA). The choice of 1 drug over the other also depends on local availability of the intravenous or enteral preparation.

Indomethacin

Indomethacin mediates its effects, at least in part, by inhibiting local prostaglandin synthesis, and evidence exists to support its use. The potential timing of indomethacin varies according to whether a prophylactic, early, or late treatment is being considered.

The relative benefits of early versus late therapy have not yet been clearly established. Indomethacin therapy in the first 2 days of life is associated with higher rates of PDA closure than later treatment. Among 8 trials including 264 infants, early symptomatic treatment was associated with lower risks for BPD (odds ratio [OR], 0.39; 95% confidence interval [CI], 0.21–0.76) or NEC (OR, 0.24; 95% CI, 0.06–0.96), and need for surgical ligation (OR, 0.37; 95% CI, 0.20–0.68). On the other hand, in a large population-based cohort of infants born prior to 27 weeks of gestation, no differences in the rates of PDA surgery or death were observed following early (0–2 days), intermediate (3–6 days), or late (≥ 7 days) pharmacological treatment.

The optimal dosing strategy of indomethacin remains unknown. Although plasma drug concentrations correlate poorly with drug efficacy and side effects, some advocate for the potential value of obtaining plasma drug levels to titrate treatment regimens; however, data on optimal levels are not widely available and this approach is poorly characterized. A multicenter RCT that investigated the correlation between indomethacin levels and continued ductal patency showed that higher serum indomethacin concentrations were not associated with lower rates of patency, but were linked with higher rates of nephrotoxicity and retinopathy of prematurity.

Because indomethacin provides only transient suppression of prostaglandin synthesis, multiple courses of indomethacin are typically provided to achieve ductal constriction; however, data to support the provision of multiple courses of indomethacin are lacking, and prolonged treatment courses have been associated with an increased risk of NEC and renal impairment. These observations are consistent with a recent RCT that reported no difference in ductal closure rates but did report trends toward lower rates of NEC (1.4% vs 7%), following therapy with 3 versus 6 doses of indomethacin. These studies suggest that prolonged courses of indomethacin treatment should not be part of the routine management of the PDA during neonatal period.

Ibuprofen

Among the COX inhibitors, it has been advocated that ibuprofen should be the preferred agent for PDA closure. Some studies have shown that while the rate of successful ductal closure with ibuprofen and indomethacin are similar, the rates of NEC and renal insufficiency, and the duration of mechanical ventilation are lower with ibuprofen. However, data on the safety of ibuprofen versus indomethacin are mixed. A recent study of over 6300 infants born at < 28 weeks of gestation reported similar short-term safety outcomes following exposure to indomethacin versus ibuprofen. Moreover, data shows that ibuprofen, but not indomethacin, interferes with binding of serum albumin at standard doses of the drugs, which would increase the risk of kernicterus among infants treated with ibuprofen.

Regarding the optimal route for administration, it has been shown that orogastric administration is as effective as intravenous administration. However, questions regarding the optimal dosing strategy, early versus expectant administration, and continuous versus intermittent dosing, remain unanswered. Data showing that prophylactic ibuprofen does not appear to be effective in preventing IVH, and may be associated with an increased risk of pulmonary hypertension, have led experts to recommend against the use of prophylactic ibuprofen.

Complications and Contraindications of COX Inhibitors

A number of studies have examined the short-term adverse effects of COX inhibitors in preterm infants. Indomethacin decreases gastrointestinal, renal, and cerebral blood flow. Consequently, treatment with indomethacin has been associated with spontaneous intestinal perforation, necrotizing enterocolitis, renal insufficiency, and altered cerebral perfusion. Importantly, healthcare providers should be aware of the high risk for isolated gastrointestinal perforation among preterm infants treated concurrently with indomethacin and dexamethasone. Despite the lack of consensus, a number of contraindications to COX inhibitors are shown in Table 58-3.

Enteral Feeding during Therapy with COX Inhibitors

Indomethacin decreases intestinal blood flow and disrupts the gastrointestinal mucosal barrier function. While ibuprofen has less effect on gastrointestinal blood flow than does indomethacin, gastrointestinal permeability is altered following ibuprofen therapy. These findings have led to concerns that initiation or advancement of enteral feedings during PDA treatment with indomethacin or ibuprofen may increase the risk of adverse health outcomes in infants. However, a study of 177 infants born at < 31 weeks of gestation reported that infants randomized to trophic feeds (15 mL/kg/day) required less time to achieve the feeding goal (120 mL/kg/day) than did infants who remained nil per os (NPO) while receiving indomethacin or ibuprofen therapy. These observations are consistent with the growing body of literature on the safety and efficacy of enteral feeding during PDA treatment with COX inhibitors. Questions remain on optimal feeding regimens for high-risk subgroups of infants, including those with echocardiographic evidence of reversal or absence of perfusion in mesenteric arteries and those with intrauterine growth restriction. In these high-risk subgroups, a more judicious feeding regimen may be prudent.

Failure to Respond to Initial Medical Management with COXI Inhibitors

Initial medical therapy with COX inhibitors will close the PDA in 50% to 75% of infants. Among infants failing to respond to initial therapy and with persistent HSPDA, a second course of treatment can be provided. Despite limited data, available evidence suggests a second course of pharmacological therapy results in closure rates of 40% to 56%. If a patient fails to responds to 2 courses of therapy, the success of additional drug treatment in closing the PDA is unlikely. In this clinical scenario, healthcare providers must reassess the indications for PDA closure and the potential hemodynamic consequences of the persistent ductus. In settings of a persistently small PDA, a recent study suggests that approximately 75% will spontaneously close on their own without pharmacological treatment.

Acetaminophen

Data on the risks of indomethacin and ibuprofen have led to interest in alternative medical therapies for PDA closure, including acetaminophen (paracetamol). While small, observational reports suggest that acetaminophen may be an effective treatment for PDA, data on the risk/benefit profile of acetaminophen among preterm infants remain unknown. A recent Cochrane review identified only 2 RCTs involving acetaminophen for PDA closure in preterm infants. No differences in the rates of ductal closure between infants treated with acetaminophen versus oral ibuprofen were observed, but a lack of high-quality studies to guide evidence-based decision-making was recognized. A recent systematic review and meta-analysis included 16 studies, 14 of which were uncontrolled studies, comparing acetaminophen versus indomethacin, ibuprofen, or placebo. The authors reported ductal closure rates with acetaminophen of 49% (95% CI, 29–69%) and 76% (95% CI, 61–88%), respectively, following 3 and 6 days of treatment. Moreover, a recent study showed that, among infants born at < 32 weeks of gestation with a confirmed PDA, the ductus closed faster in the acetaminophen group than the placebo group (177 hours versus 338 hours for acetaminophen versus placebo, respectively) without detectable side effects. However, studies in animal models raise concerns that acetaminophen may have adverse effects on brain maturation, thus its routine use cannot be recommended prior to a more robust understanding of potential longer-term risks, including neurocognitive status.

Nonpharmacologic Intervention

Surgical Ligation

Surgical ligation provides definitive PDA closure. While early surgical mortality is low, known risks of PDA ligation include pneumothorax, bleeding, and left vocal cord paralysis. Preterm infants are at risk for post-ligation cardiac syndrome (PLCS), characterized by hypotension requiring inotropic support and respiratory failure within 12 hours after surgery. Potential physiological mechanisms of PLCS include a sudden increase in SVR and reduction in cardiac preload, leading to decreased myocardial performance and compromised systemic and cerebral blood flow. Moreover, growing evidence of anesthesia-related long-term morbidities is concerning.

While no causal link has been established, infants undergoing surgical ligation have greater risk of short-term morbidities than do infants not treated surgically. While data on longer-term outcomes remain mixed, 2 large studies have shown that infants treated with surgical ligation had higher odds of neurodevelopmental impairment (NDI) at long-term follow-up. These observations have led to growing uncertainty on the safety and efficacy of surgical PDA ligation during infancy. In most contemporary settings, surgical ligation is reserved for infants with evidence of PDA-related morbidity (HSPDA) following failed drug therapy, or in clinical settings in which drug therapy is contraindicated.

A recent meta-analysis of 40 studies including 32,345 preterm infants < 32 weeks of gestation investigated the association of surgical ligation with the risks of short- and long-term morbidity and mortality, including NDI at 2 years of age, and compared these to alternative treatment strategies, including medical management. Surgical ligation was associated with greater risks of chronic lung disease, severe retinopathy of prematurity, and NDI than were observed with medical management, but also was associated with a lower mortality. The authors acknowledge that “sicker” infants, who may be at higher risk of NDI, may have been more likely to be assigned to surgical ligation. Of note, recent studies have reported no increase in mortality after moving away from an aggressive, early surgical approach to more conservative management.

Catheter-Based PDA Closure

Catheter-based closure of the PDA is among the safest of interventional cardiac procedures and is considered by some to be the procedure of choice for PDA closure beyond the neonatal period. Although limited to observational and retrospective studies, more centers are reporting the safety and feasibility of catheter-based PDA closure in preterm infants. While data on short and longer-term outcomes of catheter-based PDA closure are limited, evidence of increased risk of vascular compromise (thrombus necessitating systemic anticoagulation) in lower weight infants is concerning. Although there are risks in any cardiac catheterization, certain technical challenges may be exaggerated in very low–weight infants; thus, the risks of catheter-based PDA closure in these patients have to be weighed against those associated with alternative surgical and medical management options. Presently, among infants who warrant PDA closure, the absence of studies comparing catheter-based versus surgical closure precludes determination of the optimal strategy for ductal closure.

Screening and Prophylaxis

While some investigators have suggested that selective echocardiography among infants with PDA-related symptoms is more cost-effective without evidence of increased harm, others advocate for universal screening to provide early detection of a PDA prior to potential ductal sequelae. A recent population-based study among over 1500 preterm infants born at < 29 weeks of gestation reported that, compared to no PDA screening, early screening (first 3 days of life) was associated with lower in-hospital mortality (14.2% vs 18.5%; OR, 0.73) and lower incidence of pulmonary hemorrhage (5.7% vs 8.4%; OR, 0.60). However, the study was observational, and caution is warranted in the interpretation of findings. The optimal screening approach for the detection of a PDA in preterm infants has yet to be defined.

PDA Prophylaxis with COX Inhibitors

In the Trial of Indomethacin Prophylaxis in Preterm Infants (TIPP), investigators randomized 1202 infants with birth weights of 500 to 1000 grams to receive prophylactic treatment (0.1 mg/kg of birth weight) or placebo intravenously, commencing within hours after birth, for 3 days. Infants receiving prophylactic indomethacin had a lower incidence of PDA (24% vs 50% in the placebo group; OR, 0.3) and of severe periventricular and IVH (9% vs 13% in the placebo group; OR, 0.6). There were no differences in the primary composite outcome of death or neurocognitive status at 18 months corrected age. These observations are consistent with a meta-analysis of 19 clinical trials including over 2800 infants that indicated that prophylactic indomethacin may have short-term benefits, including a reduction in the incidence of symptomatic PDA, the need for surgical ligation, and the incidence of severe IVH; however, these benefits did not translate into reductions in mortality or improvements in longer-term neurodevelopment. Thus, it could be argued that healthcare providers should not provide indomethacin prophylaxis with the expectation of increased survival or improved longer-term outcomes.

In a follow-up analysis of the TIPP trial, infants without a PDA who were treated with prophylactic indomethacin had a higher incidence of chronic lung disease (43%, 170/391) than did those given placebo (30%, 78/257; P = 0.015). After adjusting for baseline prognostic factors, indomethacin prophylaxis among infants without a PDA was associated with a greater weight loss and need for supplemental oxygen. While prophylactic indomethacin decreases the incidence of a PDA, additional comorbidities of preterm birth (eg, BPD, NEC) remain unchanged or may, in fact, be increased following exposure.

CONCLUSION

Medical and surgical interventions are effective in closing the PDA in most infants, but neither individual clinical trials nor meta-analyses have shown that closing the PDA results in improved long-term outcomes. Although many strategies for management of the persistent PDA have been proposed, none have been subject to systematic evaluation in RCTs that have been powered to detect differences in long-term outcomes. In the absence of clear data that adverse outcomes can be averted by closing the PDA, trials comparing closure to non-intervention are greatly needed. For a treatment to be the standard of care, the risks associated with a persistent PDA must outweigh the risks undertaken by the treatment; therefore, the primary endpoint for studies comparing PDA management strategies should not be successful closure, but rather mortality and longer-term outcomes.

SUGGESTED READINGS

INTRODUCTION

Hemodynamics, or the movement of blood, is necessary to delivery oxygen to the tissues of the body. Oxygen delivery is determined by several factors, and failure of any these can lead to organ injury and/or death. Oxygen delivery is determined by (1) the oxygen level and carrying capacity of the blood, both of which are easy to measure and monitor and (2) the flow rate of blood around the systemic and pulmonary circulations. Both the circulations are driven by the heart, and the flow is measured by the output of each of the ventricles. In a mature circulation, the right ventricular output (RVO) drives the pulmonary blood flow and the left ventricular output (LVO) drives the systemic blood flow. The efficiency with which each ventricle does this is determined by the volume of blood entering the ventricle (the preload), the health (or maturity) of the myocardium (contractility), and the resistance against which the ventricle is pumping (afterload).

In preterm infants, many of the morbidities, particularly the cerebral morbidities, are thought to have an ischemic origin, and the goal of circulatory support has always been prevention of neurological morbidity. It can be seen that there are many levels at which the delivery of oxygen to tissues can fail, and sorting out this complexity in a clinical scenario is an unresolved challenge. This chapter will focus specifically on the “movement of blood” component of this physiological cascade.

THE TRANSITIONAL CIRCULATION

An understanding of neonatal circulatory pathology is intrinsically linked to an understanding of circulatory adaptation to extrauterine life. The right-sided dominance of the fetal circulation, with blood shunting right to left through the ductus arteriosus and foramen ovale to bypass the lungs, has to change rapidly to the left-sided dominance of the extrauterine circulation, with the whole cardiac output flowing through the lungs (see also Chapter 45).

At birth, this sequence must change in a very short timeframe. The lungs expand with the first breaths, the pulmonary arterioles dilate, right heart pressures fall, and blood pours into the pulmonary circulation to collect oxygen from the inhaled air. The removal of the low-resistance placenta from the systemic circulation increases resistance and pressure on the left side of the circulation, while the pulmonary blood flow increases the left heart preload. The result is a dramatic increase in the workload of the left heart. The muscle in the wall of the ductus arteriosus constricts powerfully in response to rising oxygen levels, closing functionally within the first 24 hours after birth and structurally after several days, to leave the ligamentum arteriosum, the fibrous ductal remnant that we all have.

During the last trimester of pregnancy, much of the fetal cardiopulmonary development is preparing for the major changes that accompany birth, so babies born prematurely have exquisite circulatory vulnerability during this period of the transitional circulation. More mature babies are also vulnerable if born in a compromised condition or if they become unwell shortly after birth.

CLINICAL ASSESSMENT OF THE NEONATAL CIRCULATION

Clinical Signs

Physical signs such as poor color, increased heart rate, prolonged capillary refill, low urinary output, and altered blood parameters such as low pH and rising lactate are some of the clinical signs that are typically used to assess the circulatory performance. All of these signs are useful in identifying the neonate with severe circulatory compromise, but have limited value for the early detection of circulatory compromise.

Blood Pressure

If there is intra-arterial access, blood pressure (BP) can be accurately measured and continuously monitored. Because this is relatively easy, BP has traditionally featured very highly in neonatal circulatory assessment and many circulatory interventions have traditionally focused on increasing the BP. There exists good data on what is a normal range of BP but there is controversy within the field of neonatology as to what constitutes an “adequate” blood pressure, in other words, the BP below which organ injury can result. Further, there is an emerging body of evidence that challenges the accuracy of BP as a gold standard for circulatory well-being. There has been a common assumption in neonatology that BP equals blood flow, particularly cerebral blood flow (CBF), and this caused the early focus in diagnosis and treatment of circulatory problems in neonates to be directed solely at BP. Basic physiology tells us that this thinking must be flawed some of the time because pressure is the product of flow and resistance. Changes in either of these parameters can affect BP. This pressure-based thinking was always based on limited evidence. In very preterm babies, there are data showing an association between low early BP and ultrasound evidence of brain injury, such as intraventricular hemorrhage and periventricular leukomalacia. However, there are less data to support any relationship between hypotension in preterm babies with neurodevelopmental outcome, with several studies showing no association between them.

Measurement of Blood Flow

Blood flow, rather than BP, provides a much more accurate marker of oxygen delivery. However, this is much more difficult to measure than BP. From a range of proposed methods, the 2 that are most widely used in both the research and clinical arena are Doppler cardiac ultrasound and near-infrared spectroscopy (NIRS).

Doppler Cardiac Ultrasound

Blood flow is the product of mean velocity and vessel cross-sectional area. Blood velocity can be measured with Doppler ultrasound (Fig. 59-1), and vessel size can be measured with 2-dimensional ultrasound (Fig. 59-2), as long as the vessel is of a reasonable size. This method in turn allows estimates of ventricular outputs and flow in other major vessels such as the superior vena cava (SVC). Because of this, Doppler ultrasound flow measurements in neonates are usually limited to the great vessels close to the heart: the main pulmonary artery for the RVO, the ascending aorta for the LVO and, for reasons that will be discussed below, measures of upper body and brain blood returning to the heart via the SVC. Because they are small, Doppler measures in the smaller organ blood vessels usually are limited to velocity and flow direction pattern, which may not accurately represent flow. The advantage of Doppler ultrasound is that it measures the hemodynamic variable that most closely represents oxygen delivery. The disadvantages are that it needs expensive equipment and a clinician with the skills to use it, and there are significant intrinsic errors, particularly in vessel size measurement. Further, because of the complexities of the shunts in the transitional circulation, ventricular outputs may not accurately represent systemic and pulmonary blood flows.

