hypercortisolism in patients with diabetes

 A recent article explored the relationship between difficult-to-control diabetes and endogenous hypercortisolism, finding improved glycemic control with the use of mifepristone in the CATALYST study (see dm hypercortisolism mifepristone rx DiabCare2025 in dropbox, or https://doi.org/10.2337/dc25-1055), in the CATALYST study

Details:

-- 136 individuals who had type II diabetes (T2D) and hemoglobin A1c between 7.5% and 11.5% on multiple medications were then randomized in this prospective, multicenter, double-blind study of  the glucocorticoid receptor antagonist mifepristone (300 to 900 mg daily) vs placebo for 24 weeks, stratified by the presence or absence of an adrenal imaging abnormality

    -- specifically, patients had to be on least three glucose lowering medications, or taking insulin and any other glucose lowering medication(s), or taking at least two glucose lowering medications and having at least one microvascular or macrovascular complication, or taking at least two glucose lowering medications and at least two blood pressure lowering medications

-- the CATALYST study was performed in 36 sites across United States, and participants were recruited from outpatient general medicine and endocrinology clinics

-- patients were considered to have hypercortisolism based on a 1-mg overnight DST (dexamethasone suppression test) with a non-suppressed morning cortisol of at least 1.8 µg/dL and a dexamethasone level at least 140 ng/ml; those with common causes for false-positive DST results were excluded

    -- other exclusions included those with systolic blood pressure >160 mmHg or diastolic blood pressure >100 mmHg, serum potassium that could not be corrected to at least 4 mmol/L, poorly controlled hypothyroidism or hyperthyroidism, renal insufficiency with a creatinine-based eGFR< 30, ALT or AST >3x upper limit of normal or bilirubin >1.5x upper limit of normal, or being pregnant or breast-feeding

-- patients meeting the criteria for inclusion had a starting dose of mifepristone of 300 mg a day vs placebo, increased to 600 mg a day versus placebo; there was an optional increase of mifepristone to 900 mg a day at weeks 8 or 12

    -- two weeks after both initiation or escalation of the mifepristone, participants had monitoring for potassium, blood pressure, and glycemic control

-- an abdominal CT was performed for identification of possible adrenal imaging abnormalities, and ACTH and DHEAS were assessed

-- of the 136 participants, 91 were randomly assigned to mifepristone and 45 to placebo, and all took at least one dose of the study medication or placebo

-- baseline characteristics of the participants were balanced:

    -- mean age 63, 61% male, 81% white/15% Black/93% non-Hispanic

    -- body weight 99 kg, waist circumference 114 cm, BMI 33

    -- hemoglobin A1c 8.6%

    -- abnormal adrenal CT in 28%, post DST (dexamethasone suppression test) cortisol levels: 3.6 mcg/dL, post DST dexamethasone level 422 ng/dL

    -- ACTH 19 pg/mL, DHEAS 86 mcg/dL

    -- at baseline 38% of the participants were taking fast-acting insulin, 65% were on long or intermediate acting insulin, and 27% were taking sulfonylurea

-- Primary endpoint was change in hemoglobin A1c from baseline to week 24

-- secondary endpoints included change in glucose lowering medications, weight, waist circumference, fasting plasma glucose, lipids, blood pressure, as well as safety

Results:

primary endpoints:

-- mean hemoglobin A1c at week 24:

    -- mifepristone group: decreased from 8.62% to 7.12%, percent change -1.47% (-1.79% to -1.14%)

    -- placebo group: decreased from 8.41% to 8.36%, percent change -0.15% (-0.56% to 0.27%)

        -- difference between the arms: -1.32% (-1.81% to -0.83%), P<0.001

-- mean hemoglobin A1c change in those with versus without adrenal abnormalities on CT:

    -- with adrenal abnormalities:

        -- A1c changes: -1.41% (-2.34% to -0.48%)

    -- without adrenal abnormalities:

        -- A1c changes -1.29% (-1.88% to -0.70%)

-- sensitivity analyses of participants on mifepristone:

    -- A1c while the participants were on treatment (n= 48): -1.58% (-2.10% to -1.05%), p<0.001

    -- A1c of participants who completed the study (n=47): -1.56% (-2.12% to -1.00%), p<0.001

    -- no outcome difference by excluding those on 900mg of mifepristone

 

key secondary endpoints:

-- reductions or discontinuations of insulin in the first 12 weeks:

    -- mifepristone group versus placebo: 30% versus 11% decreased fast-acting insulin, 49% versus 13% decreased long-acting insulin: and 22% versus 11% decreased sulfonylureas

-- body weight change at week 24:

