Correlation between dysthyroidism and efficacy in patients treated with anti PD-1 or anti PDL-1 agents.

Abstract

67 Background: Immune checkpoints inhibitors have clearly improved patients’outcomes in different tumors, but definitive predictive biomarkers are still missing. Since thyroiditis have been reported, we have evaluated whether clinical responses are more frequent in patients who underwent anti PD-1 or anti PDL-1 agents and experienced dysthyroidism. Methods: We retrospectively evaluated 135 metastatic solid tumor patients, treated at our Institute with anti PD-1/PDL-1 antibodies since 2013. Thyroid toxicity was defined according to CTCAE version 4.0 and disease control (DC: CR+PR+SD) was chosen as efficacy endpoint. Correlation between dysthyroidism and DC rate was assessed by Fisher’s exact test. Results: Of 135 patients, 76 (56.3%) were treated by anti PD-1 and 59 (43.7%) by anti PDL-1 agents. Population was heterogeneous, including patients from the 1st to the 9th line of therapy, affected by the following cancers: 57 (42.2%) NSCLC, 18 (13.3%) melanoma, 13 (9.6%) RCC, 7 (5.2%) bladder and urothelial, 6 (4.5%) mesothelioma, 5 (3.7%) SCLC, 5 (3.7%) sarcoma, 5 (3.7%) biliary tract, 5 (3.7%) head and neck, 3 (2.2%) gastric, 3 (2.2%) colon, 2 (1.5%) Merkel cells and one each for thyroid, HCC, ovary, cervix, anal carcinoma and germ cells tumor. Median follow up was 8.6 months. Best responses were: 5 CR, 30 PR, 51 SD and 49 PD; 86 (63.7%) patients achieved DC. Overall, 38 patients developed dysthyroidism (subclinical, G1, G2 hypo/hyperthyroidism): 18/76 (23.7%) in anti PD-1, 20/59 (33.9%) in anti PDL-1 group; 31/38 (81.6%) of them achieved DC. The median time of dysthyroidism appearance was the 6th cycle of therapy (range 1st-22th). Of 97 patients who did not develop thyroid toxicity, 55 (56.7%) achieved DC. Dysthyroidism significantly correlated with DC rate (p=0.009). Conclusions: We found a significant correlation between dysthyroidism and clinical responses in solid tumor metastatic patients treated by anti PD-1 or anti PDL-1 antibodies. Our observation suggests that evaluation of thyroid function should be regularly performed during therapy with these agents. These retrospective findings must be confirmed in a prospective trial powered to see whether thyroid dysfunction is a surrogate marker of response.