Serum Cortisol and 17-Hydroxyprogesterone Interrelation in Classic 21 Hydroxylase Deficiency: Is Current Replacement Therapy Satisfactory?
 
Evangelia Charmandari, David R. Matthews, Atholl Johnston, Charles G D Brooke and Peter C. Hindmarsh
 
London Center for pediatric Endocrinology, University College London (E.C., C.G.D.B., P.C.H.), London WIT 3AA, United kingdom; Diabetes Research Laboratories, Radcliffe Infirmary (D.R.M.), Oxford OX2 6HE, United Kingdom; and Department of Clinical Pharmacology, St Bartholomew’s and the Royal London School of medicine and Dentistry (A.J.), London EC1M 6BQ, United Kingdom.
 
One of the main aims in the management of patients with congenital adrenal hyperplasia due to 21-hydroxylase deficiency is to archive adequate suppression of the adrenal cortex with the smallest possible dose of glucocorticoid substitution.  To evaluate the administration schedule of current replacement therapy regimens, we investigated the cotisol-17-hydroxyprogesterone interrelation in 36 patients (13 males and 23 females median age, 12.3 yr, range, 6.1-18.8 yr) with salt-wasting congenital adrenal hyperplasia.  As sufficient variation in 17-hydroxyprogesterone concentrations was required to allow analysis of the cortisol-17-hydroxyprogesterone interrelation, patients were divided into 2 groups depending on the adequacy of hypothalamic-pituitary adrenal axis suppression.  The first group consisted of 17 patients with suppressed 17-hydroxyprogesterone concentrations (group 1), and the second group consisted of 19 patients with nonsuppressed 17-hydroxyprogesterone concentrations (group 2).
 
 We determined serum cortisol and 17-hydroxyprogesterone concentrations at 20-min intervals for a total of 24 h while patients were receiving their usual replacement treatment with hydrocortisone and 9a fludrocortisone.  We also determined the lowest dose of dexamethasone required to suppress the 0800 h serum ACTH concentrations when administered as a single dose (0.3 or 0.5 mg/m2) the night before.
 
   Mean 24-h cortisol and 17-hydroxyprogesterone concentrations were 3.9 ug/dl (SD = 2.1) and 66.2 ng/dl (SD =92.7), respectively, in group 1 and 4.1 ug/dl (SD=2.5) and 4865.7 ng/dl (SD=6951) in group 2.  The 24-h 17 hydroxyprogesterone concentrations demonstrated circadian variation, with peak values observed between 0400-0900 h.  In group 2, 17-hydroxyprogesterone concentrations decreased gradually in response to the rise in cortisol concentrations between 1600-2000 h.  Mean 0800 h androstenedione concentrations correlated strongly with integrated 17-hydroxyprogesterone concentrations (r =0.81; P  < 0.0001), but not with integrated cortisol concentrations.  There was a significant negative correlation between  cortisol and 17-hydroxyprogesterone by these time intervals.  Finally, 0800 h serum ACTH concentrations were sufficiently suppressed after dexamethasone dose of 0.3 mg/m2 in all but three patients.
 
   These findings indicate that in classic 21-hydroxylase deficiency, hydrocortisone should be administered during the period of increased hypothalamic-pituitary-adrenal axis, between 0400-1600 h, with the biggest dose given in the morning.  Blood investigations performed as part of monitoring of congenital adrenal hyperplasia patients should include androstenedione and 17 hydroxyprogesterone concentrations determined in the morning before the administration of hydrocortisone.  It should be emphasized that blood investigations are only complementary to the overall assessment of these patients, which is primarily based on the evaluations of growth and pubertal progress.
 
   CONGENITAL ADRENAL HYPERPLASIA  (CAH) due to 21-hydroxylase deficiency is an autosomal recessive condition in which deletions or mutations of the cytochrome P450 21 hydroxylase gene cause glucocorticoid and often mineralocorticoid deficiency.  Substitution therapy with glucocorticoids is offered in an attempt to suppress the excessive secretion of CRH and ACTH by the hypothalamus and anterior pituitary, respectively, and to reduce the circulating concentrations of adrenal androgens and androgen precursors (1-3).  The preferred glucocorticoid option is hydrocortisone, because its short half-life minimizes growth suppression as well as other adverse side-effects of more potent, longer-acting glucocorticoids (4-10).
 
