Data-Driven Thyroglobulin Cutoffs for Low- and Intermediate-Risk Thyroid Cancer Follow-Up: ITCO Real-World Analysis
Giorgio Grani,Silvia D'Elia,Efisio Puxeddu,Silvia Morelli,Emanuela Arvat,Alice Nervo,Giovanna Spiazzi,Nicoletta Rolli,Maria Chiara Zatelli,Maria Rosaria Ambrosio,Graziano Ceresini,Michela Marina,Chiara Mele,Gianluca Aimaretti,Maria Giulia Santaguida,Camilla Virili,Anna Crescenzi,Andrea Palermo,Ruth Rossetto Giaccherino,Letizia Meomartino,Maria Grazia Castagna,Fabio Maino,Matteo Trevisan,Simone De Leo,Maria Grazia Chiofalo,Luciano Pezzullo,Clotilde Sparano,Luisa Petrone,Giulia Di Dalmazi,Giorgio Napolitano,Dario Tumino,Umberto Crocetti,Francesco Bertagna,Maurilio Deandrea,Alessandro Antonelli,Caterina Mian,Antonella Carbone,Salvatore Monti,Tommaso Porcelli,Giulia Brigante,Daniele Barbaro,Marco Alfò,Umberto Ferraro Petrillo,Sebastiano Filetti,Cosimo Durante
DOI: https://doi.org/10.1210/clinem/dgae559
2024-08-16
Abstract:Context: The utility of thyroglobulin (Tg) in the follow-up of differentiated thyroid cancer (DTC) patients has been well-documented. Although third-generation immunoassays have improved accuracy, limitations persist (interfering anti-Tg antibodies and measurement variability). Evolving treatment strategies require a reevaluation of Tg thresholds for optimal patient management. Objective: To assess the performance of serum Tg testing in two populations: patients receiving total thyroidectomy and radioiodine remnant ablation (RRA), or treated with thyroidectomy alone. Design: Prospective observational study. Setting. Centers contributing to the Italian Thyroid Cancer Observatory (ITCO) database. Patients: We included 540 patients with 5 years of follow-up and negative anti-Tg antibodies. Interventions: Serum Tg levels assessed at 1-year follow-up visit. Main outcome measure: Detection of structural disease within 5 years of follow-up. Results: After excluding 26 patients with structural disease detected at any time point, the median Tg did not differ between patients treated with or without radioiodine. Data-driven Tg thresholds were established based on the 97th percentile of Tg levels in disease-free individuals: 1.97 ng/mL for patients undergoing thyroidectomy alone (lower than proposed by the MSKCC protocol and ESMO Guidelines, yet demonstrating good predictive ability, with a negative predictive value (NPV) of 98%) and 0.84 ng/mL for patients receiving post-surgical RRA. High sensitivity and NPV supported the potential of these thresholds in excluding structural disease. Conclusions: This real-world study provides evidence for the continued reliability of 1-year serum Tg levels. The data-driven Tg thresholds proposed offer valuable insights for clinical decision-making in patients undergoing total thyroidectomy with or without RRA.