Suresh, Malavika

Abstract Background: Evaluation for adrenal insufficiency (AI) includes dynamic testing that is cumbersome, expensive, and may not always be accessible. Our objective was to evaluate the diagnostic accuracy of dehydroepiandrosterone sulfate (DHEAS) in diagnosing AI. Methods: This is a single-center retrospective study of patients evaluated for AI between July 10th, 2012 to January 7th, 2020 who underwent Cosyntropin stimulation testing (CST) or Insulin tolerance test (ITT), and had baseline measurement of DHEAS. Diagnosis of adrenal insufficiency was based on peak cortisol of <18 ug/dL during either CST or ITT. Results: We identified 491 patients (379 [77.2%] women, median age 42 years [range 18–84]) who met inclusion criteria. Primary adrenal insufficiency (PAI) was diagnosed in 16 (3%) patients, secondary adrenal insufficiency (SAI) was diagnosed in 211 (43%) patients, and AI was excluded in 264 (54%) patients. CST was performed in 478 (97%) patients, ITT was done in 45 (9%) patients, with 32 patients having both CST and ITT performed. Using a peak cortisol of <18 mcg/dl during CST to diagnose AI, ROC analysis of DHEAS demonstrated an area under the curve of 0.79, p<0.0001. A DHEAS cutoff of 100 mcg/dL excluded AI with a sensitivity of 93%, specificity of 41%, and a false negative rate of 3%. A DHEAS cut-off of 25 mcg/dl diagnosed AI with a false positive rate of 9% (Sensitivity 57% and specificity 85%). Notably, of 32 patients who had both CST and ITT performed, the diagnosis of AI was concordant in only 5 patients, while additional 27 patients had discordant results (26 were diagnosed with SAI based on ITT, but with a normal CST, and 1 patient was diagnosed with AI on CST but not on ITT). When excluding the 27 patients with discordant results, the overall performance of DHEAS in diagnosis of AI based on CST has not changed. Conclusion: DHEAS is a valuable baseline diagnostic test for AI. DHEAS >100 mcg/dl predicts a normal response during CST with a false negative rate of 3%, while DHEAS <25 mcg/dl is reflective of underlying adrenal insufficiency and predicts an abnormal CST with a false positive rate of 9%. In these situations, CST can be avoided.

Abstract Context: Quality of life (QoL) is impaired in patients with chronic adrenal insufficiency (AI) despite standard glucocorticoid (GC) replacement therapy. Current evidence on the determinants of QoL is scarce and limited in very few European countries, and how it relates to AI subtypes remains underexplored. Objective: We conducted the first survey study in patients with AI in the USA to determine the correlations between clinical parameters, adverse outcomes, patient education, socioeconomic factors, and QoL in different subtypes of AI. Design, Setting and Participants: Cross-sectional survey study of 529 patients with AI between 2015 and 2020, at three tertiary centers in the USA. Intervention: Patient-centered questionnaires. Main Outcome Measures: QoL scores using Short-Form 36. Results: Of 529 participants, 223 (42.2%) had primary AI (PAI), 190 (35.9%) had secondary AI (SAI), and 116 (21.9%) had glucocorticoid induced AI (GIAI). Median age at the time of survey was 58 years (IQR: 43–68), 342 (64.8%) were women and 483 (91.3%) were Caucasians. Median duration of AI was 6 years (IQR: 3–14.5), longest in patients with PAI (11 vs 4 years in SAI and GIAI, p=0.0001). Overall, Physical Composite Summary (PCS) score was lower than the Mental Composite Summary (MCS) (38.1±12.9 vs 46.5±11.8). Across the eight dimensions, each individual decade- and sex-adjusted Z-score (using the normative data of USA population) in patients with PAI was significantly higher than that of the other two AI subtypes (all p<0.05), although all Z-scores were still below 0. In multivariate analyses, patients were more likely to report a worse PCS score (<40) if they were women (OR: 3.3, CI 95%: 1.8–6.0), had SAI or GIAI (OR: 2.5, CI 95%: 1.4–4.3), had shorter duration of AI (<6 years) (OR: 2.0, CI 95%: 1.1–3.6), were treated with >25 mg hydrocortisone equivalent daily (OR: 2.3, CI 95%: 1.2–4.6), had more comorbidities related to GC excess (OR: 2.3, CI 95%: 1.3–4.0), reported higher financial burden due to AI (OR: 2.1, CI 95%: 1.3–3.6), and reported difficulties with AI management (OR: 2.5, CI 95%: 1.2–5.2). Women (OR: 2.1, CI 95%: 1.08–4.0), shorter duration of AI (OR: 2.4, CI 95%: 1.4–4.3), higher financial burden due to AI (OR: 2.3, CI 95%: 1.3–4.0), reporting difficulties with AI management (OR: 2.6, CI 95%: 1.4–4.9), and lack of family support during adrenal crisis (OR: 9.1, CI 95%: 2.3–33.3), were predictors of a worse MCS score (<40). Conclusions: Patients with AI have substantially impaired QoL despite GC replacement therapy. Certain determinants of QoL are modifiable and achievable, such as avoiding GC over-replacement, offering detailed hands-on education in self-management, more comprehensive insurance coverage, and more robust domestic support. Our study calls for a multidimensional effort from patients, clinicians, and society to improve QoL in this vulnerable patient population.