A (3:15)

What do we know about the relationship between type 2 diabetes and hypercortisolism?

B (3:21)

We recognize this in clinical practice a lot, mainly related to iatrogenic hypercortisolism, which is people taking steroids. And you almost immediately see a rise in blood glucose. For people without diabetes, that rise might be small because they can compensate by producing more insulin. But for people with diabetes, it often leads to poor control, necessitating changes in therapy, addition of additional therapy and so on. I just want to address what Dr. Findling just said about pseudo cushing's when you start looking for it. Some of it relates to tests that were done improperly, such as not taking the dexamethasone properly, or so it doesn't suppress or taking things that might interfere with the testing. So for example, a high estrogen because of oral contraceptive pills, or a night shift worker, or people with end stage renal disease and so on. So it's important to recognize that and we very carefully excluded that in the Catalyst study. So I don't think we included a lot of people with so called pseudo cushings.

A (4:29)

The study mentions that mifepristone was tested as a therapy to improve glycemic control. So why mifepristone?

C (4:37)

Mifepristone was approved for use in 2012. It was the first drug approved for the management of hypocortisolism. And it was shown in the Seismic trial published in 2012 that there was a dramatic improvement in glycemic control even before patients lost weight when they were started on mifepristone for neoplastic hypocortisolism. And when it was approved by the fda, it was approved for treatment of patients who had either impaired glucose tolerance, pre diabetes or frank diabetes. So it was known ahead of time that this drug was particularly useful in improving glycemic control in patients with diabetes. And just to elaborate on the dexamethrowed suppression test, which was kind of the segue into getting into the trial and many patients, and all the patients, or 24% of the patients studied, had abnormal dexamethasone suppression tests. So the prevalence of of cortisol excess in this group of patients was much higher, I think, than most of the investigators. Had expected. And it's really quite similar to a condition called max, mild autonomous cortisol secretion, which now has become very clear that those patients have significant cardiometabolic improvements with therapy, either adrenalectomy or pharmacotherapy. So I think it's becoming more and more clear that even patients with mild degrees of cortisol excess, often illustrated by an abnormal dexamethasone suppression test, may benefit from some cortisol directed therapy.

A (6:08)

We've been talking about the study a bit now, but let's just really dive into it. Can you just walk us through the design of the study and the methods?

B (6:15)

The study was done in two parts. One was to determine the prevalence of hypercortisolism in people with difficult to control diabetes in the manner that I just described the definition. There may be others who could also respond, but we chose this particular group definition in the hope that we would pick up more people with problems such as Max. And just to build on what Dr. Findling said, we see this in practice with people taking modest doses of prednisone. If they have diabetes, it gets worse. So this is analogous to that in some way. We excluded people who might have had a false positive, such as use of oral contraceptive, pills of excess alcohol consumption, shift workers, kidney failure, and so on. And we did a dexamethasone suppression test. In the first part, we gave 1 milligram of dexamethasone and the next morning measured the Cortisol level before 9am the criteria we used for determining hypo, or defining hypocortisolism, was a cortisol level the next morning or greater than 1.8 micrograms per deciliter. And we measured the dexamethasone level to ensure that the patient had taken the dexamethasone. We wanted to see that it was over 140 nanograms per deciliter. In fact, we got much higher levels in the actual study. And those who had hypercortisolism were further evaluated with labs such as acth, dhea, as an adrenal CT scan. And we'll discuss that when we discuss the results. But just to outline the second phase, we took people who had hypercortisol, as defined as having an abnormal dexamethasone suppression test. They were randomized to mifepristone or placebo. Now, we only randomized people who were willing to be randomized, did not want surgery, didn't have any contraindication to mifepristone, et cetera. So we had Adequate power to test whether mifepristone was superior to placebo. And the endpoint was hemoglobin A1C change. So change in glycemic control. So that's a broad outline of the design of the study.

A (8:25)

So now I'm sure everybody wants to know what you found. So what were the results of the study?