        -- mifepristone group -4.40 kg (-6.27 to -2.525) versus placebo group 0.72 kg (-1.838 to 3.272)

            -- difference of -5.12 kg (-8.2 to -2.0)

    -- also, essentially no patients in either group started a GLP-1 or tirzepatide

-- difference in BMI:

    -- mifepristone group -1.47 (-2.1 to -0.8) versus 0.28 (-0.58 to 1.13) in the placebo group

        -- difference -1.75 (-2.8 to -0.7)

    -- difference in waist circumference:

        -- mifepristone group -5.2 cm (-7.25 to -3.2) versus -0.1 cm (-2.7 to 2.5) in the placebo group

            -- difference of -5.1 cm (-8.2 to -2) 

-- difference in mean blood pressure:

    -- mifepristone group -1.47 mmHg (-2.1 to -0.8) versus 0.28 mmHg (-0.58 to 1.13) in the placebo group

        -- difference -1.75  mHg (-2.8 to -0.7)

    --  overall there was a difference of 10.1 mmHg (3.62 to 16.59) in systolic blood pressure and 3.7 mmHg (-0.29 to 7.72) in diastolic, favoring mifepristone

-- difference in fasting plasma glucose:

        -- mifepristone group -30.7 mg/dL (-45.1 to -16.3) versus -10.7 mg/dL (-28.3 to 7) in the placebo group

        -- difference of -20 mg/dL (-41 to 1.3) 

-- difference in HDL cholesterol:

        -- mifepristone group -3.8 mg/dL (-6.7 to -1.0) versus 1.1 (-2.5 to 4.7) in the placebo group

        -- difference of -4.9 mg/dL (-9.1 to -0.8) 

-- difference in LDL cholesterol:

        -- mifepristone group -6.8 mg/dL  (-14.8 to -16.3) versus 2.0 (-7.8 to 12) in the placebo group

        -- difference of -8.8 mg/dL (-20.8 to 3.17) 

 

Safety:

--54% of those on mifepristone completed the 24 weeks of treatment, compared to 82% in the placebo group

-- the most common reasons for treatment discontinuation were adverse effects in the mifepristone versus placebo groups (62% versus 38%), study withdrawal by the patient (12% versus 25%), being lost to follow-up (7% versus 13%), and sponsor request (10% versus 0%)

-- at least 10% of patients on mifepristone had adverse effects of: hypokalemia (23% vs 0%), fatigue (15% vs 9%), peripheral edema (11% vs 2%), nausea (10% vs 1%); other common problems included hypoglycemia (7% vs 3%), diarrhea (7% vs 5%), headache (4% vs 9%), and dizziness (6% vs 2%)

-- serious treatment-emergent adverse events in at least 2% of participants: hypokalemia (6% vs 0), euglycemic ketoacidosis (3% vs 0), acute kidney injury (2% vs 0), and heart failure (2% vs 0)

    -- the 3 cases of euglycemic DKA were in people taking SGLT2 inhibitors were eating less food and with no reduction in insulin doses; all were on mifepristone, though there are insufficient data suggesting a direct connection with the mifepristone

-- most of these adverse events were grade 1 or 2 and manageable

Commentary:

-- a prior CATALYST report of 1057 patients aged 18 to 80 with difficult-to-control T2D and having A1c between 7.5% to 11.5%, who were on 2 or more diabetes meds with or without micro/macrovascular complications or taking multiple BP meds and then were screened with a 1mg dexamethasone suppression test (see https://gmodestmedblogs.blogspot.com/2025/07/hypercortisolism-common-in-hard-to.html or for full text: https://diabetesjournals.org/care/article/48/12/2012/158180/Prevalence-of-Hypercortisolism-in-Difficult-to) found:

    -- 23.4% had unsuppressed DST cortisol (ie, had autonomous hypercortisolism)

    -- 219 of 252 participants (87%) had CT scans, with 35% having adrenal image abnormalities, of whom about 70% had unilateral adrenal nodules

    -- 33% of participants with hypercortisolism had cardiac disorders, vs 21% without hypercortisolism (p<0.0001)

    -- 38% of participants who were on at least 3 antihypertensives had hypercortisolism, vs 23% without hypercortisolism (p<0.0001)

    -- use of SGLT2 inhibitors, maximal-dose GLP1s or tirzepatide, or higher number of antihypertensives, older age, BMI<30, non-Latino ethnicity, and use of fibrates or analgesics were associated with higher prevalence of hypercortisolism

        -- for example, 23% of those on maximal dose of GLP-1s had hypercortisolism vs 17% without hypercortisolism (p=0.0260}

-- this current CATALYST study found several things:

