Dose\p=n-\responseaspects in the clinical assessment of the hypothalamo\x=req-\
`pituitary-adrenal axis, and the low-dose adrenocorticotropin test
`W Oelkers
`
`Division of Endocrinologg, Department of Medicine, Klinikum Benjamin Franklin (Steglitz), Freie Universität Berlin, Berlin, Germany
`
`Clinical tests of the hypothalamo-pituitary-adrenocortical
`axis in cases of suspected adrenocortical insufficiency are
`based almost exclusively on the stimulation of pituitary
`ACTH release or adrenocortical
`ACTH\x=req-\
`release of
`dependent steroids. The most widely applied tests of the
`hypothalamo-pituitary-adrenal axis in clinical practice
`(SAT),
`the insulin
`are the short ACTH injection test
`hypoglycemia test
`(IHT),
`the short metyrapone test
`(SMT) and the corticotropin-releasing hormone (CRH)
`test (1). In this review,
`the reliability of these tests in
`special clinical settings will be discussed under the aspect
`of dose\p=m-\responserelationships between plasma ACTH
`and cortisol in normal man.
`
`Problems with the conventional SAT
`If a patient is suspected to suffer from primary adrenal
`insufficiency (Addison's disease), the test for excluding
`or strengthening the suspicion is the SAT, comprising
`the injection (im or iv) of 250\g=m\gof ACTH (1\p=n-\24)
`(Synacthen\s=r\, Cosyntropin\s=r\or Cortrosyn\s=r\) and the
`measurement of plasma or serum cortisol 30 or 60 min
`after the injection. A basal or post-ACTH plasma cortisol
`of 550 nmol/l or greater is regarded as safely excluding
`insufficiency (1-3). In this disorder,
`primary adrenal
`basal plasma cortisol
`is decreased or low-normal and
`does not respond to ACTH because basal ACTH levels
`are markedly elevated, thus stimulating the remaining
`adrenocortical cells to 100% of their maximal secretory
`capacity (3). In cases of incipient (preclinical) primary
`adrenal insufficiency, however, it has been shown that
`measuring basal ACTH levels is even more sensitive
`than performing the SAT (4). Circulating antibodies
`directed against the adrenal ACTH receptor may also
`contribute to refractoriness of the adrenal cortex to
`exogenous ACTH (5).
`Because in severe forms of secondary adrenal
`levels the, cortisol
`insufficiency with low cortisol
`response to ACTH may also be absent due to adrenal
`atrophy and down-regulation of ACTH receptors (6),
`differentiation between primary and secondary adrenal
`insufficiency requires the demonstration of elevated
`plasma ACTH levels in primary insufficiency (3, 4) or
`the performance of a prolonged ACTH infusion test (1 ),
`
`*This paper is based on a symposium lecture presented to the
`Endocrinological Society, June 1995, in Washington, DC.
`
`which results in some increase of cortisol in secondary
`adrenal insufficiency.
`The facts reported up to now are not a matter for
`debate, but the question of whether the SAT is a good
`clinical test ("screening test") for any form of adrenal
`insufficiency and particularly for mild forms of second¬
`insufficiency has been unresolved in the
`ary adrenal
`past, with several statements for and against made by
`experienced endocrinologists. The detailed data in this
`regard have recently been reviewed competently by
`Grinspoon and Biller (1). Briefly, protagonists of the SAT
`(7-9) found a highly positive
`as a screening test
`correlation in patients with pituitary disease between
`the results of the'SAT and those of the IHT, which is still
`being regarded by many endocrinologists as the "gold
`standard" for the detection of any form of adrenal
`insufficiency. A normal IHT requires a normal activation
`of hypothalamic corticotropin-releasing mechanisms
`and normal responses of ACTH and cortisol. Antagonists
`of the SAT as a screening test (10-16) found discre¬
`pancies between the SAT and the IHT or the SMT in
`patients with pituitary disorders in the sense that the
`SAT is not sensitive enough to detect mild degrees of
`secondary adrenal
`insufficiency. Underdiagnosis of
`secondary adrenal insufficiency by the SAT may expose
`some patients to life-threatening complications in situa¬
`(e.g. major surgery) when
`tions of severe stress
`intermittent steroid replacement
`therapy would have
`been necessary. Antagonists and protagonists agree in
`the view that
`the SAT is not
`indicated in cases of
`suspected acute secondary adrenal
`insufficiency (e.g. a
`few weeks after pituitary surgery), because adrenal
`atrophy (the main reason for adrenal hyporesponsive-
`ness to ACTH in SAT) did not have sufficient time to
`develop (1).