   Achieving and maintaining adrenal suppression in patients with classic 21 hydroxylase deficiency are far more challenging than preventing adrenal crisis and often require increased doses of glucocorticoid substitution, which may produce an unacceptable and undesirable degree of hypercortisolism (11).  Treatment efficacy reflects the adequacy of adrenocortical suppression and is assessed by monitoring annualized growth velocity, the rate of skeletal maturation, and the serum concentrations of 17-hydroxyprogesterone (17OHP) and androstenedione (1,12-15).
 
   Although the circadian variation in 17OHP concentrations in patients with CAH has been well documented (16-19), to our knowledge no information exists on the relationship between cortisol and 17OHP concentrations in these patients.  The aim of the present study was to determine the cortisol-17OHP interrelation in patients with classic CAH and to evaluate current replacement therapy in the light of these findings. 
 
                                                        Subjects and Methods
 
Subjects
 
   Thirty-six patients (13 males and 23 females; median age, 12.3 yr; range, 6.1 18.8 yr) with salt-wasting CAH attending the London center for pediatric Endocrinology were studied prospectively.  They consisted of 14 prepubertal (5 males and 9 females; median age, 9.6 yr; range, 6.1-11.0 yr; Tanner stage 1), 19 pubertal (7 males and 12 females; median age, 13.3 yr; range, 10.6-16.8 yr; Tanner stage lll-V), and 3 postpubertal (1 male and 2 females; median age, 18.4 yr; range, 18.1-18.8 yr) patients.  The clinical characteristics of all patients are shown in Table 1.
 
   All patients were receiving standard doses of replacement therapy in the form of oral hydrocortisone given twice (n=22) or three times (n=14) daily, with the largest portion of the daily dose given in the morning and 9a-fludrocortisone given once daily.  Patients were excluded from the study if there was evidence of central precocious puberty or their associated endocrine disorder.  No patient had clinical or biochemical evidence of hepatic or renal disease, and none was taking drugs known to alter corticosteroid-binding globulin concentrations or to induce mixed function oxidase enzymes.
 
   Because sufficient variation in 17OHP concentrations was required to allow analysis as well as interpretation of the cortisol-17OHP interrelated findings, patients were divided into two groups depending on the adequacy of hypothalamic-pituitary-adrenal (HPA) axis suppression, as defined by the 0800 h ACTH concentrations (<75 pg/ml) and serial measurements of 17OHP concentrations (<662 ng/dl) determined on the first day of the study.  The first group consisted of 17 patients (3 males and 14 females; median age, 12.8 yr; range, 6.2-18.8 yr) with nonsuppressed 17OHP concentrations (group 2).  Daily hydrocortisone and 9a-fludrocortisone doses were similar in the 2 groups of 8 patients (Table 1).   In group 1, hydrocortisone was given twice daily in 8 patients and 3 times daily in 9 patients.  In group 2, 14 patients were receiving hydrocortisone twice daily, and 5 patients were on a 3 times daily regime. 
 
   The study was approved by the university College London Hospitals committees on the ethics of human research.  Written informed consent was obtained in all cases from a parent, and assent was given by children older than 7 yr.
 
Methods
 
   Patients were admitted to the endocrine Unit, Middlesex Hospital, 1 d before the study, and standard anthropometric measurements, including pubertal Tanner staging (20,21), were obtained by a single trained observer.  An indwelling venous catheter for frequent blood sampling was inserted at least 12 h before sampling to allow a period of adaptation.  Patients were allowed normal ambulatory activity and received their oral hydrocortisone tablets at 0900 and 2100 h or at 0800, 1500 and 2200 h depending on whether they were on twice daily or three times daily hydrocortisone replacement.  9a-Fludrocortisone was given with the morning tablets.
 
   On the first day of the study baseline investigations, including ACTH, recumbent PRA, and androgen concentrations, were performed at 0800 h before administration of the hydrocortisone dose.  Blood samples for serum cortisol and 17 OHP were collected at 20-min intervals for a total of 24h from 0800 h.  Blood samples were centrifuged, separated, and stored at – 20C until assayed. 
 
