Intratumoral Steroid Profiling of Adrenal Cortisol-Producing Adenomas by Liquid Chromatography- Mass Spectrometry

Abstract

Endogenous Cushing syndrome (CS) is an endocrine disorder marked by excess cortisol production rendering patients susceptible to visceral obesity, dyslipidemia, hypertension, osteoporosis and diabetes mellitus. Adrenal CS is characterized by autonomous production of cortisol from cortisol-producing adenomas (CPA) via adrenocorticotropic hormone-independent mechanisms. A limited number of studies have quantified the steroid profiles in sera from patients with CS. To understand the intratumoral steroid biosynthesis, we quantified 19 steroids by mass spectrometry in optimal cutting temperature compound (OCT)-embedded 24 CPA tissue from patients with overt CS (OCS, n=10) and mild adrenal cortisol excess (MACE, n=14). Where available, normal CPA-adjacent adrenal tissue (AdjN) was also collected and used for comparison (n=8). Immunohistochemistry (IHC) for CYP17A1 and HSD3B2, two steroidogenic enzymes required for cortisol synthesis, was performed on OCT sections to confirm the presence of tumor tissue and guided subsequent steroid extraction from the tumor. LC-MS/MS was used to quantify steroids extracted from CPA and AdjN. Our data indicated that CPA demonstrated increased concentrations of cortisol, cortisone, 11-deoxycortisol, corticosterone, progesterone, 17OH-progesterone and 16OH-progesterone as compared to AdjN (p<0.05). Compared to OCS, MACE patient CPA tissue displayed higher concentrations of corticosterone, 18OH-corticosterone, 21-deoxycortisol, progesterone, and 17OH-progesterone (p<0.05). These findings also demonstrate that OCT-embedded tissue can be used to define intra-tissue steroid profiles, which will have application for steroid-producing and steroid-responsive tumors.

INTRODUCTION

Endogenous Cushing syndrome (CS) is an endocrine disorder marked by excess cortisol production¹. Hypertension, obesity, osteoporosis, and hyperglycemia are frequently associated with CS, which contribute to increased morbidity and mortality in these patients². While the most common cause of CS is pharmacological intake of glucocorticoids (exogenous/iatrogenic CS), excess secretion of adrenocorticotropic hormone (ACTH) by the pituitary gland accounts for ~80% of endogenous CS cases and is often referred to as Cushing disease. Adrenal CS constitutes the remaining ~20% of endogenous cases, wherein ACTH–independent, autonomous cortisol excess occurs from the adrenal glands¹.

Adrenal CS is commonly caused by cortisol-producing adenomas (CPA) – benign, unilateral adrenal tumors. Mild adrenal cortisol excess (MACE) is present in a subset of adrenal tumors, that are generally found incidentally during unrelated imaging procedures³. CPA formation is associated with somatic mutations that contribute to the development of benign adrenal tumors. Somatic mutations in genes of the cAMP/PKA pathway, primarily to the gene encoding the catalytic subunit of protein kinase A (PRKACA), have been found in approximately half of patients with CPA and are more common in overt CS (OCS) than in MACE^4–6. Somatic mutations in other genes including β-catenin (CTNNB1) and the α-subunit of the stimulatory G protein (GNAS) have also been linked to CPA development⁷. While these mutations have been well-documented^8–12, the differential effects of each mutation on adrenal steroid production remain poorly defined.

OCS is rare and refers to a disease state in which hallmark physical manifestations of disease, such as abnormal fat distribution, buffalo hump, moon face, and skin striae are present^13,14. Conversely, MACE affects 0.2–2% of the adult population and is the most common form of hormonal excess associated with adrenal incidentalomas. Patients with MACE display hormonal evidence for cortisol excess, but only subtle physical signs of disease^13,15–18. Nevertheless, the diagnosis of MACE is associated with an increased risk of adverse cardiovascular events^2,19.

