RESUMEN
PURPOSE: We aimed to determine the frequency of transient congenital hypothyroidism (TCH) in 17 participating centers in Türkiye, evaluate the etiological distribution in permanent congenital hypothyroidism (PCH) cases, and investigate the role of laboratory and clinical findings in predicting TCH. METHODS: This retrospective observational multicenter study included patients from 17 pediatric endocrinology centers identified by "National Newborn Screening Program" (NNSP) who were born in 2015 and followed for 6 years. Demographic, clinical, and laboratory information of the cases were compiled through the database http://cedd.saglik-network.org (CEDD-NET). RESULTS: Of the 239 cases initially treated for CH, 128 (53.6%) were determined as transient in whom a trial of levothyroxine (LT4) withdrawal was performed at a median age of 36 (34-38) months. Among the patients with PCH (n = 111), thyroid dysgenesis was diagnosed in 39.6% (n = 44). The predictive factors for TCH were: LT4 dose at the withdrawal of treatment, and initial newborn blood screening (NBS)-TSH level. Based on the receiver operating characteristic (ROC) curve analysis to predict optimal cut-offs for TCH predictors, LT4 dose < 2.0 µg/kg/day at treatment discontinuation was predictive for TCH and was associated with 94.5% specificity and 55.7% sensitivity, with an area under the curve (AUC) of 0.802. The initial NBS-TSH level value < 45 µIU/mL was predictive for TCH with 93.1% specificity and 45.5% sensitivity, with an AUC of 0.641. In patients with eutopic thyroid gland only LT4 dose < 1.1 µg/kg/day at withdrawal time was predictive for TCH with 84.7% sensitivity and 40.4% specificity, with an AUC of 0.750. CONCLUSION: According to our national follow-up data, the frequency of TCH was 53.6%. We determined the LT4 dose < 2.0 µg/kg/day at discontinuation of treatment and the initial NBS-TSH level < 45 µIU/mL as the best cut-off limits to predict TCH.