Abstract
BACKGROUND: Nivolumab, an anti-programmed cell death-1 monoclonal antibody, has improved the survival of patients with malignant melanoma. Despite its efficacy, nivolumab inconsistently induces thyroid dysfunction as an immune-related adverse event (irAE). This study aimed to evaluate nivolumab-induced thyroid dysfunction to determine the risks and mechanisms of thyroid irAEs. METHODS: After excluding 10 patients, data of 24 patients with malignant melanoma (aged 17-85 years; 54% female) were retrospectively analyzed. RESULTS: Thyroid irAEs were observed in seven patients (29%). Three patients had hypothyroidism after preceding transient thyrotoxicosis, and the other four patients had hypothyroidism without thyrotoxicosis. Levothyroxine-Na replacement was required in three patients. Antithyroid antibody (ATA) titer was elevated in one of four assessable patients. The average (±SD) time to onset of thyroid irAE was 33.6 (±21.9) weeks. The administration period of nivolumab was longer in patients with thyroid irAEs than in those without thyroid irAEs (P < 0.01). There were no significant differences between patients with and without thyroid irAEs regarding age, sex, tumor stage, response to nivolumab therapy, baseline thyroid function, antithyroid peroxidase antibody (anti-TPO Ab) and antithyroglobulin antibody (anti-Tg Ab). CONCLUSIONS: Thyroid dysfunction was a common irAE of nivolumab in malignant melanoma. Neither anti-TPO Ab nor anti-Tg Ab was associated with the risk for nivolumab-induced thyroid dysfunction. A conventional ATA-independent mechanism might be involved in thyroid irAEs. Further studies are required to clarify the mechanism and identify the predictive factors of thyroid irAEs.