Ketogenic Diet Initially Masks Symptoms of Hypercortisolism in Cushing’s Disease

Abstract

Cushing’s syndrome (CS) is a diagnosis used to describe multiple causes of serum hypercortisolism. Cushing’s disease (CD), the most common endogenous subtype of CS, is characterized by hypercortisolism due to a pituitary tumor secreting adrenocorticotropic hormone (ACTH). A variety of tests are used to diagnose and differentiate between CD and CS. Hypercortisolism has been found to cause many metabolic abnormalities including hypertension, hyperlipidemia, impaired glucose tolerance, and central adiposity. Literature shows that many of the symptoms of hypercortisolism can improve with a low carb (LC) diet, which consists of consuming <30 g of total carbohydrates per day. Here, we describe the case of a patient with CD who presented with obesity, hypertension, striae and bruising, who initially improved some of his symptoms by implementing a LC diet. Ultimately, as his symptoms persisted, a diagnosis of CD was made. It is imperative that practitioners realize that diseases typically associated with poor lifestyle choices, like obesity and hypertension, can often have alternative causes. The goal of this case report is to provide insight on the efficacy of nutrition, specifically a LC diet, on reducing metabolic derangements associated with CD. Additionally, we will discuss the importance of maintaining a high index of suspicion for CD, especially in those with resistant hypertension, obesity and pre-diabetes/diabetes.

1. Introduction

Cushing’s syndrome (CS) is a rare disorder of hypercortisolism related to exposure to high levels of cortisol (>20 mcg/dL between 0600–0800 or >10 mcg/dL after 1600) for an extended period [1,2]. CS affects 10 to 15 people per million and is more common among those with diabetes, hypertension, and obesity [3]. The metabolic derangements associated with CS include visceral obesity, elevated blood pressure, dyslipidemia, type II diabetes mellitus (T2DM) and insulin resistance [4]. CS physical exam findings include round face, dorsal fat pad, central obesity, abdominal striae, acne, and ecchymosis [3]. Other symptoms associated with CS include low libido, headache, change in menses, depression and lethargy [2,3,5]. The most common features of CS are weight gain, which is found in 82% of cases, and hypertension, which is found in 50–85% of cases [6]. CS can be caused by exogenous glucocorticoids, known as iatrogenic CS, ectopic ACTH secretion (EAS) from sources like a small cell lung cancer or adrenal adenoma, known as EAS CS, or excess production of ACTH from a pituitary tumor, known as CD [3]. In CD, ACTH subsequently causes increased production of cortisol from the adrenal glands. CD accounts for 80–85% of endogenous cases of CS [3]. Other conditions including alcoholism, depression, severe obesity, bulimia and anorexia nervosa can lead to a Cushing-like state, although are not considered true CS [3]. Many studies have demonstrated that LC diets can ameliorate some of the most common metabolic derangements seen in CD, namely hyperglycemia, weight gain, hypertension and insulin resistance.
A LC diet is a general term for diets which lower the total carbohydrates consumed per day [4]. A ketogenic diet is a subtype of LC that is described as having even fewer carbohydrates, typically less than 30 g/day. By reducing carbohydrate intake and thus limiting insulin production, the body achieves ketosis by producing an elevated number of ketones including β-hydroxybutyric acid, acetoacetic acid, and acetone, in the blood [7]. A carnivore diet, a specific type of a ketogenic diet, is defined as mainly eating animal food such as meat, poultry, eggs and fish. Contrarily, a standard American diet (SAD) is defined as a diet high in processed foods, carbs, added sugars, refined fats, and highly processed dairy products [8]. There are several therapeutic applications for LC diets that are currently supported by strong evidence. These include weight loss, cardiovascular disease, T2DM, and epilepsy. LC diets have clinical utility for acne, cancer, polycystic ovary syndrome (PCOS), and neurologic deficits [9].
In this case report, the patient endorsed initially starting a LC diet to address weight gain and high blood sugars that he noted on a glucometer. The patient noted a 35 pounds (lbs.) weight loss over the first 1.5 years on his LC diet, as well as improved blood pressure and in his overall health. He then adopted a carnivore diet but found that weight loss was difficult to maintain, although his body composition continued to improveand his clothes fit better. Later, he noted that his blood pressure would at times be poorly controlled despite multiple medications and strict dietary adherence. The patient reported “being in despair” and “not trusting his doctors” because they did not understand how much his diet had helped him. Despite strict adherence, his symptoms of insulin resistance and hypertension persisted. In this report, we will describe how his symptoms of CD were ameliorated by the ketogenic diet. This case report also highlights that when patients are unable to overcome hormonal pathology, clinicians should not blame patients for lack of adherence to a diet, but instead understand the need to evaluate for complex pathology.

2. Detailed Case Description

A male patient in his thirties, of Asian descent, had a past medical history of easy bruising, central obesity, headaches, hematuria, and hypertension and past family medical history of hypertension in his father and brother. In 2015, he was at his heaviest weight of 179 lbs. with a body mass index (BMI) of 28 kg/m2, placing him in the overweight category (25.0–29.9 kg/m2). At that time the patient reported he was following a SAD diet and was active throughout the day. The patient stated he ate a diet of vegetables, fruits and carbohydrates, but he was not able to lose weight. The patient stated that he switched to a LC diet, to address weight gain and hyperglycemia, and he reported that he lost approximately 35 lbs. in 1.5 years. The patient described his LC diet as eating green leafy vegetables, low carb fruits, fish, poultry, beef and dairy products. The patient then later switched to a carnivore diet. He noted despite aggressively adhering to his diet, that his weight-loss had plateaued, although his waist circumference continued to decrease. The patient noted his carnivore diet consisted of eating a variety of different meats, poultry, fish and eggs.
The metabolic markers seen in Table 1 were obtained after the patient had started a carnivore diet. The patient’s blood glucose levels decreased overtime despite impaired glucose metabolism being a known side effect of hypercortisolism [4]. The patient’s high-density lipoprotein (HDL) remained in a healthy range (40–59 mg/dL) and his triglycerides stayed in an optimal range (<100 mg/dL), despite dyslipidemia being a complication of CD [4]. When the patient was consuming a SAD diet, he was not under the care of a physician and was unable to provide us with previous biomarkers.
Table 1. Patient’s metabolic markers on a carnivore diet. Glucose (70 to 99 mg/dL), total cholesterol (desirable <200 mg/dL, borderline high 200–239 mg/dL, high >239 mg/dL), triglycerides (optimal: <100 mg/dL), HDL (low male: <40 mg/dL), low density lipoprotein (LDL) (Optimal: <100 mg/dL).
Table
Despite strict adherence to his diet and initial improvement in his weight, his blood pressure and his blood sugar levels, in October of 2021 the patient was admitted to the hospital for hypertensive urgency, with a blood pressure of 216/155. His complaints at the time were unexplained ecchymosis, hematuria and significant headaches that were resistant to Excedrin (acetaminophen-aspirin-caffeine) use. At the hospital, the patient underwent a computed tomography (CT) scan of the head and radiograph of the chest, and both images were negative for acute pathology. During his hospital admission, the patient denied any changes in vision, chest pain or edema of the legs. Ultimately, the patient was told to eat a low-salt diet and to follow-up with a cardiologist. At discharge, the patient was placed on hydrochlorothiazide, labetalol, amlodipine and lisinopril. The patient was then seen by his primary care physician in November of 2021 and his urinalysis at that time showed 30 mg/mL (Negative/Trace) of protein in his urine, without hematuria. The patient’s primary care physician discontinued his hydrochlorothiazide and started the patient on furosemide. Additionally, the primary care physician reinforced cutting out salt and limiting his calories to prevent any further weight gain, which his physician explained would contribute further to his hypertension. He was referred to hematology and oncology in November of 2021 for his symptoms of hematuria and abnormal ecchymosis to his abdomen, thighs and arms. The patient’s coagulation and platelet counts were normal, and his symptoms were noted to be improving. His hematuria and ecchymosis were attributed to his significant Excedrin use from the past 1–2 months, secondary to his headaches, and their anti-platelet effect. It was noted that the patient had significant hemolysis during his hospital admission. However, in his follow up examination, there were no signs of hemolysis, and it was attributed to his hypertensive urgency. Again, a low-salt, calorie-limited diet was recommended. The patient was referred to cardiology where he was evaluated for secondary hypertension, because despite his weight loss and his strict adherence to his diet, his blood pressure was still uncontrolled on multiple medications. He had a normal echocardiogram and renal ultrasound which showed no signs of renal artery stenosis bilaterally. At that time the patient’s serum renin, aldosterone and urine metanephrine levels were all normal. His cardiologist increased his lisinopril, and continued him on amlodipine, furosemide and labetalol and reinforced the recommendations of lowering his salt and preventing weight gain.
The patient first contacted our office in January of 2022. At that time his blood pressure was noted to be 160/120 despite being compliant with current blood pressure medications. The patient reported strict adherence to his carnivore diet by sharing his well-documented meals on his social media accounts. Given the persistent symptoms, despite his significant change in diet and weight loss, we were concerned that a hormonal etiology may be driving his symptoms. The patient was seen in-person, in our office, in March of 2022. At the request of the patient, we again reviewed his social media profile to assess his meal choices and diet. While the patient was eager to show us his carnivore meals, what we incidentally noted in his photos was despite weight loss and strict diet adherence, he had developed moon facies (Figure 1a,b). On the physical exam, we noted his prominent abdominal striae (Figure 2). Several screening tests for Cushing’s syndrome were ordered. A midnight salivary cortisol was ordered, with values of 0.884 ug/dL (<0.122 ug/dL) and 0.986 ug/dL (<0.122 ug/dL) and a urinary free cortisol excretion (UFC) was ordered, with values of 8.8 ug/L (5–64 ug/L). At this point our suspicion was confirmed that the patient had inappropriately elevated cortisol.
Metabolites 12 01033 g001 550
Figure 1. The patient’s progression of moon facies, (a) photo from 2019 after initial weight loss (b) photo from office visit in 2022.
Metabolites 12 01033 g002 550
Figure 2. The arrows demonstrate early striae visualized on the lower abdomen bilaterally, unclear in image due to poor office lighting.
Based on screening tests and significant physical exam findings, we referred the patient to endocrinology for a low dose dexamethasone suppression test (DST). They performed a low dose DST revealing a dehydroepiandrosterone (DHEA) of 678 ug/dL (89–427 ug/dL) and ACTH of 23.9 pg/mL (7.2–63.3 pg/mL). The low dose DST and midnight salivary cortisol were both positive indicating hypercortisolism. To begin determining the source of hypercortisolism, the plasma ACTH was evaluated and was 27.2 pg/mL (7.2–63.3 pg/mL). While ACTH was within normal range, a plasma ACTH > 20 pg/mL is suggestive of ACTH-dependent CS, so a magnetic resonance imaging (MRI) of the brain was ordered [2]. The MRI revealed a 4 mm heterogeneous lesion in the central pituitary gland which is suspicious of a cystic microadenoma. To confirm that a pituitary tumor was the cause of the patient’s increased cortisol, the patient was sent for inferior petrosal sinus sampling (IPSS). The results of the IPSS indicated an increase in ACTH in both inferior petrosal sinuses and peripheral after corticotropin-releasing hormone (CRH) stimulation (Figure 3a–c), which was consistent with hypercortisolism.
Metabolites 12 01033 g003a 550Metabolites 12 01033 g003b 550
Figure 3. (a) Right IPS venous sampling values for ACTH and prolactin after CRH stimulation over multiple time intervals. (b) Left IPS venous sampling values for ACTH and prolactin after CRH stimulation over multiple time intervals. (c) Peripheral sampling values for ACTH and prolactin after CRH stimulation over multiple time intervals.
Lab results from the patient’s IPSS venous sampling can be seen above. The graphs depict the lab values of ACTH (7.2–63.3 pg/mL) and prolactin (PRL) (2.1–17.7 ng/mL) before and after CRH stimulation during IPSS. PRL acts as a baseline to indicate successful catheterization in the procedure [10].
Using the ACTH levels from our patient’s IPSS we calculated a ratio of inferior petrosal sinus to peripheral (IPS:P). These results can be seen below (Table 2). The right IPS:P was calculated as 3.60 at 10 min and the left IPS:P as 7.65 at 10 min. These ratios confirmed that the hypercortisolism was due to the pituitary tumor, as it is higher than the 3:1 ratio necessary for diagnosis of CD [11]. The patient is currently scheduled to undergo surgical resection of the pituitary microadenoma.
Table 2. Right and left petrosal sinus to peripheral serum ACTH ratios.
Table

3. Clinical Evaluation for CS

In this case, the patient presented with uncontrolled hypertension, weight gain despite a strict diet, hyperglycemia, abdominal striae and moon facies. Despite evaluation, both inpatient and outpatient, a diagnosis of CS was not yet explored. When CS is suspected based on clinical findings, the use of exogenous steroids must first be excluded as it is the most common cause of hypercortisolism [3]. If there is still concern for CS, there are three screening tests that can be done which are sensitive but not specific for hypercortisolism. The screening tests include: a 24-h UFC, 2 late night salivary cortisol tests, low dose (1 g) DST [3]. To establish the preliminary diagnosis of hypercortisolism two screening tests must be abnormal [2].
The first step to determine the cause of hypercortisolism is to measure the plasma level of ACTH. Low values of ACTH < 5 pg/mL indicate the cause is likely ACTH-independent CS and imaging of the adrenal glands is warranted as there is a high suspicion of an adrenal adenoma [2,3]. When the serum ACTH is elevated >/20 pg/mL it is likely an ACTH-dependent form of CS [2]. To further evaluate an ACTH-dependent hypercortisolism, an MRI should be obtained as there is high suspicion that the elevated cortisol is coming from a pituitary adenoma. If there is a pituitary mass >6 mm there is a strong indication for the diagnosis of CD [2]. However, pituitary tumors can be quite small and can be missed on MRIs in 20–58% of patients with CD [2]. If there is still a high suspicion of CD with an inconclusive MRI, a high dose DST (8 g) is done. Patients with CD should not respond and their ACTH and DHEA, a steroid precursor, should remain high. Similarly, CRH stimulation test is done and patients with CD should have an increase in ACTH and/or cortisol within 45 min of CRH being given. If the patient has a positive high-dose DST, CRH-stimulation test and an MRI with a pituitary tumor >6 mm no further testing is needed as it is likely the patient has CD [2]. If either of those tests are abnormal, the MRI shows a pituitary tumor < 6 mm, or there is diagnostic ambiguity, the patient should undergo IPSS with ACTH measurements before and after the administration of CRH [4]. IPSS is the gold standard for determining the source of ACTH secretion and confirming CD. In this invasive procedure, ACTH, prolactin, and cortisol levels are sampled prior to CRH stimulation and after CRH stimulation. PRL acts as a baseline to indicate successful catheterization in the procedure [12]. To confirm CD, a ratio of IPS:P is calculated for values prior to and after CRH stimulation. A peak ratio greater than 2.0 before CRH stimulation or a peak ratio greater than 3.0 after CRH stimulation is indicative of CD. In comparing the right and left petrosal sinus sample, an IPS:P ratio greater than 1.4 suggests adenoma lateralization. However, due to high variability, IPSS should not be used for diagnosing lateralization [13].

4. Discussion

Surgical intervention remains the primary treatment for CD [4]. However, remission is not guaranteed as symptoms and metabolic diseases have been shown to persist afterwards. In the literature it has been shown that nutrition can have a powerful impact on suppressing, or even reversing metabolic disorders and comorbidities associated with CD. A LC diet has been shown to promote significant weight loss, reduce hypertension, improve dyslipidemia, reverse T2DM and improve cortisol levels (2, 14–15, 18–21).
There are reports of weight loss on a LC diet in the literature. A LC significantly reduced weight and BMI of 30 male subjects [14]. In a group of 120 participants over 24 weeks who followed a LC versus low fat (LF) diet, showed a greater weight loss in the LC group vs. the LF group [15]. Patients diagnosed and treated for CD found that their weight remained largely unchanged even after treatment [6]. In many cases, surgical treatment does not always resolve the associated comorbidity of central adiposity in CD. In such cases, a LC diet can be used before, during and after treatment, as an adjunct, to decrease associated weight gain and comorbidities.
Nutritional intervention can be a powerful adjunct to reduce comorbidities associated with CD. As seen in this case report, the patient’s symptoms of CD, especially hypertension and weight gain, improved with dietary changes despite him having a pituitary microadenoma. Multiple studies showed that a LC diet was able to decrease blood pressure parameters. In a group of 120 participants over 24 weeks who followed a LC versus a LF diet showed a greater decrease in both systolic and diastolic blood pressure in the LC group vs. the LF group [15]. Other literature which studied the effect of a LC diet on hypertension demonstrated the reduction of blood pressure and is thought to be due to ketogenesis. It is thought the production of ketones have a natriuretic effect on the body therefore lowering systemic blood pressure [16].
A LC diet improves lipid profiles and inflammatory markers associated with metabolic syndrome [14]. Literature shows that a LC diet has a greater impact on decreasing triglyceride levels and increasing HDL levels, when compared to a LF diet [15]. Triglyceride levels in patients in CD remission remained high [17]. Therefore, it can be hypothesized that a LC diet would be beneficial, in addition to standard CD treatment, to lower the associated comorbidity of hypertriglyceridemia and metabolic syndrome.
Insulin resistance, a precursor to T2DM, is a common comorbidity of hypercortisolism which can be treated with a LC diet. One study showed that in subjects with T2DM, a decrease in A1c and a reduction in antidiabetic therapy were seen with consumption of a LC diet [18]. Additionally, a cohort of 9 participants following a LC diet were able to collectively lower their A1c on average by 1% while concurrently discontinuing various antidiabetic therapies including insulin [19].
Literature shows that a LC diet can minimize systemic cortisol levels through various mechanisms. Current treatment of CD includes medications which block cortisol production and/or cortisol secretion [2]. LC can imitate similar results seen through medication intervention for CD. Carbohydrate restriction can lower cortisol levels, as carbohydrates stimulate adrenal cortisol secretion and extra-adrenal cortisol regeneration [4]. A ketogenic diet can lower the level of ghrelin, a peptide produced in the stomach that has orexigenic properties [20,21]. Literature shows that ghrelin increases levels of serum cortisol [22]. Therefore, implementing a ketogenic diet would decrease ghrelin, and subsequently minimize the effects of increased ghrelin on serum cortisol. A LC diet decreases visceral fat which itself is an endocrine organ and can increase the synthesis of cortisol [14]. Therefore, decreasing visceral fat also decreases the production of cortisol. A LC was shown to significantly reduced weight, BMI and cortisol levels of 30 obese male subjects [14]. Further, a LC diet excludes foods with a high glycemic index which cause increased stress on the body which subsequently leads to the activation of the hypothalamic-pituitary-axis which causes increased levels of cortisol [14].
This case report illustrated how a LC diet was initially successful at ameliorating the patient’s associated symptoms of hypertension and obesity, making his diagnosis of CD go undetected. Literature shows that while the prevalence of CS on average is a fraction of a percent, it is much higher among patients with poorly controlled diabetes, hypertension and early onset osteoporosis [3]. Two hundred patients with diabetes mellitus were studied and 5.5% were found to have CS [23]. Another study discovered that in subjects with CD, 36.4% were found to have hyperlipidemia, 73.1% with hypertension, and 70.2% with impaired glucose metabolism [17]. It can be concluded that a higher index of suspicion and lower threshold for screening for CS may be necessary in obese and diabetic patient populations. A lower threshold for screening can allow for earlier diagnosis for many patients, and therefore provide better outcomes for those diagnosed with CS.
It is important for clinicians to consider alternative pathology for patients combating metabolic derangements. As depicted in this case, the patient lost 35 lbs. while on a LC diet, despite having hypercortisolism, presumably for months to years prior to the diagnosis of his condition. The patient noted a tendency to gain weight, have elevated blood sugar and blood pressure which prompted him to begin self-treatment with increasingly strict carbohydrate restriction. The patient was able to keep his symptoms of hypercortisolism managed, potentially making the diagnosis difficult for his team of clinicians. From a diagnostic perspective, it’s important to understand that strict dietary adherence can have profound impacts on even the most severe hormonal pathology. Ultimately, this case serves as a reminder of the power of nutrition to address metabolic derangements and simultaneously as a reminder to diagnosticians to never rely on lack of dietary adherence as a reason for persistent metabolic symptoms. The reflexive advice to “not gain weight” and “lower salt intake” in retrospect appears both dogmatic and careless. In this case, the patient had seen several doctors and was even hospitalized and yet his disease state remained unclear and the dietary messaging cursory.

5. Conclusions

Many chronic diseases, including diabetes, hypertension and obesity, are generally thought to be caused by dietary and lifestyle choices. However, as exemplified in this report underlying medical problems, such as endocrine disorders, can be the cause of such metabolic derangements. It is critical that practitioners consider other causes of metabolic derangements, as assuming that they are due to poor dietary adherence, can allow them to go undiagnosed. While there is extensive literature on LC diets and their effect on the metabolic derangements associated with hypercortisolism, there needs to be further research on LC as an adjunctive therapy to conventional CD treatment. Ultimately, nutrition can have a powerful impact on suppressing, or even reversing metabolic disorders. As depicted in this case study, a LC diet is powerful enough to temporarily suppress symptoms of CD.

Author Contributions

M.K.D., E.-C.P.-M. and T.K. equally contributed to this case report. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Written informed consent has been obtained from the patient to publish this paper.

Data Availability Statement

The data presented in this study are available in article.

Acknowledgments

We would like to thank our patients and the Society of Metabolic Health Practitioners.

Conflicts of Interest

T.K. is an unpaid member of the Board of Directors of the Society of Metabolic Health Practitioners and a producer of podcasts on health and nutrition, with all proceeds donated to humanitarian charities; his spouse has ownership interest in a food company. The other author reports no conflicts of interest.

