TSH and Pregnancy: Treat the Test or the Patient?

Susan is newly pregnant. She had some thyroid tests done earlier this year and was diagnosed with subclinical hypothyroidism. Her TSH was 6 mIU/ml with a normal thyroxine and negative anti thyroid antibodies. She has read in a forum online that she needs to be treated with thyroxine but she is worried about taking medication when pregnant. What do you advise her?

Hypothyroidism has long been associated with miscarriage, preterm delivery and other adverse outcomes in pregnancy. Standard pre-conception advice for women with hypothyroidism is to check TFTs before conception, ensure they are in the euthyroid range and to escalate the dose of thyroxine as soon as possible in pregnancy. Guidelines vary locally, but increasing the thyroxine dose by 30% as soon as pregnancy is confirmed is typical. TSH reference ranges differ in pregnancy, and the recommended upper limit of TSH is 2.5 in the first trimester and 3.0 mIU/ml in the subsequent trimesters BJGP2016;66:538. (although a TSH <4 is recommended in the latest 2017 US Guidelines)

So far, so (relatively!) uncontroversial. But what about women like Susan with subclinical hypothyroidism? Do they benefit from treatment as well? This is a big question, as we know that it affects 3% to 15% of pregnant women BMJ2014;349:g4929. Current guidelines advise that all women with subclinical hypothyroidism who are pregnant or intending conception should be treated with levothyroxine BJGP2016;66:538, but what does the primary evidence tell us? When you scratch under the surface a complicated picture for Susan emerges...

Observational studies do consistently show that subclinical hypothyroidism in pregnancy is associated with multiple adverse maternal and neonatal outcomes, but a systematic review in 2016 concluded that the value of thyroxine treatment in preventing these outcomes was uncertain. In 2015 a BMJ uncertainties paper BMJ2015;351:h4726 raised the question: are we over treating subclinical hypothyroidism in pregnancy? 

So, some recently published new evidence is welcome BMJ2017;356:i6865. This was a large cohort study of over 5,000 pregnant women with a TSH of between 2.5 and 10 mIU/L, and they compared women who were treated with thyroxine with untreated women. The results were very interesting. Treated women had a significant reduction in pregnancy loss compared to untreated women, from 13.5% to 10%. However treated women also had higher risks of other adverse outcomes compared to non-treated women, including pre-eclampsia (5% to 6%), pre-term delivery (5% to 7%) and gestational diabetes (9% to 12%). Intriguingly the authors suggest that since the observed benefit of treatment was in preventing early pregnancy loss, treatment may only be needed in the first trimester. The study raises yet more questions and presses the case for an urgent need for the prospective randomised controlled trials that we currently lack to answer them.  

So, what do we advise Susan? Our role, as always, is to facilitate shared informed decision-making and to treat the patient and not the test. We need to repeat the TFTs, explain that subclinical hypothyroidism is associated with an increased risk of adverse outcomes, that guidelines do recommend treatment and to offer her a specialist opinion…but also to explain that there is real uncertainty regarding the outcome of treatment. We should encourage her to do her own research but if she does it is likely that Susan, like me, may then find she ends up ‘better informed but more confused’!

If you found this case study helpful you will love our Hot Topics course. Click here for information about face-to-face events and online courses.

This blog was originally published via the NB Medical Hot Topics Blog, on 23rd August 2018.