The role of levothyroxine in obstetric practice

Abstract

Thyroid hormones play a pivotal role in somatic growth, metabolic regulation and neurodevelopment. There is growing evidence regarding adverse obstetric and perinatal consequences of maternal thyroid hypofunction during early stages of pregnancy. These include: early pregnancy loss, preterm delivery and lower intelligence quotient (IQ) in children. Different clinical guidelines have been published by scientific societies for the management of thyroid diseases during pregnancy and levothyroxine (LT₄) has become a therapeutic agent increasingly prescribed by obstetricians. The aim of this work was to search for both similarities and controversial clinical aspects from the currently available literature. Guidelines published from 2011 onwards have been analysed and compared, in order to clarify the evidence about the involvement of thyroid dysfunction in pregnancy complications and the impact of LT₄ use in their prevention and/or treatment. This review summarizes the most updated knowledge about the effectiveness of LT₄ for pregnancy complications, the current recommendations and its application into clinical practice.

KEY MESSAGES

• The use of levothyroxine in obstetric practices requires a correct diagnosis and to consider the specific recommendations for each thyroid dysfunction entity.

• The effectiveness and safety of levothyroxine treatment in preventing adverse perinatal events in pregnant women with clinical hypothyroidism is supported by all the current guidelines.

• Levothyroxine therapy is strongly recommended in all cases of overt hypothyroidism and in cases of subclinical hypothyroidism associated to positive thyroid autoimmunity.

Keywords:

• Thyroid dysfunction
• pregnancy
• obstetric outcomes

Introduction

It is well established that the developing embryo/foetus is dependent on the maternal supply of thyroid hormones (TH), particularly in the first trimester of pregnancy, with the foetal thyroid only becoming fully functional in the second half of gestation [1]. Traditionally, the prevalence of thyroid hypofunction (Table 1) in women of childbearing age, was around 3% comprised of overt hypothyroidism (OH = 0.3–0.5%) [2] and subclinical hypothyroidism (SCH = 2–2.5%) [2]. However, the last national study in the United States showed that up to 15% of pregnant women had SCH [3]. Moreover, between 8 and 17% [4] of women have positive thyroid antibodies with normal thyroid function (Thyroid autoimmunity, TAI), and approximately 1.5% might present as isolated hypothyroxinemia (IH). Taken together, thyroid hypofunction is likely to be detected during pregnancy.

Table 1. Current definitions for thyroid dysfunction entities.

	Definition
Overt hypothyroidism (OH)	TSH > upper limit of the pregnancy-specific reference range^a plus FT4 < 2.5–5th percentile of the reference range^b.
Subclinical hypothyroidism (SCH)	TSH > upper limit of the pregnancy-specific reference range with normal FT4 concentration.
Isolated hypothyroxinemia (IH)	Normal maternal TSH concentration with FT4 < 2.5–5th percentile of the reference range.
Thyroid autoimmunity (TAI)	Presence of antithyroid antibodies that can be:  – Antiperoxidase (TPOAb)  – Anti-thyroglobulin (TgAb)

^a The ATA recommends population-based trimester-specific references ranges for serum TSH (obtained from pregnant women with no known thyroid disease, optimal iodine intake and negative TPOAb status. In abscence of own reference range, the ATA suggest a TSH upper reference limit of 4.0 mUI/L within weeks 7–12 of gestation.

^b The ATA also recommends assay method-specific and trimester-specific pregnancy references ranges for serum Free T4.

Whilst overt thyroid disease has been repeatedly and consistently been associated with adverse outcomes, even mild degrees of maternal thyroid dysfunction [5,6] have also been associated with a wide range of obstetrical complications such as infertility, recurrent miscarriage, preterm delivery, intrauterine growth retardation, placental abruption and/or impairment of cognitive function in the progeny including lower IQ, and attention deficit hyperactivity disorders [7,8].

The aim to achieve optimal pregnancy outcomes has lately generated a growing interest in the thyroid screening in pregnancy [9], which consequently highlights the importance of reliable diagnostic criteria and indications for treatment [10]. In this regard, the concurrence of clinical guidelines from different scientific societies [11–13] has contributed to an increasing controversy about the need to detect and treat thyroid dysfunction in pregnancy. More recently, the American Thyroid Association (ATA) has published new guidelines [14] for the diagnosis and management of thyroid disease during pregnancy and the postpartum, which updates evidence-based recommendations taking into account the newest scientific advances in this field, although these guidelines did not form an opinion regarding universal thyroid screening.