References

  1. Nieman, L.K. UpToDate. Available online: https://www.uptodate.com/contents/measurement-of-cortisol-in-serum-and-saliva?search=cortisol%20level&source=search_result&selectedTitle=1~150&usage_type=default&display_rank=1 (accessed on 27 September 2022).
  2. Feelders, R.; Sharma, S.; Nieman, L. Cushing’s Syndrome: Epidemiology and Developments in Disease Management. Clin. Epidemiol. 2015, 7, 281. [Google Scholar] [CrossRef] [PubMed]
  3. Guaraldi, F.; Salvatori, R. Cushing Syndrome: Maybe Not so Uncommon of an Endocrine Disease. J. Am. Board Fam. Med. 2012, 25, 199–208. [Google Scholar] [CrossRef] [PubMed]
  4. Guarnotta, V.; Emanuele, F.; Amodei, R.; Giordano, C. Very Low-Calorie Ketogenic Diet: A Potential Application in the Treatment of Hypercortisolism Comorbidities. Nutrients 2022, 14, 2388. [Google Scholar] [CrossRef] [PubMed]
  5. Nieman, L.K. UpToDate. Available online: https://www.uptodate.com/contents/epidemiology-and-clinical-manifestations-of-cushings-syndrome?search=cushings%20diagnosis%20symptoms&source=search_result&selectedTitle=2~150&usage_type=default&display_rank=2 (accessed on 27 September 2022).
  6. Schernthaner-Reiter, M.H.; Siess, C.; Gessl, A.; Scheuba, C.; Wolfsberger, S.; Riss, P.; Knosp, E.; Luger, A.; Vila, G. Factors Predicting Long-Term Comorbidities in Patients with Cushing’s Syndrome in Remission. Endocrine 2018, 64, 157–168. [Google Scholar] [CrossRef] [PubMed]
  7. Giordano, C.; MarchiÃ2, M.; Timofeeva, E.; Biagini, G. Neuroactive Peptides as Putative Mediators of Antiepileptic Ketogenic Diets. Front. Neurol. 2014, 5, 63. [Google Scholar] [CrossRef]
  8. Standard American Diet (SAD). Available online: https://piviohealth.com/knowledge-bank/glossary/standard-american-diet-sad/ (accessed on 2 October 2022).
  9. Paoli, A.; Rubini, A.; Volek, J.S.; Grimaldi, K.A. Beyond Weight Loss: A Review of the Therapeutic Uses of Very-Low-Carbohydrate (Ketogenic) Diets. Eur. J. Clin. Nutr. 2013, 67, 789–796. [Google Scholar] [CrossRef] [PubMed]
  10. Sharma, S.T.; Nieman, L.K. Is Prolactin Measurement of Value during Inferior Petrosal Sinus Sampling in Patients with ACTH-Dependent Cushing’s Syndrome? J. Endocrinol. Investig. 2013, 36, 1112–1116. [Google Scholar] [CrossRef]
  11. Kline, G.; Chin, A.C. Chapter 5—Adrenal disorders. In Endocrine Biomarkers: Clinical Aspects and Laboratory Determination; Elsevier: Amsterdam, The Netherlands, 2017; Available online: https://www.sciencedirect.com/science/article/pii/B9780128034125000057 (accessed on 18 October 2022).
  12. Ghorbani, M.; Akbari, H.; Griessenauer, C.J.; Wipplinger, C.; Dastmalchi, A.; Malek, M.; Heydari, I.; Mollahoseini, R.; Khamseh, M.E. Lateralization of Inferior Petrosal Sinus Sampling in Cushing’s Disease Correlates with Cavernous Sinus Venous Drainage Patterns, but Not Tumor Lateralization. Heliyon 2020, 6, e05299. [Google Scholar] [CrossRef]
  13. Knecht, L. Inferior Petrosal Sinus Sampling in the Diagnosis of Cushing’s Disease. Available online: https://csrf.net/doctors-articles/inferior-petrosal-sinus-sampling-diagnosis-cushings-disease/ (accessed on 18 October 2022).
  14. Polito, R.; Messina, G.; Valenzano, A.; Scarinci, A.; Villano, I.; Monda, M.; Cibelli, G.; Porro, C.; Pisanelli, D.; Monda, V.; et al. The Role of Very Low Calorie Ketogenic Diet in Sympathetic Activation through Cortisol Secretion in Male Obese Population. J. Clin. Med. 2021, 10, 4230. [Google Scholar] [CrossRef] [PubMed]
  15. Yancy, W.S.; Olsen, M.K.; Guyton, J.R.; Bakst, R.P.; Westman, E.C. A Low-Carbohydrate, Ketogenic Diet versus a Low-Fat Diet to Treat Obesity and Hyperlipidemia. Ann. Intern. Med. 2004, 140, 769. [Google Scholar] [CrossRef] [PubMed]
  16. Khan, S.S.; Ning, H.; Wilkins, J.T.; Allen, N.; Carnethon, M.; Berry, J.D.; Sweis, R.N.; Lloyd-Jones, D.M. Association of Body Mass Index with Lifetime Risk of Cardiovascular Disease and Compression of Morbidity. JAMA Cardiol. 2018, 3, 280–287. [Google Scholar] [CrossRef]
  17. Sun, X.; Feng, M.; Lu, L.; Zhao, Z.; Bao, X.; Deng, K.; Yao, Y.; Zhu, H.; Wang, R. Lipid Abnormalities in Patients with Cushing’s Disease and Its Relationship with Impaired Glucose Metabolism. Front. Endocrinol. 2021, 11, 600323. [Google Scholar] [CrossRef] [PubMed]
  18. Bolla, A.; Caretto, A.; Laurenzi, A.; Scavini, M.; Piemonti, L. Low-Carb and Ketogenic Diets in Type 1 and Type 2 Diabetes. Nutrients 2019, 11, 962. [Google Scholar] [CrossRef] [PubMed]
  19. Norwitz, N.G.; Soto-Mota, A.; Kalayjian, T. A Company Is Only as Healthy as Its Workers: A 6-Month Metabolic Health Management Pilot Program Improves Employee Health and Contributes to Cost Savings. Metabolites 2022, 12, 848. [Google Scholar] [CrossRef] [PubMed]
  20. Ebbeling, C.B.; Feldman, H.A.; Klein, G.L.; Wong, J.M.W.; Bielak, L.; Steltz, S.K.; Luoto, P.K.; Wolfe, R.R.; Wong, W.W.; Ludwig, D.S. Effects of a Low Carbohydrate Diet on Energy Expenditure during Weight Loss Maintenance: Randomized Trial. BMJ 2018, 363, k4583. [Google Scholar] [CrossRef] [PubMed]
  21. Marchiò, M.; Roli, L.; Lucchi, C.; Costa, A.M.; Borghi, M.; Iughetti, L.; Trenti, T.; Guerra, A.; Biagini, G. Ghrelin Plasma Levels after 1 Year of Ketogenic Diet in Children with Refractory Epilepsy. Front. Nutr. 2019, 6, 112. [Google Scholar] [CrossRef] [PubMed]
  22. Kärkkäinen, O.; Farokhnia, M.; Klåvus, A.; Auriola, S.; Lehtonen, M.; Deschaine, S.L.; Piacentino, D.; Abshire, K.M.; Jackson, S.N.; Leggio, L. Effect of Intravenous Ghrelin Administration, Combined with Alcohol, on Circulating Metabolome in Heavy Drinking Individuals with Alcohol Use Disorder. Alcohol. Clin. Exp. Res. 2021, 45, 2207–2216. [Google Scholar] [CrossRef] [PubMed]
  23. Catargi, B.; Rigalleau, V.; Poussin, A.; Ronci-Chaix, N.; Bex, V.; Vergnot, V.; Gin, H.; Roger, P.; Tabarin, A. Occult Cushing’s Syndrome in Type-2 Diabetes. Available online: https://academic.oup.com/jcem/article/88/12/5808/2661485 (accessed on 27 September 2022).
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Levoketoconazole Treatment in Endogenous Cushing’s Syndrome

Objective: This extended evaluation (EE) of the SONICS study assessed effects of levoketoconazole for an additional 6 months following open-label, 6-month maintenance treatment in endogenous Cushing’s syndrome.

Design/Methods: SONICS included dose-titration (150–600 mg BID), 6-month maintenance, and 6-month EE phases. Exploratory efficacy assessments were performed at Months 9 and 12 (relative to start of maintenance). For pituitary MRI in patients with Cushing’s disease, a threshold of ≥2 mm denoted change from baseline in largest tumor diameter.

Results: Sixty patients entered EE at Month 6; 61% (33/54 with data) exhibited normal mean urinary free cortisol (mUFC). At Months 9 and 12, respectively, 55% (27/49) and 41% (18/44) of patients with data had normal mUFC. Mean fasting glucose, total and LDL-cholesterol, body weight, body mass index, abdominal girth, hirsutism, CushingQoL, and BDI-II scores improved from study baseline at Months 9 and 12. Forty-six patients completed Month 12; 4 (6.7%) discontinued during EE due to adverse events. The most common adverse events in EE were arthralgia, headache, hypokalemia, and QT prolongation (6.7% each). No patient experienced ALT or AST >3× ULN, QTcF interval >460 msec, or adrenal insufficiency during EE. Of 31 patients with tumor measurements at baseline and Month 12 or follow-up, largest tumor diameter was stable in 27 (87%) patients, decreased in 1, and increased in 3 (largest increase 4 mm).

Conclusion: In the first long-term levoketoconazole study, continued treatment through 12-month maintenance period sustained the early clinical and biochemical benefits in most patients completing EE, without new adverse effects.

Read the whole article at https://eje.bioscientifica.com/configurable/content/journals$002feje$002faop$002feje-22-0506$002feje-22-0506.xml?t%3Aac=journals%24002feje%24002faop%24002feje-22-0506%24002feje-22-0506.xml&body=pdf-45566

Development of Human Pituitary Neuroendocrine Tumor Organoids to Facilitate Effective Targeted Treatments of Cushing’s Disease

Abstract

(1) Background: Cushing’s disease (CD) is a serious endocrine disorder caused by an adrenocorticotropic hormone (ACTH)-secreting pituitary neuroendocrine tumor (PitNET) that stimulates the adrenal glands to overproduce cortisol. Chronic exposure to excess cortisol has detrimental effects on health, including increased stroke rates, diabetes, obesity, cognitive impairment, anxiety, depression, and death. The first-line treatment for CD is pituitary surgery. Current surgical remission rates reported in only 56% of patients depending on several criteria. The lack of specificity, poor tolerability, and low efficacy of the subsequent second-line medical therapies make CD a medical therapeutic challenge. One major limitation that hinders the development of specific medical therapies is the lack of relevant human model systems that recapitulate the cellular composition of PitNET microenvironment.
(2) Methods: human pituitary tumor tissue was harvested during transsphenoidal surgery from CD patients to generate organoids (hPITOs).
(3) Results: hPITOs generated from corticotroph, lactotroph, gonadotroph, and somatotroph tumors exhibited morphological diversity among the organoid lines between individual patients and amongst subtypes. The similarity in cell lineages between the organoid line and the patient’s tumor was validated by comparing the neuropathology report to the expression pattern of PitNET specific markers, using spectral flow cytometry and exome sequencing. A high-throughput drug screen demonstrated patient-specific drug responses of hPITOs amongst each tumor subtype. Generation of induced pluripotent stem cells (iPSCs) from a CD patient carrying germline mutation CDH23 exhibited dysregulated cell lineage commitment.
(4) Conclusions: The human pituitary neuroendocrine tumor organoids represent a novel approach in how we model complex pathologies in CD patients, which will enable effective personalized medicine for these patients.

1. Introduction

Cushing’s disease (CD) is a serious endocrine disorder caused by an adrenocorticotropic hormone (ACTH)-secreting pituitary neuroendocrine tumor (PitNET) that stimulates the adrenal glands to overproduce cortisol [1,2,3,4]. The WHO renamed pituitary adenomas as PitNETs [5]. While PitNETs have been defined as benign, implying that these tumors cause a disease that is not life threatening or harmful to health, in fact chronic exposure to excess cortisol has wide-ranging and detrimental effects on health. Hypercortisolism causes increased stroke rates, diabetes, obesity, depression, anxiety, and a three-fold increase in the risk of death from cardiovascular disease and cancer [4,6,7,8].
The first-line treatment for CD is pituitary surgery, which is followed by disease recurrence in 50% of patients during the 10-year follow-up period after surgery in the hands of an experienced surgeon [9,10,11]. Studies have demonstrated that surgical failures and recurrences of CD are common, and despite multiple treatments, biochemical control is not achieved in approximately 30% of patients. This suggests that in routine clinical practice, initial and long-term disease remission is not achieved in a substantial number of CD patients [7,12]. Hence, medical therapy is often considered in the following situations: when surgery is contraindicated or fails to achieve remission, or when recurrence occurs after apparent surgical remission. While stereotactic radiosurgery treats incompletely resected or recurrent PitNETs, the main drawbacks include the longer time to remission (12–60 months) and the risk of hypopituitarism [3,13,14]. There is an inverse relationship between disease duration and reversibility of complications associated with the disease, thus emphasizing the importance of identifying an effective medical strategy to rapidly normalize cortisol production by targeting the pituitary adenoma [4,7,12]. Unfortunately, the lack of current standard of care treatments with low efficacy and tolerability makes CD a medical therapeutic challenge.
The overall goal of medical therapy for CD is to target the signaling mechanisms to lower cortisol levels in the body [15,16]. The drugs offered for treatment of CD vary in the mechanism of action, safety, tolerability, route of administration, and drug–drug interactions [15,16]. In the era of precision medicine [17], where it is imperative to identify effective therapies early, there is an urgent need to accelerate the identification of therapies targeted to the ACTH-secreting pituitary tumor which are tailored for each individual patient. The absence of preclinical models that replicate the complexity of the PitNET microenvironment has prevented us from acquiring the knowledge to advance clinical care by implementing therapies specifically targeting the tumor, which would have a higher efficacy and tolerability for CD patients. In this instance, organoids can replicate much of the complexity of an tumor. An “organoid” is defined as a three-dimensional cell structure, grown from primary cells of dissociated pituitary tumors in Matrigel matrix, which proliferate, and differentiate in three dimensions, eventually replicating key biological properties of the tissue [18]. While pituitary cell lines predominantly represent hormonal lineages, these cultures do not reproduce the primary pituitary tissue because of the tumor transformation and non-physiological 2D culture conditions [19,20,21]. Pituitary tissue-derived organoids have been generated from mouse models [22,23]. While several human and rat pituitary spheroid/aggregate/tumoroid models have been reported, these cultures consist of poorly differentiated cells with high replicative potential which can affect drug response and produce data that poorly translate to the clinic [24,25]. In this study, we developed an organoid model derived from human PitNETs that replicated much of the cellular complexity and function of the patient’s tumor. Organoids derived from corticotroph PitNETs retained the genetic alterations of the patient’s primary tissue.

2. Materials and Methods

2.1. Generation and Culture of Human Pituitary Neuroendocrine Tumor (PitNET) Organoids

Patients with planned transsphenoidal surgery for pituitary tumors were identified in the outpatient neurosurgery clinics. Tissues were collected under the St. Joseph’s Hospital and Barrow Neurological Institute Biobank collection protocol PHXA-05TS038 and collection of outcomes data protocol PHXA-0004-72-29, with the approval of the Institutional Review Board (IRB) and patient consent. Samples were de-identified and shipped to the Zavros laboratory (University of Arizona) for processing.
Pituitary tumor tissue was collected in Serum-Free Defined Medium (SFDM) supplemented with ROCK inhibitor (Y27632, 10 µM), L-glutamine (2 mM), A83-01 (activin receptor-like kinase (Alk) 4/5/7 inhibitor, 0.5 mM), penicillin/streptavidin (1%), kanamycin (1%), amphotericin/gentamycin (0.2%), CHIR-98014 (4 mM), and thiazovivin (TZV, 2.5 mM). Tissues that contained red blood cells were incubated with Red Blood Cell (RBC) Lysis Buffer according to the manufacturer’s protocol (Thermo Fisher Scientific, San Fransisco, CA, USA). Tissues were dissected into small pieces, transferred to digestion buffer (DMEM/F12 supplemented with 0.4% collagenase 2, 0.1% hyaluronic acid, 0.03% trypsin-EDTA) and incubated for 5–10 min at 37 °C with gentle shaking. Tissue was further incubated with Accutase™ (Thermo Fisher Scientific) for 5 min at 37 °C. Enzymatically dissociated cells were pelleted and washed in DPBS supplemented with antibiotics at a 400 relative centrifugal force (RCF) for 5 min. Dissociated adenoma cells were resuspended in Matrigel™, and Matrigel™ domes containing the cells were then plated in culture dishes and overlaid with pituitary growth media (Supplemental Table S1). The culture was maintained at 37 °C at a relative humidity of 95% and 5% CO2. Organoid growth medium was replenished every 3–4 days and passaged after 15 days in culture.

2.2. Generation of Induced Pluripotent Stem Cells (iPSCs)

Induced pluripotent stem cell lines (iPSC lines) were generated from control individuals (no reported disease) or CD patients according to published protocols by the University of Arizona iPSC Core [26]. All human iPSC lines were tested and found to be negative for mycoplasma contamination using the Mycoalert Mycoplasma testing kits (LT07-318, Lonza), and no karyotype abnormalities were found (KaryoStat+, Thermo).

2.3. Pituitary Organoids Generated from iPSCs

Six well culture plates were coated with 2 mL/well 0.67% Matrigel (diluted in E8 media, UA iPSC core, 151169-01) and incubated at 37 °C at a relative humidity of 95% and 5% CO2 overnight. The iPSC lines were reprogrammed from the blood of either a healthy donor (JCAZ001) or a CD patient (iPSC7 and iPSC1063) at the University of Arizona iPSC Core. Passage 12 iPSCs were plated onto the coated plates and incubated at 37 °C at a relative humidity of 95% and 5% CO2. At 70% confluency, cells were passaged to freshly coated 24 well plates at a ratio of 1:8 and grown to 85–90% confluency before beginning the directed differentiation schedule. From days 0 to 3, cells were cultured in E6 media supplemented with 1% penicillin/streptomycin, 10 μM SB431542, and 5 ng/mL BMP4. BMP4 was withdrawn from the culture at day 3. Starting on day 4, the cells were cultured in E6 media, supplemented with 10 μM SB431542, 30 ng/mL human recombinant SHH, 100 ng/mL FGF8b, 10 ng/mL FGF18, and 50 ng/mL FGF10. Fifteen days after culture, the cells were harvested in cold E6 media by pipetting and resuspended in Matrigel™ (20,000 cells/50 mL Matrigel™). Matrigel™ domes containing the cells were plated in culture dishes and overlaid with differentiation media containing E6 media which was supplemented with 10 μM Y-27632, 30 ng/mL human recombinant SHH, 100 ng/mL FGF8b, 10 ng/mL FGF18, and 50 ng/mL FGF10 (Supplemental Table S2). Organoids were cultured for a further 15 days at 37 °C at a relative humidity of 95% and 5% CO2.

2.4. Spectral Flow Cytometry (Cytek™ Aurora)

The multicolor flow cytometry panel was designed using the Cytek® Full Spectrum Viewer online tool to calculate the similarity index (Supplemental Figure S1). The organoids were harvested in cold SFDM media and centrifuged at 400× g for 5 min. Supernatant was discarded and organoids were dissociated to single cells using Accutase® (Thermo Fisher Scientific 00-4555-56). The enzymatic reaction was stopped using prewarmed DPBS, and cells were then centrifuged at 400× g for 5 min and incubated with fluorochrome-conjugated/unconjugated primary surface or cytoplasmic antibodies (Supplemental Figure S1) at 4 °C for 30 min. Cells were then washed with Cell Staining Buffer (BioLegend # 420-201) and incubated with secondary antibodies (Supplemental Figure S1) at 4 °C for 30 min. Cells were fixed using Cytofix/Cytoperm™ Fixation/Permeabilization Solution (BD Biosciences # 554714) at 4 °C for 20 min, followed by washing with Fixation/Permeabilization wash buffer. Cells were labeled with fluorochrome-conjugated/unconjugated intracellular primary antibodies (Supplemental Figure S1) at 4 °C for 30 min, then washed and incubated with secondary antibodies at 4 °C for 30 min. Cells were resuspended in cell staining buffer and fluorescence and measured using the Cytek Aurora 5 Laser Spectral Flow Cytometer. An unstained cell sample was fixed and used as a reference control. UltraComp eBeads™, Compensation Beads (Thermo Fisher Scientific # 01-2222-42) were stained with the individual antibodies and used as single stain controls for compensation and gating. Data were acquired using the Cytek™ Aurora and analyzed using Cytobank software (Beckman Coulter, Indianapolis, IN, USA).

2.5. Whole Mount Immunofluorescence

Organoids were immunostained using published protocols by our laboratory [27,28,29]. Proliferation was measured by using 5-ethynyl-2′-deoxyuridine (EdU) incorporation according to the Manufacturer’s protocol (Click-IT EdU Alexa Fluor 555 Imaging Kit, Thermo Fisher Scientific C10338). Co-staining was performed by blocking fixed organoids with 2% donkey serum (Jackson Immuno Research, # 017-000-121) diluted in 0.01% PBST for 1hr at room temperature. Organoids were then incubated overnight at 4 °C with primary antibodies, followed by secondary antibodies and Hoechst (Thermo Fisher Scientific H1399, 1:1000 in 0.01% PBST) for 1 h at room temperature. Human specific primary antibodies used included: rabbit anti-ACTH (Thermo Fisher Scientific 701293, 1:250), rabbit anti-Synaptophysin (Thermo Fisher Scientific PA5-27286, 1:100), species PIT1 (Thermo Fisher Scientific PA5-98650, 1:50), rabbit anti-LH (Thermo Fisher Scientific PA5-102674, 1:100), mouse anti-FSH (Thermo Fisher Scientific MIF2709, 1:100), mouse anti-PRL (Thermo Fisher Scientific CF500720, 1:100), Alexa Flour conjugated GH (NB500-364AF647, 1:100), and mouse anti-CAM5.2 (SIGMA 452M-95, 1:250). The secondary antibodies used included Alexa Fluor 488 Donkey Anti Rabbit IgG (H+L) (Thermo Fisher Scientific A21206, 1:100) or Alexa Fluor 647 Donkey Anti Mouse IgG (H+L) (Thermo Fisher Scientific A31571, 1:100). Organoids were visualized and images were acquired by confocal microscopy using the Nikon CrestV2 Spinning Disk (Nikon, Melville, NY, USA). Fluorescence intensity and percentage of EdU positive cells of total cells, were calculated using Nikon Elements Software (Version 5.21.05, Nikon, Melville, NY, USA).

2.6. Nuclear Morphometric Analysis (NMA)

Nuclear Morphometric Analysis (NMA) using treated organoids was performed based on a published protocol that measures cell viability based on the changes in nuclear morphology of the cells, using nuclear stain Hoechst or DAPI [30]. Images of organoid nuclei were analyzed using the ImageJ Nuclear Irregularity Index (NII) plugin for key parameters, which included cell area, radius ratio, area box, aspect, and roundness. Using the published spreadsheet template [30], the NII of each cell was calculated with the following formula: NII = Aspect − Area Box + Radius Ratio + Roundness. The area vs. NII of vehicle-treated cells were plotted as a scatter plot using the template, and was considered as the normal cell nuclei. The same plots were generated for each condition, and the NII and area of treated cells were compared to the normal nuclei, and classified as one of the following NMA populations: Normal (N; similar area and NII), Mitotic (S; similar area, slightly higher NII), Irregular (I; similar area, high NII), Small Regular (SR; apoptotic, low area and NII), Senescent (LR; high area, low NII), Small Irregular (SI; low area, high NII), or Large Irregular (LI; high area, high NII). Cells classified as SR exhibited early stages of apoptosis, and cells classified as either I, SI, or LI exhibited significant nuclear damage. The percentage of cells in each NII classification category were calculated and plotted as a histogram using GraphPad Prism.

2.7. ELISA

Concentration of secreted ACTH in conditioned media that was collected from organoid cultures was measured using the Human ACTH ELISA Kit (Novus Biologicals, NBP2-66401), according to the manufacturer’s protocol. The enzyme–substrate reaction was measured spectrophotometrically (BioTek Gen5 Micro Plate Reader Version 3.11, Santa Clara, CA, USA) at a wavelength of 450 nm, and the ACTH concentration (pg/mL) was interpolated by a standard curve with a 4-parameter logistic regression analysis, using GraphPad Prism (Version 9.2.0, San Diego, CA, USA).

2.8. Drug Assay

Patient adenoma-derived pituitary organoids were grown in 96-well plates and treated with 147 small molecules taken from the NCI AOD9 compound library for 72 h. (https://dtp.cancer.gov/organization/dscb/obtaining/available_plates.html (accessed on 22 August 2021)). Drugs were diluted from 10 mM DMSO stock plates into 100 M DMSO working stocks with a final concentration of 1μM. All vehicle controls were treated with 0.1% DMSO. Organoid proliferation was measured using a CellTiter 96® AQueous One Solution Cell Proliferation Assay kit (MTS, Promega, G3582, Madison, WI, USA) according to the manufacturer’s instruction. Organoid death was calculated based on the absorbance readings at 490 nm, collected from the MTS assay relative to the vehicle controls. Drug screens were performed with biological replicates in the same screen. Drugs were selected based on their ability to target key signaling pathways as well as clinical relevance to the treatment. Drug sensitivity is represented by cell viability, and is significant at <0.5 suppressive effect of the drugs. The percent of cell viability relative to the vehicle control was calculated. Correlation coefficients across each organoid were calculated using the Pearson method to assess confidence in replication. The variance component was detected for each drug across all organoids. A random effect model was run with a single random factor for each drug, and estimated variance was calculated by rejecting the null hypothesis that variation was not present among samples. The drug responses were grouped by variance factor, into large (vc > 100), median (100 > vc > 50), and small (vc < 50). A heatmap was used to display the differential responses in cell viability for the drugs.
Drugs that clustered together and showed response within corticotrophs were investigated further based on their mode of action. Pathways (Kegg and Reactome) and gene ontology mapping were conducted for the genes that were being targeted by the drugs, in order to evaluate the key responses in cellular processes. A network was constructed in Cytoscape v 3.8.2 (San Diego, CA, USA) for the purpose of association between the drugs and genes.

2.9. Drug Dose Responses

Organoids were grown in Matrigel™ domes within 96-well round-bottom culture plates. Recombinant human SHH was removed from the pituitary organoid growth media, 24 h prior to drug treatment. Organoids were treated with either vehicle (DMSO), cabergoline (Selleckchem S5842), ketoconazole (Selleckchem S1353), roscovitine (Selleckchem S1153), GANT61 (Stemcell Technologies 73692), pasireotide (TargetMol TP2207), mifeprostone (Selleckchem S2606), etomidate (Selleckchem S1329), mitotane (Selleckchem S1732), metyropane (Selleckchem S5416), or osilodrostat (Selleckchem S7456) at concentrations of 0, 1, 10, 100, 1000, and 10,000 nM, for 72 h. The percentage of cell viability was measured using an MTS assay (Promega G3580). Absorbance was measured at 490 nm and normalized to the vehicle. Concentrations were plotted in a logarithmic scale, and a nonlinear dose response curve regression was calculated using GraphPad Prism. An IC50 value for each drug treatment was determined based on the dose response curve, using GraphPad Prism analysis software.

2.10. Calculation of Area under the Curve (AUC)

AUC (area under the curve) was determined by plotting the normalized % cell viability versus transformed concentration of the drugs, using a trapezoidal approximation for the area [31]. The formula was based on splitting the curve into trapezoids with bases equal to the % viability (V) and height equal to the interval length (difference in concentrations (C), and then summing the areas of each trapezoid:

n0(Vn+Vn1)2(CnCn1)

2.11. Quantitative RT PCR (qRT-PCR)

RNA was collected from patient-derived organoid cultures using the RNeasy Mini Kit (Qiagen). cDNA was generated from the extracted RNA, and then pre-amplified using TaqMan PreAmp Master Mix (Thermo Fisher Scientific 391128). The primers used were human-specific GAPDH (Thermo Fisher Scientific, Applied Biosystems Hs02786624_g1), NR5A1 (SF1) (Thermo Fisher Scientific, Hs00610436_m1), PIT1 (Thermo Fisher Scientific, Hs00230821_m1), TPit (Thermo Fisher Scientific, Hs00193027), and POMC (Thermo Fisher Scientific, Hs01596743_m1). Each PCR reaction was performed using a final volume of 20 µL, composed of 20X TaqMan Expression Assay primers, 2X TaqMan Universal Master Mix (Applied Biosystems, TaqMan® Gene Expression Systems), and a cDNA template. Amplification of each PCR reaction was conducted in a StepOne™ Real-Time PCR System (Applied Biosystems, Foster City, CA, USA), using the following PCR conditions: 2 min at 50 °C, 10 min at 95 °C, denaturing for 15 s at 95 °C, and annealing/extending for 1 min at 60 °C, for a total of 40 cycles. Relative fold change was calculated using the 2 − ∆∆Ct method [32], where CT = threshold cycle. Results were analyzed as the average fold change in gene expression compared to the control, and GAPDH served as an internal control.

