Take a deep breath: the important role of vitamin A in neonatal lung development and visiting the potential of aerosolized delivery for bronchopulmonary dysplasia prevention

1. Introduction

Despite ongoing improvements in perinatal care and the increasing survival of extremely premature infants, the prevalence of bronchopulmonary dysplasia (BPD) continues to increase and remains the most common serious morbidity of prematurity [1,2]. BPD more than doubles the initial hospital per-patient costs and contributes to lifelong sequalae including chronic lung disease and cardiac and neurological comorbidities, all associated with significant socioeconomic effects [3,4]. Although numerous therapeutic interventions have been investigated, only a handful have consistently prevented BPD in clinical trials – most notably intramuscular (IM) vitamin A [5-13]. However, it is widely recognized that BPD prevention and treatment remain a significant and growing unmet medical need.

2. Vitamin A and human lung evolution

Retinoids, or colloquially ‘vitamin A,’ have played a crucial evolutionary role in mammalian physiology, especially in facilitating water to land transition, and in regulating critical organ development and function both in utero and early development [14,15]. Vitamin A plays a key role in normal vision, immune function, and lung development including airway branching and subsequently, alveolarization, endothelial differentiation, vascularization, and surfactant production [16,17]. In placental mammals, where fetal lung maturation must be synchronized with birth to enable immediate gas exchange, vitamin A accumulation in the third trimester reflects an adaptive mechanism to prepare the lungs for effective air breathing at birth [18].

3. Perinatal biology of vitamin A

The biochemistry and cell biology of vitamin A are vastly complex and beyond the scope of this editorial; the absorption, transport, regulation, and fetal storage of vitamin A have been well studied [19]. Fetal vitamin A accumulation primarily occurs in the third trimester. As a result, preterm infants have reduced plasma retinol and hepatic retinyl ester stores and often low circulating vitamin A levels throughout infancy. In plasma, vitamin A circulates bound to retinol-binding protein (RBP), which complexes with transthyretin in a 1:1 ratio. Infants who develop BPD often have lower initial plasma retinol levels, although this association is not consistently observed due to numerous confounding factors [20-23]. Plasma retinol levels of <0.7 μM (equiv. to <200 ug/L) are indicative of vitamin A deficiency, with levels half of that (<0.35 μM, or <100 ug/L) reflecting severe deficiency, depleted hepatic retinol stores, low circulating RBP levels, and low retinol/RBP molar ratios.

4. Vitamin A and BPD prevention

Of the many BPD preventive strategies, including early surfactant administration, restriction of fluid intake, early patent ductus arteriosus closure, minimizing ventilatory support, anti-inflammatory agents such as systemic and inhaled corticosteroids, caffeine, and vitamin A supplementation, IM vitamin A is among few treatments shown in randomized trials to reduce the incidence of BPD. The most notable data on IM vitamin A for preventing BPD comes from the landmark, double-blind, placebo-controlled Tyson study [5] and supporting meta-analyses of many smaller clinical studies [24-26]. The use of IM vitamin A remains limited, however, influenced in part by variability in level of clinical evidence and practitioner preferences.

Several meta-analyses highlight the challenges posed by the lack of standardization across studies evaluating vitamin A for BPD prevention [24-27]. These challenges include variability in BPD disease definitions, patient populations studied, and dosing regimens employed. Nevertheless, contemporary NICU guidelines from several highly regarding institutions include IM vitamin A as part of the standard-of-care for preterm infants, with few cautions or contraindications noted on its use in this population (e.g. [28,29]). Enteral administration, although appealing due to its noninvasive nature, has proven ineffective in reducing incidence of BPD in clinical trials, using dosing regimens comparable (on a per-unit basis) to those used for IM administration [25,27,30].

Some reports have weighed the invasiveness of IM injections and concerns related to pain, bruising, and risk of infection as well the emotional toll of such procedures on caregivers against the greater burden of subsequent procedures and interventions typically required for infants who develop BPD [31,32]. An argument can be made that the tangible risks of IM dosing may seem more immediate than the relatively abstract benefit of preventing one case of BPD per 14 infants treated, the estimated number needed-to-treat (NNT) based on Tyson’s findings [5]. Such an assertion seems consistent with differing opinions regarding prioritization of improvements in neonatal care solicited from a broad range of stakeholders including physicians, nurses, caregivers, and family members, which showed that immediacy of topic can outweigh future less-definite outcomes (e.g. BPD ‘might’ develop vs. observation of the immediate response to an invasive treatment like airway intubation) [33].

Although the shortage of parenteral vitamin A that began in the previous decade has since been resolved, the return of commercial availability came with significant price increases. However, even at higher costs, vitamin A supplementation has been deemed cost-effective in the NICU setting when weighed against the long-term expenses associated with BPD care [6]. Ultimately, even those who support the use of IM vitamin A likely agree that preventing BPD remains a critically important and largely unmet medical need.

