Endocrine Abstracts

Background: Potentially fatal adrenal crises (AC) still occur in educated patients with adrenal insufficiency (AI). Our study investigates clinical and biochemical fingerprints of increased susceptibility to AC.

Material and methods: 199 patients with chronic AI (disease duration ≥6 years), classified into tertiles according to the frequency of AC/patient-years (group 1 no AC, group 2 ≤0.07 AC/py, group 3 >0.07 AC/py). Following parameters were assessed: 24 h urinary cortisol, salivary cortisol day profile, predefined glucocorticoid (GC) replacement score based on clinical and biochemical parameters of GC exposure, plasma and urinary catecholamines, polymorphisms (SNPs) of the glucocorticoid receptor (NR3C1), mineralocorticoid receptor (NR3C2), FKBP5, HSD11B1 and HSD11B2 and crisis management.

Results: One-hundred patients (50%) never experienced AC, whereas sixty-seven patients (34%) had frequent AC (group 3). This group received significantly higher GC replacement doses (13±5 mg hydrocortisone-equivalent /m²/day) compared to group 1 (11±3 mg hydrocortisone-equivalent /m²/day, P<0.01) and group 2 (11±3 mg hydrocortisone-equivalent /m²/day, P<0.01). On the contrary, mean GC under-replacement score was lowest in patients with frequent AC (−2.3±1.7) compared to both group 1 (−1.4±1.4, P=0.017) and group 2 (−1.0±1.1, P=0.012). 24 h-urinary metanephrine levels were lower, whereas 24 h-urinary normetanephrine levels were higher in patients from groups 2 and 3 compared to those without AC. Among patients with frequent AC, crises frequency correlated with HSD11B2 activity assessed by salivary cortisol/cortisone ratios Two NR3C1 SNPs (rs1005297, rs860457) linked to glucocorticoid resistance and three NR3C2 SNPs (rs6810951, rs11099680, rs4835493) linked to blood pressure regulation were associated with AC frequency. No significant differences in crisis management were observed between groups 2 and 3. Gastrointestinal and psychiatric comorbidities were more frequent among patients from group 2 (24% and 21%, respectively) and group 3 (22% and 21%, respectively) compared to patients without AC (10%, P=0.08 and 7%, P=0.04, respectively).

Conclusion: Higher glucocorticoid replacement doses in patients with frequent AC fit to previous observations and could simply reflect increased caution but might also be an indicator of increased vulnerability, as also suggested by the lower GC replacement scores and by the correlation with HSD11B2 activity. Gastrointestinal comorbidities might impair cortisol absorption and increase susceptibility to AC. The association between AC frequency and SNPs of NR3C1 and NR3C2also implies a genetic susceptibility to AC. These observations require validation in prospective studies.