OR21-07 The 2015 American Thyroid Association Risk Stratification System Is a Predictor of Persistent Disease in Real-World Clinical Practice

Background. Management and follow-up of differentiated thyroid cancer (DTC) are guided by the likelihood of disease persistence or recurrence. The American Thyroid Association (ATA) practice guidelines provide a risk-estimation system based on data mainly derived by retrospective, single-center, and small cohorts. Aim. To validate the ATA risk-stratification system in predicting persistent structural disease. Methods. We analyzed data from the Italian Thyroid Cancer Observatory’s observational, web-based database, which prospectively enrolls newly diagnosed DTC patients in 40 Italian centers. For the present study we selected consecutive cases satisfying the inclusion criteria: 1) histological diagnosis of DTC, including papillary, follicular, and poorly differentiated tumors; 2) registration in the ITCO database between January 1, 2013 and April 23, 2019; 3) clinical evaluation between 6 and 18 month after primary treatment, including enough data to estimate the response to the initial treatment. Exclusion criteria were: histological diagnosis of NIFTP, medullary, or anaplastic thyroid cancer. The response to the initial treatment was categorized as excellent, biochemical incomplete, structural incomplete, or indeterminate based on imaging findings (neck ultrasound and other imaging studies, if performed), basal or stimulated serum thyroglobulin levels, and anti-Tg antibody levels. To model the response to treatment, we used a cumulative link model; given the hierarchical structure of the data, with patients nested within centers, we used a mixed-effect model, with a center-specific intercept summarizing unobserved center-specific characteristics. Results. Complete data about initial treatment and response to treatment after 6-18 months since initial treatment was available for 2071 patients. According to the ATA system, 1109 patients (53.6%) were classified as low-risk, 796 (38.4%) as intermediate, and 166 (8.0%) as high-risk. Excellent response was recorded in 1576 (76.1%) patients, indeterminate in 376 (18.2%), biochemical incomplete in 33 (1.6%), and structural incomplete in 86 (4.2%).The ATA risk stratification system is a significant predictor of response to treatment after 6-18 months: classification as intermediate- and high-risk increased the likelihood of a response worse than excellent (OR 1.68 [95% confidence intervals, CI 1.34-2.10] and 3.23 [95% CI 2.23-4.67], respectively), and a persistent structural disease (OR 4.67 [95% CI 2.59-8.43] and 16.48 [95% CI 7.87-34.5], respectively. In both analyses, the effect of the center (taking into account center-specific features) was negligible and not statistically significant. Conclusion. The 2015 ATA risk stratification system is a reliable predictor of short-term outcomes in patients with DTC, also if applied in a real-world setting consisting of several different clinical sites.