Suppressed ACTH Is Frequently Unrelated to Autonomous Cortisol Secretion in Patients With Adrenal Incidentalomas

The Journal of Clinical Endocrinology & Metabolism, Feb 2019

ACTH is considered a weak marker for autonomous cortisol secretion (ACS) in patients with adrenal incidentalomas (AIs). Our aim was to investigate suppressed basal ACTH as a marker of ACS and to elucidate why this criterion is of limited value.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://academic.oup.com/jcem/article-pdf/104/2/506/27378241/jc.2018-01029.pdf

Suppressed ACTH Is Frequently Unrelated to Autonomous Cortisol Secretion in Patients With Adrenal Incidentalomas

Abstract Objective ACTH is considered a weak marker for autonomous cortisol secretion (ACS) in patients with adrenal incidentalomas (AIs). Our aim was to investigate suppressed basal ACTH as a marker of ACS and to elucidate why this criterion is of limited value. Methods Basal ACTH and cortisol after overnight dexamethasone suppression test (cortisolONDST) were measured in 198 patients with unilateral AI and at 2-year follow-up. Basal ACTH was measured in 100 control subjects. Results In patients with cortisolONDST <50 nmol/L (n = 145), ACTH was <2 pmol/L in 19%, compared with 4% in control subjects (P < 0.001). ACTH and size of AI correlated negatively (P = 0.002). Among patients with cortisolONDST ≥50 nmol/L, ACTH was <2 pmol/L in 53%. The patients were grouped according to whether cortisolONDST was <50 or ≥50 nmol/L and whether ACTH was <2.0 or ≥2.0 or pmol/L. At follow-up, these four groups were still separated with statistically significant differences in ACTH and cortisolONDST. Conclusions This study identifies a previously unrecognized group of patients defined by suppressed ACTH despite normal cortisolONDST. This suppression of ACTH by a factor other than ACS may explain the limitation of suppressed ACTH as a marker for ACS. We suggest increased cortisol secretion in response to ACTH by the AI to be an additional factor. The need to identify subclinical hypercortisolism in patients with adrenal incidentalomas (AIs) has recently been underlined because the condition has been linked to increased cardiovascular morbidity and mortality (1–3). However, it has been difficult to establish generally accepted biochemical criteria, as demonstrated by the different definitions used to define subclinical hypercortisolism (4–6). In the European Society of Endocrinology (ESE) guidelines issued in 2016, the term “subclinical hypercortisolism” has been replaced with “autonomous cortisol secretion” (ACS). The diagnosis of ACS is based primarily on increased cortisol after 1-mg overnight dexamethasone suppression test (cortisolONDST) (7). Patients with cortisolONDST ≥140 nmol/L are considered to have ACS, and patients with levels between 50 and 140 nmol/L are considered to have possible ACS. The ESE guidelines recommend further hormonal evaluation of patients with cortisolONDST ≥140 nmol/L and of patients with cortisolONDST between 50 and 140 nmol/L and metabolic complications (7). It is suggested that ACS should be verified by measuring ACTH, urinary cortisol, and possibly late-evening salivary cortisol, but no guidance is given regarding the evaluation of these complementary hormonal tests. Consequently, there is no generally accepted approach for verification or exclusion of ACS in patients with cortisolONDST ≥50 nmol/L. There are several possible explanations why it is difficult to establish reliable criteria for the diagnosis of ACS. The main marker of ACS, cortisolONDST, may be elevated also by nonadrenal factors, such as low dexamethasone concentration, advanced age, and impaired renal function (8–10). Furthermore, suppressed basal ACTH has low reliability as a marker of ACS (7, 11). This low reliability could be explained by the circadian rhythm of the hypothalamic-pituitary-adrenal (HPA) axis activity, the pulsatile pattern of ACTH secretion, and the inability of some ACTH assays to detect low levels (12, 13). Furthermore, factors other than ACS may influence ACTH levels, such as increased cortisol response to ACTH (14, 15). Thus, we investigated why suppressed basal ACTH is of limited value as a marker of ACS. Subjects and Methods Consecutive patients examined for unilateral AI at the Department of Internal Medicine, Helsingborg Hospital, Sweden, between 1 January 2005 and 15 September 2015, were considered. Exclusion criteria were metastatic malignancy, AI <1 cm, nonadenoma lesions such as cysts and hemorrhage, biochemical results supporting pheochromocytoma, primary hyperaldosteronism or clinical Cushing syndrome, oral corticoid treatment with more than single doses in the past 3 months, use of inhalation steroids or medication affecting dexamethasone metabolism, body mass index <19.0 kg/m2, age >75 years, and estimated glomerular filtration rate <60 mL/min/1.73 m2. Medical history was collected from patient records. Basal blood samples were collected at 8:00 am to measure cortisol, ACTH, dehydroepiandrosterone (DHEAS), aldosterone, and renin. Patients