Successful use of an artificial placenta–based life support system to treat extremely preterm ovine fetuses compromised by intrauterine inflammation

Abstract

Exvivo uterine environment therapy is an experimental intensive care strategy for extremely preterm infants born between 21 and 24 weeks of gestation. Gas exchange is performed by membranous oxygenators connected by catheters to the umbilical vessels. The fetus is submerged in a bath of synthetic amniotic fluid. The lungs remain fluid filled, and pulmonary respiration does not occur. Intrauterine inflammation is strongly associated with extremely preterm birth and fetal injury. At present, there are no data that we are aware of to show that artificial placenta-based systems can be used to support extremely preterm fetuses compromised by exposure to intrauterine inflammation. To evaluate the ability of our exvivo uterine environment therapy platform to support extremely preterm ovine fetuses (95-day gestational age; approximately equivalent to 24 weeks of human gestation) exposed to intrauterine inflammation for a period of 120 hours, the following primary endpoints were chosen: (1) maintenance of key physiological variables within normal ranges, (2) absence of infection and inflammation, (3) absence of brain injury, and (4) gross fetal growth and cardiovascular function matching that of age-matched in utero controls. Ten ewes with singleton pregnancies were each given a single intraamniotic injection of 10-mg Escherichia coli lipopolysaccharides under ultrasound guidance 48 hours before undergoing surgical delivery for adaptation to exvivo uterine environment therapy at 95-day gestation (term=150 days). Fetuses were adapted to exvivo uterine environment therapy and maintained for 120 hours with constant monitoring of key vital parameters (exvivo uterine environment group) before being killed at 100-day equivalent gestational age. Umbilical artery blood samples were regularly collected to assess blood gas data, differential counts, biochemical parameters, inflammatory markers, and microbial load to exclude infection. Ultrasound was conducted at 48 hours after intraamniotic lipopolysaccharides (before surgery) to confirm fetal viability and at the conclusion of the experiments (before euthanasia) to evaluate cardiac function. Brain injury was evaluated by gross anatomic and histopathologic investigations. Eight singleton pregnant control animals were similarly exposed to intraamniotic lipopolysaccharides at 93-day gestation and were killed at 100-day gestation to allow comparative postmortem analyses (control group). Biobanked samples from age-matched saline-treated animals served as an additional comparison group. Successful instillation of lipopolysaccharides into the amniotic fluid exposure was confirmed by amniotic fluid analysis at the time of administration and by analyzing cytokine levels in fetal plasma and amniotic fluid. Data were tested for mean differences using analysis of variance. Six of 8 lipopolysaccharide control group (75%) and 8 of 10 exvivo uterine environment group fetuses (80%) successfully completed their protocols. Six of 8 exvivo uterine environment group fetuses required dexamethasone phosphate treatment to manage profound refractory hypotension. Weight and crown-rump length were reduced in exvivo uterine environment group fetuses at euthanasia than those in lipopolysaccharide control group fetuses (P<.05). There were no biologically significant differences in cardiac ultrasound measurement, differential leukocyte counts (P>.05), plasma tumor necrosis factor α, monocyte chemoattractant protein-1 concentrations (P>.05), or liver function tests between groups. Daily blood cultures were negative for aerobic and anaerobic growth in all exvivo uterine environment group animals. No cases of intraventricular hemorrhage were observed. White matter injury was identified in 3 of 6 lipopolysaccharide control group fetuses and 3 of 8 vivo uterine environment group fetuses. We report the use of an artificial placenta-based system to support extremely preterm lambs compromised by exposure to intrauterine inflammation. Our data highlight key challenges (refractory hypotension, growth restriction, and white matter injury) to be overcome in the development and use of artificial placenta technology for extremely preterm infants. As such challenges seem largely absent from studies based on healthy pregnancies, additional experiments of this nature using clinically relevant model systems are essential for further development of this technology and its eventual clinical application.