Hello, everyone, and welcome to the session on the hyperthyroidism in pregnancy. My name is Sun Lee. I'm an endocrinologist at Boston University School of Medicine in Boston, Massachusetts, and it's my pleasure. I'll be moderating the session. So after the session, if you can come to the mic to ask questions, that would be great. And it is my pleasure to introduce our speaker, Dr. Chrysoula Dosiu, who received her medical degree from Harvard Medical School, and then did her training in endocrinology at Stanford University, as well as her training in clinical epidemiology. She is currently a clinical professor in the division of endocrinology in Stanford University School of Medicine, as well as the medical director of the Stanford Thyroid Eye Clinic, and program director of the endocrinology fellowship at Stanford University. Welcome. Hi. Good morning, everybody. Thank you, Sun, for the kind introduction. I'm very excited to be with you today to speak about one of my favorite topics, the evaluation and management of hyperthyroidism in pregnancy. I would like to also thank the organizers and Dr. Liang for inviting me to speak to you today. So I'm going to start us off with a case. This is a case of a woman who was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. She was diagnosed with hyperthyroidism in her first year of pregnancy. This is a 44-year-old female with infertility and a new diagnosis of Graves' disease that comes shortly after egg retrieval. She's planning an embryo transfer ASAP. Her TSH is undetectable, free T4 is high at 4.2, and she has an elevated TSI at 4.6. And she wants to know specifically, what is the safest treatment for her hyperthyroidism with respect to her baby? And she emphasized to me that really, here we're looking at this purely from baby's point of view, future baby's point of view. So before I get started, I just want to get a show of hands among the audience. Who would advise this woman that at this time, the safest treatment for her hyperthyroidism with respect to her baby is PTU? Can I get a show of hands? Okay. How about methamazole? Nobody? Okay. How about radioactive iodine? She's not pregnant yet. Nobody. Oh, a couple of people. Okay. And how about thyroidectasis? How about thyroidectomy? All right. Excellent. So keep in mind what you answered, and check in with yourself at the end of the talk and see if you still have the same answer or not. So today, my goal will be to initially cover briefly some background on maternal and fetal thyroid physiology, as well as the prevalence and differential diagnosis of maternal hyperthyroidism. Then discuss the evaluation of hyperthyroidism in pregnancy specifically, some of the complications of maternal hyperthyroidism, both for the mom as well as the fetus. And then spend the second half of the talk on treatment. And hopefully we'll have some time for some Q&A at the end. So there's a lot of changes that happen during pregnancy that affect thyroid physiology. And I'm just going to focus on the two most important ones. The first is the rise in HCG that happens with a peak in the first trimester. And HCG, as you all know, is an alpha-beta glycoprotein with very high homology to TSH. In fact, the alpha subunit is exactly the same. So it can interact with the TSH receptor at high concentrations and cause stimulation of the TSH receptor, leading to a small rise in free T4 and a reciprocal fall in TSH. So that's the first major change. The second major change is that under the influence of high levels of estradiol, TBG levels significantly increase. TBG is the major carrier protein for T4 and T3. And to meet this extra demand imposed by these new binding sites on TBG that are unfilled, the thyroid has to bump up its production of thyroid hormones significantly. So the total T4 levels tend to rise by about 50%. The fetal thyroid forms from the anterior pharyngeal floor at 16 to 17 days of gestation. It assumes its final position and shape by 8 to 9 weeks. It starts trapping iodine at around 12 weeks of gestation, but T4 is not secreted until 18 to 20 weeks by the fetal thyroid at the time of full maturation of the HP axis. And the prevalence of maternal hyperthyroidism is about 2.4%, 1.6% subclinical hyperthyroidism and 0.6% overt hyperthyroidism. The differential diagnosis is actually the same as we have in the non-pregnant patients, but by far the most common diseases that we deal with in terms of diagnosis in this population are Graves' disease or gestational transient thyrotoxicosis, both usually presenting in the first trimester. So I'm going to spend a little bit of time discussing how we distinguish between those two entities. Graves' disease affects about 0.2% of pregnancies. It accounts for 85% of cases of hyperthyroidism in pregnancy, so by far the most common etiology. And the disease can present, recur, or be exacerbated in the first trimester. So typically we see this happen around 10 to 15 weeks. And subsequently there's significant improvement, especially in the third trimester, presumably due to the impact of high levels of progesterone on the immune system. And this really beautiful study from Anderson's group in Denmark looked at the incidence of Graves' disease around the time of pregnancy in women of reproductive age and showed kind of a flat incidence in the few years before pregnancy. There's this little peak that you see in the first part of pregnancy with a fall in incidence, as I mentioned in the second part of pregnancy, and then a pretty big peak in the postpartum phase where we know that a lot of autoimmune diseases flare, presumably again to do the rapid changes in progesterone levels. So how do we distinguish GDT from Graves'? So gestational thaloxycosis is much more common, remember, so it affects about 2% of pregnancies versus Graves' disease that affects 0.2% of pregnancies. GDT typically does not have a prominent goiter, but Graves' disease has usually a prominent goiter. GDT has no antithyroid or TSH receptor antibodies, whereas Graves' disease does. And then GDT resolves typically by 20 weeks of gestation and is associated with no adverse outcomes if left untreated, which is very important to note. Graves' disease can be associated with ophthalmopathy and dermopathy in patients, although those typically are not seen in the setting of pregnancy, but they can be seen, in which case that is a pathognomonic finding on exam. It more commonly has T3 elevation, so typically the ratio of T3 to T4 is 20 to 1 or higher in Graves' patients, and Graves' disease importantly has adverse outcomes to the mom and the fetus if left untreated. So how do we evaluate hyperthyroidism in pregnancy? So the typical tests that we use in non-pregnant patients have value here, but there are a few caveats that we need to be aware of. So first, we need to use pregnancy and trimester-specific reference ranges for both TSH and 54 if we're going to use these tests. And remember what we talked about regarding TSH, that it significantly falls during pregnancy. We now know that there