Shortcomings in the Low-Dose (1 µg) ACTH Test for the Diagnosis of ACTH Deficiency States

David H. P. Streeten

Professor Emeritus of Medicine SUNY Health Science Center Syracuse, New York 13210

Address correspondence and requests for reprints to: David H. P. Streeten, Dept. of Medicine, SUNY Health Science Center, 750 E. Adams St., Room CWB-322, Syracuse, New York 13210.


    Introduction
 Top
 Introduction
 What should we do...
 References
 
Severe and long-standing hypocortisolism may result from primary or secondary adrenocortical failure to secrete sufficient cortisol to sustain normal health either in the absence or in the presence of stress. In patients with primarily adrenal failure the diagnosis can be made reliably, quickly, and relatively inexpensively by measurement of the serum cortisol response to a single injection (250 µg) of ACTH1–24 (cosyntropin, tetracosatrin, Cortrosyn (Organon, W. Orange, NJ), or Synacthen (Novartis, Basel, Switzerland): all identical products), administered either i.v. or i.m. as an out-patient procedure. This diagnosis can even be made in many patients by measuring plasma concentrations of endogenous adrenocorticotropin (ACTH) and cortisol in a single blood sample (1). There is also no doubt that profound ACTH deficiency that has been present for more than a few weeks will, by causing severe adrenocortical atrophy or impairment of cortisol secretion, almost always result in a subnormal response in serum cortisol concentration to the "short ACTH test" (2).

However, it has been recognized for several years that, in some patients who had adrenal insufficiency secondary to hypothalamic or pituitary failure, the normal result of the short ACTH test with 250 µg did not necessarily indicate normal function of the entire hypothalamic-pituitary-adrenal (HPA) system. This evidence and the fact that plasma ACTH levels rose to extremely high levels after 250 mg of ACTH1–24 led to attempts to determine whether the elevation of plasma ACTH to more physiologically increased concentrations by the injection of lower doses of ACTH1–24, might provide a more reliable means of diagnosing moderate or mild forms of corticotropin deficiency. The fact that 1 µg and 250 µg doses of ACTH1–24 resulted in very similar plasma levels of glucocorticoids at 30 min (3) was substantiated in several subsequent measurements, particularly those of Dickstein et al. (4). These clinical investigators found, in six patients who had been on long-term steroid therapy, that serum cortisol concentration rose to 649 (SD 91) nmol/L, similar to the responses of normal subjects (to 707, SD 55, P > 0.3) after the 250 µg dose of ACTH1–24, but rose subnormally after 1 µg of ACTH1–24 in five of the six steroid-treated patients, to 458 (SD 86) compared with 704 (SD 72) nmol/L in normal subjects (P < 0.01). In contrast to these abnormal responses to the 1 µg dose of ACTH1–24, only one of the six steroid-treated patients responded subnormally to the 250 µg dose. These impressive data stimulated several more recent publications by Dickstein’s and other groups (5, 6, 7, 8), in which it was reported that some patients with suspected or established ACTH deficiency responded subnormally to the 1 µg dose of ACTH1–24 (in some instances confirming the conclusions from an abnormal insulin tolerance test) but normally to the 250 µg dose. Such results might reasonably have been expected now that it has been shown that, unlike the usual up-regulation of end-organ receptors in response to reduced concentrations of the appropriate agonist, the ACTH receptors on human adrenocortical cells—at least during culture in vitro—are up-regulated when exposed to increasing ACTH concentrations (10) and would be expected therefore, to be down-regulated by the subnormal plasma concentrations of endogenous ACTH in patients with ACTH deficiency.

The article in this issue of JCE&M by Abdu et al. (see page 0000) (9) reports another comparison of cortisol responses to the 1 and 250 µg ACTH1–24 tests and to the insulin tolerance test in a series of 64 patients with suspected or proven pituitary diseases of various types. Using 600 nmol/L (22 µg/dL) as the lower limit of the normal serum cortisol response to 1 µg ACTH1–24, the authors found a sensitivity of 100% compared with responses to the insulin tolerance test. Eleven percent of their patients were abnormal by the low dose test without other clinical or biochemical evidence of cortisol deficiency. The low dose test was found to be superior to the 250 µg ACTH test, and the authors concluded that, as a screening procedure, the low dose test was an attractive alternative to the insulin tolerance test that could be reserved for patients in whom careful clinical evaluation failed to dispel reasonable doubts of possible HPA axis impairment. Used in this way, as a screening test for HPA insufficiency and with the recognition that the insulin tolerance test would be required in patients with borderline responses, the low dose test might eventually prove to be a useful diagnostic procedure.

