Overt hypothyroidism, diagnosed and treated with L-thyroxine pre-pregnancy, affects 1% of pregnancies. Untreated maternal hypothyroidism is associated with adverse obstetric and neonatal outcomes, and even when partially treated the risks of miscarriage and prematurity remain elevated compared to euthyroidism. With increased understanding of the normal physiological changes to thyroid function during pregnancy, there should be a predictive approach to thyroxine dose adjustments in order to prevent the development of abnormalities in thyroid function, which could impact upon fetoplacental development. The vast majority of women will require a 3050% dose increase by 58 weeks of gestation. Gestation specific reference ranges should be used, and these vary between assay platforms and patient populations. Since subclinical hyperthyroidism is not associated with any adverse obstetric outcomes, and given that even within the normal reference range there are correlations between serum TSH and free T4 concentrations with perinatal loss and fetal malposition, the aim of treatment should be to maintain TSH within the lower part and free T4 within the upper part of the respective normal reference ranges.
The roles of universal screening for and treatment of previously undiagnosed subclinical hypothyroidism (defined as elevated serum TSH accompanied by normal free T4; 2.5% of pregnancies) and isolated hypothyroxinaemia (normal serum TSH accompanied by low free T4; 2% of pregnancies) in pregnancy remain controversial. Although these biochemical abnormalities have been associated with increased obstetric risks and impaired offspring neurodevelopment in some studies, recently completed intervention trials of L-thyroxine treatment have not yielded conclusive results. The Wales controlled antenatal thyroid screening (CATS) study which assessed the impact of antenatal screening and L-thyroxine treatment on offspring neurodevelopment found no significant differences in mean IQ scores at aged 3 years between experimental and control groups. Another study conducted in Italy comparing universal screening with targetted case finding in early pregnancy found no differences in obstetric outcome between groups. However, in both these studies, sub-analyses suggest that L-thyroxine may still be beneficial in some cases. It is possible that screening and treatment has been commenced too late to make a full impact on the pregnancy.
A further 1015% of pregnant women have thyroid autoimmunity (anti-thyroid peroxidase (TPO) positivity). Despite the majority being biochemically euthyroid, they are also at increased risk of miscarriage and preterm delivery. Two small clinical trials have suggested that L-thyroxine treatment could halve these risks. Our group in Birmingham has just received EME-funding to commence a multi-centred (21 UK centres) randomised controlled trial (TABLET) to confirm this effect in a substantially larger group of women.
Two key issues which remain to be answered are i) whether screening for subclinical thyroid dysfunction and thyroid autoimmunity, and the commencement of L-thyroxine treatment PRECONCEPTION could normalize obstetric outcomes and offspring neurodevelopment, and ii) whether there are specific subgroups of women who would definitely benefit from L-thyroxine treatment.