Abstract

Background

Endocrine dysfunction is a common toxicity of immune-checkpoint inhibitors. Unlike other immune-related-adverse events, endocrinopathies present with non-specific symptoms of insidious onset that are often irreversible and permanent. A plan-do-study-act (PDSA) methodology was employed to identify and improve shortcomings in the detection and monitoring of immunotherapy-induced endocrinopathies. 

Method

A preliminary survey was carried out on all patients treated with immunotherapy for metastatic melanoma between 2016 and 2018 (n=76). Hormone profiling was used to identify the incidence of hypocortisolism and the proportion of those that underwent appropriate follow-up with brain imaging and endocrine review. Subsequent steps were taken to improve patient and staff awareness and promote early recognition with prompt referral for patients with hormonal disturbances. Following this intervention, repeat data was collected one year onwards with patients on immunotherapy between 2018 and 2019 (n=45).

Results

The initial study found that 6.6% of immunotherapy patients (n=5) developed adrenal insufficiency (defined as serum cortisol<50nmol/L). 60% of these patients were referred to endocrine physicians, whilst 80% underwent pituitary MRIs. Repeat study showed a 16% incidence of hypocortisolism (n=7). Only one patient (14%) was referred for a specialist endocrine opinion, whilst 57% underwent pituitary imaging with MRI. 

Conclusion

Data from the past year suggests that further work is necessary to improve follow-up of patients with immunotherapy-induced hypocortisolism. Meetings between medical oncologists and endocrinologists are currently underway to help develop shared management protocols and promote integration of services. Application of current Trust guidelines on managing endocrinopathies may eliminate the need for endocrine input in straightforward cases. It remains vitally important, however, that patients are educated regarding ‘sick-day’ rules and are provided with ‘back-up’ parenteral steroids in the event of an acute illness. Future studies should ascertain whether hypocortisolism is centrally-driven in all cases, and whether it is associated with central hypothyroidism and hypogonadism.