SUN-438 - Corticosteroid Use Is Independently Associated with Radiomic Measures of the Adrenal Glands in A Population of Smokers

BACKGROUND

Respiratory diseases such Chronic Obstructive Pulmonary Disease (COPD) are commonly treated with inhaled or oral glucocorticoids (GC). This class of medications is a known disruptor of the Hypothalamic Pituitary Adrenal Axis (HPA) and can lead to secondary Adrenal Cortical Insufficiency (ACI) The risk of HPA disruption increases across doses and forms of GC, suggesting that steroid use can serve as a surrogate for GC-related HPA disruption risk. HPA disruption may result in structural changes to the adrenal glands that can be potentially assessed via computed tomography (CT). The purpose of this study was to evaluate associations between GC exposure and objective radiomic features of the adrenal glands derived from non-contrast chest CT in a large observational study of smokers with and without COPD.

METHODS

COPDGene collected high resolution chest CTs from 10,129 participants, as well as data on current and past treatment with inhaled and oral GC. Subsequently automatic segmentation of the adrenal glands was accomplished using Total Segmentator, an open-source deep learning tool that identifies numerous major organs in CT images. Gland Volume, mean and standard deviation of CT attenuation values within the glands (Volume, MA, STDattenuation) were calculated using segmentation masks. If the segmentation mask for either gland met the edge of the CT that observation was excluded; only participants with complete segmentation of both glands were retained in this analysis. The cross-sectional association between GC exposure and radiomic measures were performed with SAS 9.4 using general linear models.

RESULTS

A total of 6,843 scans were retained for analysis; while radiomic results were similar for left and right glands, left glands were evaluated. 177 participants reported oral GC and 413 reported inhaled GC. Scans were evaluated visually at acquisition and observations with incidental findings of masses were removed (n=128). In univariate analysis Volume and STDattenuation were skewed and a log transformation was applied. Age was associated with reduced MA, while height was associated with increased Volume and decreased STDattenuation; BMI, sex, and race were associated with all three radiomic measures. Multivariable models predicting each of the outcomes adjusted for all five of the covariates showed independent effects on each outcome except in the model of MA, where height was not associated with MA. Adding oral or inhaled GC use to the models revealed independent effects on each of the three outcomes (all p<=0.02) except inhaled steroids on STDattenuation (p=0.08).

CONCLUSIONS

The use of oral and inhaled steroid medications is associated with changes in CT-derived radiomics of adrenal glands, notably in attenuation and volume. Results of this cross-sectional observational study suggest that radiomic analysis may be useful in exploring HPA effects of steroid treatment.

*Unless otherwise noted, all abstracts presented at ENDO must not be released to the press or the public until the date and time of presentation. For oral presentations, the abstracts are embargoed until the session begins. The Endocrine Society reserves the right to lift the embargo on specific abstracts that are selected for promotion prior to or during ENDO.*