LT₄ is the first option for the treatment of hypothyroidism because of its effectiveness, costs and safety profile [14,15]. An insufficient maternal thyroid function may, therefore, be optimized with the prescription of an inexpensive, available and well-established therapy [16,17]. Nevertheless, the findings of some recent interventional studies [18,19] highlight the potential benefits and risks of the use of LT₄ in pregnant women with subclinical thyroid conditions.

Identifying thyroid dysfunction in pregnancy

Although the American Congress of Obstetricians and Gynecologists (ACOG), American Thyroid Association (ATA), Endocrine Society, American Association of Clinical Endocrinologists (AACE) and European Thyroid Association(ETA) do not at present recommend universal screening, the debate is still ongoing [20,21].

Whilst scientific societies strongly support aggressive case finding, to identify and test high-risk women for elevated TSH concentrations by the ninth week or at the time of their first visit before or during pregnancy, it is also recognized that in most situations ascertainment of the individual’s risk, with a growing list of criteria, may not be feasible [11].

At the same time, all endocrine, thyroid, and obstetrical societies recommend initiating treatment for overt thyroid disease detected in pregnancy. A conservative estimate (obtained from various large studies) is that per 100,000 pregnant women screened, between 200 and 300 cases of overt hypothyroidism would be detected [22]. These unrecognized cases of overt hypothyroidism in otherwise apparently healthy pregnant women would by itself indicate a potential need for screening [22,23].

But these dilemmas should not mask the real concern: how to identify properly thyroid dysfunction in pregnant women. Recent studies have pointed out the practicalities of the screening: since the standardization of thyroid function tests is still so far [24], all laboratories are being compelled to establish their own reference ranges, adapted to their populations [14]. The use of imported reference ranges is not currently recommended: the accuracy of diagnosis is seriously limited with potential risk of over diagnosis and subsequent overtreatments [25].

Another important fact is that thyroid function is dynamic through the three trimesters of gestation [25]. A recent study has demonstrated that significant changes of thyroid hormone levels occur even within the first trimester [26]. Therefore, any strategy to early detect abnormal values of thyroid parameters in first trimester needs to be related to the exact gestational age at the time of determination. If a possible intervention with LT₄ could benefit obstetric outcomes and foetal neurological development, it should be started as soon as possible, which in turn reinforces the necessity to screen early in the first trimester [27].

Finally, an effective screening policy must include people in charge to interpret the results and decide the necessary therapeutic options [28]. The identification and management of thyroid diseases as well as the monitoring of their treatment has never fallen into the competencies of obstetricians [29]. However, the strong evidence regarding the involvement of thyroid function in obstetric complications is exerting pressure on gynaecologists to acquire skills in this field [30]. One logical approach might be to have more joint endocrine and obstetric clinics.

All these factors add complexity to the diagnostic criteria of thyroid dysfunction and its implementation in clinical settings.

Obstetric complications related to thyroid dysfunction

The association between OH and negative maternal and/or neonatal outcomes has been extensively described from both retrospective and prospective studies [31,32]. In fact, one of the most convincing arguments for universal screening is the urgent need to detect and treat OH at early stages of foetal development or, ideally, at childbearing age before conception [22,27].

Additionally, over the last decade many studies (observational studies, systematic reviews and meta-analyses) have evaluated the relative risks for obstetrical complications in case of mild degrees of thyroid dysfunction such as SCH and/or TAI (Table 2) [5,6,33–43]. Although a myriad of complications, from breech presentation [44] to postpartum depression [45] have been associated with thyroid dysfunction, the most convincing evidence is in the association between thyroid dysfunction and recurrent miscarriage and preterm delivery. Less robust evidence is available for the impact of thyroid dysfunction on growth restriction, hypertensive disorders and gestational diabetes [5].

Table 2. Meta-analysis and observational studies relating thyroid dysfunction and pregnancy complications.

