Think Like a Doctor: Red Herrings Solved!

Posted on January 11, 2026 by MaryO

By LISA SANDERS, M.D.

On Thursday we challenged Well readers to take the case of a 29-year-old woman with an injured groin, a swollen foot and other abnormalities. Many of you found it as challenging as the doctors who saw her. I asked for the right test as well as the right diagnosis. More than 200 answers were posted.

The right test was…

The dexamethasone suppression test,though I counted those of you who suggested measuring the cortisol in the urine.

The right diagnosis was…

Cushing’s disease

More than a dozen of you got the right answer or the right test, but Dr. Davin Quinn, a consultant psychiatrist at the University of New Mexico Hospital, was the first to be right on both counts. As soon as he saw that the patient’s cortisol level was increased, he thought of Cushing’s. And he had treated a young patient like this one some years ago as a second year resident.

The Diagnosis:

Cushing’s disease is caused by having too much of the stress hormone cortisol in the body. Cortisol is made in the adrenal glands, little pyramid shaped organs that sit atop the kidneys. It is normally a very tightly regulated hormone that helps the body respond to physical stress.

Sometimes the excess comes from a tumor in the adrenal gland itself that causes the little organ to go into overdrive, making too much cortisol. More often the excess occurs when a tumor in the pituitary gland in the brain results in too much ACTH, the hormone that controls the adrenal gland.

In the body, cortisol’s most fundamental job is to make sure we have enough glucose around to get the body’s work done. To that end, the hormone drives appetite, so that enough fuel is taken in through the food we eat. When needed, it can break muscle down into glucose. This essential function accounts for the most common symptoms of cortisol excess: hyperglycemia, weight gain and muscle wasting. However, cortisol has many functions in the body, and so an excess of the hormone can manifest itself in many different ways.

Cushing’s was first described by Dr. Harvey Cushing, a surgeon often considered the father of modern neurosurgery. In a case report in 1912, he described a 23-year-old woman with sudden weight gain, mostly in the abdomen; stretch marks from skin too thin and delicate to accommodate the excess girth; easy bruising; high blood pressure and diabetes.

Dr. Cushing’s case was, it turns out, a classic presentation of the illness. It wasn’t until 20 years later that he recognized that the disease had two forms. When it is a primary problem of an adrenal gland gone wild and producing too much cortisol on its own, the disease is known as Cushing’s syndrome. When the problem results from an overgrown part of the pituitary making too much ACTH and causing the completely normal adrenal glands to overproduce the hormone, the illness is called Cushing’s disease.

It was an important distinction, since the treatment often requires a surgical resection of the body part where the problem originates. Cushing’s syndrome can also be caused by steroid-containing medications, which are frequently used to treat certain pulmonary and autoimmune diseases.

How the Diagnosis Was Made:

After the young woman got her lab results from Dr. Becky Miller, the hematologist she had been referred to after seeing several other specialists, the patient started reading up on the abnormalities that had been found. And based on what she found on the Internet, she had an idea of what was going on with her body.

“I think I have Cushing’s disease,” the patient told her endocrinologist when she saw him again a few weeks later.

The patient laid out her argument. In Cushing’s, the body puts out too much cortisol, one of the fight-or-flight stress hormones. That would explain her high blood pressure. Just about everyone with Cushing’s disease has high blood pressure.

She had other symptoms of Cushing’s, too. She bruised easily. And she’d been waking up crazy early in the morning for the past year or so – around 4:30 – and couldn’t get back to sleep. She’d heard that too much cortisol could cause that as well. She was losing muscle mass – she used to have well-defined muscles in her thighs and calves. Not any more. Her belly – it wasn’t huge, but it was a lot bigger than it had been. Cushing’s seemed the obvious diagnosis.

The doctor was skeptical. He had seen Cushing’s before, and this patient didn’t match the typical pattern. She was the right age for Cushing’s and she had high blood pressure, but nothing else seemed to fit. She wasn’t obese. Indeed, she was tall (5- foot-10) and slim (150 pounds) and athletic looking. She didn’t have stretch marks; she didn’t have diabetes. She said she bruised easily, but the endocrinologist saw no bruises on exam. Her ankle was still swollen, and Cushing’s can do that, but so can lots of other diseases.

The blood tests that Dr. Miller ordered measuring the patient’s ACTH and cortisol levels were suggestive of the disease, but many common problems — depression, alcohol use, eating disorders — can cause the same result. Still, it was worth taking the next step: a dexamethasone suppression test.