   On the second and third days of the study, an overnight, single dose, dexamethasone suppression test was performed after randomization to two different doses of dexamethasone (0.3 and 0.5 mg/m2).  The dexamethasone dose (0.3 and 0.5 mg/m2 the following night, or vice versa) was administered instead of the evening hydrocortisone dose at 2200 h on d 2 and 3.  There was no alteration in the administration schedule of 9a-fludrocortisone.  A blood sample for measurement of ACTH concentrations was taken the following morning before the administration of hydrocortisone.  Blood samples were centrifuged and separated immediately after collection and were stored at –20C until assayed. 
 
   We arbitrarily divided the 24-h period into two 12-h intervals, from 0400-1600 h (daytime) and from 1600-0400 h (nighttime), because we expected maximal increases in 17OHP concentrations in the early morning hours.  Cortisol concentrations were expected to be maximal during the same period (0400-1600 h), because all patients had the largest hydrocortisone dose given in the morning.  In this way we could examine the main peaks of both cortisol and 17OHP in parallel.
 
 
Statistical analysis
 
24-h cortisol and 17OHP profiles.  Nonnormally distributed data were logarithmically transformed before statistical analysis.  Comparisons between two groups were performed using the t test.  The relationship between baseline 0800 h and androstenedione concentrations and integrated 17OHP and cortisol concentrations was investigated by linear regression and calculation of Pearson’s correlation coefficient.  Integrated 17OHP and cortisol concentrations were calculated as the area under the serum concentrations vs. time curve using the trapezoid method. 
 
Cross-correlation studies.  To search for a time-ordered relationship between cortisol and 17OHP, analysis of correlations between the absolute values of the time series of each hormone was computed at 20-min intervals at various time lags covering the 24-h period of the study.  Only data from the group of CAH patients with non-suppressed 17OHP concentrations were used for cross correlation studies because they demonstrated significant variation in 17OHP concentrations to allow such analysis.  All data were stationarized and processed as previously described (23,24).  Cross-correlation was computed after lagging (shifting) the concentration-time series of 17OHP relative to the concentration-time of cortisol.  If rt is the coefficient of correlation between cortisol and 17OHP at lag time t for one patient, then the mean r, for all patients was considered significant when it exceeded zero by more that 2 SE.  The SE was calculated from the individual values of rt for all patients at the lag time t. 
 
Single dose dexamethasone test.  The proportion of patients who achieved suppression of 0800 h ACTH concentrations after administration of the two different dexamethasone doses was determined and compared between the two groups of patients.
 
                                                                                      Results
 
24-h cortisol and 17OHP profiles
 
   The 24-h, daytime and nighttime mean concentrations of cortisol and 17OHP in both groups of patients are listed in table 2 and are depicted in Figs. 1 and 2.  There was no significant difference in mean 24-h cortisol concentrations between the two groups of CAH patients (group 1 vs. group 2, 3.9 = 2.1 vs. 4.1 = 2.5 ,ug/dl ).  Also, no difference in daytime (7.5 = 5.9 vs. 6.7 = 5.0 ,ug/dl) or nighttime (6.0 = 4.4 vs. 5.0 = 3.0 ,ug/dl) mean cortisol concentrations was noted between groups (Table 2).
 
As expected, group 2 patients had significantly higher mean 24-h 17OHP concentrations (4865.7 = 6951 vs.  66.2 =  92.7 ng/dl; P < 0.0001) as well as mean daytime (6653.1 = 9466.6 vs.  36.4 = 39.7 ng/dl; P < 0.0001) 17OHP concentrations than patients in group 1.   In both groups, mean 17OHP concentrations were higher during the daytime than during the nighttime; however these differences did not reach statistical significance. 
 
   Peak 17OHP concentrations were observed between 0400-0900 h in both groups of patients (Figs. 1 B and 2B).  In group 2, 17OHP concentrations decreased gradually in response to the rise in cortisol concentrations during the daytime, but remained low during the nighttime despite the almost undetectable cortisol concentrations observed between 1600-2000 h (Fig. 2B).
 
