Hello, everyone, and welcome to the session, Cyclic Cushing Syndrome. I am Adriana Iacomescu, professor of endocrinology and neurosurgery at Emory University in Atlanta, Georgia, and I will be serving as your moderator today. It is my absolute pleasure to introduce our distinguished speaker today, Dr. Glynette Nieman, a senior investigator at the National Institute of Health in Bethesda, Maryland, and the head of the endocrine consult service at the institution. Dr. Nieman is an active clinical investigator, and among her research interests are those of cortisol excess, which is what our session is about. Dr. Nieman has authored more than 280 publications and sponsored three investigational new drug applications, one of which was licensed in both U.S. and Europe. Dr. Nieman has received numerous awards, including the NIH Director's Award, NIH Clinical Teacher Award, and the Endocrine Society Distinguished Physician Award. Dr. Nieman was the president of the Endocrine Society in 2017. With that, I would like to welcome Dr. Nieman for her presentation. Good afternoon. My name is Glynette Nieman. I'll be speaking to you today about cyclic cushion syndrome, and I'd like to thank the organizers of the meeting for inviting me to this Meet the Professor. The objectives of this session are as follows. I hope that when you finish this, you'll be able to recognize patients who should be evaluated for cyclic cushion syndrome, to describe diagnostic approaches to identify those patients with cyclic cushion syndrome, and to be able to plan a successful evaluation for the differential diagnosis of patients with this condition. I have no disclosures. This is what we're going to talk about today. First, I'd like to define cyclic cushion syndrome, talk a little bit about clinical and laboratory presentations, and even more in depth on laboratory presentations. I want to show some patients, some actual patients, and describe them, and challenge you to think about how you would manage them. I want to then talk about strategies to diagnose this condition, approaches to the differential diagnosis as well. Let's start with the definition. This slide is primarily taken from the reference below, which was a compendium of a number of case reports. It was a meta-analysis of sorts of case reports in the literature. Dr. Mainardi showed that and defined this as periods of intermittent hypercortisolism alternating with periods of eucortisolism. What they found was that the eucortisolism duration was quite variable from 1 to 2160 days, but the median was about 30 days, and that the hypercortisolism duration was a little shorter at 21 days median, but in a shorter range of 3 to 510 days. The periods, the predictability of the periods is not very good. They're usually unpredictable, but they may be regularly spaced. The onset and offset, so the beginning of hypercortisolism and the return to eucortisolism may be either gradual or rapid, and I'll refer to this rapid change as a square wave, and it can be as rapid as today I'm normal, tomorrow I'm 10 times normal. You'll see in the literature that cyclic Cushing syndrome is also referred to as intermittent or periodic Cushing syndrome, but very importantly this should not be confused with variability in hypercortisolism severity. So we know that people with Cushing syndrome vary quite a bit in how hypercortisolemic they are, but in general people with Cushing syndrome do not have periods of normal cortisol levels or eucortisolism. So we're going to be talking about the people who have high levels interspersed with normal levels. And here's a slide that shows you some examples of the temporal patterns of alternating hypercortisolism and eucortisolism. This is taken from a paper that looked at 10 patients who presented to a single site over one year and were suspected of having cyclic Cushing syndrome. In the end of the 10, only three were found to have cyclic Cushing syndrome, and the patterns of these three patients are shown on this slide. What you're seeing on the y-axis is the cortisol to creatinine ratio in urine that was provided as part of the first void each day against the salivary cortisol on the other y-axis, which was the salivary cortisol at 11 p.m. So we're looking at 11 p.m. and then a morning urine cortisol. And what you can see here is the salivary cortisols in the dashed lines are quite remarkable in terms of showing variability, and the normal range is shown by the dotted line. You can see that some one of these patients has very few urine cortisols that are normal. All salivaries are abnormal. Here's another patient with all abnormal salivary cortisols and just a few normal urine cortisols. So these were all considered to be cyclic because primarily the urine cortisols showed this temporal pattern of normal C. And when we get to this later, we'll recognize that this is probably related to the fact that salivary cortisols are required to be high in order for urine cortisols to be high, but that salivary cortisol is one point in time that you can think of it as part of a pulse, and so that you might have a normal salivary cortisol but abnormal urine cortisol and vice versa. In the same paper, the patients were subdivided as to the cause of Cushing's syndrome. You can see that the largest proportion was almost half or a little more than half had pituitary Cushing's disease. About a quarter had ectopic, and the other quarter were divided between adrenal and unclassified patients. As we know, most patients with ectopic ACTH secretion have neuroendocrine tumors of one sort or another. What may be somewhat striking, although the numbers are small, is that the higher proportion of thymic carcinoids compared to bronchial carcinoids. Generally, there are many more pulmonary carcinoids or pulmonary neuroendocrine tumors than there are thymic in terms of those tumors that secrete ACTH. This was just a case series, as I showed you. I went back to look at our patients over a five-year period or so of time when we were using a urine-free cortisol immunoassay with an upper limit of 90, shown in the dotted line here, and in looking at these patients and determining whether they had at least one normal urine-free cortisol, I found that 10% of our patients in that period of time did fit that definition. That gives a sense of how often this occurs in a large case series that is from a single institution, albeit a somewhat biased referral pattern at the NIH. Here's another study conducted in the UK