2.12. Whole Exome Sequencing

WES was performed by the University of Arizona Center for Applied Genetics and Genomic Medicine. Isolated DNA from patient adenoma tissue will be quantified using the Qubit quantitation system with standard curve, as per the supplier protocol (Thermo Fisher Scientific). All samples were further tested for quality using the Fragment Analyzer (Advanced Analytical), following the manufacturer-recommended protocols. Whole exome sequencing (WES) was performed by array capture and approximately 60 Mb of exome target sequence, using the SureSelectXT Human All Exon V6 enrichment (Agilent) or equivalent (which one was used). All exome library builds were quantified via qPCR and subsequently sequenced to a minimum 20X coverage, using paired-end chemistry on the Illumina NovaSeq platform. Whole exome sequencing (WES) was performed by hybridization capture of approx. 35 Mb of the exome target sequence, using the Swift Exome Hyb Panel (Swift Biosciences 83216). All exome library builds were quantified via qPCR and subsequently sequenced to a minimum 20X coverage, using paired-end chemistry on the Illumina NextSeq500 or NovaSeq platform (Illumina). DNA reads were trimmed, filtered by quality scores and aligned to the human genome (hg38) with Burrows–Wheeler Aligner with default parameters. Picard (http://broadinstitute.github.io/picard (accessed on 22 December 2021)) was used to mark duplicates. Germline single nucleotide variants (SNV) were called using the Genome Analysis Tool Kit (GATK), using the given guidelines. Mutations were annotated using ANNOVAR for coding sequences. Variants that passed the quality filter were further investigated for similarity. Concordance between tissue and organoids was calculated using Jaccard similarity index (Jij = Mij/(Mi + Mj − Mij) where Mi is the number of variants in tissues, Mj is the number of variants in organoids, and Mij is the number of identical variants in both tissue and organoid.

2.13. Single Cell RNA Sequencing (scRNA-Seq)

Cultures were collected on day 15 of the pituitary directed differentiation schedule, and cells were dissociated into a single-cell suspension using Cell Dissociation Buffer (Thermo Fisher Scientific 13151014). Cells (15,000 cells/sample) were resuspended in the sample buffer (BD Biosciences 65000062), filtered using cell strainer (40 microns), and loaded into a BD Rhapsody cartridge (BD Biosciences 400000847) for single-cell transcriptome isolation. Based on the BD Rhapsody system whole-transcriptome analysis for single-cell whole-transcriptome analysis, microbead-captured single-cell transcriptomes were used to prepare a cDNA library. Briefly, double-stranded cDNA was first generated from the microbead-captured single-cell transcriptome in several steps, including reverse transcription, second-strand synthesis, end preparation, adapter ligation, and whole-transcriptome amplification (WTA). Then, the final cDNA library was generated from double-stranded full-length cDNA by random priming amplification using a BD Rhapsody cDNA Kit (BD Biosciences, 633773), as well as the BD Rhapsody Targeted mRNA and WTA Amplification Kit (BD Biosciences, 633801). The library was sequenced in PE150 mode (paired-end with 150-bp reads) on NovaSeq6000 System (Illumina). A total of 80,000 reads were demultiplexed, trimmed, mapped to the GRCh38 annotation, and quantified using the whole transcriptome analysis pipeline (BD Rhapsody™ WTA Analysis Pipeline v1.10 rev6, San Jose, CA, USA) on the Seven Bridges Genomics platform (https://igor.sbgenomics.com (accessed on 4 April 2022)), prior to clustering analysis in Seurat. For QC and filtration, read counting and unique molecular identifier (UMI) counting were the principal gene expression quantification schemes used in this single-cell RNA-sequencing (scRNA-seq) analysis. The low-quality cells, empty droplets, cell doublets, or multiplets were excluded based on unique feature count (less than 200 or larger than 2500), as they may often exhibit either an aberrantly high gene count or very few genes. Additionally, the mitochondrial QC metrics were calculated, and the cells with >5% mitochondrial counts were filtered out, as the percentage of counts originating from a set of low-quality or dying cells often exhibit extensive mitochondrial contamination. After the removal of unwanted cells from the single cell dataset, the global-scaling normalization method LogNormalize was employed. This method normalizes the feature expression measurements for each cell by the total expression, multiplies this by a scale factor (10,000), and log-transforms the result. The molecules per gene per cell, based on RSEC error correction (RSEC_MolsPerCell file) matrix files from iPSCctrl and iPSCCDH23 samples, were imported into Seurat v4, merged, and processed (as stated above) for UMAP reduction, cluster identification, and differential marker assessment using the FindAllMarkers function within Seurat.

2.14. Statistical Analyses

Sample size was based on assessment of power analysis using SigmaStat software. Data collected from each study from at least 4 in vitro technical replicates were analyzed by obtaining the mean ± standard error of the mean (SEM), unless otherwise stated. The significance of the results was then tested using commercially available software (GraphPad Prism, GraphPad software, San Diego, CA, USA).

3. Results

3.1. Generation and Validation of Human PitNET Tissue Derived Organoids

Human PitNET tissue was harvested during endoscopic transsphenoidal pituitary surgery from 35 patients in order to generate organoids. These cultures are referred to as human PitNET tissue derived organoids (hPITOs). Supplementary Table S3 summarizes the neuropathology reports and clinical diagnosis from these cases. In summary, 12 corticotroph (functional, CD), and 3 silent corticotroph tumors (nonfunctional tumors), 9 gonadotroph tumors, 8 lactotroph tumors, and 3 somatotroph tumors (acromegaly) were used to generate hPITOs (Supplementary Table S3).
Bright-field microscopy images of hPITOs that were generated from corticotroph adenomas from patients diagnosed with CD (Figure 1a–e). Silent/nonfunctioning tumors (Figure 1f,g) revealed morphological diversity among the organoid lines between individual patients and amongst subtypes. Confocal microscopy was used to capture a z-stack through the hPITO38, immunofluorescently stained for CAM5.2 (red), ACTH (green), and Hoechst (nuclear staining, blue) and emphasizes the 3D cellular structure of the hPITOs (Supplemental Video S1). Lactotroph, gonadotroph, and somatotroph adenomas were used to generate hPITOs, and showed the same morphological divergence amongst subtypes and between each patient line (Supplemental Figure S2). Proliferation was measured within the cultures using 5-ethynyl-2′-deoxyuridine (EdU) uptake and showed that the percentage of EdU+ve cells/total Hoechst+ve nuclei directly correlated with the pathology MIB-1 (Ki67) score (red, R2 = 0.9256) (Figure 1a–g, Supplemental Figure S2). ACTH concentration, which was measured by ELISA using organoid conditioned culture media collected from each hPITO line, showed the highest expression in the corticotroph adenoma organoids generated from CD patients (Figure 1h).
Cells 11 03344 g001 550
Figure 1. Morphology and function of corticotroph hPITOs. (ag) Brightfield images, immunofluorescence staining using antibodies specific for CAM5.2 (red), ACTH (green), and EdU (magenta, inset) of organoid cultures generated from patients with Cushing’s disease (hPITOs 1, 7, 10, 33, 35) or nonfunctional corticotroph adenomas (hPITO8, 12). Quantification of %EdU positive cells/total cell number is shown and compared to the Ki67 score given in the pathology report (Supplemental Table S3). An ELISA was performed using conditioned media collected from (h) corticotroph hPITO cultures and (i) lactotroph, somatotroph, and gonadotroph hPITO cultures for the measurement of ACTH secretion (pg/mL).

3.2. Characterization of Cell Lineages in Pituitary Adenoma-Derived Organoids by Spectral Cytek™ Aurora Analysis

In order to validate the similarity in cell lineages identified between the organoid line and the patient’s tumor, we compared the immunohistochemistry from the neuropathology report (Supplemental Table S3) to the expression pattern of pituitary adenoma-specific markers, which were measured using Cytek™ Aurora spectral flow cytometry (Figure 2). The location of cells that are found in each cluster based on the highly expressed antigens are shown in the representative tSNE (viSNE) maps (Figure 2a). Compared to nonfunctional adenoma-derived hPITOs, organoids derived from corticotroph adenomas of CD patients highly expressed proliferating (Ki67+) T-Pit+ ACTH cells (Figure 2a). Interestingly, there was an increase in SOX2+ cells within the total cell population, associated with Crooke’s cell adenoma hPITOs (Figure 2a). Within the total cell population, cell clusters expressing CD45 and vimentin were also measured (Figure 2a). Data for the analysis of corticotroph hPITOs, derived from CD patients and individuals with nonfunctional adenomas, were summarized in a heatmap for each subtype organoid line based on quantified cell abundance (percent of total cells) using spectral flow cytometry (Figure 2b).
Cells 11 03344 g002 550
Figure 2. Cell heterogeneity of corticotroph hPITOs. (a) viSNE maps define spatially distinct cell populations using pituitary specific cell lineage, stem cell, and transcription factor markers. Cell populations were quantified in organoids generated from CD patients with corticotroph adenomas (sparsely granulated and Crooke’s cell adenoma) or patients with nonfunctional corticotroph adenomas. (b) Quantification of the abundance of cells expressing pituitary specific markers as a percent total. viSNE maps define spatially distinct cell populations in organoid cultures generated from CD patient with (c) corticotroph adenoma (hPITO37, Crooke’s cell adenoma) and adjacent normal tissue (hPITO37N), or (d) sparsely granulated corticotroph adenomas (hPITO38) and adjacent normal tissue (hPITO38N).
Organoid cultures derived from pituitary adenomas (hPITO37 and hPITO38) were compared to organoids derived from adjacent normal pituitary tissue (hPITO37N and hPITO38N) (Figure 2c,d). While Pit1 lineages including cells expressing GH and PRL, as well as SF1 lineages expressing FSH and LH, were detected in the hPITO37N and hPITO38N organoid cultures, these cell populations were significantly reduced within the patient’s matched adenoma tissue (Figure 2c,d). Overall, hPITOs derived from CD patients expressed increased stem and progenitor cell markers, including CXCR4, SOX2, and CD133 (Figure 2). Collectively, our findings of the characterization of the hPITO cultures support our prediction that this in vitro model recapitulates much of the patient’s adenoma pathophysiology.

3.3. Inherent Patient Differences to Drug Response Is Reflected in the Organoid Culture

Tumor recurrence can occur in as many as 30–50% of CD patients after successful surgical treatment [10,33,34]. Unfortunately, bilateral adrenalectomy is the chosen surgical treatment for patients with persistent CD [35]. Bilateral adrenalectomy leads to the increased risk for development of Nelson’s syndrome (progressive hyperpigmentation due to ACTH secretion and expansion of the residual pituitary tumor). Although the risk of developing Nelson’s syndrome following adrenalectomy can be reduced by 50% with stereotactic radiotherapy [35], there is a need to develop medical therapies that directly target the pituitary adenoma. Thus, we established a high-throughput drug screening assay using patient-derived PitNET organoids. After 72 h of treatment, cell viability was measured using an MTS assay, and data were represented as a heatmap whereby blue indicated higher cell death, and red suggested higher cell viability. The replicates behaved consistently with the drug response, with correlation scores of >0.8 for these samples (Figure 3a). We estimated the variance component for each drug across all organoids. Variation among samples was found to be significant (p ≤ 0.05) for each of the 83 drugs. The drug responses were grouped by variance factor into large, median, and small. The larger the variance, the more variable the drug response was across the organoids. We noted a set of drugs that showed a significant differential response across the functional corticotroph organoids. Unsupervised clustering of drug responses across organoids shows a pattern that relates to our statistically calculated results (Figure 3a,c), and the replicates for each independent organoid cluster together. The drugs with higher variance components across all the functional corticotrophs cluster together as a group (Figure 3a). These drugs show cell viability of 10% to 60% across different organoids. Analyzing the pattern more closely, we observe that, within a pathologically defined group, there was a differential organoid response to drugs as well as inherent patient differences to drugs within this group. Figure 3 demonstrates a variation in drug responsiveness amongst the organoid lines generated from individual patients. Importantly, there was further divergence in drug responsiveness amongst the individual organoid lines within each pathologically defined corticotroph subtype. These data clearly demonstrate that the inherent patient difference to drug response which is often observed among CD patients is reflected in the organoid culture.
Cells 11 03344 g003 550
Figure 3. Drug screen using hPITOs generated from CD patients. (a) High-throughput drug screening of hPITOs reveals sensitivities to a range of therapeutic agents. Cell viability with high values (indicating resistance) are depicted in red, and low values (indicating sensitivity) are in blue in the clustered heatmap. (b,c) Clusters showing response to therapeutic agents with the most variance across the organoids. (d) Network of drugs from the clusters b and c and their gene targets, showing their participation in signaling pathways and cellular processes.
Drugs that clustered together and showed correlated responses were investigated further for their mode of action based on target genes (Figure 3d). The genes were analyzed for their associations in cellular pathways and gene ontology functional processes. Identified drug–gene pairs were interconnected by cellular pathways that are known to regulate cell cycle, WNT signaling, hedgehog signaling, and neuroactive ligand-receptor interaction signaling pathways (Figure 3d). These identified genes are also known to be influenced by multiple cellular functions, such as cytokine–cytokine receptor interactions and Notch signaling. Proteosome 20S subunit genes PSMAs/PSMBs and the HDAC gene family are involved in many cellular functions. The ephrin receptors (EPHs), adrenoceptor alpha receptors (ADRs), dopamine receptors (DRDs), and the 5-hydroxytryptamine serotonin receptors (HTRs) gene families influence neuronal functions and are targeted by multiple drugs in our focused cluster. These data reveal potential therapeutic pathways for CD patients.
Divergent half maximal inhibitory concentration (IC50) values, as documented by an MTS cell viability assay, were observed in response to drug treatment among hPITOs lines 28, 33, 34, 35, and 37. Note that a shift of the curve to the right indicates a higher IC50 (i.e., more resistant to that drug). Cell viability assays were normalized to vehicle-treated controls in order to ensure that toxicity was specific to the drug effects (Figure 4). Dose response curves for organoid 33 and organoid 34 showed better responses at lower doses for cabergoline compared to Metyrapone and osilodrostat, but different for organoid 35, where Metyrapone and osilodrostat gave better responses than Cabergoline (Figure 4a–h). For the drugs mifepristone and GANT61, 33 and 34 had the same level of response to both the drugs. However, when the two organoid responses were compared, 34 had a better response than 33 (Figure 4a–h). Similar divergent drug responses were observed in hPITO lines 37 and 38 (Figure 4i,k). However, organoids generated from adjacent normal pituitary tissue from patients 37 and 38 were nonresponsive to the same standard of care of investigational drugs for CD (Figure 4j,l). These data were consistent with observation made in the drug screen (Figure 3a–c), and demonstrate that there was an inherent difference to drug response within the organoid cultures of the same corticotroph subtype.
Cells 11 03344 g004 550
Figure 4. Drug dose responses by hPITOs generated from CD patients. Dose responses to mifepristone, GANT61, cabergoline, and osilodrostat. (a,e) hPITO28, (b,f) hPITO33, (c,g) hPITO34, and (d,h) hPITO35. Dose responses to cabergoline, ketoconazole, roscovitine, GANT61, pasireotide, mifepristone, etomidate, mitotane, metyrapone, and osilodrostat in (i) hPITO37, (j) organoids generated from adjacent normal pituitary tissue (hPITO37N), (k) hPITO38, (l) hPITO38N, and (m) hPITO39. (n) IC50 and integrated area under the curve in response to mifepristone, ketoconazole, and pasireotide using hPITO39 cultures. Nuclear morphometric analysis of hPITO39 cultures in response to (o,p) vehicle, (q,r) mifepristone, (s,t) pasireotide, and (u,v) ketoconazole. Morphometric classification of NII was based on the normal (N), small (S), small regular (SR), short irregular (SI), large regular (LR), large irregular (LI), and irregular (I) nuclear morphology. Representative Hoechst staining of organoids in response to drug treatments for the calculation of the nuclear irregularity index (NII) are shown in the insets in (p,r,t,v).
In addition to cell viability, Nuclear Morphometric Analysis (NMA) using treated organoids was performed based on a published protocol that measures cell viability according to the changes in nuclear morphology of the cells, using nuclear stain Hoechst or DAPI [30]. Nuclear Irregularity Index (NII) was measured based on the quantification of the morphometric changes in the nuclei in response to the standard-of-care drugs mifepristone, pasireotide, and ketoconazole in hPITO39 (Figure 4o–v). The area vs. NII of vehicle-treated cells were plotted as a scatter plot using the template, and considered as the normal cell nuclei (Figure 4o). The same plots were generated for mifepristone (Figure 4q), pasireotide (Figure 4s), and ketoconazole (Figure 4u). The NII and area of treated cells were compared to those of the normal nuclei, and classified as one of the following NMA populations: Normal (N; similar area and NII), Mitotic (S; similar area, slightly higher NII), Irregular (I; similar area, high NII), Small Regular (SR; apoptotic, low area and NII), Senescent (LR; high area, low NII), Small Irregular (SI; low area, high NII), or Large Irregular (LI; high area, high NII) (Figure 4p,r,t,v). Cells classified as SR exhibited early stages of apoptosis, and cells classified as either I, SI, or LI exhibited significant nuclear damage. Data showed that mifepristone induced significant apoptosis in hPITO39 cultures (Figure 4r), compared to responses to pasireotide (Figure 4t) and ketoconazole (Figure 4v). These responses were consistent with the IC50 and the total area under the curve in response to drugs (Figure 4m,n). Measurement of NII is an approach which may be used to confirm potential drug targets identified from the drug screen.

3.4. Organoid Responsiveness to Pasireotide Correlates with SSTR2 and SSTR5 Expression

Organoid lines hPITO28, 31, 33, 34, and 35 exhibited divergent IC50 values in response to SSTR agonist pasireotide (Figure 5a). hPITO34 was the most responsive to pasireotide, with a low IC50 value of 6.1 nM (Figure 5a). Organoid lines hPITO33 and hPITO35 were the least responsive, with IC50 values of 1.2 µM and 1 µM, respectively, in response to pasireotide (Figure 5a). The expression of SSTR subtypes 1–5 among the different organoid lines were measured by qRT-PCR and IHC (Figure 5b). One of the least responsive organoid lines, hPITO28, exhibited lower differential expression in SSTR2 and SSTR5 compared to the highly responsive hPITO34 line (Figure 5a,b). Gene expression levels of SSTR2 and SSTR5 within hPITO28 and 34 correlated with protein levels within the patient’s tumor tissue (Figure 5c–f). Given the greater binding affinity for SSTR5 compared to SSTR2 by pasireotide, these data were consistent with greater responsiveness to the drug by hPITO34 in comparison to hPITO28 (Figure 5a,c–f). The expression of SSTR subtypes 2 and 5 within the organoid cultures correlated with the expression patterns of the patient’s tumor tissues (Figure 5a,c–f).
Cells 11 03344 g005 550
Figure 5. SSTR1-5 expression in hPITOs and patient’s PitNET tissue. (a) Dose response of hPITO28, 31, 33, 34, and 35 lines to pasireotide. (b) Differential expression of SSTR subtypes 1–5 (SSTR1, SSTR2, SSTR3, SSTR4, SSTR5) in hPITO28, hPITO31, hPITO33, hPITO34, and hPITO35. Immunohistochemistry of (c,e) SSTR2 and (d,f) SSTR5 expression in patient PitNET tissue (Pt28 and Pt34), from which hPITO28 and 34 were generated.

3.5. Organoids Derived from Pituitary Corticotroph Adenomas Retain the Genetic Alterations of the Patient’s Primary Tumor

In order to identify the genetic features of the organoids derived from pituitary adenomas of CD patients, we performed whole-exome sequencing (WES) of hPITOs and the corresponding primary adenoma tissues. We performed WES analysis of each hPITO line, and compared the results with those for the corresponding primary adenoma tissues. We showed the concordance rate of exonic variants between the primary tumor tissues obtained from CD patients and the corresponding organoid line. We identified, on average, approximately 5000 mutations across each of the 14 paired samples of organoids and tissues. For the variants detected, all seven pairs showed a Jaccard index ranging from 0.5 to 0.8. Out of seven pairs, five (hPITO24, 25, 28 and 35) pairs had a Jaccard score of 0.8, while hPITO33 and 34 pairs had 0.7, and hPITO1 had 0.5. In order to investigate the similarity across the SNV (single nucleotide variation) sites, we calculated the Jaccard index of exon sites for synonymous and non-synonymous events, and found scores for all pairs ranging from 0.8 to 0.9. Furthermore, for only non-synonymous events, Jaccard scores also ranged from 0.8 to 0.9, except for hPITO1, which showed overall lower concordance, and had a score of 0.4 to 0.5. Figure 6 shows non-synonymous mutations found in organoid and tissue pairs for some of the key genes that are known to be involved in pituitary adenoma disease. Concordance indices between organoids and the matched patient’s adenoma tissues is reported in Figure 6. Therefore, WES data demonstrated that organoids derived from pituitary corticotroph adenomas retained the genetic alterations of the patient’s primary tumor tissue.
Cells 11 03344 g006 550
Figure 6. Genomic landscape of hPITOs recapitulates genetic alterations commonly found PitNETs. Overview of single nucleotide variation events detected in hPITOs in genes commonly altered in PitNETs. The mutation frequency across the organoid population is depicted on the right. Color coding of the figure shows that organoid lines are derived from the same patient tumor tissue. ORG: organoid line, TIS: matched patient’s PitNET tissue.

3.6. IPSC Pituitary Organoids Generated from a CD Patients Expressing Familial Mutations Reveal Corticotroph Adenoma Pathology In Vitro

Extensive research has revealed the role of somatic and germline mutations in the development of CD adenomas [36,37]. Pituitary organoids were developed from iPSCs generated from the PBMCs of CD patients and carrying germline mutations that were identified by WES (Supplemental Figure S4). Chromosomal aberrations were not found when comparing against the reference dataset in the iPSCs generated from the CD patients (Supplemental Figure S3a,b). PBMCs isolated from patients diagnosed with CD were analyzed by WES in order to determine the expression of germline mutations. WES revealed the expression of a more recently identified gene predisposing patients to CD, namely cadherin-related 23 [38] (Supplemental Figure S5).
Pituitary organoids were then developed from iPSCs which were generated from the PBMCs of patients with CD (iPSCCDH23 and iPSCMEN1) and a healthy individual (iPSCctrl). Expression of PIT1 (pituitary-specific positive transcription factor 1), ACTH (adrenocorticotropic hormone), GH (growth hormone), FSH (follicle-stimulating hormone), LH (luteinizing hormone), PRL (prolactin), and synaptophysin (synaptophysin) with co-stain Hoechst (nuclei, blue) was measured by immunofluorescence, using chamber slides collected at 15 of the differentiation schedules (Supplemental Figure S6). While pituitary tissue that was differentiated from iPSCctrl expressed all major hormone-producing cell lineages (Supplemental Figure S6a), there was a significant increase in the expression of ACTH and synaptophysin, with a concomitant loss of PIT1, GH, FSH, LH, and PRL in iPSCsMEN1 (Supplemental Figure S6b,c). Interestingly, iPSCCDH23 cultures exhibited a significant increase in the expression of ACTH, GH, LH, and synaptophysin, with a concomitant loss of PIT1, FSH, and PRL (Supplemental Figure S6b,c). Immunofluorescence of iPSCs collected on the fourth day of the differentiation schedule revealed no expression of PIT1, ACTH, GH, FSH, LH, or PRL in (data not shown). Compared to control lines, iPSC lines expressing mutated CDH23 secreted significantly greater concentrations of ACTH earlier in the differentiation schedule (Supplemental Figure S7a). The upregulated expression of pituitary corticotroph adenoma-specific markers in iPSCCDH23 and iPSCMEN1 demonstrates that the iPSC-derived organoids represented the pathology of corticotroph adenomas in vitro.

