Immune-related adrenal insufficiency: A single center retrospective analysis.

Abstract

2659 Background: Immune checkpoint inhibitors (ICI) became widely used for treatment of various malignancies. They may cause different immune-related adverse events (irAEs), including adrenal insufficiency (irAI). IrAI was reported to be comparatively uncommon but potentially life threatening. We conducted a retrospective single center study to identify and describe the incidence and characteristics of AI associated with ICI. Methods: We reviewed the medical records of patients treated at an oncological department at Loginov Moscow Clinical Scientific Center between Sep’2019 and Jan’2023 with immunotherapy under direct cortisol monitoring every 6-8 weeks. The main criteria for AI was low serum cortisol in the absence of glucocorticoid intake. Results: A total 101 patients (pts) treated by ICI or ICI in combination with chemotherapy or targeted therapy for various malignancies were included in our retrospective analysis. 16 patients (15.8%) were diagnosed with AI: 7 pts with renal cell carcinoma (RCC), 6 pts with melanoma (2 in adjuvant setting), 1 with non-small-cell lung cancer (NSCLC), 1 with small-cell lung cancer (SCLC), 1 with MSI-high colorectal cancer (CRC). Median age was 65 (range 46-79), the majority (10 pts, 62.5%) were male. The median time of irAI onset from ICI start was 5 months (range 2-15). Assessment of serum cortisol was performed every 6-8 weeks routinely and off-schedule in patients who had new onset AI-like symptoms. The most frequent symptoms were unspecific and included fatigue and weakness (100%), loss of appetite/anorexia (69%), hypotension (50%), weight loss (44%), insomnia (38%), chills (31%), nausea/vomiting (31%), muscle pain or cramps (31% and 25% respectively), dizziness (25%), dry mouth (25%), headache (13%), fever (6%). Remarkably, laboratory detected serum cortisol decline preceded any AI symptoms in 7 patients (43.7%). Other 7 patients (43.7%) had concurrent secondary hypothyroidism which suggested panhypopituitarism. Hydrocortisone replacement led to rapid symptoms recovery in 15/16 patients. Additional treatment with fludrocortisone was required for only one patient who had primary/mixed AI due to bilateral metastases of RCC in adrenal glands. All patients resumed immunotherapy after symptoms resolved. Only 1 patient needed treatment modification (ipilimumab cessation). By 12th February 2023 median follow-up from irAI onset is 7 months (range 0.5-37), 12 patients are alive, they all remain on permanent hydrocortisone replacement therapy. Conclusions: Routine cortisol assessment during immunotherapy allowed us to detect irAI in 15.8% of patents that is higher than usually reported. Detection of irAI before clinical presentation let us prevent severe consequences including life threating. We consider that routine cortisol assessment during immunotherapy should be implemented in clinical practice.