In our study of premature patients with BPD, we found a high incidence of pathologic spirometry (39.2%) between 5 and 7 years of age, exceeding the rates reported in patients less than 32 weeks of age without BPD (14.1%) and in term controls (5–10%) [12–15].
Although we observed a correlation between the severity of BPD, as defined by the NIH 2001 classification [2], and a decline in lung function values, our study did not find any significant differences between BPD type 2 and BPD type 3. However, upon examination of the requirement for oxygen exceeding 30% at 36 weeks PMA, an increase in the incidence of pathological spirometry became evident. It is noteworthy that none of the patients requiring positive pressure support with ≥ 30% oxygen at 36 weeks PMA met the criteria for normal lung function, in contrast to the 57.7% of patients requiring positive pressure support with < 30% oxygen. Consequently, in our population, the more severe forms of BPD, particularly those that exhibit parenchymal involvement as evidenced by a need for more than 30% oxygen at 36 weeks PMA, are associated with persistent impairment of lung function throughout the school-age years.
One of the most debated aspects of BPD lies in its diagnostic criteria and classification, which depend on the respiratory support received at a specific time. This results in significant variation in long-term predictive capabilities for morbidity across different centers, which can be attributed to differences in respiratory support protocols and oxygen utilization [5, 16]. Furthermore, the diagnosis of BPD coincides with the early stage of alveolar development. This implies that factors encountered during infancy can significantly impact its lung function trajectory and affect the long-term prognosis of affected individuals.
Although there have been several proposed classifications in recent years [5, 6], a unanimous consensus on the most suitable classification has yet to be reached [7].
In addition, the shift in the pathophysiology of this disease as a consequence of extremely premature birth adds complexity to determining the level of impairment that constitutes the diagnosis of BPD. Patients born during the canalicular or saccular phase of pulmonary development will inevitably experience some degree of alteration in lung development. This is particularly evident in studies of premature infants born before 27–28 weeks of gestation, where those who do not meet the diagnostic criteria for BPD still exhibit lung function abnormalities by school age, albeit to a lesser extent than their BPD-diagnosed counterparts. [17, 18]. Therefore, it's crucial to consider BPD not as a simple binary condition but as a spectrum of severity.
Recent proposals for classification [5, 6, 19] suggest a transition towards considering only the need for respiratory support at 36 weeks of gestational age as a diagnostic criterion for BPD. This approach removes the notion of mild BPD (type 1) from the conventional classification, which previously included cases in which oxygen support > 21% was required for more than 28 days, even if no support was required at 36 weeks PMA. This classification practice has been adopted in many studies, often grouping these patients with those without BPD [19]. In several studies that differentiate Type 1 BPD, significant lung function alterations are observed in these patients compared to premature infants without BPD. However, when patient characteristics are examined, it is often found that the type 1 BPD group has a lower gestational age [20, 21]. As a result, distinguishing between the impact of prematurity and the effect of mild BPD diagnosis on long-term morbidity becomes challenging.
The Spanish Research Group on Bronchopulmonary Dysplasia decided to include in the registry patients requiring more than 28 days of respiratory support, acknowledging the inherent heterogeneity in lung involvement within this population. This is because the need for respiratory support in premature infants can arise from factors beyond respiratory issues alone, including central or obstructive apnea, muscle weakness, coexisting conditions and is also influenced by the management protocols employed at each center. The rationale behind this decision is to enhance understanding of the pathophysiology of different phenotypes of BPD patients and analyze both perinatal and childhood risk factors affecting long-term lung function and respiratory morbidity.
In our cohort, even though the differences in lung function abnormalities between those under and over 26 weeks in each BPD group do not reach statistical significance, patients born under 26 weeks had a similar proportion of pathological spirometry as those born over 26 weeks but with a higher grade in the BPD classification. For instance, patients classified as having type 1 BPD exhibited a similar proportion of pathological lung function (37.5%) as those with BPD type 2 born at or after 26 weeks (35%), highlighting the importance of follow-up and evaluation for patients with BPD type 1, particularly those with lower gestational age, to elucidate the impact of these milder forms of BPD on long-term outcomes.
In premature patients with BPD, the most common pathological pattern observed is the obstructive pattern, as reported in previous studies [22–24]. Consistent with these findings, our study shows that 28.2% of cases have an obstructive pattern. In addition, 10.1% have a restrictive pattern and 2.1% have a mixed pattern. Notably, there is a progressive increase in the occurrence of the restrictive-mixed pattern with the severity of BPD, reaching its peak at 50% in the most severe cases when oxygen requirements are considered. This correlation between the restrictive pattern and severe forms of BPD has also been documented in recent studies [12, 20, 25].
In line with the findings reported by Lai et al. [25], a decline in FVC Z-score correlates with exposure to and duration of mechanical ventilation. Furthermore, within our cohort, the presence of a restrictive-mixed pattern is additionally associated with the administration of more than 30% oxygen at birth, diagnosis of pulmonary hypertension, nosocomial pneumonia, length of mechanical ventilation exposure and the necessity for oxygen therapy upon discharge. These correlations suggest an increased level of structural lung damage contributing to the development of a restrictive pattern.