is a lot of variation according to geography and different populations studied. I have here on the bottom left just a figure from one of the initial papers that highlighted this variation. This is from Hong Kong in 2001, has longitudinal data on patients serially tested during pregnancy, and as you can see, the reference ranges in the population of women that did not have any thyroid disease, you can see there how much lower they are than the non-pregnant reference ranges and how they changed during the three trimesters. The second point is that immunoassays for 54 performed very poorly in pregnancy. This is illustrated by the figure I have on the right bottom side there by Sapan, who took about 30 women who had no thyroid disease in the third trimester and just tested their serum for 54 using either equilibrium dialysis in the first bar and then a bunch of different immunoassays in the other bars shown there. And in the white boxes you see the normal reference range for non-pregnant patients. And you can see that uniformly all these women cluster in the bottom end of these reference ranges, and a lot of them actually fall off below the normal reference range using the non-pregnant reference range. So it is important to realize that, and that's another reason why we need to use trimester and assay-specific reference ranges that are obtained during pregnancy. I'm going to just make a point here that if one really wants to measure free T4 as accurately as possible, one way that one can do that is by getting a free T4 index that takes into account both the total T4 and TBG, and usually that's a send-out test that, however, is available in most labs. What about other tests? Total T4 and total T3 I think are pretty valuable as well during assessment of maternal hyperthyroidism in pregnancy, especially after 16 weeks of gestation when one can use the adjusted range for pregnancy that is 1.5 times the non-pregnant reference range. And another important point is that serial TFTs in pregnancy can be important, especially when you're grappling with a differential diagnosis, and as I mentioned, GTT tends to improve dramatically after the first trimester. And finally, TRAB and TSI are quite important. We're going to spend quite a bit of time on this later on in the talk, but these can help in the differential diagnosis. As I mentioned earlier, Graves would have positive antibodies versus GTT that would be negative, but also realize that levels decrease as pregnancy progresses. And this is shown nicely in this paper by Amino that looked at serial levels of TBBI or TSI in pregnancy, and at the bottom you see the one, two, and three are the three different trimesters, and PP is the postpartum phase, and we can see a nice drop of the antibody titers during pregnancy, again, presumably due to the effects of progesterone on the immune system, and then a rebound in the postpartum phase. Okay, so what are the complications of maternal hyperthyroidism? And the first paper that I could find in the literature discussing these was published in Lancet in 1929, and this described 71 patients with Graves' disease. This was in women of reproductive age in the prime years of reproductive age, and the pregnancy rates were only 54% lower than expected, and of those that were able to get pregnant, more than half had a miscarriage, preterm birth, or were stillborn. And the results are, the conclusion of the study is kind of interesting, I just put it up there to see how people were thinking back in the 1920s about this disease. So to a large extent, the poor results of pregnancy in the present series, miscarriage or premature birth in 50%, may be attributed to shocks, mental stress, and inadequate rest and treatment, for where satisfactory conditions were attained and mental stress was absent, the results were generally good, if only it were so simple. So since then, we've come a long way regarding the understanding of complications and how we can avoid them. I'm just going to reference a couple of recent papers. This was published in Clinical Endocrinology in 2020, it was a retrospective assessment of about 2,000 women in Finland, and highlighted the excess risk for certain outcomes, such as gestational hypertensive disorders with an odds ratio of 1.3. C-section rates were up, odds ratio of 1.3, placental abruption about 2, and fetal outcomes such as preterm birth, neonatal ICU, and small fetal gestational age were all elevated. Interestingly, in this study, I've highlighted with little asterisks there, some of these outcomes were also elevated in women who were no longer on anti-therapy drug therapy, but had a prior history of treated Graves' disease. That's kind of an area of growing interest as to whether that confers any extra risk in pregnancies, even after treatment. The second study was just published, much larger, 17,000 women, a retrospective study in the U.S., that again reports a lot of these outcomes. So again, we see gestational hypertension, preeclampsia, as well as venous rhomboembolism risks increased in the mother, increased rates of premature delivery, placental previa, and preterm premature rupture of membranes, and then increased risk for intrauterine fetal death and small fetal gestational age. It's important to note that in contrast to overt hyperthyroidism, subclinical hyperthyroidism, defined as traditionally as low TSH with normal T4 and T3 levels for pregnancy, is not associated with adverse obstetrical outcomes. And the data for this mainly come from the big Casey study in 2006 in the U.S. of 26,000 pregnant women, of whom about 1.7% had subclinical hyperthyroidism. And there was absolutely no difference in 11 neonatal and 6 maternal outcomes in these women compared to women with normal TSH. And there have been a couple of recent meta-analyses on this topic that again found no association of subclinical hyperthyroidism with gestational hypertension, preeclampsia, or IUGR. So I want to highlight another really wonderful review on this topic. This is a paper by Carolyn Nguyen at USC and a Messman's group. So this looked at maternal Graves' disease and tried to highlight all the complications that occur and what treatment can improve. So we see the maternal and obstetrical outcomes that we just discussed, but also the paper highlights nicely the fetal and neonatal complications. And for the next few slides, I'm just going to spend some time discussing potential effects on neurodevelopment, which is an area of growing interest, and then fetal and neonatal hyperthyroidism. So before we do that, I think it really helps to think of what actually crosses over the placenta. So I think of it in terms of things that increase fetal exposure to thyroid hormone or things that decrease fetal exposure to thyroid hormone. So the things that actually increase fetal exposure to thyroid hormone would be maternal thyroid hormone that crosses over the placenta, and then stimulating TSH receptor antibodies, or TSI. What decreases the fetal exposure to thyroid hormone would be antithyroid