Near-Infrared Spectroscopy

NIRS uses infrared light to measure oxygenated and deoxygenated hemoglobin and, from this, blood flow in the superficial tissues of the brain (and other organs). Organ blood flow is derived using the Fick principle from change in oxyhemoglobin in response to an increase inspired oxygen and indices of tissue oxygenation can be derived from the ratios between oxyhemoglobin and deoxyhemoglobin. In preterm babies, very low readings (< 50%) of regional cerebral oxygenation (rScO₂) during the first 1 to 3 days of life have been associated with poorer neurodevelopmental outcome and, in 1 of these studies, high rScO₂ saturation levels were also associated with poor outcome. Near-infrared spectroscopy also can be used in conjunction with BP to assess for loss of cerebral autoregulation. Infants who show high correlation between changes in mean blood pressure (MBP) and rScO₂ are more likely to have cerebral ultrasound abnormalities. However, in a moderate-sized pilot clinical trial, treatment directed on the basis of unblended rScO₂ reading did not improve clinical outcomes. Further trials are being planned to examine this more robustly.

NEONATAL HEMODYNAMIC PATHOPHYSIOLOGY

The most commonly encountered hemodynamic disturbance is that seen in the transitional period in the very preterm infant. But there are also a variety of other clinical scenarios in which hemodynamic pathology is common. Given that the clinical recognition of neonatal circulatory compromise can be challenging, it is important to recognize that the following clinical situations are at high risk and to proactively exclude circulatory compromise.

The Preterm Transitional Circulation

The preterm neonate is an exteriorized fetus, whose ventricular function is adapted to the low resistance of the placental circulation. The preterm myocardium is immature with fewer contractile components than a term neonate, so it may not adapt to the higher afterloads of extrauterine life. The fetal channels of the ductus arteriosus and foramen ovale also are less likely to close in preterm infants and the resultant left-to-right shunts may divert blood away from the systemic back into the pulmonary circulation. The circulation may be further compromised by the fact that preterm infants need positive pressure ventilation, and ventilation can also compromise cardiac performance. When an infant is born prematurely, the stage is set for circulatory compromise and, indeed, the intrapartum period and the first 24 hours after birth are a period of particular circulatory vulnerability for preterm neonates.

The intrapartum period is difficult to study, but in the early hours after birth, low BP, low systemic blood flow (SBF), and low CBF are all associated with ultrasound evidence of brain injury and adverse neurodevelopmental outcome. Thinking in this area has predominantly focused on autoregulation of the cerebral circulation or the ability of the circulation to maintain CBF through redistribution of blood flow once the BP drops below a critical value. Much of the study of CBF has focused on exploring this pressure/flow relationship. Although the results of these studies are somewhat inconsistent, NIRS studies suggest there is a subgroup of preterm infants in whom autoregulation is compromised and who, in turn, are at higher risk of ultrasound evidence of brain injury. It is unclear whether absent autoregulation is the primary problem or whether it is an intermediate response to another primary insult. These CBF data have tended to have little information about the system driving that blood flow, the cardiovascular system. This pressure-based thinking has been questioned by Doppler cardiac ultrasound studies that have focused on transitional hemodynamics in the heart and central measures of SBF. These studies showed that the usual measures of SBF, the ventricular outputs, can be confounded by left-to-right shunts through the fetal channels, which may be significant, even early after birth. These studies showed that, contrary to conventional thinking, shunts through the patent ductus arteriosus (PDA) could be quite large, even in the early postnatal hours, and these shunts were predominantly left-to-right. Further the patent foramen ovale may also be a site of significant left-to-right shunting. This impacts on the usual measure of SBF, the ventricular outputs. These will be confounded by these shunts, left-to-right ductal shunts increase LVO, and those through the foramen ovale increase RVO. For this reason, the Doppler measure of SVC flow was used to study the natural history of preterm SBF (upper body and brain) during the early postnatal period. These studies showed that low SVC flow occurs in about a third of babies born before 30 weeks, developing within a predictable timeframe of the first 12 hours after birth and is followed by improvement of flow by 24 to 48 hrs. There is a dose-dependent relationship between the severity of this low flow and grade of intraventricular hemorrhage (IVH), which develops as or after flow improves. There is also a significant association between lower average SVC flow in the first 24 hours and poor 3-year developmental outcome. The causes of this low-flow state are not completely clear but immaturity (lower gestational age) appears to be the main risk factor. Of note, at the time of measurement, a larger PDA (shunting blood away from the systemic circulation) and higher ventilatory pressures were also associated with lower flow, particularly in studies that were performed within 6 hours of birth.

Of clinical importance, there is only a weak relationship between BP and SBF, whether measured by SVC flow or ventricular outputs (Fig. 59-3), which is reflected by the close inverse relationship between flow and calculated vascular resistance. Although this could be compensatory, it may also be causative. There is consistent evidence that the preterm myocardium has a limited ability to respond to an increase in afterload, suggesting that the transition from the low-resistance intrauterine state to the high-resistance extrauterine state is the part of the primary problem. When this transition is compounded by large ductal shunts out of the systemic circulation (which are not uncommon in the early postnatal hours) and positive pressure ventilation, critically low blood flow to all organs of the body (not just the brain) results. This hypothesized model of preterm circulatory compromise has not yet been proven. However, these data emphasize that assessment of the neonatal circulation is more complicated than simply a measurement of BP.

Other Causes of Preterm Hypotension

The physics of fluid dynamics dictate that BP can be low because either flow or resistance, or both, are low, so hypotension is a symptom with many causes in preterm and more mature infants. After the period of circulatory transition, it is more usual that hypotension is associated with normal or high SBF; thus, the problem is more commonly loss of resistance. This has been best studied in neonatal sepsis; see below.

Hypovolemia

Hypovolaemia is an uncommon cause of preterm circulatory compromise, and studies have demonstrated a lack of relationship between blood volume and BP. When seen as a primary pathology, hypovolemia is usually due to fetal blood loss and presents in the immediate postnatal period. These cases are usually associated with intrapartum blood loss from a vasa previa or fetoplacental transfusion, for example, after early clamping of a nuchal cord or from an acute fetomaternal hemorrhage. This is most often seen in term infants and can be clinically difficult to differentiate from the post resuscitation pallor that typically accompanies asphyxia. The clinician should be alert to the classic features of circulatory compromise with pallor, tachycardia, hypotension, low hematocrit, and a typical appearance on cardiac ultrasound of poorly filled ventricles.

Patent Ductus Arteriosus

Patent ductus arteriosus during the first week is associated with both lower systolic and diastolic BPs, and hence mean BP. This association occurs because the PDA exposes the systemic circulation to the lower resistance of the pulmonary circulation throughout the cardiac cycle. Most PDAs are clinically silent during the first 3 postnatal days, so this needs to be considered in neonates with persistent hypotension. Diagnosis with cardiac ultrasound is quick and easy (Fig. 59-4). As discussed above, in the very early transitional period, a larger ductal diameter is significantly associated with lower SBF. After this time (and contrary to popular belief) our data suggests that many infants will protect their systemic circulation well in the presence of a significant PDA (see also Chapter 58).

Sepsis

Neonatal sepsis can commonly present with signs of circulatory compromise. Little data exists about the hemodynamics of neonatal septic shock. In early onset sepsis, the problems are probably a combination of the transitional hemodynamics (described above) and pulmonary hypertension associated with pneumonia (described below). There are now studies that show that circulatory compromise associated with late onset sepsis in preterm infants is usually due to vasodilatory shock. These infants usually have normal or high SBF, indicating loss of vascular tone. In advanced septic shock, evidence of tissue ischemia can persist despite normalization of systemic hemodynamics. Although little studied in the newborn, this may be due to the microvascular dysfunction described in older subjects.

Inotrope-Resistant Hypotension

Hypotension that is resistant to vasopressor support seems to be associated with poor adrenocortical function. However, blunted cortisol responses are not universally found in these infants, so the etiology is likely to be more complex than this. Our clinical observations and 1 more systematic study suggest that hypotension is associated with vasodilation, with the usual hemodynamic finding of normal or high SBF. The mechanistic complexity of vasodilatory shock has been described in older subjects. Whether these observations apply in the neonate is not known.

The Term Infant with Asphyxia, Pulmonary Hypertension, or Severe Respiratory Distress

Term and near-term infants with asphyxia, pulmonary hypertension, or severe respiratory distress are a heterogenous group who also have a high risk of low SBF in the first 24 hours and, like the preterm infants, the incidence of low SBF becomes less with postnatal age. The causes of this are complex, and hemodynamic findings vary widely among infants. There is probably a varying degree of influence from all the factors already discussed, particularly the effect of high ventilatory pressures. In infants with perinatal asphyxia, hypoxic-ischemic damage to the myocardium is important in the etiology. One factor often not considered is the potential for high resistance in the pulmonary circulation to compromise the systemic circulation. The left ventricle can only pump into the systemic circulation what it receives from the pulmonary venous return. If pulmonary vascular resistance is high enough to restrict pulmonary blood flow and the ductus arteriosus and foramen ovale are closed or restricting (as they often are in these more mature infants), then the SBF will be similarly compromised. It has been suggested that myocardial contractility is poor in these infants but low left ventricular (LV) preload will also make the LV contractility appear poor. Therefore, this may be a secondary rather than a primary phenomenon. There can be quite dramatic increases in cardiac output in infants treated with nitric oxide (Fig. 59-5).

Circulatory Collapse

Sudden circulatory collapse in infants is rare but catastrophic when it happens. It is characterized by rapid onset of all the typical clinical signs described above. Recognizing the problem is not typically difficult, but diagnosing the cause can be more problematic. Such infants are often assumed to be septic, and while sepsis should always be considered, it must never be assumed until other causes are excluded. Congenital heart lesions with ductal-dependent systemic circulations, such as critical coarctation or hypoplastic left heart syndrome, are the most important diagnoses to exclude. However, other rarities, such as cardiac tamponade resulting from a silastic long line catheter with the tip being left in the heart or acute viral myocarditis, can present unexpectedly in this way. Many of these diagnoses will be quickly apparent on cardiac ultrasound, thus emphasising the importance of ready access to point-of-care ultrasound in good neonatal care.

Summary

Circulatory pathology is common during the first 24 hours in very preterm and sick term infants. The etiology is multifactorial but cardiac maladaptation to higher ex utero vascular resistance may be 1 factor. This can manifest as low BP and/or low Doppler-measured SBF, and the measurement of SBF together with BP will improve understanding of an infant’s circulatory pathology. Low SBF is uncommon after 24 hours. Most infants with hypotension beyond the transitional period (and a few within the transitional period) have normal or high SBF, pointing to vasodilation as the dominant pathology. In some infants with persistent pulmonary hypertension of the newborn (PPHN), restricted pulmonary blood flow will in turn restrict systemic blood flow. Circulatory collapse may be due to sepsis, but this should not be assumed, and other primary cardiac causes should be looked for with cardiac ultrasound.

CIRCULATORY SUPPORT IN THE PRETERM INFANT

Volume expansion, catecholamines, and increasingly, hydrocortisone are the main agents used in neonatal circulatory support. Empirically, each agent differs in its effect on the circulation: Volume expansion restores normovolemia in a hypovolemic infant and increases preload and hence cardiac output in a normovolemic infant. Inotropes increase SBF by increasing cardiac rate and myocardial contractility or by reducing afterload. Some inotropes have predominantly vasoconstrictive effects and modify preload and SBF by constricting the venous bed. Arterial constriction improves BP, but excessive constriction can reduce SBF by causing afterload compromise.

Dopamine and epinephrine are catecholamines, and have cardiac stimulant (inotropic) effects and predominantly pressor (vasoconstrictive) effects peripherally, particularly at higher doses. Dobutamine is primarily a β-adrenergic catecholamine with inotropic, chronotropic, and peripheral vasodilator effects Dopamine is a naturally occurring precursor to epinephrine and norepinephrine that has β, α, and dopaminergic effects. Each of these effects is more likely (in the order shown) to be stimulated as the dose increases. Over 10 mcg/kg/min, the α vasconstrictive effects on BP predominate, but in the very immature infant, these α effects may be apparent at lower doses. Epinephrine is an endogenous catecholamine and has broad α- and β-adrenergic effects and, like dopamine, vasoconstricts at higher doses. Hydrocortisone increases BP, but little data exists about how it does this in the neonate. The limited information there is suggests mixed central and peripheral effects with the latter probably dominating.

The Evidence Base for Pharmacological Interventions

Clinical trials in this area have been mainly designed to look at changes in a physiological parameter (mainly BP) rather than effects on important clinical outcomes. These studies have demonstrated that routine early, or perinatal volume expansion does not change long-term preterm outcomes; that dopamine is better than dobutamine at increasing BP, but dobutamine may be better than dopamine for improving cardiac output; that epinephrine has similar effects to dopamine on BP and CBF; and that hydrocortisone will improve BP. There is no evidence that supports an effect of any of these interventions on important clinical outcomes.

Evidence for Inotrope Use

Dopamine is better than dobutamine at increasing BP, although this does not translate into better short-term outcomes. Epinephrine and dopamine have similar efficacy in increasing BP. None of the studies to date has been powered or designed to look at clinical outcomes. In most of these studies, infants were enrolled on the basis of BP below an intervention threshold, and change in BP was the main outcome measure. In a preterm infant study using Doppler measurement of LV output, an increase was shown with dobutamine, while LV output decreased with dopamine.

The author’s group has, to their knowledge, performed the only randomized study of dopamine and dobutamine that enrolled preterm infants on the basis of low blood flow. Dobutamine produced significantly better increases in SBF (as measured by SVC flow) than dopamine, while dopamine produced better increases in BP. At 10 mcg/kg/min, there was no difference between the 2 drugs, but in those that needed to be increased to 20 mcg/kg/min, dobutamine produced more flow benefit, which was not seen in those on dopamine, even though BP continued to increase. However, these physiological differences did not translate into any significant differences in outcome. SVC flow will incorporate CBF, but the two are not interchangeable, so what effect do inotropes have on CBF? The effect of dobutamine on CBF in preterm infants has not been studied. Using the Xe clearance technique, it was shown that dopamine, while increasing LVO and MBP, did not increase CBF. Another study showed no effect on middle cerebral artery pulsatility index after dopamine in a group of preterm infants with normal BP but clinical evidence of poor perfusion. In an open label observational study, NIRS was used to show an increase in mean CBF after dopamine was given to a cohort of 12 preterm infants with an MBP of less than 30 mm Hg. In a double-blind randomized trial, also using NIRS, it was shown that both dopamine (at doses up to 10 mcg/kg/min) and epinephrine (at doses up to 0.5 mcg/kg/min) produced similar increases in regional cerebral intravascular oxygenation, which in turn was correlated with increases in MBP.

Evidence for Volume Expansion (Including Delayed Umbilical Cord Clamping)

Systematic reviews have demonstrated that routine early volume expansion with an intravenous fluid bolus does not improve outcomes in preterm infants. There is likely no advantage in using a colloid compared to a crystalloid, and volume is not as good as dopamine at increasing BP. There is less data about the effect of volume on organ blood flows. Studies have shown increases in LV output or SVC flow with volume that did not translate to an increase in CBF. In all these studies, measurements were taken immediately after the volume expansion, so it is not known if these increases are sustained. There is some evidence that the fluid used in volume expansion redistributes quickly out of the vascular compartment.

The “natural” method to expand an infant’s volume at birth is to delay the clamping of the cord to allow fetal blood in the placental compartment to drain back into the infant. The current evidence suggests that this is advantageous for the preterm infant, with systematic review showing delayed cord clamping associated with better circulatory stability, less IVH, and lower risk for necrotizing enterocolitis. The evidence is limited because it includes many small trials, and there are limited numbers of very preterm infants. In some parts of the world, delayed cord clamping is being introduced as a standard of care. A large prospective randomized trial to investigate the potential benefit of cord clamping in infants born before 30 weeks’ gestation is under way to address this important question.

Evidence for Hydrocortisone Administration

Much of the data on hydrocortisone is observational and relates to its use in inotrope-resistant hypotension. One randomized trial showed hydrocortisone at 2.5 mg/kg had similar efficacy to dopamine in improving BP in hypotensive preterm babies. There was no difference in other clinical outcomes. A Doppler ultrasound study has shown increases in BP, LV output, and systemic vascular resistance after 2 mg/kg of hydrocortisone in 14 infants with inotrope-resistant hypotension. However, the effects of hydrocortisone on CBF are not known.

Recommendations

There is no clinical outcome–based evidence on which to recommend therapeutic recommendations. No study has shown any improvement in any meaningful clinical outcome, short- or long-term, in response to a specific inotrope or other intervention. This lack of evidence for effect on clinical outcomes means that the approach to therapy therefore must be empirical and pragmatic. It is the author’s view that many of the complexities of the transitional preterm circulation are not reflected in BP (or other commonly used clinical signs of circulatory compromise). Treatment targeted at BP alone will be appropriate in some infants, unnecessary in others, and miss (or deliver late therapy) in some who really need it. BP must be important, but it is the author’s view that to target circulatory support requires measures of both pressure and flow. Despite this opinion, the reality is that measuring flow is difficult and many newborn intensive care units do not have access to appropriate equipment or skills and so will remain dependent on BP and other clinical parameters to guide therapy. Reflecting this reality, this chapter presents these clinical suggestions pragmatically in 3 parts: general measures and volume, a pressure-based approach, and a pressure- and flow-based approach.

General Measures

Basic preventative strategies are important but are easily overlooked. Antenatal steroids mediate their effects on both the respiratory and cardiovascular systems. Infants born after maternal steroid treatment have higher BP and less need for inotropes, and are less likely to develop low systemic blood flow. There are few valid contraindications for administration of antenatal steroids and it remains probably the most effective therapeutic intervention in preterm neonatology. As discussed above, it is reasonable to consider delayed cord clamping in very preterm infants. Avoiding overventilation is also important both because of the direct negative effects on the circulation of high intrathoracic pressure but also because of the effects of low partial pressure of carbon dioxide (pCO₂) in reducing CBF.