were given 1 mg dexamethasone orally at 11:00 pm, and cortisol levels were determined at 8:00 am the following morning (cortisolONDST). Primary aldosteronism was excluded using the aldosterone/renin ratio. Pheochromocytoma was determined by determining catecholamines and metoxycatecholamines in two 24-hour urinary samples or plasma metanephrines. A second dexamethasone suppression test was performed after ∼2 years. After September 2015, further data collection was inhibited due to an exchange of our cortisol assay. One hundred blood donors served as a control group. ACTH was analyzed in blood collected between 7:00 and 9:00 am. The study was approved by the Ethics Committee, Lund, Sweden. Imaging AI size was defined as the maximal axial diameter on CT scans. Assays Plasma cortisol was analyzed using a one-step competitive immunoassay (Roche Diagnostics, Mannheim, Germany). The reference range was (171 to 536 nmol/L), the coefficient of variation (CV) was 2.1% at 94.9 nmol/L, and the detection limit was 0.5 nmol/L. Plasma ACTH was analyzed using a two-step immunometric sandwich assay (Cobas®, Roche Diagnostics). The reference range was 1.6 to 13.9 pmol/L, the CV was 5.4% at 1.1 pmol/L, and the detection limit was 0.23 pmol/L. DHEAS was analyzed using a two-step immunometric competitive assay (Cobas). The reference ranges of DHEAS were as follows: for female patients, 2.7 to 9.2 µmol/L (25 to 34 years), 1.6 to 9.2 µmol/L (35 to 44 years), 0.96 to 7.0 µmol/L (45 to 54 years), 0.51 to 5.6 µmol/L (55 to 64 years), 0.26 to 6.7 µmol/L (65 to 74 years), and 0.33 to 4.2 µmol/L (≥75 years); for male patients, 4.3 to 12 µmol/L (25 to 34 years), 2.4 to 12 µmol/L (35 to 44 years), 1.2 to 9.0 µmol/L (45 to 54 years), 1.4 to 8.0 µmol/L (55 to 64 years), 0.91 to 6.8 µmol/L (65 to 74 years), and 0.44 to 3.3 µmol/L (≥75 years). The CV was 6% at 7.0 µmol/L. The detection limit was 0.003 µmol/L, but all results <0.81 µmol/L were given as <0.81 µmol/L. Plasma creatinine was analyzed using an IDMS standardized enzymatic colorimetric assay (Cobas). Reference ranges were 60 to 105 µmol/L for male patients and 45 to 90 µmol/L for female patients, and the CV at 70 µmol/L was 1.4%. Data analysis Suppressed ACTH was defined as a level <2.0 pmol/L. Statistics Statistical analyses were performed using SPSS 22 (IBM, Chicago, IL). All hormone levels and anthropometric measures in patients with AI were nonnormally distributed. Continuous variables are given as medians and ranges. Multivariate linear regression analysis was used to compare ACTH in the four groups of patients vs control subjects, adjusted for differences in sex. The reference interval for ACTH in control subjects was calculated by the bootstrapping method in RefVal 4.11 (16). Continuous data were compared with the Mann-Whitney test and categorical data with the χ2 test or Fisher exact test as appropriate. Linear regression analysis was used to determine the correlation between cortisol and cortisolONDST to ACTH. Multivariate linear regression analysis was used to determine the difference in DHEAS between groups of patients, adjusted for differences in sex and age. Multivariate logistic regression analysis was used to examine the prevalence of treated hypertension, treated diabetes, and treated dyslipidemia in the patient groups, adjusted for differences in age and sex. Among patients with cortisolONDST <50 pmol/L, univariate linear regression was used to investigate the correlation between ACTH and the size of the AI. ACTH, DHEAS, cortisol, and cortisolONDST were ln-transformed in all the regression analyses. A P value <0.05 was considered statistically significant. Results We screened 375 patients with unilateral AIs, 177 of whom were excluded based on the exclusion criteria, leaving 198 patients included in the analyses. CortisolONDST was ≥50 nmol/L, and ACTH was <2.0 pmol/L in 53 (27%) and 56 (28%) of the patients, respectively. Five patients had cortisolONDST >140 nmol/L. Cortisol correlated positively with ACTH (r = 0.19, P = 0.008). Figure 1 shows ACTH in relation to cortisolONDST. CortisolONDST was negatively correlated to ACTH. ACTH levels in patients with cortisolONDST <50 nmol/L and in control subjects are shown in Fig. 2. Patients with cortisolONDST <50 nmol/L had lower ACTH than control subjects (P < 0.001), also if adjusted for differences in sex (P < 0.001). Age was omitted from this analysis because ACTH was not correlated to age. In patients with cortisolONDST <50 nmol/L, ACTH was below the lower reference range stated by the manufacturer (1.6 pmol/L) in 12% and not above the higher reference range in any. The prevalence of ACTH <2.0 pmol/L was higher in patients with cortisolONDST <50 nmol/L compared with control subjects (19% vs 4%; P < 0.001). The prevalence of ACTH <1.8 pmol/L was also higher in patients than in control subjects (16% vs 1%; P < 0.001). Patients with cortisolONDST <40 nmol/L had a higher prevalence of ACTH <2.0 than control subjects (16% vs 4%; P < 0.001). This difference in prevalence of ACTH <2.0 pmol/L was present also after