Decreased Adrenomedullary Function in Infants With Classical Congenital Adrenal Hyperplasia

Abstract

Context:

Classical congenital adrenal hyperplasia (CAH) due to 21-hydroxylase deficiency can cause life-threatening adrenal crises as well as severe hypoglycemia, especially in very young children. Studies of CAH patients 4 years old or older have found abnormal morphology and function of the adrenal medulla and lower levels of epinephrine and glucose in response to stress than in controls. However, it is unknown whether such adrenomedullary abnormalities develop in utero and/or exist during the clinically high-risk period of infancy and early childhood.

Objective:

The objective of the study was to characterize adrenomedullary function in infants with CAH by comparing their catecholamine levels with controls.

Design/Settings:

This was a prospective cross-sectional study in a pediatric tertiary care center.

Main Outcome Measures:

Plasma epinephrine and norepinephrine levels were measured by HPLC.

Results:

Infants with CAH (n = 9, aged 9.6 ± 11.4 d) had significantly lower epinephrine levels than controls [n = 12, aged 7.2 ± 3.2 d: median 84 [(25th; 75th) 51; 87] vs 114.5 (86; 175.8) pg/mL, respectively (P = .02)]. Norepinephrine to epinephrine ratios were also significantly higher in CAH patients than controls (P = .01). The control infants had primary hypothyroidism, but pre- and posttreatment analyses revealed no confounding effects on catecholamine levels.

Conclusions:

This study demonstrates for the first time that infants with classical CAH due to 21-hydroxylase deficiency have significantly lower plasma epinephrine levels than controls, indicating that impaired adrenomedullary function may occur during fetal development and be present from birth. A longitudinal study of adrenomedullary function in CAH patients from infancy through early childhood is warranted.

Classical congenital adrenal hyperplasia (CAH) due to 21-hydroxylase deficiency is the most common cause of primary adrenal insufficiency in children, affecting 1:10 000–25 000 newborns (1). In this condition, there is deficient cortisol and aldosterone production, potentially leading to life-threatening adrenal crises. Although typically thought of as a disease of the adrenal cortex, CAH also affects the morphology and function of the adrenal medulla in children as young as 4 years old (2–4). Normal development of the medulla, the major site of production of epinephrine, depends on exposure to cortisol produced from the cortex, with important crosstalk occurring between cells of the cortex and medulla (5, 6). From early fetal life onward, intraadrenal cortisol induces phenylethanolamine N-methyltransferase, the enzyme responsible for converting norepinephrine to epinephrine in the chromaffin cells of the medulla (7, 8). Epinephrine and cortisol are both part of the stress response and, as counterregulatory hormones, also help to prevent hypoglycemia.

Lower levels of epinephrine in CAH children correlate with increased rates of hospitalization for adrenal crises (2). When subjected to the stress of moderate- to high-intensity exercise, CAH adolescents with a partial epinephrine deficiency show lower epinephrine and glucose levels as a physiological response than controls (9, 10). Like children with CAH, the 21-hydroxylase-deficient mouse exhibits severely impaired catecholamine production, as well as structural abnormalities of the medulla and cortex, as early as 1 week of age (11). Thus, both human and animal studies suggest that CAH children are likely to have abnormal adrenomedullary development from an early age. The neonate has a physiological catecholamine surge in response to the stress of normal delivery and exposure to cold. Although the surge consists predominantly of norepinephrine, epinephrine is known to be the major regulator of gluconeogenesis and glycogenolysis in the counterregulatory response to hypoglycemia in newborns (12). Further studies are needed to understand adrenomedullary morphology and function in very young children, especially those with CAH, who are at greater risk for seizures, cognitive impairment (13–15), and increased mortality (16) due to severe hypoglycemia associated with illness. Therefore, we undertook the first study of plasma catecholamine levels in infants with and without CAH. We hypothesize that infants with CAH, due to impaired fetal adrenal development, have lower epinephrine levels than age-matched controls.