`
`The SAT in the light of dose-response
`relationships between plasma ACTH and
`cortisol
`Early studies by Landon et al. (17) made it clear that the
`infusion of 4/ig per hour of ACTH(l-24) was sufficient
`to stimulate the adrenal cortex maximally. In spite of
`this observation, which obviously fell into oblivion until
`the SAT has been performed for about 30
`recently,
`years with a dose of 250pg of ACTH(l-24). Other
`authors (18, 19) confirmed the results of Landon et al.
`
`

`

`the course of the SAT, the adrenal cortex is exposed to
`excessively high ACTH levels. Although the plasma
`ACTH-cortisol dose-response curve is rather flat at
`ACTH levels greater than 100ng/l in the normal, it is
`possible that a partially atrophie adrenal with a
`decreased number of ACTH receptors per cell in patients
`with secondary adrenal
`insufficiency exhibits a dose-
`response curve that is different from that of a normal
`gland in the sense that an insufficient cortisol response
`to ACTH levels in the physiological
`range can be
`overcome by
`a huge overdose of ACTH. These
`considerations led to clinical tests with greatly reduced
`dosages of ACTH.
`
`Low-dose SAT
`Dickstein et al. (14) and Broide et al. (22) from Haifa
`recently developed a low-dose ACTH(l-24) test (LD-
`SAT). In normal subjects, the cortisol response 20 min
`after injection of 0.5 pg per 1.73 m2 of ACTH was not
`different from that after a single dose of 2 50 pg of ACTH,
`while 45 min after injection the cortisol response was
`significantly lower (and falling) after 0.5 µg of ACTH
`than after 250pg (still rising). When results of the
`conventional SAT and of the LD-SAT in 46 asthmatic
`children and young adults who chronically inhaled
`topically active glucocorticoids (budesonide or beclo-
`methasone diproprionate) were compared with those in
`33 age-matched controls,
`the following differences
`emerged: only one patient had a subnormal cortisol
`response in the 250-/¿g SAT, while 16 patients failed to
`reach a peak cortisol response to 500 nmol/1 or greater
`in the LD-SAT. Such a response was failed by only one of
`the control subjects in the LD-SAT. An additional
`criterion of normality in the LD-SAT was an increment
`of cortisol from baseline by at least 200 nmol/1. Patients
`with a subnormal cortisol response in the LD-SAT had
`significantly lower urinary free cortisol excretion levels
`(with a large overlap to normal) than patients with a
`normal response or control subjects (Fig. 2). These data
`seem to indicate that the LD-SAT can detect mild forms
`of secondary adrenal insufficiency that escape detection
`by the conventional SAT.
`Another article from Israel
`(23) comes to similar
`conclusions: Tordjman et al. compared the cortisol
`response to 250, 5 and 1 µg of ACTH(l-24) in seven
`normal subjects (group 1), 10 patients with pituitary
`macroadenomas and pathological steroid responses in
`the IHT or SMT (group 2) and nine patients with
`IHT and SMT
`pituitary macroadenomas and normal
`results (group 3). Basal morning cortisol
`levels in the
`three groups were similar, and no patient required
`regular hydrocortisone substitution. A plasma cortisol
`response to a level 497 nmol/1 (18 pg/d\) or greater was
`regarded as "pass" in each variant of the SAT and in the
`IHT. Groups 1 and 3 were essentially indistinguishable
`from each other with the three types of SAT. However,
`the cortisol responses to 1 pg were at all time points
`
`(17), but studies with intact human ACTH(l-39) and
`measurement of plasma ACTH levels after injection
`have not been performed. Could the marked overdose of
`ACTH be a cause of the relative insensitivity of the SAT
`in patients with mild secondary adrenal insufficiency?
`For investigating acute dose-response relationships
`between plasma ACTH and cortisol in normal man, we
`injected synthetic human ACTH(l-39) subcutaneously
`into eight normal male subjects and measured the
`course of plasma ACTH and cortisol with sensitive
`radioimmunoassays within the next 2 h (20). Doses of
`ACTH between 2.5 and 30 pg were used in a
`randomized order. The studies were performed in the
`afternoon, when endogenous ACTH and cortisol levels
`are relatively low, and each subject
`received five
`different doses of ACTH on different days with a time
`least 3 days.