Hypovolaemia is rare, but it does happen. It is difficult to diagnose clinically but easy to fix if it is present. For these reasons, volume expansion should be part of the initial approach to circulatory support with any inotrope strategy. The authors use normal saline at 10 mL/kg over 20 to 30 minutes but would not repeat that without a convincing response to treatment (falling heart rate and improving BP) or strong clinical and/or cardiac ultrasound evidence of hypovolemia. Overdoing volume expansion is as potentially harmful as underdoing it.

A Blood Pressure Intervention Threshold

If treatment is being guided mainly by BP, then the first issue is determining what intervention threshold to use. In infants born before 30 weeks’ gestation, most neonatologists would either aim to maintain MBP above 30 mm Hg or above the gestation (in weeks) of the infant. There is no outcome-based evidence to say which approach is correct. Empirically, because gestation is an important independent determinant of BP, it likely makes more sense to use the latter, in that it reflects the range of values in the population rather than using a “1 size fits all” approach.

Dobutamine increases the BP of many infants and seems better than dopamine at improving SBF. Therefore, within the first 24 hours after birth, when low SBF is common, there is an empirical logic in using dobutamine as a first-line treatment, with the pressor inotropes used as a second-line treatment in infants with an inadequate BP response. For a more consistent effect on BP, dopamine or epinephrine should be the first choice. Despite similar hemodynamic effects, there is more clinical experience with dopamine, so it would in general be used before epinephrine. When using a pressor inotrope in the first 24 hours, the risks of afterload compromise of SBF from vasoconstriction should be considered. Also, the principle that more BP may not necessarily be better was emphasised by studies that showed some infants with quite dramatic rises in estimated CBF in response to an increase in BP. Hyperperfusion after hypoperfusion seems important in the pathogenesis of IVH, so is probably best avoided. These risks can probably be minimized by starting at a low dose (5 μg/kg/min for dopamine or 0.01–0.02 μg/kg/min for epinephrine) and titrating in careful steps to a minimally acceptable BP. It is recommended to aim for MBP within 5 mm Hg of the therapeutic threshold and to be prepared to wean the infusion rate quickly in infants whose BP overshoots this level.

Targeting Blood Pressure and Flow

A more complete understanding of the underlying cardiovascular hemodynamic state provides an effective basis for therapy, more than what BP alone can provide. This understanding requires point-of-care cardiac ultrasound, which should be targeted on the basis of postnatal age (3 to 9 hours of age is the highest risk time for low SBF in preterm babies) or clinical concern about the circulation (usually low BP). The 3 cardiac ultrasound measures, in order of importance, are (1) an indirect measure of systemic blood flow (SVC flow or RV output), (2) the size and direction of ductal shunt, and (3) assessment of degree of pulmonary hypertension. Intervention thresholds of 50 mL/kg/min for SVC flow and 150 mL/kg/min for RV output are recommended, and pathologically low measures of these 2 variables would be less than 40 and 120 mL/kg/min, respectively. Estimating blood flow requires measures of velocity and vessel size and can be time consuming to derive. Large atrial shunts are not common in the first 48 hours, so for clinical purposes, RV output is a reasonably accurate surrogate for SBF in the early postnatal period. Velocity in the main pulmonary artery is the dominant determinant of RV output and so measuring the maximum velocity in the pulmonary artery (PA V_max) provides a simple way to screen for low SBF (Fig. 59-6). If the PA V_max is higher than 0.45 m/sec, low SBF is unlikely. If the PA V_max is less than 0.35 m/sec, most infants have low SBF. Between 0.35 and 0.45 m/sec is a gray zone where discriminatory accuracy is not as good (Fig. 59-7). In practice, it is recommended to screen with PA V_max, followed by full RV output and/or SVC flow measures in those with a PA V_max less than 0.45 m/sec.

In infants with large PDAs (> 2 mm in diameter) and predominantly left-to-right shunts, consideration should be given to closing the PDAs medically, particularly if there is low SBF or low MBP. Otherwise, infants with low SBF, as defined above, can be treated with volume and dobutamine, starting at 10 μg/kg/min and increasing to 20 μg/kg/min, depending on response and regardless of the MBP. However, if the MBP remained persistently below the number of weeks’ gestation despite dobutamine, dopamine can be added at 5 μg/kg/min and titrated up to achieve a minimally acceptable MBP. It is rare to warrant escalating to doses of dopamine that are higher than 10 μg/kg/min, but if the SBF is known to be normal, higher doses than this could be considered.

After the first 24 hours, it is much more likely that the SBF will be normal or high. Hypotension that presents beyond this time frame typically indicates a loss of vascular tone and a pressor inotrope would be the logical therapeutic choice. Recommended treatment begins with dopamine at 5 μg/kg/min and titrating the dose up to achieve a minimally acceptable BP. There is no physiological reason why epinephrine should not be just as effective in this situation, though there is less clinical experience with it. Therefore, current recommendations are to use epinephrine as a second line to dopamine. If SBF is normal and other markers of circulatory status are satisfactory, clinicians can tolerate borderline low BPs (within 2 or 3 mm Hg of threshold). In infants where the BP drops lower than this or there are other clinical concerns, as long as the SBF was normal, dopamine should be titrated up to 20 μg/kg/min.

Inotrope Resistance

There are 2 facets to inotrope resistance. The first is the resistant hypotension discussed above, and the second is the less well-recognized resistant low SBF, which may or may not be associated with hypotension. Inotrope resistance was observed in 40% of the low SBF infants enrolled in the authors’ inotrope study and is not an uncommon finding in clinical practice. However, there is no reliable strategy that can be recommended to reverse the low SBF state in these infants. Neither epinephrine infusion nor hydrocortisone has been shown to be very effective for treating this. Systemic blood flow often increases gradually over the subsequent 12 to 36 hours.

Most infants with inotrope-resistant hypotension seem to be in a vasodilatory state, except in the first 24 hours when they overlap with a resistant low SBF group. The management of this resistant hypotension is empirical. Some would add in an epinephrine infusion, which can increase BP in infants already on maximum doses of dopamine. Increasingly, hydrocortisone (at doses of 1–2 mg/kg) is being used in this situation. While considerably more is known about the hemodynamic effects of epinephrine than of hydrocortisone, the association between low cortisol levels and resistant hypotension provides an empirical logic to using physiological doses of hydrocortisone in infants at risk of adrenal insufficiency. Our approach with resistant hypotension would depend on the clinical situation. Some would recommend that an epinephrine infusion be commenced before using hydrocortisone, but in the very premature infant it might be more effective to administer hydrocortisone. With the unresolved concerns about possible adverse neurological effects of early postnatal use of systemic corticosteroids, it seems wise to minimize the dose and duration of hydrocortisone when used in this way, starting with 1 to 2 mg/kg for 1 to 3 days, depending on the response. Septic shock is also often vasodilatory and can be very resistant to inotropes. Whether hydrocortisone should be used if there is suspected sepsis is open to question. This vasopressor-resistant loss of vascular tone (or vasoplegia) has been recognized as a feature of septic shock in older subjects and a variety of therapeutic strategies have been explored, including nitric oxide synthetase inhibitors and vasopressin. There is little evidence about these agents in resistant shock in newborns, although vasopressin has been tested in a small randomized trial as a primary inotrope against dopamine and has been found to have similar efficacy in improving BP.

CIRCULATORY SUPPORT IN THE TERM INFANT

The early recognition of circulatory compromise is key to its subsequent management. Where possible, for example with hypovolemia, PPHN or ductal-dependent congenital heart lesions, treatment should be directed at the known cause. In general, the same principles should be applied in relation to pressure and flow, as discussed above. The effects of pulmonary vasoconstriction on systemic blood flow should always be considered in term infants. In infants with cardiac ultrasound–estimated pulmonary artery pressure above or similar to systemic pressure, there should be a low threshold for using inhaled nitric oxide (where available) as a circulatory support measure in this population. When using vasopressors in infants with severe lung disease and/or PPHN, it is important to consider that these drugs can constrict both the systemic and pulmonary vasculatures. This has not been explored in term infants, but in preterm infants, dopamine has a balanced effect on both circulations with quite wide interindividual variation. Therefore, the assumption that vasopressors will automatically squeeze blood into the pulmonary circulation may not be correct. For example, there is some observational data that norepinephrine may have a preferential pressor effect on the systemic circulation with some vasodilatory effect on the pulmonary circulation.

When the SBF appears low, it is recommended to use dobutamine as a first-line treatment, with dopamine or epinephrine as a back-up if the BP remains low, using a similar protocol to that described for preterm infants. When MBP is low and SBF is normal or oxygenation is borderline, dopamine or adrenaline are often better to use as a first-line treatment. The data described above suggests that norepinephrine may have advantages in this situation.

FUTURE DIRECTIONS

This is an area of neonatology that is more defined by unknowns rather than by knowns. While it is important to show that circulatory support treatments actually produce the physiological effects according to their known mechanisms of action, unless studies can show that these effects improve clinical outcome, such research just becomes phenomenology. Circulatory support treatments have been in routine clinical practice for over 25 years, yet it is unknown whether they achieve the primary goal, which is to reduce neurological, neurodevelopmental, and neurobehavioral morbidities. Closing this knowledge gap is the challenge for the future.

SUGGESTED READINGS

INTRODUCTION

Necrotizing enterocolitis (NEC) is a serious gastrointestinal disease that occurs primarily in preterm infants. The disease is characterized by the rapid onset of intestinal inflammation and, in severe cases, can lead to intestinal necrosis and multiorgan dysfunction that results in death. Since the 1950s, when neonatal intensive care was introduced at many centers around the world with survival of very low–birth-weight (VLBW) infants, NEC has been reported as a complication of premature birth. It remains the most common gastrointestinal complication in preterm infants, and is a major cause of morbidity and mortality in neonates, accounting for 10% of deaths in the neonatal intensive care unit (NICU). Reports from the United States, the United Kingdom, Sweden, and the Netherlands indicate that NEC is increasing as a cause of death, as survival from lung disease has improved for preterm infants.

EPIDEMIOLOGY

Incidence and Recent Trends

The incidence of NEC varies among studies, depending on the population studied, case definition, and geographic region. Among a large network of NICUs in the United States encompassing 58,555 infants weighing 500 to 1500 g, the combined incidence of medical and surgical NEC was 3.9% in 2013 (2.6% medical NEC, 1.2% surgical), which had significantly declined from 6.6% in 2007. Similarly, the incidence of NEC in the Neonatal Research Network for infants with birthweights between 401 and 1500 g and born at 22 to 28 weeks’ gestation decreased from 13% in 2008 to 9% in 2012 (P < 0.01). This figure had previously increased between 1993 and 2008. The incidence of Bell stage 2 or 3 NEC for infants who survived for more than 12 hours after birth ranged from 15% at 23 weeks’ gestation to 8% at 28 weeks’ gestation. In a recent population-based cohort in the United Kingdom, the incidence of severe NEC was 3.2% (95% confidence interval [CI], 2.9–3.4; Fig. 60-1).

Age at Onset

The age at onset of NEC varies depending on the gestational age of the infant. More immature preterm infants present later in the hospital course compared to more mature preterm or term infants. NEC in term infants typically presents early, with a reported mean age of onset of 4 days. This observation has suggested a potential developmental window of highest disease risk, although there is substantial variation in both the postnatal age and postmenstrual age (PMA) at onset, even among distinct gestational age subgroups. Among infants less than 32 weeks’ gestation who required surgery for NEC, the PMA of onset was a median of 30.5 weeks (interquartile range 27.9–32.7 weeks) based on a large population-based study in the United Kingdom.

CLINICAL PRESENTATION

Signs and Symptoms

The clinical findings of NEC are variable and can involve both abdominal and systemic signs and symptoms. Infants can have rapid changes in findings within a short period of time and require close monitoring for deterioration if NEC is suspected. In 1978, Bell et al proposed a method to clinically stage the severity of NEC based on both clinical and radiographic findings (Table 60-1), and this staging system was subsequently expanded to a modified staging system. Although the Bell staging of NEC is commonly used, particularly in case definitions of NEC for clinical studies, there is no clear and universally utilized case definition of NEC. Hence, current definitions likely encompass heterogeneous clinical entities that share a similar clinical presentation. Recently, investigators from the United Kingdom proposed a gestational age–specific definition of NEC scored on clinical and radiographic findings. Non-gastrointestinal signs and symptoms include temperature instability, lethargy, apnea, and bradycardia, while more severe and progressive cases demonstrate instability of vital signs, respiratory failure, and shock. Gastrointestinal signs and symptoms include poor feeding, increases in pregavage residuals, emesis, abdominal distention, and bloody stools. As NEC worsens in severity, gastrointestinal symptoms become more pronounced with marked abdominal distention, gastrointestinal hemorrhage, and abdominal wall discoloration or erythema (Fig. 60-2). Abdominal wall discoloration is often an ominous sign, and a predictor of the severity of NEC. The differential diagnosis of NEC also requires consideration of spontaneous intestinal perforation, which presents with pneumoperitoneum. This is considered to be a distinct entity from NEC, and occurs earlier in the hospital course than NEC, often in the setting of no or minimal feeding; and is more common among extremely preterm infants (< 28 weeks’ gestation). NEC in term infants accounts for approximately 10% of cases and these infants manifest symptoms similar to those of preterm infants, although they are likely to have a different pathophysiology of the disease, as discussed later.

Radiographic Findings

The primary radiographic finding in NEC is pneumatosis intestinalis, a radiographic appearance of gas within the wall of the small or large intestine (Fig. 60-3). This is the hallmark of NEC and often establishes the diagnosis (Table 60-1). Other radiographic findings may include abdominal distention with evidence of ileus, portal venous gas, or pneumoperitoneum (Fig. 60-4). In addition, bowel loops that remain unchanged on serial radiographic films for 24 to 48 hours raise the concern for ischemic bowel. Infants with evidence of pneumoperitoneum require surgery. The presence of portal venous gas may be associated with a higher eventual need for surgery although its role as a prognostic marker for mortality is unclear.

Diagnostic Evaluation

The diagnostic evaluation of infants with suspected NEC includes evaluation for bloodstream infection, which can accompany NEC and is associated with worse outcome. Urine culture should also be obtained, as a urinary tract infection may precede the diagnosis of NEC. Patients with NEC should have a complete blood count with differential and serum chemistry monitored. Hematologic changes that occur with NEC include thrombocytopenia, coagulopathy, anemia, neutropenia or neutrophilia, eosinophilia, and lymphopenia; further, the presence of significant hematologic abnormalities is often associated with more severe disease. In addition, infants with NEC may develop hyponatremia, hyperkalemia, and metabolic acidosis. Infants with suspected NEC should have abdominal radiographs taken that include both anteroposterior views and either left-lateral decubitus or cross-table images to evaluate for free intraperitoneal air (Fig. 60-4). Abdominal radiographs should be performed serially until the patient has demonstrated improvement in radiographic findings and clinical symptoms. Blood gas measurements should be performed at routine intervals to monitor for acidosis and respiratory deterioration. A lower serum pH (ie, metabolic acidosis) is associated with advanced disease. If infants appear likely to need surgery or demonstrate worsening clinical status, coagulation studies should be obtained, as disseminated intravascular coagulation is often present in advanced cases of NEC.

TREATMENT

Medical Management

The treatment of NEC is based on the severity and involves bowel decompression, bowel rest, antimicrobial therapy, and monitoring of serial abdominal radiographs and laboratory parameters (Table 60-1). Bloodstream infection develops in up to one-third of infants with NEC. Therefore, broad-spectrum antimicrobial therapy is indicated. Most bloodstream infections are caused by gram-negative bacteria, most commonly Escherichia coli and Klebsiella.

Antibiotic Selection and Duration

There exists wide variation in the choice of antibiotic treatment for NEC, with a combination of vancomycin and gentamicin being the 2 most commonly used antibiotic agents in this population. Other commonly used antibiotics include ampicillin, cephalosporins, clindamycin, metronidazole, and carbapenems. In a small randomized controlled trial of 42 infants, the addition of anaerobic coverage with clindamycin to a standard regimen of ampicillin and gentamicin, compared to ampicillin and gentamicin alone, did not reduce the risk of intestinal perforation or death. However, surviving infants in the group receiving clindamycin, compared to ampicillin and gentamicin alone, had a higher incidence of intestinal stricture (40% vs 5.6%; P = 0.02). In a larger propensity-matched study of 2780 infants from 348 US NICUs, infants receiving initial anaerobic antimicrobial therapy at the onset of NEC, compared to those not exposed, had a higher risk of strictures (odds ratio [OR], 1.73; 95% CI, 1.11–2.72) but a lower risk of mortality (OR, 0.71; 95% CI, 0.52–0.95). In addition, this study found a decreasing use of clindamycin and increasing use of metronidazole during the study period, from 1997 to 2012. Therefore, the risks of intestinal stricture with anaerobic antimicrobial therapy should be balanced with the potential benefits in survival, particularly among infants with risk factors for high-case fatality (discussed in the next section). There is a paucity of data to guide the optimal duration of antibiotic treatment for NEC, although treatment for 7 to 10 days is common for definite NEC and often longer for surgical NEC or when specific pathogens are isolated (Table 60-1).

Surgical Management

Indications

Surgical treatment is necessary in 30% to 35% of infants with NEC, and mortality at 18 to 22 months of follow-up is as high as 50% among these infants. Pneumoperitoneum is the most frequent indication for surgery, although additional surgical criteria can include portal venous gas, abnormal paracentesis, abdominal wall erythema, palpable abdominal mass, or “fixed” intestinal loop on x-ray. A deterioration in clinical and laboratory parameters, as previously discussed, should also be considered when deciding whether surgery is indicated. Small studies have suggested that abdominal ultrasound may be useful in identification of bowel necrosis in infants with NEC, although this is currently not widely used.