The use of liquid chromatography with tandem mass spectrometry (LC-MS/MS) as a tool for quantitative measurement of steroid hormones has continued to expand over the last decade²⁰. A limited number of studies have used LC-MS/MS to measure steroids in plasma/serum from patients with CS ^19,21–24. The present investigation is the first use of LC-MS/MS to measure intratumoral levels of 19 steroids in optimal cutting temperature compound (OCT)-embedded CPA tissue specimens obtained from patients with CS. In this proof-of-concept study, we have also sought to identify distinguishing steroid patterns for adrenal tumors causing OCS and MACE.

MATERIALS AND METHODS

Surgical collection of tumor and normal adrenal tissues and OCT preparation

Twenty-four patients diagnosed with ACTH-independent CS who underwent unilateral adrenalectomy at University of Michigan were included in the study. The diagnosis of CS was established as per the Endocrine Society’s clinical practice guideline²⁵ or the institutional consensus at that time. OCS was distinguished from MACE if the patients presented with the hallmark physical characteristics of CS such as skin striae, buffalo hump, or moon face. The availability of archival OCT-embedded blocks of resected CPA determined the inclusion of the patients in the study. This study was approved by the institutional review board at the University of Michigan (HUM00106798) and use of OCT-embedded CPA tissue in this study was consented by the patients.

Immunohistochemistry to define the tumor region in CPA

Immunohistochemistry (IHC) on FFPE sections of CPA was performed as previously described in aldosterone-producing adenomas^26–28. Unlike FFPE sections that were typically intact, the frozen OCT tissue slices were small fragments selected from the surgical resection of the CPA. Where available, normal adrenal tissue adjacent to the CPA (AdjN) was also separately collected and embedded in OCT. For patients for whom there was available FFPE and OCT samples, IHC was performed on both samples. IHC of 5 μm thick OCT-embedded CPA and AdjN tissue was performed to confirm the tumor phenotype of the slide. OCT sections were thawed at room temperature and fixed in chilled acetone at −20°C for 10 min. IHC for the key steroidogenic enzymes in the cortisol biosynthesis pathway, namely 17α-hydroxylase/17,20-lyase (CYP17A1) and 3β-hydroxysteroid dehydrogenase (HSD3B2) (Fig. 1), was performed on both CPA and AdjN samples. Following fixation with acetone and peroxidase blocking of the sections, OCT sections were incubated for 1 hour at room temperature in a humidity chamber with anti-human rabbit polyclonal antibodies against CYP17A1 (diluted 1:2000; LifeSpan Biosciences, Cat # LS-B14227) and HSD3B2 (diluted 1:5000, provided kindly by Dr. Richard Parker)²⁹. FFPE sections were incubated with an anti-human mouse monoclonal HSD3B2 antibody (diluted 1:5000, provided kindly by Dr. Celso Gomez-Sanchez) and the same CYP17A1 anti-human rabbit antibody as used for OCT sections. The Polink-2 HRP Plus Rabbit DAB System (GBI Labs, Bothell, WA) was used to stain the sections. Slides were counterstained with Harris hematoxylin for 10-20 seconds, dehydrated, cover slipped, and scanned.

Figure 1.

Comparison of the Immunohistochemistry for key steroidogenic enzymes in the cortisol biosynthesis pathway in FFPE CPA sections (panels A-F) and their respective OCT (panels G-L) sections. The FFPE and OCT CPA (panels A-L), and the OCT AdjN (panels G-O) slides were stained for CYP17A1, HSD3B2 and hematoxylin and eosin (HE). The FFPE and the corresponding OCT sections have been derived from the same OCS and MACE CPA respectively. While AdjN indicates a zone-based expression pattern of CYP17A1 and HSD3B2 (panels M-O), both OCS and MACE CPA demonstrate co-localization of these two enzymes (panels G-L) throughout the section. This enzyme co-expression in the CPA facilitates the excess synthesis of cortisol and related glucocorticoid precursors. OCS, Overt Cushing’s Syndrome; MACE, mild adrenal cortisol excess; AdjN, Adjacent normal adrenal tissue. Scale, 5mm.

