Radiographic thoracic area in newborn infants with Down’s syndrome

Introduction

Down’s syndrome (DS) is the commonest chromosomal abnormality with an incidence of approximately 1 in 700 live births [1]. The syndrome affects multiple systems and organs such as the gastrointestinal tract, the musculoskeletal system and the heart, while cognitive and immunological abnormalities are frequently encountered [2]. Respiratory problems are frequent in DS, accounting for more than half of all hospital admissions [3]. While the lung manifestations of DS have been well described in adults and children [4], there are few studies on respiratory disease in newborns with DS. Lung hypoplasia has been reported in DS and is assumed to originate from the early phases of intrauterine lung development, as affected individuals have a reduced number of alveoli and a decrease in the branch generation number [4]. Lung hypoplasia can present as a clinical problem in DS, but is often not prioritised relative to more pressing diagnoses, such as major congenital heart disease or gastrointestinal atresia. It is not known, however, whether lung hypoplasia in DS can affect outcomes such as the duration of ventilation or supplemental oxygen.

The assessment for lung hypoplasia can be undertaken by measurement of the functional residual capacity (FRC) or by segmentation of magnetic resonance images (MRI). These methods though, are not accessible to most neonatal units caring for infants with DS. We have previously described an alternative method to estimate lung volumes by the chest radiographic thoracic area (CRTA), which is the measurement of the areas corresponding to the lungs on chest radiography. We reported CRTA values in normal newborn infants and in infants with congenital diaphragmatic hernia (CDH) [5]. In infants with CDH the CRTA had a high ability to predict survival to discharge from neonatal care [5], and these findings were subsequently validated by two external cohorts from high-volume surgical centres [6], 7].

We hypothesised that the CRTA would be lower in infants with Down’s syndrome compared to term-born controls and that the CRTA would be significantly related to the duration of invasive ventilation and the duration of supplemental oxygen therapy. Our aims were to test these hypotheses.

Subjects and methods

Chest radiographs

The chest radiographs in the first 24 h after birth were reviewed for each infant and the one with the highest CRTA was included in the analysis. The chest radiographs were anterio–posterior, in the supine position obtained at end-inspiration and at a standard distance of one m above the infant. Rotated radiographs and radiographs with evidence of pneumothorax were excluded from analysis. The radiographs were imported as digital image files by Sectra PACS software (Sectra AB, Linköping, Sweden) which automatically adjusted for magnification errors. Free-hand tracing of the perimeter of the thoracic area as outlined by the diaphragm and the rib cage was undertaken and the CRTA was calculated by the software (Figure 1). The repeatability of the method has been previously reported with an inter- and intra-observer coefficient of repeatability of 1.06 and 1.0 cm² respectively [10].

Figure 1:

A chest radiograph of a newborn infant with Down’s syndrome and pulmonary hypoplasia (A) and a control infant without respiratory disease (B) on the first day of life. The method of free hand tracing of the perimeter of the chest radiographic thoracic area excluding the mediastinal shadow is presented.

Results

In the study period, 67 infants with DS were admitted to the Neonatal Unit at KCH. Eight were excluded because there was no chest radiograph in the first 24 h of life and a further 19 because they were born before 36 weeks of gestation. Forty infants (18 male) with DS were included for measurement and analysis. Their median (IQR) maternal age was 37(33–42) years, duration of rupture of membranes 7(0–24) hours, gestational age 38(36–39) weeks and birth weight 3.11(2.49–3.60) kg (Table 1). Their Apgar score at 5 min was 9(8–10). Seven infants were ventilated at the time of the CRTA measurement for a duration of 5(2–6) days. The duration of supplemental oxygen was 10(2–36) days. Two infants were treated with inhaled nitric oxide and five received inotropes during their neonatal stay. Eight infants (20 %) were diagnosed with a gastrointestinal atresia (none with exomphalos or gastroschisis) and 16(40 %) were diagnosed with structural congenital heart disease. The median IQR duration of stay was 20(9–29) days and all but one infant survived to discharge from neonatal care. Eighty infants (42 male) with a gestational age of 39(38–40) weeks and birth weight of 3.30(2.93–3.66) kg were included as controls. The controls did not differ from DS in birth weight (p=0.273) but had a significantly higher gestational age (p=0.001).

The median (IQR) CRTA was significantly lower in infants with DS [1,922(1,571–2,261) mm²] compared to controls [2,495(2,108–2,908) mm², p<0.001]. The CRTA in the infants with DS was not significantly related to the duration of rupture of membranes (rho=0.289, p=0.192), the duration of invasive ventilation (rho=0.060, p=0.711), the duration of supplemental oxygen (rho=0.195, p=0.227) or the total duration of stay in neonatal care (rho=0.227, p=0.330). When the subgroups of infants with DS without atresia and without congenital heart disease where examined separately, the CRTA was not related to any of the duration of ventilation or supplemental oxygen. The CRTA in the infants with DS was not significantly different in infants with DS who were ventilated vs. not (p=0.423), who had gastrointestinal atresia vs. not (p=0.892) or congenital heart disease vs. not (p=0.847).

Discussion

We have demonstrated that the radiographic thoracic area was lower in newborn infants with Down’s syndrome compared to term controls without respiratory disease, but the CRTA was not significantly associated with a longer duration of ventilation or supplemental oxygen.

Our results of lower thoracic areas in infants with DS compared to healthy controls are in agreement with previous experimental and epidemiological studies, which describe varying degrees of lung hypoplasia in DS [11]. The histological basis of lung hypoplasia in DS appears to persist into adulthood, suggesting that the pulmonary complications are related to early developmental insufficiency [12]. The respiratory issues remain present in later life in children and adults with DS with a variety of presenting symptoms ranging from recurrent lower respiratory infections to persistent pulmonary hypertension [13].

In our population of infants with DS, there was no association between the CRTA measured shortly after birth and later clinical outcomes, such as the duration of ventilation, duration of supplemental oxygen or the duration of stay in neonatal care. This observation possibly highlights that a fixed element of lung hypoplasia is present in the majority of infants with DS, which is not significant enough to prolong their period of respiratory support, compared to possibly stronger effectors such as major abdominal surgery for gastrointestinal atresias or the presence of significant congenital heart disease. It is possible that lower lung volumes might predispose these infants to a higher future risk of respiratory complications such as bronchiolitis in infancy. We should note that KCH is a surgical referral, and not a cardiac referral centre, and as such caters for a population of infants with surgical rather than major cardiac pathology. It might be useful, thus, for a study to be undertaken in the future in another independent cohort, possibly in a centre with predominantly cardiac pathology.

Similar to this study, we have also previously reported that a small CRTA did not translate to a longer duration of ventilation in infants with CDH [5]. It is possible that the duration of ventilation in infants with either DS or CDH is not determined only by the severity of lung hypoplasia as measured by the CRTA, but that other parameters might also influence this relationship. Such parameters in CDH might be the degree of pulmonary hypertension or post-operation complications, and in DS the presence of concomitant pathologies, which might influence cardiorespiratory interactions or major events such as laparotomy for gastrointestinal atresias. Our median CRTA of 1,922 mm² in infants with DS was significantly lower compared to healthy term (median CRTA 2,589 mm²) [9] but higher compared to our previously reported median value of 1,780 mm² in infants with CDH [5].

Methodologically, the assessment of lung hypoplasia can be undertaken by more elaborate assessments than the CRTA, such as FRC by helium dilution, which is considered the gold standard method in neonates [14]. The use of the CRTA, however, as an index of pulmonary hypoplasia has been previously validated by studies in CDH and preterm infants, which reported that CRTA exhibited a strong correlation with the FRC by helium dilution [10], 15]. Other methodologies can also be applied to assess lung hypoplasia, such as lung MRI, which unfortunately is rarely available as an in-house examination in neonatal intensive care units, or the estimation of the end expiratory lung volume by electrical impedance tomography (EIT). It should be noted that EIT also comes with some methodological limitations such as a relatively low spatial resolution, and a lack of standardisation of fundamental experimental parameters such as even the exact position of the measuring electrodes [16].

It was interesting that in our study, infants with DS and gastrointestinal atresias did not have lower CRTAs compared with infants with DS and no atresia. This finding appears to contradict what we have previously described, that lower CRTAs were measured in infants with gastroschisis and exomphalos compared to term controls [17], but in our study none of the infants had gastroschisis or exomphalos suggesting possibly a different pathophysiological mechanism.

Our study has strengths and some limitations. To our knowledge, this is the first neonatal study to describe and quantify lung hypoplasia in admitted newborn infants with DS. We recruited a population of 40 infants with DS and 80 controls, which demonstrated significant differences in the CRTA between the two groups. A possible limitation was that some of our infants were ventilated and some were not, there was however, no significant difference in the CRTA in ventilated infants compared to the ones who were not ventilated. We should also note that in our study we did not perform a sample size calculation but included all eligible infants in our unit. This might be the reason why we did not detect some possible associations: for example a smaller CRTA in DS infants with surgical malformations. Our single centre approach though, guaranteed uniformity of clinical care which would not had been a given in a larger multicentre study with variable intubation criteria and ventilation strategies. Finally, in our study we only included admitted infants, while some infants with uncomplicated DS might not be admitted to neonatal care following birth. If however they had abnormal respiratory signs, they would have been admitted to the neonatal unit – thus we feel we did not miss infants with significant pulmonary hypoplasia.

In conclusion, we report a lower chest radiographic thoracic area in newborn infants with Down’s syndrome, compared to healthy term controls, but this finding was not associated with a longer duration of ventilation or supplemental oxygen therapy.