MON-189 Evaluation of Adrenal Insufficiency and Recovery in Rheumatology Patients on Long-Term Glucocorticoid Therapy

Background Long-term glucocorticoid (GC) therapy is commonly used to treat rheumatological conditions. This may result in tertiary adrenal insufficiency, as a result of suppression of the HPA axis, when GC doses are weaned/withdrawn. There is little published data about tertiary adrenal insufficiency in this group. This study aims to further evaluate prevalence, characteristics and recovery of adrenal insufficiency in these patients at a large UK teaching hospital. Methods We retrospectively identified patients seen in outpatient clinics between January 2014 and September 2019 who had received tapering doses of long-term GC to treat their underlying condition (largely patients with polymyalgia rheumatica, giant cell arteritis or other vasculitis) and had either had a 9am cortisol or short synacthen test (SST). Data were collected using a standardised proforma. Results There was a total of 238 patients, median age of 71 years with a female preponderance (75%). Mean duration of glucocorticoid use was 63.3 months. Mean peak dose of glucocorticoid was 29.2mg. 142 patients had 9am cortisol as the first line test to assess adrenal function. 65% of these were considered sub-optimal based on local protocol (cortisol <350nmol/L). 38% of these patients went on to have SST, of which 56% continued to show evidence of sub-optimal cortisol production. All patients where baseline 9am cortisol was <100nmol/L failed to reach stimulated cortisol of >500nmol/L on SST, whereas 31% failed SST if 9am cortisol was 250-350 nmol/L. In total 138 SSTs were performed of which 51% (n=70) were abnormal (cortisol <500nmol/L post synacthen). When baseline cortisol was <100nmol/L on SST, all patients had a suboptimal peak response. However, where baseline cortisol on SST was >350 nmol/L only 3% had a sub-optimal peak cortisol. 32 of these patients with an abnormal baseline SST went on to have a repeat SST within 2 years. 50% (n=16) continued to be suboptimal. Of the 32 patients, 38% (n=12) were switched to hydrocortisone with 33% showing complete adrenal recovery, average time to recovery of 25 months. 62% (n=20) patients did not switch, with 60% demonstrating recovery within the same time period (p=0.05). Mean ACTH levels in patients who had sub-optimal SST were 23.1 ng/L (n=19). ACTH levels were not different between those who recovered and those who did not (p=0.23). Conclusion Our study suggests that tertiary adrenal insufficiency is highly prevalent in this cohort of patients with rheumatological conditions requiring long-term glucocorticoid therapy. A 9am cortisol threshold of greater than 350nmol/L excludes most patients with adrenal insufficiency. These data also suggest no significant difference in adrenal recovery if switched to hydrocortisone versus continuing on prednisolone. ACTH levels were not fully suppressed in patients with adrenal insufficiency and did not predict recovery.