Processing of adrenal tissue for steroid extraction

OCT blocks of CPA and AdjN were used to prepare 5 μm serial sections. For each sample, three consecutive slides were used for hematoxylin and eosin (HE), CYP17A1, and HSD3B2 IHC. The subsequent five to seven slides were used for steroid extraction, depending on the size of the adrenal tumor. IHC confirmed the presence of CYP17A1 and HSD3B2 within the tumor and identified the tumor region of the sample on the slide. CPA and AdjN tissue was scraped using a chilled scalpel and placed in 850 μL cold phosphate buffered saline (PBS) within a Lysing Matrix D tube (MP Biomedicals, Solon, OH). Using a FastPrep bead-milling homogenizer, the tissue was ground and subsequently centrifuged at 4°C. The supernatant was stored on ice and 50 μL of the supernatant was aliquoted and stored at −20°C for protein quantification. The remaining PBS homogenate (400 μL for Δ4 steroids, 200 μL for Δ5 steroids) was mixed with internal standard steroids of known concentration in 675 μL and 500 μL of methanol respectively. The mixture was stored at −80°C until the time of steroid extraction. Nineteen steroids were quantified in each sample using LC-MS/MS (Fig. 2).

Figure 2.

Adrenal steroidogenesis pathway. Steroids measured by LC-MS/MS in this study are boxed. The key steroidogenic enzymes responsible for cortisol biosynthesis, namely CYP17A1 and HSD3B2, which were used for IHC to identify and confirm the tumor region, are marked in blue and red respectively.

Steroid extraction and quantification by LC-MS/MS

Steroids were extracted by the Folch extraction method as previously described³⁰. Unlabeled and deuterium-labeled steroid standards were obtained from Sigma Aldrich (St. Louis, MO, USA), Cerilliant (Round Rock, TX, USA), C/D/N Isotopes (Pointe-Claire, QC, Canada), Cambridge Isotope Laboratories (Tewksbury, MA, USA), and Steraloids (Newport, RI, USA). For the extraction of Δ4 steroids, 1250 μL chloroform was added to the thawed PBS homogenate-methanol mix on ice. This mixture was then vortexed at 1500 rpm for 30 min and centrifuged at 2500 rpm for 5 min. The lower organic phase was separated and dried under nitrogen gas at room temperature. The dried steroids were dissolved with 1 mL ethyl acetate and washed with 300 μL 10% sodium chloride. The mixture was vortexed at 1500 rpm for 5 min and centrifuged at 2500 rpm for 5 min. The upper organic phase was collected and dried under nitrogen at room temperature. The dried residue of each sample was stored at −80°C until LC-MS/MS analysis. For steroid quantification, the residue was reconstituted in 100 μL of methanol: water (1:1) and transferred to a 250 μL vial insert.

For the extraction of Δ5 steroids, 1000 μL chloroform was added to the thawed PBS homogenate-methanol mix on ice. As described above, after vortexing and centrifuging the mixture, the lower organic phase was separated and dried under nitrogen gas at room temperature. The dried steroids were dissolved with 1 mL ethyl acetate and washed with 300 μL 10% sodium chloride. The upper organic phase was collected and dried under nitrogen at room temperature. The dried residue was incubated in 100 μL reagent containing 2-picolinic acid, 2-methyl-6-nitrobenzonic anhydride, and 4-(dimethylamino) pyridine (25 mg, 20 mg, and 10 mg per 1 mL of acetonitrile, respectively) at room temperature for 90 minutes³¹. The reaction was then quenched with 350 μL water, steroids were extracted with 1 mL methyl tert-butyl ether (MTBE) and dried under nitrogen at room temperature. The dried residue of each sample was stored at −80°C until LC-MS/MS analysis. For LC-MS/MS steroid quantification, the residue was reconstituted in 300 μL of methanol: water (1:1) and transferred to a 350 μL vial insert.

