Kapadia VS, Chalak LF, Dupont TL, Rollins NK, Brion LP, Wyckoff MH: Perinatal Asphyxia with Hyperoxemia within the First Hour of Life Is Associated with Moderate to Severe Hypoxic-Ischemic Encephalopathy. The Journal of pediatrics 2013.
If you are acidotic at birth, and then you become hyperoxic, it seems that you are more likely to have more severe encephalopathy than if hyperoxia is avoided. This is the kind of thing that you can only ever find from observational studies. No-one would suggest a prospective RCT; even though this kind of study is often criticized, this is a perfect example of why such things should, at least sometimes, be published.
If you develop HIE, does magnesium supplementation help? This systematic review suggests that maybe it does, but maybe it increases mortality; the numbers of babies really tested for this are tiny, so we can have almost no confidence in any statement about either benefit or risk. Magnesium therapy has also never been tested since we started cooling babies, so there is room for another study. This is something that could be tested in low resource countries, whereas xenon, which is much more sexy, will never be tried in those places.
One of the ‘biomarkers’ of the severity of asphyxia is the count of nucleated RBCs. The more severely asphyxiated you are the more your bone marrow (in general terms) will be stimulated to release red cells which are are not yet mature. This article from the brain research group in Cork, Ireland shows that this ‘biomarker’ of the severity of brain injury is no longer reliable in infants treated with therapeutic hypothermia. It points out what I think should be a general rule, biomarkers may be unreliable if therapies change.
I understand the desire to find ‘biomarkers’ (what we used to call surrogate outcomes) we have to find ways to make clinical research easier and cheaper and produce results without waiting for years of follow up, but we have to be extremely careful before basing changes in clinical management on a demonstrated change in ‘biomarkers’ without proof that real clinically important outcomes are affected. Even when we think that a specific surrogate outcome is reliable, then it has often been demonstrated that things which improve the surrogate don’t necessarily improve the real, clinically important outcomes. We have been down this path before (we just didn’t call them ‘biomarkers’ before). Hypertension, for example, is reliably decreased by beta-blockers, but they have no effect on other clinically important outcomes, such as stroke.