Mifepristone-Induced Hypothyroidism

Abstract

Medications are known to affect the thyroid physiology and are a known cause of hypothyroidism. There is an ever-growing list of medications that affect the thyroid by 1 or more mechanisms. Mifepristone is presently used for the treatment of mild autonomous cortisol secretion (MACS). Hypothyroidism is not a known side effect of this medication. We present a 71-year-old woman with newly diagnosed impaired fasting glucose, dyslipidemia, and osteopenia presenting with a 3-year history of unintentional 15-pound weight gain (despite exercise and a good diet) and increased anxiety. Her physical examination was pertinent for mild lower extremity edema, easy bruising, and skin thinning. Workup revealed adrenocorticotropic hormone (ACTH)-independent MACS from bilateral micronodular hyperplasia of the adrenals. Since she was not a surgical candidate, medical management with mifepristone was chosen. While on mifepristone, she complained of excessive fatigue, a workup done revealed new-onset hypothyroidism. Given her symptoms and bloodwork, she was started on levothyroxine. After stopping mifepristone, she was biochemically and clinically euthyroid and was eventually off levothyroxine. The mechanism by which mifepristone induces hypothyroidism is unknown. Except for a multicenter case series suggesting that mifepristone increases thyroid hormone requirements in patients with central hypothyroidism, to the best of our knowledge, the literature on euthyroid patients developing hypothyroidism secondary to mifepristone is scarce. In conclusion, while the hypothyroidism seems reversible our case highlights the importance of getting baseline thyroid function tests (TFTs) and repeating them while on the medication. Treatment of hypothyroidism is based on symptoms and bloodwork.

Introduction

Medications frequently affect the thyroid axis. The mechanism by which different medications could affect the thyroid is as follows: (1) inhibition of synthesis and/or release of thyroid hormones, (2) immune mechanisms, (3) drug-induced thyroiditis, (4) inhibition of thyroid-stimulating hormone (TSH), or (5) treatment interactions leading to decreased levothyroxine (LT4) absorption or alteration in transport and metabolism of thyroid hormones. ¹ Some of the examples of medications that cause hypothyroidism through 1 or more of the mechanisms mentioned above are amiodarone, lithium, glucocorticoid, bromocriptine, anti-interferon alpha (eg, infliximab), tyrosine kinase inhibitors (eg, imatinib), anti-PD1 (eg, pembrolizumab), and anti-CTLA4 (eg, Ipilimumab). ¹ Mifepristone could also be added to this list.

In 2012, the US Food and Drug Administration approved mifepristone to mainly control hyperglycemia secondary to Cushing syndrome (CS) in patients with type 2 diabetes mellitus (DM) or glucose intolerance and who have failed surgery or are not surgical candidates. ² Presently, it is also used to treat MACS.

Mild autonomous cortisol secretion is characterized as an ACTH-independent cortisol secretion by an adrenal adenoma/macronodular hyperplasia (very rarely micronodular hyperplasia) that is insufficient to produce typical signs or symptoms of CS. It rarely progresses to overt CS, but these patients have increased rates of type 2 DM, hypertension (HTN), dyslipidemia, obesity, cardiovascular events, metabolic bone disease (osteopenia, osteoporosis, and fractures), and increased mortality. ³ Hence, treatment is necessary, which could be surgical or medical. One of the medications used for management is mifepristone. The most common adverse effects (AEs) are nausea, fatigue, headache, hypokalemia, arthralgias, peripheral edema, vomiting, HTN, dizziness, decreased appetite, and endometrial thickening in women. ⁴ Hypothyroidism is not a commonly known AE.

Here, we report a case of hypothyroidism secondary to mifepristone use for the treatment of ACTH-independent MACS from micronodular hyperplasia Graph 1 and Table 1.

Graph 1.

Graphical representation of TFT in relation to mifepristone.

Abbreviation: TFT, thyroid function test.

Table 1.