	Subclinical hypothyroidism (SCH)	Thyroid autoimmunity (TAI)
Pregnancy loss Miscarriage	Meta-analysis Maraka [6] (2016) OR: 2.01 Zhang [33] (2017) RR: 1.90	Meta-analysis Van den Boogaard [5] (2011) OR: 3.73 Thangaratinam [34] (2011) OR: 3.90
Preterm delivery	Meta-analysis Maraka [6] (2016) OR: 1.20 Observational studies Arbib [35] (2016) OR: 1.81	Meta-analysis Van den Boogaard [5] (2011) OR: 1.9 Thangaratinam [34] (2011) OR: 2.07 He [36] (2012) RR: 1.41 Observational studies Korevaar [37] (2013) OR: 2.05 Karakosta [38] (2012) OR: 1.70
Growth restriction	Meta-analysis Maraka [6] (2016) OR: 2.14 Tong [39] (2016) OR: 0.98 Observational studies Karakosta [38] (2012) OR: 3.10 Chen [40] (2014) OR: 3.34	Meta-analysis Tong [39] (2016) OR: 1.57
Hypertensive disorders, pre-eclampsia	Meta-analysis Van den Boogaard [5] (2011) OR: 1.70 Maraka [6] (2016) OR:1.22 Observational studies Chen [40] (2014) OR: 2.24	Meta-analysis Van den Boogaard [5] (2011) OR: 1.2
Gestational diabetes	Meta-analysis Van den Boogaard [5] (2011) OR: 1.40 Toulis [41] (2014) OR: 1.39 Maraka [6] (2016) OR: 1.28 Gong [42] (2016) OR: 1.56 Observational studies Karakosta [38] (2012) OR: 4.30 Ying [43] (2016) OR: 1.37	Observational studies Ying [43] (2016) OR: 1.70

Note: Observational studies included in meta-analysis have not been reported in this table.

Finally, isolated hypothyroxinemia (IH) has emerged as a new potential cause for concern [46]. Although it has not been universally accepted as a separate thyroid disease, different studies have pointed out the solid association between maternal IH and impaired neurodevelopment in the progeny [47]. Traditionally, IH was considered in relation to maternal iodine deficiency [48], since pregnant women from iodine-deficient areas were more likely to have low FT4 concentrations. More recent data suggest that IH can also be prevalent in regions with adequate nutritional iodine status [7].

However, the identification of IH is substantially influenced by the assay method used to determine thyroxine levels. In this regard, the ATA guidelines stresses the need of assay method-specific and trimester-specific pregnancy reference ranges when measuring free T4 in pregnant women [14]. Total T4 measurement or free thyroxine index have been proposed as alternative means of estimating hormone concentration during pregnancy [14].

Overall, it is difficult to discern if thyroid hormones and/or thyroid autoimmunity act as causal agents as trials of LT₄ treatment commenced during pregnancy have been lacking.

When to treat thyroid dysfunction?

All the current guidelines [11–15,49] advocate for treating OH at any stage of life and, particularly, during pregnancy in order to prevent serious adverse effects to the foetus [US Preventive Task Force (USPSTF) recommendation level: A; evidence: good]. For women with known OH, it would be most desirable to have a careful preconception adjustment of LT₄ to optimize thyroid function before pregnancy. However, a worrying percentage of LT₄ treated women are not being advised to use contraception until achievement of euthyroidism before conceiving [50,51].

When OH is diagnosed during pregnancy, thyroid function tests should be normalized as rapidly as possible [52]. A correct thyroid replacement therapy has demonstrated to reduce the risk of pregnancy loss, predominantly miscarriages [53] and preterm deliveries [54] in patients with overt hypothyroidism.

For SCH, the current guidelines show different recommendations: for ACOG [13] there is no evidence that identification and treatment of subclinical hypothyroidism during pregnancy improves the outcomes. The Endocrine Society [11] and European Thyroid Association (ETA) [12] guidelines endorse LT₄ replacement independently of the presence of thyroid antibodies.

Different LT4 doses have been proposed according the initial TSH levels, such as: 1.20 µg/kg/d for SCH with TSH ≤4.2 mU/L, 1.42 µg/kg/d for SCH with TSH between 4.2–10 and 2.33 µg/kg/d for OH [52].

