This Article Abstract Full Text (PDF) Submit a related Letter to the Editor Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Streeten, D. H. P. Articles by Jones, C. Search for Related Content PubMed PubMed Citation Articles by Streeten, D. H. P. Articles by Jones, C.

Suppressibility of Plasma Adrenocorticotropin by Hydrocortisone: Potential Usefulness in the Diagnosis of Cushing’s Disease¹

Section of Endocrinology, Diabetes and Metabolism, Department of Medicine, State University of New York Health Science Center, Syracuse, New York 13210

Address all correspondence and requests for reprints to: David H. P. Streeten, Department of Medicine, SUNY Health Science Center, 750 East Adams Street, Syracuse, New York 13210.

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Repeatedly normal cortisol suppressibility by dexamethasone in 2 patients with Cushing’s disease led to the present study of the prevalence of this phenomenon in 58 patients with otherwise incontrovertible evidence of Cushing’s disease. Because as many as 23% of these patients manifested this phenomenon, we investigated the suppressibility of plasma ACTH: 1) during iv infusion of hydrocortisone, after a priming dose (7 mg), at 3 mg/h in 8 patients and 8 normal controls; and 2) for 2 h, after oral hydrocortisone, 0.25 mg/kg, in 13 patients and 16 controls. The data showed invariable suppression of plasma ACTH to 10 pg/mL (2.2 pmol/L) after 120 min of the infusion or at 90 min after oral hydrocortisone in 16 fasting normal subjects given oral hydrocortisone between 0800 and 0830 h. Plasma ACTH exceeded 10 pg/mL (2.2 pmol/L) at the same times in 14/14 patients with active Cushing’s disease, including 3 patients whose cortisol suppressibility by dexamethasone had been misleadingly normal and in 4/7 patients with intermittent hypercortisolism. Occasional variations in plasma cortisol elevations after the oral dose require that plasma cortisol concentration be monitored at 60 min after the oral hydrocortisone dose, because the present evidence supports the validity of the conclusion that a plasma ACTH concentration below 10 pg/mL excludes Cushing’s disease only when plasma cortisol concentration at 60 min lies between 16 and 38 µg/dL. Further evaluation of ACTH suppressibility by cortisol would be worthwhile, to confirm its potential value in facilitating positive diagnosis of Cushing’s disease when dexamethasone suppressibility seems misleading.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
MEASUREMENT of the suppressive action of administered dexamethasone on the excretion of urinary 17-hydroxycorticosteroids (17-OHCS) was introduced by Liddle in 1960 (1) as a diagnostic test for Cushing’s syndrome. This procedure and its modifications involving measurements of urinary free cortisol (2) and of plasma cortisol concentration after overnight dexamethasone administration (3) have been simple, frequently very useful, and widely used tests for the diagnosis of hypercortisolism ever since. However, most endocrinologists have probably encountered a minority of patients, said to be very rare (4), with obvious clinical features of Cushing’s syndrome, and elevated urinary excretion, and/or plasma concentrations of cortisol, yet with normal suppressibility of plasma cortisol concentration or urinary cortisol excretion after low-dose dexamethasone administration. This confusing finding is probably usually attributable to an excessively prolonged half-life of dexamethasone caused by retarded metabolism or excretion of this steroid (5). Repeatedly normal overnight, low-dose dexamethasone suppression tests in two patients with otherwise unequivocal evidence of persistently elevated plasma cortisol concentrations (measured every 2 h for 24 h) and urinary 17-OHCS and cortisol excretion (Table 1) led us to determine the prevalence of such conflicting results in 58 patients with clinically obvious Cushing’s disease, strongly supported by other steroid evidence of hypercortisolism. The prevalence of confusingly normal evidence of steroid suppressibility by dexamethasone, as high as 23% for some measures of this phenomenon, raised the possibility that plasma ACTH suppression by iv or oral hydrocortisone administration might be more reliable than cortisol suppression by dexamethasone in the diagnosis of Cushing’s disease. The results have shown that ACTH suppressibility by hydrocortisone administration was abnormal, in agreement with steroid excretion and mean plasma cortisol concentrations in 14/14 patients with persistently active Cushing’s disease, and in 4 of 7 patients with demonstrably intermittent hypercortisolism, a disorder associated with frequently normal and often excessive steroid levels in the absence of evident stress. These patients included 3 in whom suppression of plasma and urinary cortisol by dexamethasone had been found to be normal before excision of ACTH-containing pituitary adenomas.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Plasma ACTH suppression by hydrocortisone administration

By iv infusion. Starting at 0800 h, 15–30 min after inserting plastic cannulae into a forearm vein on each upper limb, blood was drawn for baseline plasma cortisol and ACTH measurements. Hydrocortisone was then administered in an initial bolus of 7 mg, followed by a continuous infusion of 3 mg/h for 10 h, into one of the veins. Blood was drawn from the contralateral vein at 1 h and every 2 h, for a total of 10 h in recumbency, for plasma cortisol and ACTH measurements.

