Investigation of patients with atypical or severe hyperandrogenaemia including androgen-secreting ovarian teratoma.

Approximately 7% of women of reproductive age manifest polycystic ovary syndrome (PCOS) and <0.5% have other causes of hyperandrogenism including congenital adrenal hyperplasia (CAH), androgen-secreting tumour of an ovary or an adrenal gland, Cushing's syndrome or hyperthecosis. The presence of features atypical of PCOS should prompt more extensive evaluation than that usually undertaken. Features atypical of PCOS include the onset of symptoms outside the decade of 15-25 years, rapid progression of symptoms, the development of virilization and a serum testosterone concentration in excess of twice the upper limit of the reference range. Ethnic background, family history and specific clinical findings, e.g. Cushingoid appearance, may inform a focused investigation. Otherwise, patients should have measurement of 17-hydroxyprogesterone (17-OHP) under basal conditions ideally in the early morning, and if abnormal, they should have measurement of 17-OHP one hour after the administration of synthetic ACTH, 250 microg i.v., to screen for CAH, which is present in approximately 2% of hyperandrogenic patients. The overnight cortisol suppression test employing 1 mg dexamethasone at midnight is a sensitive test for Cushing's syndrome. Coronal tomographic (CT) scanning of the adrenals and transvaginal ultrasonography of the ovaries are the investigations of choice when screening for tumours in these organs. Less frequently required is catheterization and sampling from both adrenal and ovarian veins, which is a technically demanding procedure with potential complications which may provide definitive diagnostic information not available from other investigations. Illustrative case reports highlight some complexities in the investigation of hyperandrogenic patients presenting with features atypical of PCOS and include only the ninth case report of an androgen-secreting ovarian teratoma.

Should macroprolactin be measured in all hyperprolactinaemic sera?

Macroprolactin is a nonbioactive prolactin isoform usually composed of a monomer of prolactin and a IgG molecule which has a prolonged clearance rate similar to that of the immunoglobulins. Macroprolactinaemia, hyperprolactinaemia entirely accounted for by the presence of macroprolactin, is estimated to account for approximately 10% of all hyperprolactinaemia coming to clinical attention in the United Kingdom and the United States. Failure to recognize that macroprolactinaemia can explain hyperprolactinaemia, leads to unnecessary investigation, incorrect diagnosis and inappropriate treatment. Screening of hyperprolactinaemic sera for the presence of misleading concentrations of macroprolactin is readily performed in biochemistry laboratories although the procedures have not been automated. The most widely employed method is to treat the hyperprolactinaemic sera with polyethylene glycol which precipitates out high-molecular weight constituents including immunoglobulins. Re-assay of the sera for prolactin will then identify those sera which yield values within the relevant normal range indicative of macroprolactinaemia and not true hyperprolactinaemia. The case for the routine screening of all hyperprolactinaemic sera for macroprolactin is compelling. The consequences of failure to recognize macroprolactinaemia are significant, the problem is frequently encountered, the means of addressing it are immediately available and it is cost effective.

Characterization of macroprolactin and assessment of markers of autoimmunity in macroprolactinaemic patients.

OBJECTIVE: It has been reported that macroprolactin is a complex of PRL and an immunoglobulin G (IgG). This study further characterizes macroprolactin and evaluates for other markers of autoimmunity using a cohort of macroprolactinaemic sera. PATIENTS AND NORMAL SUBJECTS: Following treatment of hyperprolactinaemic sera (n = 58) with polyethylene glycol (PEG), PRL values fell from 524-13 546 mU/l (Range) to 452-8455 mU/l, while in macroprolactinaemic sera (n = 41), PRL concentration fell from 525-5747 to 98-378 mU/l (PEG treated normoprolactinaemic reference range, 68-230 mU/l in males, 70-390 mU/l in females). DESIGN: PRL was measured in sera prior to and following gel filtration chromatography, ultrafiltration, treatment with protein A-sepharose, protein G-sepharose, antihuman IgG-agarose and sodium thiocyanate (NaSCN). The binding of radio-labelled PRL in macroprolactinaemic sera was also measured. Sera were assayed for antithyroid and antinuclear antibodies. C-reactive protein (CRP) and CD5 positive B cells were also measured. Comparisons were made between values obtained in normal, hyperprolactinaemic and macroprolactinaemic sera. Results Macroprolactinaemic sera indicated the presence of an IgG molecule and/or IgG fragments with one or more molecules of PRL. In 97% of the sera macroprolactin had a molecular weight of 204 kDa. Treatment of macroprolactinaemic sera with NaSCN caused dissociation of macroprolactin, releasing monomeric PRL. Macroprolactinaemic sera did not yield evidence of an increase in markers of autoimmunity when compared with hyperprolactinaemic or normal sera. CONCLUSIONS: Comprehensive analysis of macroprolactin confirmed its composition as an IgG molecule or fragment with a PRL molecule. The occurrence of macroprolactin does not appear to be associated with autoimmunity.