Surgical Approaches

The surgical treatment of NEC involves 1 of 2 approaches. Traditionally, management has involved an exploratory laparotomy with resection of affected bowel. However, especially in infants with a birth weight of less than 1000 g, peritoneal drainage has become a more commonly used method to provide abdominal decompression until the need for a definitive laparotomy can be reassessed. In a multicenter trial of 117 infants with perforated NEC, infants were randomized to primary peritoneal drainage or laparotomy with bowel resection. There was no difference in the 90-day postoperative mortality between laparotomy and primary peritoneal drainage treatment arms (35.5% vs 34.5%; relative risk, 1.03; 95% CI, 0.63–1.69), with similar findings in the subgroup of infants weighing less than 1000 g. In a multicenter observational study of 156 extremely low–birth-weight (ELBW) infants with either spontaneous intestinal perforation or severe NEC, 76% were treated with an initial peritoneal drain and did not need a subsequent laparotomy. However, of concern was a higher associated risk of death or neurodevelopmental impairment among infants receiving peritoneal drainage compared to initial laparotomy. Currently, a multicenter randomized trial is being conducted in the United States to compare these 2 surgical approaches with the primary outcome of death or neurodevelopmental impairment at 18 to 22 months of age.

OUTCOMES

Mortality

Cause-specific mortality from NEC is high and varies depending on the gestational age and birth weight of the infant, with case-fatality rates ranging from 65% at 23 weeks’ gestation to 12% among term infants (Fig. 60-1). In a study of 7099 infants with NEC in a large multicenter cohort of NICUs, of whom 1505 died, factors associated with a higher risk of death included a lower gestational age, lower birth weight, treatment with assisted ventilation on the day of diagnosis, treatment with vasopressors at the time of diagnosis, and black race. The majority of deaths from NEC occur within 7 days of diagnosis.

Long Term Outcomes

Neurodevelopmental Impairment

Neurodevelopmental impairment is common among preterm infants with NEC, particularly those infants that required surgery or had a comorbid bloodstream infection. In a UK study that followed 119 (77%) of 157 infants with NEC until 7 years of age, and compared them to 6496 controls, those with a history of NEC had a higher incidence of cerebral palsy (6.8% vs 2.5%; unadjusted OR, 2.77; 95% CI, 1.33–5.77), attention-deficit/hyperactivity disorder (15.3% vs 7.3%; unadjusted OR, 2.24; 95% CI, 1.34–3.72), and decreased educational attainment. However, after adjusting for antenatal and neonatal factors, many of these differences were no longer significantly different between groups. If ELBW infants with NEC undergo surgery, they are at higher risk for significant long-term cognitive and motor impairment: 24% of these infants are diagnosed with cerebral palsy at 18 to 22 months corrected age and 57% have a degree of neurodevelopmental impairment, as compared to an incidence of impairment in 44% of infants with NEC not requiring surgery and 40% of infants without NEC. In addition, ELBW infants with sepsis and NEC have a similarly high incidence of neurodevelopmental impairment of 53% compared to an incidence of 29% among ELBW infants without infection. The highest risk of adverse outcomes appears to be in infants with both surgical NEC and late bacteremia, who were found to have 8-fold higher odds of cerebral palsy compared to infants without NEC or late bacteremia.

Short-Bowel Syndrome

Short-bowel syndrome is the leading cause of intestinal failure in infants, accounting for 44% of cases. It results from the surgical resection of so much affected bowel that there remains insufficient intestinal mass to maintain growth, hydration, or electrolyte balance. The risk of short-bowel syndrome depends on the amount of bowel resected and the presence of an ileocecal valve. With bowel lengthening and tapering procedures, infants with only 35 cm of residual bowel have a 50% probability of being able to wean off of parenteral nutrition and achieve enteral autonomy. In a multicenter cohort of children with intestinal failure, each additional centimeter of residual bowel length was associated with higher odds of enteral autonomy (OR, 1.04; 95% CI, 1.02–1.06), as was having an ileocecal value present (OR, 2.80; 95% CI, 1.63–4.83). Even among children who remain on parenteral nutrition beyond 1 year of age, 64% are able to wean from it, and a diagnosis of NEC is associated with a greater probability of weaning off parental nutrition compared to other causes of short-bowel syndrome. However, among those infants unable to wean off parenteral nutrition, liver disease, sepsis, and re-hospitalization are common, and intestinal and liver transplantation may be necessary for survival. In the previously mentioned UK study with 7-year follow-up of infants with NEC, the incidence of gastrointestinal problems persisted for many infants. Compared to 3.2% of infants without NEC, 15.3% of infants with NEC reported ongoing bowel problems at 7 years of age and 5.1% had a bowel stoma present.

PATHOGENESIS

The major underlying factor involved in the pathogenesis of NEC is prematurity, indicated by an increasing incidence and severity of NEC as gestational age falls (Fig. 60-1). Preterm infants have an immature epithelial barrier with a greater tendency for inflammation, a leakier gut, and decreased innate mucosal defenses. Studies of the intestinal microbiota have identified a number of differences in infants with NEC, compared to those without (discussed below). An important consideration is that NEC likely encompasses a spectrum of heterogeneous diseases that lead to intestinal injury and present with similar signs and symptoms. Therefore, the pathogenesis of “classical” NEC that predominately affects VLBW infants may differ from NEC in full-term infants, which typically presents earlier in the hospital course and predominately affects the colon. Multiple risk factors for NEC in term infants have been reported, including intrauterine growth retardation, birth asphyxia, congenital heart disease, gastroschisis, polycythemia, hypoglycemia, sepsis, exchange transfusion, blood transfusions, the use of umbilical catheters, milk-protein allergy, premature rupture of membranes, chorioamnionitis and gestational diabetes.

Role of the Microbiome

The intestine must balance protection against harmful environmental antigens, such as pathogenic bacteria, while housing beneficial commensal bacteria. This balance is important in maintaining intestinal homeostasis. If disrupted, intestinal inflammation and injury may play a role in NEC. Commensal bacteria help protect against intestinal injury, maintain intestinal homeostasis and immune function, and support digestion of food. Preterm infants have delayed acquisition of beneficial commensal bacteria such as Bifidobacterium and are more likely to be colonized with virulent pathogens. A number of factors may influence this colonization. Decreased human milk feeding can result in decreased colonization of beneficial commensal bacteria such as Bifidobacterium, Lactobacillus, and Bacteroides. Prolonged empiric antibiotic therapy has been associated with an increased risk of NEC in several studies, and antibiotic treatment after birth results in decreased Bifidobacterium colonization. In addition, cesarean delivery may result in decreased Bifidobacterium, Lactobacillus, and Bacteroides colonization. Several studies have shown that infants who go on to develop NEC have changes in the prevalence of Clostridium perfringens, Firmicutes, Proteobacteria, and Enterobacter before the onset of NEC, as well as decreases in Bifidobacterium, with some changes occurring soon after birth. In a longitudinal study of 122 VLBW infants, those who went on to develop NEC had a relative abundance of Gram-negative facultative bacilli (Gammaproteobacteria) and a paucity of strict anaerobic bacteria, particularly Negativicutes. Additionally, changes in the microbiome that precede NEC may occur several weeks before NEC onset, and infants with NEC have lower microbial diversity. Finally, Toll-like receptors (TLRs) recognize microbial associated molecule patterns (MAMPs) such as lipopolysaccharide and peptidoglycan, which are part of the cell wall of bacteria. Toll-like receptor 4 activation has been implicated as a major pathway of NEC and can impair intestinal perfusion through changes in endothelial nitric oxide synthase signaling, which is ameliorated by administration of nitrate, a component of breast milk. In addition, commensal and probiotic bacteria have been shown to decrease TLR-mediated inflammation.

PREVENTION

A number of systematic reviews and meta-analyses of randomized studies evaluating treatments that include NEC as either a primary or secondary outcome have been conducted, and are summarized in Table 60-2.

Breast Milk

Breastfeeding of maternal milk has long been established as a key, if not most important, part of a strategy to prevent NEC. The risk of NEC is inversely related to the amount of breast milk exposure. There are a number of components of human milk, including oligosaccharides, lactoferrin, probiotics, epidermal growth factor, long-chain polyunsaturated fatty acids, platelet activating factor acetylhydrolase, immunoglobulins (IgA and IgG), arginine, and glutamine, that may mediate the benefits of breast milk in reducing NEC. When maternal breast milk is not available, donor human milk also reduces the risk of NEC in VLBW infants, as demonstrated in a large, multicenter randomized trial in Canada. In this trial, the incidence of NEC (Bell stage 2 or greater) was lower in infants fed donor milk using bovine-based fortifiers compared to infants fed preterm formula (1.7% vs 6.6%; P = 0.02). The trial showed no differences in the primary outcome of mean cognitive score at 18 months corrected age or in growth parameters. However, infants receiving donor milk, compared to infants receiving formula, were more likely to have a cognitive score less than 85 (27.2% vs 16.2%; P = 0.02) in analysis of secondary outcomes, raising potential concerns for the adverse long-term effects of donor human milk. Another randomized trial found an exclusive human milk diet with human milk–based fortifiers, compared to preterm formula, resulted in a decrease in the risk of severe NEC requiring surgery. Similar findings of a decrease in NEC with a human milk diet compared to a diet containing bovine products was reported in a multicenter observational study. In a systematic review completed before these studies, infants receiving donor milk demonstrated decreased weight, length, and head circumference compared to those receiving formula, although the comparisons were limited by inadequate fortification of donor milk. Another recent multicenter trial in the Netherlands found no significant effect of donor milk on severe infections or on mortality, and additional trials are ongoing.

Probiotics

Given that an abnormal intestinal microbiome has been described in the pathogenesis of NEC, modifying the microbiome through the administration of bacteria that are contained in probiotic preparations or by administering prebiotics that support the growth of so-called health-promoting bacteria is a logical approach to prevention of NEC. Preclinical studies have shown that certain probiotics decrease proinflammatory mediators; increase anti-inflammatory mediators; support pathways important in growth, differentiation, and cytoprotection; decrease apoptosis; generate reactive-oxygen species important in regulating apoptotic, proliferative, and inflammatory signaling; and support intestinal barrier function.

Pooled results of randomized trials of probiotic supplementation in preterm infants show a decreased risk of NEC, mortality, and late-onset sepsis. However, questions regarding the optimal dose and preparation of probiotic therapy, lack of regulated commercial products, and risk of probiotic-associated sepsis have limited widespread use. A recent survey of 500 NICUs in the United States found that 14% were using probiotics in VLBW infants, with single-strain preparations being the predominant products utilized, many of which have not been extensively studied in preterm neonates. In addition, a recent large, multicenter trial found no effect of Bifidobacterium breve on the risk of NEC. This trial highlights the continued uncertainty regarding the effect of specific probiotic preparations on the risk of NEC. There is an inherent fallacy when one argues for the use of probiotics, encompassing a broad range of different preparations for the prevention of NEC. Each preparation should be considered individually rather than as a group. Caution in applying pharmacologic standards is needed when using these preparations in a vulnerable population such as preterm infants.

Additional Strategies

Feeding Strategies

There is limited evidence to support any single feeding strategy to reduce the risk of NEC, aside from the use of human milk, which is already a standard. Numerous studies in animals show that lack of enteral feeding leads to intestinal atrophy, exaggerated inflammatory responses, increased bacterial translocation, and risk for intestinal damage. Total parenteral nutrition, delayed initiation of enteral feeding, and a longer duration to reach full enteral feeding are also associated with increased risk for late-onset sepsis in VLBW infants, which may be due, in part, to increased intestinal microbial translocation. A recent study suggests that later onset of feedings is associated with greater inflammatory response as well as increased adverse outcomes, such as chronic lung disease and retinopathy of prematurity. Meta-analysis of randomized trials found no benefit of early trophic feeding, compared to no feeding, or slow advancement of feeding, compared to rapid advancement of feeding, on the risk of NEC (Table 60-2). However, there was substantial clinical heterogeneity among the study populations with small sample sizes and imprecise estimates of treatment effects in the pooled estimates. The use of feeding protocols in observational studies is associated with a reduction in the incidence of NEC, although feeding protocols differ in approaches. The role of aspiration to evaluate gastric residuals as part of feeding is controversial, although larger volumes of gastric residuals have been associated with NEC in preclinical and clinical studies. A recent pilot trial that randomized 61 preterm infants to routine monitoring of gastric residuals versus no routine monitoring demonstrated no differences in NEC or other adverse outcomes.

Avoidance of Acid Suppression Medications and Prolonged Antibiotic Exposure

Use of acid-suppression medications such as H₂ antagonists are associated with NEC, as is prolonged empiric antibiotic therapy in the setting of negative blood cultures. These H₂ antagonists reduce gastric pH and may lower bacterial diversity and increase the relative abundance of pathogenic bacteria by reducing the bactericidal effect of stomach acid on microbes entering the gastrointestinal tract. Of interest is that H₂ antagonist use is associated with increases in intestinal Proteobacteria, a phylum that has been shown to expand in abundance prior to NEC.

Similarly, antibiotic duration for greater than 3 days has been associated with a higher risk of colonization with pathogenic Gram-negative bacteria such as Klebsiella, Enterobacter, or Citrobacter. Although data from randomized trials are lacking, reducing the use of H₂ antagonists and other acid-suppression medications, and limiting prolonged empiric antibiotic therapy given negative cultures may help to decrease NEC. Use of quality improvement approaches can reduce the use of antibiotics and H₂ antagonists. Pragmatic randomized trials of the effects of shorter courses of antibiotics on the microbiome as well as short-term adverse outcomes are likely to provide important new information in this area.

Delayed Cord Clamping, Oxygen Delivery, and Anemia

Additional potential risk factors for NEC include impaired intestinal oxygen delivery from anemia, abnormal regulation of intestinal blood flow, or lower oxygen saturation targets. The role of red blood cell transfusion in the pathogenesis of NEC is controversial and may be a marker of illness severity among infants who develop NEC rather than a causal factor. However, delayed clamping of the umbilical cord to allow for placental transfusion of blood may reduce the risk of NEC. However, more precise estimation of the effect of delayed cord clamping on NEC and longer-term data regarding safety in ELBW infants is needed.

Biomarkers

Multiple biomarkers for NEC have been studied in the stool, serum, and urine in preterm infants. Some of the biomarkers that have been evaluated to date include urinary intestinal fatty acid–binding protein, urinary claudin-3, serum amyloid A, and fecal calprotectin. More recently an “electronic-nose” technique to detect fecal volatile organic compounds has been suggested as being potentially efficacious in the early diagnosis of NEC. All these techniques await rigorous validation prior to routine clinical use.

CONCLUSION

Although recent studies suggest the incidence of NEC has been decreasing in the United States, it remains a major cause of death and serious morbidity in preterm infants. New approaches have provided insight into the role of the microbiome in the pathogenesis of NEC. Clinical studies have provided additional data to support the use of human milk and avoid harmful medications. However, additional studies are needed to better define and diagnose NEC and to identify biomarkers effective in identifying high-risk infants as well as safe, effective, and reliable prevention strategies in order to improve outcomes among infants who are diagnosed with NEC.

SUGGESTED READINGS

INTRODUCTION

Neonatal germinal matrix hemorrhage/intraventricular hemorrhage (GMH/IVH) is the most common form of intracranial hemorrhage in preterm infants, and an important cause of long-term morbidities among survivors of neonatal intensive care. In its least severe form, the hemorrhage may be restricted to the germinal matrix area or subependymal zone. However, it frequently extends beyond the germinal matrix into the cerebral ventricles, following rupture of the germinal matrix hemorrhage into the ventricular space.

INCIDENCE

Germinal matrix hemorrhage/intraventricular hemorrhage occurs most frequently among preterm infants. Its incidence decreases as gestational age and birth weight increase. Because of improvement in perinatal and neonatal care, the incidence of GMH/IVH has decreased over the past several decades. However, this decline is of a lesser extent than that in neonatal mortality. Approximately 3 decades ago, GMH/IVH was noted in 45% of very low–birth-weight (VLBW) infants and in greater than 60% of those with birth weight of 750 grams or less. By the early 1990s, the incidence of GMH/IVH decreased to 30% of VLBW infants, and to 40% among the extremely low–birth-weight (ELBW) survivors, and by the mid 1990s had decreased further to 24%. Since that time, however, the incidence has remained static, remaining about 25% among all VLBW infants based upon data from the Vermont Oxford Network database (Fig. 61-1A). The more severe hemorrhages, grades III and IV, occur in 12% to 17% of ELBW infants and in as high as 20% of those with birth weight of 750 grams or less (Fig. 61-1B). These severe hemorrhages occur in 17% to 28% of those born at 22 to 26 weeks’ gestation and in 9% to 14% of those 27 to 32 weeks’ gestation.

PATHOGENESIS AND RISK FACTORS

The pathogenesis of GMH/IVH is complex and involves multiple risk factors and disorders, with biochemical, inflammatory, and hemodynamic alterations and predisposing to the development of hemorrhage. Ultimately, the mechanisms for predisposition to hemorrhage, and the relationship among GMH/IVH and white matter and neuronal injury may be linked to the generation of free radicals, reactive oxygen and nitrogen species, and gene-environment interactions.

Hemorrhage originates in the germinal matrix which is a prominent structure in the preterm brain during the second and early third trimesters that subsequently undergoes involution as gestation reaches term. The germinal matrix structure contains neuronal precursor cells prior to 20 weeks of gestation. As development progresses, the differentiating glioblasts give rise to oligodendroglial cells, which are important for myelination. The germinal matrix is supplied by a complex vascular network but is a low blood flow structure. The hemorrhage appears to originate from the endothelial-lined vessels of the matrix, in particular, vessels in communication with the venous circulation, including capillary-venule junctions and small venules. These thin-walled germinal matrix capillaries are susceptible to rupture, for several potential reasons. Their large diameter offers lesser resistance to changes in intravascular pressure, to the immature endothelial cell tight junctions, and to low levels of structurally stabilizing proteins such as fibronectin and collagen in the extracellular matrix. Rupture of the capillaries results when the endothelial wall loses integrity and/or when there is imbalance of pressure between the intravascular lumen and the extravascular space.