In this regard, the ATA’s guidelines consider that LT₄ dose should be adjusted to target a TSH in the lower half of the trimester specific reference range. When this is not available, it is reasonable to target maternal TSH concentrations below 2.5 mU/L [14]. Although concerns exist regarding possible consequences of over-treatment [55], the current guidelines do not include specific criteria to define this condition. But, it seems obvious that the availability of population-based reference ranges may add precision to the targets of treatment.

In patients found to have subclinical hypothyroidism and coexisting iodine deficiency, there is no need to initiate iodine supplementation if they are already taking LT4 [14].

Additionally, the ATA guidelines [14] recommend, firstly, the evaluation of TPO antibody status in pregnant women with TSH concentrations >2.5 mUI/L, as the combination of SCH and TAI has been found to have poorer obstetrical outcomes [56,57].

LT₄ therapy is recommended for women who are positive for TPO-Abs with TSH greater than the pregnancy specific reference range (strong recommendation, moderate quality evidence) and may be considered with TSH concentrations >2.5 mUI/L and below the upper limit of the pregnancy specific reference range (weak recommendation, moderate quality evidence) [14]. With these new recommendations, the ATA tries to identify a subgroup of patients who might be considered within or at the edge of normal thyroid function, according to their TSH values, but whose positivity for TPO-Abs makes them likely to have adverse pregnancy outcomes, such as pregnancy loss of preterm delivery.

In cases of TAI with normal thyroid function (TSH within the pregnancy specific reference range, or ≤4.0 mUI/L if unavailable), Endocrine Society guideline [11] found insufficient evidence to recommend for or against screening and subsequent LT₄ treatment; and the ATA guideline [14] does not currently recommend it.

But the most substantial difference between the current guidelines is related to maternal IH. The ATA affirms that IH should not be routinely treated in pregnancy [14], the Endocrine Society does not include recommendations for IH [9], whereas the ETA suggests that LT₄ therapy may be considered to prevent neuropsychological impairment in children [12].

Indications to screen thyroid function by obstetricians

TH play an active role in the female reproductive system: from the regulation of the hypothalamic-pituitary-ovary axis, ovulatory pattern, endometrial proliferation and mechanisms regulating implantation and early foetal development [58]. Thyroid dysfunction, especially overt hypothyroidism, has been identified as a preventable cause of ovulatory problems and infertility; however, the threshold TSH values that confer implantation success is unknown [59].

According to these guidelines (ATA, Endocrine Society and ETA), screening for thyroid function is recommended in women with infertility and a prior history of miscarriage or preterm delivery. Nevertheless, the ACOG Practice Bulletin [13] recommends testing during pregnancy only for women with a personal history of thyroid disease or symptoms of thyroid disease (who are at increased risk of overt hypothyroidism).

The Practice Committee of the American Society for Reproductive Medicine (ASRM) [60] include recommendations for the screening for thyroid abnormalities to evaluate recurrent pregnancy loss, but they do not establish an upper limit for TSH in pregnancy and they also found insufficient evidence to recommend routine thyroxine (T₄) testing or screening for anti-thyroid antibodies.

Other reviews about recurrent miscarriage [61,62] do not include thyroid dysfunction as a common cause to be investigated for potential treatment to improve outcomes. They propose instead to search for thrombophilia, antiphospholipid syndrome, uterine malformations and chromosomal rearrangements but do not recommend routine testing for thyroid disease nor LT₄ treatment even in cases of raised thyroid antibody status with normal thyroid function tests (Evidence level III) [62].

For preterm delivery, a recent review [63] highlights risk factors, enumerates the currently recommended screening tests and summarizes preventive strategies in both low-risk and high-risk women, but thyroid diseases are missing. Although it is accepted that the prediction of preterm delivery remains modest, most of the preterm delivery risk score do not take notice of the involvement of thyroid dysfunction [64].

Therefore, it seems that the perceived need to assess thyroid function in pregnant women widely varies according to the scientific society that was consulted: obstetricians who follow the ACOG recommendations [13] will be less likely to test thyroid function than those who consult other guidelines from the ATA [14] or ASRM [60].

Levothyroxine in obstetric practices: when should we use it?

