Clinical presentation and significance of thyroid dysfunctions secondary to PD-1/PD-L1 blockade cancer immunotherapy

Background: Thyroid dysfunctions are among the most common immune-related adverse events (irAEs) following the administration of immune checkpoint inhibitors (ICIs) for malignancies. The pathogenesis of thyroid irAEs remains unclear and the clinical course can be multifaceted. Aim: Aims of this study were a) to describe the incidence and the clinical course of thyroid irAEs b) to determine the association between thyroid irAEs and overall survival (OS). Methods: We performed a single-center retrospective study of cancer patients treated with anti-PD-1/PD-L1 from January 2018 to December 2020. Demographic data, thyroid function tests (serum thyrotropin, TSH; free thyroxine FT4; free triiodothyronine, FT3) and thyroid ultrasonographic findings (if available) were retrieved at baseline and at regular intervals after starting ICIs. Patients were excluded if a) they had abnormal thyroid function at baseline b) were on anti-thyroid drugs or levothyroxine (LT4) replacement c) had missing data. Results: One hundred sixty-six cancer patients were considered for potential enrollment, and after assessment of inclusion and exclusion criteria, 112 were enrolled. The mean age was 67.9 (10.6) years, and 82 patients (73.2%) were males. Lung cancer accounted for 65.8% of all cancers, followed by melanoma (16.2%), squamous cell carcinoma (9.9%), genitourinary cancers (6.3%) and other cancers (1.8%). Among all patients, 97 (86.6%) were treated with a PD-1 blockade, whereas 15 (14.4 %) received a PD-L1 inhibitor. Previous treatments had been performed in 72 patients (79.1%). During the study period, 25 patients (22.3%) developed thyroid irAEs with a median time to onset of 5.1 months (iqr 6.7). Two of them (8%) initially presented with thyrotoxicosis and 23 (92%) with hypothyroidism. Patients with thyrotoxicosis had an earlier median time to onset compared to those who had hypothyroidism (1.3 vs 5.2 months, P=0.045). Overall, 19 patients (76%) required LT4 replacement. Systemic steroids were not required in all cases. Thyroid ultrasonography, performed in 19 patients at thyroid irAEs onset, revealed a slightly increased thyroid volume in patients with thyrotoxicosis and a reduced volume in hypothyroid patients (18.1 vs 8.4 mL, P=0.01). Multivariable Cox regression analysis revealed that the occurrence of thyroid irAEs was independently associated with better OS (H.R. 0.3, CI 95% 0.1- 0.7, P=0.006). Conclusion: This study confirms that thyroid irAEs occur with a high frequency in routine clinical practice and with heterogenous clinical presentation. It also supports that thyroid irAEs may represent a predictive biomarker of better response to ICIs.