Changes in the TFT of the Patient Before, During, and After Stopping Mifepristone.

Mifepristone use	Time period	TSH (0.45-4.5 uIU/mL)	FT4 (4.5-12 μg/dL)
Before medication	August 2021	2.6	5.7
2-months on medication	October 2021	8.3	5.1
1-month off medication	November 2021	9.1	5.7
3-months off medication	January 2022	3.6	6.6

Abbreviations: TFT, thyroid function test; TSH, thyroid-stimulating hormone; FT, function test.

Case Description

Seventy-one-year-old woman with newly diagnosed impaired fasting glucose, dyslipidemia, and osteopenia initially presented with a 3-year history of unintentional weight gain of 15 pounds despite a good diet and physical activity, as well as increased anxiety. Her family history was positive for hypothyroidism in her sibling. Physical examination was relevant for mild lower extremity edema, easy bruising, and skin thinning, but negative for supraclavicular or dorsocervical fat pads, or abdominal striae. Given her symptomatology, there was a concern for CS, for which a workup was initiated. Blood work revealed a cortisol level after 1 mg dexamethasone suppression test (DST) of 2.8 (<1.8 μg/dL), ACTH of 2.3 (7.2-63.3 pg/mL), and 24 hours urinary free cortisol of 42.1 (4-50 mcg/24 h), and low DHEAS. Furthermore, bloodwork suggested hemoglobin A_1c (HbA_1c) of 5.8% (<5.6%), total cholesterol of 297 (100-199 mg/dL), low-density lipoprotein (LDL) of 191 (0-99 mg/dL), and TSH of 2.2 (0.45-4.5 uIU/mL). Magnetic resonance imaging of the abdomen and pelvis showed normal-appearing adrenals with no suggestion of adenoma. The working diagnosis was ACTH-independent MACS from bilateral micronodular hyperplasia. She wanted a second opinion at the Mayo Clinic, where the tests were repeated. Similar results (DST >1.8 μg/dL and low ACTH <5 pg/mL, low Dehydroepiandrosterone sulfate (DHEAS), and no adrenal adenoma in imagining) confirmed the diagnosis. She was not a surgical candidate; therefore, the decision was taken to start therapy with metyrapone 250 mg daily.

After 3 months of treatment, she had a weight loss of 10 pounds and an improvement in her HbA_1c: 5.3% and LDL: 69 mg/dL. She was receiving the medication from Europe with difficulty. Therefore, she wanted an alternative; hence, we discussed switching her to mifepristone. She decided to continue metyrapone for 2 years; however, her symptoms returned while on the medication. Therefore, it raised concerns about the effectiveness of the medication and the possibility of increasing the dose. Due to these reasons and the difficulty in obtaining the medication, the discussion of switching to mifepristone was re-entertained. Bloodwork was repeated before starting mifepristone, and the AEs of the medication were discussed with her. Bloodwork suggested basal cortisol of 12 (6.2-19.4 μg/dL), ACTH of <5, potassium of 5 (3.5-5.2 mmol/L), HbA_1c of 5.9%, LDL of 159 mg/dL, TSH of 2.6 uIU/mL, and free T4 (FT4) of 5.7 (4.5-12 μg/dL), and thyroid peroxidase (TPO) and thyroglobulin (Tg) antibodies were negative. Clinically and biochemically, it suggested a recurrence of her MACS. Mifepristone was started at 200 mg (2 tablets) daily. The physical examination showed a weight loss of 10 pounds, a decrease in swelling of the lower extremities, and a decrease in skin thinning. At the second month of follow-up, she complained of feeling more tired than usual. Her repeated bloodwork revealed 8 am basal cortisol of 5.7 (5-25 μg/dL), HbA_1c of 5.6%, potassium of 4.3 mmol/L, elevated TSH of 8.3 uIU/mL, and FT4 of 5.1 μg/dL. The differential for fatigue was the AE of the medication that caused adrenal insufficiency or new-onset hypothyroidism. We decided to repeat the blood work in 1 week again to check the thyroid function test (TFT) with the TPO antibodies and electrolytes. Her symptoms continued to worsen, and the repeat TSH of 8.9 uIU/mL with negative TPO was suggestive of new-onset hypothyroidism; thus, 50 mcg daily of LT4 was started with a follow-up in 1 month. Mifepristone was held since there was a high suspicion that this medication triggered her new-onset hypothyroidism. On follow-up, she continued with fatigue and repeat TSH now 9.1 uIU/mL. At this time, the dose of LT4 was increased to 75 mcg daily. Five weeks later, she reported feeling better, and her TSH levels normalized to 3.6 uIU/mL. The diagnosis of mifepristone-induced hypothyroidism was confirmed. After months of LT4 treatment, we were able to decrease the dose and eventually discontinued LT4. In her last office visit, she continued to be clinically and biochemically euthyroid without LT4 but presented with some clinical features of hypercortisolism again.

