The management of patients with adrenocorticotropic hormone-independent Cushing’s syndrome and bilateral adrenal masses is challenging. Adrenal venous sampling (AVS) has been used to identify functional lesions in previous studies, but it is not always reliable. The present study aims to address the variability of cortisol in the adrenal veins of patients without excessive cortisol secretion and investigate the use of adrenal androgens to correct the cortisol lateralization ratio in AVS. Thirty-seven patients with primary aldosteronism underwent successful AVS. Patients with normal cortisol secretion exhibited a wide range of cortisol concentrations in the right (601-89, 400 nmol/l) and left (331-35, 300 nmol/l) adrenal veins. The median cortisol gradients between adrenal venous and peripheral venous samples were 15.25 and 10.14 in the right and left sides, respectively, and the cortisol lateralization ratio (high side to low side) was as high as 9.49 (median 1.54). The mean plasma levels of cortisol in the adrenal venous and peripheral venous samples decreased from t-15 to t0. Significant positive correlations were observed between the cortisol concentrations and both androstenedione and dehydroepiandrosterone concentrations in the right and left adrenal veins. After correcting for androstenedione or dehydroepiandrosterone levels, the cortisol lateralization ratio was less than 2 in most adrenal venous samples. The present study demonstrated the wide variation in cortisol concentrations in the adrenal veins of patients with normal cortisol secretion. The adrenal androgens might be ideal analytes used as normalizers when assessing the cortisol lateralization of AVS in normal or hypercortisolism cases.
Adrenocorticotropic hormone (ACTH)-independent Cushing’s syndrome (CS) is occasionally caused by bilateral adrenocortical lesions. Such patients may have a unilateral cortisol-secreting adenoma with a contralateral nonfunctioning cortical adenoma, bilateral cortisol-secreting adenomas, or bilateral ACTH-independent macronodular adrenal hyperplasia (AIMAH) mimicking bilateral single adenomas [
Adrenal venous sampling (AVS) has been successfully used to lateralize the source of aldosterone hypersecretion in patients with primary aldosteronism [
According to the literature and our experience using AVS in primary aldosteronism (PA) cases, the following factors might interfere with the assessment of lateralization in ACTH-independent CS: (1) a stress reaction involving increased cortisol release, (2) fluctuating levels of cortisol induced by ACTH secretion, and (3) different dilutional effects in the right and left adrenals due to the adrenal venous (AV) anatomy [
The adrenal androgens androstenedione and dehydroepiandrosterone (DHEA) are mainly produced in the zona reticularis. Recently published studies from our group and others suggested that adrenal androgens are useful for assessing the selectivity of AVS in PA [
We consecutively recruited PA patients undergoing AVS among patients referred to the endocrinology department at our hospital for suspected secondary hypertension. The diagnosis of PA was confirmed by an intravenous saline infusion or a captopril test. A 1 mg dexamethasone suppression test (DST) was performed to exclude autonomous cortisol secretion, indicated by a post-DST cortisol level less than 50 nmol/l. We performed twenty-four-hour urinary catecholamine measurement to exclude pheochromocytoma. Patients were offered AVS according to the guidelines of the US Endocrine Society [
AVS was performed as described previously [
The PCC was measured with a commercially available kit (Immulite 2000 Cortisol, Siemens Healthcare Diagnostics Products Limited, Gwynedd LL55 4EL, United Kingdom). The intra- and interassay coefficients of variation (CVs) for PCCs were 4.6% and 6.8%, respectively. The plasma concentrations of androstenedione and DHEA were measured using a commercial ELISA kit (DRG International, Inc., USA). The intra-assay CVs for androstenedione ranged from 0.35% for high plasma concentrations to 1.50% for low concentrations. The interassay CVs for DHEA ranged from 0.71% for high plasma concentrations to 2.85% for low concentrations. The intra-assay and interassay CVs for this assay were 5.2% and 9.8%, respectively.
Data are expressed as the means and SDs or, in the case of skewed distributions, as medians and ranges. Kruskal-Wallis and Mann-Whitney U tests were used to assess the significance of differences in variables at the three sampling sites or between groups. Relationships among cortisol, androstenedione, and DHEA were assessed by one-tailed Spearman’s correlation coefficient (
Patients with normal cortisol secretion exhibited a wide range of cortisol concentrations in the right (601-89, 400 nmol/l) and left (331-35, 300 nmol/l) adrenal veins. The median cortisol gradients between AV and peripheral venous (PV) samples were 15.25 and 10.14 on the right and left sides, respectively. Consistent with the cortisol concentrations, considerably higher plasma androstenedione and DHEA concentrations were detected in the right and left AV samples than in PV samples (P < 0.01). The androstenedione and DHEA gradients between AV and PV samples were approximately 2-3 times higher than those of cortisol. Although no significant difference was observed in cortisol levels between the right and left adrenal veins, the cortisol lateralization (high-side to low-side) was as high as 9.49 (median 1.54). Similar findings were also observed for androstenedione and DHEA (Table
Adrenal vein cortisol and adrenal androgen measurements in patients with normal cortisol secretion undergoing adrenal venous sampling [M (25-75%)].