adjusting for sex at both cortisolONDST cutoff levels (P = 0.002 and P = 0.02, respectively). Figure 3 shows ACTH in relation to AI size in the 145 patients with cortisolONDST <50 nmol/L. ACTH was negatively correlated to AI size (r = −0.25, P = 0.002). Figure 1. View largeDownload slide CortisolONDST in relation to ACTH in patients with AI. CortisolONDST correlated negatively with ACTH (r = −0.45, P < 0.001, ln-transformed parameters). The prevalence of cortisolONDST ≥50 nmol/L increased with decreasing ACTH over the whole ACTH scale, and cortisolONDST ≥50 nmol/L may be present also at ACTH levels >6 pmol/L. One patient (ACTH 19.0 pmol/L, cortisolONDST 67 nmol/L) is not depicted. Figure 1. View largeDownload slide CortisolONDST in relation to ACTH in patients with AI. CortisolONDST correlated negatively with ACTH (r = −0.45, P < 0.001, ln-transformed parameters). The prevalence of cortisolONDST ≥50 nmol/L increased with decreasing ACTH over the whole ACTH scale, and cortisolONDST ≥50 nmol/L may be present also at ACTH levels >6 pmol/L. One patient (ACTH 19.0 pmol/L, cortisolONDST 67 nmol/L) is not depicted. Figure 2. View largeDownload slide Distribution of ACTH in 145 patients with cortisolONDST <50 nmol/L and 100 control subjects. ACTH was lower in patients with cortisolONDST <50 nmol/L than in control subjects (P < 0.001). The patients had an increased prevalence of low ACTH levels and an absence of ACTH levels >9.0 pmol/L. Figure 2. View largeDownload slide Distribution of ACTH in 145 patients with cortisolONDST <50 nmol/L and 100 control subjects. ACTH was lower in patients with cortisolONDST <50 nmol/L than in control subjects (P < 0.001). The patients had an increased prevalence of low ACTH levels and an absence of ACTH levels >9.0 pmol/L. Figure 3. View largeDownload slide ACTH in relation to size of AI in 145 patients with cortisolONDST <50 nmol/L. The dashed line indicates ACTH = 2.0 pmol/L. The solid line is the regression fit to the data. ACTH was negatively correlated to sizes of the AIs (r = −0.25, P = 0.002). The correlation was similar if the patient with a 51-mm AI was considered an outlier and excluded from the calculation (r = −0.27, P = 0.001). Figure 3. View largeDownload slide ACTH in relation to size of AI in 145 patients with cortisolONDST <50 nmol/L. The dashed line indicates ACTH = 2.0 pmol/L. The solid line is the regression fit to the data. ACTH was negatively correlated to sizes of the AIs (r = −0.25, P = 0.002). The correlation was similar if the patient with a 51-mm AI was considered an outlier and excluded from the calculation (r = −0.27, P = 0.001). The control subjects were 45 (range, 19 to 70) years of age, and 52% were male. ACTH was lower in female controls than in male controls [3.9 pmol/L (range, 1.8 to 17.4 pmol/L) and 5.2 pmol/L (range, 1.6 to 18.2 pmol/L), respectively; P = 0.002] but was not correlated to age. ACTH was <2.0 pmol/L, which is the level defined as suppressed ACTH, in 4% of the control subjects. Calculated reference intervals for male and female controls were 1.7 to 17.3 and 1.8 to 15.6 pmol/L, respectively. The 90% CI for the lower reference intervals for male and female controls were 1.6 to 2.2 and 1.8 to 2.1 pmol/L, respectively. The non–sex-specific ACTH reference interval was 1.7 to 16.8 pmol/L (90% CI, 1.6 to 2.2 and 12.3 to 18.2 pmol/L, respectively). All included patients were divided into four subgroups, named NormalACTH/ONDST<50, LowACTH/ONDST<50, NormalACTH/ONDST≥50, and LowACTH/ONDST≥50, according to whether ACTH was <2.0 or ≥2.0 or pmol/L and whether cortisolONDST was <50 or ≥50 nmol/L. Patient characteristics are given in Table 1. ACTH was lower in all four groups compared with control subjects adjusted for differences in sex (P = 0.002, P < 0.001, P = 0.004, and P < 0.001). Both male and female subjects in the LowACTH/ONDST<50 group had lower DHEAS than male and female subjects in the NormalACTH/ONDST<50 Group (P = 0.03 and P = 0.02, respectively). Five patients were on systemic estrogen treatment: two in the NormalACTH/ONDST<50, one in the NormalACTH/ONDST≥50, and two in the LowACTH/ONDST≥50 group. If these patients are omitted from the calculations, DHEAS, adjusted for age and sex, and ACTH at follow-up were still lower in the LowACTH/ONDST<50 group than in the NormalACTH/ONDST<50 group (P = 0.002 and P < 0.001, respectively). In the NormalACTH/ONDST<50 group and the LowACTH/ONDST<50 group, the prevalence of treated hypertension, diabetes, and dyslipidemia were 49 vs 18%, 16 vs 11%, and 28 vs 14% (P = 0.02, P = 0.68, and P = 0.24, respectively, adjusted for differences in age). Table 1. Anthropometric, Biochemical, Hormonal, and Radiological Characteristics of the 198 Patients with AI Divided Into Groups According to Whether Their CortisolONDST Level Was <50 or ≥50 nmol/L and Whether Their ACTH Level Was <2.0 or ≥2.0 or pmol/L Characteristic All Patients Normal ACTH/ONDST<50 Group (cortisolONDST <50, ACTH ≥2.0) LowACTH/ONDST<50 Group (cortisolONDST <50, ACTH <2.0) NormalACTH/ONDST≥50 Group (cortisolONDST ≥50, ACTH ≥2.0) LowACTH/ONDST≥50 