Serum total prolactin and monomeric prolactin reference intervals determined by precipitation with polyethylene glycol: evaluation and validation on common immunoassay platforms.

BACKGROUND: Macroprolactin is an important source of immunoassay interference that commonly leads to misdiagnosis and mismanagement of hyperprolactinemic patients. We used the predominant immunoassay platforms for prolactin to assay serum samples treated with polyethylene glycol (PEG) and establish and validate reference intervals for total and monomeric prolactin. METHODS: We used the Architect (Abbott), ADVIA Centaur and Immulite (Siemens Diagnostics), Access (Beckman Coulter), Elecsys (Roche Diagnostics), and AIA (Tosoh) analyzers with samples from healthy males (n = 53) and females (n = 93) to derive parametric reference intervals for total and post-PEG monomeric prolactin. Concentrations of immunoreactive prolactin isoforms in serum samples from healthy individuals were established by gel filtration chromatography (GFC). We then used samples from 22 individuals whose hyperprolactinemia was entirely attributable to macroprolactin and 32 patients with true hyperprolactinemia to compare patient classifications and prolactin concentrations measured by GFC with the newly derived post-PEG reference intervals. RESULTS: Parametric reference intervals for post-PEG prolactin in male and female serum samples, respectively, were (in mIU/L): 61-196, 66-278 (Centaur); 63-245, 75-381 (Elecsys); 70-301, 92-469 (Access); 72-229, 79-347 (Architect); 73-247, 83-383 (AIA); and 78-263, 85-394 (Immulite). Concordance between GFC and immunoassay-specific post-PEG reference intervals was observed in 311 of 324 cases and for 31 of 32 patients with true hyperprolactinemia and 17 of 22 patients with macroprolactinemia. Results leading to misclassification occurred in a few analyzers for 5 macroprolactinemia patient samples with relatively minor increases in post-PEG prolactin (mean 61 mIU/L). CONCLUSIONS: Our validated normative reference data for sera pretreated with PEG and analyzed on the most commonly used immunoassay platforms should facilitate the more widespread introduction of macroprolactin screening by clinical laboratories.

Androgens, exercise capacity, and muscle function in cystic fibrosis.

BACKGROUND: It is well established that muscle strength and exercise capacity are impaired in adults with cystic fibrosis (CF). Numerous factors have been implicated in this process including nutritional factors, disuse, and corticosteroid use. It is also known that serum testosterone levels are frequently low in adolescent male patients with CF. As testosterone is known to have an effect on peripheral muscle, we aimed to assess the relationship among circulating androgen levels, muscle strength, and exercise capacity in adults with CF. METHODS: A prospective observational study was performed. Participants underwent maximal exercise testing and peripheral muscle strength assessment (quadriceps, shoulder flexion, handgrip, maximal inspiratory pressure, and maximal expiratory pressure). Venous blood samples were obtained for the measurement of total testosterone, sex hormone-binding globulin, and albumin. The free testosterone (FT) level was calculated. RESULTS: A total of 15 subjects completed the study. The patients who were studied had impaired muscle strength and exercise capacity but normal mean androgen levels. Quadriceps strength, SF, and peak oxygen uptake (Vo(2)max) were 40.1 +/- 9.4%, 59.3 +/- 20.3%, and 77.4 +/- 22.2% predicted, respectively. One patient had a low testosterone level, with low FT levels evident in two subjects. There was no correlation between testosterone or FT level with any of the muscle strengths calculated. Neither testosterone or FT level was correlated with exercise capacity or pulmonary function. CONCLUSIONS: Male CF patients with impaired skeletal muscle strength and exercise capacity have normal testosterone levels. Hypogonadism was rare in this group of adult men with CF. These findings would not support the role of testosterone in muscle dysfunction or impaired exercise capacity in CF patients.