3.7. ScRNA-seq Reveals the Existence of Unique Proliferative Cell Populations in iPSCCDH23 Cultures When Compared to iPSCsctrl

Using Seurat to identify cell clusters, as well as Uniform Manifold Approximation and Projection 9UMAP, clustering analysis identified 16 distinct cell populations/clusters consisting of known marker genes. Clusters 1, 5, and 7 of the iPSCsCDH23 were distinct from the iPSCctrl cultures (Figure 7a,b). Pituitary stem cells were characterized in iPSCctrl and iPSCCDH23 cultures (Figure 7b). Clusters 1 and 5 expressed markers consistent with the corticotroph subtype cell lineage (Figure 5c). Markers of dysregulated cell cycles and increased proliferation were identified in cell cluster 7 (Figure 7c). Expression of the E2 factor (E2F) family of transcription factors, which are downstream effectors of the retinoblastoma (RB) protein pathway and play a crucial role in cell division control, were identified in distinct cell cluster 7, which was identified within the iPSCCDH23 cultures (Figure 7c). Stem cell markers were also upregulated in cell cluster 7, and identified within the iPSCCDH23 cultures (Figure 7c). Using Cytobank software to analyze organoids collected 30 days post-differentiation, cells were gated on live CK20 positive singlets, and 9000 events per sample were analyzed by the viSNE algorithm. ViSNE plots are shown in two dimensions with axes identified by tSNE- 1 and tSNE-2, and each dot representing a single cell positioned in the multidimensional space (Figure 7d). Individual flow cytometry standard files were concatenated into single flow cytometry standard files, from which 12 spatially distinct populations were identified (Figure 7e). Overlaying cell populations identified by traditional gating strategies onto viSNE plots identified unique cell populations within the iPSCCDH23 cultures (Figure 7e). There were distinct cell populations between the iPSCctrl and iPSCCDH23 organoids, in addition to expression of hormone and cell lineage markers such as ACTH, TPit, PRL, and PIT1 (Figure 7e). The cell populations that exhibited high expression of Ki67 within the iPSCctrl organoid cultures included SOX2+ and PIT1+ populations (Figure 7f). The highly proliferating cell populations within the iPSCCDH23 organoid cultures included those that expressed CD90+/VIM+/CXCR4+ (mesenchymal stem cells), CXCR4+/SOX2+ (stem cells), TPit+ (corticotroph cell lineage), CD133+/CD31+ (endothelial progenitor cells), and CK20+/VIM+/CXCR4+ (hybrid epithelial-mesenchymal stem cells) (Figure 7f). Overall, the iPSCCDH23 organoids were significantly more proliferative compared to the iPSCctrl cultures (Figure 7f). Immunofluorescence staining of iPSCCDH23 organoids revealed increased mRNA expression of TPit and POMC, which correlated with increased ACTH protein compared to iPSCsctrl (Supplemental Figure S6). As shown in Supplemental Figure S6b,c, iPSCCDH23 cultures also exhibited a significant increase in the expression of GH and LH (Supplemental Figure S6b,c).
Cells 11 03344 g007 550
Figure 7. Single cell analysis of iPSCctrl and iPSCCDH23 cultures 15 and 30 days post-directed differentiation. (a) UMAP plots showing identified cell clusters 0–16 in iPSCctrl and iPSCCDH23 cultures 15 days post-directed differentiation. (b) Violin plots of representative identified markers of the corticotroph cell lineage, where 2 subpopulations were observed among iPSCctrl and iPSCCDH23 cultures. Arrows highlight clusters 1, 5, and 7. (c) Violin plots showing expression of genes representative of stem cells, Wnt, NOTCH, Hh and SST signaling, anterior pituitary (corticotroph) cell lineage, and cell cycle in clusters 1, 5, and 7 of iPSCCDH23 cultures. Plot width: cell number, plot height: gene expression. (d) viSNE maps showing concatenated flow cytometry standard files for both samples and iPSCctrl and iPSCCDH23 organoids 30 days post-directed differentiation. (e) Overlay of manually gated cell populations onto viSNE plots. (f) Fluorescent intensity of Ki67 of viSNE maps for both samples and iPSCctrl and iPSCCDH23 organoids. iPSCctrl = 22518 events; iPSCCDH23 = 17542 events.
Collectively, Figure 7 demonstrates that the development of pituitary organoids generated from iPSCs of CD patients may reveal the existence of cell populations which, potentially, contribute to the support of adenoma growth and progression, as well as an expansion of stem and progenitor cells that may be the targets for tumor recurrence.

4. Discussion

Our studies demonstrate the development of organoids generated from human PitNETs (hPITOs) can potentially be used to screen for the sensitivity and efficacy of responses to targeted therapies for CD patients that either fail to achieve remission or exhibit recurrence of disease after surgery. In addition, we have documented that induced pluripotent stem cells (iPSCs) generated from a CD patient expressing germline mutation CDH23 (iPSCCDH23) reveals the disease pathogenesis under directed differentiation. Many early in vitro experiments have used pituitary cell lines, spheroids, aggregates, and/or tumoroids that do not replicate the primary PitNET microenvironment [19,20,21], and lack a multicellular identity [39,40]. The development of PitNET tissue-generated organoids is limited to the use of transgenic mouse models as the source [22,23,41]. The recent organoid cultures reported by Nys et al. [42] have been generated from single stem cells isolated from PitNET tissue, and are claimed to be true organoids due to their clonality. However, multicellular complexity was not validated by the protein expression or hormone secretion from pituitary cell lineages in these cultures [42]. According to the National Cancer Institute (NCI, NIH), an ‘organoid’ is defined as “a tiny, 3-dimensional mass of tissue that is made by growing stem cells (cells from which other types of cells develop) in the laboratory” [43]. The hPITOs reported here begin from single and/or 3–4 cell clusters dissociated from the PitNET tissue that harbors the stem cells. Supplemental Video S2 demonstrates a process of ‘budding,’ as well as lumen formation as organoids grow and differentiate. We document differentiation and function by comprehensive spectral flow cytometry, ELISA, and response to standard of care drugs. The growth of PitNET organoids reported in the current study is consistent with that of gastrointestinal tissue derived cultures that begin from cell clusters, crypts, or glands [27,44,45].
Our studies report a PitNET tissue organoid culture with a multicellular identity consisting of differentiated cell lineages, stem/progenitor cells, and immune and stromal cell compartments, which replicates much of the patient’s own adenoma pathology, functionality, and complexity. We have also demonstrated that iPSCs, derived from the blood of a CD patient, can be directly differentiated into pituitary organoids that resemble similar characteristics to the tumor tissue. Many investigators have proposed the use of organoids in personalized medicine, but have focused these efforts on targeted treatment of cancers [27,46,47,48]. The findings reported in these studies are the first to implement this approach for the potential treatment of PitNETs. Collectively, we have developed a relevant human in vitro approach to potentially advance our knowledge as well as our approach to studies in the field of pituitary tumor research. Both the hPITOs and the iPSCCDH23 may be implemented in studies that strive to (1) define the molecular and cellular events that are crucial for the development of PitNETs leading to CD, and (2) accelerate the identification of effective targeted therapies for patients with CD.
While published studies have advanced our understanding of the molecular mechanisms of the pathogenesis of corticotroph adenomas and elucidated candidate therapeutic targets for CD, these reports fall short of directly informing clinical decisions for patient treatment. Using organoids to screen potential drugs and compounds can potentially improve therapeutic accuracy. Figure 3 demonstrated a variation in drug responsiveness amongst the organoid lines generated from individual patients. Importantly, there was further divergence in drug responsiveness amongst the individual organoid lines within each pathologically defined corticotroph subtype. For example, hPITOs generated from patients with sparsely granulated corticotroph adenomas (hPIT0s 10, 25, 34, 35) and Crooke’s cell adenomas (hPITOs 7, 33) showed variable responses regardless of similar pathologically defined subtypes. In addition, the response of the tumor cells within the organoids to the standard of care drugs that directly target the pituitary in the body, including mifepristone and cabergoline, was only 50% in hPITO34 and hPITO35, and almost 0% in the other lines, including hPITO7, 10, and 25. These data clearly demonstrate that the inherent patient difference to drug response that is often observed among CD patients is reflected in the organoid culture. This culture system may be an approach that will provide functional data revealing actionable treatment options for each patient. Patient-derived organoids from several tumors have served as a platform for testing the efficacy of anticancer drugs and predicting responses to targeted therapies in individual patients [27,46,48,49,50]. An example of the use of organoids in identifying drug responsiveness within an endocrine gland is that of papillary thyroid cancer [51]. Organoids developed from PTC patients were used as a preclinical model for studying responsiveness to anticancer drugs in a personalized approach [51]. However, our study is the first report of the use of hPITOs for drug screening. Connecting genetic and drug sensitivity data will further categorize corticotroph subtypes associated with CD. WES analysis of each hPITO line was compared to the results for the corresponding primary adenoma tissues. We showed the concordance rate of exonic variants between the primary tumor tissues obtained from CD patients and the corresponding organoid line. On average, approximately 80% of the variants observed in the CD patients’ adenoma tissues were retained in the corresponding hPITOs.
Pituitary organoids were also developed from iPSCs generated from PBMCs of a CD patient expressing a germline genetic alteration in cadherin-related 23 CDH23 (iPSCCDH23), a CD patient expressing an MEN1 mutation (iPSCMEN1), and a healthy individual (iPSCctrl). Foundational studies performed by investigators at the genome level have revealed significant knowledge regarding the pathophysiology of CD [36,37,52,53]. In some instances, CD is a manifestation of genetic mutation syndromes that include multiple endocrine neoplasia type 1 (MEN1), familial isolated pituitary adenoma (FIPA), and Carney complex [54,55]. CDH23 syndrome is clinically associated with the development of Usher syndrome, deafness, and vestibular dysfunction [56]. Several mutations in CDH23 are associated with inherited hearing loss and blindness [57]. However, none of the variants found in this study were linked to any symptoms of deafness or blindness. A possible explanation is that deafness-related CDH23 mutations are caused by either homozygous or compound heterozygous mutations [57]. In a study that linked mutations in CDH23 with familial and sporadic pituitary adenomas, it was suggested that these genetic alterations could play important roles in the pathogenesis of CD [38]. Genomic screening in a total of 12 families with familial PitNETs, 125 individuals with sporadic pituitary tumors, and 260 control individuals showed that 33% of the families with familial pituitary tumors and 12% of individuals with sporadic pituitary tumors expressed functional or pathogenic CDH23 variants [38]. Consistent with the expected pathology and function of a PitNET from a patient with CD, iPSCCDH23 organoids exhibited hypersecretion of ACTH, and expression of transcription factors and cell markers were reported in the pathology report for corticotroph PitNETs. Collectively, these findings warrant further investigation to determine whether carriers of CDH23 mutations are at a high risk of developing CD and/or hearing loss. Specifically, clinical investigation is required to determine whether pituitary MRI scans should be adopted in the screening of CDH23-related diseases, including Usher syndrome and age-related hearing loss.
Pituitary organoids generated from iPSCs of a CD patient revealed the existence of cell populations that potentially contribute to the support of PitNET growth and disease progression, as well as an expansion of stem and progenitor cells that may be the targets for tumor recurrence. Organoids derived from both pituitary adenomas and iPSCs exhibited increased expression of stem cell and progenitor markers at both the protein and transcriptomic levels. Unique clusters that were proliferative in the iPSCCDH23 organoids expressed a hybrid pituitary cell population which was in an epithelial/mesenchymal state (CK20+/VIM+/CXCR4+/Ki67+). In support of our findings, a similar report of a hybrid epithelial/mesenchymal pituitary cell has been made as part of the normal developmental stages of the human fetal pituitary [58]. Previous studies have suggested that pituitary stem cells undergo an EMT-like process during cell migration and differentiation [59,60,61]. Consistent with our findings are extensive studies using single cells isolated from human pituitary adenomas to show increased expression of stem cell markers SOX2 and CXCR4 [22,23,41,62,63]. Within the clusters identified in the iPSCCDH23 culture were cell populations expressing stem cell markers, including SOX2, NESTIN, CXCR4, KLF4, and CD34. The same iPSCCDH23 cell clusters, 4, 8, 9, and 11, co-expressed upregulated genes of NOTCH, Hedgehog, WNT, and TGFβ signaling, which are pivotal not only in pituitary tumorigenesis and pituitary embryonic development, but also in ‘tumor stemness’ [22,23,41,62,63,64]. We also noted that clusters of cell populations 5 and 14 unique within the iPSCCDH23 cultures expressed upregulated genes which were indicative of high proliferation. We observed upregulated expression of the E2F family of transcription factors (E2Fs) E2F1 and E2F7. These findings are of significance, given that there is evidence to show that upregulation of E2Fs is fundamental for tumorigenesis, metastasis, drug resistance, and recurrence [65]. Within the pituitary adenoma microenvironment, whether these stem cells directly differentiate into pituitary tumors or support the growth of the adenoma is largely unknown. In addition, whether pituitary stem cell populations become activated in response to injury is also understudied. Although the role of stem cells has been identified using a mouse model through implantation of the cells within the right forebrain [66], the identification of pituitary tumor-initiating stem cells using in vivo orthotopic transplantation models is impossible in mice. Pituitary tumors harboring the stem cells may require engraftment within the environment from which the cells are derived in order to enable growth and differentiation of the tumor. However, it is technically impossible to implant cells orthotopically in the murine pituitary. The pituitary tumor organoid cultures presented in these studies may offer an approach by which isolation, identification, and characterization of this stem cell population is possible. Therefore, we would gain knowledge on the mechanisms of pituitary tumor pathogenesis and reveal potential novel targets for therapeutic interventions by using the iPSC generated pituitary organoid culture.
PitNETs associated with the development of CD cause serious morbidity due to chronic cortisol exposure that dysregulates almost every organ system in the body. Overall, existing medical therapies remain suboptimal, with negative impact on health and quality of life, including considerable risk of therapy resistance and tumor recurrence. To date, little is known about the pathogenesis of PitNETs. Here, we present a human organoid-based approach that will allow us to acquire knowledge of the mechanisms underlying pituitary tumorigenesis. Such an approach is essential to identify targeted treatments and improve clinical management of patients with CD.

5. Conclusions

Cushing’s disease (CD) is a serious endocrine disorder caused by an adrenocorticotropic hormone (ACTH)-secreting pituitary neuroendocrine tumor (PitNET), which stimulates the adrenal glands to overproduce cortisol. The absence of preclinical models that replicate the PitNET microenvironment has prevented us from acquiring the knowledge to identify therapies that can be targeted to the tumor with a higher efficacy and tolerability for patients. Our studies demonstrate the development of organoids generated from human PitNETs or induced pluripotent stem cells as an essential approach to identifying targeted therapy methods for CD patients.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/cells11213344/s1, Figure S1: Antibodies used and Cytek® Full Spectrum Viewer showing calculated similarity indices; Figure S2: Morphology and proliferation of lactotroph, somatotroph, and gonadotroph hPITOs; Table S1: Pituitary Growth Media; Table S2: Components used for pituitary organoids generated from iPSCs; Table S3: clinical characteristics of pituitary adenoma samples used for the generation of organoids; Table S4: Average correlation of replicates reported in Figure 3; Table S5: pituitary cell lineage or stem cell markers used in the scRNA-seq analysis; Video S1: hPITO38 EdU ACTH 3.

Author Contributions

Conceptualization, Y.Z.; methodology, J.C., Y.Z., J.M.C., B.N.S., S.M. and K.W.P.; software, J.C., Y.Z., J.M.C., S.M., Y.C., P.M. and R.P.; validation, Y.Z., J.C., J.M.C., A.S.L., K.C.J.Y. and R.P.; formal analysis, J.C., Y.Z., J.M.C., R.P., Y.C., S.M. and P.M.; investigation, Y.Z.; resources, Y.Z., J.C., J.E., C.A.T., B.H. and A.S.L.; data curation, J.C., Y.Z., J.M.C., R.P. and S.M.; writing—original draft preparation, Y.Z., J.C, S.M., J.M.C., Y.C., B.H. and R.P.; writing—review and editing, Y.Z., J.C., J.M.C., A.S.L., K.C.J.Y., S.M., J.E., C.A.T., K.W.P., B.H., Y.C., P.M., B.N.S. and R.P.; visualization, Y.Z., J.C., J.M.C., A.S.L., K.C.J.Y. and R.P.; supervision, Y.Z.; project administration, Y.Z.; funding acquisition, Y.Z. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by the Department of Cellular and Molecular Medicine (University of Arizona College of Medicine) startup funds (Zavros). This research study was also partly supported by the National Cancer Institute of the National Institutes of Health under award number P30 CA023074 (Sweasy).

Institutional Review Board Statement

The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of St. Joseph’s Hospital and Barrow Neurological Institute Biobank collection protocol PHXA-05TS038, and collection of outcomes data protocol PHXA-0004-72-29, and patient consent (protocol date of approval).

Informed Consent Statement

Written informed consent was obtained from all subjects involved in the study.

Data Availability Statement

The datasets generated during the analysis of the present study are available in the ReDATA repository, https://doi.org/10.25422/azu.data.19755244.v1. The datasets generated in the current study are also available from the corresponding author on reasonable request. All data generated or analyzed during this study are included in this published article (and its Supplementary Information Files).