looking at 201 patients, again retrospectively looking at patients with cyclicity, and they found that 13.4%, so not very different from the NIH group, had cyclicity defined as two abnormal values separated by one normal value. There is some discussion in the literature as to whether you should have two sets of two abnormal values separated by a single normal value between each of them in order to define a true cyclic Cushing syndrome. This is a less conservative definition. The length of these cycles were 3.8 years with quite a large range, as we saw in the previous slide, and when they looked at the cycles themselves, they had 54 individual cycles, and they defined eight of them, or about 15%, as being primarily clinical, and some being clinical and biochemical, and some being primarily biochemical without clinical features. So what are these clinical features? Well, patients can present a large number of ways, and when you think about the fact that sometimes they're hypercortisolemic and sometimes they're eucortisolemic, that sort of explains why we have so many different kinds of presentations. So someone could come in and look Cushingoid and have symptoms but have normal labs. They may have Cushingoid appearance, no Cushingoid symptoms, but rather GI symptoms and hypocortisolism. They may have a history suggestive of a very distant Cushing syndrome, but they look very normal and have normal labs. They might come in, and as you follow them, they have a variable appearance, they have variable symptoms, and variable labs, and when you speak to them about the history over time, they may be very vague about what changes occurred, or they may be very clear it was an extremely abrupt change, and it's important, very important when taking the history to get a sense of that. So when you take this history, you want it to be very detailed about the onset of signs and symptoms, whether they wax and wane, if there are any periods where everything just sort of magically went away and a patient felt normal again. Again, from the same paper, review paper, this is just the clinical characteristics of these patients, and I would not put too much weight on this because this is not from a single institution where you would expect to see the usual proportions of patients, but rather this is the case series, and of course not all clinical features would have been sought and reported, and so we have to take this a little bit with a grain of salt, but what I think we should point out is that with the adrenal patients, they seem to be younger, although that may reflect primary pygmy metanodular adrenal disease, which is a younger population, and in the pink rectangle here, you can see that the proportion of patients with mood disorders seems to be lower. Acne is about what we usually see. Hirsutism perhaps a little higher, and amenorrhea lower than what we usually see, so these patients don't necessarily present with the same historical clinical features of Cushing's as what we see with patients with consistent hypercortisolism. So how about more on the laboratory presentation of these patients? The patients with the square wave onset tend to have ectopic ACTH secretion with very severe hypercortisolism and hypokalemia, and that hypokalemia can alert you to the fact that they are entering a new cycle, and they will often talk to clinical features associated with that, such as muscle weakness, and so you can ask them to report back and get labs checked when they develop that muscle weakness. The UFCs in the late night salivary cortisol might be normal. If they're eucortisolemic, they might be mildly elevated or they might be very severely increased, so this spectrum is anything that you can think of may occur with these values. Hyperglycemia and increased white count, as we see with Cushing's syndrome, may be present intermittently, and that may help you to recognize that this patient has intermittent cycle of Cushing's, and here is a nice graph from the paper where the 10 patients were followed. This shows the values from all of those patients, and what we have here is the log of salivary cortisol on the y-axis, the log of urinary cortisol with creatinine ratio on the x-axis, and you can see that there's a pretty good correlation overall with our value of 0.79. When we look at the group as a whole, you can see some patients have good correlation with a high salivary cortisol and a high urinary free cortisol, a good creatinine ratio, whereas other patients have a high salivary cortisol and a normal urinary free urine cortisol, and others have a high urine cortisol and a low salivary cortisol, just to say that you may not find perfect correlation between all of these, between these two parameters. I'd also like to point out that the regression of the non-cycling patients, the patients who of the 10, the 6 who did not have Cushing's, had Cushing's syndrome but not cycling, was a much higher regression than those who were cycling, and that may well be because the people may have cycled out from one day to the next so that one number would be normal and the other one abnormal. So let's talk about some patients that might present to us. First of all, a Cushingoid-appearing patient comes for evaluation. The UFC's late night salivary and one milligram overnight dexamethasone suppression test are normal, and as we go through these I will come back to them and talk about what you might be thinking about, but you might think as we're, I'm talking now about how you would approach a patient like that. There's a second patient who comes for evaluation. The urin-free cortisol's late night salivary cortisol and one milligram dexamethasone suppression tests are also normal, but there's a recent history of nausea, diarrhea, and dizziness. And here's another patient who comes who's diabetic and referred because of waxing and waning moodiness, weight gain, and hypoglycemic episodes, and when she comes through she does not want to have phlebotomy for fear of getting COVID-19 while she's waiting to get her laboratory testing done. And here's the final patient. We'll discuss who comes for evaluation of a one-month weight gain of 10 pounds, new hypertension, with extremely elevated laboratory results. ACTH is twice normal. IPSS a few weeks later is quite abnormally normal, given her previous values, showing a peripheral ACTH value of 25 picograms per ml with a peak at 35 and pertussal levels that are quite low for pertussal ACTH values of 55 peaking at 175. So in