The association between infertility and thyroid dysfunction seems to be evident for OH [65], but is relatively inconsistent for SCH and evidence remains limited for TAI [66]. The 2017 Guidelines of the ATA affirms that LT₄ treatment may be considered in case of SCH, thyroid-antibody negative women who attempt natural conception [not undergoing Assisted Reproductive Techniques (ART)]. In spite of the fact that there exists insufficient evidence to determine if LT₄ improves fertility in these cases, is recommended in order to prevent progression to significant hypothyroidism if pregnancy occurs (weak recommendation, low quality evidence). For infertile women with TAI and normal thyroid function who attempt natural conception no recommendation can be made for LT₄ therapy [14] (insufficient evidence).

For women undergoing ART (IVF or ICSI) the ATA’s recommendations for LT₄ treatment are stronger in SCH women seeking pregnancy with ART (undergoing IVF or ICSI) for any TSH elevation >2.5 mUI/L. This therapy may also be considered for TPOAb positive women with normal thyroid function (weak recommendation, low quality evidence) [14].

Additionally, in recent years, we have witnessed different studies where screening strategies have been used and LT₄ therapy was administered in cases of SCH and/or TAI in order to assess its effectiveness in preventing obstetrical complications and trying to increase the rate of successful pregnancies (Table 3) [67–86].

Table 3. Interventional studies with levothyroxine in subclinical hypothyroidism or thyroid autoimmunity during pregnancy.

	Study design	Intervention	Results	Comments
Vissenberg et al. (2012) [67]	Meta-analysis	Included Abalovich et al. (2002) [54] and Negro et al. (2010) [56] for SCH. Five studies reported on the effect of LT4 therapy for TAI.	SCH: ↓ Miscarriage ↓ Preterm delivery TAI: Miscarriage (not significant) ↓ Preterm delivery.	Conclusion: For SCH and TAI, evidence is insufficient to recommend treatment with thyroxine.
Yan et al. (2012) [68]	Prospective non-randomized	53 pregnancies from 34 patients TPO Ab (+) with recurrent miscarriage: — 17 pregnancies with LT4 treatment. — 36 pregnancies without treatment.	NO significant difference in the outcome between groups in live birth rate.	Empirical thyroxine therapy in TPOAb(+) pregnant women did not seem to improve outcome.
Lepoutre et al. (2012) [69]	Retrospective	96 TPO Ab (+) pregnant women — 49 treated with LT4. — 47 no treated.	↓ Miscarriage rate	Potential benefit of universal screening and LT4 treatment in pregnant women with TAI.
Reid et al. (2013) [70]	Meta-analysis	For SCH included Rotondi et al. (2004) [71], Negro et al. (2006) [72], Negro et al. (2007) [73] and Yassa (2010) [16].	↓ Preterm delivery. No effect on pre-eclampsia.	There was a trend towards reduced risk of miscarriage with LT4, but did not reach statistical significance.
Velkeniers et al. (2013) [74]	Meta-analysis	Women with SCH undergoing ART. Included Negro et al. (2005) [75], Abdel Rahman et al. (2010) [76] and Kim et al. (2011) [77].	↑ Delivery rate. ↓ Miscarriage rate.	In an ART setting, no data are available on the effects of LT4 treatment on preterm delivery or pre-eclampsia.
Bernardi et al. (2013) [78]	Prospective non-randomized	Women with SCH and history of recurrent early pregnancy loss — 24 treated with LT4. — 15 no treated.	NO significant difference in the outcome between groups in live birth rate.	The prevalence of SCH in this recurrent early pregnancy loss (REPL) cohort was 19%.
Bartáková et al. (2013) [79]	Prospective non-randomized	Cost-effectiveness analysis of screening and treatment after spontaneous abortion (SpA) — 73 SCH and/or TAI treated with LT4. 38 SCH and/or TAI untreated.	↑ Successfully completed subsequent pregnancies.	Screening for thyroid disorders in women after SpA and treatment with LT4 is cost-saving and it improves the subsequent pregnancy rate.
Lata et al. (2013) [80]	Prospective non-randomized	Women with two or more consecutive miscarriages. — 31 with TAI and 27 with SCH were treated with LT4. Compared to 100 healthy women without a history of miscarriage.	Following LT4 treatment, NO difference in miscarriage rate between hypothyroid and euthyroid individuals in TPO Ab (+) women.	The prevalence of TAI was higher in pregnant women with a history of recurrent miscarriage compared with healthy pregnant control population.
Ma et al. (2015) [81]	Prospective non-randomized	Screening group (675 pregnancies) was compared to control group (996 pregnancies). 105 SCH from screening group were treated with LT4. 252 SCH from control group did not receive treatment.	↓ Miscarriage rate. ↓ Foetal macrosomia risk. ↑ Caesarean risk.	Screening and intervention of SCH can reduce the risk of miscarriage. No significant differences were observed in other pregnancy outcomes between the two groups.
Maraka et al. (2016) [82]	Retrospective	82 women with SCH were treated. 284 women with SCH did not received treatment with LT4.	↓ Risk of low birth weight. ↓ Risk of low Apgar score.	Other pregnancy-related adverse outcomes were similar between the two groups.
Negro et al. (2016) [83]	Prospective Randomized	198 euthyroid, TAI (+) treated with LT4. 195 euthyroid, TAI (+) untreated. 197 euthyroid, TAI (−) untreated.	NO significant difference in miscarriage or preterm delivery rate between the three groups.	Levothyroxine intervention had no impact on the rate of miscarriage or preterm delivery in euthyroid TAI (+) women.
Nazarpour et al. (2016) [84]	Prospective Randomized	65 TPO Ab (+) women treated with LT4. 66 TPO Ab (+) women untreated. 131 TPO Ab (−) women untreated.	↓ Preterm delivery rate.	The number needed to treat (NNT) for preterm birth was 1.7
Ju et al. (2017) [85]	Prospective non-randomized	184 women with SCH were treated. 273 women with SCH did not received treatment with LT4.	↓ Gestational diabetes. ↓Foetal macrosomia.	The earlier the gestational week when the target TSH is achieved through treatment, the lower incidence of complications.
Blumenthal et al. (2017) [86]	Prospective non-randomized	28 women with SCH were treated. 111 euthyroid women (control group).	NO adverse outcomes in women diagnosed and treated for SCH.	No significant differences were observed between the two groups.

