Adrenal steroid metabolites and bone status in patients with adrenal incidentalomas and hypercortisolism

Abstract

Background: Autonomous cortisol secretion (ACS), resulting from cortisol-producing adenomas (CPA), causes endogenous steroid-induced osteoporosis (SIOP). However, the risk of endogenous SIOP cannot be explained by cortisol excess alone, and how other steroid metabolites affect bone status is unclear.

Methods: ACS was diagnosed as serum cortisol ≥1.8 μg/dL after the 1-mg dexamethasone suppression test (DST-cortisol). Using liquid chromatography tandem mass spectrometry, 21 plasma steroid metabolites were measured in 73 patients with ACS and 85 patients with non-functioning adrenal tumors (NFAT). Expression of steroidogenic enzymes and relevant steroid metabolites were analyzed in some of CPA tissues.

Findings: Discriminant and principal component analyses distinguished steroid profiles between the ACS and NFAT groups in premenopausal women. Premenopausal women with ACS exhibited higher levels of a mineralocorticoid metabolite, 11-deoxycorticosterone (11-DOC), and lower levels of androgen metabolites, dehydroepiandrosterone-sulfate, and androsterone-glucuronide. In premenopausal women with ACS, DST-cortisol negatively correlated with trabecular bone score (TBS). Additionally, 11-DOC negatively correlated with lumbar spine-bone mineral density, whereas androsterone-glucuronide positively correlated with TBS. The CPA tissues showed increased 11-DOC levels with increased expression of CYP21A2, essential for 11-DOC synthesis. Adrenal non-tumor tissues were atrophied with reduced expression of CYB5A, required for androgen synthesis.

Interpretation: This study demonstrates that unbalanced production of adrenal steroid metabolites, derived from both adrenal tumor and non-tumor tissues, contributes to the pathogenesis of endogenous SIOP in premenopausal women with ACS.

Funding: JSPS KAKENHI, Secom Science and Technology Foundation, Takeda Science Foundation, Japan Foundation for Applied Enzymology, AMED-CREST, JSTA-STEP, JST-Moonshot, and Ono Medical Research Foundation.

Keywords: Adrenal incidentalomas; Autonomous cortisol secretion; Cortisol-producing adenomas; Steroid metabolites; Steroid-induced osteoporosis.

Fig. 1

Study design and steroid pathways. (A) The flowchart depicts the process of patient selection for the study. (B) Steroid pathways of mineralocorticoids, glucocorticoids, and androgens are shown, and the steroid metabolites used in this study are highlighted. Abbreviations: ACS, autonomous cortisol secretion; NFAT, non-functioning adrenal tumors; DST, dexamethasone suppression test; Pheo, pheochromocytoma; ACC, adrenocortical carcinomas; APA, aldosterone-producing adenomas; CPA, cortisol-producing adenomas; NFA, non-functioning adrenocortical adenomas; RNA-seq, RNA-sequencing; IHC, immunohistochemistry; 11-DOC, 11-deoxycorticosterone; DHEA-S, dehydroepiandrosterone-sulfate.

Fig. 2

Plasma steroid profiles in patients with ACS and NFAT. PLS-DA and PCA were performed to determine whether steroid profiles distinguished between the ACS and NFAT groups. A VIP score analysis was performed to rank the overall contribution of each variable to the PLS-DA model, and variables with VIP scores >1 were considered significant. The analysis was categorized into premenopausal women (ACS, n = 18; NFAT, n = 12), postmenopausal women (ACS, n = 26; NFAT, n = 34), and men (ACS, n = 29; NFAT, n = 39). Abbreviations: ACS, autonomous cortisol secretion; NFAT, non-functioning adrenal tumors; PLS-DA, partial least squares-discriminant analysis; VIP, variable importance in projection; 11-DOC, 11-deoxycorticosterone; DHEA-S, dehydroepiandrosterone-sulfate; PCA, principal component analysis; PC, principal component.

Fig. 3

Comparison of each steroid metabolite between patients with ACS and NFAT. Cohen’s d values were calculated to compare each steroid metabolite between patients with ACS and NFAT. Steroid metabolites with P < 0.05 are colored and those with FDR < 0.1 are indicated by asterisks. FDR < 0.1 was considered statistically significant. The analysis was categorized into premenopausal women (ACS, n = 18; NFAT, n = 12), postmenopausal women (ACS, n = 26; NFAT, n = 34) and men (ACS, n = 29; NFAT, n = 39). Abbreviations: ACS, autonomous cortisol secretion; NFAT, non-functioning adrenal tumors; 11-DOC, 11-deoxycorticosterone; DHEA-S, dehydroepiandrosterone-sulfate.

Fig. 4

Correlation of steroid metabolites with bone status in patients with ACS. We evaluated the correlation between ACS-related steroid metabolites and BMD Z-scores and TBS in patients with ACS and categorized them into three subgroups: 17 premenopausal women, 26 postmenopausal women, and 27 men. FDR < 0.1 was considered statistically significant. Abbreviations: ACS, autonomous cortisol secretion; NFAT, non-functioning adrenal tumors; VIP, variable importance in projection; LS-BMD, lumbar spine bone mineral density; FN-BMD, femoral neck bone mineral density; TBS, trabecular bone score; 11-DOC, 11-deoxycorticosterone; DHEA-S, dehydroepiandrosterone-sulfate.

Fig. 5

Expression of steroidogenic enzymes and measurement of steroid metabolites in CPA tissues. (A) Volcano plot of RNA-seq data displays DEGs between CPA (n = 16) and NFA (n = 3). DEGs are defined as genes with an absolute logFC > 1 and an FDR < 0.05. Upregulated genes in CPA are shown in red and downregulated genes in blue. (B) The Z-scores of normalized counts (the trimmed mean of M-values) for genes of interest were compared between the two groups. ∗, FDR < 0.05. (C) IHC for CYP21A2 and HE staining were performed on CPA (n = 6) and a control (NFA, n = 1). Representative images are shown. (D) IHC for CYB5A and HE staining were performed on adrenal non-tumor tissues adjacent to CPA (n = 7) and controls (n = 7, five adjacent to Pheo and two adjacent to APA). Representative images are shown. (E) Two steroid metabolites (cortisol and 11-DOC) were measured in fresh frozen tissues from CPA (n = 8) and controls (n = 4, 2 NFA and, 2 ACC). ∗, P < 0.05. All Tissue analyses were collectively performed on premenopausal women, postmenopausal women, and men. Abbreviations: CPA, cortisol-producing adenomas; NFA, non-functioning adrenocortical adenomas; RNA-seq, RNA-sequencing; DEGs, differentially expressed genes; logFC, log2 fold-change; HE, hematoxylin and eosin; CYP21A2, cytochrome P450 21A2; CYB5A, cytochrome b5 type A; zG, zona glomerulosa; zF, zona fasciculata; zR, zona reticularis; M, medulla; 11-DOC, 11-deoxycorticosterone.

Fig. 6

Overview scheme of this study. Abbreviations: ACS, autonomous cortisol secretion; CPA, cortisol-producing adenomas; 11-DOC, 11-deoxycorticosterone; CYP21A2, cytochrome P450 21A2; CYB5A, cytochrome b5 type A; DHEA-S, dehydroepiandrosterone-sulfate; SIOP, steroid-induced osteoporosis.