Clinical risk factors associated with IVH include lower gestational age, lack of antenatal corticosteroids, clinical chorioamnionitis, male sex, significant delivery room resuscitation, mechanical ventilation, and hypotension requiring multiple inotropes. In Table 61-1, the predisposing cerebrovascular factors and alterations in cerebral hemodynamics that may result from a variety of antenatal and postnatal conditions are listed. As seen in the schematic diagram in Figure 61-2, a single factor or a combination of factors may disrupt the cerebral circulation, leading to GMH/IVH and its complications.

CLINICAL MANIFESTATIONS AND DIAGNOSIS

Germinal matrix hemorrhage/intraventricular hemorrhage is usually detected in the first 4 to 5 days of life with approximately 40% to 50% occurring on the first day of life, and up to 90% within the first 72 hours. Approximately 20% of early-onset hemorrhage may evolve to become more severe, with the maximum extent typically seen within the first week of life. The majority of hemorrhages, in particular, small ones, are asymptomatic and are only detected by routine cranial ultrasound. However, larger cases of GMH/IVH may present with sudden clinical deterioration, especially when there is a significant blood loss. Other clinical manifestations include anemia or failure; seizures; tense, full, and/or bulging fontanels; split and wide sutures; apnea and bradycardia with desaturation episodes; poor perfusion; hypotension; severe metabolic acidosis; increase in oxygen requirement; and increase in ventilator support.

Imaging studies will establish the diagnosis of GMH/IVH. The current standard approach to diagnosis is the use of cranial ultrasonography. Findings on ultrasound have been interpreted with a high level of agreement among practitioners. Ultrasound can be performed at the bedside, thus avoiding the risks of transporting the infant for either computed tomography (CT) or magnetic resonance imaging (MRI) scans. Table 61-2 shows the grading of GMH/IVH, based on Papile criteria from CT scanning still employed by clinicians and researchers for cranial ultrasound findings. Others have added findings such as cystic periventricular leukomalacia (PVL) to indicate associated white matter injury and measurements of the lateral ventricles to define mild to severe ventriculomegaly. Volpe proposed a modification to the grading system for cranial ultrasound findings. This newer modified grading system takes into consideration the location of hemorrhage and the amount of blood detected in the ventricles (Fig. 61-3).

COMPLICATIONS AND NEUROPATHOLOGICAL ASSOCIATIONS

Germinal matrix hemorrhage/intraventricular hemorrhage may cause the destruction of the germinal matrix, periventricular hemorrhagic infarction, ventriculomegaly or hydrocephalus, white matter injury, and loss of cerebellar volume—each of which may have implications for the prognosis of the infant.

During the second trimester, the ganglionic eminence and subventricular zone, which are part of and immediately associated with the germinal matrix, respectively, are rich locations for neuronal and glial proliferation. Therefore, destruction of the germinal matrix during this vulnerable period leads to a reduction in proliferation and migration of neuronal and glial precursor cells, which, in turn, may impact brain growth and, potentially, the developmental outcome of these infants.

Periventricular hemorrhagic infarction (PVHI) can be seen as large echodensities on cranial ultrasound within the parenchyma adjacent to the GMH/IVH. Papile initially described this as extension of the IVH into the parenchyma. However, subsequent study has indicated that while PVHI is strongly associated with IVH, it is not an extension of the IVH; rather, it occurs due to obstruction of venous drainage leading to a venous hemorrhagic infarct. The area of infarction evolves into tissue loss such that a porencephalic cyst may be noted on later follow-up cranial ultrasound.

Periventricular leukomalacia (PVL) is a form of white matter injury that is commonly associated with GMH/IVH. It occurs in the periventricular arterial border zones. Although the exact mechanism of injury is not fully defined, it is thought to occur secondary to ischemia and inflammation, with associated glial activation, and damage of the preoligodendrocytes.

PVL may be either focal or diffuse. Focal PVL is classically described as macroscopic areas of necrosis. These initially appear as echodense lesions in the periventricular area with or without blood in the ventricles. After several weeks, these echodense areas can become cystic, and are referred to as cystic PVL (Fig. 61-4). Cystic PVL occurs in less than 5% of very low–birth-weight infants, and represents the minority of PVL. More commonly, focal PVL consists of microscopic areas of necrosis, and is termed noncystic PVL. Diffuse PVL is characterized by a loss of preoligodendrocytes, rather than areas of necrosis. It leads to hypomyelination and is associated with ventriculomegaly due to decreased cerebral tissue volume. Cranial ultrasound has limited use in identifying noncystic and diffuse PVL, with MRI being a superior neuroimaging modality to screen for these types of injuries. The MRI findings among these infants include abnormalities on diffusion-weighted imaging and diffuse signal abnormalities.

About 2 to 3 weeks after the initial detection of GMH/IVH, some infants may present with increasing head circumference, separation of sutures, full and tense fontanels, and sun-setting appearance. A repeat cranial ultrasound in the presence of these manifestations of increased intracranial pressure will reveal prominence and enlargement of the cerebral ventricles (ventriculomegaly), and thus a diagnosis of posthemorrhagic hydrocephalus. Posthemorrhagic hydrocephalus results from a disturbance in cerebrospinal fluid (CSF) dynamics because of (1) the obstruction of the CSF pathway by blood clots in the posterior fossa cisterns, aqueduct of Sylvius, or foramen of Monro, and (2) postinflammatory changes in the arachnoid villi that may impair CSF absorption. Obstruction and delayed absorption of CSF leads to a progressive increase in ventricular size. Among those with GMH/IVH who survive for more than 14 days, 50% go on to develop ventricular dilation, which will progress in half of these infants. In 62% of those with progressive ventricular dilation, spontaneous arrest occurs, while the remaining 38% ultimately require nonsurgical or surgical treatment. Surgical intervention to relieve the increase in intracranial pressure and ameliorate associated clinical manifestations will be necessary for rapidly increasing head size, frequent apnea and bradycardia, and the need for increasing ventilatory support.

Ventriculomegaly may also occur without increased intracranial pressure (hydrocephalus ex-vacuo). This differs from posthemorrhagic hydrocephalus and may be a result of brain atrophy from periventricular white matter injury. It rarely requires surgical treatment. Preterm infants with GMH/IVH without obvious parenchymal involvement may have evidence of white matter injury detected by MRI diffusion-weighted imaging at term postmenstrual age; diffuse, excessive, high signal intensity is noted in the cerebral white matter on T2-weighted images.

Recent studies demonstrate that GMH/IVH is associated with reductions in cerebellar volume. This is thought to occur secondary to (1) a direct toxic effect of hemosiderin on the surface of the cerebellum, impairing the proliferation of the external granular layer, and (2) the loss of cortical neuronal inputs leading to underdevelopment of the cerebellum on the contralateral side from the site of the cortical injury (a maturation-distinctive form of diaschisis).

MANAGEMENT

The treatment of GMH/IVH is primarily supportive. Management strategies include initiation of mechanical ventilation or increase in ventilatory support, and/or administration of oxygen in order to maintain optimal levels of the arterial partial pressure of carbon dioxide (PaCO₂) and the arterial partial pressure of oxygen (PaO₂); treatment of hypotension with slow volume expansion and then cautious use of pressors if unresponsive; blood transfusion to correct anemia from blood loss; correction of metabolic acidosis; anticonvulsant therapy for seizures; and administration of fresh frozen plasma, platelets, and other products if there is associated coagulopathy. Progression or evolution of GMH/IVH is monitored by serial cranial ultrasounds, especially for decision-making and parental counseling in the face of rapid clinical deterioration.

During early posthemorrhagic hydrocephalus, interventions such as diuretics and intraventricular fibrinolytic therapy have been trialed and shown to have no benefit in small randomized trials. The infants must be monitored with frequent cranial ultrasounds to examine changes in ventricular size. If there is increasing ventriculomegaly and raised intracranial pressure, the management is then directed at drainage of the CSF fluid to reduce the ventricular size. This can be done either by serial lumbar puncture or surgical intervention with the creation of a ventricular reservoir, or drainage of CSF by ventriculostomy. The optimum timing of intervention is unclear; however, retrospective data suggests that earlier intervention with serial therapeutic lumbar punctures may avoid the need for surgical intervention and improve neurodevelopmental outcome. A multicenter randomized trial of early versus late ventricular intervention study (ELVIS; ISRCTN43171322 or NCT00875758) is currently ongoing to validate this retrospective data.

Ventriculoperitoneal shunt is the definitive surgical treatment when there is continued progression of ventriculomegaly accompanied by increase in intracranial pressure. Shunt obstruction, malfunction, and infection can complicate shunt placement and long-term function. Endoscopic third ventriculostomy with choroid plexus cauterization may avoid the necessity for a shunt among those requiring surgical intervention. However, while the data is encouraging in hydrocephalus due to alternate etiologies, it has been less successful for the management of posthemorrhagic hydrocephalus in preterm infants, and ventriculoperioneal shunt remains the gold standard.

PROGNOSIS AND LONG-TERM OUTCOMES

In severe GMH/IVH, sudden deterioration unresponsive to escalating support may be observed. Mortality in severe GMH/IVH, especially with associated periventricular hemorrhagic infarction, is 40%. Among those who survive for weeks after a hemorrhage with complicating progressive and persistent ventricular dilation, death occurs in 18%. In extremely low–birth-weight infants, after excluding deaths due to extreme immaturity, 7% to 9% of the deaths are attributable to hemorrhage.

Forty percent of ELBW infants who survive following a grade III or IV GMH/IVH have moderate-to-severe neurosensory impairments, including cerebral palsy, which occurs in 30% of survivors; severe neurodevelopmental impairment (developmental quotients 2 standard deviations below the mean) in 15% to 20%; bilateral deafness in 8%; and blindness in 2%. Among infants with lower grades of IVH (grade I and II), there is a 2-fold increase in the risk for neurodevelopmental impairment and a 2.6-fold increase in the risk of cerebral palsy, as compared to children of similar birth weight but normal cranial ultrasound. An alternate large cohort study of infants less than 29 weeks’ gestation, reported that 20% of those with grade I or II IVH had moderate-to-severe neurosensory impairment, with 10% having cerebral palsy, and 8% having severe neurodevelopmental impairment (developmental quotients 2 standard deviations below the mean). However, even in the absence of abnormalities on cranial ultrasound, 10% of extremely low–birth-weight children are reported to have a moderate-to-severe neurosensory impairment at 18 to 22 months of age.

PREVENTION

Several studies have addressed the prevention of IVH from the antenatal through the postnatal periods. Antenatal prevention is directed toward prevention of preterm birth through use of tocolytics, treatment of maternal complications (bleeding, chorioamnionitis, conditions that may predispose to preterm delivery), intrapartum fetal surveillance, the use of epidural rather than general anesthesia, controlled vaginal delivery, and administration of antenatal steroids. Antenatal steroids in particular have been repeatedly shown to significantly reduce the incidence of severe IVH. Data from the Neonatal Research Network finds the odds ratio for a severe IVH following a complete course of antenatal steroids is 0.39 (95% confidence interval, 0.27–0.57).

Postnatal preventive strategies include supportive measures, such as resuscitative measures at delivery with a goal of preventing hyperoxia; maintenance of optimal oxygenation and acid–base balance; gentle ventilation to prevent pneumothoraces and other air-leak syndromes; minimizing abrupt hemodynamic alterations, including slow volume expansion for low blood pressure and judicious use of inotropic agents (there are risks associated with sudden changes in systemic pressures in the presence of a pressure-passive cerebral circulation); and careful surfactant administration to improve respiratory status and oxygenation. Studies of pharmacologic agents such as pancuronium, ethamsylate, vitamin E, and phenobarbital do not demonstrate clear benefit.

Indomethacin, a prostaglandin H synthase inhibitor, inhibits generation of oxygen free radicals, promotes maturation of the germinal matrix, stabilizes cerebral blood flow, and attenuates the hyperemic response to hypoxia and hypercapnia. Indomethacin reduces the incidence of severe IVH, but despite this, its use has not resulted in an improvement in neurodevelopmental outcome. Additionally, administration of indomethacin has been associated with an increased risk of renal insufficiency, ileal perforation, and chronic lung disease. Therefore, despite the clear evidence for a reduction in severe IVH, due to the potential for adverse side effects and no clear impact on developmental outcome, the use of prophylactic indomethacin has remained limited. Recently, some groups have begun to explore a more targeted approach for prophylactic indomethacin, providing it to those at highest risk of a severe IVH, and therefore most likely to benefit.

SUGGESTED READINGS

INTRODUCTION

Bronchopulmonary dysplasia (BPD) is the most common chronic complication associated with preterm birth. In the United States, it affects between 10,000 and 15,000 infants annually, including as many as 50% of extremely low–birth-weight (ELBW; birth weight < 1000 g) infants. BPD is a leading cause of mortality after the first month of life in extremely preterm infants and is a strong predictor of chronic respiratory and cardiovascular impairments, growth failure, and neurodevelopmental delay.

The first descriptions of BPD appeared in the literature soon after the advent of modern day neonatology. In the late 1950s and early 1960s, infant respiratory distress syndrome (RDS) was a leading cause of neonatal death among preterm infants. Increased use of mechanical ventilation improved survival rates for infants with severe RDS. However, prolonged treatment with invasive respiratory support and supplemental oxygen therapy was often necessary. These exposures contributed to chronic lung injury in surviving moderate preterm infants and the development of what Northway and colleagues first termed bronchopulmonary dysplasia in 1967.

As the use of antenatal corticosteroids, noninvasive continuous positive airway pressure (CPAP), and exogenous surfactant therapy increased during the subsequent decades, the epidemiology and pathophysiology of BPD evolved. It is now an infrequent complication among infants born with birth weights greater than 1500 g and gestational ages exceeding 32 weeks. In the presurfactant era, prominent airway injury, epithelial metaplasia, smooth muscle hypertrophy, and alternating parenchymal fibrosis and emphysema were the most common histological findings. This disease phenotype is often termed “old” BPD. The most common lung histology in BPD now commonly displays a homogenous disease pattern marked by reduced numbers of alveoli and pulmonary capillaries, and minimal areas of hyperinflation and focal collapse. This is termed “new” BPD. Importantly, the pathologic characteristics of both old and new BPD can develop in extremely preterm infants who require prolonged intubation and mechanical ventilation. Old and new BPD therefore likely represent a continuum of disease severity and may not be fully distinct disease entities.

INCIDENCE

The incidence of BPD is highest among infants born with the lowest gestational ages and birth weights. The National Institutes of Child Health and Human Development’s (NICHD) Neonatal Research Network (NRN) reported rates of supplemental oxygen use at 36 weeks postmenstrual age (PMA) among infants who survived to that time point. BPD rates ranged from 85% for those born at 22 weeks’ gestation to 23% for those born at 28 weeks’ gestation (Fig. 62-1). Among preterm infants born in Israel between 2000 and 2010, 31% of surviving ELBW infants received supplemental oxygen at 36 weeks PMA compared to 13.7% of very low–birth-weight (VLBW; birth weight < 1500 g) infants. BPD severity is also inversely proportional to gestational age and birth weight. Fifty-six percent of infants born at 22 weeks’ gestation in the NRN cohort were diagnosed with severe BPD based on the National Institutes of Health (NIH) consensus definition, compared to only 8% of those born at 28 weeks’ gestation (Fig. 62-1).

Finally, BPD rates vary depending on the definition of BPD applied to the cohort. The NRN compared 3 different definitions of BPD among infants with gestational ages between 22 and 28 weeks. Rates based on the NIH consensus definition were the highest (68%), as this includes infants who received supplemental oxygen for at least 28 days but were breathing in room air by 36 weeks PMA. Forty-two percent of the evaluated extremely preterm infants were diagnosed with BPD based on supplemental oxygen use at 36 weeks PMA and 40% by the physiologic definition.

CURRENT DIAGNOSTIC CRITERIA

Along with the disease epidemiology, the diagnostic criteria for BPD have also evolved. The first widely used definition required supplemental oxygen exposure for at least 28 days and radiograph evidence of chronic lung injury. Subsequently, an oxygen requirement at 36 weeks PMA was shown to better predict adverse pulmonary outcomes at 2 years of life among VLBW infants. Many reports continue to define BPD using this criterion. In 2000, an NIH workshop recommended a graded diagnosis for BPD that incorporated both an oxygen requirement for greater than 28 days and the respiratory support utilized at 36 weeks PMA (Table 62-1). The presence of abnormal radiographic findings was not required. To address variability in target oxygen saturation ranges used in clinical practice, Walsh and colleagues proposed a “physiologic” definition for BPD to more objectively determine an infant’s “need” for supplemental oxygen. This definition requires an oxygen reduction test to determine dependency at 36 weeks PMA among those receiving 30% or less supplemental oxygen. Unfortunately, the physiologic test is not consistently performed in clinical or research settings. Finally, recent data indicate that many former extremely preterm infants who do not meet the current diagnostic criteria for BPD experience deficits in respiratory health through childhood and into adulthood. These observations prompt continued efforts to develop more reliable means to diagnose and quantify chronic respiratory disease associated with extreme preterm birth.

PATHOPHYSIOLOGY

In the presurfactant era, BPD was primarily attributed to chronic, postnatal lung injury from prolonged mechanical ventilation and supplemental oxygen therapy. In the current era, new BPD is increasingly recognized as a consequence of arrested lung alveolarization and pulmonary capillary formation following birth during the late canalicular or early saccular stages of lung growth. Although the terminal bronchial epithelium has sufficiently thinned to permit life-sustaining ex-utero gas exchange by this time, alveolar septation has not begun. Postmortem lung histology in extremely preterm infants with BPD demonstrates alveolar simplification, reduced secondary septation, and abnormal microvascular growth. In addition, multiple antenatal and postnatal factors, including infection, hyperoxia, and volu-/baro-/atelectrauma, contribute to the development of BPD by triggering inflammatory lung injury and cellular apoptosis. Abnormalities in extracellular matrix turnover and impaired structural remodeling follow these exposures and result in disordered elastin deposition, saccular wall fibrosis, and greater alveolar simplification.