By oral intake. After an overnight fast, blood was drawn into a vacutainer, containing EDTA solution and fitted with a siliconized stopper, through an indwelling iv cannula, for plasma cortisol and ACTH assays at 0800–0830 h. Patients and normal subjects were given hydrocortisone tablets by mouth, in amounts of 0.25 mg/kg BW, immediately after the initial blood sample had been obtained. Blood was then drawn every 30 min, for 120 min, for plasma cortisol and ACTH measurements.

The importance of performing this procedure in the fasting state has been indicated by evidence of delayed absorption of hydrocortisone administered after breakfast (see Table 5).

Laboratory measurements

Plasma cortisol concentration. Plasma cortisol concentration has been measured by RIA using kits from Amersham (Arlington Heights, IL) and Diagnostic Systems Laboratory (Webster, TX).

Urinary free cortisol output. Urinary free cortisol output was measured by RIA, using the same kits as for plasma concentrations, usually on at least two 24-h urines, since these measurements became available.

Urinary 17-OHCS output. All urines were tested for glucose. If glycosuria was present, sodium bisulfite was added in the amount of 250 mg to each tube, to prevent otherwise excessive color development, during the 16-h incubation with ß-glucuronidase, which preceded spectrophotometric measurement at 410 Å (8).

Plasma ACTH concentrations. Plasma ACTH concentrations were determined by immunoradiometric assay using kits supplied by Nichols Institute Diagnostics (San Juan Capistrano, CA). The lower limit of ACTH measurements with this procedure was 3 pg/mL.

The EDTA tubes in which blood for plasma ACTH determinations was collected were immediately placed in cups containing ice and centrifuged within 15–30 min at 4 C. Plasma was separated, immediately frozen at -70 C before thawing for cortisol and ACTH assays, both on the same day, within 8 weeks. Previous study revealed no significant changes in plasma ACTH concentration after storage at -70 C for 8 weeks.

Statistical methods

Statistical data have been expressed as mean, SEM, and 95% confidence intervals (11).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Steroid measurements in the 58 patients are shown in Table 2. Results that fell within the reference ranges are marked with asterisks. It is evident that: 1) the mean of 2-hourly plasma cortisol concentrations was within the normal range in only one of 55 determinations (1.8%); 2) basal urinary 17-OHCS excretion was below 6.5 mg/g creatinine in 9 of 123 determinations on urine from 6 of the 58 (10.3%) patients but was always excessive in at least one measurement in every patient; urinary 17-OHCS fell normally (to below 1.7 mg/g creatinine) in 6 patients (10.3%) after low-dose dexamethasone suppression; 3) urinary free cortisol excretion was normal in 8 of the 34 patients (23.5%) in whom it was measured and fell below 20 µg/day during low-dose dexamethasone administration in 7 of 30 patients (23%); 4) suppression of plasma cortisol by low-dose dexamethasone showed a normal response (to below 5 µg/dL) in 7 of 40 patients (17.5%).

It is evident from Table 3 that the infused hydrocortisone raised plasma cortisol concentrations to levels that were not statistically significantly different at 120 min after the onset of hydrocortisone infusion in the normal subjects (: 16.2; 95% CI, 12.9–19.5) and the Cushing’s patients (: 21.1; 95% CI, 15.2–27.0 µg/dL) and that this lack of significant difference was still present after 8 h.

Plasma ACTH concentrations, shown in Table 4, reveal reductions to: : <4.5; 95% CI, <3.0–6.2 pg/mL at 2 h, and : <3.1; 95% CI, <3.0–3.4 pg/mL at 8 h, after the onset of hydrocortisone infusion in the normal subjects. The plasma ACTH concentrations in the Cushing’s patients were higher in every measurement in all 8 patients and were significantly above the normal ranges at 2 h (: 59.4; 95% CI, 13.2–105.6 pg/mL) and at 8 h (: 63.1; 95% CI, 17.8–108.4 pg/mL), equivalent to a P value < 0.05.

Table 5 shows that hydrocortisone, in an oral dose of 0.25 mg/kg BW, raised plasma cortisol concentrations at 60 min to comparable levels in the normal subjects (: 24.5; 95% CI, 19.3–29.7 µg/dL) and in the groups of patients with active hypercortisolism (: 31.7; 95% CI, 18.9–44.9 µg/dL), and those with intermittent hypercortisolism (: 26.1; 95% CI, 15.6–36.6 µg/dL).