Group (cortisolONDST ≥50, ACTH <2.0) n, % 198 (100) 117 (59) 28 (14) 25 (13) 28 (14) Female, % 54 47 64 44 79a Age, y 62.8 (24.7–74.9) 62.8 (24.7–74.2) 58.3b (44.4–70.1) 64.3 (34.7–74.9) 64.7 (30.4–72.8) BMI, kg/m2 27.8 (19.0–43.9) 27.7 (20.6–43.8) 27.6 (21.8–41.7) 28.1 (20.3–43.9) 28.0 (19.0–43.0) eGFR, mL/min/1.73 m2 86 (60–151) 87 (60–151) 87 (66–117) 84 (63–143) 84 (61–115) AI size, cm 2.0 (1.0–5.1) 1.8 (1.0–5.1) 2.2b (1.2–3.8) 2.1 (1.5–4.0) 2.3b (1.0–4.5) Cortisol, nmol/L 445 (131–1007) 444 (131–967) 402b (157–610) 504 (307–973) 484 (177–1007) ACTH, pmol/L 3.0 (0.23–19.0) 3.9 (2.0–8.6) 1.5c (0.34–1.9) 3.2b (2.0–19.0) 1.3c (0.23–1.9) DHEAS, µmol/L 1.9 (<0.81–11.0) 2.4 (<0.81–10.0) 1.5a (<0.81–3.8) 1.8 (<0.81–5.8) 1.3b (<0.81–11.0) Cortisol/ACTH ratio, nmol/L/pmol/L 150 (32–2078) 114 (32–312) 267c (129–812) 164a (45–274) 349c (104–2078) CortisolONDST, nmol/L 37 (12–310) 30 (12–48) 38c (28–49) 67c (50–310) 77c (51–157)  Characteristic All Patients Normal ACTH/ONDST<50 Group (cortisolONDST <50, ACTH ≥2.0) LowACTH/ONDST<50 Group (cortisolONDST <50, ACTH <2.0) NormalACTH/ONDST≥50 Group (cortisolONDST ≥50, ACTH ≥2.0) LowACTH/ONDST≥50 Group (cortisolONDST ≥50, ACTH <2.0) n, % 198 (100) 117 (59) 28 (14) 25 (13) 28 (14) Female, % 54 47 64 44 79a Age, y 62.8 (24.7–74.9) 62.8 (24.7–74.2) 58.3b (44.4–70.1) 64.3 (34.7–74.9) 64.7 (30.4–72.8) BMI, kg/m2 27.8 (19.0–43.9) 27.7 (20.6–43.8) 27.6 (21.8–41.7) 28.1 (20.3–43.9) 28.0 (19.0–43.0) eGFR, mL/min/1.73 m2 86 (60–151) 87 (60–151) 87 (66–117) 84 (63–143) 84 (61–115) AI size, cm 2.0 (1.0–5.1) 1.8 (1.0–5.1) 2.2b (1.2–3.8) 2.1 (1.5–4.0) 2.3b (1.0–4.5) Cortisol, nmol/L 445 (131–1007) 444 (131–967) 402b (157–610) 504 (307–973) 484 (177–1007) ACTH, pmol/L 3.0 (0.23–19.0) 3.9 (2.0–8.6) 1.5c (0.34–1.9) 3.2b (2.0–19.0) 1.3c (0.23–1.9) DHEAS, µmol/L 1.9 (<0.81–11.0) 2.4 (<0.81–10.0) 1.5a (<0.81–3.8) 1.8 (<0.81–5.8) 1.3b (<0.81–11.0) Cortisol/ACTH ratio, nmol/L/pmol/L 150 (32–2078) 114 (32–312) 267c (129–812) 164a (45–274) 349c (104–2078) CortisolONDST, nmol/L 37 (12–310) 30 (12–48) 38c (28–49) 67c (50–310) 77c (51–157)  Statistical differences were calculated between the NormalACTH/ONDST<50 group and the other three groups. Statistical differences for DHEAS were adjusted for differences in age and sex. Abbreviations: BMI, body mass index; eGFR, estimated glomerular filtration rate. a P < 0.01 compared with the NormalACTH/ONDST<50 group. b P < 0.05 compared with the NormalACTH/ONDST<50 group. c P < 0.001 compared with the NormalACTH/ONDST<50 group. View Large Table 1. Anthropometric, Biochemical, Hormonal, and Radiological Characteristics of the 198 Patients with AI Divided Into Groups According to Whether Their CortisolONDST Level Was <50 or ≥50 nmol/L and Whether Their ACTH Level Was <2.0 or ≥2.0 or pmol/L Characteristic All Patients Normal ACTH/ONDST<50 Group (cortisolONDST <50, ACTH ≥2.0) LowACTH/ONDST<50 Group (cortisolONDST <50, ACTH <2.0) NormalACTH/ONDST≥50 Group (cortisolONDST ≥50, ACTH ≥2.0) LowACTH/ONDST≥50 Group (cortisolONDST ≥50, ACTH <2.0) n, % 198 (100) 117 (59) 28 (14) 25 (13) 28 (14) Female, % 54 47 64 44 79a Age, y 62.8 (24.7–74.9) 62.8 (24.7–74.2) 58.3b (44.4–70.1) 64.3 (34.7–74.9) 64.7 (30.4–72.8) BMI, kg/m2 27.8 (19.0–43.9) 27.7 (20.6–43.8) 27.6 (21.8–41.7) 28.1 (20.3–43.9) 28.0 (19.0–43.0) eGFR, mL/min/1.73 m2 86 (60–151) 87 (60–151) 87 (66–117) 84 (63–143) 84 (61–115) AI size, cm 2.0 (1.0–5.1) 1.8 (1.0–5.1) 2.2b (1.2–3.8) 2.1 (1.5–4.0) 2.3b (1.0–4.5) Cortisol, nmol/L 445 (131–1007) 444 (131–967) 402b (157–610) 504 (307–973) 484 (177–1007) ACTH, pmol/L 3.0 (0.23–19.0) 3.9 (2.0–8.6) 1.5c (0.34–1.9) 3.2b (2.0–19.0) 1.3c (0.23–1.9) DHEAS, µmol/L 1.9 (<0.81–11.0) 2.4 (<0.81–10.0) 1.5a (<0.81–3.8) 1.8 (<0.81–5.8) 1.3b (<0.81–11.0) Cortisol/ACTH ratio, nmol/L/pmol/L 150 (32–2078) 114 (32–312) 267c (129–812) 164a (45–274) 349c (104–2078) CortisolONDST, nmol/L 37 (12–310) 30 (12–48) 38c (28–49) 67c (50–310) 77c (51–157)  Characteristic All Patients Normal ACTH/ONDST<50 Group (cortisolONDST <50, ACTH ≥2.0) LowACTH/ONDST<50 Group (cortisolONDST <50, ACTH <2.0) NormalACTH/ONDST≥50 Group (cortisolONDST ≥50, ACTH ≥2.0) LowACTH/ONDST≥50 Group (cortisolONDST ≥50, ACTH <2.0) n, % 198 (100) 117 (59) 28 (14) 25 (13) 28 (14) Female, % 54 47 64 44 79a Age, y 62.8 (24.7–74.9) 62.8 (24.7–74.2) 58.3b (44.4–70.1) 64.3 (34.7–74.9) 64.7 (30.4–72.8) BMI, kg/m2 27.8 (19.0–43.9) 27.7 (20.6–43.8) 27.6 (21.8–41.7) 28.1 (20.3–43.9) 28.0 (19.0–43.0) eGFR, mL/min/1.73 m2 86 (60–151) 87 (60–151) 87 (66–117) 84 (63–143) 84 (61–115) AI size, cm 2.0 (1.0–5.1) 1.8 (1.0–5.1) 2.2b (1.2–3.8) 2.1 (1.5–4.0) 2.3b (1.0–4.5) Cortisol, nmol/L 445 (131–1007) 444 (131–967) 402b (157–610) 504 (307–973) 484 (177–1007) ACTH, pmol/L 3.0 (0.23–19.0) 3.9 (2.0–8.6) 1.5c (0.34–1.9) 3.2b (2.0–19.0) 1.3c (0.23–1.9) DHEAS, µmol/L 1.9 (<0.81–11.0) 2.4 (<0.81–10.0) 1.5a (<0.81–3.8) 1.8 (<0.81–5.8) 1.3b (<0.81–11.0) Cortisol/ACTH ratio, nmol/L/pmol/L 150 (32–2078) 114 (32–312) 267c (129–812) 164a (45–274) 349c (104–2078) CortisolONDST, nmol/L 37 (12–310) 30 (12–48) 38c (28–49) 67c (50–310) 77c (51–157)  Statistical