RESUMEN
The variable presence of adrenal insufficiency (AI) due to hypocortisolemia (HC) in patients with thalassemia is well established; however, the prevalence of adrenocortical hypofunction (ACH) in the zona glomerulosa and zona reticularis of the adrenal cortex is unknown. To establish the prevalence of ACH, we examined the cortisol response to 1-µg and 250-µg ACTH tests, plasma aldosterone (A)/plasma renin activity (PRA) ratio, and serum dehydroepiandrosterone sulfate (DHEAS) levels in a large cohort of patients with thalassemia, and to investigate the impact of total body iron load (TBIL) on adrenocortical function. The setting used was University hospital and government-based tertiary care center. One hundred twenty-one (52 females) patients with ß-thalassemia major (ß-TM) and 72 healthy peers (38 females) were enrolled. The patients underwent a 250-µg cosyntropin test if their peak cortisol was <500 nmol/L in a 1-µg cosyntropin test. Magnetic resonance imaging (MRI) was performed to assess the MRI-based liver iron content and cardiac MRI T2* iron. The associations between ACH and TBIL were investigated. The patients with thalassemia had lower ACTH, cortisol, DHEAS, and A/PRA values compared with the controls (p < 0.001). Thirty-nine patients (32.2 %) had HC [primary (n = 1), central (n = 36), combined (n = 2)], and 47 (38.8 %) patients had reduced DHEAS levels; 29 (24.0 %) patients had reduced A/PRA ratios. Forty-six (38.0 %) patients had hypofunction in one of the adrenal zones, 26 (21.5 %) had hypofunction in two adrenal zones, and 9 (7.4 %) had hypofunction in all three zones. Patient age and TBIL surrogates were significant independent parameters associated with ACH. Cardiac MRI T2* iron was the only significant parameter that predicted the severity of ACH at a cut-off of 20.6 ms, with 81 % sensitivity and 78 % specificity. Patients with thalassemia have a high prevalence of AI due to HC and zona glomerulosa and zona reticularis hypofunction. TBIL surrogates can predict ACH, but cardiac iron was the only surrogate that was adequately sensitive to predict the severity of ACH.
