Response to initial therapy of differentiated thyroid cancer predicts the long-term outcome better than classical risk stratification systems.

Objective. Although differentiated thyroid cancer (DTC) usually has an indolent course, some cases show a poor prognosis; therefore, risk stratification is required. The objective of this study is to compare the predictive ability of classical risk stratification systems proposed by the European Thyroid Association (ETA) and American Thyroid Association (ATA) with the system proposed by Tuttle et al. in 2010, based on the response to initial therapy (RIT). Methods. We retrospectively reviewed 176 cases of DTC with a median follow-up period of 7.0 years. Each patient was stratified using ETA, ATA, and RIT systems. Negative predictive value (NPV) and positive predictive value (PPV) were determined. The area under receiver operating characteristic (ROC) curve was calculated in order to compare the predictive ability. Results. RIT showed a NPV of 97.7%, better than NPV of ETA and ATA systems (93.9% and 94.9%, resp.). ETA and ATA systems showed poor PPV (40.3% and 41%, resp.), while RIT showed a PPV of 70.8%. The area under ROC curve was 0.7535 for ETA, 0.7876 for ATA, and 0.9112 for RIT, showing statistical significant differences (P < 0.05). Conclusions. RIT predicts the long-term outcome of DTC better than ETA/ATA systems, becoming a useful system to adapt management strategies.