thinking about these patients, what we have to really get back to is what does the hypothalamic pituitary adrenal axis look like when people are normal and what does it look like when people have Cushing's of the various sorts. So let's go through this. I know all of you know it, but just as a refresher so that we have this in our mind as we're thinking about those four patients. So here we have the normal HPA axis with CRH being produced by the hypothalamus going and stimulating the pituitary corticotrophs to make ACTH. ACTH in turn then stimulating adrenal production of cortisol. And what's extremely important about this is the negative feedback effect of cortisol to inhibit CRH and ACTH, helping in part to give us the normal diurnal rhythm that we all as normal people experience. So what about when we're stressed? Well, something is happening above the hypothalamus and we think, although we have no absolute proof of that in humans, that CRH is increased, then causing an increase in ACTH as depicted by this thicker line, which then causes an increased amount of cortisol production. But very importantly, in normal human beings, this negative feedback restrains this response to whatever that stressor was, so that the ACTH production is sort of caught in between the push of the CRH and the restraint of the cortisol feedback. And in general, these patients have urine free cortisol values less than three times normal. So let's then turn to the people who have Cushing's syndrome. What does the HP axis look like? Let's start over on the left here with Cushing's disease. Here we have an ACTH producing pituitary tumor making a lot of ACTH. Again, cortisol production is increased. There is some restraint of that tumor by excessive cortisol production. But what is really primarily restrained is the normal CRH and ACTH secretion, so that this line coming from normal corticotrosis is very thin. When we have ectopic ACTH production from a non-pituitary tumor, we have the same high ACTH and high cortisol with negative feedback and also inhibition of ACTH secretion. And with autonomous adrenal tumors, they have the independent secretion of cortisol without requiring ACTH stimulation. So ACTH levels are quite low because they only reflect what's coming out of the pituitary, which has been suppressed by the high cortisol values. So what is consistent amongst all of these causes of Cushing's syndrome is that ACTH levels from the corticotroph are low. Measurable ACTH levels may be quite high or normal in the ACTH dependent forms, but the corticotroph is suppressed. And it's suppressed because of long-standing hypercortisolism. With very short-standing hypercortisolism, such as when we give a medrel pack, as you know, patients will not have suppression of their normal corticotrophs. That's what we have to keep in mind as we're thinking about people who are cycling in and out of high and low cortisol values. So let's put the normal person into the mix here. We've just talked about the causes of Cushing's syndrome. The normal person will have a higher ACTH value than what's coming from the corticotrophs of the patients with any of the causes of Cushing's syndrome. So when we look at the circulating levels, the normal person will have normal levels. This person on the right with the adrenal causes will have low or very suppressed levels, and the patients with the ACTH dependent Cushing's will have higher ACTH levels. So that's the backdrop on which we're going to think about what happens with dexamethasone. So in a normal person, we are going to have the negative feedback of dexamethasone acting similarly to cortisol with a decrease of ACTH. I'm sorry, my arrow has dropped off here. And in patients with all of the causes of Cushing's syndrome, this dexamethasone is not going to inhibit at least a one milligram dose. There's not going to be much of a change. So in normals, we expect cortisol to go down. In people with Cushing's syndrome, we expect the cortisol will not go down with one milligram of dexamethasone. However, remember if these corticotrophs have not been inhibited sufficiently, the patient who may have cyclic Cushing's syndrome will behave like a normal person and will have inhibition of ACTH and cortisol secretion. So let's think about late-night salivary cortisol and how that might relate to the diagnosis of cyclic Cushing's syndrome. This is a slide depicting the normal scenario where people sleep usually at night and we know that the cortisol rhythm is very tightly entrained to sleep so that salivary cortisol and serum cortisol levels reach their nadir slightly after the onset of sleep and as you know they rise so that we have this kind of diurnal rhythm of ACTH and cortisol and that rhythm because of the dysregulated or unregulated ACTH and or cortisol response and the various causes of Cushing's syndrome is really lost and so that the late-night salivary cortisol is a really good reflection of how of the status of a patient in terms of how hypercortisolemic they are. So let's go back again to our patients. Here we have a cushionoid-appearing patient who comes for evaluation. The urine-free cortisol is late-night salivary cortisol. One milligram dex tests are normal. So in this setting what are we thinking about? Who is coming in and has a cushionoid-appearing look to them and perhaps some signs and symptoms. However, their laboratory tests are all normal and so you're going to be wondering is this a normal person with some features consistent with Cushing's syndrome or is this someone with cyclic Cushing's syndrome. Now if it's a normal person with some features consistent with Cushing's syndrome that person is not going to have features of Cushing's syndrome that are really classic for severe Cushing's syndrome such as purple stretch marks or proximal myopathy. If you have a patient of that sort with very severe signs and symptoms of Cushing's you're very much going to be thinking that they have cyclic Cushing's because there's not a very good explanation for having those other more severe signs and symptoms. But if you have a more mild phenotype that's consistent with Cushing's and normal labs then you have to really think about cyclic Cushing's as perhaps not the ultimate diagnosis but this might just be a normal person. So here to try to make this distinction you're going to ask again about careful history, asking about signs and symptoms, the complete spectrum, when they started, were they consistent, did they