Note: Observational studies included in meta-analysis have not been reported in this table.

Although the results substantially vary (taking in account the groups or cohort studied, different cut-off for TSH levels, risk of bias and/or the selected outcomes as endpoints), LT₄ treatment seems to be effective in reducing miscarriage and/or preterm delivery rates in SCH, especially when TAI is present. Nevertheless, no potential benefits have been found in reducing hypertensive disorders or gestational diabetes.

In spite of the fact that the ATA admits that it seems reasonable to recommend or consider levothyroxine treatment for specific sub-groups of pregnant women with SCH (history of miscarriage or preterm delivery) the current recommendations have only taken into account TSH concentrations and TPO antibody status. However, for obstetricians, the use of LT₄ should be considered differently according to low or high-risk perinatal factors or history of adverse reproductive events (Table 4).

Table 4. Summary of current evidence for the use of levothyroxine in thyroid dysfunction.

	Indications	Warnings
Overt hypothyroidism (OH)	— All the cases must be treated	— Treatment should be carefully monitored.  — Avoid sub/over treatments.
Subclinical hypothyroidism (SCH)	— Treatment reduces the risk of:    – Miscarriage.    – Preterm delivery.    – Assisted reproductive techniques (ART).  — No evidence of effectiveness for:    – Gestational diabetes.    – Hypertensive disorders    – Infant cognitive function.	— The diagnosis must be based on reliable and own trimester-specific reference range.  — If reference levels are not available, the indications for treatment should be considered with caution (TSH >4 mUI/L or TPOAb positive with TSH >2.5 mUI/L).
Thyroid autoimmunity (TAI)	— Treatment can be considered in cases of:    – Assisted reproductive techniques (ART).    – Recurrent miscarriage.    – Preterm delivery^a.  — No evidence of benefits in:    – Any other obstetric complications.	— The use of thyroxine must be offered individually in cases of high-risk factors or history of adverse reproductive events.  — The use of thyroxine in absence of obstetric risk factors is not currently justified.
Isolated hypothyroxinemia (IH)	— There is no available evidence of benefits from thyroxine therapy.	— Use of iodine supplementation can be useful in iodine-deficient areas.  — Interventional studies in cases of IH are lacking.
Normal thyroid function	— There is no indication for use of thyroxine in normal conditions.	— Potential risks of iatrogenic use of thyroxine in pregnancy:  – Growth restriction  – Abnormal brain morphology in children.