Discussion

Mild autonomous cortisol secretion (previously known as Subclinical Cushing’s) is found in 5% to 20% of adrenal incidentaloma and has an estimated prevalence of 79 cases per 100 000 persons. ⁵ It is more common than classic CS. Patients without obvious signs of Cushing but with a value of >1.8 mcg/dL on 1 mg DST are diagnosed with MACS. ⁶ The diagnosis is based on the evaluation of the hypothalamic-pituitary-adrenal axis. The American Association of Clinical Endocrinologists (AACE), the Endocrine Society guidelines, and the European Society of Endocrinology recommend the use of 1 mg DST with a cutoff <1.8 μg/dL of cortisol as a primary test to rule out autonomous cortisol secretion, while >5 μg/dL to confirm it. Further testing with morning ACTH/DHEAS is required to confirm the diagnosis and to assess the degree of cortisol secretion. ⁷ Treatment is important given its increased association with cardiovascular events, metabolic disease, bone disease, and increased mortality. Management is usually surgical resection of functioning adenomas or adrenalectomy (bilateral or unilateral), but in patients with bilateral macronodular or micronodular adrenal hyperplasia surgery may not be feasible and in such cases, medical management is advisable. Nowadays, medical therapy includes adrenal steroidogenesis-inhibiting agents such as ketoconazole, metyrapone, mitotane, etomidate, osilodrostat, relacorilant, and glucocorticoid receptor (GR) blockers like mifepristone. ⁸ Our patient was initially on metyrapone which was switched to mifepristone due to concerns about the medication’s effectiveness given the recurrence of symptoms and difficulty obtaining the medication in the United States.

Mifepristone was shown to be beneficial in the treatment of CS in the 6-month, multicenter phase 3 Study of the Efficacy and Safety of Mifepristone in the Treatment of Endogenous Cushing Syndrome (SEISMIC) trial. About 60% of the patients had significant improvement in glycemic control; 38% had a reduction in diastolic blood pressure; and 87% of patients had an overall improvement in clinical status.^4,8 A retrospective case series study by Cohan et al ⁹ suggested mifepristone as a safe and effective treatment option for patients with hypercortisolism from bilateral adrenal macronodular hyperplasia. Therefore, we chose this medication for our patient with micronodular hyperplasia.

Mifepristone is a synthetic steroid. It is a competitive GR antagonist and selective progesterone receptor (PR) antagonist. At low doses, it blocks PR while at higher doses it blocks GR. It has little to no affinity for mineralocorticoid receptors and has 3 times more affinity for GR than dexamethasone.^4,8 Competitive antagonism causes cortisol levels to increase, but its effects are blocked. Hence, dose adjustments are primarily based on clinical assessment of signs or symptoms (blood pressure, weight, blood glucose, waist circumference, etc) and tolerability. It has a rapid onset of action and is prescribed with a maximum daily dose of 1200 mg. Given its long half-life of 85 hours, this medication can be prescribed once a day. ⁸ It is metabolized in the liver primarily by cytochrome P450 3A4 (CYP3A4); therefore, medication reconciliation is extremely important to check drug-drug interactions, especially in patients who take medications that are CYP3A4 inhibitors. ²