Plasma ACTH concentrations after oral administration of hydrocortisone are shown in Table 6. A gradual reduction in plasma ACTH concentration is evident in all of the normal subjects, except in study no. 17 in a subject who had mistakenly eaten breakfast before the test. When this subject was retested in the fasting state (study no. 16), his plasma ACTH fell at 90 min, as in all of the other studies, to 10 pg/mL or less. The results in the normal subjects at 90 min (: 5.6; 95% CI, 3.3–7.9 µg/dL) showed a significantly greater decline in plasma ACTH concentration than in the patients with active hypercortisolism (: 34.5; 95% CI, 18.8–51.2 pg/mL), but there was some overlap of the 95% CIs between the plasma ACTH results in the normal subjects and those in the patients with intermittent hypercortisolism as a group (: 11.6; 95% CI, 4.2–19.0 pg/mL). In four of the patients with intermittent hypercortisolism (patients no. 59, 63, 64, and 65), plasma ACTH remained above 10 pg/mL at 90 min after the oral hydrocortisone. After completion of these studies, we tested three other patients, who were unusually obese (>90 kg), and in whom the large dose of hydrocortisone administered raised plasma cortisol concentration to between 50.9 µg/dL (1.4 fmol/L) and 62.6 µg/dL (1.72 fmol/L) at 60 min, i.e. approximately twice the concentrations observed in some of our previous patients. By repeating the oral suppression study in two of these patients, with hydrocortisone in a lower dose (0.17–0.20 mg/kg), we reduced the level of plasma cortisol at 60 min to 24 and 34 µg/dL (660 and 940 nmol/L); and we observed normal suppression of plasma ACTH in one, but a clearly abnormal response in the other (plasma ACTH falling only to 20.8 pg/mL at 90 min). These observations may provide a caveat for the future study of ACTH suppressibility by cortisol, which is clearly a dose-dependent phenomenon.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The basis of Liddle’s introduction of the low- and high-dose dexamethasone suppression tests was his recognition of the fact that excessive secretion of ACTH, in the overwhelming majority of patients with Cushing’s disease, was partially suppressible by glucocorticoid administration in a dose-dependent manner. The present studies have shown that plasma ACTH concentrations fall, to some extent, in several patients with Cushing’s disease but consistently fall less and more slowly in the patients than in healthy subjects. Thus, the experimentally raised plasma cortisol levels present at 60 min after oral hydrocortisone, in a dose of 0.25 mg/kg, were not significantly different in the two groups of individuals, but there was complete separation between the levels to which plasma ACTH had fallen 90 min after the same inhibitory stimulus in the healthy individuals, in the patients with consistently evident hypercortisolism, and even in 4 of the 7 patients whose inconsistently excessive cortisol excretion indicated intermittent hypercortisolism (9, 10). If ACTH suppressibility by oral hydrocortisone administration is used as a procedure to confirm the presence of Cushing’s disease, therefore, it is essential that the usual absorption of the administered hydrocortisone be confirmed by finding a plasma cortisol concentration above the normal level, preferably between 16 and 38 µg/dL, at 0900–0930 h, 60 min after administration of the steroid, and in the absence of known stress. Studies no. 17 and 19 in the normal subjects, as shown in Table 5, indicate that after breakfast, plasma cortisol sometimes failed to show the usual rise 60 min after oral hydrocortisone administration, changing only from 10 to 9 and from 5 to 6 µg/dL, respectively, with no consequent fall in plasma ACTH at 90 min in these studies (Table 6). These data have confirmed the reasonable expectation that the absorption of hydrocortisone given by mouth might be delayed by a meal and, since breakfast and lunch are usually associated with a spontaneous rise in plasma cortisol concentration (13, 14), it is wise to perform this procedure after an overnight fast. Thus, in one of our normal subjects, plasma ACTH was elevated (25 pg/mL) 90 min after oral hydrocortisone, when he had mistakenly eaten breakfast before the test, but fell to 10 pg/mL (2.2 pmol/L) on a second occasion, when he had not eaten before the test. This individual is clinically normal, and his urinary free cortisol has been normal on 2 occasions (66 and 71 µg/day or 182 and 195 nmol/day).

The misleadingly normal suppressibility of cortisol by dexamethasone probably results from an even longer half-life of dexamethasone (indicated by higher plasma dexamethasone levels, when measured simultaneously with cortisol at 0800 h) in these patients than in others with Cushing’s disease. Whether other factors might be involved (such as different sites of action of the steroid on the HPA axis, or their different affinities for the glucocorticoid receptors in the pituitary adenomas) is not known.