increase or decrease or go away completely, did more accumulate over time. The accumulation of signs and symptoms is very consistent with Cushing's syndrome. Is there a change in cognition with the ability to synthesize information, is there a change in mood or memory. Short-term memory as you all know is impaired in Cushing's syndrome. Emotional liability as Monica Starkman showed many years ago is one of the most consistent features of Cushing's syndrome so it's really important to ask about that because all of this will then help give you a better sense of the pretest probability of the patient having Cushing's syndrome or not and may guide your decision as to how much testing you're going to do, how much you're going to really pursue a diagnosis of Cushing's syndrome. So here's our second patient, again another Cushingoid appearing patient who's coming who again has normal laboratory results but there's a recent history of nausea, diarrhea, and dizziness. So in this patient you're going to be asking yourself is this a normal person with some features compatible with Cushing's or is this someone with cyclic Cushing's and intercurrent GI illness or is this somebody with cyclic Cushing's syndrome and adrenal insufficiency and of course those are really important distinctions to make particularly if the patient has adrenal insufficiency because you would like to consider whether or not that needs to be treated. So again for all of these patients history is so important but in addition to the history we just reviewed you're going to ask about other signs and symptoms of adrenal insufficiency such as joint aches and pains, especially lower back pain, real fatigue, really debilitating fatigue, perhaps headaches, and weight loss. Weight loss obviously is a very discriminating feature if it's present to suggest adrenal insufficiency and in the context of weight gain, previous weight gain that may have represented Cushing's syndrome. Here is another patient, this is the diabetic who's referred because of waxing and waning moodiness, weight gain, and hypoglycemic episodes and who doesn't want to have phlebotomy because of worrying about getting COVID. Again the history is extremely important. What you want to think about and ask about is is there a correlation? Does she notice that her weight gain and diabetes came at the same time but she stopped gaining weight when she was having hypoglycemia and she lost weight? What about glycometer readings? Is there any documentation that you can use to look for some sort of a pattern and are there any other signs and symptoms that come and go that might be more characteristic for Cushing's syndrome? So you're really trying to build up in your mind a sense of what's coming, what's going, what's getting worse, are more and more symptoms coming to decide. Again the big decision here is how much workup are you going to do and how much you're going to pursue this. So then our last patient comes for evaluation after one month weight gain of 10 pounds and new hypertension. The urine-free cortisol is late night salivary cortisols and one milligram dex suppression tests are extremely elevated. ACT is twice normal but the IPSS a few weeks later has peripheral ACTH levels of 25 so quite normal ACTHs peaking at 35 so sort of perhaps even within the normal range response. Certainly not the response we expect to see with patients with Cushing's disease and the pitrose levels are 55 peaking at 175 so even though there is a borderline three-to-one response we also know that normal individuals have responses to CRH in IPSS testing so we can't say this makes the diagnosis of Cushing's disease. So again what are we going to ask about? Very important to ask about the history. Has she had this kind of an episode before? Did she gain weight? If she had any other symptoms with her hypertension such as headaches did she have that before? Was she weak during the time that she had that weight gain? Were there any other signs or symptoms and does she have any recent signs and symptoms of adrenal insufficiency? The reason why you're going to ask about that is if she had very severe hypercortisolism for one month she may have suppressed her normal corticotrophs and then the appearance of adrenal insufficiency after her values would decrease would suggest that that was really true Cushing's syndrome and not something else you're trying to sort out. You know in reality when you can see a patient like this if these numbers are truly so elevated and you've got the correlation between urinary cortisol late nights elevated cortisol one milligram dexamethasone with a high ACTH this is Cushing's syndrome and so what your true differential is is this cyclic Cushing's syndrome most likely and what is the cause and so the problem is how are you going to figure out the cause if she only has one month of hypercortisolism the next time that she becomes hypercortisolemic? So what do we do to diagnose cyclic Cushing's syndrome? So after you've done your history and physical you're going to be thinking about the pre-test probability so let's first think about person who has a high pre-test probability but normal or discrepant screening test results. So what's a high pre-test probability? This is a person who has features of Cushing's syndrome that really give you the sense that that's what they have not the person who has things that are common or prevalent in the usual general population but people who have more unusual things or things that are unexpected for their age such as you know a young person who gets a fracture with very little trauma, a young person who develops hypertension, a woman in her 30s who becomes amenorrheic or develops irregular menses who doesn't have features to suggest another reason for that so for example she doesn't look like she has PCOS. So people who have things that really make you think about Cushing's syndrome as a very clear probability. So for those people you're really going to be thinking Cushing's so you need to catch the episodes because all the tests will be normal during the interval so the question is how often do you test them and what do you test them with? So there are many different ways to approach this. One approach is to determine by your history if the person thinks they have waxing and waning or on and off episodes to test when the patient feels worst and when they feel best. I found that this this is a nice approach when you have somebody where you don't really think they have