Enhanced predictive validity of integrative models for refractory hyperthyroidism considering baseline and early...

Area of Science:

• Endocrinology
• Internal Medicine
• Clinical Research

Background:

• Graves' disease (GD) can lead to refractory hyperthyroidism, complicating treatment strategies.
• Identifying patients at high risk for refractory GD early is essential for effective management.

Purpose of the Study:

• To develop and validate predictive models for refractory Graves' disease using baseline characteristics and early treatment indicators.
• To assess the efficacy of different antithyroid drug (ATD) dosage regimens in managing refractory GD.

Main Methods:

• A prospective cohort study of 597 adult GD patients treated with methimazole (MMI) from 2018-2022.
• Development of predictive models (Model A, B, C) incorporating baseline data and 3-month MMI therapy parameters.
• Analysis of thyroid function tests, Thyrotropin Receptor Antibody (TRAb) levels, and ATD dosage.

Main Results:

• 49.7% of analyzed patients developed refractory GD, characterized by recurrence and sustained TRAb positivity.
• Baseline factors (younger age, Graves' ophthalmopathy, goiter size, higher initial fT3/fT4/TRAb) predicted refractory GD (Model A).
• Combined models (B & C) integrating 3-month data showed superior predictive ability (AUC) and risk classification accuracy compared to Model A.

Conclusions:

• Integrating baseline and early MMI therapy data significantly enhances prediction of refractory GD outcomes.
• These predictive models offer valuable tools for refining risk assessment and guiding personalized treatment decisions in GD management.