We sought to characterize thyroid dysfunction and its association with baseline clinical and demographic characteristics, as well as progression-free survival (PFS), in a multiethnic cohort of lung cancer patients treated with ICIs. A retrospective chart review of lung cancer patients receiving an anti-PD1 or PD-L1 agent was performed. Multivariate Cox proportional hazards were fitted to compare time to thyroid dysfunction among race subgroups controlling for age, gender, treatment type, and duration. Thyroid dysfunction was based on laboratory testing; clinical symptoms were not required. PFS at a 24-week landmark analysis point among patients with and without thyroid dysfunction was compared using a log-rank test. We identified 205 subjects that received ICIs, including 76 (37.1%) who developed thyroid dysfunction. Rates of thyroid dysfunction by one year occurred at similar frequencies among all races (p = 0.92). Gender and concurrent chemotherapy showed no significant association with thyroid dysfunction (p = 0.81 and p = 0.67, respectively). Thyrotoxicosis occurred at higher rates in Black (25, 31.6%) subjects than in White (7, 16.7%) and Hispanic (8, 12.7%) subjects when employing the log-rank test (p = 0.016) and multivariate Cox regression (HR 0.48, p = 0.09 for White and HR 0.36, p = 0.01 for Hispanic compared to Black subjects). PFS was similar among subjects with and without thyroid dysfunction when applying the log-rank test (p = 0.353). Gender, concurrent treatment with chemotherapy, and PFS were not associated with thyroid dysfunction in patients receiving ICIs; however, Black race was a risk factor for thyrotoxicosis. The mechanisms underlying the role of race in the development of irAEs warrant further study.
Keywords: NSCLC; SCLC; checkpoint inhibitors; immune-related adverse events; immunotherapy; thyroid.