drugs and inhibitory TSH receptor antibodies. And then we've known for quite some time that low maternal FT4 poses a risk to fetal neurodevelopment, and of course severe maternal hypothyroidism can be linked to that through this mechanism. But what about high maternal FT4? And there was a provocative paper about three, four years ago on this topic that suggested a link of really high maternal FT4 with lower IQ in the offspring. And this has spurred a number of other studies, and finally this meta-analysis that was published a few years ago looking at the data of three large cohort studies with about 9,000 altogether, 9,000 mother-child pairs, and tried to see if there's any association of really high FT4 with adverse neurodevelopmental outcomes. And they did not find an association with IQ, verbal, I'm showing here the data on the left for verbal IQ, there was no association, or for non-verbal IQ. But they actually interestingly did find a small association with the risk for autistic traits with an odds ratio of 1.9. And in recent years there have been a number of other studies, I've highlighted here the ones I thought were the most important that came out since 2018, that looked at this. So the first one was a prospective cohort study from the UK of about 4,000 mother-child pairs that showed no association of maternal hypothyroidism with child standard assessment test scores in the UK from 54 months up to 15 years of age. So the kids performed just as well as their peers. There was a Danish case cohort study of about 2,000 cases that did show some association with outcomes such as epilepsy in the first year of life with an adjusted hazard ratio of three, autism spectrum disorder with an adjusted hazard ratio of two, and no association with ADHD. And finally there have been two meta-analyses published that show some quite mild but statistically significant associations with ADHD and epilepsy. So I think this is an area of growing investigation and we'll have to see what better designed prospective studies will show in the next few years. So fetal hypothyroidism affects 1% of pregnancies in patients with current or previous Graves disease and it is caused by TSH receptor antibodies that stimulate directly the fetal TSH receptor after it becomes fully functional after 18 weeks of gestation. And the manifestations can be fetal goiter, tachycardia, IUGR, and accelerated bone maturation. There's increased risk of intrauterine death, stillbirth, and skeletal developmental abnormalities in fetuses who have this disease. And antithyroid drug administration to the mother improves the prognosis. So you can use a mother as a vehicle to get this medication over to cross with the placenta and treat the fetal thyroid. But to diagnose this and to have a suspicion for it, you have to check maternal TSH receptor antibodies. So this is critical. We're going to talk about this a little more extensively. So first I want to show some pictures of a fetal goiter. So here we can see a profile picture of the neck, longitudinal view. So you can see the face here, the chest here, and here you can see the flat outline of the neck. And then in a fetal goiter, you can see the protuberance of the thyroid over there. And here we have a transverse view of the thyroid, nice fetal normal thyroid, and here you can have an image of a fetal goiter with both increased vascularity as well as significant growth. So sorry, I could not see very well. This is actually the normal contour, and this is the bump. You can see over there in the neck. It's very hard to see the screen from over here. Okay, so neonatal hyperthyroidism occurs in 1 to 5% of neonates born to mothers with Graves' disease. It affects 1 to 25,000 to 1 in 50,000 of newborns. It's mostly present by day 10 to 14 of life. And the manifestations of this disease include tachycardia, irritability, poor feeding, sweating, sleeping difficulty, craniosynostosis, and microcephaly if severe. It is associated with significant morbidity and mortality, so 12 to 20% if left untreated. And the treatment is methimazole as well as beta blockers with very frequent monitoring of the baby. The average length of treatment is two to three months, and the antibody is fully clear by four to six months postpartum. And the risk of neonatal hyperthyroidism really depends on the levels of these antibodies in the mom, especially in the second and third trimester. So TBII levels over five or TSI levels over 400 have been found to be extremely sensitive in predicting neonatal hyperthyroidism. And in a recent systematic review of about 20 studies, the one by Van Dyck that I'm quoting there at the bottom, a TRAB level over 3.7 times the upper limit of normal with whatever assay you're using was the lowest level associated with neonatal hyperthyroidism. So what are the guidelines that we have for measuring these antibodies and when? So I'm gonna go through a couple of different scenarios to illustrate. So in a euthyroid pregnant woman who has a past history of Graves' disease, but is currently on no medication, so she could have been treated previously with antithyroid drugs, but currently has her thyroid in place and is euthyroid or no meds, measurement is not necessary. And that is because this woman has a functioning biosensor, right? Her own thyroid is there, so if she has normal TFTs, that means that really she does not have clinically significant TSI titers. In mothers who are untreated or have ATD-treated Graves, the recommendation is to check when pregnancy is confirmed in the beginning, and if still taking ATDs later in pregnancy, to repeat testing at 18 to 22 weeks and 30 to 34 weeks. And the reason why we have two time points is the 18 to 22 week time point captures the risk for fetal hyperthyroidism, and the 30 to 34 week time point captures the risk for neonatal hyperthyroidism. And my recommendation is, if you're gonna be checking at 18 to 22 weeks, to check on the earlier side of that TSI, and depending on what lab you use, TSI antibodies can take a while to come back, and the fetal goiter can develop if you wait, if the baby is to develop, and it can develop pretty fast if it's meant to be, so I'd rather know earlier rather than later. In women who have a previous history of Graves that has been already treated with definitive therapy, such as RA or thyroidectomy, or who have a prior newborn with hyperthyroidism, the recommendation is again to check in early pregnancy, and if elevated, repeat at 18 to 22 weeks. If still elevated at that time, repeat at 30 to 34 weeks. If it's undetectable or low in early pregnancy, there's no further testing needed, because remember, the antibodies fall as pregnancy is progressing, so if you already are starting with really low antibodies, then there's no point in checking again. And what is abnormal? So in general, I would say three times elevation with whatever assay you're using is considered abnormal, and that would indicate the need for further fetal monitoring, such as with pulse rate, growth rate, and assessment for