`interval of at
`In order to compare the
`response of the adrenal to injected ACTH(l-39) with
`that to ACTH released by iv administration of human
`the experimental subjects also received 30 and
`CRH,
`100 pg of hCRH on different days. Mean peak plasma
`ACTH levels were then plotted against mean peak
`increments of plasma cortisol, as shown in Fig. 1. Mean
`basal plasma cortisol
`levels in this study were around
`200 nmol/1. There is a very steep initial segment ofthe
`dose-response curve. If we define the cortisol response
`to 30 pg of ACTH (plasma ACTH between 300 and
`350ng/l) as 100%,
`it is obvious that a near-maximal
`cortisol
`response (77%) was obtained with plasma
`ACTH levels as low as 60-70 ng/1. These levels are
`slightly higher
`than normal plasma ACTH in the
`morning (5-45 ng/1).
`In a previous study (21) we
`measured plasma ACTH in several normal subjects
`about 10 min after im injection of 2 50µ£ of ACTH(1-
`24) and found it to be invariably higher than 2000 ng/1
`(ACTH-level data not published previously). Thus,
`in
`
`10 µ ACTH (77%),_I
`f _^-—--1 20 Mg ACTH (94%)
`ilL
`¿£ 100 Mg CRH (76%)
`
`1 300
`
`200
`
`g ¿
`
`
`"5
`
` 100
`
`o
`
`,·, 2.5 Mg ACTH 11 2%)
`
`50
`
`100
`
`150
`
`200
`
`250
`
`300
`
`350
`
`peak ACTH levels ng/1
`
`Fig. 1. Dose-response relationships between plasma ACTH and delta
`cortisol after sc injection of different dosages of human ACTH(l-39)
`or iv injection of 30 and 100 ¿¿g of human CRH. The peak response of
`ACTH and cortisol to 30 µg of ACTH is arbitrarily defined as 100%.
`Mean values ± sem are shown. To convert ACTH values in ng/1 to
`pmol/1, multiply by 0.22. To convert cortisol values in nmol/1 to /ig/dl,
`multiply by 0.036 (Ref. 20).
`
`

`

`groups perform the test in the morning and others in the
`late afternoon when endogenous ACTH and cortisol is
`low. Some groups inject 1 /ig/kg body weight of human
`or ovine CRH iv and others use a standard dose of
`100 µg irrespective of body weight (25-2 7). As an
`example of normal test results in 50 healthy adults, the
`data of Schlaghecke et al. (26) are of interest. They
`injected 100 pg of hCRH between 8.00 and 9.00 a.m.
`and found an increase from a basal cortisol
`level of
`333 ± 101 sd to a maximum of 568 ± 188 nmol/1. The
`increased from 24 ± 17 to a maximum of
`ACTH level
`49 ±26 ng/1.
`injection of 30 and 100µg of
`As shown in Fig. 1,
`hCRH into normal men leads to plasma ACTH peaks of
`about 32 and 57ng/I..respectively, and the cortisol
`responses to these increments in ACTH falls exactly into
`the dose-response curve that has been constructed
`from results ofthe ACTH(l-39) injection study. Thus,
`bolus injections of CRH seem to stimulate cortisol
`exclusively through ACTH and not by other cleavage
`products of proopiomelanocortin. Orth et al. (28) tested
`the response of ACTH and cortisol to a wide range of
`oCRH dosages in normal man. The highest dose (30 pg/
`kg body weight) that caused unpleasant side effects did
`not stimulate plasma ACTH to mean levels greater than
`80 ng/1. The ACTH response to 1 pg/kg body weight in
`Orth's study (28) was about 40-50ng/1. The cortisol
`response to the highest dose of CRH was only slightly
`greater than that to 1 pg/kg. This would fit well into the
`flattening part of our dose-response curve (Fig. 1). The
`moderate response of plasma ACTH to a very large dose
`of CRH is probably due to a strong feedback of rising
`cortisol levels on the corticotroph cells of the pituitary,
`where glucocorticoids markedly blunt
`the ACTH
`response to CRH (25). Thus, even very high doses of
`CRH induce only a submaximal response of cortisol,
`while the standard tests for secondary adrenal
`insuffi¬
`ciency (the IHT and the SMT) lead to plasma ACTH
`levels of greater than 150 ng/1 (29, 30) and to an almost
`response. Nevertheless,
`maximal acute cortisol
`the
`results of
`the CRH test and the IHT correlated
`remarkedly well
`in a large number of
`(r = 0.82)
`patients on long-term glucocorticoid therapy, although
`plasma ACTH and cortisol responses in the IHT were
`more pronounced than in the CRH test (26). Further¬
`the CRH test has been an extremely valuable
`more,
`research tool in studying the cortisol-ACTH feedback in
`pathological states like Cushing's disease, anorexia
`nervosa and hypercortisolism associated with depres¬
`sion (25, 31).
