Prolonged ventilation and postnatal growth of preterm infants

Introduction

Extremely premature infants often need invasive respiratory support from birth, but have low nutritional reserves and high metabolic demands that can make optimising their nutritional status challenging. Under-malnutrition can detrimentally affect outcomes. A study of newborn rats demonstrated that body growth, lung growth and lung DNA levels were significantly reduced by both undernutrition and hyperoxia [1]. Malnutrition can delay the development of new alveoli and also have a detrimental effect on diaphragmatic and intercostal muscle strength, hence, in animal studies, prolonging the need for mechanical ventilation [2]. Poor nutritional status is also implicated in bronchopulmonary dysplasia (BPD) development [3] and can lead to loss of neuronal cells [4], and suboptimal nutrition during the sensitive stages in early brain development may have long-term effects on cognitive function [5]. Poor head growth during neonatal admission in preterm infants has been strongly associated with adverse neurodevelopmental outcomes and correlated with subsequent poor cognition, particularly if the inadequate growth persists post discharge [6]. A higher rate of head circumference growth from birth to discharge has been associated with a lower incidence of cerebral palsy and neurodevelopmental impairment [7]. Furthermore, head circumference growth from birth to discharge has been shown to be a predictor of neurodevelopmental outcome and gross motor development at 5 years of age [8].

It is, therefore, important to determine which modifiable factors in extremely preterm infants affect body weight and head circumference growth. Very low birth weight infants with severe BPD have been demonstrated to have insufficient weight gain. Infants with BPD are frequently tachypnoeic due to a reduced lung compliance [9], which would increase their energy expenditure [10]. The binary outcome of BPD, however, often fails to capture the whole spectrum of on-going respiratory morbidity. Indeed, preterm infants can suffer chronic respiratory morbidity independent of a diagnosis of BPD [11]. Although clinicians anecdotally appreciate that prolonged ventilation is associated with impaired growth, this relationship has rarely been studied. In one study, there was a significant correlation between the duration of mechanical ventilation, poor head growth and adverse neurodevelopmental outcomes that persisted at 2 years of age [6].

We hypothesised that quantified growth indices, defined as the difference in both weight and head circumference z-scores from birth to discharge, would be negatively associated with respiratory disease severity as assessed by the duration of invasive ventilation. The aim of this study was to test that hypothesis as such data would further encourage practitioners to develop efficacious management strategies to achieve successful early extubation.

Materials and methods

The records of ventilated infants born at less than 28 weeks of gestation between 1/1/2012 and 1/12/2016 and solely cared for at a tertiary neonatal intensive care unit were reviewed. Infants who were ventilated for less than 7 days and those who died before discharge home were excluded from the analysis. We excluded infants who were ventilated for less than 7 days in order to assess postnatal growth in the most at-risk group of infants, as a requirement for mechanical ventilation after 1 week of age has been shown to be a predictor of the development and severity of BPD [12]. The study was registered as a service evaluation with the Clinical Governance Department. The Health Research Authority Toolkit of the National Health System, United Kingdom confirmed that the study would not be considered as research and would not need regulatory approval by a Research Ethics Committee.

According to the unit’s routine policies, preterm infants born at less than 28 weeks of gestation with respiratory distress, who failed to stabilise with continuous positive airway pressure via facemask, were intubated and given surfactant in the delivery suite [13]. Further doses of surfactant were given on the neonatal unit according to clinical need, i.e. a fraction of inspired oxygen concentration (FiO₂) >0.3 despite adequate ventilatory pressures. Infants in the delivery suite were started on a peak inflating pressure (PIP) of 20–25 cm H₂O and an FiO₂ of 0.21–0.3 [14], which were altered according to chest wall rise and to keep oxygen saturation levels between 90% and 94%. Infants subsequently received volume-targeted or pressure-controlled time-cycled ventilation using either the SLE5000 or SLE6000 neonatal ventilators (SLE, Croydon, UK). High-frequency oscillation (HFOV) was considered on the neonatal unit in infants with homogeneous lung disease and severe respiratory distress. Extubation on the neonatal unit was considered if the FiO₂ was less than 0.4 and the infant had acceptable blood gases (pH >7.25 and PaCO₂ <8.5 kPa) (conversion factor to mm Hg is 7.5) and an adequate respiratory drive with a breathing rate above the set ventilator rate. All infants received a loading dose and subsequent maintenance doses of caffeine citrate. As per Trust guidelines, infants born less than 34 weeks of gestation were prescribed a loading dose (20 mg/kg) of intravenous caffeine citrate before day 3 after birth. Twenty-four hours after the loading dose, a daily maintenance dose (5 mg/kg) was administered until 34 weeks of postmenstrual age.

