Long-term risk of incident thyroid cancer after COVID-19 in patients with thyroid disorder: a multicenter propensity score–matched cohort study

Abstract

Background Prior studies examining the relationship between coronavirus disease 2019 (COVID-19) and thyroid cancer in general populations may be influenced by surveillance bias, complicating interpretation of their findings. To address this limitation, we evaluated the association between severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and incident thyroid cancer specifically among adults with pre-existing thyroid disease. Methods We used the TriNetX Global Collaborative Network to conduct a propensity score–matched cohort study of adults aged ≥18 years with thyroid disorders between 2020 and 2022. Patients with documented COVID-19 were compared with matched controls without recorded SARS-CoV-2 infection. A 24-month landmark design excluded cancers diagnosed during early post-infectious evaluation. The primary outcome was incident thyroid cancer, estimated using Cox proportional hazards models, and all-cause mortality was assessed as a secondary outcome. Follicular cysts of skin served as a negative-control outcome. E-values were calculated to assess the robustness of the observed association to potential unmeasured confounding. Sensitivity, subgroup, alternative landmark, hospitalization-based severity-stratified, competing-risk, and secondary generalizability analyses were performed. Results After matching, 603,364 patients were included in each cohort. COVID-19 was associated with a higher risk of incident thyroid cancer (hazard ratio [HR], 1.61; 95% confidence interval [CI], 1.49–1.75; E-value, 2.60). The association persisted across sensitivity analyses (HRs, 1.70–1.95), sex and age strata, baseline thyroid-disease subtypes, hospitalization status, and 1- to 3-year landmark analyses. Similar findings were observed in a secondary cohort of adults with type 2 diabetes and no prior thyroid disease (HR, 1.78; 95% CI, 1.53–2.07). All-cause mortality was higher in the COVID-19 cohort compared with controls (HR, 1.70; 95% CI, 1.66–1.73). The negative-control outcome showed no increase in cumulative incidence (risk ratio, 0.95) despite a modestly higher hazard (HR, 1.16; 95% CI, 1.13–1.19). Conclusion Among adults with pre-existing thyroid disease, COVID-19 was associated with a delayed increase in incident thyroid cancer. This association was not readily explained by measured healthcare utilization alone, but residual surveillance bias cannot be excluded. These findings represent an epidemiologic signal rather than proof of causation and do not support routine thyroid cancer screening based solely on prior COVID-19.