Study Participants and Methods

The study was cross-sectional and approved by the Children's Hospital Los Angeles Institutional Review Board. Parents of infants gave written informed consent.

Nine infants with CAH were enrolled. CAH was diagnosed based on positive newborn filter paper screening (California Department of Health Services) with elevated confirmatory serum 17-hydroxyprogesterone (17-OHP) level, increased serum androgen and plasma renin levels, hyperpigmentation, and ambiguous genitalia in eight females. Genotyping was not performed due to limitations on blood volume (2 mL/kg) with the clinically indicated blood draw. Our control group consisted of infants with congenital primary hypothyroidism (n = 12), diagnosed by a positive newborn screen (elevated TSH) and confirmatory serum testing, with no other medical problems. Of note, our institutional review board did not allow blood draws in completely healthy infants unless clinically indicated. Participants were recruited at the pediatric endocrinology clinic or ward. Both groups consisted of full-term infants (except for one 36.3 week old CAH baby), appropriate for gestational age, with no statistically significant difference in age, weight, or length (birth/study visit), head circumference, or vital signs (Table 1). There was a female preponderance in both groups; all CAH females had ambiguous genitalia (12% Prader 2, 25% Prader 3, and 63% Prader 4). Four CAH infants had mild electrolyte disturbances (Na 132.5 ± 1.3 mEq/L, K 6.23 ± 0.9 mEq/L) but were not acutely ill. One CAH baby diagnosed at age 1 week had a congenital heart defect repair at age 2 weeks (3 wk prior to the study visit) but was not acutely ill or on medications known to interfere with the catecholamine assay.

Table 1.

Clinical Characteristics of Newborns With CAH due to 21-Hydroxylase Deficiency and Controls

	CAH	Control
n	9	12
Gender (female)	8	12
Gestational age, wka	38.9 ± 0.5	39.6 ± 0.4
Birth weight, kga	3.27 ± 0.51	3.52 ± 0.47
Birth length, cma	50.06 ± 1.74	50.11 ± 1.97
Age at visit, da	9.56 ± 11.35	7.17 ± 3.16
Visit weight, kga	3.17 ± 0.47	3.56 ± 0.42
Visit length, cma	50.04 ± 2.14	49.63 ± 1.58
Head circumference, cma	34.8 ± 1.75	35.27 ± 3.23
Heart rate, beats/mina	130.2 ± 14.69	130.9 ± 19.32
Systolic blood pressure, mm Hga	78.63 ± 10.82	89.58 ± 17.5
Diastolic blood pressure, mm Hga	48.00 ± 5.2	55.50 ± 9.15
Newborn screen 17-OHP, ng/dLb	6256 (4708; 14899)	<30

^aNo significant difference between groups was seen (mean ± SD).

^bMedian with interquartile range (25th; 75th percentiles); to convert 17-OHP to SI units, multiply by 0.0302 = nanomoles per liter.

All infants had a history and physical examination, including length, weight, head circumference, heart rate, and blood pressure measurements. Plasma catecholamines and serum cortisol were drawn in the supine position at the time of a clinically indicated blood test. Four participants had cortisol treatment initiated prior to the study visit.

In addition to the main cross-sectional comparison, longitudinal follow-up of the control group was performed to assess hypothyroid infants after treatment with levothyroxine (LT4; 10–12 μg/kg·d). They returned for follow-up in 48 ± 19 days for a history, physical examination, and repeat blood tests.

Fractionated plasma catecholamine levels were quantified by HPLC (Quest Diagnostics Nichols Institute). Interassay and intraassay coefficients of variation were less than 5%. The limit of detection was less than 40 pg/mL with normative ranges derived from published data (17). Serum cortisol was quantified by liquid chromatography/tandem mass spectrometry (Quest Diagnostics Nichols Institute).

