Lung function in very low birth weight infants after pharmacological and surgical treatment of patent ductus arteriosus - a retrospective analysis

A persistently patent ductus arteriosus (PDA) affects about one-third of premature infants with very low birth weight (VLBW) and two-thirds of those with extremely low birth weight (ELBW). Left-to-right shunting through the PDA impairs systemic perfusion, and pulmonary overcirculation may lead to pulmonary edema, increased duration and intensity of mechanical ventilation, impaired alveolar development, and ultimately bronchopulmonary dysplasia (BPD) [1].

This provides the rationale for PDA treatment, although there are different opinions regarding the indications and treatment strategies based on consideration of the potential benefits and adverse effects [1, 2]. Pharmacological treatment with a cyclooxygenase (COX) inhibitor such as indomethacin or ibuprofen leads to PDA closure with a similar efficacy reported for both drugs [1, 3]. When medical treatment is contraindicated or unsuccessful, PDA ligation provides definitive closure. Although the morbidity and mortality rates following PDA ligation are generally low, previous studies have linked this procedure with severe complications such as left ventricular and respiratory compromise, pneumothorax, chylothorax, increased risk of BPD, thoracic scoliosis, and neurosensory impairment [4, 5]. However, it is unclear whether these adverse outcomes are causally associated with surgery itself or merely reflect the more severe illness and/or longer exposure to PDA-related circulatory effects in these patients [6]. Another potentially underestimated complication of PDA surgery is injury to the recurrent laryngeal nerve. This condition may result in voicing disorders and impediments to gas flow in the proximal airways with possible persistence of symptoms into adulthood [7].

The unfavorable short-term effects of PDA on pulmonary mechanics, formation of pulmonary edema, and higher ventilator dependence are well established. However, a direct role of PDA in the development of chronic lung disease/BPD has not yet been proven unequivocally. Several studies have assessed the short-term effects of pharmacological or surgical PDA closure on pulmonary function in animal models [8, 9] and in ventilated infants [10‐12]. We hypothesize that different treatment effects can also be seen in VLBW infants by postnatal lung function testing (LFT). To our knowledge, no previous studies have compared LFT results of spontaneously breathing infants after pharmacological PDA treatment vs. secondary surgical closure during early infancy. The aim of this study was to compare LFT results of infants with PDA who achieved closure following successful pharmacological or secondary surgical treatment after failed pharmacological treatment.

There is an ongoing debate on the potential harms and benefits of PDA treatment in neonates. An hsPDA is thought to negatively affect pulmonary outcomes, however, the different treatment options for PDA (pharmacological, surgical, interventional) themselves exhibit potential risks and unclear benefits [1, 2, 19]. Chen et al. recently provided important data showing that preterm infants who received conservative management for PDA had a higher percentage of BPD than the infants without PDA, but LFT results were comparable between these two groups at discharge [20]. However, these authors did not investigate the effects of specific pharmacological or surgical PDA treatment on LFT. We herein retrospectively analyzed LFT data of a large cohort of VLBW infants at 48 postconceptional weeks after successful pharmacological PDA treatment or PDA ligation secondary to failed pharmacological treatment.

Our data shows that infants who underwent PDA ligation exhibited alterations in their lung function parameters as compared to those who received successful pharmacological treatment. Specifically, the PDA ligation group had reduced respiratory compliance, lower V’max_FRC, increased airway resistance (R_aw), tidal volume, forced expiratory flow, and impaired blood gases.

One major limitation of our study lies in the strong differences between the two groups: At birth the infants who received PDA ligation were more immature and sicker (with significantly higher CRIB scores), and had a longer duration of mechanical ventilation. These factors are associated with higher failure rates of pharmacological intervention for PDA in infants [21]. This is a common problem in the literature on PDA treatment, and due to a lack of randomized trials encompassing a non-treatment arm, researchers have to rely on retrospective analysis which impairs the possibility to establish causal relationships [1, 21]. In order to somewhat circumvent this problem, we conducted a multivariate analysis accounting for the differences in birth weight, gestational age, and durations of hsPDA and mechanical ventilation. This analysis showed that PDA ligation itself was an independent risk factor for increased airway resistance while all other parameters were not independently affected by PDA ligation.

