The Prevalence of Adrenal Insufficiency in Kidney Transplant Recipients

Abstract

Abstract Long term glucocorticoid therapy is a key component of immunosuppression for kidney transplant recipients (KTR), leading to a significant cumulative glucocorticoid exposure in this cohort. However, assessment of adrenal function and steroid education do not feature in clinical practice guidelines for KTR. This study aims to investigate the prevalence of adrenal insufficiency (AI) in KTR on maintenance prednisolone and develop a simple screening algorithm to identify patients at highest risk of AI. 67 KTR receiving prednisolone were recruited, with a median age of 52 yrs (IQR 39–64) and median time since transplantation of 79 months (IQR 46–146). Patients with known hypothalamic-pituitary-adrenal axis pathology, exposure to oral oestrogens or impaired graft function (creatinine >170 umol/l) were excluded. Participants underwent a short synacthen test [SST, Roche Elecsys II immunoassay, pass cortisol > 430 nmol/l, CV assay 1.8-2,1%)], assessment of lifetime glucocorticoid exposure and an assessment of metabolic health.72% (n=48) of participants failed the SST(30 min cortisol < 430nmol/l). Participants with AI had a higher daily prednisolone dose (4.9 vs 4.2mg/day, p=0.002), a greater cumulative glucocorticoid exposure (289 vs 111 mg/kg, p=0.03) and lower estimated glomerular filtration rate (63 vs 81 ml/min/1.73m2, p=0.03) than those with intact adrenal function. There was no difference in duration of glucocorticoids post transplantation, previous transplant history, pre-transplant glucocorticoid exposure, immunosuppression regime, exposure to other glucocorticoid preparations, BMI and cortisol binding globulin concentration between those who passed or failed a SST. Participants with AI had a lower baseline cortisol (143 vs 303 nmol/l, p<0.001), lower baseline serum ACTH (14 vs 20 pg/ml), a lower peak stimulated cortisol (325 vs 490 nmol/l, p< 0.001) and lower cortisol response (139 vs 189 nmol/l, p=0.013) to synacthen testing than those with normal adrenal function. A morning cortisol of <294 nmol/l predicted AI result with 100% sensitivity (95% CI 92–100%, p< 0.001) and 63% specificity, therefore AI could be excluded if morning cortisol >300nmol/l, reducing the need for stimulatory testing. Our results suggest KTR are at higher risk for AI than previously reported, highlighting a need for greater patient and physician awareness regarding AI. We suggest that KTR with a morning cortisol < 300nmol/l are at high risk of AI, and dynamic testing should be considered for these patients.