A simple and cost-effective approach to assessment of pituitary adrenocorticotropin and growth hormone reserve: combined use of the overnight metyrapone test and insulin-like growth factor-I standard deviation scores.

CONTEXT: The insulin tolerance test (ITT) is the gold standard for assessment of ACTH and GH reserve in patients with suspected hypopituitarism. It is labor intensive and costly. OBJECTIVE: The objective of the study was to determine whether use of the overnight metyrapone test (OMT) and plasma IGF-I sd scores (SDS) could provide a cost-effective alternative to the ITT. DESIGN: This was a retrospective chart review. SETTING: The study was conducted at a teaching hospital. PARTICIPANTS AND INTERVENTION: Charts from 100 patients with organic pituitary disorders were reviewed. All underwent the OMT unless 0900 h plasma cortisol was less than 80 or greater than 450 nmol/liter when ACTH deficiency or ACTH sufficiency, respectively, was diagnosed. Patients were considered GH deficient if the age-related IGF-I SDS was less than -3 or if they had three or more other pituitary hormone deficiencies. Patients were considered GH sufficient if age-related IGF-I SDS was greater than the 95th centile established from patients with known GH deficiency. Thirty-three underwent an ITT. MAIN OUTCOME MEASURES: The proportion of patients in whom ACTH and GH reserve could be assessed using OMT/IGF-I SDS was measured. The concordance with results was obtained from ITT. RESULTS: Fifty-five patients were ACTH sufficient and 45 were ACTH deficient. Twenty-one were GH sufficient and 33 were GH deficient based on IGF-I SDS and other pituitary hormone deficiencies, whereas 46 could not be classified. There was near-uniform concordance between OMT/IGF-I SDS and ITT. Initial investigation using OMT/IGF-I SDS resulted in a significant cost saving. CONCLUSIONS: ACTH and GH reserve can be accurately and cost-effectively investigated using OMT/IGF-I SDS in approximately 50% of patients with organic pituitary disorders.

Technology insight: measuring prolactin in clinical samples.

Measurement of prolactin is one of the most commonly undertaken hormonal investigations in evaluating patients with reproductive disorders. Hyperprolactinemia is found in up to 17% of such cases. Diagnostic evaluation of hyperprolactinemia is difficult but is facilitated by a logical approach where a thorough patient history is obtained, secondary causes of hyperprolactinemia are excluded, and the limitations of current prolactin assays are appreciated. Once hyperprolactinemia has been confirmed, attempts to establish the underlying cause can start. Given current workloads, laboratories rely on automated platforms to measure prolactin, most of which employ two-site immunoassay sandwich methods. Although generally robust and reliable, such immunoassays are susceptible to interference, and good collaboration between clinicians and the laboratory helps to minimize problems. A major challenge facing laboratories is correct differentiation of patients with true hyperprolactinemia from those with macroprolactinemia. Macroprolactin is a high-molecular-mass, biologically inactive form of prolactin that is detected to varying degrees by all prolactin immunoassays. Conservative estimates suggest that the presence of macroprolactin leads to misdiagnosis in as many as 10% of all reported instances of biochemical hyperprolactinemia. In the absence of specific testing, macroprolactin represents a diagnostic pitfall that results in the misdiagnosis and mismanagement of large numbers of patients.

Specificity and clinical utility of methods for the detection of macroprolactin.