Acknowledgments

We acknowledge the technical support of Maga Sanchez in the Tissue Acquisition and Cellular/Molecular Analysis Shared Resource (TACMASR University of Arizona Cancer Center) for assistance with embedding and sectioning of organoids. We would also like to acknowledge Patty Jansma (Marley Imaging Core, University Arizona) and, Douglas W Cromey (TACMASR imaging, University of Arizona Cancer Center) for assistance in microscopy. The authors thank the patients who consented to donate pituitary tumor tissues and blood for the development of the organoids. Without their willingness to participate in the study, this work would not be possible.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Cushing, H. Posterior Pituitary Activity from an Anatomical Standpoint. Am. J. Pathol. 1933, 9, 539–548.19. [Google Scholar] [PubMed]
  2. Cushing, H. The basophil adenomas of the pituitary body and their clinical manifestations (pituitary basophilism) 1932. Obes. Res. 1994, 2, 486–508. [Google Scholar] [CrossRef] [PubMed]
  3. Ironside, N.; Chen, C.J.; Lee, C.C.; Trifiletti, D.M.; Vance, M.L.; Sheehan, J.P. Outcomes of Pituitary Radiation for Cushing’s Disease. Endocrinol. Metab. Clin. N. Am. 2018, 47, 349–365. [Google Scholar] [CrossRef]
  4. Loriaux, D.L. Diagnosis and Differential Diagnosis of Cushing’s Syndrome. N. Engl. J. Med. 2017, 377, e3. [Google Scholar] [CrossRef]
  5. Asa, S.L.; Mete, O.; Perry, A.; Osamura, R.Y. Overview of the 2022 WHO Classification of Pituitary Tumors. Endocr. Pathol. 2022, 33, 6–26. [Google Scholar] [CrossRef]
  6. Nishioka, H.; Yamada, S. Cushing’s Disease. J. Clin. Med. 2019, 8, 1951. [Google Scholar] [CrossRef] [PubMed]
  7. Feelders, R.A.; Hofland, L.J. Medical treatment of Cushing’s disease. J. Clin. Endocrinol. Metab. 2013, 98, 425–438. [Google Scholar] [CrossRef]
  8. Limumpornpetch, P.; Morgan, A.W.; Tiganescu, A.; Baxter, P.D.; Nyawira Nyaga, V.; Pujades-Rodriguez, M.; Stewart, P.M. The Effect of Endogenous Cushing Syndrome on All-cause and Cause-specific Mortality. J. Clin. Endocrinol. Metab. 2022, 107, 2377–2388. [Google Scholar] [CrossRef]
  9. Ciric, I.; Zhao, J.C.; Du, H.; Findling, J.W.; Molitch, M.E.; Weiss, R.E.; Refetoff, S.; Kerr, W.D.; Meyer, J. Transsphenoidal surgery for Cushing disease: Experience with 136 patients. Neurosurgery 2012, 70, 70–80; discussion 71–80. [Google Scholar] [CrossRef]
  10. Alexandraki, K.I.; Kaltsas, G.A.; Isidori, A.M.; Storr, H.L.; Afshar, F.; Sabin, I.; Akker, S.A.; Chew, S.L.; Drake, W.M.; Monson, J.P.; et al. Long-term remission and recurrence rates in Cushing’s disease: Predictive factors in a single-centre study. Eur. J. Endocrinol. 2013, 168, 639–648. [Google Scholar] [CrossRef]
  11. Sonino, N.; Zielezny, M.; Fava, G.A.; Fallo, F.; Boscaro, M. Risk factors and long-term outcome in pituitary-dependent Cushing’s disease. J. Clin. Endocrinol. Metab. 1996, 81, 2647–2652. [Google Scholar] [CrossRef] [PubMed]
  12. Van der Pas, R.; Feelders, R.A.; Gatto, F.; De Bruin, C.; Pereira, A.M.; Van Koetsveld, P.M.; Sprij-Mooij, D.M.; Waaijers, A.M.; Dogan, F.; Schulz, S.; et al. Preoperative normalization of cortisol levels in Cushing’s disease after medical treatment: Consequences for somatostatin and dopamine receptor subtype expression and in vitro response to somatostatin analogs and dopamine agonists. J. Clin. Endocrinol. Metab. 2013, 98, E1880–E1890. [Google Scholar] [CrossRef]
  13. Kondziolka, D. Cushing’s disease and stereotactic radiosurgery. J. Neurosurg. 2013, 119, 1484–1485; discussion 1485. [Google Scholar] [CrossRef] [PubMed]
  14. Mehta, G.U.; Sheehan, J.P.; Vance, M.L. Effect of stereotactic radiosurgery before bilateral adrenalectomy for Cushing’s disease on the incidence of Nelson’s syndrome. J. Neurosurg. 2013, 119, 1493–1497. [Google Scholar] [CrossRef] [PubMed]
  15. Tritos, N.A. Adrenally Directed Medical Therapies for Cushing Syndrome. J. Clin. Endocrinol. Metab. 2021, 106, 16–25. [Google Scholar] [CrossRef] [PubMed]
  16. Gheorghiu, M.L.; Negreanu, F.; Fleseriu, M. Updates in the Medical Treatment of Pituitary Adenomas. Horm. Metab. Res. 2020, 52, 8–24. [Google Scholar] [CrossRef]
  17. Kaiser, U.B. Cushing’s disease: Towards precision medicine. Cell. Res. 2015, 25, 649–650. [Google Scholar] [CrossRef]
  18. Bissell, M.S.a.M.J. Organoids: A historical perspective of thinking in three dimensions. J. Cell Biol. 2017, 216, 31–40. [Google Scholar] [CrossRef]
  19. Danila, D.C.; Zhang, X.; Zhou, Y.; Dickersin, G.R.; Fletcher, J.A.; Hedley-Whyte, E.T.; Selig, M.K.; Johnson, S.R.; Klibanski, A. A human pituitary tumor-derived folliculostellate cell line. J. Clin. Endocrinol. Metab. 2000, 85, 1180–1187. [Google Scholar] [CrossRef]
  20. Bjoro, T.; Torjesen, P.A.; Ostberg, B.C.; Sand, O.; Iversen, J.G.; Gautvik, K.M.; Haug, E. Bombesin stimulates prolactin secretion from cultured rat pituitary tumour cells (GH4C1) via activation of phospholipase C. Regul. Pept. 1987, 19, 169–182. [Google Scholar] [CrossRef]
  21. Bjoro, T.; Sand, O.; Ostberg, B.C.; Gordeladze, J.O.; Torjesen, P.; Gautvik, K.M.; Haug, E. The mechanisms by which vasoactive intestinal peptide (VIP) and thyrotropin releasing hormone (TRH) stimulate prolactin release from pituitary cells. Biosci. Rep. 1990, 10, 189–199. [Google Scholar] [CrossRef] [PubMed]
  22. Cox, B.; Laporte, E.; Vennekens, A.; Kobayashi, H.; Nys, C.; Van Zundert, I.; Uji, I.H.; Vercauteren Drubbel, A.; Beck, B.; Roose, H.; et al. Organoids from pituitary as a novel research model toward pituitary stem cell exploration. J. Endocrinol. 2019, 240, 287–308. [Google Scholar] [CrossRef] [PubMed]
  23. Vennekens, A.; Laporte, E.; Hermans, F.; Cox, B.; Modave, E.; Janiszewski, A.; Nys, C.; Kobayashi, H.; Malengier-Devlies, B.; Chappell, J.; et al. Interleukin-6 is an activator of pituitary stem cells upon local damage, a competence quenched in the aging gland. Proc. Natl. Acad. Sci. USA 2021, 118, e2100052118. [Google Scholar] [CrossRef] [PubMed]
  24. Zhang, D.; Hugo, W.; Redublo, P.; Miao, H.; Bergsneider, M.; Wang, M.B.; Kim, W.; Yong, W.H.; Heaney, A.P. A human ACTH-secreting corticotroph tumoroid model: Novel Human ACTH-Secreting Tumor Cell in vitro Model. EBioMedicine 2021, 66, 103294. [Google Scholar] [CrossRef] [PubMed]
  25. Tsukada, T.; Kouki, T.; Fujiwara, K.; Ramadhani, D.; Horiguchi, K.; Kikuchi, M.; Yashiro, T. Reassembly of anterior pituitary organization by hanging drop three-dimensional cell culture. Acta. Histochem. Cytochem. 2013, 46, 121–127. [Google Scholar] [CrossRef] [PubMed]
  26. Narsinh, K.H.; Jia, F.; Robbins, R.C.; Kay, M.A.; Longaker, M.T.; Wu, J.C. Generation of adult human induced pluripotent stem cells using nonviral minicircle DNA vectors. Nat. Protoc. 2011, 6, 78–88. [Google Scholar] [CrossRef]
  27. Steele, N.G.; Chakrabarti, J.; Wang, J.; Biesiada, J.; Holokai, L.; Chang, J.; Nowacki, L.M.; Hawkins, J.; Mahe, M.; Sundaram, N.; et al. An Organoid-Based Preclinical Model of Human Gastric Cancer. Cell. Mol. Gastroenterol. Hepatol. 2019, 7, 161–184. [Google Scholar] [CrossRef]
  28. Bertaux-Skeirik, N.; Feng, R.; Schumacher, M.A.; Li, J.; Mahe, M.M.; Engevik, A.C.; Javier, J.E.; Peek, R.M.J.; Ottemann, K.; Orian-Rousseau, V.; et al. CD44 plays a functional role in Helicobacter pylori-induced epithelial cell proliferation. PLoS Pathog. 2015, 11, e1004663. [Google Scholar] [CrossRef]
  29. Feng, R.; Aihara, E.; Kenny, S.; Yang, L.; Li, J.; Varro, A.; Montrose, M.H.; Shroyer, N.F.; Wang, T.C.; Shivdasani, R.A.; et al. Indian Hedgehog mediates gastrin-induced proliferation in stomach of adult mice. Gastroenterology 2014, 147, 655–666.e9. [Google Scholar] [CrossRef]
  30. Filippi-Chiela, E.C.; Oliveira, M.M.; Jurkovski, B.; Callegari-Jacques, S.M.; da Silva, V.D.; Lenz, G. Nuclear morphometric analysis (NMA): Screening of senescence, apoptosis and nuclear irregularities. PLoS ONE 2012, 7, e42522. [Google Scholar] [CrossRef]
  31. Gagnon, R.C.; Peterson, J.J. Estimation of confidence intervals for area under the curve from destructively obtained pharmacokinetic data. J. Pharm. Biopharm. 1998, 26, 87–102. [Google Scholar] [CrossRef] [PubMed]
  32. Livak, K.; Schmittgen, T. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001, 25, 402–408. [Google Scholar] [CrossRef] [PubMed]
  33. Hinojosa-Amaya, J.M.; Varlamov, E.V.; McCartney, S.; Fleseriu, M. Hypercortisolemia Recurrence in Cushing’s Disease; a Diagnostic Challenge. Front. Endocrinol. 2019, 10, 740. [Google Scholar] [CrossRef] [PubMed]
  34. Patil, C.G.; Prevedello, D.M.; Lad, S.P.; Vance, M.L.; Thorner, M.O.; Katznelson, L.; Laws, E.R., Jr. Late recurrences of Cushing’s disease after initial successful transsphenoidal surgery. J. Clin. Endocrinol. Metab. 2008, 93, 358–362. [Google Scholar] [CrossRef]
  35. Katznelson, L. Bilateral adrenalectomy for Cushing’s disease. Pituitary 2015, 18, 269–273. [Google Scholar] [CrossRef]
  36. Reincke, M.; Sbiera, S.; Hayakawa, A.; Theodoropoulou, M.; Osswald, A.; Beuschlein, F.; Meitinger, T.; Mizuno-Yamasaki, E.; Kawaguchi, K.; Saeki, Y.; et al. Mutations in the deubiquitinase gene USP8 cause Cushing’s disease. Nat. Genet. 2015, 47, 31–38. [Google Scholar] [CrossRef]
  37. Chen, J.; Jian, X.; Deng, S.; Ma, Z.; Shou, X.; Shen, Y.; Zhang, Q.; Song, Z.; Li, Z.; Peng, H.; et al. Identification of recurrent USP48 and BRAF mutations in Cushing’s disease. Nat. Commun. 2018, 9, 3171. [Google Scholar] [CrossRef]
  38. Zhang, Q.; Peng, C.; Song, J.; Zhang, Y.; Chen, J.; Song, Z.; Shou, X.; Ma, Z.; Peng, H.; Jian, X.; et al. Germline Mutations in CDH23, Encoding Cadherin-Related 23, Are Associated with Both Familial and Sporadic Pituitary Adenomas. Am. J. Hum. Genet. 2017, 100, 817–823. [Google Scholar] [CrossRef]
  39. Ikeda, H.; Mitsuhashi, T.; Kubota, K.; Kuzuya, N.; Uchimura, H. Epidermal growth factor stimulates growth hormone secretion from superfused rat adenohypophyseal fragments. Endocrinology 1984, 115, 556–558. [Google Scholar] [CrossRef]
  40. Baek, N.; Seo, O.W.; Kim, M.; Hulme, J.; An, S.S. Monitoring the effects of doxorubicin on 3D-spheroid tumor cells in real-time. Onco. Targets 2016, 9, 7207–7218. [Google Scholar] [CrossRef]
  41. Laporte, E.; Nys, C.; Vankelecom, H. Development of Organoids from Mouse Pituitary as In Vitro Model to Explore Pituitary Stem Cell Biology. J. Vis. Exp. 2022. [Google Scholar] [CrossRef] [PubMed]
  42. Nys, C.; Lee, Y.L.; Roose, H.; Mertens, F.; De Pauw, E.; Kobayashi, H.; Sciot, R.; Bex, M.; Versyck, G.; De Vleeschouwer, S.; et al. Exploring stem cell biology in pituitary tumors and derived organoids. Endocr. Relat. Cancer 2022, 29, 427–450. [Google Scholar] [CrossRef]
  43. Available online: https://www.cancer.gov/publications/dictionaries/cancer-terms/def/organoid (accessed on 20 September 2022).
  44. Mahe, M.M.; Aihara, E.; Schumacher, M.A.; Zavros, Y.; Montrose, M.H.; Helmrath, M.A.; Sato, T.; Shroyer, N.F. Establishment of Gastrointestinal Epithelial Organoids. Curr. Protoc. Mouse Biol. 2013, 3, 217–240. [Google Scholar] [CrossRef] [PubMed]
  45. Schumacher, M.A.; Aihara, E.; Feng, R.; Engevik, A.; Shroyer, N.F.; Ottemann, K.M.; Worrell, R.T.; Montrose, M.H.; Shivdasani, R.A.; Zavros, Y. The use of murine-derived fundic organoids in studies of gastric physiology. J. Physiol. 2015, 593, 1809–1827. [Google Scholar] [CrossRef] [PubMed]
  46. Holokai, L.; Chakrabarti, J.; Lundy, J.; Croagh, D.; Adhikary, P.; Richards, S.S.; Woodson, C.; Steele, N.; Kuester, R.; Scott, A.; et al. Murine- and Human-Derived Autologous Organoid/Immune Cell Co-Cultures as Pre-Clinical Models of Pancreatic Ductal Adenocarcinoma. Cancers 2020, 12, 3816. [Google Scholar] [CrossRef] [PubMed]
  47. Boj, S.F.; Hwang, C.I.; Baker, L.A.; Chio, I.I.C.; Engle, D.D.; Corbo, V.; Jager, M.; Ponz-Sarvise, M.; Tiriac, H.; Spector, M.S.; et al. Organoid models of human and mouse ductal pancreatic cancer. Cell 2015, 160, 324–338. [Google Scholar] [CrossRef]
  48. Tiriac, H.; Belleau, P.; Engle, D.D.; Plenker, D.; Deschenes, A.; Somerville, T.D.D.; Froeling, F.E.M.; Burkhart, R.A.; Denroche, R.E.; Jang, G.H.; et al. Organoid Profiling Identifies Common Responders to Chemotherapy in Pancreatic Cancer. Cancer Discov. 2018, 8, 1112–1129. [Google Scholar] [CrossRef]
  49. Driehuis, E.; van Hoeck, A.; Moore, K.; Kolders, S.; Francies, H.E.; Gulersonmez, M.C.; Stigter, E.C.A.; Burgering, B.; Geurts, V.; Gracanin, A.; et al. Pancreatic cancer organoids recapitulate disease and allow personalized drug screening. Proc. Natl. Acad. Sci. USA 2019 116, 26580–26590. [CrossRef]
  50. Jung, Y.H.; Choi, D.H.; Park, K.; Lee, S.B.; Kim, J.; Kim, H.; Jeong, H.W.; Yang, J.H.; Kim, J.A.; Chung, S.; et al. Drug screening by uniform patient derived colorectal cancer hydro-organoids. Biomaterials 2021, 276, 121004. [Google Scholar] [CrossRef]
  51. Chen, D.; Tan, Y.; Li, Z.; Li, W.; Yu, L.; Chen, W.; Liu, Y.; Liu, L.; Guo, L.; Huang, W.; et al. Organoid Cultures Derived From Patients With Papillary Thyroid Cancer. J. Clin. Endocrinol. Metab. 2021, 106, 1410–1426. [Google Scholar] [CrossRef]
  52. Reincke, M.; Theodoropoulou, M. Genomics in Cushing’s Disease: The Dawn of a New Era. J. Clin. Endocrinol. Metab. 2021, 106, e2455–e2456. [Google Scholar] [CrossRef] [PubMed]
  53. Ma, Z.Y.; Song, Z.J.; Chen, J.H.; Wang, Y.F.; Li, S.Q.; Zhou, L.F.; Mao, Y.; Li, Y.M.; Hu, R.G.; Zhang, Z.Y.; et al. Recurrent gain-of-function USP8 mutations in Cushing’s disease. Cell. Res. 2015, 25, 306–317. [Google Scholar] [CrossRef] [PubMed]
  54. Melmed, S. Pathogenesis of pituitary tumors. Nat. Rev. Endocrinol. 2011, 7, 257–266. [Google Scholar] [CrossRef] [PubMed]
  55. Stratakis, C.A.; Tichomirowa, M.A.; Boikos, S.; Azevedo, M.F.; Lodish, M.; Martari, M.; Verma, S.; Daly, A.F.; Raygada, M.; Keil, M.F.; et al. The role of germline AIP, MEN1, PRKAR1A, CDKN1B and CDKN2C mutations in causing pituitary adenomas in a large cohort of children, adolescents, and patients with genetic syndromes. Clin. Genet. 2010, 78, 457–463. [Google Scholar] [CrossRef]
  56. Mouchtouris, N.; Smit, R.D.; Piper, K.; Prashant, G.; Evans, J.J.; Karsy, M. A review of multiomics platforms in pituitary adenoma pathogenesis. Front. Biosci. 2022, 27, 77. [Google Scholar] [CrossRef] [PubMed]
  57. Bolz, H.; von Brederlow, B.; Ramirez, A.; Bryda, E.C.; Kutsche, K.; Nothwang, H.G.; Seeliger, M.; del, C.S.C.M.; Vila, M.C.; Molina, O.P.; et al. Mutation of CDH23, encoding a new member of the cadherin gene family, causes Usher syndrome type 1D. Nat. Genet. 2001, 27, 108–112. [Google Scholar] [CrossRef] [PubMed]
  58. Zhang, S.; Cui, Y.; Ma, X.; Yong, J.; Yan, L.; Yang, M.; Ren, J.; Tang, F.; Wen, L.; Qiao, J. Single-cell transcriptomics identifies divergent developmental lineage trajectories during human pituitary development. Nat. Commun. 2020, 11, 5275. [Google Scholar] [CrossRef]
  59. Cheung, L.Y.; Davis, S.W.; Brinkmeier, M.L.; Camper, S.A.; Perez-Millan, M.I. Regulation of pituitary stem cells by epithelial to mesenchymal transition events and signaling pathways. Mol. Cell. Endocrinol. 2017, 445, 14–26. [Google Scholar] [CrossRef]
  60. Shintani, A.; Higuchi, M. Isolation of PRRX1-positive adult pituitary stem/progenitor cells from the marginal cell layer of the mouse anterior lobe. Stem Cell. Res. 2021, 52, 102223. [Google Scholar] [CrossRef]
  61. Yoshida, S.; Nishimura, N.; Ueharu, H.; Kanno, N.; Higuchi, M.; Horiguchi, K.; Kato, T.; Kato, Y. Isolation of adult pituitary stem/progenitor cell clusters located in the parenchyma of the rat anterior lobe. Stem Cell. Res. 2016, 17, 318–329. [Google Scholar] [CrossRef]
  62. Laporte, E.; Vennekens, A.; Vankelecom, H. Pituitary Remodeling Throughout Life: Are Resident Stem Cells Involved? Front. Endocrinol. 2020, 11, 604519. [Google Scholar] [CrossRef] [PubMed]
  63. Vankelecom, H.; Roose, H. The Stem Cell Connection of Pituitary Tumors. Front. Endocrinol. 2017, 8, 339. [Google Scholar] [CrossRef] [PubMed]
  64. Mertens, F.; Gremeaux, L.; Chen, J.; Fu, Q.; Willems, C.; Roose, H.; Govaere, O.; Roskams, T.; Cristina, C.; Becu-Villalobos, D.; et al. Pituitary tumors contain a side population with tumor stem cell-associated characteristics. Endocr. Relat. Cancer 2015, 22, 481–504. [Google Scholar] [CrossRef] [PubMed]
  65. Chen, H.Z.; Tsai, S.Y.; Leone, G. Emerging roles of E2Fs in cancer: An exit from cell cycle control. Nat. Rev. Cancer 2009, 9, 785–797. [Google Scholar] [CrossRef] [PubMed]
  66. Xu, Q.; Yuan, X.; Tunici, P.; Liu, G.; Fan, X.; Xu, M.; Hu, J.; Hwang, J.Y.; Farkas, D.L.; Black, K.L.; et al. Isolation of tumour stem-like cells from benign tumours. Br. J. Cancer 2009, 101, 303–311. [Google Scholar] [CrossRef] [PubMed]
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Evaluation of Ketoconazole As a Treatment for Cushing’s Disease in a Retrospective Cohort

Objective: The first-line treatment for Cushing’s disease is transsphenoidal surgery, after which the rates of remission are 60 to 80%, with long-term recurrence of 20 to 30%, even in those with real initial remission. Drug therapies are indicated for patients without initial remission or with surgical contraindications or recurrence, and ketoconazole is one of the main available therapies. The objective of this study was to evaluate the safety profile of and the treatment response to ketoconazole in Cushing’s disease patients followed up at the endocrinology outpatient clinic of a Brazilian university hospital.

Patients and methods: This was a retrospective cohort of Cushing’s disease patients with active hypercortisolism who used ketoconazole at any stage of follow-up. Patients who were followed up for less than 7 days, who did not adhere to treatment, or who were lost to follow-up were excluded.

Results: Of the 172 Cushing’s disease patients who were followed up between 2004 and 2020, 38 received ketoconazole. However, complete data was only available for 33 of these patients. Of these, 26 (78%) underwent transsphenoidal surgery prior to using ketoconazole, five of whom (15%) had also undergone radiotherapy; seven used ketoconazole as a primary treatment. Ketoconazole use ranged from 14 days to 14.5 years. A total of 22 patients had a complete response (66%), three patients had a partial response (9%), and eight patients had no response to treatment (24%), including those who underwent radiotherapy while using ketoconazole. Patients whose hypercortisolism was controlled or partially controlled with ketoconazole had lower baseline 24-h urinary free cortisol levels than the uncontrolled group [times above the upper limit of normal: 0.62 (SD, 0.41) vs. 5.3 (SD, 8.21); p < 0.005, respectively] in addition to more frequent previous transsphenoidal surgery (p < 0.04). The prevalence of uncontrolled patients remained stable over time (approximately 30%) despite ketoconazole dose adjustments or association with other drugs, which had no significant effect. One patient received adjuvant cabergoline from the beginning of the follow-up, and it was prescribed to nine others due to clinical non-response to ketoconazole alone. Ten patients (30%) reported mild adverse effects, such as nausea, vomiting, dizziness, and loss of appetite. Only four patients had serious adverse effects that warranted discontinuation. There were 20 confirmed episodes of hypokalemia among 10/33 patients (30%).

Conclusion: Ketoconazole effectively controlled hypercortisolism in 66% of Cushing’s disease patients, being a relatively safe drug for those without remission after transsphenoidal surgery or whose symptoms must be controlled until a new definitive therapy is carried out. Hypokalemia is a frequent metabolic effect not yet described in other series, which should be monitored during treatment.

Introduction

Cushing’s disease (CD) results from a pituitary tumor that secretes adrenocorticotropic hormone (ACTH), which leads to chronic hypercortisolism. It is a potentially fatal disease with high morbidity and a mortality rate of up to 3.7 times than that of the general population (14) associated to several clinical–metabolic disorders caused by excess cortisol and/or loss of circadian rhythm (5). In general, its management is a challenge even in reference centers (67).

Transsphenoidal surgery (TSS), the treatment of choice for CD, results in short-term remission in 60 to 80% of patients (8). However, recurrence rates of 20 to 30% are found in long-term follow-up, even in those with clear initial remission (9). Drug therapies can help control excess cortisol in patients without initial remission, in cases of recurrence, and in those with contraindications or high initial surgical risk (10).

Nevertheless, specific drugs that act on the pituitary adenoma, which could directly treat excess ACTH, have a limited effect, and only pasireotide is approved for this purpose in Brazil (1112). In this scenario, adrenal steroidogenesis blockers are important. One such off-label medication is the antifungal drug ketoconazole, a synthetic imidazole derivative that inhibits the enzymes CYP11A1, CYP17, CYP11B2, and CYP11B1. Because of its hepatotoxicity and the availability of other drugs, it has been withdrawn from the market in several countries (13). In Europe, it is still approved for use in CD, although in the United States, it is recommended for off-label use almost in CD (1416). Due to the potential benefits for hypercortisolism, ketoconazole has been replaced by levoketoconazole, which the European Union has recently approved for CD with a lower expected hepatotoxicity (17).

Thus, when adrenal inhibitors are used as an alternative treatment for CD, information about the outcomes of drugs such as ketoconazole are important. Clinical studies on these effects in CD are scarce, mostly retrospective, multicenter, or from developed countries (1418). A recent meta-analysis on the therapeutic modalities for CD included only four studies (246 patients) that evaluated urinary cortisol response as a treatment outcome and eight studies (366 patients) describing the prevalence of some side effects: change in transaminase activity, digestive symptoms, skin rash, and adrenal insufficiency. Hypokalemia was not mentioned in this meta-analysis (19).

The objective of this study was to evaluate the safety profile of and treatment response to ketoconazole in CD patients followed during a long term in the endocrinology outpatient clinic of a Brazilian university hospital.

Patients and methods

Patients

We retrospectively evaluated 38 patients (27 women) diagnosed with CD. These patients, whose treatment included ketoconazole at any time between 2004 and 2020, are part of a prospective cohort series from the Hospital de Clínicas de Porto Alegre neuroendocrinology outpatient clinic.

The diagnostic criteria for hypercortisolism were based on high 24-h urinary free cortisol levels (24-h UFC) in at least two samples, non-suppression of serum cortisol after low-dose dexamethasone testing (>1.8 µg/dl), and/or loss of cortisol rhythm (midnight serum cortisol >7.5 µg/dl or midnight salivary cortisol >0.208 nmol/L). CD was diagnosed by normal or elevated ACTH levels, evidence of pituitary adenoma >0.6 cm on magnetic resonance image (MRI), and ACTH central/periphery gradient on inferior petrosal sinus catheterization when MRI was normal or showed an adenoma <0.6 cm.

CD was considered to be in remission after the improvement of hypercortisolism symptoms or clinical signs of adrenal insufficiency, associated with serum cortisol within reference values, normalization of 24-h UFC and/or serum cortisol <1.8 μg/dl at 8 am after 1 mg dexamethasone overnight, and/or normalization of midnight serum or salivary cortisol. In patients with active disease, to evaluate the ketoconazole treatment response, 24-h UFC was used as a laboratory parameter, as recommended in similar publications (14162021), but in some cases, we considered elevated late night salivary cortisol and/or 1 mg dexamethasone overnight cortisol (even with normal 24-h UFC), given the greater assessment sensitivity seen through these two methods in the detection of early recurrence when compared with 24-h UFC (22).

Inclusion criteria

We included patients with CD and active hypercortisolism who used ketoconazole either as primary treatment, after TSS without hypercortisolism remission, or after a recurrence.

Exclusion criteria

We excluded patients with CD and active hypercortisolism who used ketoconazole but had <7 days of follow-up, irregular outpatient follow-up, treatment non-adherence, and incomplete medical records or those who were lost to follow-up.

Evaluated parameters

Prior to ketoconazole treatment, all patients underwent an assessment of pituitary function and hypercortisolism, including serum cortisol, ACTH, 24-hour UFC, cortisol suppression after 1 mg dexamethasone overnight, midnight serum cortisol, and/or midnight salivary cortisol. The evaluated parameters were sex, age at diagnosis, weight, height, prevalence and severity of hypertension and DM, pituitary tumor characteristics, prior treatment (surgery, radiotherapy, or other medications), symptoms at disease onset, biochemical tests (renal function, hepatic function, and lipid profile), number of medications used to treat associated comorbidities, data on medication tolerance, and reasons for discontinuation, when necessary.

The clinical parameters observed during treatment were control of blood pressure and hyperglycemia, anthropometric measurements (weight, height, and body mass index), jaundice, and any other symptoms or adverse effects reported by patients.

The biochemical evaluation included fasting glucose, glycated hemoglobin, lipid profile (total cholesterol, high-density lipoprotein, low-density lipoprotein, and triglycerides), markers of liver damage (transaminases, bilirubin, gamma-glutamyl transferase, and alkaline phosphatase), electrolytes (sodium and potassium), and renal function (creatinine and urea). Hypecortisolism was accessed preferentially by 24-h UFC, however, late-night salivary cortisol and cortisol after 1 mg overnight dexamethasone could also be used.

Study design

This retrospective cohort study included patients with CD who were followed up at the Hospital de Clínicas de Porto Alegre Endocrinology Division, with their medical records from the first outpatient visit and throughout clinical follow-up collected. This study was approved by the Hospital de Clínicas de Porto Alegre Research Ethics Committee (number 74555617.0.0000.5327).

Outcomes

Hypercortisolism was considered controlled when the 24-h UFC and/or late-night salivary cortisol (LNSC) and/or overnight 1 mg dexamethasone suppression test (DST) levels were normalized in at least two consecutive assessments. Hypercortisolism was considered partially controlled when there was a 50% over-reduction in 24-h UFC and/or LNSC and/or DST levels but still above normal. A reduction lower than 50% in these parameters was considered as non-response.

We also assessed the ketoconazole doses that resulted in 24-h UFC normalization, maximum dose, medication tolerance, adverse effects, and changes in liver, kidney, and biochemical function. Due to the characteristics of this study, these outcomes were periodically evaluated in all patient consultations, which occurred usually every 2 to 4 months.

Data collection

This retrospective cohort evaluated outpatient medical records and any tests indicated by the attending physician as a pragmatic study. Ketoconazole use followed the department’s care protocol, which is based on national and international guidelines (4), and all patients received a similar care routine: the recommended initial prescription was generally taken in two to six doses at 100 to 300 mg/day. It was then increased by 200 mg every 2 to 4 months until hypercortisolism was controlled or side effects developed, especially those related to liver function. The maximum prescription was 1,200 mg/day. Clinical follow-up of these patients was performed 30 days after starting the medication and every 2–4 months thereafter (23).

Clinical, anthropometric, laboratory, and other exam data were collected through a review of the hospital’s electronic medical records for the entire follow-up period. Data from the first and last consultation were considered in the final analysis of all parameters.

Statistical analysis

Baseline population characteristics were described as mean and standard deviation (SD) or median with interquartile ranges (25–75) for continuous variables. The chi-square test was used to compare qualitative variables, and Student’s t-test or ANOVA was used to compare the quantitative variables. The Mann–Whitney U-test was used for unpaired data. P-values <0.05 were considered significant. Statistical analysis was performed in SPSS 18.0 (SPSS Inc., Chicago, IL, USA) and R package geepack 1.3-1.

Results

Treatment with ketoconazole was indicated for 41 of the 172 CD patients. In 3/41 patients, ketoconazole was unallowed due to concomitant liver disease, and 38 received ketoconazole during CD treatment between 2004 and 2020. Of these, five were excluded due to insufficient data to determine the response to ketoconazole (short treatment time, irregular follow-up, incomplete medical records, or lost to follow-up). The baseline characteristics of every sample are shown in Table 1. Thus, 33/41 patients were included in the final analysis. The patients were predominantly women (84.2%) and white (89.5%); 11 had microadenoma, 15 had macroadenoma, and 11 had no adenoma visualized. In 12/33 patients, pituitary imaging was not performed immediately before starting ketoconazole. Hypertension was observed in 26 patients (78%) and DM in 12 patients (36%). The mean age at CD diagnosis was 31.7 years.

Table 1
www.frontiersin.orgTABLE 1 Baseline clinical data of Cushing’s disease patients treated with ketoconazole.

Of the 33 patients with complete data, 26 (78%) underwent TSS prior to starting ketoconazole, five of whom (15%) had also undergone radiotherapy. Thus, seven patients used ketoconazole as primary treatment since performing a surgical procedure was impossible at that time. Of these, four had no response to ketoconazole, one had a partial response, and two had a complete response. At follow-up, four of these patients underwent their first TSS, and three continued the ketoconazole therapy, achieving full UFC control. Among those who used ketoconazole after TSS (n = 26), 20 had a complete response, two had a partial response, and four had no response. Figure 1 shows the study flow chart and patient distribution throughout the treatment.

