Distinct serum steroid profiles between adrenal Cushing syndrome and Cushing disease

Abstract

Background: Differentiating between adrenal Cushing syndrome (adrenal CS) and Cushing disease (CD) can be challenging if there are equivocal or falsely elevated adrenocorticotropic hormone (ACTH) values. We aim to investigate the diagnostic value of serum steroid profiles in differentiating adrenal CS from CD.

Method: A total of 11 serum steroids in adrenal CS (n = 13) and CD (n = 15) were analyzed by liquid chromatography with tandem mass spectrometry (LC-MS/MS). Age- and gender-specific steroid ratios were generated by dividing the actual steroid concentration by the upper limit of the relevant reference range. A principal component analysis (PCA) and an orthogonal partial least squares discriminant analysis (OPLS-DA) were performed.

Results: The PCA and OPLS-DA analyses showed distinct serum steroid profiles between adrenal CS and CD. Dehydroepiandrosterone sulfate (DHEA-S), dehydroepiandrosterone (DHEA), and androstenedione ratios were identified as biomarkers for discrimination by variable importance in projection (VIP) in combination with t-tests. The sensitivity and specificity of DHEA-S ratios <0.40 were 92.31% (95% CI 64.0%-99.8%) and 93.33% (95% CI 68.1%-99.8%), respectively, in identifying adrenal CS. The sensitivity and specificity of DHEA ratios <0.18 were 100% (95% CI 75.3%-100.0%) and 100% (95% CI 78.2%-100.0%), respectively, in identifying adrenal CS.

Conclusion: Our data support the clinical use of the DHEA-S and DHEA ratios in the differential diagnosis of adrenal CS and CD, especially when falsely elevated ACTH is suspected.

Keywords: Cushing syndrome; dehydroepiandrosterone sulfate (DHEA-S); mass spectrometry; orthogonal partial least squares discriminant analysis (OPLS-DA); serum steroids.

Figure 1

Flow chart of identification of study population.

Figure 2

Principal component analysis (PCA) score plot of adrenal Cushing syndrome (adrenal CS) group and Cushing disease (CD) group. The x- and y-axes represent the first principal component and the second principal component after PCA, respectively. The model parameters were R ² X = 0.79 and Q ² = 0.36. PCA, principal component analysis.

Figure 3

The orthogonal partial least squares discriminant analysis (OPLS-DA) model for the classification of adrenal Cushing syndrome (adrenal CS) and Cushing disease (CD). (A) Score plot of first predictive component (t[1]) vs. first orthogonal component (to[1]) of established OPLS-DA model based on the adrenal CS vs. CD dataset. The model parameters were as follows: R ² X = 0.788, R ² Y = 0.899, and Q ² Y = 0.807. (B) Permutation test of 200 permutations of the OPLS-DA model. OPLS-DA, orthogonal partial least squares discriminant analysis.

Figure 4

Loading plot and S-plot for identifying putative steroid biomarkers between adrenal Cushing syndrome (adrenal CS) and Cushing disease (CD). (A) P1 loading plot of the OPLS-DA model. p[1] provided confidence intervals of each loading value. (B) S-plot of OPLS-DA for adrenal CS and CD. The p[1]-axis describes the magnitude of each variable. The p(corr)[1]-axis represents the reliability of each variable. Ideal steroid biomarkers have high magnitude and high reliability. T, testosterone; DHEA, dehydroepiandrosterone; P, progesterone; 17-OHP, 17-OH-progesterone; A4, androstenedione; DOC, 11-deoxycorticosterone; 11-DOC, 11-deoxycortisol; DHEA-S, dehydroepiandrosterone-sulfate.