Meeting NewsPerspective

Serum steroid levels can help differentiate adrenal disorders

LOS ANGELES — Three widely available serum steroid precursor measurements are highly predictive for diagnosing adrenal cortical carcinoma and are also useful to differentiate patients with subtypes of Cushing’s syndrome and nonfunctional adrenal tumors, according to study data presented at the AACE Annual Scientific and Clinical Congress.

Shobana Athimulam

“Patients with adrenal disorders typically present with a wide variety of symptoms; however, imaging, preliminary labs and clinical appearance might not tailor us toward the correct diagnosis,” Shobana Athimulam, MBBS, a clinical fellow in the division of endocrinology, diabetes and nutrition at the Mayo Clinic in Rochester, Minnesota, told Endocrine Today. “It is sometimes more complicated, but these serum steroid precursors can be widely measured in all reference laboratories, and they can help us differentiate important clinical dilemmas in managing adrenal disorders.”

Athimulam and colleagues analyzed data from 370 adults with any type of Cushing’s syndrome or adrenal mass who were not taking exogenous steroids. Participants were prospectively enrolled over 3 years at the Mayo Clinic (62% women; median age, 59 years; median BMI, 31.78 kg/m²). Researchers measured levels of 11-deoxycortisol, 17-hydroxypregnenolone, 17-hydroxyprogesterone, pregnenolone and androstenedione via liquid chromatography-mass spectrometry, and levels of dehydroepiandrosterone sulfate (DHEAS) and cortisol via automated chemiluminescent competitive immunoassay.

Within the cohort, 128 had nonfunctioning adrenal tumors, 63 had mild autonomous cortisol secretion (subclinical Cushing’s syndrome), 34 had aldosteroma, 33 had pituitary Cushing’s syndrome, 26 had pheochromocytoma, 22 had other benign adrenal masses, 21 had adrenal Cushing’s syndrome, 19 had adrenocortical carcinoma, 17 had other adrenal malignancies and seven had ectopic Cushing’s syndrome.

Researchers found that levels of all serum steroid were between 2.5 and 12 times higher in patients with adrenocortical carcinoma vs. patients with other adrenal tumors (P < .0001 for all). In a multivariable model, three steroid precursors accurately predicted adrenocortical carcinoma (receiver operating characteristic area under the curve, 0.99; P < .0001), Athimulam said.

“In our 19 patients with adrenocortical carcinoma, we found significant elevation — between 2 and 12 times — of multiple steroid precursors,” Athimulam said. “The combined utility of just three of these precursors, which are 17-hydroxypregnenolone, pregnenolone and 11-deoxycortisol, accurately predicted [adrenocortical carcinoma].”

In patients with Cushing’s syndrome, levels of 17-hydroxypregnenolone (P = .03) and DHEAS (P = .0003) were lowest and 11-deoxycortisol levels were highest (P < .0001) when compared with patients with mild autonomous cortisol secretion and nonfunctioning adrenal tumors, Athimulam said.

“The rationale behind this is when you have an adrenal tumor producing cortisol excess, the tumor mass itself is what produces 11-deoxycortisol, and that is elevated,” Athimulam said. “But 17-hydroxypregnenolone and DHEAS are actually produced by the normal adrenal cortex, and this production is suppressed when there is cortisol excess.”

In patients with mild autonomous cortisol secretion, 17-hydroxypregnenolone level was a predictor of the condition when compared with patients with nonfunctioning adrenal tumors (P = .04), Athimulam said.

Furthermore, in assessing patients with adrenocorticotropic hormone-dependent Cushing’s syndrome, patients with ectopic Cushing’s syndrome had higher levels of 11-deoxycortisol (P = .005) and 17-hydroxypregnenolone (P = .04), with the findings for 11-deoxycortisol persisting after adjusting for cortisol concentrations (P = .02), Athimulam said.

“We were very excited with the findings,” Athimulam said in an interview. “These are common dilemmas that endocrinologists face when managing adrenal disorders. What we found was that all the serum precursors were significantly elevated. It can help with managing patients for all clinical endocrinologists, not only someone in a research setting or a tertiary center.” – by Regina Schaffer

Reference:

Bancos I, et al. Clinical applicability of serum steroid precursors in the diagnosis of adrenal tumors and Cushing’s syndrome. Presented at: AACE Annual Scientific and Clinical Congress; April 24-28, 2019; Los Angeles.

Disclosure: Athimulam reports no relevant financial disclosures.

LOS ANGELES — Three widely available serum steroid precursor measurements are highly predictive for diagnosing adrenal cortical carcinoma and are also useful to differentiate patients with subtypes of Cushing’s syndrome and nonfunctional adrenal tumors, according to study data presented at the AACE Annual Scientific and Clinical Congress.

Shobana Athimulam

“Patients with adrenal disorders typically present with a wide variety of symptoms; however, imaging, preliminary labs and clinical appearance might not tailor us toward the correct diagnosis,” Shobana Athimulam, MBBS, a clinical fellow in the division of endocrinology, diabetes and nutrition at the Mayo Clinic in Rochester, Minnesota, told Endocrine Today. “It is sometimes more complicated, but these serum steroid precursors can be widely measured in all reference laboratories, and they can help us differentiate important clinical dilemmas in managing adrenal disorders.”

