Diffusion tensor imaging detects ventilation-induced brain injury in preterm lambs

Early administration of umbilical cord blood cells following brief high tidal volume ventilation in preterm sheep: a cautionary tale

BACKGROUND

Umbilical cord blood (UCB) cells are a promising treatment for preterm brain injury. Access to allogeneic sources of UCB cells offer the potential for early administration to optimise their therapeutic capacities. As preterm infants often require ventilatory support, which can contribute to preterm brain injury, we investigated the efficacy of early UCB cell administration following ventilation to reduce white matter inflammation and injury.

METHODS

Preterm fetal sheep (0.85 gestation) were randomly allocated to no ventilation (SHAM; n = 5) or 15 min ex utero high tidal volume ventilation. One hour following ventilation, fetuses were randomly allocated to i.v. administration of saline (VENT; n = 7) or allogeneic term-derived UCB cells (24.5 ± 5.0 million cells/kg; VENT + UCB; n = 7). Twenty-four hours after ventilation, lambs were delivered for magnetic resonance imaging and post-mortem brain tissue collected. Arterial plasma was collected throughout the experiment for cytokine analyses. To further investigate the results from the in vivo study, mononuclear cells (MNCs) isolated from human UCB were subjected to in vitro cytokine-spiked culture medium (TNFα and/or IFNγ; 10 ng/mL; n = 3/group) for 16 h then supernatant and cells collected for protein and mRNA assessments respectively.

RESULTS

In VENT + UCB lambs, systemic IFNγ levels increased and by 24 h, there was white matter neuroglial activation, vascular damage, reduced oligodendrocytes, and increased average, radial and mean diffusivity compared to VENT and SHAM. No evidence of white matter inflammation or injury was present in VENT lambs, except for mRNA downregulation of OCLN and CLDN1 compared to SHAM. In vitro, MNCs subjected to TNFα and/or IFNγ displayed both pro- and anti-inflammatory characteristics indicated by changes in cytokine (IL-18 & IL-10) and growth factor (BDNF & VEGF) gene and protein expression compared to controls.

CONCLUSIONS

UCB cells administered early after brief high tidal volume ventilation in preterm fetal sheep causes white matter injury, and the mechanisms underlying these changes are likely dysregulated responses of the UCB cells to the degree of injury/inflammation already present. If immunomodulatory therapies such as UCB cells are to become a therapeutic strategy for preterm brain injury, especially after ventilation, our study suggests that the inflammatory state of the preterm infant should be considered when timing UCB cells administration.

Preterm lung and brain responses to mechanical ventilation and corticosteroids

Mechanical ventilation is necessary to maintain oxygenation and ventilation in many preterm infants. Unfortunately, even short periods of mechanical ventilation can cause lung and airway injury, and initiate the lung inflammation that contributes to the development of bronchopulmonary dysplasia (BPD). The mechanical stretch leads to airway cell differentiation and simplification of the alveoli, and releases cytokines that cause systemic response in other organs. Mechanical ventilation also leads to brain injury (IVH, white and gray matter) and neuronal inflammation that can affect the neurodevelopment of preterm infants. In efforts to decrease BPD, corticosteroids have been used for both prevention and treatment of lung inflammation. Corticosteroids have also been demonstrated to cause neuronal injury, so the clinician must balance the negative effects of both mechanical ventilation and steroids on the brain and lungs. Predictive models for BPD can help assess the infants who will benefit most from corticosteroid exposure. This review describes the lung and brain injury from mechanical ventilation in the delivery room and chronic mechanical ventilation in animal models. It provides updates on the current guidelines for use of postnatal corticosteroids (dexamethasone, hydrocortisone, budesonide, budesonide with surfactant) for the prevention and treatment of BPD, and the effects the timing of each steroid regimen has on neurodevelopment.

Investigating Pathways of Ventilation Induced Brain Injury on Cerebral White Matter Inflammation and Injury After 24 h in Preterm Lambs

Initiation of respiratory support in the delivery room increases the risk and severity of brain injury in preterm neonates through two major pathways: an inflammatory pathway and a haemodynamic pathway. The relative contribution of each pathway on preterm brain injury is not known. We aimed to assess the role of the inflammatory and haemodynamic pathway on ventilation-induced brain injury (VIBI) in the preterm lamb. Fetal lambs (125 ± 1 day gestation) were exteriorised, instrumented and ventilated with a high tidal-volume (V_T) injurious strategy for 15 min either with placental circulation intact to induce the inflammatory pathway only (INJ_INF; n = 7) or umbilical cord occluded to induce both the inflammatory and haemodynamic pathways (INJ_INF+HAE; n = 7). Sham controls were exteriorised but not ventilated (SHAM; n = 5) while unoperated controls (UNOP; n = 7) did not undergo fetal instrumentation. Fetuses were returned in utero following intervention and the ewe allowed to recover. Arterial blood gases and plasma were sampled periodically. Twenty-four hours following intervention, lambs were delivered and maintained on non-injurious ventilation for ∼40 min then brains were collected post-mortem for immunohistochemistry and RT-qPCR to assess inflammation, vascular pathology and cell death within white matter regions. Compared to INJ_INF lambs, INJ_INF+HAE lambs achieved a consistently higher V_T during injurious ventilation and carotid blood flow was significantly lower than baseline by the end of ventilation. Throughout the 24 h recovery period, systemic arterial IL-6 levels of INJ_INF+HAE lambs were significantly higher than SHAM while there was no difference between INJ_INF and SHAM animals. At 24 h, mRNA expression levels of pro-inflammatory cytokines, tight junction proteins, markers of cell death, and histological injury indices of gliosis, blood vessel protein extravasation, oligodendrocyte injury and cell death were not different between groups. Injurious ventilation, irrespective of strategy, did not increase brain inflammation or injury 24 h later when compared to control animals. However, the haemodynamic pathway did influence carotid blood flow adaptations during injurious ventilation and increased systemic arterial IL-6 that may underlie long-term pathology. Future studies are required to further characterise the pathways and their long-term effects on VIBI.

