Adrenal Insufficiency following Prolonged Exogenous Steroid Treatment for Graft Versus Host Disease


Year:

Session type:

Ann Tivey1, Henry Simons1, John Murray1, Claire Higham1
1The Christie NHS Foundation Trust

Abstract

Background

Nearly 50% of patients who undergo allogeneic stem-cell transplantation will develop graft-versus-host disease (GvHD) necessitating long term steroid therapy. Patients taking courses greater than 5mg prednisolone daily (or equivalent) are at risk of adrenal insufficiency (AI) due to hypothalamic-pituitary-axis suppression, risking life-threatening adrenal crisis.  No guidelines exist for assessing adrenal function when GvHD treatment with steroids is completed. 

Method

Over two years all patients at a tertiary centre completing long-term steroid therapy for GvHD underwent a short synacthen test (SST). 30 minute cortisol >440 nmol/l was considered a ‘pass’ (local threshold). Demographic, laboratory and clinical data was extracted from electronic records. Relationship between time from GvHD diagnosis (steroid exposure surrogate) to SST and baseline cortisol was assessed using Pearson’s correlation coefficient.

Results

Thirty patients, median age 51 (18–72) were included. 27% (8/30) failed at least one SST indicating AI. These patients continued physiological steroid replacement. Four patients had two SSTs. One passed initially but resumed steroids and failed a subsequent SST. One failed initially but following weaning over 9 months passed an SST. No correlation was observed between time from GvHD diagnosis and baseline cortisol (r = −0.03, P = 0.89) or SST result (r = −0.10, P = 0.60). Descriptive analysis of clinical/laboratory features did not suggest any predictive factors for SST result (Table 1).


Total (n=30)

Passed (22)

Failed (8)

Female

14 (47%)

12 (55%)

2 (25%)

Median Age (years)

51

50

51

Mean systolic BP (mmHg)

125

124

128

BMI (kg/m2)

26.3

26.3

26.3

Unexplained weight loss

6 (20%)

4 (13%)

2 (25%)

Hyponatraemia

1 (3%)

1 (3%)

0

Hyperkalaemia

0

0

0

Conclusion

One in four patients had AI at steroid cessation. Identifying these patients, ensuring physiological steroid replacement and patient education is critical to prevent morbidity and mortality. There were no factors predictive of inadequate SST response. Patients with AI need endocrinology referral for long-term monitoring. It is possible for patients who fail an initial SST to be subsequently successfully weaned.

Impact statement

Given the high incidence of AI in our cohort and lack of predictive factors, we recommend SSTs be considered for all patients receiving prolonged steroids for GvHD.