RISK FACTORS

Epidemiological studies have identified numerous individual risk factors for BPD (Table 62-2). However, BPD likely results from the combined effect of multiple perinatal and postnatal exposures on the developmentally immature lung rather than from a single risk factor.

Perinatal Risk Factors

Prematurity

Prematurity is the most important risk for BPD. Fewer than 5% of infants born at or beyond 30 weeks’ gestation develop BPD compared to 50% or more of surviving infants born at 24 to 25 weeks’ gestation. Fetal growth restriction and being small for gestational age further increase the risk for BPD. The odds of BPD are 2- to 3-fold higher among VLBW infants born with birth weights less than the 10th percentile for their gestational age. Finally, preterm male infants are twice as likely to develop BPD compared to female infants born at similar gestational ages and birth weights.

Chorioamnionitis

Chorioamnionitis predisposes to preterm birth and is a proposed risk factor for BPD. In animal models, chorioamnionitis arrests lung alveolar and vascular development producing a BPD-like phenotype. However, the available evidence from large human studies is inconclusive as to whether chorioamnionitis is a true BPD risk factor. A 2011 meta-analysis found that preterm infants born to mothers with histologically but not clinically diagnosed chorioamnionitis were at increased risk for BPD. However, after adjusting for potential confounders, the association was no longer statistically significant. An important distinction may be the bacterial agent responsible for chorioamnionitis. A meta-analysis of 39 studies found that preterm infants with respiratory tract colonization by Ureaplasma spp, the most common bacterial organisms cultured in chorioamnonitis, were at increased risk for supplemental oxygen use at 28 days and 36 weeks PMA.

Postnatal Risk Factors

Mechanical Ventilation and Supplemental Oxygen Exposure

Mechanical ventilation and supplemental oxygen exposure are the best-described postnatal risk factors for BPD. Animal models show a clear link between mechanical ventilation and oxygen free-radical–induced lung injury that mimics human BPD. However, epidemiological evidence of a causal link between these exposures and BPD is less robust. Meta-analyses of randomized controlled trials show that the avoidance of mechanical ventilation in extremely preterm infants only leads to a small reduction in BPD risk. Importantly, BPD also develops in infants who have never required tracheal intubation and only had low supplemental oxygen requirements. Moreover, clinical trials have failed to demonstrate any impact of protective measures such as minimizing supplemental inspired oxygen during resuscitation, or targeting a lower saturation range, on the risk of BPD in extremely preterm infants.

Patent Ductus Arteriosus

The presence of a patent ductus arteriosus (PDA) beyond the first days of life in preterm infants is another well-established risk factor for BPD. Whether this relationship is causal is unclear. Randomized controlled trials of prophylactic medical and/or surgical closure of a PDA showed no benefit for prevention of BPD. For example, the Trial of Indomethacin Prophylaxis in Preterms (TIPP) found no reduction in BPD risk following prophylactic indomethacin despite a decrease in the incidence of PDA by more than 50%. In a small RCT, surgical ligation of a PDA within the first 24 hours of life compared to conservative management increased the risk for supplemental oxygen use and mechanical ventilation at 36 weeks PMA. Finally, a systematic review of all RCTs evaluating surgical and pharmacological methods of PDA closure at any time in the neonatal period found no reduction in BPD or the composite outcome of death or BPD with any therapy.

Sepsis

Observational studies implicate postnatal sepsis as an independent risk factor for subsequently developing BPD. Although the direct role of infection is unclear, systemic inflammation and increased vascular permeability may cause immediate alveolar injury and long-term disruption of alveolarization. Of note, the risk for BPD may be further increased among infants diagnosed with postnatal sepsis who were exposed to chorioamnionitis in utero. Perinatal and postnatal airway colonization with Ureaplasma spp and Gram-negative rod bacteria are also associated with increased risk for BPD and later adverse respiratory outcomes.

CLINICAL PRESENTATION

Exposure to invasive mechanical ventilation and supplemental oxygen during the first days and/or weeks of life is common among extremely preterm infants who develop BPD. However, some infants with minimal early respiratory support needs also develop BPD. The physical examination in BPD is highly variable and depends on the extent of pulmonary disease and the respiratory support. Classical signs of respiratory insufficiency such as tachypnea, chest wall retractions, and grunting or expiratory braking may be present. Auscultation may also reveal scattered areas of poor aeration, rales, crackles, and/or wheezing, depending on the presence and distribution of pulmonary edema, atelectasis, and airway narrowing. Importantly, the respiratory physical examination may also be largely normal in BPD, even in severe disease, when adequate lung aeration is achieved using positive airway pressure.

Radiographic findings in BPD vary according to the severity of the illness and the stage of the disease. Initial chest radiographs in extremely preterm infants commonly show low lung volumes with a homogenous fine granular or “ground glass” appearance consistent with RDS. With time, a more heterogeneous pattern marked by diffuse haziness, coarse interstitial densities, and small cyst-like structures may appear. In mild to moderate BPD, lung volumes appear normal or decreased depending on the presence and severity of atelectasis (Fig. 62-2). In contrast, the lungs of infants with severe BPD are often hyperinflated with larger cystic areas and coarse interstitial densities (Fig. 62-2). Acquired lobar emphysema and right ventricular enlargement from cor pulmonale may become apparent. The radiographic abnormalities seen in surviving infants with BPD often improve over time, although residual scarring can persist into adulthood.

Chest computed tomography (CT) is increasingly utilized to assess lung disease in BPD due to its greater sensitivity for identifying focal abnormalities such as acquired lobar emphysema and localized atelectatic or infectious consolidations. Chest CT with angiography can also complement echocardiography and cardiac catheterization assessment of pulmonary hypertension.

PREVENTION

Prematurity

It is unlikely that a single, postnatal intervention aimed at preventing BPD will be fully effective. As a result, any comprehensive effort to reduce BPD must include strategies to prevent extreme preterm birth. Although birth rates near the limits of viability improved slightly over the past decade in the United States, additional evidence-based interventions are needed to produce the substantial reductions in extreme preterm birth required to eradicate BPD.

Avoidance of Mechanical Ventilation

Successful transition after birth requires lung aeration. Unfortunately, most extremely preterm infants are unable to independently clear fetal lung fluid and fully aerate the lungs. Weak respiratory muscles, a highly compliant chest wall, underexpression of transepithelial sodium channels, and deficient production of surfactant complicate this transition. Consequently, many require positive airway pressure and supplemental oxygen after birth. The optimal means to provide these potentially lifesaving interventions while minimizing lung injury remains an area of intense research.

Application of noninvasive CPAP immediately after birth facilitates lung recruitment while preventing stretch injury caused by the higher distending pressures often administered with invasive ventilation. Several large RCTs compared a strategy of early noninvasive CPAP to immediate intubation and surfactant administration. Meta-analyses of these trials found a small but statistically significant reduction in the risk for death or BPD at 36 weeks PMA among CPAP-treated infants. The 2 largest trials also showed a significant reduction in duration of mechanical ventilation and a trend toward shorter duration of oxygen exposure with early CPAP use.

Less-Invasive Surfactant Administration

Less-invasive surfactant administration (LISA) is intended to provide the potential benefits of surfactant therapy without the harm of intubation. This instillation method utilizes a thin catheter (eg, an angiocatheter or small feeding tube) instead of a standard endotracheal tube to administer surfactant. Two meta-analyses of small RCTs and a Bayesian random-effects network meta-analysis suggest that LISA may be a more optimal approach than other early respiratory management strategies, such as CPAP alone, for preventing BPD.

Pharmacological Therapy

Caffeine

Caffeine stimulates breathing in preterm infants by increasing minute ventilation, CO₂ sensitivity, and diaphragmatic activity. The international Caffeine for Apnea of Prematurity (CAP) trial showed that caffeine reduced the risk for BPD and neurodevelopmental impairment. Caffeine is now the most commonly prescribed medication after antibiotics in the neonatal intensive care unit (NICU).

Postnatal Corticosteroids

Dexamethasone, a synthetic corticosteroid and potent anti-inflammatory, is effective at reducing BPD. However, concerns for increased risk of neurodevelopmental impairment, particularly cerebral palsy, limit its use. Meta-analysis of available RCTs show that dexamethasone should not be used during the first week of life. Such “early” treatment increases the risk for gastrointestinal perforation, hypertrophic cardiomyopathy, cerebral palsy (CP), and major neurosensory disability. In contrast, dexamethasone use after the first week of life has not been found to increase the risk for CP. Adverse effects of this “late” dexamethasone therapy include transient hyperglycemia, glycosuria, and hypertension, as well as a possible increase in the risk for severe retinopathy of prematurity. Two meta-regressions by Doyle and colleagues highlight the importance of balancing the risks of neurodevelopmental impairment due to corticosteroids against those of BPD, itself a strong predictor of neurologic impairment. If the risk for BPD in the control population was less than approximately 33%, corticosteroid therapy significantly increased the risk for death or CP; whereas when the risk for BPD exceeded 60%, they reduced the risk for death or CP. Combined with the meta-analysis data, these results suggest that the adverse long-term effects of dexamethasone when used the first week of life and in infants at low risk for BPD, outweigh its benefits. However, among infants who continue to require invasive mechanical ventilation after the first 1 to 2 weeks of life, the balance may favor the use of corticosteroids. Importantly, the optimal dosing regimen and timing of initiation after the first week of life have not been determined.

Systemic hydrocortisone and inhaled budesonide have been studied as alternative anti-inflammatory therapies that may reduce lung disease and avoid the harmful long-term consequences of dexamethasone. Existing meta-analyses of RCTs evaluating these drugs failed to show clear benefit. More recently, 2 large multicenter RCTs reported reductions in BPD risk with these drugs. However, for inhaled budesonide, the risk for the composite outcome of death or BPD was of borderline statistical significance, owing to a trend toward higher rates of death in the steroid-treated infants. Long-term follow-up data and incorporation of the recent trial results into existing meta-analyses are necessary to fully assess the safety and efficacy of these 2 therapies.

Vitamin A

Vitamin A is involved in a variety of processes in the human body, including growth of epithelial cells in the respiratory tract. Preliminary studies suggest that preterm infants who develop BPD have lower plasma vitamin A levels than infants who do not develop BPD. A multicenter RCT published in 1999 demonstrated reduced risk for BPD among ELBW infants who were treated with intramuscular injections of vitamin A during the first 4 weeks of life. Subsequently, 2 recent large observational studies reported similar rates of BPD among preterm infants treated with vitamin A and untreated controls. Although these observational results provide lower quality data than the RCT, they do raise questions about the efficacy of vitamin A to prevent BPD in the current era. Ongoing shortages of this drug limit its current clinical use. A planned international RCT of oral vitamin A supplementation in extremely preterm infants may provide important, contemporary data.

MANAGEMENT OF ESTABLISHED BRONCHOPULMONARY DYSPLASIA

There are limited rigorous data to guide treatment during the chronic and resolution phases of BPD. This evidence gap, in turn, has led to considerable variation in the treatment of infants with established, severe BPD.

Respiratory Management

The primary respiratory management strategy in RDS and evolving chronic lung disease is avoidance of mechanical ventilation by using noninvasive modalities (eg, nasal CPAP, high-flow nasal cannula) and permissive hypercapnia. For most infants with mild to moderate respiratory insufficiency secondary to established BPD, continued emphasis on noninvasive support with gradual weaning as permitted is appropriate. However, for some infants with severe lung disease, prolonged use of noninvasive modalities provides inadequate respiratory support. This may result in poor somatic and alveolar/vascular growth and persistent ventilation/perfusion (V/Q) mismatch, leading to the development of pulmonary hypertension. In this subset of infants, increased respiratory support and even invasive mechanical ventilation may be necessary to promote long-term recovery.

Pharmacological Therapy

Medications such as corticosteroids, bronchodilators, and diuretics are commonly used in the management of established BPD. Unfortunately, RCTs evaluating the safety and efficacy of many of these are lacking. Several of the most common classes of drugs used in established BPD are listed in Table 62-3. When selecting a medication, including its dose and duration, clinicians should carefully consider the risks and benefits and be willing to discontinue therapies when clinical benefit is not observed.

Diagnosis and Management of Pulmonary Arterial Hypertension

Pulmonary arterial hypertension (PAH) is a common cardiovascular complication in infants with BPD. Although the true incidence of PAH in BPD is unknown, several small retrospective reports estimate the rates to be between 25% and 45%, with a higher incidence among those with more severe lung disease. In a single center series, only 2% of infants with mild BPD had evidence of PAH on echocardiogram compared to 36% with moderate BPD and 50% with severe BPD. Infants diagnosed with BPD-related PAH have a higher risk of mortality than do those with BPD alone, however most survivors with BPD and PAH demonstrate improvement in PAH after the first 1 to 2 years of life.

There are no well-established, evidence-based guidelines to inform the appropriate diagnostic approach and management of PAH in BPD. At present, screening echocardiography performed at 36 weeks PMA in infants with severe BPD should be considered. A peak tricuspid valve regurgitant jet velocity > 2.8 m/s observed on Doppler echocardiography is diagnostic of PAH in BPD. When this measurement cannot be obtained, right atrial enlargement, interventricular septal flattening, right ventricular hypertrophy or dilation, and a short right ventricular ejection time are commonly used. Cardiac catheterization is used infrequently in BPD but is the gold standard to confirm the diagnosis of PAH and accurately assess its severity. Serum brain-type natriuretic peptide is also used to assess for PAH, but its validity in BPD needs confirmation.

The most common pharmacological therapies used in BPD related PAH are pulmonary vasodilators such as inhaled nitric oxide and systemic sildenafil. At present, only minimal published data support their efficacy in BPD, although survey data suggests that their use is common. As pulmonary vascular resistance can depend heavily on the degree of V/Q matching in the lung, ensuring adequate lung aeration is essential to the management of BPD-associated PAH. Reducing V/Q mismatch through effective ventilation may ameliorate the need for pharmacological PAH therapy.

Long-Term Mechanical Ventilation

Approximately 1% to 3% of extremely preterm infants who develop BPD ultimately require a tracheostomy to manage severe, persistent lung disease and/or associated airway abnormalities such as subglottic stenosis and tracheobronchomalacia. The appropriate timing of tracheostomy in this population remains controversial. However, among those infants in whom long-term invasive ventilation is highly likely, tracheostomy placement prior to approximately 4 months of life may improve neurodevelopment by facilitating earlier initiation of developmentally appropriate care. Following discharge from hospital, tracheostomy-dependent infants should ideally be followed-up within a multidisciplinary home ventilator program that includes otolaryngology, pediatric pulmonology, speech and nutrition therapy, as well as close neurodevelopmental monitoring.

LONG-TERM OUTCOMES FOR INFANTS WITH BRONCHOPULMONARY DYSPLASIA

Broncopulmonary dysplasia is a strong predictor of several adverse health outcomes, including chronic cardiopulmonary impairment, growth failure, hearing and vision deficits, rehospitalization for respiratory illness in the first years of life, neurodevelopmental delay, and later mortality. Although most long-term follow-up data were obtained in patients from the presurfactant era, accumulating data indicate that new BPD also predisposes survivors to chronic morbidities that persist into adulthood.

Respiratory Outcomes

Preterm infants diagnosed with BPD require hospital readmission for respiratory reasons during early childhood at approximately twice the rate of similar premature infants without BPD. Although rehospitalization rates decline after the second and third year of life, survivors remain at increased risk for respiratory-related hospitalizations well into adulthood. Lung function abnormalities and persistent respiratory symptoms, including exercise intolerance, dyspnea, cough, and wheeze, are also more common among children and adult survivors diagnosed with BPD during infancy.

Neurodevelopmental Outcomes

A diagnosis of BPD is a strong predictor of future neurodevelopmental impairment (NDI) including deficits in psychomotor function, language, cognitive, behavior, vision, and hearing. In a Swiss national cohort of ELBW infants, a history of BPD increased the risk-adjusted odds of death or severe NDI at 2 years of life nearly 3-fold. The NRN found rates of NDI to be increased with greater severity of BPD, from 28.1% in ELBW infants without BPD to 61.9% in infants with severe BPD. Protracted mechanical ventilation beyond 36 weeks PMA is associated with an additional stepwise increase in the risk for both NDI and death among ELBW infants.

SUMMARY

BPD is a multifactorial disease that results from a complex interaction between the developmentally immature lung and multiple perinatal and postnatal exposures. Since the first descriptions of BPD over 40 years ago, the epidemiology and pathophysiology of the disease evolved. New BPD primarily affects infants born less than 1500 g and 32 weeks gestation. Survivors with BPD are at increased risk for multiple chronic sequelae, including high rates of rehospitalization, prolonged deficits in pulmonary function, neurodevelopmental impairment, and postneonatal mortality. Emerging evidence indicates that many of these complications persist into adulthood. Unfortunately, BPD rates among infants born near the current limits of viability have not improved over past 2 to 3 decades. This demands new paradigms in BPD prevention and high quality, multicenter research to enable evidence-based strategies for the management of established BPD.

SUGGESTED READINGS

INTRODUCTION

Several congenital anomalies presenting in the newborn period require immediate diagnosis and treatment. Certain conditions constitute medical and surgical emergencies in the delivery room and are best handled by anticipation and a multidisciplinary team approach to improve chances of successful outcomes. Basic principles of resuscitation should be applied in all cases by a team skilled in airway management and circulatory support. Recent advances in prenatal diagnostic technologies facilitate prenatal counseling as well as delivery room strategies and coordination of neonatal care.

RESPIRATORY EMERGENCIES

The transition from fetus to newborn is the most complex physiologic adaptation that occurs in human life. Inadequate transition from fetal to neonatal life is one of the most common causes of cardiopulmonary insufficiency in the newborn.