Athimulam and colleagues analyzed data from 370 adults with any type of Cushing’s syndrome or adrenal mass who were not taking exogenous steroids. Participants were prospectively enrolled over 3 years at the Mayo Clinic (62% women; median age, 59 years; median BMI, 31.78 kg/m²). Researchers measured levels of 11-deoxycortisol, 17-hydroxypregnenolone, 17-hydroxyprogesterone, pregnenolone and androstenedione via liquid chromatography-mass spectrometry, and levels of dehydroepiandrosterone sulfate (DHEAS) and cortisol via automated chemiluminescent competitive immunoassay.

Within the cohort, 128 had nonfunctioning adrenal tumors, 63 had mild autonomous cortisol secretion (subclinical Cushing’s syndrome), 34 had aldosteroma, 33 had pituitary Cushing’s syndrome, 26 had pheochromocytoma, 22 had other benign adrenal masses, 21 had adrenal Cushing’s syndrome, 19 had adrenocortical carcinoma, 17 had other adrenal malignancies and seven had ectopic Cushing’s syndrome.

Researchers found that levels of all serum steroid were between 2.5 and 12 times higher in patients with adrenocortical carcinoma vs. patients with other adrenal tumors (P < .0001 for all). In a multivariable model, three steroid precursors accurately predicted adrenocortical carcinoma (receiver operating characteristic area under the curve, 0.99; P < .0001), Athimulam said.

“In our 19 patients with adrenocortical carcinoma, we found significant elevation — between 2 and 12 times — of multiple steroid precursors,” Athimulam said. “The combined utility of just three of these precursors, which are 17-hydroxypregnenolone, pregnenolone and 11-deoxycortisol, accurately predicted [adrenocortical carcinoma].”

In patients with Cushing’s syndrome, levels of 17-hydroxypregnenolone (P = .03) and DHEAS (P = .0003) were lowest and 11-deoxycortisol levels were highest (P < .0001) when compared with patients with mild autonomous cortisol secretion and nonfunctioning adrenal tumors, Athimulam said.

PAGE BREAK

“The rationale behind this is when you have an adrenal tumor producing cortisol excess, the tumor mass itself is what produces 11-deoxycortisol, and that is elevated,” Athimulam said. “But 17-hydroxypregnenolone and DHEAS are actually produced by the normal adrenal cortex, and this production is suppressed when there is cortisol excess.”

In patients with mild autonomous cortisol secretion, 17-hydroxypregnenolone level was a predictor of the condition when compared with patients with nonfunctioning adrenal tumors (P = .04), Athimulam said.

Furthermore, in assessing patients with adrenocorticotropic hormone-dependent Cushing’s syndrome, patients with ectopic Cushing’s syndrome had higher levels of 11-deoxycortisol (P = .005) and 17-hydroxypregnenolone (P = .04), with the findings for 11-deoxycortisol persisting after adjusting for cortisol concentrations (P = .02), Athimulam said.

“We were very excited with the findings,” Athimulam said in an interview. “These are common dilemmas that endocrinologists face when managing adrenal disorders. What we found was that all the serum precursors were significantly elevated. It can help with managing patients for all clinical endocrinologists, not only someone in a research setting or a tertiary center.” – by Regina Schaffer

Reference:

Bancos I, et al. Clinical applicability of serum steroid precursors in the diagnosis of adrenal tumors and Cushing’s syndrome. Presented at: AACE Annual Scientific and Clinical Congress; April 24-28, 2019; Los Angeles.

Disclosure: Athimulam reports no relevant financial disclosures.

    Perspective
    Richard Auchus

    Richard Auchus

    This is a great study and a good start. We all need to get our minds thinking in the direction of profiling the panel to determine tumor type and trajectory. The main issue is distinguishing adrenal cancer. We waste an enormous amount of money on repeat CT scanning with patients to rule out cancer when only .01% of them are. It would be great to do a blood test and say, ‘No, you don’t need a scan.’ These data look promising; however, there were only 19 patients with adrenal cancer in the cohort. We also know that adrenal cancers are heterogenous. This study needs to be validated with a larger number of patients with adrenal cancer.

    Distinguishing subtypes of Cushing’s syndrome using serum steroid measurements may prove more difficult. Ultimately, it is the quantity of cortisol made that matters. Perhaps, in combination with dynamic testing, this is something that may be useful. When it comes to cortisol levels, we still struggle with who to send to surgery. The milder it is, the harder it is to demonstrate an improvement. When we perform surgery, it is often not to fix a problem that exists, but instead to prevent problems from occurring later. A good test would answer two key questions: Does this person have cortisol excess, and is this person going to benefit from surgery? I find it hard to understand how determining intermediate measures is going to answer such questions, but this research is a step in the right direction.

    • Richard Auchus, MD, PhD
    • Professor of Pharmacology and Internal Medicine
      Division of Metabolism, Endocrinology and Diabetes
      The University of Michigan

    Disclosures: Auchus reports he is a consultant for Quest Diagnostics.

    See more from American Association of Clinical Endocrinologists Annual Meeting