Extracted CPA and AdjN samples (20 μL) were injected via autosampler and resolved with a pair of Agilent 1260/ 1290 binary pump HPLCs via 2D liquid chromatography. Two columns were used: a 10 mm x 3 mm, 3 μm particle size Hypersil GOLD C4-loading column (Thermo Scientific, Waltham, Massachusetts) and a Kinetex 150 mm x 2.1 mm, 2.6 μm particle size biphenyl resolving column (Phenomenex, Torrance, CA). The mobile phases consisted of 0.2 mmol/L aqueous ammonium fluoride (A) and methanol with 0.2 mmol/L ammonium fluoride (B). After elution, steroids were directed into the source of an Agilent 6495 triple quadrupole mass spectrometer using electrospray ionization in positive ion mode and analyzed using multiple reaction monitoring (MRM) mode. Steroid quantification was performed as previously described^32,33. While the double extraction method (Folch and ethyl acetate extraction) allowed better resolution of the chromatography peaks, the recoveries of all the steroids from the OCT-embedded adrenal control tissue ranged from 19-58%. The recoveries were measured by quantifying the response (area under the curve) of the deuterated internal standards for each steroid in extracted OCT samples spiked with a mid-range standard vs. non-extracted standards. The extraction variability of all steroids between four aliquots of the same OCT-embedded adrenal control tissue, spiked with 1000 pg/mL of each steroid, ranged between 2-17%. The authors also evaluated the post-preparative stability of the extraction by re-injecting the extracted samples after one batch analysis and two days after being placed in the sample tray of autosampler. No significant changes were observed between the two batches.

The adrenals used in the study were embedded in OCT for varying periods - between 2 to 22 years (Supplemental Table 1). We observed that steroid content in the adrenal tissue used in the study did not correlate with the age of the OCT block. This observation argues strongly against significant steroid degradation during storage of OCT-embedded tissue at −80°C for over two decades.

Protein assay and steroid normalization

Protein concentration of the PBS homogenate was determined by a bicinchoninic acid protein assay using the microbicinchoninic acid protocol (Thermo Scientific, Waltham, MA). The protein concentrations were used to normalize steroid concentrations to units of pg/μg protein.

Data analysis and statistics

GraphPad Prism was used to perform statistical analysis on the steroid data in order to determine significant differences in steroid production between groups. The Mann-Whitney test was used to compare steroid concentrations between AdjN and all CPA; and between OCS and MACE, respectively. Wilcoxon signed rank test was utilized to compare CPA with their matched AdjN. Principal component analysis (PCA) was used to reduce the steroid profile to a two-dimensional space based on the first two principle components, which visually aids the comparison in steroids between OCS and MACE. Significance was defined by a two-tailed p value <0.05. Data are presented as medians with interquartile ranges.

RESULTS

Patient demographics and characteristics

Of 24 patients included in this study, 10 had a clinical diagnosis of overt CS and 14 were diagnosed with MACE (Table 1). AdjN samples were available in 8 cases. All patients were females with ages between 25-76 years (y) with a median age of 47.0 y at the time of adrenalectomy. Notably, a majority of the patients included in the study (71%) presented with hypertension. Serum cortisol both in the morning and following 1 mg dexamethasone suppression test and 24-hour urinary free cortisol were significantly higher in patients with overt CS compared to those with MACE (Table 1, p<0.05).

Table 1.

Patient demographics and clinical characteristics.

Parameter	Overt CS	Mild Adrenal Cortisol Excess
N	10	14
Sex (M/F)	0/10	0/14
Age (y)	44.5 (32.0-47.0)*	56.0 (46.0-64.0) a*
Race (n/%)
White	10/100.0%	13/92.9%
Black	0/0.0%	0/0.0%
Asian	0/0.0%	1/7.1%
Body Mass Index (kg/m2)	32.7 (30.3-38.3) b	32.6 (28.7-36.7) b
Hypertension (Yes/No)	8/2	9/5
Tumor Characteristics
Size (cm)	3.6 (3.1-3.9)	3.0 (2.7-4.0)
Side (Left/Right)	5/3 c	10/2 c
Serum steroids
Cortisol (μg/dL)	24.9 (17.3-27.6) d**	11.0 (9.6-16.8) d**
DST Cortisol (μg/dL)	20.7 (11.8-28.6) e*	3.7 (2.8-8.1) *
DHEA-S (μg/dL)	26.0 (15.0-33.0) f	62.5 (15.3-84.0) f
24 hour Urinary free Cortisol (μg/24h)	206.0 (158.9-407.3) g***	53.0 (26.0-60.0) g***
Morning ACTH (% of patients with concentrations ≤10 pg/mL)	9 (100.0%) h	9 (75.0%) h

Demographic patient data and routine biochemical test results for patients screened for CS according to diagnosis have been presented in the table. DST, 1 mg Dexamethasone Suppression Test. Data are represented as median with the interquartile range in parentheses.

