Clinical determinants and outcomes of hypothyroidism in immune checkpoint inhibitor (ICI) therapy.

Abstract

e14623 Background: Immune checkpoint inhibitors (ICIs) are transformative for their efficacy in cancer treatment; however, they are associated with thyroid dysfunction, including hypothyroidism. While it is known that hypothyroidism is a common immune-related adverse event, the specific risk factors, incidence rates, and long-term outcomes in diverse cancer populations remain poorly defined. This study aims to address these gaps by identifying the determinants and outcomes of ICI-induced hypothyroidism. Methods: Data were obtained from the TriNetX research network for cancer patients indicated for ICI use. Patients aged 18 years or older treated with ICIs (e.g., pembrolizumab, nivolumab, atezolizumab, durvalumab, and cemiplimab) between January 1, 2013, and December 31, 2024, were included. Exclusion criteria included prior thyroid disorders or levothyroxine use. A competing risk analysis was performed to assess the risk of hypothyroidism versus death up to 12 months after ICI use and the first-time use of levothyroxine after the diagnosis of hypothyroidism. Cox proportional hazards modeling was used to identify significant covariates associated with hypothyroidism, including age, sex, cancer type, comorbidities, and ICI type. Significant covariates were retained using backwards batchwise elimination. Results: During the study period, among 39,749 patients treated with ICI, 14.2% developed hypothyroidism within 12 months. Of these patients, 72.9% initiated levothyroxine treatment for the first time after diagnosis. The cumulative incidence of overall mortality among patients with hypothyroidism was 18.8%. Factors associated with a lower risk of developing hypothyroidism included lung cancer (hazard ratio [HR]: 0.92, p = 0.025). Conversely, an increased risk of hypothyroidism was observed in patients with endometrial cancer (HR: 1.18, p < 0.001), liver cancer (HR: 1.14, p = 0.007), and kidney cancer (HR: 1.20, p < 0.001). For ICI usage, pembrolizumab (HR: 1.7, p < 0.001), nivolumab (HR: 1.66, p < 0.001), atezolizumab (HR: 1.84, p < 0.001), durvalumab (HR: 1.80, p < 0.001), and cemiplimab (HR: 2.24, p = 0.02) were all associated with an elevated risk of hypothyroidism. Additionally, patients with adenocarcinoma NOS histology showed a slightly higher risk (HR: 1.22, p = 0.002). Conclusions: ICI-induced hypothyroidism is a significant immune-related complication, with distinct risk profiles depending on the cancer type and ICI therapy. Identifying high-risk populations based on cancer type and ICI therapy is crucial for personalized management strategies. Early identification with routine thyroid function monitoring and timely initiation of levothyroxine therapy are essential to mitigating morbidity and mortality in affected patients. Future research should focus on refining risk stratification and optimizing surveillance protocols for high-risk populations.