Discussion
In this single-center prospective cohort study of infants born from mothers with Pre-E, no differences were detected in the lung parenchymal function of gas exchange and lung volume, as assessed by DLCO, DLCO/VA, or VA, when compared to control infants born from normotensive mothers. This finding does not support our hypothesis nor agree with animal models of pre-eclampsia, which suggest impaired alveolar development in the offspring of mothers with Pre-E. However, we did find that airway function, assessed by FEFs, were actually higher in the non-severe Pre-E group compared to Control and severe Pre-E groups. Although the non-severe Pre-E group had higher FEFs, the severe Pre-E group, which did not have higher FEFs, had a lower risk for wheeze in the first year of life when adjusted for covariates including GA. These findings highlight the complex relationships among prematurity, lung function and respiratory morbidity in infants born from mothers with Pre-E.
We are not aware of other studies that have assessed DLCO as a respiratory outcome in offspring of mothers with Pre-E. The majority of infants born following maternal pre-eclampsia are preterm because labor is either induced or a Cesarean section is performed at a premature GA to protect the mother with severe Pre-E. Therefore, preterm birth, which is associated with impaired alveolar development and increased respiratory morbidity, becomes an important confounder in the assessment of respiratory outcomes of lung function and respiratory morbidity. We therefore used a control group from normotensive pregnancies that included preterm and term births. Our findings that DLCO and VA did not differ between Pre-E (Non-Severe and Severe) and normctentsive Control group suggests that Pre-E did not significantly impair alveolar development when evaluated at a mean corrected-age of 7-8 months.
Our findings related to parenchymal function do not support our initial hypothesis and are not consistent with current animal models of Pre-eclampsia. Tang et al reported that the amniotic administration of the anti-angiogernic factor FLT-1 to pregnant rats 2 days prior to delivering pups via caesarian section resulted in decreased alveolar number, reduced pulmonary vessel density, and right ventricular hypertrophy [10]. More recently, Taglauer et al used a heme oxygenase-1 null mouse model of pre-eclampsia to demonstrate disrupted alveolar formation and altered airway development. This model was also associated with a down-regulation of angiogenic and epithelial pathways, as well as an up-regulation of inflammatory and extra-cellular matrix pathways, suggesting multiple molecular pathways contributing to the observed pulmonary phenotype. It is possible that the Pre-E infants we evaluated may have demonstrated impaired alveolar development if evaluated during the neonatal period and subsequently exhibited catch-up in alveolar development prior to our evaluation. However, the animal models of Pre-E that demonstrated impaired alveolar development often evaluated animal offspring at human developmental age equivalent to our study in human infants [19]. In addition, there are currently no longitudinal data in humans to indicate that there is catch-up lung growth following preterm birth or maternal pre-E. Therefore, it remains unclear how well the current animal models of Pre-E reflect clinical Pre-E, and the various subtypes, which may result from multiple differing factors and be associated with multiple co-morbidities, such as fetal growth restriction and prematurity, which can affect lung development.
In contrast to no differences in DLCO and VA, we did find higher FEFs in infants of mothers with Non-severe Pre-E. This finding is consistent with the higher FEFs reported in older children born of mothers with Pre-E, although that study of older children was restricted to subjects born preterm with GA < 28 weeks or weighing < 1000g [20]. In both that study and ours, FVC did not differ between Pre-E and control groups, suggesting that the higher FEFs were related to differences in airway function rather than to differences in lung volume. In our study, we also found that VA did not differ between Pre-E and Control groups, again suggesting that differences in FEFs were secondary to differences in airway function rather than differences in lung volumes. The only previous study evaluating infants born to women with Pre-E was by Stokholm, et al [13]. These investigators reported that at 1-month of age, FEV0.5 and FEF50 were not significantly different comparing infants from Pre-E and non-PreE mothers; however, in that study, all infants were from mothers with asthma, which may also have an effect upon the airway function of offspring. The mechanism for our observed higher FEFs in offspring of PreE mothers is unclear and future studies might include a more direct assessment of airway size, such as high resolution computed tomography.