^aAge data was not available for 1 MACE patient.

^bBody mass index data was not available for 1 OCS and 2 MACE patients.

^cTumor side data levels was not available for 2 OCS and 2 MACE patients.

^dMorning serum cortisol levels were not available for 1 OCS and 3 MACE patients.

^eSerum cortisol levels following 1 mg DST were not available for 6 OCS patients.

^fSerum DHEA-S levels were not available for 5 OCS and 4 MACE patients.

^gUrinary free cortisol measurements were not available for 3 OCS and 5 MACE patients.

^hMorning ACTH levels were not available for 1 OCS and 2 MACE patients.

Data was compared using the Mann-Whitney nonparametric unpaired t-test,

^*p<0.05,

^**p<0.01,

^***p<0.001 between OCS and MACE.

IHC revealed a unique enzymatic signature of CPA

IHC was performed to determine the expression of CYP17A1 and HSD3B2, two key steroidogenic enzymes involved in adrenal cortisol production (Fig. 1 and 2). IHC of CYP17A1 and HSD3B2 demonstrated that both OCS CPA and MACE CPA, co-expressed CYP17A1 and HSD3B2 in the tumor, and unlike the AdjN tissue, there was no zone-specific expression pattern of these enzymes (Fig. 1). As seen in Fig. 1, immunostained FFPE sections provide a much more complete picture of the tumor while OCT sections represent the smaller pieces of tumor that were collected. While there was some slight variability between individual tumors, all tumors displayed similar patterns of co-expression of HSD3B2 and CYP17A1. Identification of the tumor region by HE staining (Fig. 1) as well as CYP17A1 and HSD3B2 co-localization provided confirmation that the surgically removed tumor tissue provided was in fact CPA and not normal adjacent adrenal tissue or adipose tissue.

Steroid profiling of OCT-embedded adrenal tissues

Comparisons in steroid concentrations were made between AdjN and all CPA (OCS and MACE combined) as well as between OCS and MACE samples. All CPA demonstrated increased concentrations of cortisol, cortisone, 11-deoxycortisol, corticosterone, progesterone, 17OH-progesterone and 16OH-progesterone as compared to AdjN (p<0.05, Table 2). Largest elevations were observed for cortisol and cortisone (5- and 5.7-fold respectively). In contrast, most CPA displayed undetectable levels of aldosterone. Comparatively, out of the 8 total AdjN samples, 3 and 6 tissues exhibited detectable levels of 18-oxocortisol and aldosterone respectively. Comparison of the 8 CPA-AdjN matched tissue pairs yielded similar observations as the all CPA vs. AdjN analysis (Supplemental table 1). CPA demonstrated significant increases in cortisol, progesterone and 18OH-corticosterone (p<0.05). The other steroids that demonstrated increased levels in all CPA vs. AdjN (Table 2), also showed an increasing trend in the CPA when compared with their matched AdjN (Supplemental Table 1). Compared to OCS, MACE patient tissue displayed significantly higher concentrations of corticosterone, 18OH-corticosterone, 21-deoxycortisol, progesterone, and 17OH-progesterone (p<0.05, Table 2).

Table 2.