B (8:30)

First of all, you know, let me give you a little bit about the demographics of the kind of people we enrolled. They were aged about 60 years. They were almost equal, men and women. They had moderate obesity, but the BMI was 33. Mean waist circumference was 44, so a little bit of central obesity. They were represented a wide range of people in the United States. 71% white, 19% black, about 24% Hispanic. Their A1 mean A1C was 8.8, but their blood pressure was reasonably well controlled. 127 over 75. 69% were taking at least three antihypoglycemic drugs, a wide range of antihyperglycemic medications, including metformin. 71% were taking insulin, 50 plus percent taking SGLT2 inhibitors, and over 50% were taking either GLP1 receptor agonists or dual agonists such as tirzepatide. And many of them were taking combinations of all these, including some taking insulin plus GLP1RA and SGLT2 inhibitors. And despite that were not controlled. So these were very challenging for people like, including myself, having people like this in our clinic were not well controlled. In addition to that, 80% were taking lipid lowering drugs, Many were taking antihypertensive drugs. Again, a wide range of that. Interestingly, a lot of them had some form of pain. They were taking analgesics. Very small number were taking opioids. A few were taking psychiatric medications, not for severe psychosis, but for mostly antidepressants for mild to moderate depression. And a few had cardiovascular disease. So the first part, as Dr. Findling pointed out, we were very surprised that 24% had hypercortisolism. We defined what we meant by it, but the mean post DST cortisol in this population was 3.5. So it's not like just over 1.8. So it was fairly well increased to a great extent despite having a dexamethasone level of 412 milligrams per decimeter. So the test was probably conducted. We did our best to exclude pseudo Cushing's, yet find such a high prevalence, which was surprising to us because we haven't really thought this through before. But has been described in the literature. Just one other thing about the characteristics of these patients. The simple thing is the more medications they were taking, the greater the likelihood that they had an abnormal test. More medicines for diabetes, more medicines for hypertension, and in fact, if people who are taking three or more drugs for diabetes and three or more drugs for blood pressure, 35% of them had an abnormal DST. This is a fairly common scenario in diabetes clinics. And we haven't really thought about this before. It was more common in whites and African Americans, perhaps a little bit less common in Hispanics for reasons we haven't worked out. And the worse your control and the more drugs you were taking, more common was the condition. The other really interesting thing that we found was when we did CT scan, we found that 66% had no imaging abnormality, which means nearly 24% of people had some abnormality, of which 23% was a unilateral adrenal adenoma. So these were people who could be surgically treated that we had missed previously, despite taking care of them for their diabetes and not being successful. I think this is a very important finding that people should pay attention to and perhaps address in clinical practice. So that's the first part. We then went on to the second part and we randomized patients to mifepristone or placebo. The baseline A1C in the mifepristone group was 8.6% and it fell by almost 1.5%. That is a highly significant fall. The drop with placebo was only 0.15, so the difference between the two was minus 1.3, which is better than many of the diabetes drugs that we have in clinical practice. The effects came on fairly quickly. Within 12 weeks, there was a significant drop in A1C and that persisted out through the 24 weeks of the study. The drop in A1C was also needs to be considered in the light of what we did for patient safety. To avoid hypoglycemia or treat hypoglycemia, many patients reduced their diabetes medications, particularly insulin. And many came off their insulin, including fast acting insulin and in some cases short acting insulin or sulfonylurea. So despite reducing the medications, the diabetes control improved very significantly. And there were similar improvements whether you had an adrenal adenoma or didn't have an adrenal adenoma, which is also very interesting. I don't think there's ever been a study before of looking at people who don't have a clear cut tumor and treating them medically. So this is another important finding. There was some reduction in body weight, modest, about 5kg. Most of that was fat around the waist. As the waist circumference declined, Maybe I'll ask Dr. Findling to discuss the side effects, many of which we anticipated based on previous experience. Yeah.

A (13:53)

So were there any adverse effects in using the fipristone that was noticed?