BACKGROUND: Increased serum concentrations of macroprolactin are a relatively common cause of misdiagnosis and mismanagement of hyperprolactinemic patients. METHODS: We studied sera from a cohort of 42 patients whose biochemical hyperprolactinemia was explained entirely by macroprolactin. Using 5 pretreatments, polyethylene glycol (PEG), protein A (PA), protein G (PG), anti-human IgG (anti-hIgG), and ultrafiltration (UF), to deplete macroprolactin from sera before immunoassay, we compared residual prolactin concentrations with monomer concentrations obtained by gel-filtration chromatography (GFC). A monomeric prolactin standard was used to assess recovery and specificity of the pretreatment procedures. RESULTS: Residual prolactin concentrations in all pretreated sera differed significantly (P < 0.001) from monomeric concentrations obtained after GFC. PEG underestimated (mean, 75%), whereas PA, PG, anti-hIgG, and UF overestimated (means, 178%, 151%, 178%, and 112%, respectively) the amount of monomer present. Of the 5 methods examined, PEG correlated best with GFC (r = 0.80) followed by PG (r = 0.78), PA (r = 0.72), anti-hIgG (r = 0.70), and UF (r = 0.61). After UF or pretreatment with anti-hIgG or PEG, recovery of monomeric prolactin standard was low: 60%, 85%, and 77% respectively. In contrast, pretreatment with PA or PG gave almost quantitative recovery. CONCLUSIONS: None of the methods examined yielded results identical to the GFC method. PEG pretreatment yielded results that correlated best and is recommended as the first-choice alternative to GFC.

Coregulatory protein-orphan nuclear receptor interactions in the human adrenal cortex.

The capacity of the adrenal to produce steroids is controlled in part through the transcriptional regulation of steroid enzymes. The orphan nuclear receptor steroidogenic factor 1 (SF-1) is central to the transcriptional regulation of all steroid hydroxylase enzymes, whereas nur77 can preferentially regulate steroid enzyme genes relevant to cortisol production. We hypothesised that, in the presence of secretagogues, SF-1 and nur77 may differentially interact with coregulatory proteins in the human adrenal cortex. Both coregulatory proteins, steroid receptor coactivator (SRC-1) and silencing mediator for retinoid and thyroid hormones (SMRT), were found to be expressed in the zona fasciculata and reticularis in the human adrenal cortex, but were largely absent from the zona glomerulosa. Both coregulatory proteins were colocalised with SF-1 and nur77. In the H295R adrenal tumour cell line, SF-1 and nur77 transcripts were increased in cells in the presence of forskolin, whereas nur77 mRNA was also induced with angiotensin II (AII). The coactivator SRC-1 mRNA was increased in the presence of both forskolin and AII. Forskolin induced recruitment of SRC-1 to the SF-1 response element and induced SRC-1-SF-1 interactions, whereas AII increased recruitment of SRC-1 to the nur77 response element and induced SRC-1-nur77 interactions. The corepressor SMRT interacted with SF-1 in the presence of AII and with nur77 in cells treated with forskolin. Orphan nuclear receptor-coregulatory protein interactions may have consequences for the regulation of key steroidogenic enzymes in the human adrenal cortex.

Insulin-induced hypoglycemia suppresses plasma parathyroid hormone levels in patients with adrenal insufficiency.