Figure 1
www.frontiersin.orgFIGURE 1 Flowchart of ketoconazole treatment in Cushing’s disease patients.

Individual patient data are described in Table 2. The duration of ketoconazole use ranged from 14 days (in one patient who used it pre-TSS) to 14.5 years. The total follow-up time of the 22 patients with controlled CD ranged from 3 months to 14.5 years, with a mean of 5.33 years and a median of 4.8 years.

Table 2
www.frontiersin.orgTABLE 2 Individual data.

Therapeutic response

Relative therapeutic response data are described in Table 3. Patients whose hypercortisolism was controlled or partially controlled with ketoconazole had lower baseline 24-h UFC than the uncontrolled group [times above the upper limit of normal: 0.62 (SD, 0.41) vs. 5.3 (SD, 8.21); p < 0.005, respectively], in addition to more frequent prior TSS (p < 0.04). In some patients (4/33), 24-h UFC was in the normal range at the beginning of ketoconazole therapy, but they were prescribed with the medication due to the clinical recurrence of CD associated to cortisol non-suppression after 1 mg dexamethasone overnight and/or abnormal midnight salivary or serum cortisol.

Table 3
www.frontiersin.orgTABLE 3 Baseline characteristics of Cushing’s disease patients according to therapeutic response to ketoconazole.

Figure 2 shows that the prevalence of uncontrolled patients remained stable over time (approximately 30%) despite dose adjustments or association with other drugs, which led to no differences. When analyzing only the results of the last follow-up visit (eliminating fluctuations during follow-up), 22 patients had a complete response (66%), three patients had a partial response (9%), and eight patients had no response to ketoconazole treatment (24%), which includes patients who underwent radiotherapy during ketoconazole treatment.

Figure 2
www.frontiersin.orgFIGURE 2 Prevalence of controlled hypercortisolism during follow-up of Cushing’s disease patients treatesd with ketoconazole.

During follow-up, no significant differences were found in blood pressure control or in dehydroepiandrosterone sulfate, cortisol, ACTH, or glucose levels. Worsening of hypertension control was observed in association with hypokalemia in some cases, as described in side effects. The ketoconazole doses ranged from 100 to 1,200 mg per day, and there were no significant dose or response differences between the groups (Table 4). Figure 3 shows the patients, their dosages, and 24-h UFC control at the first and last consultation, showing a trend toward hypercortisolism reduction in approximately 70% of the cohort (25 of 33). Only four patients used doses lower than 300 mg at the end of follow-up. One of them used before TSS and suspended its use after surgery. One patient, who has already undergone radiotherapy, discontinued ketoconazole due to intolerance, despite adequate control of hypercortisolism. Another one, who had also undergone radiotherapy, was lost to follow-up when it was controlled using 100 mg daily, and one remained controlled using 200 mg, without previous radiotherapy.

Table 4
www.frontiersin.orgTABLE 4 Final dose of ketoconazole used in patients with Cushing’s disease.

Figure 3
www.frontiersin.orgFIGURE 3 First and last consultation 24çhour UFC results vs. ketoconazole dosage in Cushing’s disease patients.

Side effects

Regarding adverse effects (Table 5), there was no significant difference between the controlled/partially controlled group and the uncontrolled group regarding liver enzyme changes or drug intolerance. Mild adverse effects, including nausea, vomiting, dizziness, and loss of appetite, occurred in 10 patients (30%). Only four patients had serious adverse effects that warranted discontinuing the medication. In two cases, ketoconazole was discontinued due to a significantly acute increase in liver enzymes (drug-induced hepatitis) during the use of 400 and 800 mg of ketoconazole. Non-significant elevation of transaminases (up to three times the normal value) was observed in three cases. A slight increase in gamma-glutamyltransferase occurred in six patients. In these nine patients with elevated liver markers, the daily dose ranged from 400 to 1,200 mg. None of those with mild increases in liver markers needed to discontinue ketoconazole.

Table 5
www.frontiersin.orgTABLE 5 Adverse effects of ketoconazole in Cushing’s disease patients treated with ketoconazole.

One female patient developed pseudotumor cerebri syndrome, which was treated with acetazolamide. She did not need to discontinue ketoconazole, having used it for more than 10 years without new side effects and achieving complete control of hypercortisolism (24). Another patient became pregnant during follow-up while using the medication, but no maternal or fetal complications occurred (25).

Hypokalemia was also observed during follow-up. Twenty episodes of reduced potassium levels occurred in 10 patients over the course of treatment. Of these episodes, six occurred in controlled patients, three in partially controlled patients, and 11 in uncontrolled patients (Table 6). The hypokalemia was managed with spironolactone (25 to 100 mg) and oral potassium supplementation.

Table 6
www.frontiersin.orgTABLE 6 Characteristics of Cushing’s disease patients who developed hypokalemia during ketoconazole treatment.

Ketoconazole and associations

Of the patients who used an association of cabergoline and ketoconazole, one did so since the beginning of follow-up, while another nine were prescribed cabergoline during follow-up due to non-response to ketoconazole alone. Of these 10 patients, two did not start the medication due to problems in obtaining the drug. Thus, in two of the nine patients on the maximum tolerated dose of ketoconazole or who could not tolerate a higher dose due to hepatic enzymatic changes, 1.5–4.5 mg of cabergoline per week was associated. In patients not controlled with ketoconazole plus cabergoline, mitotane (two patients) or pasireotide (two patients) was added. Only two of nine patients responded to the combination of cabergoline and ketoconazole. Data on these associations are shown in Table 7.

Table 7
www.frontiersin.orgTABLE 7 Effects of associating cabergoline with ketoconazole in Cushing’s disease patients.

Considering that one of the indications for the treatment of hypercortisolism may be complementary to radiotherapy, we analyzed the eight patients who underwent radiotherapy after transsphenoidal surgery. In these patients, doses of ketoconazole from 200 to 1,200 mg were used, and in six patients there was a normalization of the UFC in 1 to 60 months of treatment. Thus, the association of ketoconazole with radiotherapy was effective in normalizing the 24-h UFC in 75% of cases.

Clinical follow-up

New therapeutic approaches were attempted in some patients during follow-up: radiotherapy (eight patients), new TSS (five patients), and bilateral adrenalectomy (four patients). At the end of this analysis, 11 patients remained on ketoconazole, all with controlled hypercortisolism. Among the 11 patients who were not fully controlled by the last visit, five were using ketoconazole as pre-TSS therapy and underwent TSS as soon as possible, while three others underwent radiotherapy and two underwent bilateral adrenalectomy. One patient was lost to follow-up.

Discussion

According to the current consensus about CD, drug treatment should be reserved for patients without remission after TSS, those who cannot undergo surgical treatment, or those awaiting the effects of radiotherapy (416). Drugs available in this context may act as adrenal steroidogenesis blockers (ketoconazole, osilodrostat, metyrapone, mitotane, levoketoconazole, and etomidate), in pituitary adenoma (somatostatinergic receptor ligands—pasireotide), dopamine receptor agonists (cabergoline), or glucocorticoid receptor blockers (mifepristone) (1626). Among these alternatives, the drug of choice still cannot be determined. Thus, the best option must be established individually, considering aspects such as remission potential, safety profile, availability, cost, etc. (162728).

For over 30 years, ketoconazole has been prescribed off-label for CD patients with varied rates of remission of hypercortisolism, and it can be used in monotherapy or associated with other drugs (2930). The Brazilian public health system does not provide drugs for the treatment of CD, and among medications with a better profile for controlling hypercortisolism, such as osilodrostat, levoketoconazole, and pasireotide, only pasireotide has been approved by the national regulatory authority (ANVISA). Due to such pragmatic considerations, ketoconazole is among the most commonly used drugs in our health system, whether recently associated or not with cabergoline (7).

In this cohort, the most prevalent response type was complete (66%). Since 75% of the CD patients who used ketoconazole had a complete or partial response, there was a clear trend towards improvement in hypercortisolism. When only those who used ketoconazole post-TSS were evaluated, the rate of control increased to 76%. We found that patients with a higher initial 24-h UFC tended to have less control of excess cortisol, a difference that was not observed when analyzing ketoconazole dose or follow-up time.

In our series and at the prescribed doses, the combination of cabergoline and ketoconazole was not effective in the management of hypercortisolism since only two of nine patients (22%) had their 24-hour UFC normalized. However, it should be observed that this association was used in patients who had more severe CD and, consequently, were less likely to have a favorable response. The effects of cabergoline in CD patients remain controversial, although some studies have shown promising responses (3132).

Previous reviews found that the efficacy of ketoconazole for hypercortisolism control was quite heterogeneous, ranging from 14 to 100% in 99 patients (3334). Our cohort’s response rate was lower than that of Sonino et al. (89%) (20) but higher than that of a multicenter cohort by Castinetti et al. (approximately 50%) (14). Regarding other smaller series (3537) our results reinforce some findings that demonstrate a percentage of control greater than 50% of the cases.

Our analyses showed a trend toward a response that continued, with some oscillations, over time. The rate of uncontrolled patients remained stable over time (approximately 30%), regardless of association with other drugs (cabergoline, mitotane, or pasireotide) or dose adjustments. Speculatively, it would appear that patients who respond to ketoconazole treatment would show some type of response as soon as therapy begins.

Our cohort has the longest follow-up time of any study on ketoconazole use in CD, nearly 15 years. Our results demonstrate that patients who benefit from ketoconazole (i.e., control of hypercortisolism and associated comorbidities) can safely use it for a long term since those who did not experience liver enzyme changes at the beginning of treatment also had no long-term changes.

Another relevant information for clinical practice is the result of treatment with ketoconazole associated with radiotherapy, which demonstrated normalizing the 24-h UFC in 75% of cases, a finding that reinforces the use of this therapeutic combination, especially in cases that are more resistant to different treatment modalities.

As described in the literature, adverse effects, such as nausea, vomiting, dizziness, headache, loss of appetite, and elevated transaminases, are relatively frequent (38). In our cohort, 10 patients (30%) had mild adverse effects, and four (12%) had more serious adverse effects requiring discontinuation. In other studies, up to 20% of patients required discontinuation due to side effects (14). We documented 20 episodes of hypokalemia during ketoconazole treatment, some with worsening blood pressure control. In most cases, hypokalemia has occurred in association with the use of diuretic drugs, which may have potentiated potassium spoliation, reinforcing the need of stringent surveillance in hypertensive Cushing’s disease patients using this combination. It can also result from the enzymatic blockade that could lead to the elevation of adrenal mineralocorticoid precursors (pex. deoxycorticosterone), with consequent sodium retention and worsening hypertension. Although it has not been analyzed in other series with ketoconazole, this side effect has been observed in patients who received other adrenal-blocking drugs, such as osilodrostat and metyrapone (16). This alteration seems to be transient in some patients; in our series, it was managed by suspending drugs that could worsen hypokalemia and introducing spironolactone and/or potassium supplementation. Hypokalemia may also result from continuing intense adrenal stimulation by ACTH and changes in the activity of the 11-beta-hydroxysteroid dehydrogenase enzyme, which increase the mineralocorticoid activity of cortisol, as observed in patients with severe hypercortisolism in uncontrolled CD (39). Hypogonadism occurred in one male patient. In two adolescent patients (one female and one male), hypercortisolism was effectively controlled without altering the progression of puberty. As described in other cohorts, this effect was expected due to the high doses, which block adrenal and testicular androgen production (20).

Thus, our findings confirm previous reports in the literature and add important information about the side effects and safety of long-term ketoconazole use in CD treatment. Our data reinforce the current recommendations about ketoconazole for recurrent cases or those refractory to surgery, including proper follow-up by an experienced team specializing in evaluating clinical and biochemical responses and potential adverse effects (71840). Despite the severity of many of our CD patients, no ketoconazole-related death occurred during follow-up, including long-term observation. On the other hand, no patient progressed to definitive remission of hypercortisolism, even after many years of treatment with ketoconazole.

Conclusions

In our cohort of patients, ketoconazole proved to be an effective and safe alternative for CD treatment, although it can produce side effects that require proper identification and management, allowing effective long-term treatment. We found side effects that have been rarely described in the literature, including hypokalemia and worsening hypertension, which require specific care and management. Thus, ketoconazole is an effective alternative for CD patients who cannot undergo surgery, who do not achieve remission after pituitary surgery, or who have recurrent hypercortisolism.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors without undue reservation.

Ethics statement

The studies involving human participants were reviewed and approved by the Hospital de Clínicas de Porto Alegre Research Ethics Committee. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

Author contributions

CV and MAC created the research format. CV, RBM, and MCBC realized the search on medical records. CV performed the statistical analysis. MAC, ACVM, and TCR participated in the final data review and discussion. ACVM participated in the final data review and discussion as volunteer collaborator. All authors contributed to the article and approved the submitted version.

Funding

This work was supported by the “Coordenação de Aperfeiçoamento de Pessoal de Nı́vel Superior” (CAPES), Ministry of Health – Brazil, through a PhD scholarship; and the Research Incentive Fund (FIPE) of Hospital de Clı́nicas de Porto Alegre.

Acknowledgments

The authors would like to thank the HCPA Research and Graduate Studies Group (GPPG) for the statistical technical support provided by Rogério Borges. We also thank the Research Incentive Fund of Hospital de Clínicas de Porto Alegre and Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), by funds applied. We also thank the Graduate Program in Endocrinology and Metabolism (PPGEndo UFRGS) for all the support in the preparation of this research.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

1. Fleseriu M, Castinetti F. Updates on the role of adrenal steroidogenesis inhibitors in cushing’s syndrome: a focus on novel therapies. Pituitary (2016) 19(6):643–53. doi: 10.1007/s11102-016-0742-1

PubMed Abstract | CrossRef Full Text | Google Scholar

2. Pivonello R, De Leo M, Cozzolino A, Colao A. The treatment of cushing’s disease. Endocr Rev (2015) 36(4):385–486. doi: 10.1210/er.2013-1048

PubMed Abstract | CrossRef Full Text | Google Scholar

3. Nieman LK, Biller BMK, Findling JW, Newell-Price J, Savage MO, Stewart PM, et al. The diagnosis of cushing’s syndrome: An endocrine society clinical practice guideline. J Clin Endocrinol Metab (2008) 93(5):1526–40. doi: 10.1210/jc.2008-0125

PubMed Abstract | CrossRef Full Text | Google Scholar

4. Nieman LK, Biller BMK, Findling JW, Murad MH, Newell-Price J, Savage MO, et al. Treatment of cushing’s syndrome: An endocrine society clinical practice guideline. J Clin Endocrinol Metab (2015) 100(8):2807–31. doi: 10.1210/jc.2015-1818

PubMed Abstract | CrossRef Full Text | Google Scholar

5. Pivonello R, De Martino M, De Leo M, Lombardi G, Colao A. Cushing’s syndrome. Endocrinol Metab Clin North (2008) 37(1):135–49. doi: 10.1016/j.ecl.2007.10.010

CrossRef Full Text | Google Scholar

6. Alexandraki KI, Grossman AB. Therapeutic strategies for the treatment of severe cushing’s syndrome. Drugs (2016) 76(4):447–8. doi: 10.1007/s40265-016-0539-6

PubMed Abstract | CrossRef Full Text | Google Scholar

7. Machado MC, Fragoso MCBV, Moreira AC, Boguszewski CL, Neto LV, Naves LA, et al. A review of cushing’s disease treatment by the department of neuroendocrinology of the Brazilian society of endocrinology and metabolism. Arch Endocrinol Metab (2018) 62(1):87–105. doi: 10.20945/2359-3997000000014

PubMed Abstract | CrossRef Full Text | Google Scholar

8. Rollin G, Ferreira NP, Czepielewski MA. Prospective evaluation of transsphenoidal pituitary surgery in 108 patients with Cushing’s disease. Arq Bras Endocrinol Metabol. (2007) 51(8):1355–61. doi: 10.1590/s0004-27302007000800022

PubMed Abstract | CrossRef Full Text | Google Scholar

9. Patil CG, Prevedello DM, Lad SP, Lee Vance M, Thorner MO, Katznelson L, et al. Late recurrences of cushing’s disease after initial successful transsphenoidal surgery. J Clin Endocrinol Metab (2008) 93(2):358–62. doi: 10.1210/jc.2007-2013

PubMed Abstract | CrossRef Full Text | Google Scholar

10. Rubinstein G, Osswald A, Zopp S, Ritzel K, Theodoropoulou M, Beuschlein F, et al. Therapeutic options after surgical failure in cushing’s disease: A critical review. Best Pract Res Clin Endocrinol Metab (2019) 33(2):101270. doi: 10.1016/j.beem.2019.04.004

PubMed Abstract | CrossRef Full Text | Google Scholar

11. Zhao N, Yang X, Li C, Yin X. Efficacy and safety of pasireotide for Cushing’s disease: A protocol for systematic review and meta-analysis. Medicine (Baltimore). (2020) 99(51):e23824. doi: 10.1097/MD.0000000000023824

PubMed Abstract | CrossRef Full Text | Google Scholar

12. Pivonello R, Fleseriu M, Newell-Price J, Bertagna X, Findling J, Shimatsu A, et al. Efficacy and safety of osilodrostat in patients with cushing’s disease (LINC 3): a multicentre phase III study with a double-blind, randomised withdrawal phase. Lancet Diabetes Endocrinol (2020) 8(9):748–61. doi: 10.1016/S2213-8587(20)30240-0

PubMed Abstract | CrossRef Full Text | Google Scholar

13. Yan JY, Nie XL, Tao QM, Zhan SY, De Zhang Y. Ketoconazole associated hepatotoxicity: A systematic review and meta-analysis. Biomed Environ Sci (2013) 26(7):605–10. doi: 10.3967/0895-3988.2013.07.013

PubMed Abstract | CrossRef Full Text | Google Scholar

14. Castinetti F, Guignat L, Giraud P, Muller M, Kamenicky P, Drui D, et al. Ketoconazole in cushing’s disease: Is it worth a try. J Clin Endocrinol Metab (2014) 99(5):1623–30. doi: 10.1210/jc.2013-3628

PubMed Abstract | CrossRef Full Text | Google Scholar

15. Castinetti F, Nieman LK, Reincke M, Newell-Price J. Approach to the patient treated with steroidogenesis inhibitors. J Clin Endocrinol Metab (2021) 106(7):2114–23. doi: 10.1210/clinem/dgab122

PubMed Abstract | CrossRef Full Text | Google Scholar

16. Fleseriu M, Auchus R, Bancos I, Bem-Shlomo A, Bertherat J, Biermasz NR, et al. Consensus on diagnosis and management of cushing’s disease: a guideline update. Lancet Diabetes Endocrinol (2021) 9(12):847–75. doi: 10.1016/s2213-8587(21)00235-7

PubMed Abstract | CrossRef Full Text | Google Scholar

17. Fleseriu M, Pivonello R, Elenkova A, Salvatori R, Auchus RJ, Feelders RA, et al. Efficacy and safety of levoketoconazole in the treatment of endogenous cushing’s syndrome (SONICS): a phase 3, multicentre, open-label, single-arm trial. Lancet Diabetes Endocrinol (2019) 7(11):855–65. doi: 10.1016/S2213-8587(19)30313-4

PubMed Abstract | CrossRef Full Text | Google Scholar

18. Tritos NA. Adrenally directed medical therapies for cushing syndrome. J Clin Endocrinol Metab (2021) 106(1):16–25. doi: 10.1210/clinem/dgaa778

PubMed Abstract | CrossRef Full Text | Google Scholar

19. Simões Corrêa Galendi J, Correa Neto ANS, Demetres M, Boguszewski CL, dos S V. N. nogueira, “Effectiveness of medical treatment of cushing’s disease: A systematic review and meta-analysis,”. Front Endocrinol (Lausanne) (2021) 12:732240(September). doi: 10.3389/fendo.2021.732240

PubMed Abstract | CrossRef Full Text | Google Scholar

20. Sonino N, Boscaro M, Paoletta A, Mantero F, Zillotto D. Ketoconazole treatment in cushing’s syndrome: experience in 34 patients. Clin Endocrinol (Oxf) (1991) 35(4):347–52. doi: 10.1111/j.1365-2265.1991.tb03547.x

PubMed Abstract | CrossRef Full Text | Google Scholar

21. Costenaro F, Rodrigues TC, Rollin GAF, Czepielewski MA. Avaliação do eixo hipotálamohipófise adrenal no diagnóstico e na remissão da doença de cushing. Arquivos Brasileiros Endocrinologia e Metabologia (2012). doi: 10.1590/S0004-27302012000300002

CrossRef Full Text | Google Scholar

22. Amlashi FG, Swearinger B, Faje AT, Nachtigall LB, Miller KK, Klibanski A, et al. Accuracy of late-night salivary cortisol in evaluating postoperative remission and recurrence in cushing’s disease. J Clin Endocrinol Metab (2015) 100(10):3770–7. doi: 10.1210/jc.2015-2107

PubMed Abstract | CrossRef Full Text | Google Scholar

23. Silveiro SP, Satler F. Rotinas em endocrinologia. (Porto Alegre: Artmed) (2015).

Google Scholar

24. Costenaro F, Rodrigues TC, Ferreira NP, da Costa TG, Schuch T, Boschi V, et al. Pseudotumor cerebri during cushing’s disease treatment with ketoconazole. Arq. Bras Endocrinol Metabol (2011). doi: 10.1590/s0004-27302011000400008

CrossRef Full Text | Google Scholar

25. Costenaro F, Rodrigues TC, De Lima PB, Ruszczyk J, Rollin G, Czepielewski MA. A successful case of cushing’s disease pregnancy treated with ketoconazole. Gynecol Endocrinol (2015) 31(3):176–8. doi: 10.3109/09513590.2014.995615

PubMed Abstract | CrossRef Full Text | Google Scholar

26. Gadelha MR, Neto LV. Efficacy of medical treatment in cushing’s disease: A systematic review. Clin Endocrinol (Oxf) (2014) 80(1):1–12. doi: 10.1111/cen.12345

PubMed Abstract | CrossRef Full Text | Google Scholar

27. Fleseriu M, Petersenn S. New avenues in the medical treatment of cushing’s disease: Corticotroph tumor targeted therapy. J Neurooncol (2013) 114(1):1–11. doi: 10.1007/s11060-013-1151-1

PubMed Abstract | CrossRef Full Text | Google Scholar

28. Fleseriu M, Petersenn S. Medical management of cushing’s disease: What is the future? Pituitary (2012) 15(3):330–41. doi: 10.1007/s11102-012-0397-5

PubMed Abstract | CrossRef Full Text | Google Scholar

29. Feelders RA, De Bruin C, Pereira AM, Romijn JÁ, Netea-Maier RT, Hermus AR, et al. Pasireotide alone or with cabergoline and ketoconazole in cushing’s disease. N Engl J Med (2010) 362(19):1846–8. doi: 10.1056/NEJMc1000094

PubMed Abstract | CrossRef Full Text | Google Scholar

30. Barbot M, Albiger N, Ceccato F, Zilio M, Frigo AC, Denaro Lc, et al. Combination therapy for cushing’s disease: Effectiveness of two schedules of treatment. should we start with cabergoline or ketoconazole? Pituitary (2014) 17(2):109–17. doi: 10.1007/s11102-013-0475-3

PubMed Abstract | CrossRef Full Text | Google Scholar

31. Vilar L, Naves LA, Azevedo MF, Arruda MJ, Arahata CM, Silva LM, et al. Effectiveness of cabergoline in monotherapy and combined with ketoconazole in the management of cushing’s disease. Pituitary (2010) 13(2):123–9. doi: 10.1007/s11102-009-0209-8

PubMed Abstract | CrossRef Full Text | Google Scholar

32. Pivonello R, De Martino MC, Cappabianca P, De Leo M, Faggiano A, Lombardi G, et al. The medical treatment of cushing’s disease: Effectiveness of chronic treatment with the dopamine agonist cabergoline in patients unsuccessfully treated by surgery. J Clin Endocrinol Metab (2009) 94(1):223–30. doi: 10.1210/jc.2008-1533

PubMed Abstract | CrossRef Full Text | Google Scholar

33. Castinetti F, Morange I, Jaquet P, Conte-Devolx B, Brue T. Ketoconazole revisited: A preoperative or postoperative treatment in cushing’s disease. Eur J Endocrinol (2008). doi: 10.1530/EJE-07-0514

PubMed Abstract | CrossRef Full Text | Google Scholar

34. Loli P, Berselli ME, Tagliaferri M. Use of ketoconazole in the treatment of cushing’s syndrome. J Clin Endocrinol Metab (1986) 63(6):1365–71. doi: 10.1210/jcem-63-6-1365

PubMed Abstract | CrossRef Full Text | Google Scholar

35. Kakade HR, Kasaliwal R, Khadilkar KS, Jadhav S, Bukan A, Khare Sc, et al. Clinical, biochemical and imaging characteristics of cushing’s macroadenomas and their long-term treatment outcome. Clin Endocrinol (Oxf) (2014) 81(3):336–42. doi: 10.1111/cen.12442

PubMed Abstract | CrossRef Full Text | Google Scholar

36. Luisetto G, Zangari M, Camozzi V, Boscaro M, Sonino N, Fallo F. Recovery of bone mineral density after surgical cure, but not by ketoconazole treatment, in cushing’s syndrome. Osteoporos Int (2001) 12(11):956–60. doi: 10.1007/s001980170025

PubMed Abstract | CrossRef Full Text | Google Scholar

37. Huguet I, Aguirre M, Vicente A, Alramadan M, Quiroga I, Silva J, et al. Assessment of the outcomes of the treatment of cushing’s disease in the hospitals of castilla-la mancha. Endocrinol y Nutr (2015) 62(5):217–23. doi: 10.1016/j.endonu.2015.02.007

CrossRef Full Text | Google Scholar

38. Tritos NA, Biller BMK. Advances in the medical treatment of cushing disease. Endocrinol Metab Clin North Am (2020) 49(3):401–12. doi: 10.1016/j.ecl.2020.05.003

PubMed Abstract | CrossRef Full Text | Google Scholar

39. Torpy D, Mullen N, Ilias I, Nieman L. Association of hypertension and hypokalemia with cushing’s syndrome caused by ectopic ACTH secretion. Ann N Y Acad Sci (2002) 970:134–44. doi: 10.1111/j.1749-6632.2002.tb04419.x

PubMed Abstract | CrossRef Full Text | Google Scholar

40. Varlamov EV, Han AJ, Fleseriu M. “Updates in adrenal steroidogenesis inhibitors for cushing’s syndrome – a practical guide,”. Best Pract Res Clin Endocrinol Metab (2021) 35(1):101490. doi: 10.1016/j.beem.2021.101490

PubMed Abstract | CrossRef Full Text | Google Scholar

Keywords: Cushing’s disease, Cushing’s syndrome, hypercortisolism, treatment, ketoconazole

Citation: Viecceli C, Mattos ACV, Costa MCB, Melo RBd, Rodrigues TdC and Czepielewski MA (2022) Evaluation of ketoconazole as a treatment for Cushing’s disease in a retrospective cohort. Front. Endocrinol. 13:1017331. doi: 10.3389/fendo.2022.1017331

Received: 11 August 2022; Accepted: 06 September 2022;
Published: 07 October 2022.