differences were calculated between the NormalACTH/ONDST<50 group and the other three groups. Statistical differences for DHEAS were adjusted for differences in age and sex. Abbreviations: BMI, body mass index; eGFR, estimated glomerular filtration rate. a P < 0.01 compared with the NormalACTH/ONDST<50 group. b P < 0.05 compared with the NormalACTH/ONDST<50 group. c P < 0.001 compared with the NormalACTH/ONDST<50 group. View Large Baseline and 2.1 (1.0–3.9) year follow-up data for the four groups are given in Table 2. Groups were still separated in a similar manner at follow-up compared with baseline. CortisolONDST was higher at follow-up in the entire cohort (P < 0.001), but ACTH, cortisol, and cortisol/ACTH ratio were unchanged. In the LowACTH/ONDST<50 and LowACTH/ONDST≥50 groups, ACTH was slightly higher at follow-up most likely attributed to the regression to the mean effect (P = 0.005 and P = 0.03), which also possibly could explain why cortisolONDST in the NormalACTH/ONDST<50 group was higher at follow-up (P < 0.001). Table 2. Hormone Levels at Baseline and at Follow-Up of 106 Patients with AI, Grouped According to Whether Their CortisolONDST Level Was <50 or ≥50 nmol/L and Whether Their ACTH Level Was <2.0 or ≥2.0 pmol/L  Baseline Follow-Up NormalACTH/ONDST<50 group (n = 64)    Cortisol, nmol/L 444 (131–715) 411 (130–792)  ACTH, pmol/L 3.8 (2.0–8.5) 3.7 (1.3–25.0)  CortisolONDST, nmol/L 29 (12–47) 32 (16–88)  Cortisol/ACTH ratio, nmol/L/pmol/L 112 (32–274) 117 (16–311) LowACTH/ONDST<50 group (n = 19)    Cortisol, nmol/L 399 (232–610) 472 (189–761)  ACTH, pmol/L 1.5 (0.34–1.9)a 1.9 (1.2–4.6)a  CortisolONDST, nmol/L 40 (28–47)a 43 (24–55)b  Cortisol/ACTH ratio, nmol/L/pmol/L 261 (129–812)a 255 (65–486)a NormalACTH/ONDST≥50 group (n = 6)    Cortisol, nmol/L 508 (388–568) 489 (352–568)  ACTH, pmol/L 3.5 (2.5–4.4) 2.9 (2.0–4.6)  CortisolONDST, nmol/L 92 (50–155)a 124 (32–330)a  Cortisol/ACTH ratio, nmol/L/pmol/L 146 (114–188) 157 (105–284) LowACTH/ONDST≥50 group (n = 17)    Cortisol, nmol/L 410 (177–794) 454 (246–865)  ACTH, pmol/L 1.3 (0.23–1.9)a 1.5 (0.27–3.8)a  CortisolONDST, nmol/L 74 (51–131)a 69 (44–232)a  Cortisol/ACTH ratio, nmol/L/pmol/L 329 (104–2078)a 325 (120–911)a   Baseline Follow-Up NormalACTH/ONDST<50 group (n = 64)    Cortisol, nmol/L 444 (131–715) 411 (130–792)  ACTH, pmol/L 3.8 (2.0–8.5) 3.7 (1.3–25.0)  CortisolONDST, nmol/L 29 (12–47) 32 (16–88)  Cortisol/ACTH ratio, nmol/L/pmol/L 112 (32–274) 117 (16–311) LowACTH/ONDST<50 group (n = 19)    Cortisol, nmol/L 399 (232–610) 472 (189–761)  ACTH, pmol/L 1.5 (0.34–1.9)a 1.9 (1.2–4.6)a  CortisolONDST, nmol/L 40 (28–47)a 43 (24–55)b  Cortisol/ACTH ratio, nmol/L/pmol/L 261 (129–812)a 255 (65–486)a NormalACTH/ONDST≥50 group (n = 6)    Cortisol, nmol/L 508 (388–568) 489 (352–568)  ACTH, pmol/L 3.5 (2.5–4.4) 2.9 (2.0–4.6)  CortisolONDST, nmol/L 92 (50–155)a 124 (32–330)a  Cortisol/ACTH ratio, nmol/L/pmol/L 146 (114–188) 157 (105–284) LowACTH/ONDST≥50 group (n = 17)    Cortisol, nmol/L 410 (177–794) 454 (246–865)  ACTH, pmol/L 1.3 (0.23–1.9)a 1.5 (0.27–3.8)a  CortisolONDST, nmol/L 74 (51–131)a 69 (44–232)a  Cortisol/ACTH ratio, nmol/L/pmol/L 329 (104–2078)a 325 (120–911)a  Statistical differences were calculated between the NormalACTH/ONDST<50 group and the other three groups. a P < 0.001 compared with the NormalACTH/ONDST<50 group. b P < 0.05 compared with the NormalACTH/ONDST<50 group. View Large Table 2. Hormone Levels at Baseline and at Follow-Up of 106 Patients with AI, Grouped According to Whether Their CortisolONDST Level Was <50 or ≥50 nmol/L and Whether Their ACTH Level Was <2.0 or ≥2.0 pmol/L  Baseline Follow-Up NormalACTH/ONDST<50 group (n = 64)    Cortisol, nmol/L 444 (131–715) 411 (130–792)  ACTH, pmol/L 3.8 (2.0–8.5) 3.7 (1.3–25.0)  CortisolONDST, nmol/L 29 (12–47) 32 (16–88)  Cortisol/ACTH ratio, nmol/L/pmol/L 112 (32–274) 117 (16–311) LowACTH/ONDST<50 group (n = 19)    Cortisol, nmol/L 399 (232–610) 472 (189–761)  ACTH, pmol/L 1.5 (0.34–1.9)a 1.9 (1.2–4.6)a  CortisolONDST, nmol/L 40 (28–47)a 43 (24–55)b  Cortisol/ACTH ratio, nmol/L/pmol/L 261 (129–812)a 255 (65–486)a NormalACTH/ONDST≥50 group (n = 6)    Cortisol, nmol/L 508 (388–568) 489 (352–568)  ACTH, pmol/L 3.5 (2.5–4.4) 2.9 (2.0–4.6)  CortisolONDST, nmol/L 92 (50–155)a 124 (32–330)a  Cortisol/ACTH ratio, nmol/L/pmol/L 146 (114–188) 157 (105–284) LowACTH/ONDST≥50 group (n = 17)    Cortisol, nmol/L 410 (177–794) 454 (246–865)  ACTH, pmol/L 1.3 (0.23–1.9)a 1.5 (0.27–3.8)a  CortisolONDST, nmol/L 74 (51–131)a 69 (44–232)a  Cortisol/ACTH ratio, nmol/L/pmol/L 329 (104–2078)a 325 (120–911)a   Baseline Follow-Up NormalACTH/ONDST<50 group (n = 64)    Cortisol, nmol/L 444 (131–715) 411 (130–792)  ACTH, pmol/L 3.8 (2.0–8.5) 3.7 (1.3–25.0)  CortisolONDST, nmol/L 29 (12–47) 32 (16–88)  Cortisol/ACTH ratio, nmol/L/pmol/L 112 (32–274) 117 (16–311) LowACTH/ONDST<50 group (n = 19)    Cortisol, nmol/L 399 (232–610) 472 (189–761)  ACTH, pmol/L 1.5 (0.34–1.9)a 1.9 (1.2–4.6)a  CortisolONDST, nmol/L 40 (28–47)a 43 (24–55)b  Cortisol/ACTH ratio, nmol/L/pmol/L 261 (129–812)a 255 (65–486)a