LAUSR

Unauthorized reproduction of this article is prohibited Pediatric Critical Care Medicine www.pccmjournal.org 777 The ongoing controversy about possible adrenal insufficiency in infants and children following cardiopulmonary bypass (CPB) and heart surgery has been swirling with little resolution for years. Indeed, Flores et al (1) recently published an international survey of pediatric cardiac intensivists demonstrating a shocking lack of consensus of opinion on this topic. Given that most centers administer large corticosteroid doses to patients prior to and/or during CPB, we do not know if the derangements reported in the hypothalamic-pituitary axis (HPA) in these patients reflects true critical illness– induced adrenal insufficiency or a predictable suppression of the HPA from the exogenous steroids (2–5). So, if measuring cortisol level or performing adrenocorticotropic hormone stimulation testing, is difficult to interpret, is there another approach to solve this conundrum? In this issue of Pediatric Critical Care Medicine, Flores et al (6) report their analysis of measuring, by flow cytometry, levels of glucocorticoid receptors (GCRs) in peripheral WBCs of patients before and after CPB. In a certain light, this is a very creative approach to further characterize the HPA in these patients and may provide, at the least, more insight than simply measuring cortisol levels again and again. They found, somewhat surprisingly (see further discussion below why this is surprising), that rising GCR levels in these patients’ WBCs seemed to correlate with a longer time to weaning off vasoactive support. Paradoxically, they also identified higher GCR levels were associated with lower severity of low cardiac output syndrome (LCOS) in this cohort. So GCR levels correlated with higher severity of illness as reflected by the time to wean off inotropic support (TIF) but also with lower LCOS levels? Let us try to unravel this apparent mystery. There is evidence that alterations of WBC GCR production, location, and sensitivity are correlated with critical illness. Cohen et al (7) studied GCR messenger RNA (mRNA) expression in septic adults versus controls and in vitro dexamethasone suppression of lipopolysaccharide-stimulated leukocyte cytokine production. GCR mRNA expression was increased dramatically in the patients compared with the controls, but sensitivity to dexamethasone was not different. van den Akker et al (8) found a transient “decrease” in GCR mRNA in neutrophils of septic children. Likewise, Shibata et al (9) found critically ill children with higher severity of illness, and shock had lower GCR expression in CD4 and CD8 cells. But just measuring GCR levels may not provide a complete picture. Indyk et al (10) demonstrated that where the receptor is matters. They compared cytosolic with nuclear GCRs in WBCs of controls and patients with septic shock and traumatic brain injury; critically ill children had lower cytosolic GCR levels but equal nuclear levels compared with the controls. If GCR levels even vary within a cell, how do peripheral WBC GCR levels relate to cardiac surgery and why should we care? Presumably, in our postoperative cardiac surgery patients, the GCR pool we care about is in the vasculature or cardiac myocyte, not WBCs. We could find no studies that convincingly correlate GCRs found in WBCs with those of the endothelium, vascular smooth muscle, or cardiac myocytes. Are we looking under the lamp post (at WBCs) because that is where the light is; that is, they are readily accessible to us? Furthermore, not all GCRs are created equal. Two pediatric studies have analyzed the genotypes of the GCRs in critically ill children. Cvijanovich et al (11) genotyped 482 children with septic shock and found one particular genotype (homozygous for wild type) was associated with worse outcomes when steroids were given. Jardine et al (12) found that in 92 critically ill children, a certain single nucleotide polymorphism in the MC2R gene strongly associated with low cortisol response (the AA genotype was seven times more likely to be associated with lower free cortisol compared with the GG genotype). Therefore, it is unclear whether just quantifying GCRs in peripheral leukocytes is the right metric. *See also p. 705.