Hypoglycemia has been reported to cause suppression of parathyroid hormone (PTH) levels in serum in normal subjects. It is possible that increasing cortisol levels in response to hypoglycemia was responsible. To examine this possibility the acute PTH response to insulin administration and resulting hypoglycemia was examined in patients with adrenal insufficiency. The possible acute impact of insulin-induced hypoglycemia on bone formation and bone resorption in the absence of an endogenous cortisol response was also examined. A prospective open study was undertaken to examine the acute effects of insulin and resulting hypoglycemia on PTH levels, on bone formation as indicated by serum levels of aminoterminal propeptide of type 1 procollagen (PINP), and on bone resorption as indicated by serum levels of beta carboxy terminal telopeptide of type 1 collagen (beta-CTx). Seven patients with adrenal insufficiency participated. These patients were studied on 3 occasions under different conditions: (1) when insulin was administered to induce hypoglycemia while the patients received their routine glucocorticoid replacement; (2) when the patients received their routine glucocorticoid replacement, but were not rendered hypoglycemic; and (3) when they did not receive glucocorticoid replacement and were not rendered hypoglycemic, ie, untreated. This facilitated isolation of the PTH response to insulin and hypoglycemia from the effects of the normal increase in endogenous cortisol levels in response to hypoglycemia. Blood samples were taken at baseline and after 3 hours while the subjects continued fasting for measurement of plasma glucose, serum ionized calcium (Cai), magnesium, phosphate, PINP, PTH, and beta-CTx. Insulin 0.075 IU/kg body weight was given intravenously after the first blood sample. The usual morning glucocorticoid replacement dose was given 20 minutes after the baseline blood sample was obtained. After the administration of insulin, plasma glucose decreased from 4.8 +/- 0.5 to 2.7 +/- 0.5 mmol/L, mean +/- SD (P < .0001). PTH was not influenced by time or glucocorticoid treatment, but decreased in response to insulin-induced hypoglycemia (P < .05). Serum levels of PINP and beta-CTx decreased when untreated between 9 AM and 12 PM (P < .05), but were not independently influenced by insulin-induced hypoglycemia or glucocorticoid treatment. Serum levels of Cai increased and serum phosphate levels decreased in response to insulin-induced hypoglycemia, while serum phosphate levels were also independently influenced by time decreasing between 9 AM and 12 PM (P < .05). There was no effect of time, insulin-induced hypoglycemia, or glucocorticoid treatment on serum levels of magnesium. Possible mechanisms involved in the acute decrease in serum PTH observed include a direct effect of insulin or hypoglycemia or an indirect effect, eg, increased sympathomimetic activity on PTH secretion or on calcium or phosphate intercompartmental shifts.

Modulation of steroidogenic enzymes by orphan nuclear transcriptional regulation may control diverse production of cortisol and androgens in the human adrenal.

The capacity of the adrenal to produce cortisol is controlled in part by 21-hydroxylase (CYP21) and the production of androgens by 17-hydroxylase/17-20-lyase (CYP17), in response to secretagogues including ACTH, angiotensin-II (A-II) and insulin. In this study we examined the capacity of human adrenocortical cells to produce cortisol and androgens in response to these secretagogues and their ability to regulate the expression of CYP21 and CYP17. In H-295 cells, forskolin and A-II were found to stimulate production of cortisol relative to androstenedione and a similar pattern of steroid production was noted in primary human adrenocortical cells. Both mRNA and protein expression of CYP21 was upregulated with forskolin and A-II alone and in combination, as detected by Northern and Western blotting. Whereas expression of CYP17 mRNA and protein was up regulated in the presence of forskolin and forskolin in combination with insulin. The ability of steroidogenic factor-1 (SF-1) and nur77 to regulate transcription of these enzymes was examined. Forskolin, A-II and insulin increased the protein expression of SF-1. Increased binding of SF-1 to its response element in the presence of forskolin, A-II and insulin was observed. Nur77 was expressed primarily in the zona glomerulosa and fasciculata. Increased protein expression of nur77 and the greatest binding of nur77 to its response element was seen when cells were stimulated with A-II in combination with forskolin. These data indicate that nur77 may preferentially regulate steroid enzyme genes relevant to cortisol production and thereby regulate differential cortisol and adrenal androgen production.

The impact of different glucocorticoid replacement schedules on bone turnover and insulin sensitivity in patients with adrenal insufficiency.