Airway Anomalies

Normal airway anatomy and patency are essential for a successful transition to extrauterine life. Bilateral choanal atresia is a life threatening condition. Since newborns are obligatory nose breathers, symptoms appear immediately after birth. It is often associated with other developmental anomalies such as CHARGE (coloboma, heart defects, choanal atresia, retarded growth and development, genito-urinary abnormalities, and ear anomalies), Treacher Collins, and Tessier syndromes.

Micrognathia is usually associated with conditions such as Pierre Robin sequence. At birth, this can be severe enough to cause posterior displacement of the tongue with airway obstruction, and can be an airway emergency. In many cases, prone positioning and placement of an oral airway may relieve symptoms until surgical correction is possible, but in the extreme cases, even in the most experienced hands, intubation can be very challenging. Mandibular distraction osteogenesis has been increasingly used for neonates with this congenital anomaly.

Prenatal diagnosis has transformed the outcome of fetuses with airway obstruction. Thorough evaluation of prenatal imaging allows for categorizing fetuses with airway compromise into those who will require a special mode of delivery and those who can be delivered without any special resources. The ex utero intrapartum treatment (EXIT) approach allows accessing the airway while the fetus is still attached to placental support. This permits the fetal airway to be secured in an elective, controlled manner. Once adequate ventilation has been established, the fetus is taken off placental support. The indications for an EXIT procedure have grown over time and now include a myriad of conditions, including cervical lymphangiomas, cervical teratomas, large congenital lung malformations, and congenital high airway obstruction syndrome (CHAOS).

Acute Hypoxemic Respiratory Failure

Neonatal hypoxemic respiratory failure (HRF) can be due to a variety of etiologies, including prematurity, perinatal insults or disorders of transition, and congenital lung abnormalities. Persistent pulmonary hypertension (PPHN) of the newborn refers to a failure of the normal fall in pulmonary vascular resistance that normally occurs during the first hours and days after birth. While PPHN can occasionally be an isolated problem with known obvious precipitant, it is more commonly associated with other causes of neonatal HRF.

Respiratory Distress Syndrome

Respiratory distress syndrome (RDS) is the leading cause of respiratory insufficiency and respiratory failure in preterm infants, and its incidence increases with decreasing gestational age. It is caused by lung immaturity and surfactant deficiency. Clinical symptoms are grunting and retractions as well as progressive hypoxemia and hypercarbia. Severe cases require positive pressure ventilation and surfactant replacement therapy. In term infants, RDS most commonly presents as a result of delayed clearance of alveolar lung fluid and is most often seen with cesarean or precipitous deliveries. Other rare causes of neonatal RDS include congenital surfactant protein deficiencies.

Congenital Lung Abnormalities

Large congenital lung lesions including congenital cystic adenomatoid malformation (CCAM), pulmonary sequestrations, bronchial atresias, foregut cysts, and congenital lobar emphysemas, may cause significant respiratory compromise by early compression of the tracheobronchial tree or by progressive air trapping, which may lead to a prompt surgical resection. They can also present with symptoms of pulmonary hypoplasia and pulmonary arterial hypertension.

Congenital diaphragmatic hernia (CDH) is a rare disorder in which a posterolateral diaphragmatic defect results in herniation of the abdominal contents into the chest and compression of the intrathoracic structures. Hypoplasia of the ipsilateral and contralateral lungs, severe PPHN, and left ventricular hypoplasia and dysfunction are present at birth and contribute to increased mortality. Prenatal imaging and several individual prenatal prognostic factors have been identified and can predict the clinical course after birth as well as long-term outcomes. Fetal interventions have become available but yet remain in an experimental phase. Optimal delivery room strategy and team assembly are essential and ideally, delivery at, or otherwise immediate transfer to a level III neonatal intensive care unit is warranted. After delivery, infants should be intubated with simultaneous placement of a suction catheter for gastric decompression. Early management focuses on treatment of HRF and PPHN. Surgical repair should ideally take place following a period of stabilization.

Perinatal Asphyxia

Perinatal asphyxia is another well-known predisposing factor for PPHN. It can interfere with the circulatory transition during birth, impeding the decrease in PVR and increasing the risk for PPHN. There are various causes of respiratory failure in asphyxia, including fetal hypoxemia, ischemia, meconium aspiration, ventricular dysfunction, and acidosis.

Air Leak

Air leaks such as pneumothorax and pneumomediastinum are collections of air that are located within the visceral and parietal pleura or in the mediastinal space. They can occur spontaneously, with an incidence of 1% in term infants after delivery, or they can be secondary to structural lung disease or meconium aspiration, and most often occur in patients receiving positive pressure ventilation. The clinical presentation depends upon the volume of air in the pleural space and/or mediastinum. Small pneumothoraces may be asymptomatic, whereas large air leaks may present with symptoms of tension pneumothorax with severe work of breathing, ipsilateral lung atelectasis, mediastinal shift, and cardiovascular instability due to impaired venous return. In this case, immediate intervention by needle aspiration/decompression followed by placement of a chest tube are required.

The Role of Extracorporeal Membrane Oxygenation

Extracorporeal membrane oxygenation (ECMO) has been increasingly used to support infants with severe reversible respiratory and/or cardiac failure unresponsive to conventional medical interventions. Common diagnoses requiring ECMO in the neonatal population include meconium aspiration syndrome, CDH, sepsis, PPHN, RDS, and congenital lung abnormalities. Neonatal trials have demonstrated that ECMO is an effective treatment with acceptable cognitive and functional outcomes.

CARDIOVASCULAR EMERGENCIES

Congenital Heart Disease

Congenital heart disease is a common congenital disorder in newborns with a birth prevalence of approximately 1%. Delay in diagnosis may lead to devastating consequences. As pediatric cardiology has expanded greatly into fetal echocardiography, new surgeries and procedures have become available that carry significant improvement in long-term outcomes.

Cyanotic Congenital Heart Disease

Cyanosis requires immediate attention and evaluation, and it is vitally important to distinguish between respiratory and cardiac causes as a matter of priority. Congenital heart diseases presenting with cyanosis in the newborn period include transposition of the great arteries (TGA), tricuspid atresia, tetralogy of Fallot, pulmonary atresia, critical pulmonary stenosis, and total anomalous pulmonary venous return (TAPVR). Newborns with TGA may present with inadequate mixing through the foramen ovale and an intact ventricular septum. This may require immediate intervention by transcatheter balloon atrial septostomy to improve mixing and oxygen delivery after birth. Obstructed TAPVR usually presents soon after birth with hypoxemic respiratory failure and radiographically commonly resembles RDS.

Early postnatal management of the cyanotic infant is greatly facilitated by a prenatal diagnosis with fetal echocardiography that helps to establish an appropriate management strategy (including optimal place for delivery). If a prenatal diagnosis has been made, then prenatal counseling as well as a coordinated multidisciplinary approach in the delivery room should already be in place. This would include, when indicated, immediate initiation of a prostaglandin infusion, and transfer to a cardiac unit for ongoing evaluation, including confirmatory postnatal echocardiography.

In the cyanosed infant without a prenatal diagnosis, if there is no access to echocardiography, a hyperoxia test may help. This determines whether the patient’s cyanosis is due to lung disease or a fixed cardiopulmonary shunt that allows deoxygenated blood from the right atrium bypass the pulmonary circulation. A hyperoxia test is performed by measuring the arterial partial pressure of oxygen in the neonate breathing room air and then breathing 100% oxygen, each for 10 minutes. If the cause of the cyanosis is due to a cardiac defect, the arterial partial pressure of oxygen usually remains below 100 mm Hg. It is important to note that while it may help in the initial work-up, hyperoxia test does not completely rule out other causes for cyanosis, in particular PPHN or severe lung disease. Infants with suspected (but not yet confirmed) cardiac disease should be started on a prostaglandin E1 infusion to maintain ductal patency until additional diagnostic studies can be performed.

Acyanotic Congenital Heart Disease

Critical acyanotic heart defects often present with symptoms of congestive heart failure or circulatory shock, within hours or days after birth. The timing of the onset of symptoms is often related to closure of the ductus arteriosus, in those infants whose systemic perfusion is dependent upon this. While many patients present very acutely within minutes or hours (eg, some newborns with critical aortic stenosis, coarctation of the aorta, or hypoplastic left heart syndrome), there may be a more gradual clinical decompensation when compared with cyanotic heart defects. Again, if a diagnosis of critical acyanotic congenital heart disease is suspected, then the immediate institution of prostaglandin infusion is recommended, until a confirmatory echocardiogram has been performed.

Neonatal Shock

Shock or circulatory failure is defined as a physiologic state characterized by tissue hypoxia due to reduced oxygen delivery and/or increased oxygen consumption. It is manifested with physical findings of tissue hypoperfusion, hypotension, and metabolic acidosis. It is often reversible but it must be recognized and treated immediately to prevent progression to irreversible organ dysfunction.

Shock can be classified as hypovolemic, distributive, cardiogenic, and obstructive. Hypovolemic shock can be caused by fetomaternal hemorrhage, acute bleeding from umbilical cord rupture, or massive gastrointestinal or intracranial bleeding. Distributive shock secondary to sepsis is the most common cause of neonatal shock. Cardiogenic shock can present as a result of congenital heart diseases with obstructive left heart lesions (eg, critical aortic valve stenosis, critical coarctation, interrupted aortic arch) or as a result of cardiac arrhythmias and myocarditis. Lastly, obstructive shock occurs when extracardiac diseases lead to impaired cardiac output and is usually secondary to tension pneumothorax and cardiac tamponade. Bedside functional echocardiography or point-of-care echocardiography is a noninvasive tool that provides the neonatologist with a better understanding of the pathophysiology and allows a logical approach based on individualized hemodynamic findings. Increasingly, neonatologists perform serial echocardiographic assessments to target treatment and evaluate response.

GASTROINTESTINAL EMERGENCIES

Emesis and feeding intolerance are frequent symptoms in neonates. These must be distinguished from pathological gastrointestinal processes that may require further and detailed evaluation.

Intestinal Obstruction

Bilious emesis is synonymous with intestinal obstruction, either anatomical or functional. Any infant presenting with bilious emesis requires immediate evaluation. Malrotation with midgut volvulus is a time-sensitive neonatal emergency that requires urgent diagnosis performed by an upper gastrointestinal contrast study and prompt surgical intervention to prevent intestinal necrosis.

Congenital intestinal obstructions can affect the gastrointestinal (GI) tract from the esophagus to the rectum, and clinical symptoms are usually present within the first days of life. The classic triad of features of bowel obstruction are bilious emesis, constipation or failure to pass meconium within the first 48 hours of life, and abdominal distention. The initial management includes intravenous access to provide fluid resuscitation and the passage of a suitably sized nasogastric tube to achieve gastric decompression by continuous aspiration. Diagnosis is confirmed by x-ray and contrast studies, and surgical correction is indicated.

Hirschsprung Disease

Hirschsprung disease is a motor disorder of the bowel that is caused by failure of enteric ganglion cell precursors to migrate during intestinal development. The resulting aganglionic segment of the colon fails to relax, causing a functional obstruction that presents clinically in the neonatal period with abdominal distention and failure to pass meconium. The diagnosis may be suspected clinically by obtaining intestinal decompression as a result of rectal examination and confirmed by rectal biopsy.

Necrotizing Enterocolitis

Necrotizing enterocolitis (NEC) is the most common GI emergency in preterm infants that presents in the first 3 to 4 weeks of life. It is a multifactorial inflammatory disease of the bowel that leads to intestinal necrosis and perforation. Its incidence decreases with increasing gestational age. Although early recognition and aggressive treatment have improved clinical outcomes, NEC still accounts for substantial long-term morbidities and poor neurodevelopmental outcomes in survivors. Spontaneous intestinal perforation mostly occurs in extremely low–birth-weight infants and usually presents earlier than NEC, in the first few days of life. Once pneumoperitoneum is diagnosed, a surgical intervention, either laparotomy or peritoneal drain placement, is indicated.

Omphalocele and Gastroschisis

Omphalocele and gastroschisis represent the most frequent congenital abdominal wall defects. Early detection of these malformations and associated anomalies allows multidisciplinary counseling and planning of delivery in a center equipped with high-risk pregnancy assistance, pediatric surgery, and neonatology. Early management involves immediate covering of herniated abdominal contents, nasogastric tube placement for gastric decompression, and surgical evaluation.

NEUROLOGIC EMERGENCIES

Seizures

Seizures are the most common neonatal neurological emergency and require prompt diagnosis and treatment. When present, they may reflect a potentially treatable etiology. Immediate evaluation is essential since prolonged seizures may precipitate hypoxemia, acidemia, alterations in glucose homeostasis, and hemodynamic changes in blood pressure and heart rate, all of which can contribute to further brain injury. Although most neonatal seizures are acute and reactive, some seizures occur without an identifiable cause and are termed idiopathic. Most neonatal seizures are focal in origin because of the delay in generalization of electrical activity secondary to lack of myelination and incomplete dendritic and synaptic formation in the immature brain. Neonatal seizures can be the result of a wide range of conditions: neonatal encephalopathy, intracranial hemorrhage, metabolic disturbances, central nervous system or systemic infections, inborn errors of metabolism, and structural brain lesions. Hypoxic ischemic encephalopathy secondary to perinatal asphyxia is the most common cause of neonatal seizures. Therapeutic hypothermia over a 72-hour period has been established as an effective neuroprotective strategy and has been demonstrated to significantly improve neurodevelopmental outcomes at 24 months of age. Treatment with anticonvulsants such as phenobarbital is indicated if seizure activity is suspected. Screening includes evaluation for electrolyte disturbances and glucose, cerebral spinal fluid studies, intracranial imaging, and electroencephalography.

Neural Tube Defects

Neural tube defects must be emergently addressed unless completely covered by intact skin. The defect should be covered with a sterile, saline-soaked dressing and the infant should be placed in a prone or lateral position to avoid prolonged pressure over the lesion. Prompt neurosurgical evaluation and treatment are essential to optimize neurodevelopmental prognosis. Fetal surgery for myelomeningocele has become available over the past years. Its rationale is that early closure of the defect prevents further damage of the spinal cord and leakage of spinal fluid, which can therefore prevent or reverse herniation of the hindbrain (Chiari II malformation) and hydrocephalus. The Management of Myelomeningocele Study (MOMS) resulted in a decreased need for a cerebrospinal fluid shunt at 1 year of age and improvement in a composite score for mental developmental and motor function at 30 months of age. Fetal surgery for myelomeningocele is currently performed at highly specialized centers across North America and Europe.

SUGGESTED READINGS

INTRODUCTION

Over the past half century, important advances in obstetric and neonatal intensive care have led to dramatic reductions in neonatal mortality at all gestational ages. However, prematurity remains a significant problem, and is associated with both mortality and important morbidities. According to the March of Dimes Report Card, 9.6% of children born in the United States in 2015 were preterm, or born at less than 37 weeks’ gestation. In addition, perinatal events in full-term infants can also be associated with adverse outcomes. Children who are premature or critically ill at birth can have poor long-term medical outcomes as well as poor developmental and functional outcomes. This chapter addresses how and when to assess developmental outcomes of high-risk children, reviews the associations between specific risk factors and developmental outcomes, and finally briefly addresses future directions for the long term follow-up of high risk newborns.

IDENTIFYING THE HIGH-RISK INFANT

Infants who are anticipated to be at high risk for developmental problems require close surveillance for developmental problems during the first few years of life. The American Academy of Pediatrics has defined high-risk infants as those infants who fall into 1 of 4 categories: “(1) the preterm infant; (2) the infant with special health care needs or dependence on technology; (3) the infant at risk because of family issues; and (4) the infant with anticipated early death.” Such infants should be identified at the time of hospital discharge. Depending on the complexity of the child’s medical needs, discharge planning should include a plan for ongoing management of medical problems; appointments with a primary care physician, developmental follow-up program, and any necessary medical specialists; identification of necessary home support services including home care or home nursing; and comprehensive education of the parents. Appropriate family counseling should be provided for as many caregivers as possible, and should include guidance for optimal medical and developmental care of the infant. In some situations, the follow-up program can also include care coordination and primary care in addition to addressing developmental needs. While there is evidence that such “medical homes” decrease resource utilization among children with complex medical conditions, the practicalities of setting up such programs have led to infrequent adoption of this model. Though all children with risk factors for developmental problems should be referred for routine surveillance at the time of hospital discharge, some are missed and some do not come immediately to attention. Therefore, it is essential for the general pediatrician or general practitioner to also provide basic screening in order to identify young children who may have delays and then to refer them for comprehensive evaluations.

ASSESSING DEVELOPMENTAL OUTCOMES

Developmental outcomes evolve over the first several years of life. While some developmental outcomes can be measured in young infants, many important outcomes cannot be measured until school age. In general, surveillance should include multiple domains: motor, cognitive, language, behavior, and socioemotional skills. Assessment of outcomes serves several purposes: It provides parents with information about their children’s development, alerts caregivers to the need for support services including early intervention for infants and toddlers and school services for older children, and provides outcomes for research studies that in turn inform parental counseling and future research. The assessment tools used for an individual child depend on that child’s age and history, and the goals of the evaluation.

The earliest assessment of infant development includes a comprehensive neuromotor examination, a review of feeding and sleeping skills, and an evaluation of social interactions. Feeding is 1 of the first key skills for newborns; therefore, poor feeding skills may be an early marker of risk for slow attainment of developmental milestones. Structured motor evaluations such as the General Movements Assessment may be used to assess the risk for cerebral palsy (CP) in preterm infants.

The first few years of life represent an age range when developmental outcomes are most commonly assessed, and reported, in the context of large studies. Early developmental outcomes are most commonly assessed with the Bayley Scales of Infant and Toddler Development in the United States and many other countries. However, other assessments such as the Griffiths Mental Development Scales may also be used. Importantly, such assessments measure the extent to which the child can perform the developmental tasks typical for its age, but they are not intelligence tests. The key use of assessments during the infant and toddler years is to determine whether children would benefit from specific interventions, targeted to areas of developmental delay. For instance, emerging research suggests that early, targeted, and goal-oriented therapies may improve both motor outcomes and cognitive development among children at high risk for CP.