Parenteral nutrition was prescribed within the first 24 h post-delivery as per the unit’s guidelines. Infants were initially prescribed bags of standard parenteral nutrition, which contained both aqueous parenteral nutrition and lipid with the addition of electrolytes and trace elements as determined by laboratory results. Individualised bags of parenteral nutrition were prescribed as needed. Nutritional decisions were made daily during consultant-led ward rounds and reviewed weekly during the nutrition ward round (comprising a consultant neonatologist with a specialist interest in nutrition, a gastroenterologist, a senior dietician and a senior pharmacist), with regular monitoring of the infant’s biochemical and anthropometric status. Generally, an infant born at less than 28 weeks of gestation was commenced on parenteral nutrition on day 1 after birth and received an average daily energy intake of between 110 and 135 kcal/kg as per the European Society of Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) guidelines [15]. Infants were commenced on two hourly trophic feeds on day 1 at 10–20 mL/kg/day with advancement if tolerated by 10 mL/kg/day twice daily every 24 h as one to two hourly feeds until a total of 180 mL/kg/day enteral feeds were reached. Our practice was to achieve an intake of at least 130 kcal/kg/day with either fortified breast milk or preterm formula in keeping with current consensus guidelines for protein, carbohydrate and fat intake. The choice of milk feeds for preterm infants was maternal expressed breast milk if available or, if not available, donor expressed breast milk for infants at risk of necrotising enterocolitis (NEC). The decision to continue donor expressed breast milk beyond 7–10 days was made on a case-by-case basis. Infants at low risk of NEC were gradually changed on to a preterm formula. For very low birth weight infants solely fed on maternal expressed breast milk, once 180–200 mL/kg/day had been reached the addition of breast milk fortifier was routinely added (if tolerated).

Data on administration of antenatal corticosteroids, gender, gestational age at birth, birth weight, duration of invasive mechanical ventilation, days of non-invasive ventilation post extubation, development of BPD and weight and age at discharge were collected. BPD was defined as the need for supplemental oxygen therapy for at least 28 days [16]. Data were also collected on postnatal corticosteroid administration. Dexamethasone was given if infants could not be weaned from mechanical ventilation [17]. A 9-day course was started at a dose of 0.25 mg/kg twice daily for the first 3 days; the dose was then weaned to 0.15 mg/kg twice daily for the next 3 days and then further weaned to 0.05 mg/kg twice daily for the last 3 days. If there was no response to postnatal steroid treatment after the first 3 days, the course was discontinued.

Results

Two hundred and fourteen ventilated, preterm infants born at less than 28 weeks of gestation were cared for within the study period; 117 infants were transferred to another neonatal unit prior to discharge home, 17 infants were ventilated for less than 7 days and 25 infants died on the neonatal unit, these 159 infants were excluded from analysis.

