Hello, and welcome to Monitoring Thyroid Function in Critical Illness. The devil is in the details. My name is Michael Agus, and I'm a pediatric intensivist and endocrinologist, and I'm the chief of the Division of Medical Critical Care at Boston Children's Hospital. I'm sorry we couldn't be together in person in Puerto Rico, but this is a close second best. I have no conflicts of interest to report. I do have some ongoing funding from NICHD, as well as some internal support, none of which represents conflict. When we think about critical illness, this is kind of the classic graphic that we use, and there are many hormones which, as we get sick, they begin to drop appropriately, and as we get better, they begin to rebound, sometimes over the normal range as we get back to normal. The specific overlay of thyroid function on this general curve is broken out into the different hormones that compose thyroid hormone, that compose the thyroid axis. The one I was showing you initially was the TSH, where the TSH drops and then surges above normal to reconstitute normal thyroid function as the patient improves. Total thyroxin, or T4, tends to drop in the height of critical illness, not quite below the normal range generally. Free T4 often will drop a little bit below the normal range. Total T3 will go even further below. Where does all that T3 go? If you have a certain amount of T4 and you're making less T3, well, it has to go to reverse T3, which is the choice, the metabolic choice made as T4 is processed. It either goes into T3 or reverse T3. We have a surge in reverse T3, and then we have these binding indices, which in some hospitals are called the thyroid binding globulin index, others are T3 resin uptake, but we'll talk about those. Each of them has their own curve, and it's important to think about what moment in time we're looking at these thyroid function tests and how that relates to the overall course of the patient. The most extreme version is in my world of pediatrics, and in particular in the conversion of the gestational state to the postnatal state, where we have an increase in TSH at the time of birth to right before birth in the eight or nine milliunit per liter range, but once we expose the baby to the elements, in particular cold, the TSH will often go as high as 70, and that is thought to initiate the conversion of a large amount of T4 into T3, whereas reverse T3 has really been dominant during the greater part of intrauterine life As the baby is exposed to the world with cold and light, TSH surges and begins to convert T4 to T3 as the baby becomes responsible for maintaining its own hemostasis and thermoregulation. As I mentioned, T4 can go one of two ways. It can either be deiodinated by type 1 or type 2 deiodinases to T3, or it can be deiodinated by type 3 deiodinase to reverse T3. And therein lies all of the differences that we see where normal functioning people will have fixed relatively normal ranges of T3, but in the setting of critical illness will dramatically increase their reverse T3 while the T4 gets preferentially converted in that direction. Thyroid hormone has a range of actions, in part depending upon whether we're talking about T3 or T4. T3 is more active at the thyroid hormone receptor than T4 by a factor of 100 to 1,000. And of course, that's why it's present in the serum at much lower concentrations. The half-life of T4 is approximately a week, whereas the half-life of T3 is measured in hours. We know that the effect of thyroid hormone is to increase the resting energy expenditure and related to that, it increases temperature, CO2 production, and heart rate. All of those are very important in the monitoring of a critically ill patient. And it decreases systemic vascular resistance, which is particularly important in a cardiac patient as reduction and afterload can be a very important part of thyroid function and the absence of thyroid can increase afterload. It acts as a CNS stimulant, which is mostly obvious when it's absent and patients are abundant in the extreme state. And of course, it potentiates normal hepatic protein handling, which in particular can affect coagulation such that a severely hypothyroid patient may have abnormal clotting, which may be particularly relevant in the perioperative period. When we think about measuring thyroid function, one of the older ways to do it was with this T3 binding index that I mentioned earlier. We would measure total T4, total thyroxin, and one of these binding indices. The basic idea of these tests is that you put a resin into a test tube that has both thyroid hormone from the patient and some radiolabeled thyroid hormone. And then you pull out the resin, which binds both the native thyroid hormone and the radioactive, and you count to see how much thyroid hormone you have extracted on the resin. If there is very little thyroid hormone, endogenous thyroid hormone around, then you will extract very little of the radioactive because all of the radioactive will have bound to the available binding sites, whereas if there's an excess of endogenous thyroid hormone, when you pull out the resin, there will be more of the radioactive because there were much fewer binding sites for it to bind to and wouldn't have been available to the resin. It's complicated to think about for sure, and thankfully in this