Androstenedione and integrated 17OHP and cortisol concentrations
 
   Mean 0800 h androstenedione concentrations were significantly higher in patients with nonsuppressed 17OHP concentrations than in patients with suppressed 17OHP (583.4 = 880.9 vs. 20.0 = 25.7 ng/dl; P < 0.0001).  There was a strong positive correlation between 0800 h androstenedione concentrations and integrated 17OHP concentrations (r = 0.81; P < 0.0001), but not between 0800 h androstenedione and integrated cortisol concentrations.
 
Cross correlation studies
 
   The graph depicting the first, second (median), and third quartile coefficients of correlation from the cross-correlation analysis over the 24-h period between cortisol and 17OHP values in the group 2 patients are shown in Fig. 3.  The analysis showed a significant negative correlation between cortisol and 17OHP at lag time 0 min (r =--0.187; P < 0.0001), peaking at lag time 60 min (r = --0.187; P < 0.0001).  In addition, a significant positive correlation was observed over time between cortisol and 17OHP, peaking at lag time 6 h, 40 min (r = 0.359; P < 0.0001), with cortisol leading 17OHP by these time intervals. 
 
   When data were analyzed after dividing the 24-h profiles into daytime and nighttime 12-h profiles: There was a significant negative correlation between daytime cortisol and 17OHP concentrations (r = --0.586; P < 0.0001), with cortisol leading the 17OHP by these time intervals.  Also, there was a significant negative correlation between nighttime cortisol and 17OHP concentrations (r=- 0.436; P <0.001) at lag time 0 min, which reached a peak at lag time 20 min (r = -0.447; P <0.001).  A significant positive correlation was noted over time between daytime and nighttime cortisol and 17OHP concentrations, which reached a peak at lag time 2 h, 20 min and 3 h, respectively, with cortisol leading 17OHP by these time intervals. 
 
Single dose dexamethasone suppression test 
 
   In group 1, all patients achieved significant suppression of the 0800 h ACTH concentrations after either dose of dexamethasone.  In group 2, three (15.8%) patients did not demonstrate sufficient (<75 pg/ml) suppression of the 0800 h ACTH concentrations after 0.3 mg/m2 dexamethsone, and one (5.3%) did not show sufficient suppression even after administration of the higher dexamethasone dose of 0.5 mg/ m2.  Two of those patients had extremely high baseline ACTH concentrations (856 and 506 pg/ml).
 
                                                                                       Discussion
 
   It has long been recognized that patients with CAH due to 21-hydroxylase deficiency demonstrate circadian variation in 17OHP concentrations.  Measurement of 17OHP concentrations at 4-h intervals for 24-h before treatment or during inadequate treatment showed that peak concentrations were observed early in the morning, whereas trough concentrations were documented in the evening (16-19).  The circadian variation persists after introduction of glucocorticoid therapy, although the of therapeutic control treatment.
 
    The present study is the first to delineate in such detail the pattern of 17OHP secretion in patients with classic 21-hydroxylase deficiency and the first to elucidate the relationship between cortisol and 17OHP in these patients.  The later is illustrated better in the second group of patients whom variations in 17OHP concentrations are observed in relation to circulating cortisol concentrations achieved after administration of oral hydrcortisone.  The extremely high 17OHP concentrations early in the morning gradually declined as cortisol concentrations began to rise and remained low for as long as cortisol concentrations remained elevated, indicating that during daytime (0400-1600 h) cortisol concentrations need to be sufficiently high to result in adequate sppression of the HPA axis activity, as reflected by the magnitude of 17OHP concentrations.  On the other hand, mean 17OHP concentrations during the highttime (1600-0400) remained low despite the almost undetectable cortisol concentrations between 1600-2000 h, suggesting a period of decreased activity of the HPA axis.  Finally, 17OHP demonstrated a precipitous rise beyond 0400 h, when increased activity of HPA axis is observed and by which time a significant reduction in circulating cortisol concentrations had taken place.
 