Statistical analysis

Patient characteristics of CAH and control infants were compared using an independent t test or χ² test for continuous and categorical characteristics, respectively. Descriptive statistics and comparisons of demographics were used. Given the small sample size, Mann-Whitney U tests (Tukey box and whisker plot) were used to compare catecholamine levels between infant groups to avoid undue influence of potential outliers. Wilcoxon signed-rank tests were used to compare visits 1 and 2 within the control group. Data are reported as medians (25th percentile; 75th percentile) and were analyzed using SPSS (V.21; IBM Corp) (for all analyses, α = .05).

Results

In the CAH group, the median newborn screen 17-OHP level was 6256 ng/dL (4708; 14 899), with confirmatory 17-OHP of 8149 ng/dL (1856; 20 544) (SI unit conversion × 0.0302 = nanomoles per liter); T 52 ng/dL (21; 585) (SI unit conversion × 0.0347 = nanomoles per liter); and androstenedione 258.5 ng/dL (83.3; 11 211) (SI unit conversion × 0.0349 = nanomoles per liter). There was no correlation between epinephrine and newborn screen 17-OHP (P = .6), confirmatory 17-OHP (P = .7), T (P = .6), or (17-OHP + androstenedione)/cortisol (P = .7) prior to treatment. The CAH patients were not sick or in distress at their study visit and had cortisol 2.1 μg/dL (1.2; 5.7) (SI unit conversion × 27.6 = nanomoles per liter).

Control infants were not in distress at either visit [cortisol visit 1: 7.5 μg/dL (3.0; 9.3) vs visit 2: 3.0 (1.5; 6.0), P = .2] and were no longer hypothyroid by visit 2 on LT4 treatment [visit 1: free T₄ 0.8 ng/dL (0.2; 2.0) (reference 2.2–5.3) and TSH 160 μU/mL (75.2; 230.5) (reference 3.2–35); visit 2: free T₄ 4.0 ng/dL (1.4; 4.7) (reference 1.6–3.8 < 2 wk, 0.9–2.2 ng/dL > 2 wk), and TSH 1.2 μIU/mL (0.5; 10.8) (reference 1.70–9.10 < 1 mo, 0.80–8.20 ≥ 1 mo); SI unit conversion × 12.87 = picomoles per liter].

The main finding of this study was significantly lower plasma epinephrine levels in CAH infants than controls [Figure 1A; 84 (51; 87) vs 114.5 (86; 175.8) pg/mL, respectively, P = .02]. Plasma norepinephrine levels did not differ between groups [Figure 1B; CAH 1031 (719; 1187) vs controls 836 (774; 1145) pg/mL, P = .6], although the norepinephrine to epinephrine ratio was significantly higher in CAH infants than controls [Figure 1C; CAH 12.3 (9.1; 16.7) vs controls 7.0 (5.8; 8.8), P = .01].

Figure 1.

Baseline plasma catecholamine concentrations in newborns with classical CAH due to 21-hydroxylase deficiency compared with control newborns. A, Epinephrine. B, Norepinephrine. C, Norepinephrine to epinephrine ratio. Data are shown as box (median and interquartile range 25th to 75th percentiles)-and-whisker (Tukey method) plots. *, P < .05. To convert to SI units, multiply epinephrine by 5.454 = picomoles per liter, and norepinephrine by 5.914 = picomoles per liter.

In controls, a longitudinal analysis showed no significant changes between visits 1 and 2 in either epinephrine [114.5 (86; 175.8) vs 100 (87; 119) pg/mL, P = .8] or norepinephrine levels [836 (774; 1145) vs 1109 (838.8; 1410) pg/mL, P = .2].

Discussion

This study demonstrates that infants with CAH have significantly lower plasma epinephrine levels than age-matched controls, which carries significance for prediction of disease severity at an early stage in life (18). When stressed, infants and toddlers with CAH are at high risk for morbidity and mortality from adrenal crises and severe hypoglycemia, with epinephrine deficiency possibly contributing to further risk (19). There is a paucity of studies on catecholamine levels in infants, and to our knowledge, there are no published studies on catecholamine levels in infants with CAH.