Cushing's but they've come to you and you are the court of last resort. You know they've come to the temple of Cushing's in your office and they want to know do I have it or do I not have it and their referring physician wants you to tell him or her yes or no. So you're going to test that person because of the situation not because their pretest probability is so necessarily so high. So what I find in that setting is that if the patient is deciding when they feel the worst and you're telling them your urine's cortisol is going to be high when you have Cushing's then if they have a normal value when they have felt the worst it's easier for them to accept they probably don't have Cushing's syndrome and that it makes the decision to go forward and look for other causes of their signs and symptoms a little bit easier and so I when I have patients where I don't really think they have Cushing's I will take this approach. The other approach is to test over time and so this could be every week for somebody if they seem to have possibly frequent episodes of Cushing's. For someone who has the square wave where things come on for a while but then it just abruptly stops and it may be a while before they have their next one if you're testing every week there's a lot of testing because maybe they have a very long interval. So obviously you've heard what other people have found in their experience in terms of intervals. The longest interval that has been reported by minority was 5.5 years. The longest interval reported from the UK group was 20 I believe years. My longest interval I've seen is about three years but you can see that anywhere between three and 20 years you're not going to test every week. So you're going to probably in those cases try to identify some symptoms that the person had at the time of the episode and say test when that comes back. Test when you feel weak. Test when you're starting to really gain a lot of weight. So that is a very useful way of postponing testing but for the patients who have very frequent cycles so for example I've had someone who went from a urine cortisol of four to a urine cortisol of four to five thousand over a four to five day period for that kind of person you might even test every day or three times a week. So in general I suggest that people test once a week to start off with if they don't have a real feeling for the periodicity. So again for somebody with a sharp onset the square wave we ask about symptoms such as weakness or abnormal glucose or blood pressure if they can measure them and if they have been abnormal in the past and test for Cushing's when those sentinel signs and symptoms occur. That works pretty well for most patients. And then if you're concerned about what we used to call pseudocushing syndrome which we now are transitioning a bit to the non-neoplastic physiologic hypercortisolism term which is a little more wordy, I consider the dex suppression CRH test the combination but you can only do that when someone's hypercortisolemic because they're not going to have high values they'll look normal in the interval times. So if that is on your differential if they have urine free cortisol is less than three times normal and you're concerned about this pseudocushing's kind of situation then the dex CRH test can be helpful but again you have to do it when people have abnormal test results. So just to review again to look at what you could be testing with this urinary cortisol to creatinine ratio is something that most of us here in the United States don't do but you could imagine that these values are urine free cortisols for a 24-hour period. There are also some data in the literature about using a 12-hour overnight urine cortisol to creatinine ratio. I think the difficulties with either of these cortisol to creatinine ratios is that we don't have a very good sense of the normal range particularly if we're using tandem mass spectroscopy. So but in terms of looking here at the salivaries and the urines you could see that you know both may be abnormal and both can be done. What I generally do is to get the late night salivary cortisol within that same 24-hour period that the patient's collecting the urine cortisol. In that way I can correlate the salivary and the urine cortisol over time with various collections and see if one is always low or one is if there's a discrepancy I can think about the reasons for those discrepancies and consider whether they have pseudocutions or something else or some other reason why the results seem the way they are. The other thing to always keep in mind obviously when you're looking at urine cortisols let's say the urine cortisols are consistently high the salivaries are low. It's hard to get a high urine free cortisol without a high salivary cortisol and so probably that person is over collecting and you can check that by measuring creatinine and you can also look at the volume and then you have to remember the reasons other reasons physiologic reasons why you're going to have a high 24-hour urine cortisol which would be a very high fluid intake which will increase UFCs. So you have to go back to all of the caveats for all of these tests to figure out sometimes why there is a discrepancy. I find it very helpful to do them on the same day for that very reason because if if I do them on different days I don't know whether it's a discrepancy or just that they're periodic in their hormonogenesis. This is a nice paper looking at again another paper from a single center looking at 205 patients with Cushing's disease to determine how many patients had cyclicity and interestingly what they found was that the late night salivary cortisol was more sensitive. You can see here 77 percent for cyclic Cushing's versus the urine free cortisol and this was statistically different. You can see here the late night salivary cortisol on the top panel of someone is cyclic. The gray zone shows the normal range for both panels and you can see that this cyclic person has where the little asterisks are has a peak of salivary cortisol and then the non-cyclic person you can see below has consistently elevated salivary cortisols. So this single two patients demonstrating how useful salivary cortisol can be. It also I think demonstrates the utility of multiple tests. So for example let's look up at this panel. Had you only tested the patient at these times when the numbers are normal you would conclude the patient did not have cyclic Cushing's. So here this demonstrates