NormalACTH/ONDST≥50 group (n = 6)    Cortisol, nmol/L 508 (388–568) 489 (352–568)  ACTH, pmol/L 3.5 (2.5–4.4) 2.9 (2.0–4.6)  CortisolONDST, nmol/L 92 (50–155)a 124 (32–330)a  Cortisol/ACTH ratio, nmol/L/pmol/L 146 (114–188) 157 (105–284) LowACTH/ONDST≥50 group (n = 17)    Cortisol, nmol/L 410 (177–794) 454 (246–865)  ACTH, pmol/L 1.3 (0.23–1.9)a 1.5 (0.27–3.8)a  CortisolONDST, nmol/L 74 (51–131)a 69 (44–232)a  Cortisol/ACTH ratio, nmol/L/pmol/L 329 (104–2078)a 325 (120–911)a  Statistical differences were calculated between the NormalACTH/ONDST<50 group and the other three groups. a P < 0.001 compared with the NormalACTH/ONDST<50 group. b P < 0.05 compared with the NormalACTH/ONDST<50 group. View Large Missing data DHEAS data were missing in 3% of the NormalACTH/ONDST<50 group, 11% of the NormalACTH/ONDST≥50 group, and 4% of the LowACTH/ONDST≥50 group. In these missing cases, patient’s sex distribution, age, estimated glomerular filtration rate, size of AI, ACTH, and cortisolONDST were similar to those with available DHEAS data. At 2-year follow-up, data were limited to 106 patients. Discussion This study addressed the limitations of ACTH as a complementary criterion to cortisolONDST in the diagnosis of ACS. We have previously hypothesized that increased cortisol secretion by AI in response to ACTH may suppress ACTH (14) because an intact HPA axis secretes less ACTH to maintain the desired cortisol level if the responsiveness to ACTH is augmented. Our hypothesis challenges the current opinion of cortisol secretion from adrenal adenomas being exclusively autonomous but could explain why ACTH may be suppressed in patients with discrete ACS and in the normal range when ACS is marked (7, 11, 14). In this study, a previously unrecognized group of patients, the LowACTH/ONDST<50 group, defined by suppressed ACTH in the absence of ACS according to the ESE guidelines (cortisolONDST <50 nmol/L) (7), was identified. As much as 19% of AI patients without ACS had suppressed ACTH and thus belonged to this group, whereas only 4% among the control subjects had ACTH below this level. The LowACTH/ONDST<50 group also exhibited lower DHEAS, a suggested marker of HPA axis activity (17), and an increased cortisol/ACTH ratio, supporting decreased activity of the HPA axis and increased responsiveness to ACTH. In addition, changing the definition of normal cortisolONDST to <40 nmol/L did not change our results, further contradicting ACTH suppression due to very discrete ACS. Finally, the robustness of our results was confirmed with similar results when the patients were re-evaluated after 2 years. The negative correlation between ACTH and the size of the AI among patients without ACS indicates that the AI harbors the factor responsible for the suppression of ACTH. The high prevalence of suppressed ACTH without concomitant ACS may have consequences also for patients in the LowACTH/ONDST≥50 group. Patients in this group are commonly considered to have ACS. However, a patient with the combination of increased cortisol responsiveness to ACTH resulting in suppressed ACTH and inadequate suppression of cortisolONDST, for example due to rapid dexamethasone metabolism, may falsely be diagnosed to have ACS. Our theory is supported by several other studies. First, cells from adenomas have ACTH receptors, the number of which is related to the degree of hypercortisolism (18, 19), and the functionality of ACTH receptors has been established by in vitro studies of cortisol secretion in response to ACTH (20). Second, patients with unilateral AI exhibit increased cortisol levels following ACTH stimulation test, which are inversely related to basal ACTH (15). In another study, patients with nonfunctioning AIs defined as cortisolONDST <50 nmol/L had increased cortisol response to ACTH stimulation test (21). Contrarily, in secondary adrenal insufficiency, normal adrenocortical cells lose their ability to respond to short ACTH stimulation after a period of continuously low ACTH (22). This difference in response to ACTH stimulation tests between these two disorders is consistent with the finding of preserved amounts of ACTH receptors in adenoma cells during prolonged low ACTH but reduced in normal adrenocortical cells (18). Finally, it could be argued that another factor apart from cortisol secretion by the AI could suppress ACTH, such as a steroid metabolite other than cortisol. However, in the study by Di Dalmazi et al. (15), demonstrating an inverse relation between ACTH and the cortisol response to an ACTH stimulation test, among nine other steroid metabolites, only deoxycorticosterone was assumed to contribute to the increase in glucocorticoid effect, but no evidence was presented that it was related to ACTH. The usefulness of ACTH as a marker for the activity of the HPA axis may, besides its pulsatile secretion, be hampered by the inability of some ACTH assays to detect low values (12, 13). The capability of the ACTH assay used in our study to measure low levels is supported by our previously published study and by the low CV at ACTH 1.1 pmol/L