Parameter | LAV | RAV | PV | LAV/PV | RAV/PV | Lateralization ratio |
---|---|---|---|---|---|---|
Cortisol | 3270 | 4695 | 317 | 10.14 | 15.25 | 1.54 |
Androstenedione (ng/ml) | 48.65 | 66.69 | 1.51 | 27.62 | 43.01 | 1.49 |
DHEA | 132.74 | 160.14 | 4.24 | 35.76 | 46.05 | 1.69 |
DHEA, dehydroepiandrosterone; LAV, left adrenal vein; RAV, right adrenal vein; and PV, peripheral vein.
To investigate the variation of hormones over time, we measured cortisol levels in repeated samples from 17 PA patients obtained at 15 min intervals as described previously [
Plasma cortisol levels in the adrenal veins measured from repeated samples (t-15, t0) during AVS [M (25-75%)].
Parameter | LAV | RAV | PV | LAV/PV | RAV/PV | Lateralization ratio |
---|---|---|---|---|---|---|
Cortisol (nmol/l) t-15 | 5600 | 7860 | 309 | 18.22 | 24.39 | 1.69 |
Cortisol (nmol/l) t0 | 3420 | 3920 | 292 | 10.76 | 13.42 | 1.51 |
P (t-15 vs t0) | 0.455 | 0.035 | 0.082 | 0.209 | 0.654 | 0.904 |
Variance ratio | 0.61±0.51 | 0.64±0.37 | 0.17±0.16 |
LAV, left adrenal vein; RAV, right adrenal vein; and PV, peripheral vein.
Significant positive correlations were observed between the cortisol concentrations and both androstenedione and DHEA concentrations in the right (
Correlations among plasma androstenedione (a), plasma DHEA (b), and plasma cortisol levels for AVS. LAV, left adrenal vein; RAV, right adrenal vein; and PV, peripheral vein.
Among 54 AVS procedures, 35 AVS procedures exhibited a cortisol lateralization ratio (high-side to low-side) less than 2 between the right and left AV samples, and 19 AVS procedures had a ratio greater than 2. After correction for androstenedione levels, only 4 AVS procedures had a lateralization ratio greater than 2. After correction for DHEA, 9 AVS procedures had a lateralization ratio greater than 2 (Figure
High-side to low-side adrenal vein cortisol concentration ratios with or without adrenal androgens correction. AV, adrenal vein.
These data from patients without excessive cortisol secretion undergoing AVS demonstrated that adrenal vein cortisol concentrations exhibit great variation, including significant variation in the plasma cortisol levels over time and the cortisol gradient between the right and left adrenal veins. Furthermore, we confirmed that the adrenal androgens androstenedione and DHEA are useful in correcting for the side-to-side gradient of cortisol.
The optimum surgical treatment of ACTH-independent CS and subclinical CS is less clear when a patient appears to have bilateral adrenal cortical adenomas [
AVS is commonly used to distinguish the source of hormonal production in patients with primary hyperaldosteronism [
Subsequent studies followed the criteria of Young in interpreting the results of AVS in patients with bilateral adrenal masses and ACTH-independent CS. However, AVS has not always been reliable in differentiating the source of excessive cortisol secretion according to literature reported from different centres [
To exclude the possibility of endogenous ACTH secretion interfering with the interpretation of autonomous cortisol secretion, all patients underwent AVS while receiving dexamethasone in Young’s study [
The most reliable solution is to investigate novel analytes to correct for cortisol lateralization during AVS. In PA patients, “cortisol-corrected” aldosterone ratios are compared to determine whether a unilateral source of aldosterone exists. Previous studies attempted to use epinephrine or aldosterone to correct for the side-to-side cortisol gradient [
The adrenal androgens androstenedione and DHEA are adrenocortical products. Compared to cortisol, higher gradients of AV to PV plasma androstenedione and DHEA concentrations were detected [
In conclusion, the present study promoted the use of the plasma adrenal androgens as normalizers to assess cortisol lateralization in AVS. This technique will improve the diagnostic accuracy of AVS in the localization of autonomous hypercortisolism in the setting of ACTH-independent CS in patients with bilateral adrenal masses.
The data used to support the findings of this study are included within the article.
The authors declare no conflicts of interest.
(1) Ping Li and Dalong Zhu formed subject design. (2) Wenjing Zhang, Keying Zhu, Hongyun Li, and Yan Zhang were responsible for collecting and collating data. (3) Xuebin Zhang was responsible for adrenal venous sampling in DSA. (4) Wenjing Zhang and Keying Zhu performed data analysis. (5) Ping Li and Dalong Zhu performed result analysis. (6) Wenjing Zhang and Keying Zhu wrote the paper. Wenjing Zhang and Keying Zhu contributed equally to this paper.
This work was supported by the Natural Science Foundation of Jiangsu Province (Grant number BK20181116) and the Project of Jiangsu Provincial Medical Youth Talent.