After PDA ligation, we observed an increased airway resistance indicating upper airway obstruction that was independent of immaturity and postnatal development including durations of hsPDA and mechanical ventilation. Although patient characteristics influence the reduction of V’max_FRC and C_rs, only PDA ligation seems to increase R_aw (Table 3). This increase is only weakly reflected by R_rs, because R_aw (measured by whole-body plethysmography) primarily describes air conductivity of the upper airways, whereas R_rs (measured by an airway occlusion test) describes the resistive properties of the whole respiratory system, including the resistance of lung tissue. The increase of R_aw and the expiratory flow limitation during tidal breathing in infants who underwent PDA ligation points to problems in their upper airways. It is tempting to speculate that this proximal airway obstruction was due to unilateral vocal fold paralysis (UVFP) following injury of the recurrent laryngeal nerve. A recent meta-analysis indicated that UVFP occurred in up to 40% of children after surgical PDA closure [22]. In corroboration, adults with UVFP have increased R_aw and, accordingly decreased specific airway conductance [23]. Unfortunately, a laryngoscopic evaluation of vocal fold motility was only performed in a small number of infants in the present study, and this prevented statistical evaluation. In light of a possible increased risk of UVFP, it seems prudent to routinely perform laryngoscopic evaluation in VLBW infants after PDA ligation. LFT is an additional tool to identify infants with upper airway obstruction and to quantify the severity of the airflow impediment. Of note, LFT in infants is non-invasive, as opposed to laryngotracheoscopy, and might help detect incipient upper airway obstruction before the onset of clinical symptoms such as stridor or dyspnea.

V’max_FRC (measured by the rapid thoracoabdominal compression technique) describes small airway conductivity and is highly dependent on maturity. Previous studies showed that V’max_FRC was lower in preterm infants, even in the absence of neonatal disease or therapy [24, 25]. Of note, the differences in tidal breathing parameters were smaller than in other LFT parameters; the only difference was a reduced V_T relative to body weight in the PDA ligation group. There were no significant differences in PTEF and respiratory rate, probably due to the high inter-subject variability of tidal breathing at this age [26‐28] Also, the two groups had no significant differences in lung volume relative to body weight, lung clearance index, or in an inflammatory marker (exhaled FeNO).

In general, data on LFT in infants after PDA treatment, especially after ligation, is sparse. Only a few studies assessed the short-term effects of PDA ligation on pulmonary function in ventilated patients, and the results were contradictory. Two early case series from the late 1970s showed increased pulmonary compliance after surgical PDA closure in ventilated preterm infants [10, 11], and a more recent prospective study of 16 ventilated VLBW infants demonstrated improvements of dynamic compliance (C_dyn), V_T, and V’_E after PDA ligation, although R_aw, mean airway pressure, and peak inspiratory pressure remained unchanged [12]. In contrast, McCurnin et al. [8] investigated the effect of PDA ligation in a premature baboon model and found no significant differences in C_dyn, oxygenation index, or ventilation index (VI), but a trend towards higher Cdyn and lower VI in ligated animals. The post mortem static lung compliance was similar in both groups of baboons. Another study by the same group confirmed the modest effects of ligation on lung function parameters, but reported no improvement in alveolar surface area [9].

Our data shows that, compared to infants after successful pharmacological PDA treatment, patients after PDA ligation exhibited an increase in upper airway resistance and a decreased respiratory compliance at a postconceptional age of 48 weeks. The increased upper airway resistance seems to be associated with PDA surgery itself. In contrast, all the other impairments in pulmonary mechanics observed in ligated infants were due to the patient characteristics, and, despite all the heterogeneity between groups, several LFT parameters showed no significant differences. The differences in the LFT parameters were also reflected by blood gas analysis. We used arterialized capillary rather than arterial blood gas analysis because it is less invasive, easier to perform, and shows a good correlation with arterial blood gas analysis for most parameters.

There is an ongoing controversy regarding the indications and protocols for treatment of PDA [1, 2, 21, 29]. Herein, we provide data that might support an adverse association between some LFT parameters and PDA ligation. It is nonetheless unclear if this is a causal relationship or just an epiphenomenon due to the more severe illness of infants who required surgery. A recent study by Schena et al. found a higher incidence of BPD in infants after PDA ligation [30], but their multivariate analysis showed that a longer duration of hsPDA -- not surgery itself -- was an independent risk factor for BPD. In addition, there are contradictory results on the effects of pharmacological PDA treatment on BPD rates [29]. Furthermore, hsPDA and its treatment influences parameters that affect pulmonary outcome, but cannot be assessed by LFT, such as the pulmonary vasculature. Prospective trials that include a non-treatment arm are needed to address these issues.

In conclusion, despite the retrospective nature of our analysis and the aforementioned differences between the two groups, we provide the first large cohort descriptive data on postnatal LFT results in patients treated for hsPDA using a panel of various techniques that were performed by a single experienced investigator. We suggest that LFT after discharge might help identify infants with increased risk for upper airway obstructions. Despite the potential harms, ligation provides definite PDA closure in infants after failed pharmacological treatment. Nonetheless, the optimal indication and time point for treatment are yet to be determined, and a careful follow-up of this high-risk population is mandatory in order to identify respiratory and other sequelae after PDA ligation.