`
`The SMT in the light of dose-response
`relationships
`The SMT and the IHT have been the standard tests for
`the detection of secondary adrenal
`insufficiency. The
`advantage of the SMT over the IHT is that only one
`blood sample is required and it can be performed when
`
`significantly lower than the responses to the higher
`doses of ACTH, in variance to the findings of Dickstein
`et al. (14) and Broide et al. (22). No subject of group 2
`passed the l^g ACTH test as normal, while 7/10
`group 2 patients would have passed the 5-/¿g ACTH
`test and 9/10 the 250 µg test as normal. The data
`presented in this paper support
`the view that a LD-
`SAT seems worthy of being evaluated extensively for
`the detection of milder forms of secondary AI because
`it is easy to perform, has no risk for the patient and is
`cheaper than most other tests. Similar results with a
`LD-SAT were reported recently by Rasmuson et al.
`(24).
`
`Dose-response aspects of the CRH test
`The CRH test is a test of secondary adrenal insufficiency
`and,
`in conjunction with the IHT, a valuable tool for
`the differential diagnosis between secondary and
`tertiary (hypothalamic) adrenal insufficiency (25). The
`test procedure has not been well standardized. Some
`
`p<0.0l
`
`Patients
`
`Controls
`
`400-,
`
`sz
`
`(M
`
`~ 300-
`
`o
`
`- 200
`
`I
`
`100-
`
`.1.
`
`oe
`
`n r
`
`roa
`
`UJ
`UJ
`rr
`
`rr
`
`< o
`
`r 3
`
`— -1-1-
`Serum
`Serum
`cortisol
`cortisol
`<500nmol/L>500 nmol/L
`
`Fig. 2. Urinary free cortisol excretion in normal controls and in
`children or young adults (patients) with asthma treated with inhaled
`beclomethasone or budesonide for more than 6 months. Sixteen out of
`46 patients (left column), but only one out of 3 3 controls, failed to
`respond to 0.5/ig of ACTH per 1.73 m2 with an increase of serum
`cortisol to > 500 nmol/1. These 16 patients had significantly lower
`urinary free cortisol
`levels than controls or patients responding
`normally to 0.5/jg of ACTH per 1.73 m"
`(medium column). All
`patients except one had normal serum cortisol responses (> 500 nmol/1)
`to the standard dose of 250/ig of ACTH(l-24). To convert serum
`cortisol values in nmol/1 to µg/dl, multiply by 0.036. To convert
`urinary free cortisol values in nmol/24h to /ig/24h. multiply by 0.36
`(Ref. 22).
`
`

`

`the IHT is contraindicated, e.g. in patients with coron;n \
`heart disease or at risk for cerebral convulsions. The SMT
`comprises the administration of 30 mg/kg body weight
`of metyrapone at midnight and the measurement of
`plasma-11-deoxycortisol at 08.00h the following morn¬
`ing (32). Metyrapone blocks the adrenal enzyme 11/3-
`hydroxylase, thereby inhibiting the formation of cortisol
`and stimulating the secretion of ACTH and the cortisol
`In normal subjects, 11-
`precursor 11-deoxycortisol.
`deoxycortisol
`rises after metyrapone administration
`from almost undetectable levels to 200 nmol/1 or greater
`(12, 32, 33). If plasma ACTH is measured in addition, its
`levels are > 150 ng/1 in almost every normal subject (29).
`Feek et al. (34), using the SMT with the measurement of
`plasma ACTH exclusively, made the interesting observa¬
`tion that in a group of patients with proven pituitary
`disorders, the SMT with ACTH measurement was more
`sensitive for discovering mild forms of secondary adrenal
`insufficiency than the IHT with the measurement of
`plasma cortisol. When we had evaluated our ACTH-
`cortisol dose-response study (20), we interpreted the
`data of Feek et al. (34) in the following way: ACTH is
`stimulated in the SMT as well as in the IHT vigorously to
`levels > 150ng/1 (29, 30). Because the ACTH-cortisol
`dose-response curve becomes flat at ACTH levels greater
`than 60-70 ng/1 (Fig. 1 ), a subnormal ACTH response to
`50 or 60 ng/1 in a patient with mild secondary adrenal
`insufficiency in either test may still be able to stimulate
`11-deoxycortisol or cortisol to above the lower limit of
`normal. The subnormal response of the hypothalamic-
`pituitary unit could, therefore, be detected by measuring
`ACTH but not by steroid measurement. We tested this
`hypothesis in 20 endocrinologically healthy subjects
`(group 1) and in 95 patients with proven pituitary
`disease, 25 of whom had already been on daily
`hydrocortisone replacement
`therapy because of more
`insufficiency (group 2b),
`severe secondary adrenal
`while the others (n = 70) were not substituted (group
`2a). A standard SMT with the measurement of plasma
`11-deoxycortisol plus ACTH was done in each subject.