In an attempt to improve the sensitivity and specificity of the overnight low-dose dexamethasone suppression test, Montwill et al. (15) have recently reported that lowering the cut-off point of plasma cortisol to 3.62 µg/dL resulted in an unsatisfactory false positive rate of 12.5%, whereas raising the cut-off point to 7.24 µg/dL was associated with a false positive rate of 7.3%. It is clear, from the data presented in Table 2, that the use of such a cut-off point would have resulted in interpreting the dexamethasone test result as normal in patients no. 22, 32, 37, 38, 39, 45, 53, and 58 (or 14% of our patients), which we cannot consider acceptable.

The present studies have shown abnormal resistance to ACTH suppression in three patients with unequivocal collateral steroid evidence of hypercortisolism but with normal cortisol suppressibility by dexamethasone. These findings indicate that suppressibility of ACTH by cortisol may be more consistently abnormal than the low-dose dexamethasone test, in patients with Cushing’s disease. This possibility is strongly supported by the fact that the three patients, whose surgery had been postponed because of normal responses to dexamethasone, were apparently cured by pituitary surgery shortly after their abnormal responses to ACTH suppression by oral hydrocortisone were found. These findings suggest that the procedures here described are worthy of further study and perhaps more general use, especially in patients who have clinical features and urinary or plasma measurements that indicate hypercortisolism but are being denied appropriate therapy because of what now seems to be misplaced reliance on normal cortisol suppression by dexamethasone. The iv procedure seems to have been more reliable than the oral test and clearly need not be continued for longer than 120 min (when plasma ACTH was consistently below 10 pg/mL). However, where facilities for hydrocortisone infusion and blood sampling for cortisol measurements from the contralateral arm are either unavailable or prohibitively expensive, it may be useful to know that the oral procedure usually is also reliable. It could probably be performed in a clinic setting or in some doctors’ offices, limiting plasma cortisol measurements to those made at 0 and 60 min (normal at 60 min: 16–38 µg/dL) and the ACTH measurements to 0 and 90 min (normal at 90 min: 10 pg/mL).

Because the suppressibility of plasma ACTH by administered hydrocortisone is likely to be dose-dependent, as the cortisol response to dexamethasone is known to be, it is clearly important to measure plasma cortisol at 60 min, after oral hydrocortisone has been administered. When this is above 38 µg/dL (1045 nmol/L) at 60 min and ACTH is suppressed below 10 pg/mL (2.2 pmol/L), as may happen in extremely obese individuals, our recent observations suggest that repetition of the study with a lower dose of hydrocortisone may sometimes provide what seems to be an appropriately abnormal result. On the other hand, in patients whose baseline plasma cortisol exceeds 38 µg/dL while simultaneously measured plasma ACTH concentration exceeds 10 pg/mL in the unstressed state after 90 min in recumbency, determination of the response to cortisol administration would seem to be unnecessary.

In general, therefore, the present evidence seems to indicate that in individuals of our population at 0930–1000 h, after they have been awake for at least 2 h and recumbent in a quiet environment for 90 min, and in the absence of recognizable stress, elevation of the plasma cortisol concentration above 16–38 µg/dL (the range observed after hydrocortisone administration to normal subjects) should suppress plasma ACTH concentration to 10 pg/mL or less, unless Cushing’s syndrome of central or ectopic origin or an unrecognizable stress is present. Whether these findings will apply to other populations and to measurements of plasma cortisol and ACTH concentrations by different methods, will clearly require confirmation in far larger normal and abnormal groups of subjects.

Orth et al. (16) have emphasized the need to measure steroid excretion for the diagnosis of Cushing’s syndrome in at least 2 accurately collected 24-h urines. The importance of this recommendation has been confirmed by our findings that, in 58 patients with Cushing’s disease, 1 of 2 urinary steroid determinations was normal in 9.6%, when urinary free cortisol was measured, and 1 of 2 was normal in 6.5% of determinations, when urinary 17-OHCS excretion was measured. However, the misleading normality of the cortisol suppressibility by dexamethasone did not result from intermittently excessive cortisol secretion in at least 2 of our patients, whose urinary excretion and mean plasma cortisol concentrations were consistently abnormal, whereas the dexamethasone suppression test was misleadingly normal on almost every occasion (Table 1).