It is within this context that we began evaluating inhaled vitamin A dosing as an alternative, noninvasive approach for preventing BPD under the hypothesis that direct-to-target-organ dosing might be more efficacious [34]. Previous investigations of direct-to-lung vitamin A dosing (via tracheostomy in rats [35] or liquid intratracheal instillation in lambs [36]) focused on serum and tissue vitamin A levels, and not specifically or extensively on pulmonary developmental or injury-repair benefits of vitamin A delivered directly to the lungs.

We demonstrated in a preclinical study that an aerosolized, inhaled, optimized vitamin A formulation (devoid of chlorobutanol and citric acid) significantly mitigated hyperoxia-induced lung injury and promoted alveologenesis in an established pre-weaned rat model of BPD [34]. Consistent with the model, untreated hyperoxia yields lung with fewer, larger, and thicker walled alveoli. Notably, co-administration of inhaled vitamin A (four doses administered over a few minutes each over 7 days) yielded alveolar metrics essentially indistinguishable from healthy controls. Molecular pathways mirrored these findings: in untreated hyperoxia, proteins related to lung homeostasis and function (PPARγ and Surfactant Protein-C) drop, while apoptosis (indicated by a decrease in BCL-2/Bax ratio) and repair proteins (indicated by an increase in β-catenin, and ALK-5) rise; yet inhaled vitamin A resulted in normalizing levels similar to those observed in healthy controls. Based on these encouraging results, we expect a durable benefit (e.g. after acute hyperoxia is discontinued, mimicking aging past the more lung injury-prone period of the developing lung), from a study ongoing in our lab.

Consistent with extensive clinical literature as discussed above, modest effects were also observed in hyperoxia-exposed rats dosed with IM injections [34], with a scaled dosing regimen equivalent to the typical clinical dose 5000 U/dose, dosed 3 days a week for 4 weeks [5,37]. In our hands, both IM and inhaled dosing increased serum vitamin A toward the higher end of the normal range in observed in normoxia animals (see Figure 3D in [34]), which was not statistically different from healthy normal controls, yet with vastly different effects in mitigating hyperoxic lung damage.

The underlying mechanisms for the superior effectiveness that we observed with inhaled delivery of vitamin A in supporting lung injury repair vs IM or enteral dosing are likely attributable to direct-to-target-organ delivery and resultant more effective modulation of important molecular and pharmacokinetic factors. Inhalation delivers vitamin A directly to the airway epithelium and alveoli, the primary sites of lung injury, resulting in higher local concentrations at the site of damage avoiding first-pass metabolism or dilution in systemic circulation. For example, in an adult rat model of lung injury, inhalation of all-trans retinoic acid leads to approximately 12-fold higher lung concentrations compared to that achieved by systemic administration [38]. Moreover, direct delivery of vitamin A to the lungs circumvents the need for systemic transport mechanisms involving RBP, ensuring free diffusion/uptake of vitamin A directly by lung cells [39]. Inhaled vitamin A also likely increases effective receptor occupancy in injured pulmonary cells, resulting in more immediate and potent activation of cytosolic and nuclear signaling cascades relevant to lung development and injury repair. In contrast, systemic (IM) or enteral vitamin A requires systemic transport to reach alveolar cells and activation of critical metabolic pathways which does not appear to be adequate in terms of having a meaningful impact on preventing development of BPD.

Although definitive data are lacking, we speculate that plasma retinol is an imperfect surrogate for tissue retinoid exposure in extremely preterm infants. IM administration creates a depot that bypasses first-pass hepatic esterification and releases retinyl esters slowly, producing sustained retinol/retinol-binding-protein (RBP) micro-pulses [40]. These pulses appear to replenish pulmonary retinyl-ester stores, mirroring the rise in serum vitamin A and reduction in BPD reported in the landmark IM vitA trial [5]. In contrast, it remains uncertain whether the immature preterm gut can absorb enteral vit A in amounts sufficient to support lung development [41,42]. A recent large phase-3 trial of high-dose enteral supplementation neither increased serum vitamin A nor prevented BPD, findings that likely reflect limited intestinal absorption in this population [30].

Inhaled vitamin A also may provide lung-specific benefits by enhancing interactions between alveolar type II cells and lipofibroblasts, which are critical for lung maturation, home-ostasis, and repair [43]. We previously showed that while both IM and inhaled vitamin A increased pulmonary PPARγ under normoxic conditions, only inhaled vitamin A preserved its expression during hyperoxic stress, a common challenge encountered by preterm infants [34]. Inhaled vitamin A also maintained normal levels of key mesenchymal structural proteins (e.g. fibronectin and elastin) at later developmental timepoints we studied, suggesting that short-term inhaled therapy during the neonatal period more effectively activates retinoid signaling and epithelial–mesenchymal crosstalk essential for long-term lung development and injury repair.

5. Conclusion and future directions

Inhaled vitamin A dosing as an alternative noninvasive approach to prevent BPD with direct-to-target-organ delivery holds significant promise to address a critical unmet medical need. Additional preclinical translational studies, including evaluation of animals allowed to age into adulthood and, separately, in large animal BPD models, are ongoing to support future assessment of aerosol delivery of vitamin A in clinical trials in fragile preterm infants to prevent BPD and its long-term sequelae.