fetal goiter. Okay, so we're moving on to the second part of the talk. I'm gonna focus on treatment of maternal hyperthyroidism. And first of all, the good news is that many of the adverse outcomes that we are afraid of the most significantly improve with treatment, and we have good data for that. This is just one of the initial papers that showed the dramatic decrease in preterm labor, stillbirths, and CHF with adequate treatment. This was published in 1989. As you can see here, a very small N of these patients, but still we have a drastic decrease in the complication rates with treatment. So you can see the untreated rates, they're partially treated in the second column, and then the rates for women who are euthyroid at delivery. This is another beautiful study with much bigger numbers by Messman's group that looked at complication rates, such as low birth weight, prematurity, small for gestational age, and severe preeclampsia in three different groups of women. So in red bars, you see the rates in uncontrolled women, in yellow, the ones controlled at delivery, and in black, the ones that have been controlled throughout pregnancy, and you can see beautifully how effective treatment is in reducing these rates. So where are we in 2022 with treatment? So the topics I'm gonna try to cover are first regarding medical treatment, PTU or methimazole. We know they both cross the placenta, but are they equally safe? The topic of surgery, when is that the preferred option? The topic of radioactive iodine. We clearly know it's contraindicated during gestation, but how about its use before pregnancy? And I think it helps to take a historical trip for a second to gain some perspective about this discussion, especially with respect to antithyroid drug treatment. So antithyroid drugs were first introduced in 1943, and this here is the landmark article by Astwood that introduced treatment of hyperthyroidism with thiourea and thiouracil. If you haven't read it, I urge you to read it. It's actually pretty exciting and impressive with the lens of somebody who has been treating folks as all of us have the lens of today. In 1946, Astwood investigated 300 more compounds and discovered finally the more potent 6-N-propylthiouracil, and that led shortly thereafter to its approval by the FDA in 1947. And in 1949, methimazole was found to be more potent and less toxic thiourea drugs, so that was introduced then. And even from the very beginning, there were reports of the potent effects of antithyroid drugs on the fetal thyroid. This is from a paper from Lancet, 1945, that showed the effects of thiouracil in pregnancy on the fetal thyroid, causing a goiter. So on the left, you have longitudinal sections of a thyroid gland from a six-month-old fetus from a mom that was treated aggressively with thiouracil, and you can see the much more enlarged cross-section compared to the right slices that are from a normal seven-month-old fetus. So it was the first report of a fetal goiter. And then in 1950, Dr. Bell in Boston reported in JAMA the first outcomes of 21 women that were treated with thiouracil and Lugol's, followed by subtotal thyroidectomy. And even though the fetal loss was still high at 33%, it's much better than what we had before, right? Remember, the fetal loss before these drugs was 55% initially, right? And in the conclusion of that paper, some of the observations of Dr. Bell that induced myxedema should be avoided in patients receiving a thiouracil drug by repeated observations of the metabolic rate and blood cholesterol throughout the period of medication. I thought this was pretty cool to realize because we tend to forget that in the 1940s, there were no thyroid tests available. So that's how folks measured the efficacy of treatment. On the right here, you see from the initial JAMA article by Astwood, you see the changes in basal metabolic rate going down and body weight going up after thiouracil was initiated for a few weeks. And that's how they were measuring their outcomes. So next, a big report was 1951. This was Astwood's own experience using PTQ in pregnancy. And this was published in JCNM. Back then, it was called Journal of Clinical Endocrinology. And he reported on 19 patients that delivered 22 babies. All the babies were healthy and three were premature, but there was no fetal goiter. And he treated all these patients himself. And it's important to note that, that he was very, very careful with the titration of this drug during pregnancy. So he initially was doing PTQ 100, Q8. And as soon as there was clinical improvement by the measures I mentioned earlier, he decreased to 50 Q8 and then was stopped at birth. And his conclusion was that antithiarrheic therapy can safely be continued throughout pregnancy without harm to either the mother or child. He also suggested that the findings indicate that if congenital anomalies occur during pregnancy complicated by hyperthyroidism, these anomalies are not brought about by treatment. 1954, we have the first kind of review of the literature on outcomes after ATD treatment versus thyrodectomy. So these authors reviewed all the case series and all the case reports in the literature and had 83 women who had been treated with ATDs and 122 women treated with thyrodectomy. And then in the real world, right, not by Asschud's magical hands and regimens. So perinatal mortality with ATDs was about perinatal mortality with ATDs was 7% with neonatal goiters occurring in almost in one in five babies and one death from suffocation from a large goiter. Whereas thyrodectomy had a fetal mortality of 3% and maternal mortality of 0% in this large review. So the conclusion of that paper was that treatment with iodine and thyrodectomy is to be preferred to other treatment as it carries only a very slight risk for mother and child. Antithyroid treatment jeopardizes the life and normal development of the fetus, the more so the closer to normal delivery the treatment is continued. And it's interesting reading back this literature how keenly aware they were of the fact that fetal goiters really were much higher in women who continued antithyroid drugs till the end of delivery. So fast forward 20 years, so we're now in the 1970s. And in 1972, Milham and Elledge first reported the association of ATDs with birth defects after two out of 11 newborns with aplasia acutis were found to have been exposed to methimazole in utero. And subsequent case reports and additional defects were described. And then finally in 2010, FDA did a post-marketing review of the last 40 years and found 29 case reports of methimazole-associated birth defects with use in the first trimester. And then in 1999, this term was coined, MMI embryopathy, that is just describes a constellation of congenital anomalies such as aplasia acutis, which is a scalp defect that you see there on the picture on the left, cranial atresia that causes facial dysmorphism, tracheoesophageal fistula, hypoplastic nipples, and sacromotor delay. PTU was not having a great year either in 2010. So the FDA reviewed