ESPE Abstracts

Background: Delayed TSH elevation (dTSH) is defined by normal TSH on the initial neonatal screening followed by elevated TSH on the second screen. Several studies concluded that dTSH is associated with low birth weight (BW) and is mostly transient.

Objective and hypotheses: To elucidate clinical characteristics of dTSH in a large cohort of neonatal intensive care unit (NICU) newborns.

Method: Clinical data were gathered from a cohort of 13 201 NICU newborns born between 1^st January 2008 and 31^st October 2014 that underwent TSH measurements due to low T₄ levels on the second screen. The clinical data included gestational age (GA), BW, timing of second test, T₄ levels and short-term follow-up.

Results: 333 out of 13 201 (1:40) newborns presented with dTSH (TSH>15 IU/l), 129 (39%) of them had TSH levels >40 IU/l. dTSH had a peak incidence at GA of 37–39 weeks and 66% of the patients had BW>1500 g. There was no optimal timing to detect the disorder, as test timing was equally distributed after the third week of life. T₄ values in the 333 patients were negatively correlated with TSH levels (r=−0.505; P<0.0001) and were significantly lower than T₄ levels in 12 868 newborns with normal TSH: 5.9±2.8 vs 7.6±1.7 μg/dl; P<0.001. TSH levels were already elevated among patients with dTSH on the initial screening compared with the other newborns (8.3±5.2 vs 4.2±3.7 IU/l; P<0.001). A short-term follow-up in 193 of the dTSH patients demonstrated a persistence of TSH elevation and a need for levothyroxine treatment in 58% of them. Thyroid scans obtained in twelve patients in this subgroup were normal.

Conclusion: Unlike previous reports, we found that dTSH is most prevalent in full-term newborns with BW>1500 g. Low T₄ is the best predictor for this disorder and apparently TSH elevation is already initiated within the 1st days of life.