In our cohort, the use of oxygen exceeding 30% at birth correlates with a higher prevalence of compromised lung function in school age. However, due to the observational nature of our study, determining the precise influence of this variable presents challenges. It likely serves as an indicator of prenatal lung injury, exacerbated by known oxygen-related damage and the established correlation between even brief oxygen exposures at birth and the development of BPD [26].
Regarding the obstructive pattern, it is observed in up to 25% of patients diagnosed with type 1 BPD. Although variations in FEV1 Z-score with the severity of BPD are apparent, differences in the occurrence of the obstructive pattern do not reach statistical significance. Interestingly, it is not correlated with exposure to mechanical ventilation, but rather associated with the use of oxygen exceeding 30% at birth, histological chorioamnionitis, and oligohydramnios. This suggest the presence of distinct etiopathogenic mechanisms underlying the development of obstructive versus restrictive-mixed patterns.
Airflow obstruction is a common impairment associated with prematurity, particularly prevalent in patients diagnosed with BPD, and tends to persist into adulthood, as indicated by a recent meta-analysis [27]. This obstruction often arises from a mismatch between distal airway growth and lung volume, known as dysanapsis, resulting from disruptions in normal pulmonary development [28, 29]. In our study cohort, we found that over half of the cases with an obstructive pattern exhibited FEV1 values within the normal range. Notably, this proportion was significantly higher (88.8%) in patients diagnosed with BPD type 1 compared to those with BPD types 2 (44.4%) and 3 (50%). Thus, this pattern of mild obstruction appears to be particularly associated with milder forms of BPD.
In several studies, a gradual improvement in lung function has been observed over time among premature infants with mild forms of BPD, suggesting a potential restoration of disrupted lung development during childhood. Conversely, preterm infants with more severe BPD often exhibit persistent abnormalities in lung function, with values not only lower than those of controls but also falling below the lower limit of normality (LLN) [15, 24, 25, 30]. These findings align with growing evidence suggesting that lung function follows a defined trajectory throughout life, characterized by an increase during childhood, stabilization in adulthood, and subsequent decline with age [31]. Alterations in lung function during childhood have been found to predict adult lung function and the onset of clinical pathology, including the development of chronic obstructive pulmonary disease (COPD). The fact that premature patients, particularly those with BPD, exhibit initial lung function abnormalities makes them particularly vulnerable to the early onset of respiratory failure in adulthood [32, 33]. Recent studies have indicated a higher incidence of COPD among individuals born at younger gestational ages [1], especially among those diagnosed with BPD [33].
While some studies suggest improved long-term respiratory outcomes in contemporary cohorts of preterm infants compared with earlier populations [18, 34]. others show no differences[27, 35, 36].
Advancements in technology and understanding of lung injury mechanisms have led to improved prenatal and postnatal management strategies, including less agressive respiratory support, early minimally invasive surfactant administration, optimized nutrition, and hemodynamic management. As a result of these modifications in treatment approaches, the etiopathogenesis and lung injury profiles associated with prematurity have experience significant changes. Premature infants who can be successfully treated with non-invasive support methods typically experience less structural damage, although varying degrees of pulmonary and vascular developmental impairment may persist depending on the degree of prematurity, leading to the development of milder forms of BPD. On the other hand, modern neonatal care has significantly increased the survival rates of infants who fail non-invasive respiratory support strategies due to extreme prematurity, prenatal pathology and/or genetic factors. These infants often require prolonged mechanical ventilation and have high oxygen requirements, resulting in added structural damage to the lung parenchyma, vasculature and airways. Such patients develop severe forms of BPD, which present a distinct pathophysiology and phenotype compared to milder forms. Therefore, conducting long-term follow-up studies of patients with severe BPD is imperative, given its potential association with the development of a restrictive lung pattern and the significant morbidity and mortality linked to this pattern in adulthood [37, 38].
Limitations
This study is an observational analysis utilizing data sourced from a national database, which offers extensive coverage but also introduces certain constraints. One notable limitation is the low proportion of BPD patients with recorded lung function values within the database between 5 and 7 years of age, potentially leading to selection bias. Additionally, the limited number of patients further constrains the ability to detect statistically significant differences among the gestational age groups. Although spirometry values were carefully provided by pediatric pulmonology specialists at each participating center, the lack of available data on childhood secondhand smoke exposure, lung volumes, and bronchodilator response precluded their inclusion in the analysis. These limitations underscore the need for caution in interpreting the results and highlight areas for future research to address gaps in data availability and improve the comprehensiveness of analyses in this area.
In conclusion, this study emphasizes the high prevalence of pulmonary function abnormalities in patients with BPD and highlights oxygen supplementation greater than 30% at 36 weeks postmenstrual age (PMA) as a valuable marker of disease severity. Of particular note is the identification of a significant association between a restrictive pattern and severe forms of BPD. These findings underscore the urgent need for close patient follow-up, given the significant morbidity and mortality associated with a restrictive pattern in adulthood. Further research is essential to refine management strategies to improve outcomes for this vulnerable population.