data of the last 40 years and found an association with severe liver toxicity. So there were 34 cases of severe liver injury in adults and children, resulting in 15 deaths and 12 liver transplants. And acute liver failure was estimated to occur in one in 10,000 adults and one in 2,000 children treated. And that led to the black box warning to the PTU label in April of 2010. So shortly thereafter, two years later, we have the first large population study on investigating the association of exposure to these drugs in pregnancy with congenital malformations. And this study came out of Japan and looked at over 1,000 women treated with methimazole and over 1,000 women treated with PTU, as well as 2,000 controls that were hyperthyroid women that were not on antithyroid medication during pregnancy. And we can see in the study that the odds ratio of malformations with methimazole was about double that of either PTU or control women. And the fetal malformations that occurred were omphalocele, ophalobasenteric duct anomalies, and aplasia acutis. It is important to note how these authors did this study, this initial study highlighting this effect. This was done by giving a diagnostic questionnaire to the parent at the first visit postpartum that was typically about a month after delivery. And they asked the parent, is there any problem including any congenital malformations mentioned by a doctor with your infant? Yes or no? And the parent had to write down what that was. This was the way the congenital malformations were assessed in this first study. Subsequent to that, next year, Danish study, similar design, but assessment of outcome at two years of age through a registry, not through asking the parent. And at two years of age rather than a month post-delivery. So here we have about 560 patients exposed to MMI, about 1,000 exposed to PTU, about 3,500 without any ATD exposure in pregnancy but history of hyperthyroidism, and about 800,000 non-exposed reference group. And in this study, for the first time, we see also an association of PTU with congenital malformations. So they confirmed the increased risk in methimazole-treated moms, but they also show an odds ratio of 1.4 for malformations seen with PTU exposure. And the types of malformations seen were different. So in the MMI group, we have the classic malformations we described, but in the PTU group, it seemed to be more fascia and neck malformations as well as urinary malformations. With defects that were actually less severe and in some cases considered minor congenital defects rather than major. And later on, the authors tried to quantify kind of the frequency of these birth defects, especially compared to other feared adverse outcomes of these drugs. So we see here rates of adrenal cytosis and liver failure in the general population. So 16 cases per 10,000 ATD exposed for adrenal cytosis and three for liver failure. And in pregnant women, those rates are five cases per 10,000 exposed and five for liver failure, but 340 birth defects per 10,000 exposed. So that year, the ATA hyperthyroidism guidelines, the next version was published and they took this data into account and they have a section where they describe what they recommend for women with hyperthyroidism caused by Graves that is well-controlled on MMI and who desire pregnancy. And they highlight a number of different options. First, definitive therapy before they become pregnant, switch to PTU before trying to conceive, switch to PTU as soon as pregnancy is diagnosed, or inappropriately selected patients could withdraw from ATD therapy as soon as pregnancy is diagnosed. And if ATD therapy is withdrawn, then thyroid function should be assessed weekly throughout the first trimester and then monthly. So as we're thinking about when should we withdraw, when should we switch, I think it's important to take a step back and realize when formation of organs is happening. So typically, we consider weeks five to 10 of gestation as the period of maximum sensitivity to teratogens. The second thing that is important to note are the specific risk exposure periods for the various congenital deformations that we see with these drugs potentially. So for aplasia acutis congenita, actually the risk exposure period extends to 10 to 15 weeks, so a little longer than what we traditionally think. Quaynil atresia is four to 11 weeks, Impala seal three to eight weeks, and Impala benzenteric duct anomalies up to seven weeks. So taking into account all these data, the ATA guidelines for management of thyroid disease in pregnancy that were published in 2017 recommend to treat with PTU until 16 weeks of gestation if a woman needs treatment, and thereafter may switch to methimazole with a conversion of about one to 20. If switching, check TFTs at two weeks and then at two to four week intervals, and then realize that about 30% of women can come off antithyroid medication by the third trimester. And really, if possible, that should be a goal of treatment. Finally, they also recommend to consider discontinuation of antithyroid drugs during organogenesis in patients stable on low doses of MMI or PTU and with low TRAB titers with very frequent TFT monitoring every one to two weeks. So after that, a couple years after the guidelines, this study comes out. So this is a very large, even larger, Korean registry study with a cohort of about 13,000 women treated with antithyroid drugs, so pretty large. As you can see there, the breakdown, about 1,000 with methimazole, 9,900 with PTU, and they had a group that was treated with both at different times of pregnancy, so 1,800 with both. And compared to controls of, control women who do not have a history of Graves' disease. And we can see in this study, confirmation of the Danish study results with some increased risk for malformations with methimazole of 1.3 odds ratio, a slightly lower risk but still statistically significant increased risk with PTU of 1.16 odds ratio and 1.3 for the combination. Also this study had some other interesting findings. So switching from methimazole to PTU up to three months prior to conception in this study actually did not decrease the risk of malformations seen with methimazole. There was no association with PTU cumulative dose or duration of treatment, but there was an association with higher cumulative dose of methimazole with an odds ratio of 1.87 for a cumulative dose over 495 milligrams compared to under 127. And the number needed to harm was 91 for PTU and 46 for methimazole. So to summarize these three large cohort studies that have shown a positive association of antithiatic drugs with congenital malformations, Yoshihara was the first Japanese study that assessed kids at one month of age and found defects only with MMI. Whereas the other two studies that assessed children at 24 months of age for the Danish study and 12 months of age for the Korean study found defects with both. For the sake of completeness, I need to mention that there are a number of smaller cohort studies that have not shown an association