OBJECTIVE: Optimization of physiological replacement of glucocorticoid in patients with adrenal insufficiency is controversial. The present study was undertaken to compare the relative impact of three different regimes of glucocorticoid replacement in patients with adrenal insufficiency on parameters of bone turnover and insulin sensitivity. PATIENTS: Six female and three male patients with adrenal insufficiency and 17 female and 14 male control subjects participated. DESIGN: This was an open study conducted in a university teaching hospital. Schedule 1 (S1) consisted of hydrocortisone 10 mg with breakfast and 5 mg with lunch. S2 was similar to S1 with the addition of 5 mg hydrocortisone with the evening meal. S3 utilized dexamethasone 0.1 mg/15 kg body weight given per day with breakfast only. Each schedule was given for at least 4 weeks in random sequence to nine patients with adrenal insufficiency. METHODS: Blood was obtained at 0900 h (fasting) and at 1300 h for measurement of the ionized calcium (Cai), PTH, 25-hydroxyvitamin D and the bone formation markers intact osteocalcin and amino-terminal propeptide of type 1 procollagen (PINP). Timed urine collections were made under standardized conditions, that is while fasting between 0700 and 0900 h (basal) and between 0900 and 1300 h for measurement of the bone resorption markers, free deoxypyridinoline (FDPD) and cross-linked N-telopeptide of type 1 collagen (NTX). Blood was drawn for measurement of fasting plasma glucose and serum insulin levels. Insulin (0.075 IU/kg) was administered i.v. while the patient was fasting prior to the first glucocorticoid replacement dose on each study day. Plasma glucose was measured before and 3, 6, 9, 12 and 15 min after insulin administration to calculate the glucose disappearance rate (Kitt). Insulin resistance (IR) and beta-cell function were estimated using the homeostasis model assessment (HOMA). Glucocorticoid dosage was given according to the various schedules at approximately 0930 h. RESULTS: During all three treatment schedules the serum Cai level was significantly lower than that seen in control subjects. PTH levels in patients taking the three replacement schedules and in normal subjects were similar. Serum 25-hydroxyvitamin D levels were not suppressed in the patients during any of the three treatment schedules. The bone resorption marker urinary FDPD under basal conditions was significantly lower during S3 (dexamethasone) than during either hydrocortisone schedules, S1 or S2. Urinary NTX values were not significantly different in the three study groups. The bone formation markers intact osteocalcin and PINP were similar in the three replacement schedules. The indices of IR and beta-cell function tended to be higher during treatment with dexamethasone than with S1 or S2 but did not achieve statistical significance. CONCLUSIONS: These data indicate that all three replacement schedules were associated with low serum ionized calcium levels without evidence of a compensatory increase in PTH levels. These findings are consistent with direct or indirect suppression of the bone remodelling cycle and suppression of PTH levels. Bone turnover in patients with adrenal insufficiency treated with schedule 3, dexamethasone, was associated with lower bone turnover than patients treated with hydrocortisone schedules 1 or 2. While indices of insulin sensitivity measured during schedules 1, 2 and 3 did not achieve statistical significance, there was an obvious trend for greater insulin resistance to occur with schedules 3 using dexamethasone.

Frequent misdiagnosis and mismanagement of hyperprolactinemic patients before the introduction of macroprolactin screening: application of a new strict laboratory definition of macroprolactinemia.

BACKGROUND: Macroprolactin (big big prolactin) has reduced bioactivity and is measured by immunoassays for prolactin when it accumulates in the plasma of some individuals. We applied normative data for serum prolactin after treatment of sera to remove macroprolactin to elucidate the contribution of macroprolactin to misleading diagnoses, inappropriate investigations, and unnecessary treatment. METHODS: We reviewed records of women attending a tertiary referral center who had prolactin >1000 mIU/L. Application of a reference interval to polyethylene glycol (PEG)-treated hyperprolactinemic sera identified 21 patients in whom hyperprolactinemia was accounted for entirely by the presence of macroprolactin. Presenting clinical features, diagnoses, and treatment were compared in these patients and 42 age-matched true hyperprolactinemic patients. RESULTS: Prolactin concentrations in sera of 110 healthy individuals ranged from 78 to 564 mIU/L. The range of values for the sera after PEG treatment was 70-403 mIU/L. For macroprolactinemic samples, PEG treatment decreased mean (SD) prolactin from 1524 (202) mIU/L to 202 (27) mIU/L but decreased it only from 2096 (233) mIU/L to 1705 (190) mIU/L in true hyperprolactinemic patients (P <0.01 between groups). Oligomenorrhea or amenorrhea and galactorrhea were the most common clinical features in both groups, although they occurred more frequently in true hyperprolactinemic patients (P <0.05). Serum estradiol and luteinizing hormone concentrations were significantly higher in participants with macroprolactinemia than in those with true hyperprolactinemia (P <0.05). Among participants with retrospectively identified macroprolactinemia, pituitary imaging was performed in 93% and treatment with dopamine agonist was prescribed in 87%. CONCLUSIONS: Macroprolactin is a significant cause of misdiagnosis, unnecessary investigation, and inappropriate treatment. The use of an appropriate reference interval for the PEG immunoprecipitation procedure may be of particular importance in those patients who have an excess of both macroprolactin and monomeric prolactin.