that you don't want to do just one test. If you are really truly going to decide about cyclic Cushing's you have to do more than one test unless the first one comes back abnormal in which case you're going to say this is great I need to do a few more anyway because that's what the guidelines tell me or I want to be sure. But in patients with possible cyclic Cushing's you can definitely get normal tests and sometimes you really have to keep pursuing this for some period of time. So what would be the diagnostic strategy in a patient with a low pretest probability meaning they don't look very Cushingoid but they have some features that are consistent with Cushing's but they have discrepant screening test results. So they have some high numbers and some low numbers. In this case one of the things to consider is factitious Cushing's syndrome and here you may have various laboratory results. So for example if the person is taking hydrocortisone you will have a high UFC but you will have most likely a normal dexamethasone suppression test. If patients are taking synthetic steroids you will have a low or normal UFC, a lower normal ACTH and again a normal dexapression test. So when you have this type of discrepancy and again the salivary cortisols will follow on depending on when they're taking the high dose the high the doses of hydrocortisone. If they take it close to when they do the salivary test it'll be high. If they take it only in the morning for example it may be low at night. So you can get all sorts of combinations of discrepant results. Here you may want to do tandem mass spectrometry for measuring synthetic steroids. You may want to repeat the laboratory measurements over time and then there is one condition to think about for many of us who are using tandem mass spectrometry for urine cortisol results is that these results may be factitiously elevated by the use of skin creams. If the person is taking the Pledge It out of their mouth for salivary cortisol and putting it into the container using their fingers instead of spitting it into the container it's common in a number of skin care creams to improve itchiness that hydrocortisone is added to the lotion or the ointment. That is a known cause of discrepancy. In that setting, you can ask the laboratory if they will measure cortisone, because as you know, cortisol is inactivated to cortisone in the salivary gland. If you have a high salivary cortisol, you should have a very high salivary cortisone. If you have a high salivary cortisol and a low salivary cortisone, that suggests exogenous contamination of the saliva as it's being put into the container. Glucocorticoid resistance is another thing to think about. It's extremely uncommon, but it would be something to think about if you have laboratory evidence for glucocorticoid excess. True increases in all of the parameters, urine-free cortisol, salivary cortisol, and very little in typical clinical signs and symptoms. However, you may have, for example, hypertension but no other features of Cushing syndrome or precocious puberty, but no features of Cushing syndrome in glucocorticoid resistance. You have to consider the whole clinical picture. In these patients, what has happened is because of the abnormalities in the glucocorticoid receptor, the ACTH levels are quite high there. It's pushing the adrenal glands to make cortisol to say you have more cortisol and you're hoping to get some action out of the glucocorticoid receptor. For that reason, the diurnal pattern of cortisol is present, but it's set up. You do see a diurnal variation, but the nadir is quite high. You can measure diurnal cortisol. Also, these patients will suppress to dex, but usually it takes much more dexamethasone than a normal person, so you can do a progressive dexamethasone suppression test. How about the differential diagnosis? Well, we know that normal corticotrophs must be suppressed, so you need to know the duration of hypercortisolism before you can start doing differential diagnostic tests. We don't know the maximal severity and the minimal duration that we need, or I should say the minimal severity and duration for suppression of the corticotrophs, and we don't know the timeframe for reversal. Anecdotal experience suggests that you have to have two times normal hypercortisolism, so two times the normal UFC, for example, and four to six weeks in order to suppress normal corticotrophs. But it's possible that if you have very much higher urine-free cortisol, that it takes less time. We really just don't know. The ideal would be to do testing after at least four to six weeks of mild to moderate hypercortisolism. As I just said, this might be a minimal approach. This would be to test when you have a two-fold increase for four to six weeks. One option might be to give dexamethasone prior to petrosus sinus sampling, but the role for that is not known. It hasn't been tested very broadly, and so we can't say whether that would help or not. It might be an approach for the patient with a very short cycle and then a long period of eucortisolism, where you don't think that short cycles are occurring often enough to suppress the gland. DHEAS also, we don't know the time to suppression of that, so it may not be very useful. Otherwise, we just do the regular tests, the normal way we do them, but sometimes finding the sweet spot, when to do the testing is difficult. This is what we've talked about this afternoon. I hope that you feel that we've met those objectives in the beginning of the talk. I thank you for your attention and draw your attention to the beautiful cherry blossoms here in Washington DC occurring just a few weeks ago, but now sadly gone. Thanks a lot. Thank you, Dr. Newman, for this wonderful presentation on one of the most challenging conditions that endocrinologists encounter in their practice. The floor is now open for questions, and we see them here in the chat box. I'm going to start with some questions for you with regards to use of late-night salivary cortisone. When do you tell your patients to collect their sample in terms of the exact timing? What do you do with those patients who work third shift? Those would be my initial questions for you. Thank you. I want to thank again the organizers for inviting me to speak to everyone today. This is, I think, challenging even for endocrinologists who've seen a lot of patients with Cushing's syndrome. The salivary cortisol is also something that we can use in some of these patients, and