of antithiatic drug use with congenital malformations, and you need to be aware that they are in the literature. There are a lot of limitations in these studies. The first one you can see from methimazole does not, a single one that has more than 130 cases exposed. So there are very few numbers of cases of exposure with methimazole there. The second is that most of the assessments, as you can see in the penultimate column there, were done, again, very early on, at a time when many malformations might not be evident, especially the more subtle ones that we see with PTU, so at one month of age. And also, these studies did not have this as their primary outcome. So these were studies done for other reasons that, by the way, they're like, okay, let's take a look at whether malformations were reported. And actually, the types of malformations that they document frequently actually don't even include the types of malformations that you would see, like the facial and ear malformations that you would see with PTU. So they don't even look at some of the outcomes that we would actually expect to find findings on. And finally, the CorelIT study, I'm just gonna highlight that a little bit more, because that has close to 1,000 cases of exposed PTU exposure and has assessment at 12 months. This one only looked at urinary tract defects among the malformations that we see with PTU, did not look at the facial ones. And in the urinary tract defect one, they did see an increased risk, almost double the risk, but it barely missed statistical significance of the p-value of 0.06. When you break it down according to gender and you look at the results in male fetuses, actually it was significant. And interestingly, in the Danish study, it was only male fetuses that had the urinary tract malformations for some reason. So I think this is an interesting observation. Finally, I'm gonna just mention that there have been numerous meta-analyses on this topic and I'm just highlighting here the last three that try to summarize the findings. The first two looked at associations of both PTU or MMI and confirmed associations of both with congenital malformations. The last one only looked at PTU and had much looser entry criteria, so they did not – and included a lot of the studies, again, that had the outcome assessed very early, like at one month of age, and therefore did not find an association. The interesting part that I want to mention is that the Morales study also showed that untreated hyperthyroidism carries a slightly increased risk of congenital defects. This is something of interest and I think this is something that needs to be explored further. Going back to some of the early historical data that I mentioned, this tended to be a belief that folks had even in the 1950s, and I think there have been some mixed data on our observational studies on this regard, and I don't think we have a definitive answer yet as to whether untreated hyperthyroidism poses any risk in terms of congenital deformations. So I think this should be resolved with further study of the topic. So is it time to reconsider the guidelines with respect to antithyroid drug use? I do think that the future guidelines on this topic will try to give some more detailed information as to when to consider taking off folks of antithyroid drug therapy, and I do think that it is important, if possible, to even postpone pregnancy in women for a while to get them to that phase where possibly we can safely have them off for that critical period of a few weeks in the first trimester. And finally, important to note what our goal, maternal free T4, is when we treat women with these drugs. So I just want to highlight this really old but relevant study by Momotani in the Journal of Medicine that showed that the risk of trans and neonatal hypothyroxine is actually directly related to the maternal free T4 at the end of pregnancy. So even if you have a maternal free T4 in the upper one-third of the normal range, you have a 10% risk of trans and neonatal hypothyroxineemia. So really our goal of treatment for these moms is always to keep the free T4 kind of as high as we can for the specific reference range that we're using, pregnancy-specific reference range, or even like borderline elevated. Okay, moving on to surgery. I was extremely impressed by the lack of data on thyroidectomy during pregnancy. This is the only paper I can find, 2009 retrospective paper, cross-sectional analysis of 165 women looking at outcomes with thyroidectomy during pregnancy and comparing to the non-pregnant patients. So in this paper, there was about a 5% risk of fetal complications, about a 5% risk of maternal complications. There was a higher risk of maternal combined surgical complications with an odds ratio of two overall if you combine endocrine and non-endocrine complications together. And there was a slightly increased risk of hypoparathyroidism and recurring nerve injury as we expect also as risks in the non-pregnant population. They do not provide a lot of granularity as to the specific risks of all the different outcomes we might be interested in. It's a paper that just really tries to make the point that overall complications are higher. And one of the main predictors, of course, of complications similar to the non-pregnant population is the number of thyroidectomies that your surgeon is doing, so the scale of the surgeon. The preparation is SSKI for 10 days, beta blockers, and maintenance of euthyroidism are achieving euthyroidism as much as possible at the time of anesthesia. And I just want to highlight this recent study that I thought was very interesting. This is not in a pregnant population, but it is a beautiful study from the Annals of Surgery from 2021 that looks at various outcomes in Graves patients after various treatments. And the exact details of each of these graphs are not important. I just want you to get the big picture of the trends here. So these are Kaplan-Meier curves for a variety of outcomes, mortality, cardiovascular disease, AFib, psychological disease, diabetes, and hypertension, and the x-axis is time. And in green, we see the outcomes in folks that had surgery, in red, folks that had RAI, and in blue, folks that had antithyroid drug treatment. And universally, you can easily see that folks who had surgery tended to have better outcomes across the board. In addition, the authors did a cost analysis to try to cost this out and show annual and cumulative cost of treatment. Again, in green is surgery, so higher upfront cost. Red is RAI, blue is ATD. So as the years go by, and the last bar there is like 10 years out, the surgical line, the green line, becomes lower. And at the end of 10 years, it is the lowest cost option, even though it does not quite reach statistical significance. So when should we consider thyroidectomy? So that is to be considered if there's high levels of maintenance antithyroid drugs that are required, such as PTU over 450 a day or methimazole over 30, if patient's non-compliant or patient's allergic to medical treatment. Surgery is safest in the second trimester for all types of surgery. And what is the role before conception? There are no studies that have looked at this yet, but I would argue that there is an increasing role of surgery before conception. We have iodine. So clearly we know this is contraindicated in pregnancy. If it's inadvertently administered in the first trimester, there's concern for increased incidence of miscarriage or congenital malformations, but really there's no formal recommendations for termination. There was one study that looked at the outcomes published in the literature after inadvertent exposure in the first trimester. This is a very old study from 1976. 