^a This recommendation is based on observational studies.

At the same time, it is important to highlight that all the studies of LT₄ replacement performed until now did not include any other preventive or therapeutic approaches: data are lacking regarding the effectiveness of LT4 in combination with aspirin for recurrent miscarriages or progesterone and/or pessary for preterm delivery.

This discordance between obstetrical and endocrinological approaches might be explained by the importance of thyroid function for every specialist, their skills to identify abnormal finding and to monitor LT₄ treatment properly [87].

Levothyroxine: a double-edged sword?

LT₄ has become the most prescribed drug in the USA and the third most prescribed drug in the UK [88]. Due to the increased concern regarding the relationship between thyroid dysfunction and maternal/foetal adverse outcomes, women are more likely to be identified during pregnancy with thyroid hypofunction and treatment with levothyroxine is quite often used by gynaecologists and centres for reproductive medicine [75–77].

However, the effectiveness of LT₄ treatment in preventing adverse perinatal events in pregnant women with SCH and/or TAI needs to be assessed from large randomized-clinical trials. It is possible that pregnant women with moderate to severe thyroid dysfunctions (higher TSH levels) would benefit from LT₄ therapy, while those pregnant women who have mildly increased TSH might not find benefit or even be overtreated [89].

It seems also probable that LT₄ use would be more effective in certain populations of high-risk pregnant women: history of adverse perinatal events, such as infertility, recurrent miscarriage or preterm delivery. However, a treatment threshold should be established for these patients. A recent national survey in USA showed that thyroid hormone treatment was associated with decreased risk of pregnancy loss among women with subclinical hypothyroidism, but increased risk of other pregnancy related adverse outcomes such as preterm delivery, gestational diabetes or pre-eclampsia [90].

It has widely been suggested that women on thyroxine substitution should increase their thyroxine dose immediately on confirmation of pregnancy. However, some studies have indicated that routinely increasing the LT₄ dose may result in oversubstitution in a substantial portion of these pregnant women and, a subsequent risk of foetal loss [91].

An important issue has also been pointed out by different authors: little is known about the transference of LT₄ through the utero-placental unit, from additional LT₄ dosage in mothers to foetal blood [10]. In this regard, we currently lack reliable foetal markers obtained by non-invasive methods, to monitor the effects of maternal LT₄ therapy on foetal thyroid hormone concentrations [10,92]. Samples obtained by cordocentesis have shown that around 60% of foetuses from euthyroid mothers with TAI treated with levothyroxine might have circulating free T₄ concentrations higher than normal levels [92].

Experimental studies have demonstrated that transient thyroid hormone exposure profoundly impacts the thyrotrope population during a critical period of pituitary development and may have long-term implications for the functional reserve of thyroid-stimulating hormone (TSH) production and the TSH set point later in life [93].

International guidelines recommend treatment aims based solely on maternal TSH concentrations but serum free T₄ usually increases sharply after LT₄ therapy [94]. Results from Generation R study have shown some concerning aspects related to high maternal free T₄ concentrations: an association to low birth weight and an increased risk for small for gestational age (SGA) newborns [95] and, more recently, an association with lower child IQ and lower grey matter and cortex volume [55]. Although the relationship between maternal free thyroxine concentration and child IQ shows an inverted U-shaped association in this study, high maternal fT₄ concentrations are commonly observed after LT₄ supplementation. More recently, a similar deleterious effect on behaviour was presented at the British Thyroid Association Annual Meeting (Hales et al.) [96].

It is important to highlight that the use of LT₄ when there is no recommendation may carry the risk of growth restriction [95] or adverse child neurodevelopment outcomes [55]. Therefore, it use should ideally be monitored in specialist join obstetric/endocrine clinics.

Acknowledgements

We sincerely thank to the Virtual Library from Public Health System of Andalusia for providing us with information resources to complete this review.

Disclosure statement

The authors report no conflicts of interest.

Additional information

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.