   These observations concur with previous reports on the circadian rythym of 17OHP secretion (16-19) and indicate that administration of oral hydrocortisone in the evening may only be useful for substitution purposes and should not be expected to contribute significantly toward achieving and/ or maintaining adrenocrortical suppression.  Therefore, it may be more appropriate, although not practical, to administer a dose of hydrocortisone just before  the time o the rapid rise in 17OHP concentrations early  in the morning (25).
 
   The cross-correlation studies between cortisol and 17OHP concentrations was achieved 1 h after cortisol concentrations reached their peak, and this effect was reversed 6 h, 40 min later.  That the peak negative correlation between the two hormones was pbserved earlier during the nighttime (at lag time 20 min) than during the daytime (at lag time 60 min) suggests that 17OHP suppression is achived sooner after hydrocortisone administration during the nighttime, and provides further evidence of decreased HPA axis activity between 1600-0400 h.  Thus, the pattern of 17OHP secretion and the cortisol-17OHP interrelation along with knowledge of the main pharmokinetic properties of the oral hydrocortisone formulation used (26,27) may allow us to make appropriate schedule of hydrocortisone replacement therapy. 
 
   The finding of the present study also provide useful information about the timing of blood sampling for measurement of sndrogens and androgen precursors, which, along with clinical evaluation and monitoring of the annualized growth velocity and the rate of skeletal maturation, are often performed as part of the assessment of patients with classic 21-hydroxylase deficiency.  Given the pattern of 17OHP concentrations described above, single measurements of 17OHP concentrations may be unreliable and can be particularly  misleading if blood specimens are obtained beyond 1600 h.  Serum androstenedione concentrations, on the other hand, obtained early in the morning and before the administration of oral hydrocortisone, correlate strongly with integrated 17OHP concentrations and can be used as a reliable marker of the adequacy of adrenocortical suppression if a single blood sample is to be obtained.  In patients in whom a more detailed assessment is considered necessary,as, for example, in the case of inadequate adrenocortical suppression despite substitution therapy and adherance to treatement measurement  of cortisol and 17OHP concentrations every hour for the first 4 h after administration of hydrocortisone and every 2 h thereafter until the next dose is to be given will provide adequate information about the cortisol-17OHP dynamics and will enable physicians to adjust their treatment accordingly.  This detailed assessment can also be performed without necessitating admission to hospital, as serial measurements of blood or salivary 17OHP concentrations have been shown to be particularly useful in monitoring patients with classic 21-hydroxylase deficiency and can be obtained easily  at home (13-15, 28-30).
 
   Despite the fact that both groups of patients were receiving similar hydrocortisone and 9a-fludrocortisone doses and displayed similar 24-h cortisol concentrations, there were significant differences in ACTH, 17OHP, and androstenedione concentrations between them.  This might relate to the differences in the severity of salt wasting CAH and/ or alterations in feedback mechanisms at the hypothalamic/pituitary level,  To investigate the latter, we performed a single dose dexamethsone suppression tests after randomization to two different doses.  The smaller dexamethasone dose 0.3mg/m2, recommended for th single dose dexamethasone suppression tests (31), sufficiently suppressed the 0800 h ACTH  in all but three patients, who had very elevated baseline ACTH concentrations and required a higher dose (0.5 mg/m2).  Based on those findings, one may suggest that if a few-day course dexamethasone is required to achieve suppression of HPA axis poorly controlled CAH patients, the dose of choice should be 0.5 mg/m2 on the first day and o.3 mg/m2 on subsequent days.  This represents the dose required  to achieve suppression of HPA axis in and not the daily requirements in terms of glucocorticoid substitution.  However, it should be emphasized that longer –term administration of dexamethasone should be avoided in view of its well documented detrimental  effects on growth and the associated adverse effects of longer-acting glucocorticoids (32-35).
 
   In summary, in patients with classic 21 hydroxylase deficiency, hydrocortisone replacement therapy should be administered during the period of increased HPA activity, between 0400-1600  h.  Blood investigations performed as part of monitoring of CAH patients should include androstenedione and 17OHP concentrations obtained in the morning before oral dose of hydrocortisone is given.  However, it should be emphasized that blood investigations are only complementary to the overall assessment of these patients, which is primarily based on the evaluation of growth and pubertal progress.