Regrettably, the apparently encouraging evidence supporting the usefulness of the low dose ACTH test has not eliminated all uncertainties. At least two problems indicate that the 1-µg ACTH test cannot provide a conclusive means of excluding the presence of mild or incomplete deficiencies in the functional capacity of the HPA axis. The first problem to be addressed is the fact that the lower limit of the normal cortisol response to 1 µg of ACTH1–24 has been variously considered to be 600 nmol/L (22 µg/dL) (9), 560 nmol/L (20 µg/dL) (4), 535 nmol/L (19 µg/dL) (11), 500 nmol/L (18 µg/dL) (6), and 480 nmol/L (17 µg/dL) (5). Obviously, the interpretation of normality and of the specificity of the low dose test will be very dependent on the determination of the lower limit of the normal range. Moreover, this determination is going to vary widely from one laboratory to another and from determinations performed with one RIA kit compared with another (12). Unless and until a generally acceptable lower limit of the normal range can be defined, the low dose procedure is likely to prove more confusing than diagnostically helpful to clinicians. And body size will probably influence the results of serum cortisol responses to such low doses of ACTH, as some investigators have appreciated (11, 12, 13). Therefore, if limits of the normal range of the 30-min plasma cortisol concentration can be agreed upon, it would be preferable to have these expressed in terms of serum cortisol responses to a dose of ACTH1–24 expressed in µg/m2 or µg/1.73 m2.

There is another difficult problem that will confront physicians who allow their diagnostic decisions to depend entirely on the low dose ACTH or even the insulin tolerance test. This results from the evidence, already clearly available, that mild deficiencies of ACTH secretion do not always prevent a normal rise in serum cortisol concentration stimulated by injected ACTH, thereby failing to indicate the potential danger of adrenocortical insufficiency during exposure to severe stress. This has been strongly suggested by the findings of a normal response to the 250 µg dose of ACTH1–24 in individuals with severely abnormal responses to the insulin tolerance test (14, 15) or the metyrapone test (16). The potentially life-threatening consequences of relying on a "normal" response to 250 µg of ACTH1–24 given i.v. have been dramatically illustrated in four patients (17). The partial impairment in their ACTH response to severe stress was disguised by the presence of a plasma cortisol concentration 30 min after 250-µg injection of ACTH1–24, normal for a mildly stressed individual but inadequate for the severe stresses to which these patients had been exposed. Their hypotension (exclusively orthostatic in two and present in recumbency in the other two patients) and other evidence of adrenal insufficiency were associated with a brisk response in plasma cortisol to i.v. ACTH1–24 (250 µg) and were rapidly responsive to intravenous steroid therapy. These patients were later shown to have subnormal responses to the overnight metyrapone test. How often this clinical picture occurs in our intensive care units and other hospital beds cannot be reliably estimated. Even at autopsy in such patients it is very difficult to establish whether the presence of immunoreactive ACTH in the pituitary and its release in response to stress were adequate to prevent death from secondary adrenocortical insufficiency.

The evidence that mild deficiency of ACTH release is a real entity and, like mild deficiency of thyroxine and several other hormones, may be as common as severe deficiency, has been supported by the meticulous studies of Mayenknecht et al. (11). These investigators found, in patients with clear evidence of moderate or severe ACTH deficiency (>10% reduction below lower limits of normal responses to the insulin tolerance or the metyrapone test), that the serum cortisol responses to ACTH1–24 were subnormal in all 14 patients when low dose ACTH1–24 was used (0.5 µg/m2), and in 13 of the 14 patients when the 250 µg dose was used. However, among 9 additional patients whose insulin tolerance or metyrapone tests were only mildly impaired (<10% below the lower normal limits of these tests), responses to ACTH1–24 were normal at 30 min in 7 of the 9 patients during the 250 µg dose and in 8 of the 9 patients during the 1 µg dose of ACTH1–24. The authors concluded that there was no significant difference in the responses to the high and low dose tests. Even more important, perhaps, is their demonstration that 8 of their 23 patients (35%) with ACTH deficiencies, all with evidence of mildly subnormal responses to the insulin tolerance or metyrapone test would have been passed as having a functionally normal HPA axis if either the high- or low-dose ACTH test had been relied upon. Additional evidence that the low dose test is no more successful than the 250 µg test in identifying patients with functional impairment of the HPA axis was reported by Weintrob et al. (18).


    What should we do in the face of these diagnostic predicaments?
 Top
 Introduction
 What should we do...
 References
 
In unstressed individuals with symptoms and/or signs (including orthostatic hypotension) suggesting adrenal insufficiency, a rapid ACTH test is a simple and valuable first step. If it is abnormal, the primary or secondary origin of adrenal insufficiency should be determined, and maintenance therapy with hydrocortisone and fludrocortisone (if needed) should be initiated. Whether or not the 1-µg dose of ACTH is preferable is obviously still uncertain. If the result of the ACTH test is borderline and the signs and symptoms are impressive, an insulin tolerance test as suggested by Abdu et al. (9) or an overnight metyrapone test (requiring a night in the hospital) should be performed.