Edited by:

Luiz Augusto Casulari, University of Brasilia, Brazil

Reviewed by:

Juliana Drummond, Federal University of Minas Gerais, Brazil
Monalisa Azevedo, University of Brasilia, Brazil

Copyright © 2022 Viecceli, Mattos, Costa, Melo, Rodrigues and Czepielewski. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Mauro Antonio Czepielewski, maurocze@terra.com.br

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

From https://www.frontiersin.org/articles/10.3389/fendo.2022.1017331/full

TP53 Mutations in Functional Corticotroph Tumors Are Linked to Invasion and Worse Clinical Outcome

Abstract

Corticotroph macroadenomas are rare but difficult to manage intracranial neoplasms. Mutations in the two Cushing’s disease mutational hotspots USP8 and USP48 are less frequent in corticotroph macroadenomas and invasive tumors. There is evidence that TP53 mutations are not as rare as previously thought in these tumors. The aim of this study was to determine the prevalence of TP53 mutations in corticotroph tumors, with emphasis on macroadenomas, and their possible association with clinical and tumor characteristics. To this end, the entire TP53 coding region was sequenced in 86 functional corticotroph tumors (61 USP8 wild type; 66 macroadenomas) and the clinical characteristics of patients with TP53 mutant tumors were compared with TP53/USP8 wild type and USP8 mutant tumors. We found pathogenic TP53 variants in 9 corticotroph tumors (all macroadenomas and USP8 wild type). TP53 mutant tumors represented 14% of all functional corticotroph macroadenomas and 24% of all invasive tumors, were significantly larger and invasive, and had higher Ki67 indices and Knosp grades compared to wild type tumors. Patients with TP53 mutant tumors had undergone more therapeutic interventions, including radiation and bilateral adrenalectomy. In conclusion, pathogenic TP53 variants are more frequent than expected, representing a relevant amount of functional corticotroph macroadenomas and invasive tumors. TP53 mutations associated with more aggressive tumor features and difficult to manage disease.

Introduction

Pituitary neuroendocrine tumors are the second most common intracranial neoplasm [1]. They are usually benign, but when aggressive they may be particularly difficult to manage, accompanied by high comorbidity and increased mortality [2]. Corticotroph tumors constitute 6–10% of all pituitary tumors, but they represent up to 45% of aggressive pituitary tumors and pituitary carcinomas [2]. Functional corticotroph tumors cause Cushing’s disease (CD), a debilitating condition accompanied by increased morbidity and mortality due to glucocorticoid excess [3]. Pituitary surgery is the first line treatment, but recurrence is observed in 15–20% of cases of whom most are macroadenomas (with a size of ≥ 10 mm) [4]. Treatment options include repeated pituitary surgery, radiation therapy, medical treatment and bilateral adrenalectomy (BADX) [3]. With respect to the latter, corticotroph tumor progression after bilateral adrenalectomy/Nelson’s syndrome (CTP-BADX/NS) is a frequent severe complication and may present with aggressive tumor behavior [5,6,7].

Corticotroph tumors (including CTP-BADX/NS) carry recurrent somatic mutations in the USP8 gene in ~ 40–60% of cases [8,9,10,11,12,13]. These USP8 mutant tumors are usually found in female patients and are generally less invasive [8,9,10,11]. Additional genetic studies identified a second mutational hotspot in the USP48 gene, but no other driver mutations [14,15,16,17,18]. Focusing on USP8 wild type corticotroph tumors, we recently discovered TP53 mutations in 6 out of 18 cases (33%) [17]. Subsequent reports documented TP53 mutations in small series of mainly aggressive corticotroph tumors and carcinomas [1920].

TP53 is the most commonly mutated gene in malignant neoplasms [2122], including brain and neuroendocrine tumors [2324]. Until our previous report [17], TP53 mutations were only described in isolated cases of aggressive pituitary tumors and carcinomas, and were therefore considered very rare events [81625,26,27,28]. A link between TP53 mutations and an aggressive corticotroph tumor phenotype has been hypothesized, but the heterogeneity and small size of the studies reported did not support significant clinical associations [1719].

To address this, we determined the prevalence of TP53 variants in a cohort of 86 patients with functional corticotroph tumors, including 61 with USP8 wild type tumors, and studied the associations between TP53 mutational status and clinical features.

Methods

Patients and samples

We analyzed tumor samples of 86 adult patients: 61 USP8 wild type and 25 USP8 mutant. Sixty-six patients (46 females, 20 males) were diagnosed with CD between 1994 and 2020 in Germany (Hamburg, Munich, Erlangen, and Tübingen) and Luxembourg. Twenty additional patients (16 females, 4 males) were diagnosed with CTP-BADX/NS, operated and followed up in 7 different international centers (Nijmegen, Munich, Erlangen, Hamburg, Paris, Rio de Janeiro, and Würzburg). Twenty-three out of 86 samples were collected prospectively between 2018 and 2021, and 63 were retrospective cases (of which 42 were investigated in the context of USP8 and USP48 screenings and published elsewhere) [9121317]. Seventy-one tumors were fresh frozen and 15 were formalin fixed paraffin embedded. Paired blood was available for 12 cases. The median follow-up time after initial diagnosis was 44 months (range 2–384 months).

Endogenous Cushing’s syndrome was diagnosed according to typical clinical signs and symptoms and established biochemical procedures suggesting glucocorticoid excess. Clinical features included central obesity, moon face, buffalo hump, muscle weakness, easy bruising, striae, acne, low-impact bone fractures, mood changes, irregular menstruation, infertility and impotency. Biochemical diagnosis was based on increased 24 h urinary free cortisol (UFC) and late-night salivary cortisol levels, and lack of serum cortisol suppression after low-dose dexamethasone test. A pituitary ACTH source was confirmed by > 2.2 pmol/l (10 pg/ml) basal plasma ACTH, > 50% suppression of serum cortisol during an 8 mg dexamethasone test, and ACTH and cortisol response to corticotrophin releasing hormone stimulation.

The clinical and pathological features of our study cohort are summarized in Additional file 1: Supplementary Table 1. All patients underwent pituitary surgery. The presence of an ACTH-producing pituitary tumor was confirmed histologically after surgical resection. Biochemical remission after surgery was defined as postoperative 24 h-UFC levels below or within the normal range, or serum cortisol levels < 5 µg/dl after low-dose (1 or 2 mg) dexamethasone suppression test. Tumor control was achieved when there was no evidence of regrowth or disease recurrence. Tumor invasion was defined as radiological or intraoperative evidence of tumor within the sphenoid and/or cavernous sinuses [29]. CTP-BADX/NS was defined as an expanding pituitary tumor after bilateral adrenalectomy (BADX) following expert consensus recommendations [5].

DNA extraction, TP53 amplification and sequencing

Genomic DNA was extracted using the Maxwell Tissue DNA Kit (Promega), Maxwell Blood DNA kit (Promega) or the FFPE DNA mini kit (Qiagen), depending on the type of sample, as described previously [912]. The entire coding sequence of TP53 (including exons 9β and 9γ) as well as noncoding regions adjacent to each exon were amplified using the GoTaq DNA polymerase (Promega) and specific primers (Additional file 1: Supplementary Table 2). Amplification of USP8 hotspot region and Sanger sequencing were performed as described previously [912]. Chromatograms were analyzed using the Mutation Surveyor v4.0.9 (Soft Genetics). Samples were examined for TP53 coding and splicing variants. Variant position and pathogenicity was investigated in ENSEMBL (www.ensembl.org), the UCSC Genome Browser (http://genome-euro.ucsc.edu), the IARC TP53 database (https://p53.iarc.fr/TP53GeneVariations.aspx), the Catalogue Of Somatic Mutations in Cancer (COSMIC; https://cancer.sanger.ac.uk/cosmic), ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/), PHANTM (http://mutantp53.broadinstitute.org/), the Human Splicing Finder (HSF; http://www.umd.be/HSF3/) and VarSEAK splicing predictor (https://varseak.bio/). Variant frequencies on the general population were obtained from the Allele Frequency Aggregator (ALFA) project [30], the Genome Aggregation Database (gnomAD) [31] and the International Genome Sample Resource 1000Genome project [32]. Throughout the text, variants refer to NC_000017.11 (genomic DNA), ENST00000269305.9 (coding DNA) and ENSP00000269305.4 (protein), following the Human Genome Variation Society (HGVS) standard nomenclature system.

Statistical analysis

Statistical analysis was performed with the software package SPSS v24 (IBM). We used t-test or one-way ANOVA to analyze the association of TP53 variants with age, body mass index; Mann–Whitney U and Kruskal–Wallis to test non-parametric variables, such as tumor size, hormone levels, Ki67 index and p53 score. We corrected the analysis for multiple comparisons with the Bonferroni test. Categorical variables were analyzed using a chi-square test or Fisher exact test when needed. Survival analysis was performed using Kaplan–Meier curves with log-rank tests, and multivariate Cox regression. An exact, two-tailed significance level of P < 0.05 was considered to be statistically significant.

Results

Analysis of TP53 nucleotide variants

We analyzed all TP53 coding exons (including exons 9β and 9γ) and adjacent intronic noncoding sequences in 61 USP8 wild type tumors (49 CD and 12 CTP-BADX/NS). Of these, 13 were microadenomas (< 10 mm) and 48 macroadenomas (≥ 10 mm) at the time of the current operation. A separate group of 25 USP8 mutant tumors (17 CD and 8 CTP-BADX/NS) that were mainly macroadenomas (n = 19) was used for multiple comparison.

We found 59 variants in our cohort: 30 exclusively in USP8 wild type, 21 in USP8 mutant, and 8 in wild type and mutant tumors regardless of USP8 mutational status. No indels in the coding region of TP53 were detected. In addition, we did not find any genetic variant affecting TP53 splicing.

Nine out of 30 variants found in USP8 wild type tumors were either reported in the COSMIC database as pathogenic or absent from the common variant databases (1000Genomes, gnomAD, ALPHA) or had allele frequency < 0.0001. They were all described in cancer series: 5 as pathogenic or likely pathogenic in ClinVar, 2 as variants of uncertain significance (VUS) and 2 were not described in ClinVar (Table 1). All variants are reported to alter protein function and show clear loss of transactivation activity in a yeast based assay (Table 1) [33].

Table 1 Functionally relevant TP53 variants found in 9/86 corticotroph tumors

Seven variants target amino acids within the DNA-binding domain, essential for p53 activity, disrupting S2’ and S7 β-sheets or the L3 loop spatial conformation. The other two [c.1009C > G (p.Arg337Gly) and c.1031 T > C (p.Leu344Pro)] locate in the tetramerization domain and keep p53 protein as monomer impairing its transactivation activity [34]. From the 9 variants, 8 affect highly conserved p53 residues, while in c.1031 T > C (p.Met133Lys) the methionine alternates with leucine or valine among species. This variant alters protein folding, probably reducing DNA affinity [35], while the substitution of a methionine that acts as an alternative start codon abolishes the transcription of isoforms ∆133p53α, ∆133p53β and ∆133p53γ. The 9 variants were detected in nine cases (henceforth referred to as TP53 mutant; Table 1). Two tumors from unrelated patients (#6 and #7) carried the same variant c.818G > A (p.Arg273His), while one tumor (#4) carried two variants (c.718A > G and c.773A > C). Seven variants were found in heterozygosis, while the other two (from patients #1 and #2) in homozygosis. From these two, we only had paired blood/tumor samples from patient #1 and detected the variant only on the tumor sample, indicative of loss of heterozygosity (Additional file 1: Supplementary Fig. 1A). Similarly, we could demonstrate the somatic origin of the TP53 variants in four other patients with paired tumor/blood samples (#3, #5, #6 and #9).

The remaining 21/30 variants found in USP8 wild type and all 21 variants found in the USP8 mutant tumors were described as benign, likely benign or VUS with no evidence of affecting protein function. All tumors with these variants were considered TP53 wild type. From the 21 variants found in the USP8 wild type tumors (henceforth referred to as TP53/USP8 wild type group), 7 were non-synonymous variants, 8 synonymous variants and 6 non-coding variants without splicing effect. From the 21 variants found in the 25 USP8 mutant tumors, nine were synonymous, four non-synonymous and eight non-coding without splicing effect. In addition, eight variants were found in tumors regardless of USP8 mutational status that were not categorized as TP53 mutations. The intronic variant c.782 + 62G > A was found in heterozygosis in 6/70 samples. It was not reported in any database and is not predicted to have any splicing effect. The remaining seven are common variants classified as benign or likely benign in ClinVar and their allele frequencies were similar to those reported for the general population (ALFA, gnomAD and 1000Genome project) (Additional file 1: Supplementary Table 3).

Summarizing, all TP53 mutations were found in the USP8 wild type tumors, leading to a prevalence of 15% in this subgroup.

Clinical presentation of patients with TP53 mutant tumors

Patients with TP53 mutant tumors (n = 9) tended to be diagnosed at older age compared to TP53/USP8 wild type tumors (n = 52) (t-test P = 0.069; Table 2). This was significant after including the USP8 mutant group (n = 25) in the multiple comparison analysis (ANOVA P = 0.024, Table 2) and when TP53/USP8 wild type and USP8 mutant tumors were combined to a single group (TP53 wild type, n = 77; Additional file 1: Supplementary Table 4. We did not observe any sex specific predominance of TP53 mutations in contrast to USP8 mutants that are predominantly found in female patients. Furthermore, we did not find any statistically significant differences in ACTH and cortisol levels (Table2; Additional file 1: Supplementary Table 4).

Table 2 Clinical features of TP53 mutant versus TP53/USP8 wild type and USP8 mutant groups

Patients with TP53 mutant tumors underwent more surgeries and tumor resection was more frequently incomplete compared to TP53/USP8 wild type (Table 2). These patients also underwent a higher number of additional therapeutic procedures (radiation, n = 7; BADX, n = 4; temozolomide, n = 3; pasireotide, n = 2). Only one patient (#4) with TP53 mutant tumor, a 77 year-old man, had a single surgery without any other treatment, but his follow-up was short (< 6 months).

We observed TP53 mutations more frequently in CTP-BADX/NS (4/12, 33%) compared to CD (5/49, 10%), trending towards statistically significant difference (Fischer exact test P = 0.065 for TP53 mutant vs. TP53/USP8 wild type, P = 0.060 for comparison among the 3 groups; Table 2).

The TP53 mutant group associated with higher disease-specific mortality and shorter survival than USP8 mutant or TP53/USP8 wild type groups (log rank test, P = 0.023, Fig. 1). Three patients with TP53 mutant tumors (all CTP-BADX/NS) died of disease-related deaths: two from severe cerebral hemorrhage after surgery and stereotactic radiation and one from uncontrolled disease after five failed operations, radiotherapy (gamma knife, fractionated radiation) and chemotherapy (temozolomide, bevacizumab) at the ages of 75, 80 and 37, respectively. Ten-year survival was 27% for patients with TP53 mutant tumors, 100% for TP53/USP8 wild type and 86% for USP8 mutant. In our cohort, survival did not differ after adjusting for age (HR 7.7, 95%CI 0.6–107.7, P = 0.127).

Fig. 1

figure 1

Kaplan–Meier curve showing overall survival in patients with TP53 mutant/USP8 wild type, USP8 mutant/TP53 wild type, and TP53 wild type/USP8 wild type corticotroph tumors. The table underneath the graph shows the 10-year cumulative survival after diagnosis

Tumor samples from prior surgeries were available from one TP53 mutant case (#8, Table 1). This male patient had his first pituitary surgery for CD when he was 30 years old and was treated with γ-knife one year later. He then underwent two more pituitary surgeries and BADX until the age of 35. He developed CTP-BADX/NS with para- and retrosellar tumor extension along with panhypopituitarism and underwent two more pituitary surgeries before dying at the age of 38 due to complications of the disease. We detected the TP53 variant c.1009C > G (p.Arg337Gly) in all available tumor specimens, including his first and latest surgeries (Additional file 1: Supplementary Fig. 1B).

No statistical association was found between clinical data and any of the 8 common variants.

Characteristics of TP53 mutant corticotroph tumors

All TP53 mutations were found in macroadenomas (9/66; Table 3). TP53 mutant tumors were larger that TP53/USP8 wild type (mm median [IQR] 20.0 [14.0] vs. 15.0 [14.3]), but this did not reach statistical significance (Table 3). Multiple comparison analysis showed that the difference in tumor size is significant only comparing TP53 mutant with USP8 mutant (median [IQR] 23.3 [14.0] vs. 14 [7.3] mm; Kruskal–Wallis P = 0.019; Bonferroni corrected P = 0.018).

Table 3 Tumor features of TP53 mutant versus TP53/USP8 wild type and USP8 mutant groups

Parasellar invasion was reported in 34 out of 64 cases, for which this information was available, and it was more common in TP53 mutant tumors (100% vs. 53% and 55% for TP53/USP8 wild type and USP8 mutant, respectively; Fischer exact test P = 0.006). TP53 mutant tumors had higher Knosp grade (Kruskal–Wallis P = 0.011) with the majority being Knosp 4 (Table 3, Additional file 1: Supplementary Table 4).

Ki67 proliferation index was available for 36 cases (6 TP53 mutant). Five out of six TP53 mutant tumors had Ki67 ≥ 3% and the overall Ki67 was higher than in the wild type tumors (Kruskal–Wallis P = 0.01; Bonferroni corrected P = 0.008 for TP53/USP8 wild type) (Table 3). Ki67 ≥ 10% was reported in 6 tumors, from which 5 were TP53 mutant (Fischer exact test P < 0.0001; the remaining case was TP53/USP8 wild type).

We had information on p53 immunostaining from 9 cases (all macroadenomas), four of which TP53 mutant: 3 tumors (from patients #5, 6 and 9) showed high p53 immunoreactivity, while the one (from patient #3) carrying a nonsense variant leading to a truncated protein was p53 negative. The five TP53 wild type cases showed isolated nuclear staining in < 1–3% of cells.

Summarizing, TP53 mutations were significantly associated with features related to a more aggressive tumor behavior, such as incomplete tumor resection, more frequent parasellar invasion, higher Knosp grade, and higher Ki67 proliferation index (Table 3; Additional file 1: Supplementary Table 4).

Discussion

Herein, we investigated the prevalence of TP53 mutations by screening a large cohort of 61 functional corticotroph tumors with USP8 wild type status, and found variants altering protein function in 15% of cases. We did not detect TP53 mutations in a separate group of 25 USP8 mutant tumors, which is in concordance with previously published small next-generation sequencing series [81819].

Since we focused on USP8 wild type tumors, macroadenomas were overrepresented in our cohort. Consequently, it should be noted that the prevalence of TP53 mutations is expected to be lower in the general CD population. In fact, ~ 50% of corticotroph tumors carry USP8 mutations, which others and we have shown to be mutually exclusive. Corticotroph tumors with USP8 mutations are associated with female predominance, younger age at presentation, and less invasiveness (despite shorter time to relapse) [911131836]. In contrast, TP53 mutant tumors were diagnosed mostly at older age, did not show sex predominance and were larger and more invasive, with lower complete resection rate. None of the 19 microadenomas included in our study carried TP53 mutations. Still, we need to acknowledge that since no sample was microdissected we may have lost microadenoma cases with TP53 mutations. Instead, we found TP53 mutations in 9/66 macroadenomas (14%) and 8/34 (24%) invasive tumors, supporting the findings from smaller series [1719].

Tumor size at presentation or invasiveness do not reliably predict aggressiveness. Instead, the European Society of Endocrinology Clinical Practice Guidelines for the management of aggressive pituitary tumors and carcinomas proposed a definition of pituitary tumor aggressiveness based on rapid or clinically relevant tumor growth despite optimal therapeutic options, along with bone invasion [37]. A recent study in a series of 9 aggressive pituitary tumors and carcinomas carrying ATRX mutations reported a high frequency of missense TP53 variants (5/9, 55.6%), further suggesting a link between TP53 mutational status and unfavorable outcome [20]. We do not have exact information on changes of tumor growth for the majority of our cases, but the higher number of surgical and radiation interventions, the higher Knosp grades, and the increased mortality rate indicate that patients with TP53 mutant tumors obviously follow a more aggressive disease course.

Ki67 proliferation index together with p53 immunostaining and mitotic count have been suggested as histological markers of pituitary tumor aggressiveness [2938]. In our series, Ki67 was significantly higher in TP53 mutant tumors, reinforcing our prior observation of a higher proportion of TP53 mutant tumors in the Ki67 ≥ 3 group [17]. We had limited information on p53 immunohistochemistry, since this measure is not routinely performed in our collaborative centers. Nevertheless, in the few tumors with known p53 immunopositivity, it was higher in the TP53 mutant group, which is in concordance with a previous study reporting high p53 immunoreactivity in all TP53 mutant tumors [19].

A mutagenic action of radiation on TP53 has been hypothesized by small series on radiation-induced tumors. For instance, TP53 mutations were reported in 58% of radiation-induced sarcomas [39], while a meta-analysis reported TP53 mutations in 14/30 radiation-induced gliomas [40]. A previous study reported a case with frameshift TP53 mutation in the CTP-BADX/NS tumor, but not in the initial CD surgeries, and the mutation was therefore suspected to be induced by radiotherapy [41]. In our series, however, 4 out of 7 TP53 mutant tumors were obtained before radiation.

In their case report, Pinto et al. suggested that TP53 mutations are acquired during tumorigenesis and condition tumor evolution [41]. In contrast, Casar-Borota et al. and Uzilov et al. reported high allele fraction of TP53 mutations, indicating that they are not a late event in corticotroph tumorigenesis [1920]. In addition, Uzilov et al. reported TP53 mutations in all tumor specimens from their two TP53 mutant cases with multiple surgeries [19]. Similarly, in our series we had tissue from multiple pituitary surgeries from one patient and found the TP53 variant in all samples (CD and CTP-BADX/NS), including specimens obtained before radiotherapy. Taken together, these observations suggest that in most cases, TP53 mutations may appear early during tumor development.

A limitation of our study is the short follow-up of patients who were prospectively included. Moreover, material from repeated surgeries was lacking from most patients with TP53 mutant tumors, hampering the examination of tumor evolution in these patients. Similarly, we had limited access to blood samples, so we could not demonstrate the somatic origin for all variants. Nevertheless, the older age at initial diagnosis of CD in patients with TP53 mutant tumors (53 ± 19.5 years old, with the youngest patient diagnosed at the age of 30) and the absence of additional neoplasias during follow-up also support a somatic instead of a germline origin. Furthermore, conditions related to germline TP53 mutations, such as Li-Fraumeni syndrome, very rarely present with pituitary tumor [42]. To our knowledge, the only published case so far was a pediatric patient with an aggressive lactotroph tumor [43].

In addition to the TP53 mutations, we detected several common variants. Variants rs59758982 and rs1042522 have been associated with increased cancer susceptibility [4445]. In some cancer types, the very frequent rs1042522 c.215G > C (p.Pro72Arg) alternative variant correlated to more efficient induction of apoptosis by DNA-damaging chemotherapeutic drugs, growth suppression and higher metastatic potential [46,47,48]. In nonfunctioning pituitary tumors, alternative allele C (leading to p.Arg72) was related to early age at presentation and reduced p21 expression [49]. Very recently, an overrepresentation of the rs1042522 alternative allele C (p.Arg72) was reported in 9 out of 10 corticotroph neoplasias including 5 functional tumors (allele frequency 0.900, vs 0.714 in Latino/admixed American in gnomAD [31]) without any association with clinical features [50]. In our cohort, we did not detect different allele frequencies in any of the investigated common variants (including rs1042522) compared with public databases, nor statistical association with any clinical variable, rendering their contribution to corticotroph pathophysiology unlikely.