3% of fetuses became hypothyroid, 2% developmental deficiency, and 1% of women had miscarriage. We need to be aware, though, that administration after 12 weeks can fully ablate the fetal thyroid. So concentration of iodine in the fetal thyroid is 10 to 50 times more than that of maternal thyroid. It's a very efficient concentrator of iodine. So any fetus exposed, we can expect, will have congenital hypothyroidism and may need actually intrauterine T4 if the pregnancy is kept. This recent paper, I just want to mention it. Those of you who are interested in this topic can look it up. It's very well written and describes kind of everything we know about radioactive iodine and potential effects on fetal development and what we might be concerned if we have a patient with exposure in preimplantation, organogenesis phase, or different fetal stages. So clearly we're not going to be using ever radioactive iodine during pregnancy, but what about as a definitive therapy prior to pregnancy? And the traditional guidance from the guidelines has been you can do that, but wait at least three to six months, maybe six months at least for your thyroidism to occur and the woman to be stable on her dose of thyroid hormone. However, I think another very important part that we need to remember is the effect of radioactive iodine on TRAB, on TSH receptor antibodies. And this beautiful study by Larberg on the left really shows the time course of TRAB falling after various treatments and shows nicely that with surgery and medication, TRAB rapidly declines and at one year post-treatment, the levels in most women are close to normal. Whereas with radioactive iodine, we have this huge peak of TRAB that occurs right after. And even at one year after treatment, you're back at where you were when you first gave the treatment. So thereafter, there's a much slower decline of TRAB titers. A very recent study on the right by Kim and colleagues showed similar results. This is in refractory patients, so patients who failed first-line treatment for Graves, and shows a slightly even longer duration of the peak of TSI with levels in the dark line which reflects radioactive iodine-treated patients returning to baseline only after 18 months or so. You see in the x-axis, whereas the surgical group rapidly declines. And another nice figure from the Larberg paper just shows the percentage of women that become TRAB-negative. So again, with surgery or medication, you can see that at one year about 60-70% of patients are TRAB-negative. With radioactive iodine, only 15% are TRAB-negative a year later. And from the Kim paper, another nice table shows that the risk of having clinically significant TRAB elevation, over 4.5 in this case, relates to the time after radioactive iodine treatment. So if you only wait six months, 82% of women will have TRAB elevation, significant TRAB elevation. 12 months, 67%, so still pretty high. By two years, it's down to 2%, that rises to 22% if they had a second REI. I want to mention also this really elegant study by Yoshihara that quantified the risk of neonatal hyperthyroidism directly after radioactive iodine and showed that the risks depend on time of conception after REI, so it's about 9% if you wait only 6-12 months after REI, falls to 5% at 12-18 months, and 4% 18-24 months. The TRAB over 9.7 in the third trimester was pretty sensitive for picking this up, and a TRAB under 28 in the first trimester was the most likely to decline to a safe level by the third trimester. So treatment for maternal graves is a balancing act, and we need to balance maternal and fetal risk in every single patient. I just want to mention that the recent guidelines, both the hyperthyroidism and the pregnancy guidelines, some really nice tables for you guys to consider that are very helpful clinically. This is table 8 from the pregnancy guidelines that highlights the advantages and disadvantages of all three treatments, and then from the hyperthyroidism guidelines, very nice recommendations detailing what to do if you have a patient diagnosed during pregnancy or previously diagnosed and treated prior to pregnancy, including a recommendation to withdraw ATD therapy as soon as pregnancy is confirmed with early testing if possible. And this study by Kobaly and Mandel identifies some of the factors that can predict successful withdrawal of anti-thyroid drugs in the first trimester. So low thionamide dose, low TRAB titer, normal TSH, treatment more than six-month duration, and no evidence of ophthalmopathy. A word about beta blockers, propranolol can be given safely if given for a brief period of time. Dose needs to be reduced as clinically indicated, and in most cases, you don't need more than two to six weeks of treatment. Long-term beta blocker uses associated with IUGR, fetal bradycardia, neonatal hypoglycemia should be avoided. And finally, a word about toxic nodules, which are quite rare, but you can have the occasional patient that will conceive with a toxic nodule. So you need to remember that in these ladies, there's a greater risk for fetal hypothyroidism with ATD because you do not have a TSI. So the TSI counteracts kind of ATDs in the Graves patients, but in these patients, all the baby's fetal thyroid is seeing is the ATD therapy. So you need to really use the lowest dose of ATD possible and monitor the fetus aggressively for goiter or hypothyroidism the second half of pregnancy. I really urge these ladies to have definitive therapy before they conceive so we do not have to deal with this risk. And avoid iodine-containing supplements. And I want to just close with a final figure of the kind of, I love this figure from Messman's paper that just shows kind of what a typical well-treated Graves mom's course would look like during pregnancy. So we see here the free T4 in the y-axis and time, gestational time in the x-axis. And we can see at the initiation of therapy, free T4 was about 4. PTU was started 150 to 450 a day, gradually titrated with close monitoring with a goal of keeping free T4 around 2 or close to that. We see a brief period of beta-blockade in the beginning. We see that the patient is able to come off antithyroid drugs shortly after 30 weeks. And then we see that the TSH in the bottom remains suppressed during pregnancy. In fact, actually, rising TSH or detectable TSH might be a sign that you've