The increased risk of wheeze we found related to preterm birth, maternal smoking, family history of asthma, and antenatal steroids is consistent with previous reports in the literature [21, 22]. The lower risk for wheeze in the Pre-E group, after adjusting for other covariates related to wheeze (Table 3), was primarily driven by the severe Pre-E group (OR=0.42) (Table 4). Although the non-severe Pre-E group tended to have a lower risk of wheeze compared to Controls (OR=0.61), this was not statistically significant, which may be related to the fewer Non-Severe compared to Severe Pre-E infants evaluated (41 vs 105).
In prior studies of airway function among full-term infants without Pre-E, higher airway function during infancy was associated with lower risk for subsequent wheezing in the first year of life [23]. We found a similar relationship between higher FEFs and lower risk of wheeze only when all subjects were evaluated as a single group, but not for the individual groups (Control, Non-Severe and Severe Pre-E). Our Non-Severe Pre-E group had significantly higher FEFs; however, their lower risk of wheeze did not reach statistical significance. The Severe Pre-E group had a significantly lower risk of wheeze compared to Controls, even after adjusting for several covariates that increase the risk of wheeze and more frequent in the severe Pre-E group; however, this group did not have significantly higher FEFs. These inconsistencies may relate to differences in these two very different respiratory outcomes, the limited number of infants in the non-severe Pre-E group, as well as as the multiple factors that can contribute to wheeze. Spirometry is assessed while infants are sleeping and without any intercurrent respiratory illness, making FEFs a reproducible objective measurement of airway function when not symptomatic. In contrast, wheeze is determined from parental questionnaire as a sign of airways obstruction when the infant is ill. The mechanisms that contribute to wheeze are complex and can include the baseline airway function, as well as the inflammatory responses to stimuli, such as viruses and allergen. In addition, molecular pathways that contribute to Non-severe and Severe Pre-E may not be a continuous spectrum, but rather may represent differing pulmonary phenotypes. It is also possible that Non-Severe and Severe Pre-E, as well as prematurity, differ in their effects upon baseline airway development and function, as well as immune development and responses to stimuli. Therefore the relationship between airway function and wheezing previously observed in full-term infants may be more indirect and may not apply in our populations of infants.
Our study has several strengths and limitations. One of the strengths of our study was the ability to obtain a detailed assessment of lung function in infants. We were able to address the effect of Pre-E upon the lung parenchyma development, as well as airway function. Importantly, we evaluated Control subjects from normotensive pregnancies with a balanced mix of GA and sex, as these factors can contribute to alterations in lung growth and development. Preterm infants with chronic lung disease of infancy (CLDI) have lower DLCO, but normal VA compared to fullterm infants[24]. However, the effect of Pre-E upon alveolar development may be less than that observed with CLDI. Our study also had limitations. All of the infants were recruited from a single health system, which could have led to bias in the study population or treatment of Pre-E. Although our cohort size was large for an IPFT study, the number of infants evaluated was still relatively small, potentially rendering us underpowered to detect some associations. Although we accounted for multiple covariates in our analysis, we were unable to adjust for all potential confounders, such as respiratory viral infections. The number of very low GA infants was small, and it is possible that the impact of Pre-E on pulmonary outcomes in extremely low GA neonates would be different from our observations. Lastly, our DLCO measurement was obtained in infants sleeping, which may have limited our ability to detect smaller differences of impaired alveolar development, which may only be present under conditions of increased cardiac output, such as exercise [25].
In summary, the results of our study do not support the hypothesis that in utero Pre-E exposure leads to impaired lung parenchymal development in humans. The differences between our findings in humans and those reported from animal models may be due to differences in the impact of anti-angiogenic factors on lung development or the ability of the human lung to rapidly compensate for in utero anti-angiogenic factors. However, we separately found better airway function and decreased wheeze, but not in the same severity group of Pre-E offspring. Therefore, differing Pre-E severity may represent differing molecular pathways, which could result in differing pulmonary phenotypes. Further research is needed to obtain a more comprehensive understanding of the effect of Pre-E on lung development and respiratory morbidity.