`Patients of group 2b (between 10 and 2 5 mg of
`hydrocortisone per
`therapy-free
`observed
`day)
`a
`interval of at
`least 24 h before the test
`(35). The
`results are shown in Fig. 3. The relationship between
`ACTH and ll-deoxycortisol
`in these 115 subjects
`resembles the ACTH-cortisol dose-response curve of
`Fig. 1. All control subjects of group 1 had normal ACTH
`responses to > 150ng/1 and normal
`ll-deoxycortisol
`responses to >200nmol/1. All patients of group 2b
`had subnormal ACTH and ll-deoxycortisol responses.
`Twenty-three patients of group 2 a had completely
`normal ACTH and ll-deoxycortisol responses in the
`SMT. In these patients, secondary adrenal insufficiency
`could be excluded. Twenty-one other patients of group
`2a (30%) had subnormal 11-deoxycortisol responses, but
`47 patients (67%) had subnormal ACTH responses.
`Thus, 37% of the 70 patients in group 2a with pituitary
`disease had subnormal ACTH responses with a normal
`
`900
`
`Fig. 3. Relationship between plasma ACTH and ll-deoxycortisol at
`08.00h after administration of 30mg/kg body weight of metyrapone
`at midnight (O) 20 endocrinologieally healthy subjects (group 1 ): (D)
`70 patients with proven pituitary disease not on hydrocortisone
`substitution so far (group 2a): (*) 2 5 patients with proven pituitary
`disease receiving regular hydrocortisone substitution (group 2b).
`Hydrocortisone was omitted at least 24 h before the test. The lines
`within the graph mark the lower limits of ACTH response ( 1 50 ng/1)
`and of
`ll-deoxycortisol response (200nmol/1) to metyrapone. To
`convert ACTH values in ng/1 to pmol/1, multiply by 0.22. To convert
`ll-deoxycortisol values in nmol/1 to /¿g/dl, multiply by 0.03 5 (Ref.
`35).
`
`ll-deoxycortisol, i.e. measurement of ACTH was more
`sensitive than ll-deoxycortisol
`for detecting mild
`dysfunction of the hypothalamo-pituitary-adrenal axis.
`
`Is the IHT also affected by the dose-response
`ideology?
`Because ACTH rises to about the same high levels in the
`IHT as in the SMT (29), as mentioned before, it could be
`predicted that the IHT with measurement of plasma
`cortisol only would be blind,
`like the SMT with ll-
`deoxycortisol measurement only, to detect milder forms
`of hypothalamo-pituitary-adrenal insufficiency. To test
`this supposition, we measured intact plasma ACTH
`(IRMA method, kit from Nichols Institute, San Juan
`Capistrano, CA) and cortisol
`in the scope of the IHT
`(0.1-0.15 IU insulin/kg body weight) in 30 consecutive
`patients with pituitary tumors (pre- or postoperatively,
`excluding patients with Cushing's disease) and in six
`healthy young men and plotted the peak levels of ACTH
`against those of cortisol (Fig. 4). Only tests with a fall of
`blood glucose to <2.2 mmol/1 (<40 mg%) and a clinical
`response (sweating, tachycardia) were used. It is evident
`the relationship between the responses of
`that
`ACTH and cortisol resembles that shown in Figs 1 and
`3. The six control subjects and seven patients had
`completely normal ACTH and cortisol
`responses to
`hypoglycemia. Nine patients had a normal cortisol
`response in spite of a subnormal ACTH response, and 13
`patients had subnormal ACTH and cortisol responses.