over-treated the patient. So we really don't care if the TSH remains suppressed as we, again, are keeping the free T4 level in a high range. So to summarize, where are we in 2022? I think there's an expanded role for PTU preconception. And from methimazole later in pregnancy, I do think we should make an effort to wean off both agents preconception or stop in the first trimester in a well-controlled patient. More studies are needed regarding the teratogenicity of PTU versus untreated hyperthyroidism, looking forward to that data, as well as to the effects, I should mention, of high free T4 in neurodevelopment. Surgery, I think we need to consider it much more frequently before conception than we are doing right now. Pregnant iodine is strictly contraindicated during gestation, and I believe it should be avoided in reproductive years, especially women who are planning a pregnancy in the next two to three years. And finally, remember to check TSI or tripe titers, if that's what you have, in select women. Thank you for your attention, and I think we have time for questions. Thank you. Go ahead. Thank you very much for an excellent review. So my question about the biotin and the prenatal vitamin, what do you do? The question about, I'm sorry? Biotin. Biotin. Yeah. So with biotin, I review the content of biotin and the prenatal vitamin that my patient is taking. If it's under 35, 50 or 35, there's no evidence it affects the TFTs. For higher amounts, I just tell them to hold it for two days before we do TFTs, and then you can trust the results. It gets rapidly cleared by the urine. Yeah. Oh, sorry. Go to this mic. Yeah. Hello. My name is Khaled from Qatar. Thank you for a nice presentation. My question for those ladies, you wean them during a pregnancy from ATD. How do you follow them after delivery? Do you re-institute the medicine again? Do you do investigation to consider re-institution of the medicine? Could you repeat the question? Patients with... Yeah. Those ladies, you weaned them from ATD during the pregnancy. Yes. After delivery, there may be flare-up in the antibodies. Yes, you're absolutely right. So I follow everybody very closely after delivery. That's the peak time of flare of Graves' disease. So I follow them the way... I don't follow them any differently than I would follow everybody else with the history of Graves'. Right? So even if they come without anti-therapy drugs controlled, I still follow them the same way postpartum. So I usually check them at six weeks. And then if that is completely normal, three months and then six months postpartum. So I follow everybody pretty closely with labs. Yes. Go to the mic. Great talk. Hello. My name is Sanam Latif. I'm an endocrinologist at Mount Auburn Hospital in Cambridge. I had a question about how you wean PTU off pre-conception. Say someone's on a very low dose of PTU, what is the TRAB cutoff and TSH cutoff that you would use to say, you know, we wean them off and then follow them? So that's a great question. So before conception, they have to be stable for at least ideally three, but I say six months with a very low dose. So I would say, let's say, let's say there are 25 BID of PTU and a negative TSI, completely normal TSI. Otherwise, I keep them on the low dose and I tell them that as soon as you're pregnant, you stop the medication. So if I think they are, if they're on 50 or lower of PTU, and sometimes even on 75, depending on how their trend has been, and like it's a clinical judgment call, but definitely for 50, I tell them when I, when they call me, I said, let's stop the medication. I monitor them every one to two weeks with TFTs. I have been extremely successful having that window of time, even if it is from five to 10 weeks of gestation that you can avoid exposure. I think that's fantastic. And I have been, I have not actually had to restart anybody. Great. Just to add a follow up. So I have a patient whose TRAV is positive, it's like less than two, and the patient's only on PTU, 25 milligram, four days a week, and I'm attempting to try to get a TSH, which was less than 0.015, to what I consider would be a normal range, less than 2.5 prior to conceiving. Is there any avenue to try to stop her? So if they have a positive TRAV, I keep them on it until they conceive, but I, even if the TRAV is that low, I feel comfortable stopping it once they conceive. So I have them like, the minute you call me, and if your TRAV is like 1.8 or whatever, even though it's above 1.3, which is a cutoff, then I have them stop it in such a low dose and monitor them very closely. Because remember, TRAV will fall actually with pregnancy. So two is not going to become, you know, four or five. Thank you. Yeah. Welcome. Thank you for a great talk. My name is Dalal. I'm from McMaster University, and I have a question here. I find the radioactive iodine preconception attractive option for therapy, but do we have any concerns on radioactive iodine on fertility, as was reported in thyroid cancer patients when it affects the anti-malarian hormone level? I'm glad you asked that, because I had a slide on that, but because of time, I took it out. So this has not been extensively studied, but one of the studies that showed the effect on fertility for thyroid cancer had five patients with Graves' disease included, and showed in the N of five, no effect on AMH, which is very reassuring. That's one of the three or four studies that looked at this, is the only one that had hyperthyroid treatment doses, no effect on AMH. That's all we have. Thank you. But it's an area for further investigation. For now, we don't have data to support that, that this is an adverse effect. Yes. Last question. Hi. This is Fatima from University of Arkansas. Thank you so much for the great talk. I had a question about subclinical hyperthyroidism with a thyroid nodule. Would you do any thyroid labs during the course of pregnancy? Yes. So subclinical hyperthyroidism, I do follow during pregnancy. I don't treat it, but I do follow to make sure it does not become overt. If it's a thyroid toxic nodule, I don't follow as closely because I don't think that will change dramatically. I have not had something become overt as time goes on. Interestingly, actually, autonomous nodules, think about it, because the need for thyroid hormone goes up in pregnancy, you tend to have amelioration of their symptoms because they don't bump up. They don't increase the nodules that doesn't increase as much as the rest of the gland. So because you have more need for thyroid hormone, you actually tend to become better as pregnancy is progressing. So I still check them periodically, but not very closely. So I would say maybe every four to six weeks, TFTs. Thank you. Unfortunately, I think that is our time. Thank you very much for a great talk. Thank you all for coming. Enjoy the conference.

hyperthyroidism

pregnancy

thyroid physiology

Graves' disease

gestational transient thyrotoxicosis

thyroid-stimulating immunoglobulin

antithyroid drugs

surgery

radioactive iodine

monitoring