A normal adrenal cortex is necessary for adrenomedullary development and function in the fetus (5–8), and although older patients with CAH have an adrenal medulla that is morphologically and functionally abnormal (2, 9, 10), it is not known when this adrenomedullary disruption begins. Based on our finding of decreased plasma epinephrine levels in CAH infants shortly after birth, it is reasonable to speculate that abnormal adrenomedullary development may begin in utero and is most likely due to intraadrenal cortisol deficiency, limiting the survival and function of the developing medullary chromaffin cells (5, 8).

We and others have found that, although plasma epinephrine levels are low in CAH, norepinephrine is not affected (2, 9, 10), likely due to reduced phenylethanolamine N-methyltransferase activity in the adrenal medulla (7, 8), but also because norepinephrine is primarily derived from sympathetic nerve endings (9). A large CAH study has shown increased plasma norepinephrine compared with controls (3), consistent with increased production of norepinephrine from the sympathetic nervous system (SNS), as occurs in patients with acquired adrenal insufficiency (20). Given the complexity of the stress system [SNS, hypothalamic-pituitary-adrenal cortex, and sympathoadrenal (medulla) system], the SNS could overcompensate for impaired adrenomedullary function through the shared activation of the adrenal medulla and SNS by sympathetic neurons (21). In support of this concept, our study showed a greater norepinephrine to epinephrine ratio in infants with CAH than controls, although it is possible that this is more of a reflection of decreased epinephrine levels. The sympathoadrenal system helps adapt to stress, involving close communication with the adrenocortical system (22), and perhaps playing an even more important role during fetal and infant life than during later years (11, 12). Therefore, more studies are needed to understand how the impaired sympathoadrenal and adrenocortical components of the stress system affect the SNS and the ability of the young CAH patient to adapt to stress.

To address the potential limitation posed by congenital hypothyroidism in the control infants, we measured catecholamine levels before and after treatment with LT4. According to studies in older patients, hypothyroidism increased norepinephrine levels (23) but did not significantly affect epinephrine (24, 25). In our controls with congenital hypothyroidism, there were no differences in catecholamine levels between the hypo- and euthyroid states, thus eliminating a potential confounding effect of hypothyroidism on catecholamine levels.

Although the small cohort size is an obvious limitation, catecholamine studies in older CAH children have used similar sample sizes (9, 10) and show similarly significant findings, implying sufficient power with these sample sizes. We surmise that the 18% difference in basal epinephrine levels between CAH and control infants would become increasingly relevant during times of high metabolic stress, such as severe illness (10). Four of our subjects were on hydrocortisone treatment for a brief time prior to catecholamine collection. However, our subanalysis of subjects before and after treatment showed no difference in catecholamines (data not shown). Physiological replacement of glucocorticoids and even typical supraphysiological stress doses do not appear to affect epinephrine or glucose measurements in patients with CAH (26), and catecholamine studies in older CAH patients have been conducted with subjects on hydrocortisone treatment (9, 10, 26). Thus, it is reasonable to conclude that hydrocortisone would not have a significant effect on epinephrine levels in subjects already started on replacement therapy. Due to the limited amount of blood that could safely be obtained from infants (2 mL/kg), metanephrine and normetanephrine levels, and CYP21A2 gene analysis to assess phenotype-genotype correlation, were not performed.

We conclude that infants with CAH have impaired adrenomedullary function that may affect their early postnatal health. This study begins a needed exploration of the stress response in young patients with adrenal insufficiency to develop a more thorough understanding of catecholamine physiology in these patients, particularly during times of severe illness. A larger prospective, longitudinal study of catecholamine levels and their metabolites in CAH patients from infancy is warranted to explore potential changes in adrenal function with age and in response to stress.