A very interesting article in the PNEJM that made me question a lot of my ideas; I always like that when it happens. Boonen E, Vervenne H, Meersseman P,  et al: Reduced Cortisol Metabolism during Critical Illness. Prestigious New England Journal of Medicine 2013, 368(16):1477-1488. The authors have performed a sequence of elegant tests in critically ill adults, which show 1. as we already knew cortisol levels are much higher when you are critically ill. 2. as I certainly didn’t know, this is not mostly due to the adrenal glands making more cortisol, it is actually mostly due to decreased metabolism of cortisol. They produce several lines of evidence for this, but it looks robust to me. In fact ACTH was reduced during serious illness, probably secondary to the increase in cortisol concentrations.

I think this is potentially very important for our (or maybe it is just my) understanding of what to do with steroids in the critically ill.

As the authors note, if you give a ‘physiologic replacement dose’ of hydrocortisone to someone who has markedly reduced metabolism, you are probably giving much too much (3 times too much they calculate).

So the authors seem to think that what happens is this, in critical illness there is a transient increase in ACTH, followed by an increase in cortisol, which inhibits further ACTH release, and then other mediators which are being released, such as cytokines, inhibit the breakdown of cortisol. So the cortisol stays high.

We have long known that patients, including preterm infants, who have lower cortisol responses have worse outcomes than those with brisk responses. We have responded to that by studying replacement of cortisol, which has had very mixed success. The latest large adult studies showed no real overall benefit of low dose steroid replacement in the adult (or indeed in the ventilated preterm).

It now looks to me that we don’t understand all these factors well enough, we certainly don’t understand them in the preterm, so we need to rethink our approach.