stated by the manufacturer (14). The possibility to examine long-term metabolic complications and long-term development of cortisol secretion in the LowACTH/ONDST<50 group is limited due to the number of subjects studied and the duration of follow-up. The prevalence of treated hypertension, treated diabetes, and treated dyslipidemia in relation to the NormalACTH/ONDST<50 group suggests that increased cortisol responsiveness to ACTH is not strongly linked to metabolic complications, at least not in the short-term. However, in the study by Androulakis et al. (21), carotid intima thickness, a marker for cardiovascular risk, was found to correlate with the cortisol response to ACTH in patients with nonfunctioning AIs. The younger age of the LowACTH/ONDST<50 group may indicate that some patients later convert into another of the three groups. Our study gives no answer, but a speculation would be that they later acquire ACS and switch to the LowACTH/ONDST≥50 group. Future studies on these issues are encouraged. Further limitations are that ACTH stimulation tests, which could have demonstrated increased cortisol response, were not performed, and urinary and salivary cortisol were not available. Two-year follow-up data were also not present in some of the patients; the reason for this was in most cases an exchange of the cortisol analysis method. In summary, we have highlighted a group of patients with suppressed ACTH despite normal cortisolONDST, indicating ACTH suppression by another factor than ACS. We suggest that this additional factor is increased sensitivity to ACTH, resulting in decreased ACTH secretion due to negative feedback in otherwise intact HPA axes. Variations in ACTH sensitivity, unrelated to the degree of ACS, may explain the limitation of ACTH as a useful marker for ACS in patients with AI. Potential long-term effects of suppressed ACTH, alone or in combination with ACS, have so far not been studied and deserve attention in the future. Abbreviations: Abbreviations:   ACS autonomous cortisol secretion   AI adrenal incidentalomas   cortisolONDST cortisol after overnight dexamethasone suppression test   CV coefficient of variation   DHEAS dehydroepiandrosterone   ESE European Society of Endocrinology   HPA hypothalamic-pituitary-adrenal Acknowledgments Financial Support: M.L. was supported by Medicinska Fakulteten, Lunds Universitet. Author Contributions: H.O. and O.L. designed the study. H.O. and A.K. collected the data. H.O., A.K., and O.L. performed the data analyses and interpretation. H.O. and O.L. prepared the manuscript. All authors contributed to the revision of the manuscript and approved the final version. Disclosure Summary: The authors have nothing to disclose. References 1. Morelli V , Reimondo G , Giordano R , Della Casa S , Policola C , Palmieri S , Salcuni AS , Dolci A , Mendola M , Arosio M , Ambrosi B , Scillitani A , Ghigo E , Beck-Peccoz P , Terzolo M , Chiodini I . Long-term follow-up in adrenal incidentalomas: an Italian multicenter study . J Clin Endocrinol Metab  . 2014 ; 99 ( 3 ): 827 – 834 . Google Scholar Crossref Search ADS PubMed   2. Debono M , Bradburn M , Bull M , Harrison B , Ross RJ , Newell-Price J . Cortisol as a marker for increased mortality in patients with incidental adrenocortical adenomas . J Clin Endocrinol Metab  . 2014 ; 99 ( 12 ): 4462 – 4470 . Google Scholar Crossref Search ADS PubMed   3. Di Dalmazi G , Vicennati V , Garelli S , Casadio E , Rinaldi E , Giampalma E , Mosconi C , Golfieri R , Paccapelo A , Pagotto U , Pasquali R . Cardiovascular events and mortality in patients with adrenal incidentalomas that are either non-secreting or associated with intermediate phenotype or subclinical Cushing’s syndrome: a 15-year retrospective study . Lancet Diabetes Endocrinol  . 2014 ; 2 ( 5 ): 396 – 405 . Google Scholar Crossref Search ADS PubMed   4. Di Dalmazi G , Vicennati V , Rinaldi E , Morselli-Labate AM , Giampalma E , Mosconi C , Pagotto U , Pasquali R . Progressively increased patterns of subclinical cortisol hypersecretion in adrenal incidentalomas differently predict major metabolic and cardiovascular outcomes: a large cross-sectional study . Eur J Endocrinol  . 2012 ; 166 ( 4 ): 669 – 677 . Google Scholar Crossref Search ADS PubMed   5. Perogamvros I , Vassiliadi DA , Karapanou O , Botoula E , Tzanela M , Tsagarakis S . Biochemical and clinical benefits of unilateral adrenalectomy in patients with subclinical hypercortisolism and bilateral adrenal incidentalomas . Eur J Endocrinol  . 2015 ; 173 ( 6 ): 719 – 725 . Google Scholar Crossref Search ADS PubMed   6. Vassilatou E , Vryonidou A , Ioannidis D , Paschou SA , Panagou M , Tzavara I . Bilateral adrenal incidentalomas differ from unilateral adrenal incidentalomas in subclinical cortisol hypersecretion but not in potential clinical implications . Eur J Endocrinol  . 2014 ; 171 ( 1 ): 37 – 45 . Google Scholar Crossref Search ADS PubMed   7. Fassnacht M , Arlt W , Bancos I , Dralle H , Newell-Price J , Sahdev A , Tabarin A , Terzolo M , Tsagarakis S , Dekkers OM . Management of adrenal incidentalomas: European Society of Endocrinology Clinical Practice Guideline in collaboration with the European Network for the Study of Adrenal Tumors . Eur J Endocrinol  . 