Inflammation, electrocortical and vagal activity with acute and chronic hypoxia in near term fetal sheep

Resumé of the project

Study 1. Electrocortical activity during repetitive umbilical cord occlusions with worsening acidemia in the ovine fetus near term.

There is an ongoing need for better tools to early detect fetal compromise during labour to prevent brain injury. Little is known about the fetal brain (electrocortical, ECOG) responses to worsening hypoxic acidemia as it may occur during labour. We induced repetitive umbilical cord occlusions (UCO) in fetal sheep to determine the value of ECOG to predict worsening acidemia. We demonstrated an early emergence of a specific dynamic pattern of fetal heart rate (FHR) and ECOG spectral edge frequency onsetting at moderate pH and base excess levels in the course of the UCO well in advance of the threshold for severe acidemia and increased risk for asphyxia-mediated brain injury. Given the recent success in monitoring fetal electroencephalographic activity during human labour, our findings could improve the clinical ability to predict fetal compromise during labour in the high-risk pregnant patients using combined EEG/FHR monitoring to signal adaptive brain shut-down and thereby the need for delivery. Transdisciplinary aspect and knowledge transfer: The respective paper is currently under review in AJOG. To move this research from bench to bed we filed a provisional world-wide patent for ‘Fetal EEG monitoring’ and acquired one internal and one external proof-of-principle (POP) grants to test if fetal EEG can indeed be acquired during human labour. Currently I am analyzing data comparing ECoG to EEG recordings of fetal sleep to prove that they are similar and hence EEG can also be used to characterize fetal brain electrical activity. Our next step is application for CIHR POP funding for which the letter of intent was approved by CIHR in the spring competition. With this grant we hope to show that our finding of specific fetal ECoG changes in relation to FHR allows us to predict impending fetal acidemia under variable conditions of worsening hypoxic acidemia. These steps are necessary before we can move to a prospective clinical study.

Study 2. Cholinergic anti-inflammatory pathway is active in late gestation fetal sheep and inhibits microglial activation following inflammatory stimulation.

Recently, a cholinergic anti-inflammatory pathway (CAP) has been identified whereby increased vagal activity inhibits the release of pro-inflammatory cytokines such as IL-1β. It has been shown that (1) RMSSD (a measure of HRV reflecting vagal activity) may serve as a marker of vagal modulation of inflammatory activity in adults reflecting the activity of the CAP; and (2) systemic inflammation is an important mechanism acting independently to cause fetal brain injury or contributing to in utero asphyxial brain injury with consequences for postnatal health. However, the existence of fetal CAP and its relationship to fetal hypoxia and acidemia are unknown. We addressed this issue in late gestation, chronically-instrumented fetal sheep. We found that, similar to adult animals and humans, CAP is active in the fetus pre-term as evidenced by individual measures of vagal modulation of fetal HRV and an inverse correlation between indices of HRV and the levels of the pro-inflammatory cytokine IL-1β. Fetal inflammatory response and CAP are activated by worsening acidemia near-term, but not by intermittent chronic hypoxia pre-term. Consequently, the inhibition of cerebral innate immune response through microglia by CAP is found in near-term fetuses only. Our findings suggest that RMSSD may serve as the marker of fetal CAP activity and that boosting CAP activity may reduce microglial activation due to an inflammatory stimulus.

Transdisciplinary aspect and knowledge transfer: Such translational research has the potential to meaningfully impact therapeutic management strategies and to improve perinatal and postnatal outcomes. By reducing microglial activation, we think we can protect the foetal brain and decrease the number of babies born with increased susceptibility to develop newborn or adult neurological diseases due to inflammatory brain damage. Last year I acquired an internal grant with my supervisor Dr. Bryan Richardson to study molecular components of CAP in fetal brain and changes in glial and neuronal activity following inflammatory response to UCO induced in Study 1. I hope to present some preliminary data at the November Kingston meeting.

General aspects of the transdisciplinary approach in my studies

I try to solve clinical obstetrical problems from neuroscience/fetal physiology point of view applying biomedical engineering (signal analysis) techniques. Presentation at local, regional and international meetings allows for wider dissemination of results and feedback from other scientists.

