Thyroid Immune-related Adverse Events Following Immune Checkpoint Inhibitor Treatment

CONTEXT: Thyroid dysfunction occurs commonly following immune checkpoint inhibition. The etiology of thyroid immune-related adverse events (irAEs) remains unclear and clinical presentation can be variable. OBJECTIVE: This study sought to define thyroid irAEs following immune checkpoint inhibitor (ICI) treatment and describe their clinical and biochemical associations. METHODS: We performed a retrospective cohort study of thyroid dysfunction in patients with melanoma undergoing cytotoxic T-lymphocyte antigen-4 (CTLA-4) and/or programmed cell death protein-1 (PD-1) based ICI treatment from November 1, 2009, to December 31, 2019. Thyroid function was measured at baseline and at regular intervals following the start of ICI treatment. Clinical and biochemical features were evaluated for associations with ICI-associated thyroid irAEs. The prevalence of thyroid autoantibodies and the effect of thyroid irAEs on survival were analyzed. RESULTS: A total of 1246 patients were included with a median follow-up of 11.3 months. Five hundred and eighteen (42%) patients developed an ICI-associated thyroid irAE. Subclinical thyrotoxicosis (n = 234) was the most common thyroid irAE, followed by overt thyrotoxicosis (n = 154), subclinical hypothyroidism (n = 61), and overt hypothyroidism (n = 39). Onset of overt thyrotoxicosis occurred a median of 5 weeks (interquartile range IQR 2-8) after receipt of a first dose of ICI. Combination immunotherapy was strongly associated with development of overt thyrotoxicosis (odds ratio OR 10.8, 95% CI 4.51-25.6 vs CTLA-4 monotherapy; P < .001), as was female sex (OR 2.02, 95% CI 1.37-2.95; P < .001) and younger age (OR 0.83 per 10 years, 95% CI 0.72-0.95; P = .007). By comparison, median onset of overt hypothyroidism was 14 weeks (IQR 8-25). The frequency of overt hypothyroidism did not differ between different ICI types. The strongest associations for hypothyroidism were higher baseline thyroid-stimulating hormone (OR 2.33 per mIU/L, 95% CI 1.61-3.33; P < .001) and female sex (OR 3.31, 95% CI 1.67-6.56; P = .01). Overt thyrotoxicosis was associated with longer progression free survival (hazard ratio HR 0.68, 95% CI 0.49-0.94; P = .02) and overall survival (HR 0.57, 95% CI 0.39-0.84; P = .005). There was no association between hypothyroidism and cancer outcomes. CONCLUSION: Thyroid irAEs are common and there are multiple distinct phenotypes. Different thyroid irAE subtypes have unique clinical and biochemical associations, suggesting potentially distinct etiologies for thyrotoxicosis and hypothyroidism arising in this context.