Introduction:
Since many preterm infants have impaired lung function, oxygen is one of the most common therapies in neonatal medicine(1). However, preterm infants are also particularly sensitive to the harmful effects of oxygen(2,3). Highly reactive free oxygen radicals cause cellular injury(4). Preterm infants are not adequately protected from such direct biochemical oxidative stress(3,5–7). Oxygen saturation, blood flow and hemoglobin are the three main components of oxygen delivery, and relying solely on arterial oxygen saturation may not adequately monitor tissue exposure to either hypoxia or hyperoxia(8–10). What is the optimal target arterial oxygen saturation as measured by pulse oximetry (SpO2) in infants with hypoxemic respiratory failure (HRF) in the neonatal intensive care unit (NICU), is still an unanswered question(3,5,11). Despite many recent large-scale studies, it is unclear if either low or high target oxygen saturation is safe, provides adequate organ oxygen delivery, and at the same time limits oxygen toxicity and accumulation of reactive oxygen species(3). Accepting lower oxygen saturation targets in infants with acute lung injury may limit toxic oxygen exposure, and enable weaning of inspired oxygen over a shorter period of time(5). However the outcome using this approach is dependent upon the ability of the organs to autoregulate blood flow(12,13). Cerebral autoregulation is defined as the interaction between locally released nitric oxide from red blood cells (RBCs), and the vasodilator ability of the arterioles in tissues, in order to protect oxygen delivery during active tissue metabolism and maintain oxygen saturation of the brain during hypoxemic episodes, in a setting of acceptable hemoglobin, carbon dioxide (CO2) and normal cerebral blood flow (12,14,15). In preterm infants, we do not know if CAR can adequately protect the brain during even brief periods of desaturation or intermittent hypoxemia(16). Our group has developed an integrated approach set-up to provide us with the novel ability to longitudinally measure cerebral autoregulation in real time, in infants undergoing cerebral oxygen and hemodynamics monitoring(17–19). Although it is standard of practice in modern neonatal intensive care units to monitor SpO2 via pulse oximetry, a limited number of units (including our unit) use near infrared spectroscopy (NIRS) to monitor tissue oxygenation, and there is insufficient evidence that either mild desaturations or hypoxemia impact cerebral autoregulation. Understanding the temporal relationship between desaturation and cerebral blood flow autoregulation should lead to individualized safer weaning of oxygen while maintaining the integrity of CAR(18,20,21).
We hypothesized that CAR can compensate for short periods of hypoxemia or desaturation in preterm infants with normal hemodynamics and hemoglobin, thus maintaining brain oxygen tension, and this compensatory role of CAR can be clinically recognized as low SpO2 below target limits but crRTO is maintained without change.