Incidentally, it might be added that the data in Table 2 indicate that urinary 17-OHCS excretion, when measured appropriately (8) and expressed in mg/g creatinine, may be more reliable in the diagnosis of Cushing’s disease than urinary cortisol measurements. Thus, abnormal 17-OHCS measurements were found in 89.7% of measurements, both in the basal state and during low-dose dexamethasone administration, whereas urinary free cortisol excretion was abnormal in only 76–77% of our patients under the same conditions.

There is evidence that an erroneously normal plasma cortisol response to the overnight dexamethasone test may be recognized and corrected if the plasma dexamethasone concentration is measured simultaneously and found to be unusually elevated.

	Footnotes

 
¹ Presented, in part, at the annual meeting of the Endocrine Society, June 1997. This work was supported by the Clinical Research Unit of the SUNY Health Science Center, Syracuse, and by a research grant from the Wahrsager Foundation.

² Current address: Slocum-Dickson Medical Group, New Hartford, New York.

Received June 9, 1997.

Revised October 8, 1997.

Revised December 18, 1997.

Accepted December 24, 1997.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Liddle GW. 1960 Tests of pituitary-adrenal suppressibility in the diagnosis of Cushing’s syndrome. J Clin Endocrinol Metab. 20:1539–1560.
2. Vidal-Trecan G, Laudat MH, Thomopoulos P, Luton JP, Bricaire H. 1983 Urinary free corticoids: an evaluation of their usefulness in the diagnosis of Cushing’s syndrome. Acta Endocrinol (Copenh). 103:110–115.[Medline]
3. Nugent CA, Nichols T, Tyler FH. 1965 Diagnosis of Cushing’s syndrome: single dose dexamethasone test. Arch Intern Med. 116:172–176.
4. Crapo L. 1979 Cushing’s syndrome: a review of diagnostic tests. Metabolism. 28:955–977.[CrossRef][Medline]
5. Meikle AW, Lagerquist LG, Tyler FH. 1975 Apparently normal pituitary-adrenal suppressibility in Cushing’s syndrome: dexamethasone metabolism and plasma levels. J Lab Clin Med. 86:472–478.[Medline]
6. Streeten DHP, Anderson Jr GH, Dalakos TG, et al. 1984 Normal and abnormal function of the hypothalamic-pituitary-adrenocortical system in man. Endocr Rev. 5:371–394.[Abstract]
7. Streeten DHP, Stevenson CT, Dalakos TG, et al. 1969 The diagnosis of hypercortisolism. Biochemical criteria differentiating patients from lean and obese normal subjects and from females on oral contraceptives. J Clin Endocrinol Metab. 29:1191–1211.[Medline]
8. Louis LH, Eiler PA, Streeten DHP, et al. 1956 Correction of a major error in determining urinary Porter-Silber reactive corticosteroids in the presence of glycosuria. J Lab Clin Med. 48:922–923.
9. Atkinson AB, McCance DR, Kennedy L, Sheridan B. 1992 Cyclical Cushing’s syndrome first diagnosed after pituitary surgery: a trap for the unwary. Clin Endocrinol (Oxf). 36:297–300.[Medline]
10. Streeten DHP, Anderson Jr GH, Dalakos T, Joachimpillai AD. 1997 Intermittent hypercortisolism: a disorder strikingly prevalent after hypophyseal surgical procedures. Endocrine Practice. 3:123–129.
11. Gardner MJ, Altman DG. 1994 Statistics with confidence. Confidence intervals and statistical guidelines. BMJ.
12. Meikle AW, Lagerquist LG, Tyler FH. 1975 Apparently normal pituitary-adrenal suppressibility in Cushing’s syndrome: dexamethasone metabolism and plasma levels. J Lab Clin Med. 86:472–478.
13. Follenius M, Brandenberger G, Hietter B. 1982 Diurnal cortisol peaks and their relationship to meals. J Clin Endocrinol Metab. 55:757–760.[Abstract]
14. Quigley ME, Yen SSC. 1979 A mid-day surge in cortisol levels. J Clin Endocrinol Metab. 49:945–947.[Abstract]
15. Montwill J, Igoe D, McKenna TJ. 1994 The overnight dexamethasone test is the procedure of choice in screening for Cushing’s syndrome. Steroids. 59:296–298.[CrossRef][Medline]
16. Orth DN, Kovacs WJ, Debold CW. 1992 The adrenal cortex. In: Wilson JD, Foster DW, eds. Williams textbook of endocrinology. Philadelphia: Saunders; 548.

This Article Abstract Full Text (PDF) Submit a related Letter to the Editor Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Streeten, D. H. P. Articles by Jones, C. Search for Related Content PubMed PubMed Citation Articles by Streeten, D. H. P. Articles by Jones, C.