day and age, we really focus on the free T4, the free thyroxin. If you consult your local endocrinologist and show them that you measured a TSH and a T4, they will say, I can't help you, you either need to get a free T4, or if you insist on getting the total T4, you need to combine that with one of these indices of binding because we can't fully interpret a set of TFTs without understanding the binding. The free T4 is that direct measurement of the unbound thyroxin, and you don't have to worry about the binding. In the usual kind of classic states of primary hypothyroidism, free T4 will be low, TSH elevated. Hyperthyroidism, free T4 is elevated, TSH is suppressed. TBG deficiency, which was more relevant when we used these binding tests because you see abnormalities, but TSH is normal. Here free T4 should remain normal and TSH will also remain normal. Same thing in TBG excess, the body just overproduces T4 so that the amount of free hormone is preserved with a normal TSH. So the big take-home point here is if you measure a total T4, you need to get an index of binding along with that. The next kind of test to just be aware of are tests where the free T4 may give you a false reading, and I think one of my co-speakers is going to talk about some medication effects. But the answer to those approaches to measuring in those scenarios is to measure directly, either with equilibrium dialysis, which is the most common, or centrifugal ultrafiltration, which will achieve the same idea. I'm not going to go into the details, but suffice it to say that these are more expensive, more time-consuming, require a very high-end lab to get done, and so often it's going to end up being a send-out. It involves mass spec measurements of the final free hormone that is pulled out by the equilibrium or by the centrifugation. In all these cases, it's really not affected by changes in binding protein concentrations or presence of antibodies or even non-thorough illness. So it can be a very effective way to measure. The problem is the cost and the time to do it, and so we reserve that for very specific circumstances where we have concerns about the validity of the usual free T4 and TSH. So the practical guide here is to just remember that those two go together. TSH and free T4 are far and away the first set of tests that we look at in a critically ill patient. When we see abnormality and we want to look a little bit more, we might measure reverse T3 and total T3. Now reverse T3 is generally a send-out, so it may be several days before you get your answer, except in very special circumstances, and the total T3 is often an in-house test, but not necessarily. But again, that will give you some sense of conversion of that T4 to the active hormone, where if we're worried about non-thorough illness or the hypothyroxinemia of critical illness, we would expect to see an elevated reverse T3 and a suppressed total T3 that's suppressed a little bit beyond the free T4. Notice here we use the total T3 as kind of the next most stable measurement of regular T3. It is possible to measure free T3, but it's a much less reliable test, and so we kind of put it at the bottom of these endocrine additions, I termed it. These are tests that you're often really going to order in consultation with your endocrinologist. You might think about thyroglobulin antibodies, thyroproxasase antibodies, to look at indexes of degree of thyroiditis, which may, in extreme cases in the beginning of thyroiditis illness, for example, may leak out free T4, and so you have an elevated free T4, and it also might explain a low free T4 in the setting of inflammation that has progressed for a while and actually now decreases synthetic function of the gland. These next two antibodies are associated with Graves' disease, TSH receptor-stimulating immunoglobulins, or the TSI, or just kind of the more generic TSH receptor antibody, which is antibodies that bind to and thereby block the TSH receptor in certain circumstances or activate the TSH receptor in other circumstances, which will lead to over-stimulation of the thyroid gland. So in both of these cases, TSH receptor doesn't distinguish between blocking and activating TSH-stimulating, or TSI only represents activation at the TSH receptor. It's usually a cyclic AMP assay using a particular kind of cell, generally rabbit cells. Other ones we think about are the total T4, which I mentioned earlier, we're only going to measure in the setting of a binding index measurement as well, and maybe you'll measure actual TBG, thyroid-binding globulin. But again, these are in the category of not standard tests that we're going to order at the beginning of a workup, rather we're usually going to discuss with the local endocrinologist. I thought I'd run through a few patients just to show how some of these tests play out. I'm a pediatrician, so I'm going to give you some very quick pediatric cases, but the lessons are quite relevant across the age range. In this case, a newborn was born to a mother with Graves' disease. He was born with tachycardia, suppressed TSH, an elevated total T4, an elevated total T3, and THBR is a binding index that goes with that total that's also elevated. This is a classic set of overactive thyroid glands, and this picture is to demonstrate the importance of the physical exam where we hold the