In children older than 2 years of age, multiple assessments are available to measure full-scale intelligence quotient (IQ) or developmental progress. Examples include the Differential Ability Scales (DAS), the Wechsler Preschool and Primary Scale of Intelligence (WPPSI), and the Wechsler Intelligence Scale for Children (WISC). By school age, assessments such as the Wechsler Individual Achievement Test (WIAT) can also be used to measure academic skills. School-age assessments of IQ generally correlate well with adult intelligence. However, it is critical to note that school-age assessments do not always correlate well with earlier developmental assessments. There are several possible explanations for children’s shift in scores over time, including social class of parents and availability of educational and social supports. In addition, as described, the available assessments target different skill sets in infants and school age children. As time passes, neonatal events are less predictive of outcomes over time, and instead the child’s sociodemographic environment becomes increasingly influential.

Cerebral palsy is defined as a disorder of movement and posture that involves abnormalities in tone, reflexes, coordination, and movement; delays in motor milestone achievement; and aberration in primitive reflexes that is permanent (although the presentation may evolve over time) It is caused by a nonprogressive interference, lesion, or abnormality of the developing immature brain. CP includes both an abnormality on physical examination and a corresponding functional limitation. The most common system for classification of motor abilities in CP is the Gross Motor Function Classification System (GMFCS). The GMFCS is useful because it is valid, reliable, and stable over time, particularly in children older than 2 years of age. The GMFCS allows clinicians and parents to plan appropriate therapies and frame expectations for how motor difficulties may evolve as children grow. Though it may not be possible to categorize severity of CP with confidence before 2 years of age, most children with CP present with muscle tone or movement abnormalities at far younger ages. Therefore, the clinician must be aware of the early signs of motor disorders and should, where possible, quickly provide access to essential therapies and resources so as to optimize outcomes.

High-risk children who do not have CP remain at increased risk for developmental coordination disorder (DCD), which is a problem of motor coordination and performance that impacts activities of daily living or academic achievement. The most common assessment tool for DCD is the Movement Assessment Battery for Children, but other equivalent tests are also used. The earliest age at which DCD can be diagnosed with confidence is 5 years. DCD is common in former preterm children and is associated with multiple concomitant developmental problems including speech and language, socioemotional, and behavior problems.

Children who were born preterm or were critically ill at birth have increased risk for behavioral, socioemotional and neuropsychiatric problems. Screening for behavioral or socioemotional problems in the early childhood years can be accomplished with multiple instruments; common choices are the Brief Infant and Toddler Social and Emotional Assessment (BITSEA) or the Child Behavior Checklist (CBCL). Importantly, children with a low IQ tend to have higher rates of behavior problems, making detection and concurrent management of both problems essential. Signs of autism can be detected with the Modified Checklist for Autism in Toddlers (M-CHAT), which is a validated screening instrument for children 16 to 30 months of age. Confirmation of the diagnosis of autism requires a full assessment by a child psychologist. Similarly, high-risk children have elevated rates of attention-deficit disorder.

As children progress through school age and enter adolescence, additional important aspects of development require assessment. High-risk children may experience executive function disorders, learning problems, poor academic achievement, decreased health-related quality of life, and functional problems. Most programs for surveillance of high-risk children end between 2 and 5 years of age, and state-mandated early intervention programs transition responsibility to school districts around the same time. Therefore, the burden of identifying and addressing such complex problems falls to the primary care physician and school districts. Primary care physicians must remain vigilant so as to identify these problems in a timely manner. It is important to note that higher-level neurobehavioral and cognitive problems lead to increased resource utilization and special education costs. These problems are likely to persist, influencing functional outcomes, including educational achievement, employment, wage-earning, marriage, and family building throughout the life spectrum.

DEVELOPMENTAL OUTCOMES OF PRETERM INFANTS

Prematurity is the most common and well-recognized risk factor for poor outcomes in multiple developmental domains. The risk for most developmental outcomes is best described as a gradient, with risk increasing as gestational age decreases toward the limits of viability. In large studies, even “near term” birth at 37 to 38 weeks is associated with poor outcomes, when compared to birth at 39 to 40 weeks. Late-preterm–born children (34–36 weeks’ gestation) do not routinely qualify for most developmental surveillance programs, yet they have increased risk for poor outcomes in multiple domains. These children deserve particular focus from primary care physicians, in whose hands the responsibility for detection and referral therefore falls. Most research into the outcomes of prematurity focuses on the very preterm and extremely preterm infants, in whom risk profiles are most elevated. Below, we briefly present results of several cohort studies that evaluated outcomes of large populations of extremely preterm infants.

In recent years, several large cohort studies have reported the early childhood developmental outcomes of extremely preterm infants. These study populations were heterogeneous, but generally included infants born at less than 27 weeks’ gestational age, and were performed between 2004 and 2007. Follow-up evaluations completed between 2 and 3 years corrected age identified CP in 7% to 14% of the children in these cohorts. Severe neurodevelopmental impairment was generally defined as nonambulatory CP and/or developmental testing scores greater than 3 standard deviations below the mean. Blindness and deafness were not consistently included in the criteria for severe impairment. The overall rates of severe neurodevelopmental impairment ranged from 4% to 13%.

Premature infants who reach school age continue to experience increased risk for neurodevelopmental impairments. Nearly all those with severe impairments that were identified in early childhood continue to have moderate to severe impairments at school age. At school age, rates of CP of 9% to 10%, and cognitive disability rates of 4% to 36% have been reported in this population.

In addition to the risk of developmental delay associated with prematurity or low birth weight, additional postnatal morbidities further increase this risk. This section will discuss the impact on developmental outcomes of 6 postnatal morbidities: brain injury, bronchopulmonary dysplasia (BPD), necrotizing enterocolitis, retinopathy of prematurity, infection, and growth and nutrition. Although all of these factors increase the risk of developmental impairments, severe brain injury, severe retinopathy of prematurity, and BPD are most significantly associated with neurodevelopmental impairment, each contributing an additive increase in risk for adverse outcome. These relationships were first demonstrated at 18 months corrected age, and then validated in longer-term studies of several different cohorts.

It is also important to note that in addition to the above-mentioned biologic factors, socioeconomic factors and environment have an important impact on neurodevelopmental outcomes. Higher socioeconomic status has been suggested to be protective, improving cognitive development over the first few years of life, whereas adverse socioeconomic factors are associated with adverse developmental outcomes and behavior problems. Over time, as children get older, perinatal risk factors become less predictive and socioeconomic factors become more predictive of developmental trajectories.

Brain Injury

Premature infants are at risk for many types of brain injury, including intraventricular hemorrhage (IVH), intraparenchymal hemorrhage (IPH), and cystic periventricular leukomalacia (PVL). Historically, IPH was referred to as grade IV IVH. Grade III IVH or IPH occurs in approximately 15% of infants born at less than 28 weeks’ gestation, with PVL occurring in approximately 4%. It is well known that grade III IVH and IPH are associated with adverse developmental outcomes, whereas the impact of grade I and II IVH on developmental outcomes is still unclear, with some studies showing no impact and others demonstrating a small increase in risk for developmental delay. Presence of cystic PVL is associated with increased risk of CP. Infants with IVH or IPH, especially those with higher grade IVH, are at risk for posthemorrhagic hydrocephalus (PHH). Infants who experience PHH are at increased risk for developmental delays, with those who require ventriculoperitoneal shunt placement due to hydrocephalus at even further increased risk.

Cranial ultrasound (CUS) is the standard modality to evaluate for brain injury during the neonatal period. While early ultrasounds are helpful for acute decision-making, a near-term or 36-week corrected age CUS is more predictive of 2-year neurodevelopmental impairment or severe motor impairment. While many studies have explored the role of magnetic resonance imaging (MRI) for prediction of developmental outcomes, the diagnostic utility of this tool remains limited. Therefore, current guidelines do not recommend near-term MRI for routine preterm imaging due to insufficient evidence that the practice improves long-term outcomes. It is important to note that although IVH, IPH, and PVL are associated with increased risk of early developmental delays and CP, these findings are not as clearly diagnostic of later delays. Furthermore, a normal CUS during the neonatal period does not exclude future developmental delays, with 30% to 40% of those with “normal” CUS experiencing neurodevelopmental challenges at 18 to 30 months of age. Therefore, care must be taken when counseling parents on the impact of CUS findings in preterm infants, incorporating discussion of the predictive validity of CUS and potential impact of other known risk factors on future developmental outcomes.

Bronchopulmonary Dysplasia

Bronchopulmonary dysplasia is the most common morbidity of prematurity, occurring in approximately half of all infants born less than 28 weeks’ gestation. The most universally accepted definition of BPD requires oxygen dependence at 36 weeks’ postmenstrual age. BPD is an independent predictor of adverse development at 2 and 5 years of age. Cognitive delays associated with BPD have been shown to present as late as 8 years of age.

Few therapies have been shown to both decrease the incidence of BPD and improve developmental outcomes. Vitamin A may reduce the composite outcome of BPD or death but has no effect on developmental outcomes. Postnatal steroids decrease risk for BPD but increase risk for CP if given within 4 days of life. Steroids given after 1 week of life have not been shown to increase the risk of CP. Surfactant improves the symptoms of respiratory distress but has not been associated with decreased rates of BPD. In addition, although surfactant may reduce mild disability at 1 year of age, this result does not seem to be sustained at 2 years of age. Antenatal corticosteroids reduce developmental impairment, but it is unclear if they decrease rates of BPD. Caffeine is the single treatment that has been shown to safely reduce BPD, death, or neurodevelopmental impairment at 18 to 21 months of age, and motor impairment until at least 5 years of age.

Retinopathy of Prematurity

Retinopathy of prematurity (ROP) is a common morbidity of prematurity that results from abnormal retinal vascular growth after premature birth. Retinopathy of prematurity occurs in over half of infants born less than 28 weeks’ gestation, with approximately 15% having more severe disease. Historically, blindness secondary to retinal detachment was the most feared complication of ROP. With improvements in surveillance and available treatments including cryotherapy, laser therapy, and anti–vascular endothelial growth factor treatment, blindness is now a rare outcome in developed countries. However, in the developing world, blindness is still a very real complication of untreated ROP.

Whether or not it requires treatment, ROP is associated with visual morbidities, including blindness, strabismus, and refractory errors. In addition, ROP is an independent risk factor for poor neurodevelopmental outcome. Retinopathy of prematurity has been associated with a strategy of targeting higher oxygen saturations for premature infants; however, this must be cautiously interpreted in the setting of increased risk of death associated with low oxygen saturation targeting strategies.

Necrotizing Enterocolitis

Necrotizing enterocolitis (NEC) occurs in 7% to 13% of infants born less than 28 weeks’ gestational age. Although the precise etiology of NEC is unknown, it is thought to be an inflammatory process leading to necrosis of the bowel. Depending on the severity of disease, NEC can be treated either medically with antibiotics and bowel rest or surgically with drain placement or laparotomy.

Infants who experience NEC have increased rates of death, cognitive delay, CP, and developmental impairment at 18 months of age. In an observational study, surgical NEC, but not medical NEC, was most strongly associated with this increased risk of impairment. Both antenatal corticosteroid treatment and exclusive maternal milk nutrition have been associated with decreased risk of NEC in preterm infants.

Infection

Several factors, including immune system impairment, immaturity of the skin barriers, and the frequent need for invasive ventilations and lines, render premature infants at high risk for infection. Despite reduced rates of infection over the past few decades, with quality improvement efforts to reduce iatrogenic infections, approximately 25% of infants born less than 28 weeks’ gestation still develop culture-positive sepsis during the neonatal hospitalization. Infection is associated with increased risk of cognitive delay, CP, and visual impairment at 18 to 22 months of age. Meningitis and fungal infections pose even higher risks of adverse developmental outcomes.

Growth and Nutrition

The importance of adequate growth and nutrition in premature infants has been increasingly recognized in recent years. This has led to early and more aggressive administration of total parenteral nutrition, early introduction of trophic enteral feedings, and more rapid increases in feeding volumes and fortification of feedings. Low growth velocity during the neonatal hospitalization is associated with increased risk for CP and developmental delays. Furthermore, slow postnatal head growth is highly predictive of later developmental disabilities during childhood. Despite recent interventions to improve nutrition and growth, evidence is still lacking to show that these interventions have a positive impact on developmental outcomes.

OTHER GROUPS AT HIGH RISK FOR ADVERSE DEVELOPMENTAL OUTCOMES

Prenatal conditions and perinatal events can also increase risk for adverse developmental outcomes among infants born at or near term. Notable risk factors include intrauterine growth restriction (IUGR), congenital heart disease (CHD), hypoxic ischemic encephalopathy (HIE), and the need for extracorporeal membrane oxygenation (ECMO).

Intrauterine Growth Restriction

Intrauterine growth restriction refers to suboptimal fetal growth, with a lower-than-expected growth velocity in utero. Although placental insufficiency is a major contributor to IUGR, other factors, such as congenital infections, chromosomal abnormalities, and teratogen exposures, also can contribute to this. It has been associated with poor neurodevelopmental outcomes, including motor, cognitive, and language delays. While low birth weight and length are important, slow head growth seems to be the strongest predictor of poor outcomes.

Congenital Heart Disease

Advances in the surgical and medical management of children born with congenital heart disease have resulted in increased survival rates, creating an interest in the long-term outcomes of these children as they grow. Children with CHD are at increased risk for both motor and cognitive delays. On average, children with a history of CHD have developmental scores approximately 5 to 10 points lower than standardized norms.

Deficits in children with CHD do not appear to be attributable solely to intraoperative and postoperative events. Infants who are born with CHD often have brains that are more immature and smaller than expected for their gestational age. Brain MRI in this population has shown that brain injury, mainly white matter injury, is present both pre- and postoperatively. This suggests a pre-existing vulnerability in these infants, although the degree of white matter injury is yet to be correlated with developmental outcomes in this population.

Hypoxic-Ischemic Encephalopathy

Hypoxic-ischemic encephalopathy (HIE) results from a perinatal event that prevents an adequate blood or oxygen supply to the infant brain. The diagnosis of HIE generally requires history of an acute perinatal event (eg, cord prolapse or placental abruption), a standardized neurologic exam consistent with encephalopathy, low Apgar scores, and evidence of acidosis on umbilical cord or infant blood gas soon after birth. Infants with HIE have a significant risk of mortality and morbidity.

Numerous studies have shown a decreased risk of both mortality and survival with severe neurodevelopmental disability up to school age in children with moderate or severe HIE who undergo therapeutic hypothermia treatment, or “cooling.” Yet, despite treatment with therapeutic hypothermia, these children are still at a high risk for significant impairments. Of those infants who receive therapeutic hypothermia, only about half survive without severe impairments, with approximately 20% of survivors being diagnosed with CP and 22% with developmental delay.

Early factors associated with more severe disability include a higher degree of encephalopathy on initial exam, requiring chest compressions for more than 1 minute, onset of breathing greater than 30 minutes after birth, or a base deficit of 16 or greater. An improving clinical examination during hypothermia with a normal discharge examination, as well as a normal brain MRI after hypothermia are associated with decreased risk for death or disability at 18 to 24 months of age. However, many children with a normal MRI still develop significant impairments, making follow-up of these children as they grow a necessity.

Extracorporeal Membrane Oxygenation

Extracorporeal membrane oxygenation (ECMO) is a lifesaving therapy that provides complete respiratory support with or without cardiac support for patients with severe respiratory and/or cardiac failure that is unresponsive to conventional medical treatments. It requires 1 or 2 large, surgically placed central catheters. The ECMO circuit, which includes a pump and oxygenator, then removes the patient’s blood, artificially removes carbon dioxide and oxygenates the blood, and then returns the blood to the patient. Common conditions for which ECMO is used in neonates include meconium aspiration syndrome and congenital diaphragmatic hernia.

The risk of morbidity and mortality in infants requiring ECMO depends on a number of factors including comorbidities, the indication for ECMO, and complications experienced while on ECMO. Younger gestational age and lower birth weight are both associated with increased complications. (See also Chapter 108.)

THE FUTURE OF NEONATAL DEVELOPMENTAL CARE

Despite the many well-recognized risk factors for developmental delay or impairment, no factor or combination of factors predicts with certainty the outcome of a vulnerable infant. Biologic, genetic, perinatal, postnatal, and sociodemographic factors all play critical roles in influencing a child’s development. It is likely that early sensitive periods of heightened neuronal plasticity may provide opportunities to influence developmental trajectories. For example, as discussed above, a few early pilot trials have demonstrated that focused interventions, when delivered at the right time and with the right intensity, can improve motor outcomes of preterm-born infants at high risk for CP.

In future research, it will be essential to develop strategies to improve the neonatologist’s and the pediatrician’s ability to diagnose problems early, or to predict which children are at risk, and which specific outcomes (eg, cognitive, motor, behavioral, learning, coordination) are most likely to be affected. In-hand, focused interventions that target these risk profiles must be developed and tested. This would enable intervention before any evidence of pathology emerges. Often, children are at risk for several different types of poor outcomes; for example, children with cognitive problems are also at increased risk for behavior problems. At present, there are few proven interventions that target such mixed phenotypes.

The effects of prematurity and other high-risk conditions can influence not only childhood but also adult outcomes. Extremely preterm–born adults have higher rates of neuropsychiatric disorders, autism, and CP, and have less good functional outcomes including lower wage earning, lower rates of parenthood, lower marriage rates, and lower educational achievement. Strategies to mitigate these risks should begin during the postnatal period and continue at least through early childhood and school age. Clinicians should be acutely aware of their patients’ risk profiles, and should pay attention to early signs of evolving problems, and quickly connect the patients with appropriate services and therapies, so as to optimize the developmental outcomes of these high-risk patients.

SUGGESTED READINGS