In patients seen when under stress or who are candidates for stress (e.g. surgery), the short ACTH test, using either 250 or 1 µg doses, continues to be a reliable indicator of the need for steroid therapy if the serum cortisol response is subnormal; but unfortunately, it cannot be relied upon if the response is normal. It is known that the normal increase in adrenocortical secretion during stress is roughly proportional to the severity of the stress, as Chernow et al. (19) have shown. These authors reported that, in response to such mild surgical stresses as laparoscopy and repair of an inguinal hernia, serum cortisol rose 1 h after the incision to only slightly above the level reached normally after a short 250-µg ACTH test: 500–630 nmol/L (or 18–23 µg/dL). More stressful procedures, such as cholecystectomy, appendectomy, and vaginal hysterectomy, raised serum cortisol to mean values above 900 nmol/L (32 µg/dL), and the most stressful forms of surgery evaluated (including colectomy, gastrectomy, and aorto-femoral bypass) increased serum cortisol to approximately 1430 nmol/L (or 52 µg/dL), 1 h after the onset of surgery.

When measured hourly (in other studies) for longer periods of time after the initial incision, plasma cortisol concentrations have increased steadily from mean values of 750–950 nmol/L (27–34 µg/dL) at 30 min, to peak concentrations of 1270–1360 nmol/L (46–49 µg/dL) at or after the conclusion of uncomplicated cholecystectomy (17, 20). There is little doubt that far higher cortisol concentrations are reached after even more severe and long-lasting stress and are probably maintained for as long as the stress continues at levels considerably in excess of 500–630 nmol/L (18–23 µg/dL), which is required for a normal response to the short ACTH test. The presence of normal responses to either the short ACTH test or to the insulin tolerance test give us no assurance that a patient’s serum cortisol concentration will reach and be maintained at the far higher concentrations normally seen after severe, continuing stresses such as multiple traumatic injuries, infections, and burns. The only test that will indicate the capacity of the adrenals to raise serum cortisol concentrations to between 800 and 2200 nmol/L (29–80 µg/dL) is measurement of serum cortisol every 2 h during an i.v. infusion of ACTH1–24, 250 µg over 8 h (21). However, even a normal response of this type does not ensure adequacy of the patient’s hypothalamic and pituitary capacity to stimulate an increase in serum cortisol concentrations to these levels during severe stress. Until a test becomes available to evaluate the ability of the entire HPA axis to raise serum cortisol far above 630 nmol/L (23 µg/dL), we have no recourse other than to rely on clinical assessment of the severity of stress and an "educated" guess of the adequacy of the measured serum cortisol concentration. When the adequacy of this response is in doubt, our only potentially life-saving response is to prescribe a therapeutic trial of i.v. hydrocortisone, 200 µg/day, by continuous i.v. infusion, or by six-hourly i.v. injections, or of dexamethasone 4 mg six-hourly injections i.v., for at least 3–4 days, with reduction in the dosage if and when clinical improvement or abatement of the stress occurs.

Experience over the past 45 yr has shown that, when given such steroid therapy, most patients with Addison’s disease or those who have undergone bilateral adrenalectomy will survive severe surgical, traumatic, or infective stresses about as well as patients with intact HPA axes, and without recognizable, serious side-effects of the steroids administered. And there is evidence that this clinical, albeit "unscientific" approach, has saved lives in individuals whose partial HPA insufficiency could not be immediately substantiated with laboratory procedures but was later confirmed with a metyrapone test (17) without convincing evidence of any harmful consequences to date. On the other hand, the administration of far larger, perhaps inappropriately excessive doses of glucocorticoids (methylprednisolone 30 mg/Kg every 6 h for 1 day) have not improved the outcome in patients with severe stress and shock in the absence of known HPA failure (22).

Received December 14, 1998.

Accepted December 18, 1998.