Aberrant cortico-thalamic structural connectivity in premature-born adults

Premature birth is associated with alterations in brain structure, particularly in white matter. Among white matter, alterations in cortico-thalamic connections are present in premature-born infants, and they have been suggested both to last until adulthood and to contribute to impaired cognitive functions. To test these hypotheses, 70 very premature-born adults and 67 full-term controls underwent cognitive testing and diffusion-weighted imaging. Each cortical hemisphere was parcellated into six lobes, from which probabilistic tractography was performed to the thalamus. Connection probability was chosen as metric of structural connectivity. We found increased cortico-thalamic connection probability between left prefrontal cortices and left medio-dorsal thalamus and reduced connection probability between bilateral temporal cortices and bilateral anterior thalami in very premature-born adults. Aberrant prefronto- and temporo-thalamic connection probabilities were correlated with birth weight and days on ventilation, respectively, supporting the suggestion that these connectivity changes relate with the degree of prematurity. Moreover, an increase in left prefronto-thalamic connection probability also correlated with lower verbal comprehension index indicating its relevance for verbal cognition. Together, our results demonstrate that cortico-thalamic structural connectivity is aberrant in premature-born adults, with these changes being linked with impairments in verbal cognitive abilities. Due to corresponding findings in infants, data suggest aberrant development of cortico-thalamic connectivity after premature birth with lasting effects into adulthood.

Respiratory Support of the Preterm Neonate: Lessons About Ventilation-Induced Brain Injury From Large Animal Models

Many preterm neonates require mechanical ventilation which increases the risk of cerebral inflammation and white matter injury in the immature brain. In this review, we discuss the links between ventilation and brain injury with a focus on the immediate period after birth, incorporating respiratory support in the delivery room and subsequent mechanical ventilation in the neonatal intensive care unit. This review collates insight from large animal models in which acute injurious ventilation and prolonged periods of ventilation have been used to create clinically relevant brain injury patterns. These models are valuable resources in investigating the pathophysiology of ventilation-induced brain injury and have important translational implications. We discuss the challenges of reconciling lung and brain maturation in commonly used large animal models. A comprehensive understanding of ventilation-induced brain injury is necessary to guide the way we care for preterm neonates, with the goal to improve their neurodevelopmental outcomes.

Surfactant plus budesonide decreases lung and systemic inflammation in mechanically ventilated preterm sheep

Mechanical ventilation with normal tidal volumes (VT) causes lung and systemic inflammation in preterm sheep. Mechanical ventilation is associated with bronchopulmonary dysplasia (BPD) in preterm infants, and the addition of budesonide to surfactant decreases BPD in clinical trials. Budesonide with surfactant will decrease the lung injury from mechanical ventilation for 24 h in preterm sheep. Lambs at 126 ± 1 day gestational age were delivered and randomized to either: 1) surfactant (200 mg/kg) or 2) surfactant mixed with budesonide (0.25 mg/kg) before mechanical ventilation with VT of 7-8 ml/kg for 2, 6, or 24 h (n = 6 or 7/group). Lung physiology and budesonide levels in the plasma and the lung were measured. Lung tissue, bronchoalveolar lavage fluid (BALF), liver, and brain tissues were evaluated for indicators of injury. High initial budesonide plasma levels of 170 ng/ml decreased to 3 ng/ml at 24 h. Lung tissue budesonide levels were less than 1% of initial dose by 24 h. Although physiological variables were generally similar, budesonide-exposed lambs required lower mean airway pressures, had higher hyperoxia responses, and had more stable blood pressures. Budesonide decreased proinflammatory mRNA in the lung, liver, and brain. Budesonide also decreased total protein and proinflammatory cytokines in BALF, and decreased inducible nitric oxide synthase activation at 24 h. In ventilated preterm lambs, most of the budesonide left the lung within 24 h. The addition of budesonide to surfactant improved physiology, decreased markers of lung injury, and decreased systemic responses in liver and brain.