The salivary cortisol concentration is an excellent indicator of the plasma free cortisol concentration. To establish its normal and pathological ranges, salivary cortisol concentrations were measured in 101 normal adults, 18 patients with Cushing's syndrome, and 21 patients with adrenal insufficiency. The normal subjects had a mean (+/- SEM) salivary cortisol concentration of 15.5 +/- 0.8 nmol/L (range, 10.2-27.3) at 0800 h and 3.9 +/- 0.2 nmol/L (range, 2.2-4.1) at 2000 h (n = 20). The mean value 60 min after ACTH administration in 58 normal subjects was 52.2 +/- 2.2 nmol/L (range, 23.5-99.4), and it was 1.4 +/- 1.1 nmol/L (range, 1.6-3) at 0800 h in 23 normal subjects given 1 mg dexamethasone 8 h earlier. In patients with primary or secondary adrenal insufficiency (n = 21) the mean salivary cortisol level was 7.5 +/- 0.4 nmol/L (range, 1.9-21.8) 60 min after ACTH. In patients with Cushing's syndrome (n = 7), the mean value after the 1-mg dexamethasone suppression test was 16.1 +/- 7.8 nmol/L (range, 5.8-66.8). No overlap was found between the values in the normal subjects and those in the patients during the dynamic tests. Discrepancies between salivary and total plasma cortisol were found in 8 patients with adrenal insufficiency, which may be explained by the effects of drugs such as thyroid hormones, Op'-dichlorodiphenyldichloroethane, and psychotropic agents. We conclude that salivary cortisol measurements are an excellent index of plasma free cortisol concentrations. They circumvent the physiological, pathological, and pharmacological changes due to corticosteroid-binding globulin alterations and offer a practical approach to assess pituitary-adrenal function.

The overnight 1-mg dexamethasone suppression test is a very good screening test for subjects suspected of having Cushing's syndrome. To simplify the procedure, we evaluated the 1-mg dexamethasone suppression test with measurement of salivary cortisol. We performed this test with plasma and salivary cortisol measurements in 27 patients with Cushing's syndrome and 64 normal controls. The sensitivity and specificity of plasma cortisol measurements were 100% and 87%, respectively, for a cut-off point of 100 nmol/l, in accordance with previous studies. The results of salivary cortisol showed the absence of overlap between the two groups, with a sensitivity and specificity of 100% for a cut-off point of 2.8 nmol/l. On a larger series, however, one might occasionally miss the diagnosis of a patient with Cushing's syndrome. Therefore, we favor a cut-off point of 1.9 nmol/l, the sensitivity remaining at 100% and the specificity being 94%. In conclusion we recommend the overnight dexamethasone suppression test with measurement of salivary cortisol as a screening test for Cushing's syndrome.

Salivary cortisol is an excellent indicator of the plasma free cortisol concentration in normal and pathological situations. We took advantage of its ease of sampling, allowing multiple collections at home, to follow the course of a patient with Cushing's disease living in North Africa. This 48-year-old woman presented with a clinically moderate hypercortisolism caused by a large basophilic pituitary adenoma. Bilateral extension to the cavernous sinuses precluded surgical therapy. She went into spontaneous remission based on clinical signs as well as biochemical findings. During the following 2 years she demonstrated intermittent relapses that were treated by radiotherapy (50 Gy), followed by ketoconazole and then o-paraprime-dichloro-diphenyl-dichloroethane (Op'DDD). After a prolonged clinical remission, Cushing's syndrome again became active. Bromocriptine was started without effect and a new treatment with Op'DDD was began. Evaluation and follow-up were performed during hospitalizations and mainly through the measurements of salivary cortisol in more than 100 samples sent from North Africa by air mail to our department in Paris. Thus we were able to demonstrate intermittent overproduction of cortisol before any treatment, with periods of normal and even low values, and to follow the efficacy of therapy and to detect the relapses. We conclude that measurement of salivary cortisol is a valuable tool in difficult clinical situations such as intermittent hypercortisolism and remoteness between the patient and hospital.

Objective: Measurement of plasma ACTH is a key step for the exploration of hypothalamic±pituitary± adrenal disorders. To further improve ACTH recognition a new generation of ACTH IRMA was developed using antibodies directed towards succinylated ACTH (sACTH IRMA). Design: The usefulness of this assay was compared with that of another commercially available ACTH IRMA assay using intact ACTH (ELSA-ACTH) in various pathophysiological situations: patients with low ACTH plasma levels, high ACTH plasma levels with normal or tumoural pituitaries, or ectopic ACTH syndrome, and pregnant women with high proopiomelanocortin (POMC) plasma levels. Methods: All plasma samples were assayed simultaneously with the two different IRMAs. Comparisons were assessed by plotting the results along the theoretical line of identical values, and by the graphical method of Bland and Altman. Results: In the ELSA-ACTH IRMA, CLIP (or ACTH 18-39 ) showed true cross-reactivity, and amelanocyte-stimulating hormone and purified POMC both interfered and induced falsely lower ACTH results; in the sACTH IRMA no peptide showed any cross-reactivity, and only extremely high values of CLIP (50 000 pg/ml) interfered and induced falsely lower ACTH results. In ACTH hypersecretory syndromes, of tumoural (Cushing's disease, ectopic ACTH secretion) or non-tumoural (Addison's disease, congenital adrenal hyperplasia) origins a good agreement between the two assays was observed except for very high ACTH plasma values (above 1000 pg/ml) and in some tumours where the sACTH IRMA yielded lower results; in some cases, the presence of circulating CLIP, demonstrated by HPLC studies, may contribute to this discrepancy. It is also likely that the calibration of the ELSA-ACTH kit itself generates higher ACTH values. In normal pregnant women both IRMAs gave highly correlated values, yet lower results were obtained with the sACTH IRMA. Conclusion: These data show that the sACTH IRMA has improved qualities of specificity and usefulness for rapid assessment of ACTH plasma levels.

No abstract