Conclusion

Screening a large corticotroph tumor series revealed that TP53 mutations are more frequent than previously considered. Furthermore, we show that patients with TP53 mutant tumors had higher number of surgeries, more invasive tumors, and worse disease outcome. Our study provides evidence that patients with pathogenic or function altering variants may require more intense treatment and extended follow-up, and suggests screening for TP53 variants in macroadenomas with wild type USP8 status. Further work is needed to determine the potential use of TP53 status as a predictor of disease outcome.

Availability of data and materials

The authors declare that the relevant data supporting the conclusions of this article are included within the article and its supplementary information file. Additional clinical data are available from the corresponding authors MT and LGPR upon reasonable request.

Abbreviations

CD:
Cushing’s disease
BADX:
Bilateral adrenalectomy
CTP-BADX/NS:
Corticotroph tumor progression after bilateral adrenalectomy/Nelson’s syndrome
ACTH:
Adrenocorticotropic hormone
SD:
Standard deviation
IQR:
Interquartile range
HR:
Hazard ratio

References

  1. Ostrom QT, Gittleman H, Truitt G, Boscia A, Kruchko C, Barnholtz-Sloan JS (2018) CBTRUS statistical report: primary brain and other central nervous system tumors diagnosed in the United States in 2011–2015. Neuro Oncol 20:iv1-86

    PubMed PubMed Central Article Google Scholar

  2. McCormack A, Dekkers OM, Petersenn S, Popovic V, Trouillas J, Raverot G et al (2018) Treatment of aggressive pituitary tumours and carcinomas: results of a European society of endocrinology (ESE) survey 2016. Eur J Endocrinol 178:265–276

    CAS PubMed Article Google Scholar

  3. Fleseriu M, Auchus R, Bancos I, Ben-Shlomo A, Bertherat J, Biermasz NR et al (2021) Consensus on diagnosis and management of Cushing’s disease: a guideline update. Lancet Diabetes Endocrinol 9:847–875

    PubMed Article Google Scholar

  4. Dimopoulou C, Schopohl J, Rachinger W, Buchfelder M, Honegger J, Reincke M et al (2013) Long-term remission and recurrence rates after first and second transsphenoidal surgery for Cushing’s disease: care reality in the Munich metropolitan region. Eur J Endocrinol 170:283–292

    PubMed Article CAS Google Scholar

  5. Reincke M, Albani A, Assie G, Bancos I, Brue T, Buchfelder M et al (2021) Corticotroph tumor progression after bilateral adrenalectomy (Nelson’s syndrome): systematic review and expert consensus recommendations. Eur J Endocrinol 184:P1-16

    CAS PubMed PubMed Central Article Google Scholar

  6. Fountas A, Lim ES, Drake WM, Powlson AS, Gurnell M, Martin NM et al (2020) Outcomes of patients with Nelson’s syndrome after primary treatment: a multicenter study from 13 UK pituitary centers. J Clin Endocrinol Metab 105:1527–1537

    Article Google Scholar

  7. Kemink SA, Wesseling P, Pieters GF, Verhofstad AA, Hermus AR, Smals AG (1999) Progression of a Nelson’s adenoma to pituitary carcinoma; a case report and review of the literature. J Endocrinol Invest 22:70–75

    CAS PubMed Article Google Scholar

  8. Reincke M, Sbiera S, Hayakawa A, Theodoropoulou M, Osswald A, Beuschlein F et al (2015) Mutations in the deubiquitinase gene USP8 cause Cushing’s disease. Nat Genet 47:31–38

    CAS PubMed Article Google Scholar

  9. Pérez-Rivas LG, Theodoropoulou M, Ferraù F, Nusser C, Kawaguchi K, Stratakis CA et al (2015) The Gene of the ubiquitin-specific protease 8 is frequently mutated in adenomas causing Cushing’s disease. J Clin Endocrinol Metab 100:E997-1004

    PubMed PubMed Central Article Google Scholar

  10. Ma Z-Y, Song Z-J, Chen J-H, Wang Y-F, Li S-Q, Zhou L-F et al (2015) Recurrent gain-of-function USP8 mutations in Cushing’s disease. Cell Res 25:306–317

    CAS PubMed PubMed Central Article Google Scholar

  11. Hayashi K, Inoshita N, Kawaguchi K, Ardisasmita AI, Suzuki H, Fukuhara N et al (2016) The USP8 mutational status may predict drug susceptibility in corticotroph adenomas of Cushing’s disease. Eur J Endocrinol 174:213–226

    CAS PubMed Article Google Scholar

  12. Pérez-Rivas LG, Theodoropoulou M, Puar TH, Fazel J, Stieg MR, Ferraù F et al (2018) Somatic USP8 mutations are frequent events in corticotroph tumor progression causing Nelson’s tumor. Eur J Endocrinol 178:59–65

    Article Google Scholar

  13. Albani A, Pérez-Rivas LG, Dimopoulou C, Zopp S, Colón-Bolea P, Roeber S et al (2018) The USP8 mutational status may predict long-term remission in patients with Cushing’s disease. Clin Endocrinol (Oxf) 89:454–458

    CAS Article Google Scholar

  14. Bi WL, Horowitz P, Greenwald NF, Abedalthagafi M, Agarwalla PK, Gibson WJ et al (2017) Landscape of genomic alterations in pituitary adenomas. Clin Cancer Res 23:1841–1851

    CAS PubMed Article Google Scholar

  15. Song Z-J, Reitman ZJ, Ma Z-Y, Chen J-H, Zhang Q-L, Shou X-F et al (2016) The genome-wide mutational landscape of pituitary adenomas. Cell Res 26:1255–1259

    CAS PubMed PubMed Central Article Google Scholar

  16. Chen J, Jian X, Deng S, Ma Z, Shou X, Shen Y et al (2018) Identification of recurrent USP48 and BRAF mutations in Cushing’s disease. Nat Commun 9:3171

    PubMed PubMed Central Article CAS Google Scholar

  17. Sbiera S, Perez-Rivas LG, Taranets L, Weigand I, Flitsch J, Graf E et al (2019) Driver mutations in USP8 wild-type Cushing’s disease. Neuro Oncol 21:1273–1283

    CAS PubMed PubMed Central Article Google Scholar

  18. Neou M, Villa C, Armignacco R, Jouinot A, Raffin-Sanson ML, Septier A et al (2020) Pangenomic classification of pituitary neuroendocrine tumors. Cancer Cell 37:123-134.e5

    CAS PubMed Article Google Scholar

  19. Uzilov AV, Taik P, Cheesman KC, Javanmard P, Ying K, Roehnelt A et al (2021) USP8 and TP53 drivers are associated with CNV in a corticotroph adenoma cohort enriched for aggressive tumors. J Clin Endocrinol Metab 106:826–842

    PubMed Article Google Scholar

  20. Casar-Borota O, Boldt HB, Engström BE, Andersen MS, Baussart B, Bengtsson D et al (2021) Corticotroph aggressive pituitary tumors and carcinomas frequently harbor ATRX mutations. J Clin Endocrinol Metab 106:1183–1194

    PubMed Article Google Scholar

  21. Campbell PJ, Getz G, Korbel JO, Stuart JM, Jennings JL, Stein LD et al (2020) Pan-cancer analysis of whole genomes. Nature 578:82–93

    Article CAS Google Scholar

  22. Bouaoun L, Sonkin D, Ardin M, Hollstein M, Byrnes G, Zavadil J et al (2016) TP53 variations in human cancers: new lessons from the IARC TP53 database and genomics data. Hum Mutat 37:865–876

    CAS PubMed Article Google Scholar

  23. Horbinski C, Ligon KL, Brastianos P, Huse JT, Venere M, Chang S et al (2019) The medical necessity of advanced molecular testing in the diagnosis and treatment of brain tumor patients. Neuro Oncol 21:1498–1508

    CAS PubMed PubMed Central Article Google Scholar

  24. van Riet J, van de Werken HJG, Cuppen E, Eskens FALM, Tesselaar M, van Veenendaal LM et al (2021) The genomic landscape of 85 advanced neuroendocrine neoplasms reveals subtype-heterogeneity and potential therapeutic targets. Nat Commun 12:4612

    PubMed PubMed Central Article CAS Google Scholar

  25. Herman V, Drazin NZ, Gonsky R, Melmed S (1993) Molecular screening of pituitary adenomas for gene mutations and rearrangements. J Clin Endocrinol Metab 77:50–55

    CAS PubMed Google Scholar

  26. Levy A, Hall L, Yeudall WA, Lightman SL (1994) p53 gene mutations in pituitary adenomas: rare events. Clin Endocrinol (Oxf) 41:809–814

    CAS Article Google Scholar

  27. Tanizaki Y, Jin L, Scheithauer BW, Kovacs K, Roncaroli F, Lloyd RV (2007) P53 gene mutations in pituitary carcinomas. Endocr Pathol 18:217–222

    CAS PubMed Article Google Scholar

  28. Kawashima ST, Usui T, Sano T, Iogawa H, Hagiwara H, Tamanaha T et al (2009) P53 gene mutation in an atypical corticotroph adenoma with Cushing’s disease. Clin Endocrinol (Oxf) 2009:656–657

    Article Google Scholar

  29. Trouillas J, Roy P, Sturm N, Dantony E, Cortet-Rudelli C, Viennet G et al (2013) A new prognostic clinicopathological classification of pituitary adenomas: a multicentric case-control study of 410 patients with 8 years post-operative follow-up. Acta Neuropathol 126:123–135

    PubMed Article Google Scholar

  30. Phan J, Jin Y, Zhang H, Qiang W, Shekhtman E, Shao D et al (2020) ALFA: allele frequency aggregator: national center for biotechnology information, U.S. National Library of Medicine. Available from www.ncbi.nlm.nih.gov/snp/docs/gsr/alfa/

  31. Karczewski KJ, Francioli LC, Tiao G, Cummings BB, Alföldi J, Wang Q et al (2020) The mutational constraint spectrum quantified from variation in 141,456 humans. Nature 581:434–443

    CAS PubMed PubMed Central Article Google Scholar

  32. Fairley S, Lowy-Gallego E, Perry E, Flicek P (2020) The International genome sample resource (IGSR) collection of open human genomic variation resources. Nucleic Acids Res 48:D941–D947

    CAS PubMed Article Google Scholar

  33. Kato S, Han S-Y, Liu W, Otsuka K, Shibata H, Kanamaru R et al (2003) Understanding the function–structure and function–mutation relationships of p53 tumor suppressor protein by high-resolution missense mutation analysis. Proc Natl Acad Sci 100:8424–8429

    CAS PubMed PubMed Central Article Google Scholar

  34. Kawaguchi T, Kato S, Otsuka K, Watanabe G, Kumabe T, Tominaga T et al (2005) The relationship among p53 oligomer formation, structure and transcriptional activity using a comprehensive missense mutation library. Oncogene 24:6976–6981

    CAS PubMed Article Google Scholar

  35. Greenblatt MS, Chappuis PO, Bond JP, Hamel N, Foulkes WD (2001) TP53 mutations in breast cancer associated with BRCA1 or BRCA2 germ-line mutations: distinctive spectrum and structural distribution. Cancer Res 61:4092–4097

    CAS PubMed Google Scholar

  36. Sesta A, Cassarino MF, Terreni M, Ambrogio AG, Libera L, Bardelli D et al (2020) Ubiquitin-Specific Protease 8 mutant corticotrope adenomas present unique secretory and molecular features and shed light on the role of ubiquitylation on ACTH processing. Neuroendocrinology 110:119–129

    CAS PubMed Article Google Scholar

  37. Raverot G, Burman P, McCormack A, Heaney A, Petersenn S, Popovic V et al (2018) European society of endocrinology clinical practice guidelines for the management of aggressive pituitary tumours and carcinomas. Eur J Endocrinol 178:G1-24

    CAS PubMed Article Google Scholar

  38. Thapar K, Scheithauer BW, Kovacs K, Pernicone PJ, Laws ER (1996) p53 expression in pituitary adenomas and carcinomas: correlation with invasiveness and tumor growth fractions. Neurosurgery 38:765–70

    CAS PubMed Article Google Scholar

  39. Gonin-Laurent N, Gibaud A, Huygue M, Lefèvre SH, Le Bras M, Chauveinc L et al (2006) Specific TP53 mutation pattern in radiation-induced sarcomas. Carcinogenesis 27:1266–1272

    CAS PubMed Article Google Scholar

  40. Whitehouse JP, Howlett M, Federico A, Kool M, Endersby R, Gottardo NG (2021) Defining the molecular features of radiation-induced glioma: a systematic review and meta-analysis. Neuro-Oncol Adv 3:1–16

    Google Scholar

  41. Pinto EM, Siqueira SACC, Cukier P, Fragoso MCBVCBV, Lin CJ, De Mendonca BB et al (2011) Possible role of a radiation-induced p53 mutation in a Nelson’s syndrome patient with a fatal outcome. Pituitary 14:400–404

    PubMed Article Google Scholar

  42. Orr BA, Clay MR, Pinto EM, Kesserwan C (2020) An update on the central nervous system manifestations of Li–Fraumeni syndrome. Acta Neuropathol 139:669–87

    CAS PubMed Article Google Scholar

  43. Birk H, Kandregula S, Cuevas-Ocampo A, Wang CJ, Kosty J, Notarianni C (2022) Pediatric pituitary adenoma and medulloblastoma in the setting of p53 mutation: case report and review of the literature. Childs Nerv Syst. https://doi.org/10.1007/s00381-022-05478-8

    Article Google Scholar

  44. Granja F, Morari J, Morari EC, Correa LAC, Assumpção LVM, Ward LS (2004) Proline homozygosity in codon 72 of p53 is a factor of susceptibility for thyroid cancer. Cancer Lett 210:151–157

    CAS PubMed Article Google Scholar

  45. Sagne C, Marcel V, Amadou A, Hainaut P, Olivier M, Hall J (2013) A meta-analysis of cancer risk associated with the TP53 intron 3 duplication polymorphism (rs17878362): geographic and tumor-specific effects. Cell Death Dis 4:e492

    CAS PubMed PubMed Central Article Google Scholar

  46. Katkoori VR, Jia X, Shanmugam C, Wan W, Meleth S, Bumpers H et al (2009) Prognostic significance of p53 Codon 72 polymorphism differs with race in colorectal adenocarcinoma. Clin Cancer Res 15:2406–2416

    CAS PubMed PubMed Central Article Google Scholar

  47. Dumont P, Leu JIJ, Della Pietra AC, George DL, Murphy M (2003) The codon 72 polymorphic variants of p53 have markedly different apoptotic potential. Nat Genet 33:357–365

    CAS PubMed Article Google Scholar

  48. Basu S, Gnanapradeepan K, Barnoud T, Kung CP, Tavecchio M, Scott J et al (2018) Mutant p53 controls tumor metabolism and metastasis by regulating PGC-1α. Genes Dev 32:230–243

    CAS PubMed PubMed Central Article Google Scholar

  49. Yagnik G, Jahangiri A, Chen R, Wagner JR, Aghi MK (2017) Role of a p53 polymorphism in the development of nonfunctional pituitary adenomas. Mol Cell Endocrinol 446:81–90

    CAS PubMed PubMed Central Article Google Scholar

  50. Andonegui-Elguera S, Silva-Román G, Peña-Martínez E, Taniguchi-Ponciano K, Vela-Patiño S, Remba-Shapiro I et al (2022) The genomic landscape of corticotroph tumors: from silent adenomas to ACTH-secreting carcinomas. Int J Mol Sci. 23:4861

    CAS PubMed PubMed Central Article Google Scholar

Download references

Funding

Open Access funding enabled and organized by Projekt DEAL. The study was supported by the Deutsche Forschungsgemeinschaft (DFG) (Project number: 314061271-TRR 205 to MF, MR and MT; FA 466/5-1 to MF; DE 2657/1-1 to TD), Metiphys program of the LMU Medical Faculty (to AA), Else Kröner-Fresenius Stiftung (Project number: 2012_A103 and 2015_A228 to MR) and Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ; Project number: E-26/211.294/2021 to MRG).

Author information

Authors and Affiliations

  1. Medizinische Klinik und Poliklinik IV, Klinikum der Universität München, Ludwig-Maximilians-Universität München, Munich, GermanyLuis Gustavo Perez-Rivas, Julia Simon, Adriana Albani, Sicheng Tang, Günter K. Stalla, Martin Reincke & Marily Theodoropoulou
  2. Center for Neuropathology and Prion Research, Ludwig-Maximilians-Universität München, Munich, GermanySigrun Roeber & Jochen Herms
  3. Department of Endocrinology, Center for Rare Adrenal Diseases, Assistance Publique-Hôpitaux de Paris, Hôpital Cochin, Paris, FranceGuillaume Assié
  4. Université de Paris, Institut Cochin, Inserm U1016, CNRS UMR8104, F-75014, Paris, FranceGuillaume Assié
  5. Division of Endocrinology and Diabetes, Department of Internal Medicine I, University Hospital, University of Würzburg, Würzburg, GermanyTimo Deutschbein & Martin Fassnacht
  6. Medicover Oldenburg MVZ, Oldenburg, GermanyTimo Deutschbein
  7. Division of Endocrinology, Hospital Universitário Clementino Fraga Filho, Rio de Janeiro, BrazilMonica R. Gadelha
  8. Division of Endocrinology, Department of Internal Medicine, Radboud University Medical Centre, Nijmegen, The NetherlandsAd R. Hermus
  9. Medicover Neuroendocrinology, Munich, GermanyGünter K. Stalla
  10. Service d’Endocrinologie, Centre Hospitalier du Nord, Ettelbruck, LuxembourgMaria A. Tichomirowa
  11. Department of Neurosurgery, Universitätskrankenhaus Hamburg-Eppendorf, Hamburg, GermanyRoman Rotermund & Jörg Flitsch
  12. Department of Neurosurgery, University of Erlangen-Nürnberg, Erlangen, GermanyMichael Buchfelder
  13. Department of Neurosurgery, University of Tübingen, Tübingen, GermanyIsabella Nasi-Kordhishti & Jürgen Honegger
  14. Neurochirurgische Klinik und Poliklinik, Klinikum der Universität München, Ludwig-Maximilians-Universität München, Munich, GermanyJun Thorsteinsdottir
  15. Institute of Neuropathology, University Medical Center Hamburg-Eppendorf, Hamburg, GermanyWolfgang Saeger

Contributions

LPGR and MT designed the study. LPGR, JS, AA and ST implemented the study. LGPR did the data analysis. SR, GA, TD, MF, MRG, ARH, GKS, MAT, RR, JF, MB, INK, JH, JT, WS, JH and MR provided patient materials and data. LGPR and MT interpreted the data and composed the main draft of the manuscript. All authors have seen, corrected and approved the final draft.

Corresponding authors

Correspondence to Luis Gustavo Perez-Rivas or Marily Theodoropoulou.

Ethics declarations

Ethics approval and consent to participate

The study was performed in accordance with the Declaration of Helsinki and was approved by the ethics committee of the LMU Munich (Nr. 643-16). All patients provided written informed consent.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1 of TP53 mutations in functional corticotroph tumors are linked to invasion and worse clinical outcome

Skip to file navigationSkip to generic navigation
1
Supplementary Table 1
. Description of study cohort.
Variable
mean/median
SD/IQR
Total n
Age at diagnosis (years), mean ±SD, [total n]
42
±15.2
86
Sex (female), n (%), [total n]
62
(72%)
86
BMI (kg/m2), mean ±SD, [total n]
28.9
±6.3
74
Disease presentation, n (%), [total n]
86
Cushing
66
(77%)
Nelson
20
(23%)
Number of prior pituitary surgeries, n (%), [total n]
80
0
50
(63%)
1
23
(29%)
≥2
7
(9%)
Total
number of pituitary surgeries, n (%), [total n]
82
1
46
(56%)
2
23
(28%)
≥3
13
(16%)
Complete tumor resection, n (%), [total n]
32
(60%)
53
Postoperative remission, n (%), [total n]
46
(59%)
78
Postoperative tumor control, n (%), [total
n]
34
(60%)
57
Radiation therapy, n (%), [total n]
24
(34%)
70
Radiation therapy before sample collection, n (%), [total n]
7
(13%)
53
Bilateral adrenalectomy, n (%), [total n]
23
(27%)
86
Pharmacological treatments
a
,
n (%), [total n]
18
(42%)
43
Preoperative hormone levels
Plasma ACTH (pg/mL), median (IQR)
98
(570.4)
75
Serum cortisol (
μ
g/dl), median (range)
29.1
(168.6)
50
24h
urinary free cortisol (
μ
g/24h), median (range)
432.5
(598.3)
30
Serum cortisol after low
dose DST (
μ
g/dl),
median (IQR)
20
(20.7)
46
Postoperative hormone levels
Plasma ACTH (pg/mL), median (IQR)
20
(107.6)
57
Serum cortisol nadir (
μ
g/dl), median (range)
8.8
(19.4)
58
Tumo
r size (mm), median (IQR), [total n]
15
(13.0)
85
Microadenoma
19
(22%)
Macroadenoma
66
(78%)
Granulation, n (%), [total n]
30
Sparsely
9
(30%)
Densely
21
(70%)
Ki67 index, median (IQR), [total n]
2.0
(3.8)
36
Ki67 index ≥3%, n (%)
14
(39%)
36
p53 positivity, median (IQR), [total n]
1
(26.5)
9
Invasion, n (%),
[total n]
34
(53%)
64
Hardy grade, n (%), [total n]
61
1
13
(21%)
2
22
(36%)
3
18
(30%)
4
8
(13%)
Knosp grade, n (%), [total n]
35
0
5
(14%)
1
12
(34%)
2
3
(9%)
3
7
(20%)
4
8
(7%)
Disease
specific death, n (%), [total
n]
5
(9%)
58
a
Pharmacological treatments: pasireotide (n=6), ketoconazole (n=5), mitotane (n=5), temozolamide
(n=4) metyrapone (n=5), cabergoline (n=3), bevazizumab (n=1). Five patients received >1
pharmacological agent.
2
Supplementary Table 2
. Primers used for
TP53
amplification and Sanger sequencing.
Primer
Sequence
DNA source
TP53
1
5′
TCTCATGCTGGATCCCCACT
3′
FF, FFPE
TP53
1rv
5′
GACCAGGTCCTCAGCC
3′
FFPE
TP53
2fw
5′
GGGGGCTGAGGACCTGGT
3′
FFPE
TP53
2rv
5′
ATACGGCCAGGCATTGAAGT
3′
FFPE
TP53
2
5′
AGAGGAATCCCAAAGTTCCA
3′
FF
TP53
3
5′
GTGCCCTGACTTTCAACTC
3′
FF, FFPE
TP53
3rv
5′
GGCAACCAGCCCTGTC
3′
FFPE
TP53
4fw
5′
GCCTCTGATTCCTCACTGAT
3′
FFPE
TP53
4
5′
CAGGAGAAAGCCCCCCTACT
3′
FF, FFPE
TP53
5
5′
CTTGCCACAGGTCTCCCCAA
3′
FF, FFPE
TP53
6
5′
AGGGGTCAGAGGCAAGCAGA
3′
FF, FFPE
TP53
7
5′
TAGGACCTGATTTCCTTA
3′
FF, FFPE
TP53
7rv
5′
AGTGAATCTGAGGCATAAC
3′
FFPE
TP53
7Bfw
5′
TGGAGGAGACCAAGGGTG
3′
FFPE
TP53
7Brv
5′
CGGCATTTTGAGTGTTAGAC
3′
FFPE
TP53
8
5′
TAAGCTATGATGTTCCTTAG
3′
FF, FFPE
TP53
8rv
5′
GACTGTTTTACCTGCAATTG
3′
FFPE
TP53
9
5′
CAATTGTAACTTGAACCATC
3′
FF, FFPE
TP53
10
5′
GGATGAGAATGGAATCCTAT
3′
FF, FFPE
TP53
11
5′
TCTCACTCATGTGATGTCATC
3′
FF, FFPE
TP53
12
5′
CACACCTATTGCAAGCAAGG
3′
FF, FFPE
FF, fresh frozen; FFPE, formalin
fixed
paraffin embedded.

Additional file 1

Supplementary Table 1: Description of study cohort. Supplementary Table 2: Primers used for TP53 amplification and Sanger sequencing. Supplementary Table 3: Common TP53 variants in the study cohort. Supplementary Table 4: Comparison of TP53 mutant versus TP53 wild type group. Supplementary Figure 1. Chromatograms showing the TP53 variants found in the corticotroph tumor of patient #1 and #8 (Table 1). A. The variant c.398T>A was present in homozygocity in the tumor and absent in the blood. B. The variant c.1009C>G is detected in all available surgical specimens in this patient. First and 2nd surgeries were Cushing’s disease tumors and 4th and 5th CTP-BADX/NS.

 

%d bloggers like this: