The effect of dexamethasone on respirator-dependent very-low-birth-weight infants is best predicted by chest X-ray

Abstract

Background: Chronic lung disease (CLD) in premature infants shows a variable clinical course with different radiological manifestations.

Objective: To evaluate the correlation between parameters of trans-membrane permeability [albumin/secretory component (SC)] and oxidative stress [malondialdehyde (MDA)/SC] in tracheal aspirate fluid (TAF) and radiological findings with the effect of a 5-day course of dexamethasone (0.5 mg/kg per day).

Materials and methods: Fifty ventilator-dependent premature infants with birth weights < 1,500 g (gestational ages 23-31 weeks) and radiological signs of early chronic lung disease (CLD) were treated with dexamethasone at day of life 5-27 (median 10 days) because of respiratory deterioration. TAF was collected serially. Chest X-rays taken before and 8-10 days after dexamethasone were scored for changes of opacification, consolidation and hyperinflation/emphysema, and classified into three groups.

Results: Twenty-four infants had a positive response to dexamethasone, defined as a reduction of the ventilation index FiO2 x mean airway pressure > 40% at day 5, compared to pretreatment values. About 80% of the responders showed homogeneous lung opacification on chest X-ray, reflecting leaky lung syndrome. In contrast, seven of eight infants with predominantly emphysema on radiology were non-responders; 80% of infants with a mixed radiological picture characterized by predominance of consolidations alternating with regions of emphysema were also non-responders. Ratios of albumin/SC and MDA/SC in TAF decreased significantly within 3 days after the onset of dexamethasone. However, MDA/SC was persistently higher in non-responders compared to responders. Opaque lungs were largely improved by dexamethasone, in contrast to streaky or patchy consolidations and emphysema. In a logistic regression model, radiographic classification was the most important factor influencing the response to dexamethasone with a positive predictive value of 86%, followed by albumin/SC ratio.

Conclusions: The optimum timing of dexamethasone treatment may be determined by the stage of developing CLD and radiological findings rather than by the age of the premature infant.