2016 ; 175 ( 2 ): G1 – G34 . Google Scholar Crossref Search ADS PubMed   8. Ueland GÅ , Methlie P , Kellmann R , Bjørgaas M , Åsvold BO , Thorstensen K , Kelp O , Thordarson HB , Mellgren G , Løvås K , Husebye ES . Simultaneous assay of cortisol and dexamethasone improved diagnostic accuracy of the dexamethasone suppression test . Eur J Endocrinol  . 2017 ; 176 ( 6 ): 705 – 713 . Google Scholar Crossref Search ADS PubMed   9. Olsen H , Mjöman M . Moderately impaired renal function increases morning cortisol and cortisol levels at dexamethasone suppression test in patients with incidentally detected adrenal adenomas . Clin Endocrinol (Oxf)  . 2015 ; 83 ( 6 ): 762 – 767 . Google Scholar Crossref Search ADS PubMed   10. Wilkinson CW , Petrie EC , Murray SR , Colasurdo EA , Raskind MA , Peskind ER . Human glucocorticoid feedback inhibition is reduced in older individuals: evening study . J Clin Endocrinol Metab  . 2001 ; 86 ( 2 ): 545 – 550 . Google Scholar PubMed   11. Chiodini I . Clinical review: diagnosis and treatment of subclinical hypercortisolism . J Clin Endocrinol Metab  . 2011 ; 96 ( 5 ): 1223 – 1236 . Google Scholar Crossref Search ADS PubMed   12. Russell GM , Henley DE , Leendertz J , Douthwaite JA , Wood SA , Stevens A , Woltersdorf WW , Peeters BW , Ruigt GS , White A , Veldhuis JD , Lightman SL . Rapid glucocorticoid receptor-mediated inhibition of hypothalamic-pituitary-adrenal ultradian activity in healthy males . J Neurosci  . 2010 ; 30 ( 17 ): 6106 – 6115 . Google Scholar Crossref Search ADS PubMed   13. Pecori Giraldi F , Saccani A , Cavagnini F ; Study Group on the Hypothalamo-Pituitary-Adrenal Axis of the Italian Society of Endocrinology . Assessment of ACTH assay variability: a multicenter study . Eur J Endocrinol  . 2011 ; 164 ( 4 ): 505 – 512 . Google Scholar Crossref Search ADS PubMed   14. Olsen H , Olofsson L , Mjöman M . Cortisol secretion from adrenal adenomas discovered as incidentalomas is responsive to ACTH . Clin Endocrinol (Oxf)  . 2014 ; 80 ( 3 ): 335 – 341 . Google Scholar Crossref Search ADS PubMed   15. Di Dalmazi G , Fanelli F , Mezzullo M , Casadio E , Rinaldi E , Garelli S , Giampalma E , Mosconi C , Golfieri R , Vicennati V , Pagotto U , Pasquali R . Steroid profiling by LC-MS/MS in nonsecreting and subclinical cortisol-secreting adrenocortical adenomas . J Clin Endocrinol Metab  . 2015 ; 100 ( 9 ): 3529 – 3538 . Google Scholar Crossref Search ADS PubMed   16. Solberg HE . The IFCC recommendation on estimation of reference intervals. The RefVal program . Clin Chem Lab Med  . 2004 ; 42 ( 7 ): 710 – 714 . Google Scholar Crossref Search ADS PubMed   17. Dennedy MC , Annamalai AK , Prankerd-Smith O , Freeman N , Vengopal K , Graggaber J , Koulori O , Powlson AS , Shaw A , Halsall DJ , Gurnell M . Low DHEAS: a sensitive and specific test for the detection of subclinical hypercortisolism in adrenal incidentalomas . J Clin Endocrinol Metab  . 2017 ; 102 ( 3 ): 786 – 792 . Google Scholar PubMed   18. Reincke M , Beuschlein F , Menig G , Hofmockel G , Arlt W , Lehmann R , Karl M , Allolio B . Localization and expression of adrenocorticotropic hormone receptor mRNA in normal and neoplastic human adrenal cortex . J Endocrinol  . 1998 ; 156 ( 3 ): 415 – 423 . Google Scholar Crossref Search ADS PubMed   19. Schubert B , Fassnacht M , Beuschlein F , Zenkert S , Allolio B , Reincke M . Angiotensin II type 1 receptor and ACTH receptor expression in human adrenocortical neoplasms . Clin Endocrinol (Oxf)  . 2001 ; 54 ( 5 ): 627 – 632 . Google Scholar Crossref Search ADS PubMed   20. Beuschlein F , Fassnacht M , Klink A , Allolio B , Reincke M . ACTH receptor expression, regulation and role in adrenocortial tumor formation . Eur J Endocrinol  . 2001 ; 144 ( 3 ): 199 – 206 . Google Scholar Crossref Search ADS PubMed   21. Androulakis II , Kaltsas GA , Kollias GE , Markou AC , Gouli AK , Thomas DA , Alexandraki KI , Papamichael CM , Hadjidakis DJ , Piaditis GP . Patients with apparently nonfunctioning adrenal incidentalomas may be at increased cardiovascular risk due to excessive cortisol secretion . J Clin Endocrinol Metab  . 2014 ; 99 ( 8 ): 2754 – 2762 . Google Scholar Crossref Search ADS PubMed   22. Agha A , Tomlinson JW , Clark PM , Holder G , Stewart PM . The long-term predictive accuracy of the short synacthen (corticotropin) stimulation test for assessment of the hypothalamic-pituitary-adrenal axis . J Clin Endocrinol Metab  . 2006 ; 91 ( 1 ): 43 – 47 . Google Scholar Crossref Search ADS PubMed   Copyright © 2019 Endocrine Society


This is a preview of a remote PDF: https://academic.oup.com/jcem/article-pdf/104/2/506/27378241/jc.2018-01029.pdf

Olsen, Henrik, Kjellbom, Albin, Löndahl, Magnus, Lindgren, Ola. Suppressed ACTH Is Frequently Unrelated to Autonomous Cortisol Secretion in Patients With Adrenal Incidentalomas, The Journal of Clinical Endocrinology & Metabolism, 2019, 506-512, DOI: 10.1210/jc.2018-01029