baby back in our hand to expose the neck and to look for thyromegaly. Neonatal Graves is uncommon even among babies born to Graves' mothers, but 1% is non-trivial, and so we do watch for it in a baby born to a mother with Graves. It can last up to 3 months, which is essentially as long as the half-life of the antibodies transferred by the mother. As we talked about, the TSI or the TSH receptor antibodies are the two measures of antibodies in Graves. The second patient was an interesting one when she showed up. This is a 19-year-old who was non-adherent to her antithyroid medications and came in hyperthermic, tachycardic, a little bit hypertensive and systolic, but actually with a low diastolic and as we mentioned, thyroid hormone, T3 in particular, reduces the systemic muscular resistance. She was jittery with lid lag, thyroid brewy, hyperreflexic, and wasted due to the increased metabolic rate that she simply could not meet with her dietary intake. TSH profoundly suppressed, total T4 elevated, total T3 dramatically elevated, and THBR, one of these binding indices, also dramatically elevated. Thyroid storm we know is associated with hyperthermia, CNS activation, some in terms of agitation and some to the extent of psychosis, lethargy, seizures and coma, tachycardia that can certainly lead to high output failure, GI and liver dysfunction, and is actually associated with pneumococcal sepsis in particular, interestingly. We use beta-antagonism therapy, we use beta-antagonism, we used to use propylthiouracil in children, but with the incidence of liver failure, we no longer use it, and we stick with methimazole at this point, which also has its own side effects that we watch for in terms of agranulocytosis and liver failure, among other things. We also use iodine in the acute setting to block thyroid hormone release. It can decrease vascularity, which might be useful as we move to resecting the thyroid gland, but it only has a limited effect, and after several days, it can begin to stimulate increased thyroid hormone production, so it's only good for short-term use. It's similarly have a limited impact in terms of temporarily suppressing conversion of T4 to T3. One common, well, not common, but occasional and dramatic demonstration of excess T4 is in preparation of meat. If instead of restricting to the body of the cow, they continue to make ground beef out of the neck, you can include thyroid gland and cause thyrotoxicosis. The way we use these labs in newborn screening in Massachusetts, every state has their own algorithm, is to look at T4 first and look for a low T4. If it's present, then go ahead and measure a TSH and see if it is over the normal range, which is different for each 24-hour period after birth. The reason I show this is just to demonstrate that over time, certainly in children, the ranges can change very significantly as thyroid function is taken over from the mother, now by the baby. And then the third patient, very briefly, was also just a fascinating patient who came in with known autoimmune hypothyroidism, but had severe abdominal pain, a suspected small bowel obstruction, and was really minimally responsive. She had little hypothermic and bradycardic, a little bit of an elevated diastolic blood pressure, a firm thyroid gland, and delayed return on her reflexes, and had dramatically elevated TSH and a low total T4, free T4, and total T3. In this case, due to the concerns about potential for myxedema coma, we thought about and ended up choosing to treat with both T4 and a T3 continuous infusion. This is important before you go to the operating room for your small bowel obstruction, because hypothyroidism will affect your ability to clot and to process anesthetics. And so we started her on these medications overnight, and in fact, by morning, she was symptomatically entirely resolved, got up, and went home. So these are just some examples of how the thyroid tests guide our therapy, and I'll kind of end where I started by reminding you that where we measure these tests is just extremely important. If we measure it early on, we might have an elevated T3, sorry, an elevated reverse T3 if we were measuring it, but really what we'll see is a low T4 and low TSH and a low total T3. And usually the right thing to do, if they're not extreme, is to repeat them a week later and look at the trajectory. Trajectory between these two points may be not too much of a change, but if you get it as the patient's getting better, all of a sudden, TSH can be abnormally elevated, and free T4 can be low normal or even frankly low if it hasn't begun to rise yet. As we check the week later, the TSH might be even higher, and free T4 is just beginning to come into the normal range, and it's not until we check right before they are ready to be discharged from the ICU that we see an improvement. And so the real story here is to know what you're measuring, to understand the context as how these various hormones are reflected in clinical symptoms, and then to interpret the tests really over time. There's very few ICU thyroid tests where we make the diagnosis at the first set of tests. So hopefully I've been able to give a little bit of insight into how we think about assessing thyroid function in the ICU. Thank you so much.

thyroid function

critical illness

hormonal levels

TSH

thyroxine

T4