Two questions for feedback:

1. Which further transdisciplinary aspects should I consider for my studies?

2. Can you suggest explanations for the mediation of cerebral neuroprotection via efferent vagus nerve?

Details on Methods & Results

Study 1: 'Near-term group' - ECOG analyses

Study design. Ten near-term fetal sheep (0.89 gestation) were chronically prepared with arterial catheters and placement of an inflatable umbilical cord occluder. Following 3 days of post-op recovery and a baseline recording period, they were studied through a series of mild, moderate and severe UCO until the pH was below 7.00, this being the threshold for severe acidemia with increased risk for hypoxic-ischemic brain injury. For more details on methodology please see PDF attached. ECOG was recorded from two biparietally placed electrodes planted on the dura. 95% Spectral Edge Frequency (SEF) was calculated from ECOG in 4 second sliding windows.

Results. Repetitive cord occlusions during the mild UCO series, when fetal pH was still unchanged, markedly disrupted ECOG state activity with a ~50% decrease in the incidence of low voltage/high frequency activity and a corresponding increase in indeterminate voltage/frequency activity. At a pH of 7.24 ± 0.04 and base excess of -2.9 ± 2.2 mmol/L, a specific pattern of UCO-related decreases in ECOG amplitude and increases in ECOG frequency in relation to UCO-induced FHR decelerations (FHR_dec) emerged. We have named it the ‘ECOG-FHR_dec locked-in pattern’.  This pattern was visually recognized by the onset of abrupt increases in ECOG 95% SEF values up to 23 ± 2 Hz during each UCO-induced FHR_dec followed by a decline to 3 ± 1 Hz (p < 0.001) between FHR_dec. This ‘ECOG-FHR_dec locked-in pattern’ was evident in all animals and occured at a pH of 7.24 ± 0.04 and base excess of -2.9 ± 2.2 mmol/L as determined from the most proximate blood sample 52 ± 13 minutes prior to reaching the target pH <7.00, at which the cord occlusion insults were stopped.

Study 2: 'Near-term versus pre-term groups' - analyses of CAP activity

Study design. Pre-term group: 16 pre-term fetal sheep (0.79 gestation) were chronically prepared with arterial catheters and placement of an inflatable umbilical cord occluder. Following 3 days post-op recovery and a baseline recording period, animals were subjected to 4 consecutive days of chronic intermittent hypoxia as presented in the PDF attached. RMSSD was calculated from ECG-derived FHRV.  Near-term group was instrumented and experimented upon as described above. In both experimental groups, in addition to cytokine ELISA tests, microglia immunohistochemistry was performed as described previously (see proof PDF of Prout AJOG 2009 in press attached).

Results. Pre-term group: At baseline of day 1, IL-1β concentration was 886 {449-1508} pg/ml and the short-term fHRV measure RMSSD averaged 6.4 ± 2.6 ms. RMSSD, but none of the long-term variability linear measures SDNN, HRVtri and LF band spectral power, correlated inversely to IL-1β (RMSSD: R = -0.70, p = 0.01). Of note, neither short-term nor long-term fHRV complexity measures correlated to IL-1β.  Chronic intermittent UCO over four days did not lead to changes in arterial blood gases or metabolites and in IL-1β levels. After the last UCO at day 4, IL-1β did not correlate to RMSSD or any other fHRV measures. We found no correlation of microglia counts in any brain region to baseline or day 4 post UCO6 levels of IL-1β and fHRV measures.

Near-term group: Repetitive UCO resulted in worsening acidosis over 3 to 4 h and eventually a severe degree of acidemia, fetal pH 7.36 ± 0.03 to 6.90 ± 0.13 (p < 0.05). RMSSD and IL-1β increased ~2fold from baseline versus at a time point coinciding with nadir pH (p < 0.05) and returned toward baseline levels within 1 h of recovery. Baseline (mildly acidotic fetuses with lactate > 1.5 mmol/L), nadir pH and 1 h of recovery values of RMSSD correlated to corresponding IL-1β levels at R = 0.57 (p = 0.02). RMSSD measurements at baseline and 1h of recovery correlated inversely to corresponding white matter microglia counts at R = -0.71 and R =  -0.89, respectively (both p<0.05).