All 55 infants (28 male) who fulfilled the eligibility for the study were included in the analysis; they had a median gestational age at birth of 25.3 (range 24.3–26.7) weeks (Table 1). Twenty-nine infants (52.7%) received a full course of antenatal steroids; only six (10.9%) received no antenatal steroids. All infants included in the study developed BPD, with 31 (56.4%) being discharged on supplementary oxygen. Their median duration of mechanical ventilation was 45 (33–68) days and median duration of subsequent non-invasive ventilation was 39 (28–65) days. Twenty-eight infants (50.9%) received at least one course of postnatal steroids with 10 receiving more than one course administered at a median postmenstrual age of 30.7 [interquartile range (IQR) 29.7–33.3] weeks. Infants receiving postnatal steroids were ventilated for significantly longer than those infants who did not receive postnatal steroids (P=0.017). The median birth weight z-score of the 55 infants was −0.57 (−1.24 to −0.36) and head circumference z-score was −0.57 (−1.41–0.20). The median (IQR) ΔWz was −1.21 (−1.79 to −0.39) and ΔHz was −0.72 (−1.42–0.61). Both ΔWz and ΔHz (Figure 1) were significantly negatively related to the number of ventilation days (r=−0.345, P=0.01; r=−0.508, P=0.03, respectively). Univariate analysis showed that there were no statistically significant results between those who received postnatal steroids and those who did not in relation to both ΔWz and ΔHz (P=0.417, P=0.158). There were no significant correlations between days of non-invasive ventilation and either ΔWz (r=0.156, P=0.256) or ΔHz (r=0.069, P=0.708). Combining total days of invasive mechanical ventilation with total days of non-invasive ventilation, there was still no significant correlation with ΔWz (r=−1.11, P=0.421) or ΔHz (r=−0.238, P=0.190).

Figure 1:

The relationship between head circumference z-score from birth to discharge and days of invasive mechanical ventilation [y=0.69–0.03x (P=0.03)].

Discussion

We have demonstrated that the growth of extremely premature infants was significantly negatively associated with the number of days of invasive ventilation. Mechanical ventilation in preterm infants can be a stimulus for systemic inflammation, and a duration of invasive mechanical ventilation greater than 7 days in newborns has been positively associated with a larger postnatal systemic inflammatory response, which can subsequently lead to adverse pulmonary and neurodevelopmental outcomes [19]. Long periods of ventilation can increase circulating pro-inflammatory cytokines [19], and those cytokines have inhibitory effects on the growth hormone axis [20]. Prolonged mechanical ventilation in animal studies has resulted in a significant impairment upon diaphragmatic muscle function [21] which would increase energy requirements. Studies in adult patients exhibiting under-malnutrition have shown a reduction in neural respiratory drive and a decrease in diaphragmatic muscle mass [22]. In another study of ventilated adults, the risk of remaining ventilated for at least 3 weeks was significantly greater in those patients who exhibited atrophy of the diaphragm on imaging done within the first few days following intubation [23].

It must be considered that earlier use of non-invasive ventilation might have resulted in better postnatal growth. We, however, used a standardised protocol with regard to when to intubate and use invasive mechanical ventilation and when to extubate infants, which are in line with current protocols [13]. The severity of respiratory illness and hence a longer duration of mechanical ventilation may cause difficulties in establishing oral feeding regimes which may contribute to the postnatal decline in growth velocity [24]. Nevertheless, we optimised nutritional management for preterm infants according to current consensus guidance [15], [25].

Postnatal steroids are frequently administered to infants with respiratory morbidity to enhance successful extubation from invasive mechanical ventilation.

Dexamethasone is known to alter weight gain composition by decreasing the accretion of protein and increasing the laying down of fat [26]. A previous study found a short-term negative effect between postnatal steroid administration and poor growth (both weight and head circumference), but with no statistically significant long-term effects [27]. Inhaled steroids, such as budesonide, have been shown to have less short-term effects on the growth of very low birth weight infants [28]. In our study, all the infants received corticosteroids systematically, but we did not see any statistically significant difference in postnatal growth between those who did and did not receive corticosteroids. This lack of difference may be explained by the nature of our study population, that is we only included very immature, ventilated infants.

Our study has strengths and some limitations. It is the first to describe the association between duration of invasive ventilation and postnatal growth in extremely preterm infants routinely exposed to antenatal steroids and postnatal surfactant. We used standardised weight and head circumferences [29]. It is a retrospective review, but we present data on all eligible infants and only included infants who had their entire “neonatal” admission in our institution, thus ensuring reliable data.

In conclusion, we have demonstrated that there is a significantly negative association between prolonged ventilation and postnatal growth in extremely prematurely born infants. This may be explained by the increased respiratory load of affected patients and emphasises the need for further research to optimise nutrition in such infants.