I once wrote a paper with the last author of this article, when we were both working in Edmonton, so I will have to be nice, I guess. Seriously, we do need to know a lot more about adrenal responses to being born very preterm, whether relative adrenal insufficiency exists (I think it probably does) how to define it (I have no idea, which makes the first answer a bit redundant) how to diagnose it (nope, no idea here either) and whether it helps to treat it (not a clue).

These investigators from Vancouver have made a decent attempt to find some more answers.  Hochwald O, Holsti L, Osiovich H: The use of an early acth test to identify hypoadrenalism-related hypotension in low birth weight infants. J Perinatol 2012, 32(6):412-417. They performed ACTH stimulation tests on 40 babies from 24 to 29 weeks gestation, in the first 8 hours of life and then followed them to see if they got treated for hypotension. They also performed echos to look at cardiac function before treatment for low BP, or at 8 hours if the babies were not treated.
They noted that their practice is to only treat infants for “hypotension” if they have a mean BP less than the GA and clinical signs and symptoms of decreased capillary perfusion, and lack of response to a 10 ml/kg bolus. That is similar to what I would describe that I do, but they had 30% of the infants getting treated for hypotension, whereas I treat far fewer than 30% in the first 8 hours of life. Which points out only that some of these decisions are rather subjective, and the diagnosis of hypotension needing treatment is uncertain.

They found that babies who had less than a 12% rise in their cortisol with the test were more likely to get treated with inotropes, and that cutoff was relatively specific for getting treated (93%), with a 75% sensitivity.

They also measured left ventricular output by echo at the time of treatment, or if not treated at 8 hours of age. I am not sure why they chose to measure LVO; in this group of babies the ductus arteriosus will be open almost 100% of the time at these ages, so LVO is critically dependent on ductal shunting, and is likely to change rapidly as pulmonary vascular resistance falls, so if the non-treated group were studied a little later (as they will be by design, being studied at 8 hours of age) the ductal shunt may be larger, and thus give a higher LVO. Indeed LVO was higher in the non-treated babies: I must say though that in the treated babies, after the fluid bolus but before the inotrope, the LVOs were quite low, 66 to 120 mL/kg/min. That is very low, so probably a truly low systemic flow. Compared to between 135 and 266 ml/kg/min in the untreated babies at 8 hours of age. That very large difference is probably not all due to differences in ductal shunt, but I still don’t understand why they didn’t measure RVO instead.

When they analyzed the LVO by whether the ACTH test had a low response or a high response, there was much more overlap. Low ACTH response LVO = 118 (range 66–202) high response LVO= 179 (range 90–266) ml/kg/min. So the means were significantly different, but there is not good discrimination between the groups.

So this study suggests that having a low ACTH response makes you more likely to get treated for hypotension accompanied by poor perfusion, and is associated with a variably lower LVO.

The article is accompanied by an editorial (Aucott SW: The challenge of defining relative adrenal insufficiency. J Perinatol 2012, 32(6):397-398) which points out the difficulties in defining relative adrenal insufficiency, and the lack of good evidence to guide management, including steroid replacement.