Gross variability in the detection of prolactin in sera containing big big prolactin (macroprolactin) by commercial immunoassays.

A high molecular mass form of prolactin (PRL), macroprolactin, accumulates in the sera of some subjects. Although macroprolactin exhibits limited bioactivity in vivo, it retains immunoreactivity. We examined the frequency of macroprolactinemia in clinical practice and the ability of immunoassay systems to distinguish between macroprolactin and monomeric PRL. Of 300 hyperprolactinemic sera identified, 71 normalized following treatment of sera with polyethylene glycol, indicating that 24% of hyperprolactinemia could be accounted for by macroprolactin. Ten of these macroprolactinemic sera were circulated to 18 clinical laboratories. Two sets of PRL measurements of the 10 untreated sera were obtained from each of the nine most commonly used immunoassay systems. Across the nine assay systems, differences in the PRL estimates ranged from 2.3- to 7.8-fold. Elecsys users reported the highest PRL levels. Somewhat lower values were reported for DELFIA systems followed by Immuno-1, AxSYM, and Architect assays. The Immulite 2000 assay generated PRL levels equivalent to approximately 50% of those reported by the high-reading methods. The lowest PRL levels were reported by Access, ACS:180, and Centaur systems. To avoid confusion caused by the frequent presence of macroprolactin accounting for hyperprolactinemia, secondary screening for the presence of macroprolactin is recommended.

The low-dose ACTH test does not provide a useful assessment of the hypothalamic-pituitary-adrenal axis in secondary adrenal insufficiency.

OBJECTIVE: The 1 microg ACTH stimulation test has been introduced to improve the sensitivity of ACTH as a test of the integrity of hypothalamic-pituitary-adrenal axis (HPAA). This study aims to compare the sensitivity, specificity and diagnostic accuracy of the "low-dose" 1 microg ACTH (LDACTH) test and the "standard dose" 250 microg ACTH (SDACTH) test, with the overnight metyrapone test (OMT) which assesses the entire HPAA. DESIGN: A prospective evaluation of the performance of SDACTH and LDACTH screening tests in a diverse cohort of patients with possible adrenal insufficiency as routinely encountered in clinical practice using the OMT as the reference method. PATIENTS: A total of 51 patients (26 with asthma on inhaled glucocorticoid, nine with hypopituitarism, three with hypothyroidism, one with hyponatraemia, one with Crohn's disease, one with encephalitis and 10 with non-specific symptoms) each underwent SDACTH, LDACTH and OMT tests in random sequence at least 1 week apart. MEASUREMENTS: Blood was sampled for plasma cortisol levels at 0 and 30 min after intravenous administration of 1 microg and 250 microg of ACTH. Metyrapone 30 mg/kg body weight was taken orally at midnight, and plasma samples were taken for measurement of 11-deoxycortisol and cortisol next morning between 08.00 and 09.00 h. The OMT was deemed to be abnormal when both 11-deoxycortisol and cortisol levels were less than 200 nmol/l. RESULTS: The sensitivity and specificity at an empirical "normal" plasma cortisol threshold value of 500 nmol/l were 67% and 100% for the SDACTH test, and 73% and 81% for the LDACTH test, respectively. As the plasma cortisol cut-off value was increased to 550 nmol/l and 600 nmol/l, the sensitivity of the SDACTH test was 67% and 80% and specificity was 97% and 92%, respectively. The sensitivity of the LDACTH test increased from 93% at plasma cortisol cut-off value of 550 nmol/l to 100% at plasma cortisol cut-off value of 600 nmol/l. However, the specificity of the LDACTH test fell from 72% to 56% as the plasma cortisol cut-off value was increased from 550 nmol/l to 600 nmol/l. A receiver operating characteristic curve demonstrated that the specificity of the SDACTH test was higher than the specificity of the LDACTH test at any given level of sensitivity. CONCLUSIONS: Both the LDACTH and SDACTH tests fail to achieve acceptable levels of sensitivity and specificity to be useful as screening tests for secondary adrenal insufficiency. In this context the OMT can be safely used to assess the integrity of the entire HPAA.