C (13:57)

Yeah, the anticipated side effects of block and cortisol effect, as you blunt cortisol secretion. In patients who have had excessive cortisol, there is a withdrawal period of often nausea, muscle aches and pains, and sometimes fatigue. And those were seen in this group of patients, not unexpectedly. Obviously, it was a randomized placebo controlled trial, so the investigators nor the patients knew whether they were getting placebo or mifepristone. And then the most significant adverse effect, which was also would have been anticipated, was hypokalemia. Since mifepristone blocks the cortisol receptor, there's a increase subsequently in ACTH and cortisol secretion. So actually cortisol levels go up, which overwhelmed the kidney's ability to metabolize cortisol, thereby giving it an opportunity to bind to the marocorticoid receptor and cause caliesis, potassium wasting, and sometimes a slight increase in blood pressure, which was observed in this group of patients. And so that was seen as a major adverse event. Of course, again, because it was a randomized placebo controlled trial, the patients could not be treated preemptively with sparloctone. In real practice, when this drug is used, I almost routinely give patients spironolactone and sometimes potassium replacement if necessary. Should also be pointed out, of course, this was an older population of patients. I think most of the patients, women were postmenopausal. This drug is also a progesterone receptor antagonist. So obviously this could stop menstrual periods. And of course, women cannot get pregnant on this drug. This was not a very big concern in this older population of patients, but would certainly be a concern in younger women.

A (15:44)

At the end of the day, how does this study better inform us about the role of cortisol in diabetes?

B (15:51)

This study provides some proof of concept, the concept being that cortisol is important. Even moderate levels of cortisol above the normal, which have been associated in some studies in the past with diabetes. Those modest elevations may be contributing towards the difficulty in controlling these people. And blocking that cortisol, either through its action, could be done through some other ways. I mean, we could have had surgery in some people, which we do in practice you can use other drugs, but they have not been tested in this way in this kind of population. So this is the first one and it provides, as I said, proof of principle. And I think we are going to see a lot of interest in exploring this. We don't know why it's occurring, we don't fully understand the mechanism, but at least we've hit on a possible way to manage it.

A (16:44)

We've talked a little bit in this conversation about how some of the results from the study were surprising. So is the medical community aware of this connection between cortisol and diabetes? And if not, why?

C (16:56)

I think that every physician on earth knows that glucocorticoids, I mean, they're called glucocorticoids for a reason. Of course, increase consumed glucose in patients with diabetes, certainly disrupts their glycemic control. Everybody understands that. I think what people don't understand is that patients who have diabetes, and I would extend this beyond even just patients with poorly controlled diabetes, one of the underlying causes may be some kind of hypercausaly mixed state. And so that I think more broad screening should be done, and certainly this study amplifies that point in this group of patients with poorly controlled diabetes. The more broad screening should be done in this patient population to identify those who have dysregulated cortisol hypersecretion and who may benefit from either surgical or pharmacotherapy.

A (17:46)

You mentioned screenings, and maybe that's the answer to my next question. But what do you think are the most important take home messages for healthcare providers regarding this study?

B (17:55)

We are used to Cushing's syndrome with a floridypical textbook description. And these patients did not have that. That doesn't mean they don't have an abnormality. Their abnormalities being manifest in difficult to control diabetes and hypertension, which was controlled but requiring a lot of drugs. And that's not a happy situation for the patient. You know, you could call it Cushing's as Harvey Cushing described it. It's not, you know, you don't get Cushings overnight. Maybe they would get it if you waited, but if you waited, they'll get complications of diabetes. And with that, the imperative is to improve that outcome. I think if you're confronted with a challenging patient where you don't have an easy answer, we shouldn't just dismiss it as being related to lack of compliance or lack of dieting and so on and exclude this possibility. I mean, that is a common scenario and many patients were grateful. When you find something wrong that they hadn't thought about before and the physicians hadn't thought about before. So identifying an abnormality that you could possibly treat really contributes a great deal towards them being accepting of what's going on and looking for a treatment solution.

A (19:09)

Okay, that's all the time we have for now. Thank you both for being on the podcast. I really appreciate you taking the time to talk to us. Thank you.

C (19:15)

Thank you, Aaron.

B (19:16)

Thank you. Thank you, Aaron.

A (19:19)

And that's all for this episode. I hope you found today's topic to be as interesting as I did. You know, I'm always on the lookout for the next great issue to feature on the podcast. If you have an idea, please share it with me@podcastndocrine.org that's podcastndocrine.org we'll be back soon with another fabric. Fascinating dive into the world of endocrinology. Until then, thanks for listening.

C (19:51)

Endocrine News Podcasts are a free service of the Endocrine Society. To learn more or to become a member, visit the society's website at www.endocrine.org.