it doesn't work in other patients. Let me frame that out a little bit. When I think about salivary cortisol or any diagnostic test, I like to think of it in the context of the physiology of what's normal. We know that people who have normal or consistent sleep-wake cycles have their lowest cortisol value in the blood, and hence in the saliva, about within an hour or so of falling asleep. Salivary cortisol levels are really tightly entrained to sleep. It's not at 11 o'clock at night or 11.30 or midnight. Our bodies don't know what the clock time is. Our bodies just know sleep. What I tell my patients is, first, I ask them what their sleep schedule is. I want to know that they have a consistent pattern of going to sleep, within an hour or two perhaps. Then I tell them, try to be consistent the week or so before you do the sample. The night of the sample, give the sample just before you go to bed, just before you're planning to sleep. Don't do anything exciting to get your cortisol levels up. Don't read any fancy books or watch TV, or get into an argument with your partner, or yell at your kids. Just have a nice quiet evening, take the sample just before you go to sleep. When we think about it from that perspective, if you are a shift worker, your salivary cortisol is going to be all over the place because it takes a while to shift from one sleep-wake cycle to another sleep-wake cycle. If you have someone who is on nights all the time, then you just have them take a salivary cortisol just before they go to bed, whatever time of day that would be. But if you have someone who is on nights for five days and then goes back to living during the daytime awake, then you're going to have very likely a falsely positive result. You could try in someone who's going to be on one set of schedules for at least seven days, because it takes about that long to switch over, you could try on that last shift to get the sample before they go to bed after that last shift. But if it comes back positive, I would discard it and try some other tests. If it comes back normal, you could accept it as excluding cushions at that particular day. So that's, it's a problem. And so what I usually do for testing night shift people, if I can't trust that they're going to be consistent for at least seven days, I use another test, either urine-free cortisol or the overnight dexamethasone suppression test. Yeah, thank you so much for your answer. Also about salivary cortisol, a question that is high in the polls, how many samples do you need to say no cushioning? So we all know that, especially in these patients with tests that are all over the place and symptoms that fluctuate, we need many pairs of late night salivary cortisol tests, but when would you say, okay, I don't think you have it? Yeah, well, that's very tricky because it depends on the patient. If you have someone with consistent cushions, in the guidelines, we recommended two to three samples at a minimum. If you have someone where you suspect cyclic cushions, you need to develop your strategy. How am I going to catch this cyclic cushions? And if the person has symptoms, what I usually do is ask them to collect at least two or three samples of urine and saliva on the same 24 hour period, so you can correlate the two. And I ask them to do that while they're having symptoms, two or three, and when they're not having symptoms, two or three. If all of those come back normal, which is what usually happens because they don't really have cushions, I'll say, look, we tested you when you were having symptoms and it doesn't look like you have cushions at this time. And then I educate them about the signs and symptoms of cushions. Most of the time, they know all of that. And then I say, if you develop something else, come back and let's test you again. On the other hand, if my pretest probability is really high, someone comes in, they look cushy annoyed, I really think they might have it, then I'll continue to test them. And I usually use one of two approaches. One, the one I most commonly use when people are not very symptomatic is I'll test them once a week for a month or two, three, depending on my pretest probability. Or if that doesn't work and I still think they might have cushions, but they may have a long interval of being normal, I'll ask them to test when something changes. So perhaps I'll have them check their blood pressure or if they had any kind of a sign or symptom before, like one thing I found helpful is people who get cramps when they exercise because probably because they're hypokalemic and they have an ectopic and they have one of these square wave functions, I'll ask them to come back and get tested again when they have those cramps. So it really has to be individualized to each patient and there's no one correct answer or one right algorithm that you can always use. Yes, this is a situation where our clinical judgment, we have to use it and we have to see the patient in clinic and stay in touch with them. So it is my understanding you're using all three screening tests for diagnosis of cyclic cushioning as you would do with any patient suspected for cushioning. And is there a certain strategy? Do you do the dexamethasone test repeatedly? Like how do you, the question pretty much says, do you need two or three positive tests to confirm your diagnosis? And which one is your best choice? So I would say the worst choice is the dexamethasone suppression test because you really don't wanna keep giving dexamethasone every week or every two weeks or whatever. So I usually do the urine and the saliva on the same day. And the reason for that is that I have my two tests already. And if they correlate with each other and I get positive tests, I'm done. If I want to even further confirm, if I have a worry, then I can do a dexapression test but I have to do it at a time when the salivary and the urine is high. Otherwise they'll have cycled out and you'll get a normal response. So it's important that you do dexamethasone testing and any differential diagnosis testing when the person is in a high phase. And I know there's some concern about how do you know they're in a high phase? I personally, for every single patient I have, cyclic, not cyclic, whatever, I get a urine-free cortisol. And if I can salivary with it once a week for six weeks before I admit them, I want to know that they are hypercortisolimic enough to suppress their normal corticotrophs. So I'm not dealing with a false positive