Characterizing Immune Checkpoint Inhibitor-Induced Hypophysitis and Obstacles in Diagnosis

Background: Immune checkpoint inhibitors (ICI) are used in the treatment of multiple cancers. ICI-induced endocrinopathies are known side effects of many of these agents due to their immunogenic properties. The primary aim of this single institution analysis was to characterize the population who developed hypophysitis and secondary adrenal insufficiency (AI) related to ICI therapy, and identify diagnostic challenges. Methods: We performed a retrospective cohort study of adult cancer patients who received ICI therapy from 12/1/2012- 12/31/2019. We identified 839 patients who received CTLA-4, PD-1, PDL-1 inhibitors, or a combination during this time. Hypophysitis (H) was defined as low ACTH and low serum cortisol ± other pituitary hormone deficiencies. Results: Sixteen patients (1.9%; 16/839) developed hypophysitis after a median of 7 months from ICI initiation. The underlying cancers included: uroepithelial (1/48; 2.1 %), neuroendocrine (1/10; 10%), melanoma (9/156; 5.8%), renal cell carcinoma (4/74; 5.4%), and non-small cell lung carcinoma (1/247; 0.4%). Four patients were excluded since they had isolated secondary AI due to exogenous steroids. Two patients with hypophysitis also had exogenous steroid usage. However, these patients had central hypothyroidism + secondary AI so were included. The median age at start of ICI was 61.3 years and 57% were males. Patients who developed hypophysitis were younger compared to the non-hypophysitis (NH) patients (median age: 57 years vs. 65 years; p=0.0113). Time from initiation of ICI to death/last follow-up was longer in the H group (20 months vs. 10.8 months; p=.0013). 87.5% of group H as compared to 47.3% of NH were alive at end of data collection (p=0.0008). In the H cohort, 43.8% (7/16) received combination CTLA-4/PD-1 inhibitors, 43.8% (7/16) PD-1 monotherapy, 6.25% (1/16) CTLA-4 monotherapy, and 6.25% (1/16) PD-L1 monotherapy. Forty-six percent (6/13) of group H had radiographic evidence of hypophysitis on MRI that resolved on follow up imaging. Six patients had concurrent thyroiditis, 2/16 had concurrent type 1 diabetes, 9/16 had central hypothyroidism, 1/16 had secondary hypogonadism, and 2/16 had GH deficiency. In 15/16 patients, secondary AI presented as the first endocrinopathy, while one patient presented with central hypothyroidism. Conclusions: Development of hypophysitis following ICI therapy was associated with a longer survival. Deficiencies in multiple pituitary hormones occurred in the majority and non-pituitary endocrinopathies occurred in half the cases. High dose steroid usage can also present as secondary AI, making the diagnosis of ICI induced hypophysitis difficult. Further pituitary evaluation must be conducted to differentiate secondary AI due to exogenous steroids from hypophysitis from ICI therapy.