Endocrine Abstracts

JOINT852

Background: Immune checkpoint inhibitors (ICI) have transformed the prognosis of many solid malignancies. However, their use has also been associated with distinct endocrine immune-related adverse events (irAEs), hypophysitis being among the most common.

Purpose: To describe, in a real-world patient cohort, how ir-hypophysitis is depicted on a pituitary magnetic resonance imaging (MRI) and how these imaging findings are evolved overtime.

Methods: A retrospective analysis of pituitary MRI was conducted in cancer patients treated with ICI-based regimens who developed biochemically documented pituitary insufficiency from January 2016 to September 2024. The first MRI was performed at the time of diagnosis of ir-hypophysitis, and the second MRI during the follow-up. MRIs were evaluated by a central radiologist team blinded to the onset of ir-hypophysitis.

Results: Sixty-six ICI-treated cancer patients diagnosed with biochemically documented anterior pituitary deficiency were eligible for inclusion in our analysis. The majority of them received immunotherapy (70% anti-PD-1 or anti-PDL-1 ICI, and 19% anti-PD-1/anti-PD-L1 and anti-CTLA-4 combinations) for melanoma (90.9%), whereas 4.5% of cases for lung cancer, 1.5% colon cancer and 3% hepatocellular carcinoma. The initial pituitary MRI was available in 60 patients and performed at a median time of 2 weeks post-diagnosis of ir-hypophysitis. Abnormalities were found in 32 patients (53.3%), including enlargement (25%) or reduced enhancement of the pituitary gland (10%), empty sella turcica (8.3% ), and less commonly heterogeneous enhancement (5%), reduced dimensions of the pituitary gland (3.3%) and slight deviation of the stalk (1.7%). A 2^nd pituitary MRI assessment, after a median follow-up of 1.6 years, was available in 37 patients; 45% of them presented alteration of their initial abnormal MRI findings. Abnormalities were described in 62.2% of cases, including reduced dimensions (18.9%), enlargement of the pituitary gland (16.2%), a partially empty sella turcica (16.2%), heterogeneous enhancement of the pituitary gland (8.1%) and reduced enhancement (2.7%). No ICI-based regimen nor combination treatments were associated with a specific abnormal pituitary imaging or with the frequency of abnormal imaging findings at follow-up. Patients with multiple axes deficiencies presented an increased prevalence of MRI abnormalities compared to those with isolated corticotrope deficiency in both time assessments.

Conclusion: MRI pituitary abnormalities in patients with ir-hypophysitis were found in approximately half of them; were not specific to the underlying malignancy and the administered ICI; and persisted over time transforming their abnormal MRI imaging. The other half presented normal pituitary MRI in both assessments, keeping the challenge of imaging for this irAE.