Positive baseline thyroid autoantibodies as a predictive biomarker for thyroid immune-related adverse events: A systematic review and meta-analysis.

Abstract

e24124 Background: Thyroid dysfunction is one of the most common immune-related adverse events (irAE) by immune checkpoint inhibitors (ICI). Given the known relationship between antithyroid autoantibodies (ATA), including antithyroid peroxidase (TPO Ab) and anti-thyroglobulin (Tg Ab), and autoimmune thyroid disease, they could also play a role as a predictive biomarker for thyroid irAE (T-irAE). We performed this meta-analysis to evaluate the association of positive baseline thyroid autoantibodies and T-irAE. Methods: Eligible studies were screened using Ovid Medline and Embase databases with search terms related to ATA and thyroid irAE. Studies must be case-control or cohort studies with patients who received ICI and report the numbers of patients who have thyroid dysfunction and do not have thyroid dysfunction by the pre-ICI treatment statuses of ATA, TPO Ab, or Tg Ab. The pooled odds ratio (OR) and the associated 95% confidence interval (95% CI) were calculated using the Mantel-Haenszel method. Subgroup analysis by types of ICI (anti-PD1, anti-PD-L1, or combination treatment) and study region were also performed. The presence of publication bias was assessed by direct visualization of funnel plots. Results: After two rounds of screening, 19 case-control and five retrospective cohort studies were included in the meta-analysis. Patients who had positive baseline ATA had increased odds of developing T-irAE, hypothyroidism, and hyperthyroidism (pooled OR 10.95, 95% CI 6.65-18.02, I2= 38%; 6.82, 95% CI 2.99-15.56, I2= 27%; and 7.59, 95% CI 2.51-22.96, I2= 42%; respectively). Positive pre-treatment TPO Ab and Tg Ab were also independently associated with higher odds of hypothyroidism (pooled OR 5.01, 95% CI 2.87-8.75, I2= 4%, and 18.73, 95% CI 9.04-38.80, I2= 0%, respectively) and hyperthyroidism (pooled OR 2.94, 95% CI 1.57-5.52, I2= 25%, and 8.36, 95% CI 1.18-59.31, I2= 82%, respectively). The associations did not significantly differ between the types of ICI utilized and the study region. Visualization of funnel plots did not reveal a presence of publication bias. Conclusions: Pre-ICI treatment statuses of ATA, TPO Ab, and Tg Ab are associated with both hypothyroidism and hyperthyroidism. Thyroid autoantibodies as a baseline predictive tool can be helpful for early detection of T-irAE in patients who receive ICI therapy.