Thyroxine treatment in pregnant women with hypothyroidism

TOPIC: Treatment of pregnant hypothyroid women

Title: The magnitude of increased levothyroxine requirements in hypothyroid pregnant women depends upon the etiology of the hypothyroidism.

Authors: Loh JA, Wartofsky L, Jonklaas J, & Burman KD. Reference: Thyroid 19 (3): 269-275, 2009



In the United States, many women are on thyroid hormone replacement for hypothyroidism during pregnancy. The optimal management strategy for thyroid hormone (l-T4) dosing in hypothyroid women during pregnancy is controversial. The authors hypothesized that dosage requirements during pregnancy might differ depending upon the nature of the underlying hypothyroidism.


The authors conducted a retrospective review of 45 pregnancies from 38 women whose hypothyroidism was managed during pregnancy. Thyroid function tests were obtained when pregnancy was confirmed, then every 4-8 weeks. The goal was to maintain serum TSH levels between 0.40 – 4.10 mU/L.


On average, the entire group required a cumulative increase from baseline in l-T4 dosage of 13% in 1st trimester, 26% in 2nd trimester and 26% in 3rd trimester (all P values <0.001). Average baseline l-T4 dose for patients with primary hypothyroidism was 93 ± 32 µg/day. These patients required small cumulative dose increases of 11%, 16%, and 16% from baseline in each trimester (P values = 0.125, 0.016, 0.016). Average baseline l-T4 dose for patients with hypothyroidism resulting from treated Graves’ disease was 140 ± 62 µg/day. These patients required the largest cumulative increases in l-T4 dosage, namely 27% in 1st trimester, 51% in 2nd trimester, and 45% in 3rd trimester (all P values = 0.063). Average baseline l-T4 dose for patients with thyroid cancer was 153 ± 30 µg/day. The cumulative l-T4 dose increases for patients with thyroid cancer were 9%, 21%, and 26% in each trimester, respectively (P ≤ 0.03).


The etiology of hypothyroidism plays a pivotal role in determining the timing and magnitude of thyroid hormone adjustments during pregnancy. Patients require vigilant monitoring of thyroid function upon confirmation of conception and anticipatory adjustments to l-T4 dosing based on the etiology of their hypothyroidism.


The article by Jennifer Loh et al. is a retrospective clinical analysis of 38 women (45 pregnancies) with hypothyroidism aiming to assess the timing and extent of thyroxine treatment adjustment during gestation. Patients were subdivided into 3 groups, based on the etiology of hypothyroidism: primary hypothyroidism, presumably of autoimmune origin (group 1: 32%), hypothyroidism following radical cure of Graves’ disease (GD) after thyroidectomy or radioiodine administration (group 2: 21%), and hypothyroidism resulting from total thyroidectomy (Tx) for thyroid cancer ± remnant ablation with I131 (group 3: 47%). My first comment is that this population of hypothyroid pregnant women is somewhat unusual, since more than two thirds of these women developed hypothyroidism following surgical removal of (or I131 administration to) the thyroid, with almost one half of the study group operated for thyroid cancer. In most clinical settings (especially outside the USA), the main cause of hypothyroidism during pregnancy is primary hypothyroidism, resulting from chronic autoimmune thyroiditis (see Table at the end of text). Another noticeable feature is that these women have been managed before and during pregnancy by extremely attentive clinicians (not surprising in view of the well-known names of the coauthors). This is evidenced by the fact that, before the onset of pregnancy, mean serum TSH was 1.18 µIU/ml (for primary hypo-), 0.91 µU/ml (for hypo- after GD cure), and <0.01 µU/ml (for hypo- after Tx for cancer). With only one exception, the adjustment of thyroxine treatment before pregnancy was better than what has recently been indicated in the International Guidelines ( Abalovich et al., JCEM August 2007 ), namely that serum TSH should be kept ≤ 2.5 µU/ml in hypothyroid women before the onset of pregnancy. The main results confirmed the need to adapt thyroxine dosage (from baseline) in the 1st trimester of gestation in each group of women, with mean increments of 11% (Gr 1), 27% (Gr 2), and 9% (Gr 3). From 1st to 2nd trimester, a further dosage increment was required, but remained minimal in group 1 (~5%), while significantly larger in groups 2 & 3 (respectively, 24% and 12%). Finally, there were no further significant increases in thyroxine dosage from 2nd to 3rd trimester to maintain normal serum TSH values. Is the present study helpful to anticipate necessary adjustments of thyroxine replacement in hypothyroid pregnant women? My answer is ‘ yes and no ’. Yes, because the study dissected precisely management differences between these 3 groups of hypothyroid women, based on the etiologies. By doing so, the authors provide a more refined approach to monitoring such women, which is clinically useful. On the other hand, present study also shows that specific characteristics of hypothyroid women play an important role and, therefore, global conclusions cannot be drawn simplistically. A few examples below of what we mean:

