610 - Association between the length of mechanical ventilation and white matter injury on MRI and neurodevelopmental outcome in preterm infants

Publication Number: 610.3944

Eniko Szakmar, BC Children's Hospital, Vancouver, BC, Canada; Steven Ufkes, University of British Columbia Faculty of Medicine, Vancouver, BC, Canada; Cecil MY. Chau, University of British Columbia Faculty of Medicine, Vancouver, BC, Canada; Vann Chau, The Hospital for Sick Children, Toronto, ON, Canada; Ting Guo, The Hospital for Sick Children, Toronto, ON, Canada; Ruth Grunau, University of British Columbia Faculty of Medicine, Vancouver, BC, Canada; Steven Miller, University of British Columbia Faculty of Medicine, Vancouver, BC, Canada; Julia K. Charlton, University of British Columbia Faculty of Medicine, Vancouver, BC, Canada

Background: Preterm infants managed with prolonged ventilation are at high risk for developing bronchopulmonary dysplasia and also extended exposure to sedatives, which can lead to neurodevelopmental impairment. A better understanding of the relationship between respiratory management, associated exposures, and brain injury among preterm infants is crucial to improve long-term outcomes.

Objective: To assess the association in preterm infants between the length of invasive mechanical ventilation and significant white matter injury (WMI) on magnetic resonance imaging (MRI) at early age and neurodevelopmental delay at 18 months of age.

Design/Methods: This was a post-hoc analysis of data from a prospective cohort study of infants born ≤32 weeks of gestation in a tertiary neonatal intensive care unit in Vancouver, Canada between 2006 and 2013. Significant WMI was defined as ≥ 40 mm3 in volume on brain MRI at early age. The Bayley Scales of Infant and Toddler Development, Third Edition were applied at 18 months of age. Multivariable logistic regression was used to model the association between the length of mechanical ventilation in days and WMI and motor, cognitive and language delay (composite score < 85). Models were expanded to assess mediation effects of days of exposure to opioids (fentanyl and morphine) and midazolam. The mediation effect of significant WMI was assessed for 18 months outcome.

Results: A total of 177 infants (median [IQR] GA at birth, 28 [26- 30] weeks, birth weight, 1020 [809- 1285] grams) had early MRI at median 32 (range 24-32) weeks of postmenstrual age; 165 infants with available ventilation data completed follow up. Significant WMI occurred in 29 (16.4%). Length of ventilation was significantly associated with motor (aOR 1.028 [95% CI 1.008; 1.050], p=0.007) and cognitive delay (aOR 1.037 [95% 1.012; 1.062], p=0.003) at 18 months of age, after adjustment for GA at birth, antenatal steroids (none or full/partial), and maternal education. None of these associations were mediated by the total number of days receiving opioids or midazolam before term-equivalent age, or by significant WMI on early MRI. No association was found between the length of ventilation and WMI on early MRI. See details in Table 1.

Conclusion(s): In this cohort study, a longer duration of mechanical ventilation was associated with higher risk of motor and cognitive delay. The findings did not support a mediating effect of sedative agents or WMI injury for cognitive and motor outcome.

Table 1. Association between length of mechanical ventilation and significant white matter injury (WMI) on early MRI, and motor, cognitive and language delay at 18 months of age (composite score < 85) using multivariable regression models. Models were expanded to assess mediation effects of days of exposure to opioids (fentanyl and morphine) and midazolam. The mediation effect of significant WMI was assessed for 18 months outcome.
1 The model was adjusted for gestational age and antenatal steroids (none or full/partial).
The length of exposures including mechanical ventilation (total days requiring conventional and high-frequency ventilation), opioid administration (total days on intravenous morphine and/or fentanyl) and midazolam use (total days on intravenous midazolam) calculated up to the day of MRI.

2 The model was adjusted for gestational age, antenatal steroids (none or full/partial) and maternal education measured on a 7-point ordinal scale.
The length of exposures including mechanical ventilation (total days requiring conventional and high-frequency ventilation), opioid administration (total days on intravenous morphine and/or fentanyl) and midazolam use (total days on intravenous midazolam) calculated up to term- equivalent age.
Significant white matter injury (WMI) defined as ≥ 40 mm3 in volume on early brain MRI.