Predicting variables associated with diagnostic reevaluation of transient congenital hypothyroidism.

Purpose: Current guidelines recommend immediate treatment after diagnosis of congenital hypothyroidism and reassessment of the hypothalamic-pituitary-thyroid axis at 3 years of age. As the known incidence of transient congenital hypothyroidism (TCH) has increased, experts have suggested the possibility of early drug discontinuation. Distinguishing TCH from permanent congenital hypothyroidism (PCH) is important to avoid prolonged treatment. We aimed to investigate the factors associated with TCH and to identify markers that indicate patients suitable for early treatment discontinuation.

Methods: Participants were 167 children with congenital hypothyroidism. Subjects attempting to discontinue levothyroxine before 2 years of age were defined as the "early-off group." Cox proportional hazards models were used to identify factors associated with TCH and to determine factors predicting early drug discontinuation.

Results: Totals of 96 (57%) and 71 children (43%) were classified as having TCH and PCH, respectively. In the Cox multivariate analysis, gestational age (GA) and low levothyroxine dose at 24 months of age were statistically associated with TCH. Based on receiver operating characteristic (ROC) curve analysis, an optimal cutoff dose for levothyroxine of 3.03 µg/kg/day at 18 months of age can predict early treatment discontinuation (P<0.001; sensitivity, 75.0%; specificity, 72.9%; area under the curve, 0.778).

Conclusion: Our study showed that lower GA and lower levothyroxine doses during treatment were highly suggestive of TCH. Those requiring lower levothyroxine levels at 18 months of age could be candidates to cease medication prior to 3 years of age.