El Kawkgi Omar, Toro-Tobon David, Toloza Freddy J K, Vallejo Sebastian, Jacome Cristian Soto, Ayala Ivan N, Vallejo Bryan A, Wenczenovicz Camila, Tzeng Olivia, Spencer Horace J, Thostenson Jeff D, Li Dingfeng, Kohlenberg Jacob, Lincango Eddy, Mohan Sneha, Castellanos-Diaz Jessica, Maraka Spyridoula, Ospina Naykky Singh, Brito Juan P
Division of Endocrinology, Diabetes, and Metabolism, Department of Medicine, Mayo Clinic, Rochester, Minnesota; Department of Endocrinology, Mayo Clinic Health System, Eau Claire, Wisconsin.
Division of Endocrinology, Diabetes, and Metabolism, Department of Medicine, Mayo Clinic, Rochester, Minnesota.
Endocr Pract. 2025 Apr;31(4):455-464. doi: 10.1016/j.eprac.2024.12.011. Epub 2024 Dec 16.
Predicting recurrence after antithyroid drug (ATD) cessation is crucial for optimal treatment decision-making in patients with Graves' disease (GD). We aimed to identify factors associated with GD recurrence and to develop a model using routine pretherapeutic clinical parameters to predict GD recurrence risk during the first year following ATD discontinuation.
This electronic health records-based observational cohort study analyzed patients with GD treated with ATDs at 3 U.S. academic centers. Demographic, clinical characteristics, and GD recurrence within 1 year following ATD discontinuation were assessed. Univariable and multivariable analyses were performed. A predictive model for GD recurrence was developed and visualized as a nomogram.
Among the 523 patients included in the study, 211 (40.34%) discontinued treatment. Of these, the 142 (67.29%) that had a follow-up period exceeding 12 months after stopping ATD were used for the development of the predictive model. Among the patients included in the model, the majority were women (n = 111, 78.16%), with a mean age of 49.29 years (standard deviation 16.31) and baseline free thyroxine (FT4) levels averaging 3.39 ng/dl (standard deviation 2.25). Additionally, 79 of 211 patients (37.44%) experienced recurrence within 1 year. Multivariable analysis indicated a 31% increased risk of GD recurrence per additional decade of age (odds ratio 1.31, 95% confidence interval 1.03-1.66, P = .0258), and a 65% increased risk of GD recurrence for every 2.0 ng/dL rise in baseline FT4 (odds ratio 1.65, 95% confidence interval 1.08-2.50, P = .0192). The recurrence predictive model's area under the curve was 0.69 in the derivation dataset and 0.65 in cross-validation.
This study introduced a practical model that can be used during the initial therapeutic decision-making process. It utilizes easily accessible baseline clinical data to predict the likelihood of GD recurrence after 1 year of ATD therapy. Further research is needed to identify other factors affecting risk of recurrence and develop more precise predictive models.
预测抗甲状腺药物（ATD）停药后的复发情况对于Graves病（GD）患者的最佳治疗决策至关重要。我们旨在确定与GD复发相关的因素，并使用常规治疗前临床参数建立一个模型，以预测ATD停药后第一年的GD复发风险。
这项基于电子健康记录的观察性队列研究分析了美国3个学术中心接受ATD治疗的GD患者。评估了人口统计学、临床特征以及ATD停药后1年内的GD复发情况。进行了单变量和多变量分析。建立了GD复发的预测模型，并将其可视化为列线图。
在纳入研究的523例患者中，211例（40.34%）停止治疗。其中，142例（67.29%）在停止ATD后随访期超过12个月，用于建立预测模型。在纳入模型的患者中，大多数为女性（n = 111，78.16%），平均年龄49.29岁（标准差16.31），基线游离甲状腺素（FT4）水平平均为3.39 ng/dl（标准差2.25）。此外，211例患者中有79例（37.44%）在1年内复发。多变量分析表明，年龄每增加十岁，GD复发风险增加31%（比值比1.31，95%置信区间1.03 - 1.66，P = 0.0258），基线FT4每升高2.0 ng/dL，GD复发风险增加65%（比值比1.65，95%置信区间1.08 - 2.50，P = 0.0192）。复发预测模型在推导数据集中的曲线下面积为0.69，在交叉验证中为0.65。
本研究引入了一个可在初始治疗决策过程中使用的实用模型。它利用易于获取的基线临床数据来预测ATD治疗1年后GD复发的可能性。需要进一步研究以确定影响复发风险的其他因素，并开发更精确的预测模型。