    --there were significant reductions in A1c by 12 weeks in those on mifepristone, reinforcing that the direct pharmacologic effect of mifepristone in lowering cortisol was a reasonable approach in those with hypercortisolism (it blocks the action of cortisol by binding to the glucocorticoid receptor: https://pubmed.ncbi.nlm.nih.gov/17984235/)

    -- there was an impressive decrease in the need for diabetes meds, weight, BMI and waist circumference

        -- a different small study of 50 patients found that the weight loss was largely in visceral fat: https://academic.oup.com/jcem/article-abstract/97/6/2039/2536684

    -- the benefits of mifepristone were evident prior to significant weight loss and with the lowest mifepristone doses

    -- patients on insulin and sulfonylureas had the most rapid reductions in glucose levels 

    -- hypokalemia was common (30% in the study) and it can be associated with hypernatremia. it seems that there is benefit in co-prescribing spironolactone or other potassium-sparing diuretic :https://link.springer.com/article/10.1186/s40842-020-00105-4

        -- in fact, in this study 33% of those on mifepristone were on potassium-sparing diuretics, mostly spironolactone, vs 2.2% in the placebo arm

    -- most of the adverse events were mild and manageable, and the most common ones (fatigue, nausea, vomiting and headache) are consistent with cortisol withdrawal associated with any treatment for hypercortisolism, including both pharmacologic and surgical

        -- this raises the question that since adverse events are so common with mifepristone, perhaps they could be avoided by proactive initiation of potassium-sparing agents, proactive reduction of insulin and sulfonylurea doses, and close monitoring for the cortisol withdrawal symptoms

-- the MOMENTUM study is designed to assess the prevalence of endogenous hypercortisolism in the US population who have resistant hypertension and to enroll 1000 participants in this observational, multicenter study; endogenous hypercortisolism is defined as a cortisol level >1.8 μg/dL on the 1-mg overnight dexamethasone suppression test with adequate blood dexamethasone levels (at least 140 ng/dL): see hypercortisolism MOMENTUM study JACC ADV2026 in dropbox, or https://doi.org/10.1016/j.jacadv.2026.102596

    -- the basis for this study is that 18% of patients with hypertension have resistant hypertension

    -- resistant hypertension is associated with increased risk of adverse clinical outcomes, including cardiovascular disease (MI, heart failure) at any level of elevated blood cortisol, hypertension in up to 85% and thromboembolic events in up to 20%, and insulin resistance, diabetes, fatty liver disease, dyslipidemia, and visceral obesity

    -- endogenous hypercortisolism is less well-studied than some other secondary hypertension causes, though there are several different mechanisms by which hypercortisolism is likely associated with the diverse adverse events

-- the above paper is a description of the study. not sure when it will be underway nor when results are likely to be ready

    

Limitations:

-- one evident limitation was the high numbers of participants who had adverse events and also the high numbers of people who terminated the study early. there may be ways to reduce many of the adverse events through med adjustments, as above

-- there may have been a selection bias if patients were recruited in an endocrinology clinic, since patients referred there may well have been sicker than in a general medicine clinic

-- this was a largely white population evaluated, and there are concerns that other groups may well have very different eating/exercise patterns that affect diabetic control

 

So, this study suggests several things:

-- it is clear that hypercortisolism is common, especially in people with hard-to-treat diabetes who are taking at least three glucose lowering medications, or taking insulin and any other glucose lowering medication(s), or taking at least two glucose lowering medications and having at least one microvascular or macrovascular complication, or taking at least two glucose lowering medications and at least two blood pressure lowering medications (though we do not have granular data about which of these qualifying preconditions is primary)

-- it does seem reasonable to test for hypercortisolism in this cohort

    -- and the overnight 1mg DST is a pretty simple test that could lead to much better diabetes control and very likely improvement of many other linked adverse outcomes. treatment options include surgery or meds (ketoconazole, metyrapone, or mifepristone)

-- interesting that those requiring maximal doses of GLP-1s and tirzepatide were at significantly higher risk of having hypercortisolism. Does reverse engineering of this outcome make sense: i personally see many patients on maximal doses of GLP1's/tirzepatide who have minimal diabetes improvement and/or weight loss benefit from GLP1s/tirzepatide. should we perform a dexamethasone suppression test and treatment if positive in these individuals??? should we do that even if diabetes is not present?

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

to get access to the dropbox: https://www.dropbox.com/scl/fo/pcf4vymtpo77f49wd2ljr/ANyJqB40CuBd-DMEr67XDoI?rlkey=5dsalmuhf8ekd0svsgvn93xm1&dl=0

or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them