Steroid concentrations of AdjN and CPA as measured by LC-MS/MS

Steroid pg/μg protein	Adjacent Normal Adrenal Tissue (n=8)	All CPA (n=24)	Overt CS CPA (n=10)	MACE CPA (n=14)
Pregnenolone	54.2 (41.2-116.7)	91.7 (27.6-155.2)	87.3 (31.3-120.7)	129.7 (26.7-274.1)
17OH-Pregnenolone	18.5 (5.0-28.1)	24.7 (11.5-61.1)	18.5 (8.1-74.5)	28.0 (12.8-61.1)
Progesterone	4.5 (3.1-24.5)	12.4 (8.9-166.3) a	9.9 (5.5-15.7)	45.9 (10.5-336.8) b
17OH-Progesterone	9.6 (4.5-65.0)	38.3 (16.5-99.9) a	20.2 (14.2-58.2)	74.8 (23.0-228.2) b
16OH-Progesterone	1.5 (0.5-11.2)	6.5 (2.7-38.4) a	4.5 (2.4-9.1)	25.2 (3.4-49.3)
11-DOC	0.6 (0.4-0.9)	0.7 (0.4-1.4)	0.8 (0.4-1.1)	0.6 (0.4-1.9)
Corticosterone	7.7 (6.9-38.6)	30.5 (12.9-153.8) a	21.2 (8.7-36.1)	128.7 (18.8-211.1) b
18OH-Corticosterone	2.1 (0.9-5.7)	2.0 (0.5-5.0)	0.5 (0.0-2.3)	3.7 (1.4-8.2) b
Aldosterone	1.7 (0.1-4.6)	ND	ND	ND
21-deoxycortisol	0.0 (0.0-1.3)	0.31 (0.0-3.0)	0.0 (0.0-1.1)	1.7 (0.0-5.5) b
11-deoxycortisol	1.5 (1.0-4.5)	4.4 (2.3-8.6) a	8.5 (2.6-14.2)	3.1 (2.3-7.1)
Cortisol	39.1 (27.2-148.9)	195.2 (125.3-335.5)a	177.9 (98.15-268.0)	239.2 (128.1-371.0)
18OH-Cortisol	1.7 (0.8-3.4)	3.0 (1.4-5.6)	2.3 (0.9-5.3)	3.2 (2.3-8.8)
18-oxocortisol	0.0 (0.0-1.3)	ND	ND	ND
Cortisone	2.1 (1.4-4.4)	12.1 (5.4-21.2) a	9.8 (4.0-19.6)	12.1 (6.2-21.6)
DHEA	1.8 (0.6-7.0)	2.1 (0.9-3.2)	2.1 (0.9-5.0)	2.1 (0.8-2.4)
Androstenedione	1.2 (0.8-6.5)	3.3 (1.3-9.2)	6.7 (1.0-13.3)	2.9 (1.4-6.4)
11OH-Androstenedione	6.3 (0.0-21.7)	19.6 (7.3-40.8)	16.9 (5.8-35.7)	21.3 (12.6-51.2)
Testosterone	0.0 (0.0-0.4)	0.3 (0.0-0.6)	0.3 (0.0-0.5)	0.2 (0.0-1.0)

Data are presented as medians with interquartile ranges. Data presented as ‘All CPA’ is a combination of all OCS and MACE samples and was used as a comparison to AdjN. Data was compared using the Mann-Whitney non-parametric unpaired t-test and results were considered significant at p<0.05.

^aAdjN vs. All CPA, p<0.05;

^bOCS vs. MACE, p<0.05.

ND, Not detectable.

DISCUSSION

This is the first study to define the intratumoral steroid biosynthesis directly in OCT-embedded CPA and adjacent normal adrenal tissue. The use of IHC for key steroidogenic enzymes involved in cortisol biosynthesis, such as HSD3B2 and CYP17A1, allowed us to identify the CPA region on OCT sections, and guided the characterization of CPA steroid fingerprints in both OCS and MACE tumor tissue.

In concordance with recent reports that compared serum/plasma steroids in patients with adrenal CS vs. control population^22–24, our studies indicated that CPA tissue shows elevated intratumoral levels of cortisol and 11-deoxycortisol as compared to AdjN. While Masjkur et al. reported elevations in plasma levels of pregnenolone and the hybrid steroids, 18-hydroxycortisol and 18-oxocortisol, in patients with CPA as compared to reference individuals²³, we only observed an uptrend for increases in pregnenolone and 18OH-cortisol, albeit not significant. 18-oxocortisol, on the other hand, was poorly detected in most of the OCT tissue samples in our analysis. The aforementioned analyses reported decreased adrenal androgen concentrations (DHEA, DHEA sulfate, and androstenedione) in adrenal CS patients as opposed to the normal controls, as a result of a hypothalamus-pituitary-adrenal axis suppression by hypercortisolemia^19,21–24. Conversely, we found no significant differences in DHEA and androstenedione concentrations between CPA and AdjN tissue. This could be explained by the lack of androgen production from the CPA studied and the reduction of ACTH leading to less androgen production from the AdjN.