    References
 Top
 Introduction
 What should we do...
 References
 

  1. Oelkers W, Diederich S, Bähr V. 1992 Diagnosis and therapy surveillance in Addison’s disease: rapid adrenocorticotropin (ACTH) test and measurement of plasma ACTH, renin activity, and aldosterone. J Clin Endocrinol Metab. 75:259–264.[Abstract]
  2. Lindholm J and Kehlet H. 1987 Re-evaluation of the clinical value of the 30 min ACTH test in assessing the hypothalamic-pituitary-adrenocortical function. Clin Endocrinol. 26:53–59.[Medline]
  3. Landon J, James VHT, Wharton MJ, et al. 1967 Threshold adrenocortical sensitivity in man and its possible application to corticotrophin bioassay. Lancet. ii:697–700.
  4. Dickstein G, Shechner C, Nicholson WE, et al. 1991 Adrenocorticotropin stimulation test: effects of basal cortisol level, time of day, and suggested new sensitive low dose test. J Clin Endocrinol Metab. 72:773–778.[Abstract]
  5. Tordjman K, Jaffe A, Grazas N, et al. 1995 The role of low dose (1 µg) adrenocorticotropin test in the evaluation of patients with pituitary diseases. J Clin Endocrinol Metab. 80:1301–1305.[Abstract]
  6. Rasmuson S, Olsson T, Hägg E. 1996 A low dose ACTH test to assess the function of the hypothalamic-pituitary-adrenal axis. Clin Endocrinol 44:151–156.
  7. Thaler LM, Blevins LS Jr. 1998 The low dose (1-µg) adrenocorticotropin stimulation test in the evaluation of patients with suspected central adrenal insufficiency. J Clin Endocrinol Metab. 83:2726–2729.[Abstract/Free Full Text]
  8. Talwar V, Lodha S, Dash RJ. 1998 Assessing the hypothalamo-pituitary-adrenocortical axis using physiological doses of adrenocorticotropic hormone. Q J Med 91:285–290.
  9. Abdu TAM, Elhadd TA, Neary R, et al. 1998 Comparison of low dose short Synacthen test (1 mcgm), conventional dose short Synacthen test (250 mcgm) and insulin tolerance test for the assessment of the hypothalmo-pituitary-adrenal axis in patients with pituitary disease. J Clin Endocrinol Metab. 84:0000–0000.
  10. Lebrethon M-C, Naville D, Begeot M, et al. 1994 Regulation of corticotropin receptor number and messenger RNA in cultured human adrenocortical cells by corticotropin and angiotensin II. J Clin Invest. 93:1828–1833.[Medline]
  11. Mayenknecht J, Diederich S, Bähr V, et al. 1998 Comparison of low and high dose corticotropin stimulation tests in patients with pituitary disease. J Clin Endocrinol Metab. 83:1558–1562.[Abstract/Free Full Text]
  12. Clark PM, Neylon I, Raggatt PR, et al. 1998 Defining the normal cortisol response to the short Synacthen test: implications for the investigation of hypothalamic-pituitary disorders. Clin Endocrinol (Oxf). 49:287–292.[CrossRef][Medline]
  13. Broide J, Soferman R, Kivity S, et al. 1995 Low-dose adrenocorticotropin test reveals impaired adrenal function in patients taking inhaled corticosteroids. J Clin Endocrinol Metab. 80:1243–1246.[Abstract]
  14. Borst GC, Michenfelder HJ, O’Brian JT. 1982 Discordant cortisol responses to exogenous ACTH and insulin-induced hypoglycemia in patients with pituitary disease. N Engl J Med. 130:1462–1464.
  15. Reschini E, Catania A, Giustina G. 1982 Plasma cortisol response to ACTH does not accurately indicate the state of the hypothalamic-pituitary-adrenal axis. J Endocrinol Invest. 5:259–261.[Medline]
  16. Cunningham SK, Moore A, McKenna TJ. 1983 Normal cortisol response to corticotropin in patients with secondary adrenal failure. Arch Intern Med. 143:2276–2279.[Abstract]
  17. Streeten DHP, Anderson Jr GH, Bonaventura MM. 1996 The potential for serious consequences from misinterpreting normal responses to the rapid adrenocorticotropin test. J Clin Endocrinol Metab. 81:285–290.[Abstract]
  18. Weintrob N, Sprecher E, Josefsberg Z, et al. 1998 Standard and low-dose short adrenocorticotropin test compared with insulin-induced hypoglycemia for assessment of the hypothalamic-pituitary-adrenal axis in children with idiopathic multiple pituitary hormone deficiencies. J Clin Endocrinol Metab. 83:88–92.[Abstract/Free Full Text]
  19. Chernow B, Alexander HR, Smallridge RC, et al. 1987 Hormonal responses to graded surgical stress. Arch Intern Med. 147:1273–1278.[Abstract]
  20. Donald RA, Perry EG, Wittert GA, et al. 1993 The plasma ACTH, AVP, CRH, and catecholamine responses to conventional and laparoscopic cholecystectomy. Clin Endocrinol (Oxf). 38:609–615.[Medline]
  21. 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]
  22. Bone RC, Fischer Jr CJ, Clemmer TP, et al. 1987 A controlled clinical trial of high-dose methylprednisolone in the treatment of severe sepsis and septic shock. N Engl J Med317 :653–658.