`Thus, in 9 of 30 patients with pituitary tumors (30%)
`a mild dysfunction of
`the hypothalamo-pituitary-
`
`

`

`Conclusions and clinical relevance of modified
`hypothalamo-pituitary-adrenal tests
`Published reports and the data presented here have
`shown that
`the conventional high-dose SAT is not
`sensitive enough to reveal mild forms of secondary
`insufficiency and even preclinical primary
`adrenal
`adrenal dysfunction (4). There are only a few reports
`on the insensitivity of the IHT in the literature, and it
`can therefore be assumed that the scotoma in the eye of
`the IHT is smaller and less relevant than that of the
`high-dose SAT. The strategy to overcome the problems
`with the SAT is to use much smaller dosages of ACTH
`(LD-SAT), which probably lead to plasma ACTH levels
`that stimulate the adrenal cortex in the steep or
`flattening part of the dose-response curve, as shown
`in Fig. 1. Because such tests carry no risk for the patient,
`and because the measurement of plasma cortisol
`is
`relatively cheap, they should be evaluated carefully in
`several endocrinological centers on larger populations
`of controls and of patients who are at risk of having or
`developing secondary adrenal insufficiency. The studies
`thus far published on the LD-SAT are encouraging, but
`the number of patients examined in each single study is
`to allow final conclusions regarding the
`too small
`sensitivity and specificity of the test. Furthermore, a
`consensus on the ACTH dose in the LD-SAT and on the
`"pass" criterion is required (38, 39).
`If symptomatic primary adrenal
`insufficiency is the
`suspected diagnosis, the high-dose SAT should be used
`further together with basal ACTH measurement (3, 4).
`An established valuable test of secondary and tertiary
`adrenal insufficiency is the CRH test, which also caries
`no risk for the patient but is more expensive than the
`LD-SAT, especially if serial ACTH measurements are
`the LD-SAT was able to detect mild
`intended.
`If
`secondary adrenal
`insufficiency with high reliability,
`then the performance of IHT and SMT may be restricted
`to a minority of patients in whom special questions arise
`or in whom insulin-induced hypoglycemia is indicated
`for other
`In such cases, measurement of
`reasons.
`plasma ACTH in addition to cortisol may give valuable
`additional information that justifies the costs of ACTH
`measurement and the trouble of transporting chilled
`blood or frozen plasma samples to the laboratory. For
`the SMT, only one ACTH sample is required. We have
`used the SMT with ACTH measurement routinely for
`several years without problems and with a significant
`gain of information. We give patients with a normal 11-
`deoxycortisol response, but a clearly subnormal ACTH
`response, a steroid card with the recommendation to
`substitute glucocorticoids in stressful situations. Closer
`follow-up is also recommended in such patients.
`Finally,
`it has to be pointed out
`that
`the dose-
`response relationship presented in Fig. 1 is not applic¬
`able to patients with hyperplastic adrenal glands (e.g.
`Cushing's disease) or to those exposed to prolonged
`intensive-care patients. In the latter, very
`stress, e.g.
`
`1400
`
`1200
`
`-
`
`Stou
`
`T-1-r^— -1-1-1-1-1-1-1
`100 120 140 160 180 200 220 240 260 280 300
`60
`Maximum Plasma ACTH (ng/L)
`
`Fig. 4. Relationship between maximum plasma ACTH and cortisol
`responses to insulin-induced hypoglycemia (blood glucose <2.2 mmol/1
`or <40mg/dl) in six normal subjects (O) and in 30 patients with
`pituitary tumors pre- or postoperatively (·). The lines within the graph
`mark the lower limits of ACTH response (150 ng/1) and of cortisol
`(5 50 nmol/1)
`to hypoglycemia. Plasma ACTH (ng/1)
`response
` 0.22 = pmol/1; plasma cortisol (nmol/1) 0.036 = /¿g/dl.
`
`adrenal axis was only detected by measuring ACTH, not
`by cortisol, in the IHT. The observation by Feek et al.
`(34) mentioned above that
`the SMT with ACTH
`is more sensitive for detecting mild
`measurement
`hypothalamo-pituitary-adrenal
`insufficiency than the
`IHT with cortisol measurement, is thus easily explained
`by the fact that a normal ACTH response in the SMT
`in the IHT is markedly greater
`as well as
`than
`necessary for stimulating adrenocortical steroid secre¬
`tion to give a response in the low-normal
`range.
`(36), 17 years ago,
`Lindholm et al.
`reported on a
`"discrepancy between ACTH and cortisol responses to
`insulin-induced hypoglycemia" in 26 healthy subjects.
`They found no correlation between peak plasma ACTH
`and cortisol
`levels in the IHT, cortisol
`levels being
`equally high whether the ACTH response was low-
`normal or high-normal. They interpreted their data as
`showing an "overcapacity of ACTH secretion in
`comparison to the secretory capacity of the adrenal
`cortex" or, as we would put it now, the flatness of the
`dose-response curve at
`supraphysiological ACTH
`levels.
`There are few reports in the literature on failures of
`the IHT to detect secondary adrenal
`insufficiency.
`Tsatsoulis et al.