In the SEISMIC trial, AEs were reported in 88% of the patients on the medication. The most common AEs reported are nausea (48%), fatigue (48%), headache (44%), hypokalemia (34%), arthralgia (30%), vomiting (26%), peripheral edema (26%), HTN (24%), dizziness (22%), decreased appetite (20%), and thickening of the endometrium in women (20%). Apart from these common side effects, this medication appeared to have altered thyroid function. In the trial, 19% (8/42) of patients had a reversible increase in TSH. ⁴ Of the common side effects, our patient did have nausea, hypokalemia, and fatigue, but it was the worsening fatigue that prompted a further workup which revealed new-onset hypothyroidism.

The mechanism by which mifepristone affects thyroid function has not been fully understood, but in a small retrospective longitudinal case series of patients with Cushing disease and central hypothyroidism treated with mifepristone, it was noted that the median dose of LT4 increased 1.83 times the initial dose to achieve normal FT4. ¹⁰ A study done by Heikinheimo et al ¹¹ showed that the TSH values increased most significantly (with no detection of thyroid antibody) within the first 3 months of mifepristone therapy. This was also accompanied by a transient decrease in T4 around the same time, which was speculated to be due to the direct effect of the drug on the thyroid. The study also suggested that long-term treatment with this medication could reset the pituitary-thyroid axis. ¹¹

Regulation of TSH secretion appears to be affected by the permissive action of glucocorticoids. For example, in Addison’s disease, glucocorticoid deficiency is associated with elevated TSH levels, while exogenous glucocorticoid or hypercortisolism likely suppresses TSH and abolishes the nocturnal surge of TSH.^11-13 The thyroid receptor (TR) shares the same nuclear receptor superfamily as progesterone and glucocorticoid, but neither mifepristone nor its metabolites bind to TR in vitro. ¹¹ A study by Takiyama et al ¹⁴ showed that hydrocortisone increases iodide uptake and releases organified iodine by porcine thyroid cells in vitro, suggesting a direct effect on the thyroid gland while mifepristone inhibited the stimulatory actions of hydrocortisone on iodide uptake without a direct effect on iodide uptake or release. Based on this understanding of the likely effect of glucocorticoids on thyroid hormone homeostasis, elevated TSH is likely due to hypothalamic and pituitary GR inhibition and could likely also be in addition to reduced thyroid hormone secretion.^2,11

Hypercortisolism could mask preexisting autoimmune thyroid disease and hypothyroid symptoms, while treatment with mifepristone can unmask these symptoms/diseases.^2,15 Therefore, the recommendations are as follows: (1) all patients should get a baseline TFT before starting therapy; (2) a baseline TPO antibody is not suggested unless indicated; and (3) in case of suspected autoimmune disease and negative antibody, ultrasound may be helpful. Treatment is based on TFTs and symptoms. In case of low T4 and low or normal TSH treat with thyroid medications and continue to monitor FT4 during mifepristone therapy to make dose adjustments accordingly. In the case of only elevated TSH, it will be based on symptoms, and the treatment threshold is based on AACE/ATA (American Thyroid Association) guidelines. ² In our case, the patient’s elevated TSH and the worsening fatigue prompted us to start treatment with LT4, which helped alleviate her symptoms. Once mifepristone was stopped, we eventually stopped the LT4 since her TFTs returned to baseline and there were no signs or symptoms suggestive of hypothyroidism.

Conclusion

In conclusion, our case raises awareness that mifepristone can lead to new-onset hypothyroidism in a previously euthyroid patient. Thus, reiterating the importance of getting baseline TFT before starting this medication and monitoring the bloodwork while on the medication. The treatment of hypothyroidism will be based on the symptoms and bloodwork.