IPSS, false positive CRH response. And if people cycle out before they come, I tell them ahead of time, if you cycle out, we have to reschedule. Because the worst thing I think you can do is to test an ectopic, find them to have Cushing's disease when they really have just a normal response, and then take them to transmittal surgery that they didn't need. So I think we all have to be very careful to test in periods of hypercortisolism. Thank you. So this answers the question about the IPSS, right? You wouldn't send anyone to the IPSS unless you've proved with the screening test that the patient is on their up cycle, they're having clinical and biochemical hypercortisol MS at that time. And there is an actual case that is described here. The question has to do with use of three serum cortisol levels. So this patient is on oral contraceptive and has elevated serum cortisol as expected. And her free serum level was checked by someone else and was elevated as well two to three times up above normal. The late night salivary cortisol was normal three times. Would you expect that oral contraceptive affect free serum cortisol levels that much? I think it depends on what may have occurred here is the patient may have given their sample in the morning and not in the evening. And so it's a question of, are there morning norms for this subject and were they really two to three times elevated? Also remember, it's possible to have a pulse of cortisol that is then translated into a free cortisol level that's very high. So I would be concerned about using a free cortisol value when I didn't have good norms. And we do know that late night salivary cortisols correlate with free plasma or serum cortisol levels. So the fact that you've got some normal salivaries, some abnormal free cortisols, assuming the reference range and the timing is correct, that suggests that you either have someone who was activated in terms of being excited at the time of the serum draw or someone who has cyclic cushions. And so, again, trying to get UFCs and salivary cortisols on the same day, I have found to be very useful in getting rid of that kind of confusion that comes up. Excellent. Four more minutes, four questions. The first one could be quick, but nothing is quick with cyclic cushioning. How many of your patients present with cyclic cushioning and adrenal insufficiency? So they have these up and really low excursions. Yeah, so most of the people I've seen, I've seen a mixture. I have these square wave people. I have people who cycle very, very rapidly, and I've got things in between. What I don't see a lot is people who have sufficiently long periods of being on to render them hypocortisolemic after they cycle out. And so what I oftentimes see is a little of a partial impairment of the axis, but not a complete impairment. But it's something that I think you need to be aware of as a possibility. And when that does happen, you do need to replace them with hydrocortisone. I replace them in a less than complete basis, however, at the very lower limit of a conservative replacement dose. Thank you. So in patients with cushioning syndrome, possibly cyclic adrenal etiology, do you put more emphasis on one milligram dexamethasone than the other screening tests? So I only put that emphasis on someone when it is an adrenal incidentaloma, and I'm trying to sort out whether they have mild dysregulated cortisol excess. And in that case, the dexamethasone suppression test is better than saliva or urine cortisol levels. In someone with overt cushion syndrome and adrenal adenoma, you can rely on all three of the tests. Yes, thank you. One question about clinical presentation. So some patients constantly gain weight. Would you expect them to have periods of weight loss if they truly have cyclic cushioning? And what is going on with your periods? Like how much do you rely on irregular periods when it comes to women with cushions? Yeah, so I mean, so the classic thing is if women have irregular menses and, or amenorrhea, if that's really the case when you have moderate to severe cushions and mild cushions, you don't always have a loss of cycles or somebody with a little bit of an unstable GnRH clock is may have amenorrhea with relatively mild cushions. So I think periods are helpful if they go away, if they haven't gone away, you can't be so sure. And in terms of the weight loss, I have seen people lose weight in their interval times, but it's people don't always lose weight once they've gained weight. It's like just a regular person without cushions may not lose weight even when they initially start to quote unquote diet. So if it's present and it shows you a variability, that's helpful. If they don't lose weight, it doesn't really mean that they don't have cyclic cushions. I think we have time for one more important question. How do you determine biochemical remission after surgery for these patients? Oh, that is a really great question. So what you're doing is our usual biochemical remission occurs because you've suppressed the normal corticotrophs. So that unmasks that suppression and people are hypocortisolemic. If the corticotroph has not been suppressed, then you may not see an abnormally low value. So what I do in that setting is I just repeat all the screening tests because you're asking the question, do they have cushions? So I do a one milligram overnight dex, I do a late night serum cortisol usually because they're in the hospital, but you could do a late night pre-bedtime salivary and I do a urine free cortisol. And that has helped us in many, many cases of patients with either mild cushions or cyclic cushions or the occasional patient who has been treated medically prior to surgery. Great question. Dr. Nieman, I'm afraid this is all the time we have. I think if we stay online, we will keep getting questions. So we look forward to more sessions like this. We truly appreciate your time and sharing with us your expertise. And as someone has written here, your beautiful cherry blossoms in your presentation also contributed to everything here. Thank you again. And for the attendees, hopefully you have the rest of ACE meeting is really good and fruitful for you. Yes, I'm looking forward to it. Thank you so much. Bye-bye everybody.

Dr. Glynette Nieman

cyclic Cushing syndrome

intermittent hypercortisolism

eucortisolism

duration of eucortisolism

duration of hypercortisolism

detailed history

late-night salivary cortisol test

diagnosis of cyclic Cushing syndrome