  1. Some women in group 2 (hypo- following radical cure of GD) were probably slightly undertreated before pregnancy, and this explains their need for the largest increment in thyroxine dosage of the entire study group.
  2. The limited need to adjust thyroxine doses in group 3 (hypo- after Tx for cancer) can be explained by the relative over-treatment of such women, in order to lower serum TSH near undetectable levels. Otherwise, one would theoretically expect groups 2 & 3 women to behave in a similar fashion, since all were considered to be athyreotic.
  3. Also, women treated for thyroid cancer were presumably truly athyreotic, while it remains plausible that some women with cured GD maintained a minimal amount of functional thyroid tissue. If this assumption is correct, it should have tended (contrarily to the data reported) to reduce differences between necessary thyroxine adjustments shown for groups 2 & 3. Therefore, such differences were most probably related to the classical over-treatment of women with thyroid cancer.
  4. Finally concerning thyroxine adjustment in women in group 1 (primary hypo-), the information given by the authors indicates that the average dosage before pregnancy was 93 ± 32 µg/day. The large standard deviation implies that some women were equilibrated with relatively low thyroxine doses (as low as 25 µg/day), a factor that could impact on the final magnitude of thyroxine dosage changes required in this group.

Thus, this study confirms the need to adjust thyroxine dosage in all hypothyroid women (irrespective of the etiology) and to act rapidly after the onset of pregnancy. The study also shows that the etiology of hypothyroidism plays a pivotal role in determining the timing and magnitude of thyroid hormone adjustment. These observations further confirm the premise that athyreotic women are unable to self compensate for the increased requirements in endogenous T4 production that constitute a major physiologic characteristic of thyroidal economy in the pregnant state. Increased thyroxine requirements constitute an early event after the onset of pregnancy, occurring already within 4 to 8 weeks after conception. Similar findings have been reported in one of the pioneering studies on this topic by Michael Kaplan published in Thyroid (1992). This author investigated 65 hypothyroid pregnant women (29 with Hashimoto’s disease & 36 with thyroid ablation) and he showed that those women with ablation required a significantly greater increment in thyroxine replacement than those women who still maintained some residual functional thyroid tissue. Finally, the present study confirms the well-established notion that thyroxine adjustment in hypothyroid pregnant women cannot easily be predicted, even in athyreotic patients. This was the main reason for us to propose already long ago that thyroxine dose adjustments be always tailored individually. In summary, this interesting retrospective clinical study presents a nice and precise analysis of different etiological groups of pregnant women with hypothyroidism, and shows that each group presents specific management characteristics. As stated by the authors, vigilant monitoring is essential for hypothyroid women during pregnancy, with rapid adjustment of thyroxine dosage. Any predictive and theroretical rule of the thumb used to anticipate thyroxine adjustment should only be used as an overall practical guideline and does not replace the need to monitor carefully and closely changes in thyroid function tests, especially during the first half of gestation. Foot-note: In parentheses, the prevalence of overt hypothyroidism (2.2 – 3.4%) in pregnancy. Positive thyroid autoantibodies were found in 25% to 77% of the women (average: 46%, i.e. 4-5-fold more frequent than in controls). The table is adapted from Glinoer D: Thyroidal and Immune adaptation to pregnancy . In: Merck European Thyroid Symposium on “The Thyroid & Reproduction” (Riga, 2008, in press).

Summary and Commentary prepared by Daniel Glinoer (Related to Chapter 14 of TDM)