Sr. Manager, Pharmacoepidemiology Regeneron Pharmaceuticals, Inc., Tarrytown, NY Tarrytown, United States
Background: The incidence of adrenal insufficiency (AI) in immune checkpoint inhibitor (ICI) monotherapy treatment is reported to be 0.7% for all grades and 0.2% for grades ≥3. However, data are mostly derived from clinical trials or pharmacovigilance efforts, thus AI burden among ICI-treated patients remains poorly characterized in the real-world setting.
Objectives: To estimate the incidence of adrenal insufficiency (AI) in patients with solid malignancies treated with immune checkpoint inhibitors (ICIs), including in the adjuvant setting, and to evaluate the proportion of patients with chronic AI necessitating long-term steroid treatment.
Methods: Patients were identified between October 2015 and December 2021 in Optum Claims and included those newly treated with an ICI after being diagnosed with NSCLC, melanoma, or any solid malignancy. In addition, NSCLC or melanoma patients treated with an adjuvant ICI, defined as ICI use 60 days after surgery or radiation, were also included. Patients who received small-cell lung cancer-specific treatments in the pre-index period were excluded. We estimated ninety-day incidence rates for AI and calculated the proportion of chronic AI among patients with 6 or more months of follow-up using two diagnostic codes for AI occurring at a minimum of 6 months and a maximum of 12 months apart and a medication code for oral AI treatment occurring 30 days before or after the second diagnostic code.
Results: For patients exposed to an ICI regardless of setting, the incidence rate of AI was 10.51 (95% CI: 8.48 - 12.87) per 1,000 patient-years for those with solid malignancies (N=41,415; mean age=70.7 years), 6.02 (95% CI: 4.00 - 8.70) per 1,000 person-years for those diagnosed with NSCLC (N=21,724; mean age=71.4 years), and 24.10 (95% CI: 15.59 - 35.57) per 1,000 patient-years in melanoma (N=4,670; mean age=69.5 years). In addition, among patients with AI, 25.5% of patients with solid malignancies and 33% of melanoma patients with any ICI use had chronic AI, but no patients with NSCLC in our sample had chronic AI, likely due to the small number of patients with AI who have adequate follow-up time. Among patients exposed to an adjuvant ICI, the rate of AI was 7.12 (95% CI: 3.68 - 12.44) per 1,000 person-years in NSCLC (N=7,987; mean age=70.1 years) and 18.38 (95% CI: 7.93 - 36.21) per 1,000 patient-years in melanoma (N=1,980; mean age=68.3 years). Further, 40% of melanoma patients with AI after adjuvant ICI exposure, but 0% of NSCLC patients, developed chronic AI.
Conclusions: This study demonstrated that adrenal insufficiency frequently occurs in patients with malignancies exposed to an ICI, particularly for those with melanoma. Further, ICI-induced chronic AI a significant concern in these non-NSCLC solid malignancies.
