Frontiers in endocrinology · 2026
Factors associated with levothyroxine withdrawal and development of a prediction model in primary hypothyroidism.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Zheng Wanli, Tang Yibei, Chen Xuanyu, Ye Wang, Yu Jingyao, Sun Yihan, Zhang Bin, Wu Guangli, Zheng Jipeng, Wang Yuqing, Cui Jianchun, Lu Li, Ju Xingai
Аннотация
OBJECTIVE: To identify factors associated with successful levothyroxine (L-T4) withdrawal in patients with primary hypothyroidism and to develop a clinical prediction model.
METHODS: We retrospectively included patients with primary hypothyroidism who received L-T4 treatment at our hospital between January 2016 and October 2024 and had a documented withdrawal attempt with follow-up records. Baseline patient data were collected, including sociodemographic characteristics (age, sex, body mass index [BMI], family history of thyroid disease), medication-related indices (L-T4 dosage and treatment duration), clinical classification (subclinical vs. overt hypothyroidism), and laboratory and ultrasound imaging parameters (TSH, FT3, FT4, TPOAb, TgAb, heterogeneous echotexture, and hypoechogenicity). Successful withdrawal was defined as maintenance of euthyroidism for 12 months after stopping L-T4; otherwise, withdrawal was considered unsuccessful. Candidate predictors were screened using univariate logistic regression and entered into multivariable logistic regression to identify independent predictors and construct a nomogram. Discriminative ability was assessed using the area under the receiver operating characteristic curve (AUC). Calibration and internal validation were evaluated using the Hosmer-Lemeshow test and bootstrap resampling (1,000 iterations).
RESULTS: A total of 110 patients were included, of whom 39 (35.5%) achieved successful withdrawal and 71 (64.5%) did not. In multivariable analysis, younger age (OR = 0.901, 95% CI: 0.852-0.953, P < 0.001) and subclinical hypothyroidism (OR = 4.879, 95% CI: 1.343-17.723, P = 0.016) were associated with a higher likelihood of successful withdrawal. In contrast, TPOAb positivity (OR = 0.150, 95% CI: 0.040-0.556, P = 0.005) and heterogeneous thyroid echotexture on ultrasound (OR = 0.155, 95% CI: 0.032-0.751, P = 0.021) were associated with a lower likelihood of success. The model showed excellent discrimination (AUC = 0.899, 95% CI: 0.840-0.957, P < 0.001). At the optimal cutoff value of 0.292, sensitivity was 87.2% and specificity was 80.3%. Model fit was good (Hosmer-Lemeshow χ² = 5.448, P = 0.709), and bootstrap validation demonstrated good calibration (mean absolute error = 0.029).
CONCLUSION: Younger age, subclinical hypothyroidism, negative TPOAb, and homogeneous thyroid echotexture were associated with a higher likelihood of successful L-T4 withdrawal. The nomogram model may facilitate pre-withdrawal risk stratification and individualized follow-up planning, supporting quantitative decision-making in clinical practice.
METHODS: We retrospectively included patients with primary hypothyroidism who received L-T4 treatment at our hospital between January 2016 and October 2024 and had a documented withdrawal attempt with follow-up records. Baseline patient data were collected, including sociodemographic characteristics (age, sex, body mass index [BMI], family history of thyroid disease), medication-related indices (L-T4 dosage and treatment duration), clinical classification (subclinical vs. overt hypothyroidism), and laboratory and ultrasound imaging parameters (TSH, FT3, FT4, TPOAb, TgAb, heterogeneous echotexture, and hypoechogenicity). Successful withdrawal was defined as maintenance of euthyroidism for 12 months after stopping L-T4; otherwise, withdrawal was considered unsuccessful. Candidate predictors were screened using univariate logistic regression and entered into multivariable logistic regression to identify independent predictors and construct a nomogram. Discriminative ability was assessed using the area under the receiver operating characteristic curve (AUC). Calibration and internal validation were evaluated using the Hosmer-Lemeshow test and bootstrap resampling (1,000 iterations).
RESULTS: A total of 110 patients were included, of whom 39 (35.5%) achieved successful withdrawal and 71 (64.5%) did not. In multivariable analysis, younger age (OR = 0.901, 95% CI: 0.852-0.953, P < 0.001) and subclinical hypothyroidism (OR = 4.879, 95% CI: 1.343-17.723, P = 0.016) were associated with a higher likelihood of successful withdrawal. In contrast, TPOAb positivity (OR = 0.150, 95% CI: 0.040-0.556, P = 0.005) and heterogeneous thyroid echotexture on ultrasound (OR = 0.155, 95% CI: 0.032-0.751, P = 0.021) were associated with a lower likelihood of success. The model showed excellent discrimination (AUC = 0.899, 95% CI: 0.840-0.957, P < 0.001). At the optimal cutoff value of 0.292, sensitivity was 87.2% and specificity was 80.3%. Model fit was good (Hosmer-Lemeshow χ² = 5.448, P = 0.709), and bootstrap validation demonstrated good calibration (mean absolute error = 0.029).
CONCLUSION: Younger age, subclinical hypothyroidism, negative TPOAb, and homogeneous thyroid echotexture were associated with a higher likelihood of successful L-T4 withdrawal. The nomogram model may facilitate pre-withdrawal risk stratification and individualized follow-up planning, supporting quantitative decision-making in clinical practice.