`reported on six patients
`(37)
`previously treated by external pituitary irradiation
`presenting with excessive tiredness, who had normal
`cortisol
`in the
`responses in the IHT (n = 4) or
`glucagon test (n = 2), but low urinary free cortisol
`and rather low plasma cortisol profiles. These patients
`profited markedly from regular glucocorticoid repla¬
`cement therapy. It is very likely that the abnormality
`of the hypothalamo-pituitary-adrenal axis would have
`been discovered in these patients if ACTH had been
`measured after insulin administration in addition to
`plasma cortisol.
`
`

`

`levels (often > 1000 nmol/1) are
`high plasma cortisol
`found a few days after admission (40, 41), probably due
`to homologous up-regulation of adrenocortical ACTH
`receptors (6) by high plasma ACTH levels.
`
`Acknowledgments. The author would like to thank Mrs Margot
`the
`Mahlcke who performed many of the hormone function tests,
`colleagues of the Endocrinology Unit who allowed me to use test
`results of their patients for this presentation and my former students
`Tim Boehlke and Holger Steiner who evaluated data of the dose-
`response study and of the short metyrapone test, respectively. Plasma
`cortisol measurements were partly performed in the Institute of
`Clinical Biochemistry of Klinikum Benjamin Franklin (Professor HJ
`Dulce).
`
`References
`
`in
`
`1. Grinspoon SK, Biller BMK. Laboratory assessment of adrenal
`insufficiency. J Clin Endocrinol Metab 1994;79:923-31
`2. May ME. Carey RM. Rapid adrenocorticotropic hormone test
`practice. Am J Med 1985:79:679-84
`3. Oelkers W, Diederich S, Bahr V. Diagnosis and therapy
`surveillance in Addison's disease: rapid adrenocorticotropin test
`and measurement of plasma ACTH, renin activity and aldoster¬
`one. J Clin Endocrinol Metab 1992;75:259-64
`4. Blevins LS, Shankroff J, Moser HW, Ladenson PW. Elevated
`plasma adrenocorticotropin concentration as evidence of limited
`adrenocortical reserve in patients with adrenomyeloneuropathy.
`J Clin Endocrinol Metab 1994;78:261-5
`5. Wulffraat NM, Drexhage HA, Bottazzo GF, Wiersinga GM,
`Immunoglobulins of patients
`Jeucken P. van der Gaag RD.
`with idiopathic Addison's disease block the in vivo action of
`adrenocorticotropin. J Clin Endocrinol Metab 1989:69:231-8
`6. Lebrethon M-C, Naville D, Begeot M, Saez IM. Regulation of the
`corticotropic receptor number and messenger RNA in cultured
`human adrenocortical cells by corticotropin and angiotensin II. J
`Clin Invest 1994:93:1826-33
`7. Lindholm J, Kehlet H. Re-evaluation of the clinical value of the
`in assessing the hypothalamic-pituitary-
`30 min ACTH test
`adrenocortical function. Clin Endocrinol 1987;26:53-9
`8. Lindholm J. Metyrapone test for hypothalamic-pituitary-adreno-
`cortical function. Clin Endocrinol [letter] 1994:41:695
`9. Stewart PM, Seckl JR, Corrie J. Edwards CRW, Padfield PL. A
`rational approach for assessing the hypothalamo-pituitary-
`adrenal axis. Lancet 1988:1:1208-10
`10. Borst GC. Michenfelder HJ. O'Brian JT. Discordant cortisol
`response to exogenous ACTH and insulin-induced hypoglycemia
`in patients with pituitary disease. Engl J Med 1982:306:
`1462-4
`11. Reschini E, Catania A, Giustina G. Plasma cortisol response to
`ACTH does not accurately indicate the state of hypothalamic-
`pituitary-adrenal axis. J Endocrinol Invest 1982:5:259-61
`12. Fiad TM. Kirby JM. Cunningham SK, Mckenna TJ. The
`is a simple and reliable
`overnight single-dose metyrapone test
`index of the hypothalamic-pituitary-adrenal axis. Clin Endocrinol
`1994;40:603-9
`for hypothalamic-
`13. McKenna JT. Fiad TM. Metyrapone test
`pituitary adrenocortical function. Clin Endocrinol
`[letter.reply]
`1994;41:695-6
`14. Dickstein G, Shechner C, Nicholson WE, Rosner I, Shen-Orr Z.
`Adawi F, et al. Adrenocorticotropin stimulation test: effects of
`basal cortisol levels, time of day, and suggested new sensitive low
`dose test. I Clin Endocrinol Metab 1991;72:773-8
`15. Streeten DHP, Anderson GH, Bonaventura M. The potential for
`serious consequences from misinterpreting normal responses to
`J Clin Endocrinol Metab
`the rapid adrenocorticotropin test.