Nevertheless, in agreement with findings from Masjkur et al, we found that the OCS CPA in our study had lower, albeit insignificant, DHEA concentrations than MACE CPA. Also in line with recent studies by Di Dalmazi et al. which compared serum steroid profiles from patients with CPA vs. non-secreting adenomas^19,21,34, we found higher intratumoral corticosterone levels in CPA as compared to AdjN tissue. Principal component analysis (PCA) indicated that, while all CPA samples from the patients with OCS clustered together, 8 out of 14 MACE CPA formed a separate cluster pattern (Fig. 3). The clustering of the 6 MACE CPA to their OCS counterparts could be reflective of the spectral nature of MACE¹⁶ in terms of the severities and the related steroid output.

Figure 3.

Principal Component Analysis (PCA) 2D-scatterplot for discrimination of OCS CPA and MACE CPA for 19 steroids. Squares in red and spheres in green represent individual OCS (n=10) and MACE (n=14) CPA samples

Interestingly, we observed that the relative concentrations of individual steroids measured in OCT adrenal tissue was discordant from that found in serum^35–37. We and others have demonstrated that the most abundant steroids in human circulation are cortisol, cortisone and DHEA^35–37. In contrast, our OCT tissue steroid analysis revealed that the most abundant intratumoral steroids are cortisol, pregnenolone, 17OH-pregnenolone and 17OH-progesterone. These discrepancies (except cortisol) could be associated with the variable steroid hydrophobicity and, thereby, the retention capacity of these molecules in tissue as well as the rate of metabolism in other tissues³⁸. Sulfated steroids, including DHEA sulfate, were not included in our current tissue-related study as they have increased water solubility and are rapidly admitted into the circulation after synthesis in the tissue. The tissue levels of sulfated steroids, thereby, will not reflect the actual concentration or production of these steroids at their biosynthetic site.

It is unclear why OCT tissue from patients with MACE exhibited higher intratumoral levels of glucocorticoid precursors compared to that from patients with OCS. The underlying somatic mutations might be responsible for the varied degree of steroid synthesis in the adrenal CS subtypes. Unfortunately, the low number of samples were not sufficient to power the analysis in our study. It also might mirror a higher flux of steroid precursors to the final product that is then released and removed into the circulation. Another potential explanation for this finding might be the relative deficiency of 21-hydroxylase (CYP21A2) activity in patients with MACE as opposed to those with OCS, due to observed differences in ratios of progesterone and 17-hydroxyprogesterone to 11-deoxycorticosterone and 11-deoxycortisol, respectively. However, we did not observe any differences in the CYP21A2 mRNA expression in MACE and OCS CPA (data not shown). Recent studies have been able to discriminate CPA from other adrenal tumors such as aldosterone producing adenomas and adrenocortical carcinomas, by using genetic markers^39,40. The use of genetic markers in combination with serum and intratumoral steroid biomarkers in CPA from patients with OCS and MACE will be key to understanding the differences in pathophysiologic mechanisms of adrenal CS subtypes as well as CPA genotypes.

A limitation of our study is the small size of the OCT samples. This is primarily due to the fact that OCT-embedded tissue samples are not commonly used for tissue preservation following surgery. The second limitation of our study is the exclusively female sex of our cohort. This, however, represents the higher prevalence of adrenal CS in women⁴¹. Irrespective of these drawbacks, we were able to demonstrate that IHC-guided capture of OCT-embedded CPA tissue and adjacent normal adrenal tissue could be utilized for steroid profiling. In addition, we were able to identify a distinct intratumoral steroid signature associated with two CPA phenotypes. Moving forward, efforts should be made to associate intratumoral steroid profiles with plasma steroid profiles to bridge the gap between the source of the steroids and practical clinical biomarkers. Additionally, identification of distinct steroid fingerprints associated with CPA genotypes could potentially unveil the pathophysiology associated with different CPA-harboring mutations.