`1996:81:285-90
`16. Soule SG, Fahle-Wilson M, Tomlinson S. Failure of
`
`the
`
`short ACTH test
`to unequivocally diagnose long-standing
`hypoadrenalism. Clin Endocrinol
`secondary
`symptomatic
`1996;44:137-40
`17. Landon J, James VHT, Cryer RJ, Wynn V, Frankland AW.
`Adrenocorticotropin effects of a synthetic polypeptide-/31-2
`corticotropin in man. J Clin Endocrinol Metab 1964;24:1206-13
`18. Crowley S, Hindmarsh PC, Holownia P, Honour JW, Brook CGD.
`The use of low doses of ACTH in the investigation of adrenal
`function in man. J Endocrinol 1991;130:475-9
`19. Daidoh H, Morita H, Muñe T, Murayama M, Hanafusa J. Ni H,
`et al. Responses of plasma adrenocortical steroids to low dose
`ACTH in normal subjects. Clin Endocrinol 1995;43:311-3
`20. Oelkers W. Boelke T, Bahr V. Dose-response relationships between
`plasms adrenocorticotropin, cortisol, aldosterone, and 18-OH-
`hydroxycorticosterone after injection of ACTH(l-39) or human
`CRH in man. J Clin Endocrinol Metab 1988:66:181-6
`21. Belkien L, Baumann J, Schirpai M. Oelkers W. Propranolol
`enhances the effect of ACTH on plasma cortisol but not on
`aldosterone in man. J Endocrinol Invest 1983:6:341-5
`22. Broide J. Soferman R, Kivity S, Colander A, Dickstein G, Spirer ,
`et al. Low-dose adrenocorticotropin test reveals impaired adrenal
`J Clin
`function in patients taking inhaled corticosteroids.
`Endocrinol Metab 1995:80:1243-6
`23. Tordjman K, Jaffe A, Grazas N, Apter C, Stern N. The role of the
`in the evaluation of
`low dose (1/ig) adrenocorticotropin test
`J Clin Endocrinol Metab
`patients with pituitary disease.
`1995;80:1301-5
`24. Rasmuson S, Olsson T, Hägg E. A low dose ACTH test to assess the
`function of the hypothalamic-pituitary-adrenal axis. Clin Endo¬
`crinol 1996;44:151-6
`25. Orth DN. Corticotropin-releasing hormone in humans. Endocr
`Rev 1992:13:164-90
`26. Schlaghecke R, Kornely E, Santen RT, Ridderskamp P. The effects
`long-term glucocorticoid therapy
`on pituitary-adrenal
`of
`responses to exogenous corticotropin-releasing hormone. Engl
`JMed 1992;326:226-30
`27. Trainer PJ, Faria M, Newell-Price J. Browne P. Kopelman P,
`Coy DH, et al. A comparison of the effects of human and ovine
`corticotropin-releasing hormone on the pituitary adrenal axis.
`J Clin Endocrinol Metab 1995:80:412-7
`28. Orth DN, Jackson RV, De Cherney GS, DeBold CR. Alexander AN.
`Island DP, et al. Effect of synthetic ovine corticotropin-releasing
`factor. Dose response of plasma adrenocorticotropin and cortisol.
`J Clin Invest 1983;71:587-95
`29. Staub JJ, Noelp B, Girard J, Baumann JB, Graf S, Ratcliffe JG. The
`short metyrapone test: comparison of the plasma ACTH response
`to metyrapone and insulin-induced hypoglycemia. Clin Endocri¬
`nol 1979;10:595-601
`30. Streeten DHP, Anderson GH, Dalakos TG, Seeley D, Mallov JS,
`function of
`the
`Ensebio R, et al. Normal and abnormal
`hypothalamic-pituitary-adrenocortical system in man. Endocr
`Rev 1984;5:371-94
`31. Gold PW. Loriaux DL, Roy A, et al. Responses to corticotropin-
`releasing hormone in the hypercortisolism of depression and
`Cushing's disease. Engl J Med 1986:314:1329-35
`32. Jubiz W, Meikle W, West CD, Tyler FH. Single-dose metyrapone
`test. Arch Intern Med 1970;125:472-4
`33. Achenbach K, Oelkers W. Comparison of insulin hypoglycemia
`and short metyrapone tests in patients with pituitary disease. Klin
`Wochenschr 1985;63:769-74
`34. Feek CM, Bevan JS, Ratcliffe JG, Gray CE, Blundell G. The short
`metyrapone test. Comparison of the plasma ACTH response