Testing, Then Treatment:

The dexamethasone suppression test depends on a natural negative feedback loop whereby high levels of cortisol suppress further secretion of the hormone. Dexamethasone is an artificial form of cortisol. Given in high doses, it will cause the level of naturally-occurring cortisol to drop dramatically.

The patient was told to take the dexamethasone pills the night before having her blood tested. The doctor called her the next day.

“Are you sure you took the pills I gave you last night?” the endocrinologist asked her over the phone. The doctor’s voice sounded a little sharp to the young woman, tinged with a hint of accusation.

“Of course I took them,” she responded, trying to keep her voice clear of any irritation.

“Well, the results are crazy,” he told her and proposed she take another test: a 24-hour urine test.

Because cortisol is eliminated through the kidneys, collecting a full day’s urine would show how much cortisol her body was making. So the patient carefully collected a day’s worth of urine.

A few days later, the endocrinologist called again: her cortisol level was shockingly high. She was right, the doctor conceded, she really did have Cushing’s.

An M.R.I. scan revealed a tiny tumor on her pituitary. A couple of months later, she had surgery to remove the affected part of the gland.

After recovering from the surgery, the patient’s blood pressure returned to normal, as did her red blood cell count and her persistently swollen ankle. And she was able to once again sleep through the night.

Red Herrings Everywhere:

As many readers noted, there were lots of findings that didn’t really add up in this case. Was this woman’s groin sprain part of the Cushing’s? What about the lower extremity swelling, and the excess red blood cell count?

In the medical literature, there is a single case report of high red blood cell counts as the presenting symptom in a patient with Cushing’s. And with this patient, the problem resolved after her surgery – so maybe they were linked.

And what about the weird bone marrow biopsy? The gastritis? The enlarged spleen? It’s hard to say for certain if any of these problems was a result of the excess cortisol or if she just happened to have other medical problems.

Why the patient didn’t have the typical symptoms of Cushing’s is easier to explain. She was very early in the course of the disease when she got her diagnosis. Most patients are diagnosed once symptoms have become more prominent

By the time this patient had her surgery, a couple of months later, the round face and belly characteristic of cortisol excess were present. Now, two years after her surgery, none of the symptoms remain.

From http://well.blogs.nytimes.com/2014/01/17/think-like-a-doctor-red-herrings-solved/?_php=true&_type=blogs&_r=0

Filed under: adrenal, Cushing's, pituitary | Tagged: , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , | Leave a comment »

Metastatic Pituitary Carcinoma Successfully Treated with Radiation, Chemo.

Posted on January 10, 2026 by MaryO

A man with Cushing’s disease — caused by an adrenocorticotrophic hormone (ACTH)-secreting pituitary adenoma — who later developed metastases in the central nervous system without Cushing’s recurrence, was successfully treated over eight years with radiation and chemotherapy, according to a case report.

The report, “Long-term survival following transformation of an adrenocorticotropic hormone secreting pituitary macroadenoma to a silent corticotroph pituitary carcinoma: Case report,” was published in the journal World Neurosurgery.

Pituitary carcinomas make up only 0.1-0.2% of all pituitary tumors and are characterized by a primary pituitary tumor that metastasizes into cranial, spinal, or systemic locations. Fewer than 200 cases have been reported in the literature.

Most of these carcinomas secrete hormones, with ACTH being the most common. Though the majority of ACTH-secreting carcinomas present with Cushing’s disease, about one-third do not show symptoms of the condition and have normal serum cortisol and ACTH levels. These are called silent corticotroph adenomas and are considered more aggressive.

A research team at the University of Alabama at Birmingham presented the case of a 51-year-old Caucasian man with ACTH-dependent Cushing’s disease. He had undergone an incomplete transsphenoidal (through the nose) resection of an ACTH-secreting pituitary macroadenoma – larger than 10 mm in size – and radiation therapy the year before.

At referral in August 1997, the patient had persistent high cortisol levels and partial hypopituitarism, or pituitary insufficiency. He exhibited Cushing’s symptoms, including facial reddening, moon facies, weight gain above the collarbone, “buffalo hump,” and abdominal stretch marks.

About two years later, the man was weaned off ketoconazole — a medication used to lower cortisol levels — and his cortisol levels had been effectively reduced. He also had no physical manifestations of Cushing’s apart from facial reddening.

In May 2010, the patient reported two episodes of partial seizures, describing two spells of right arm tingling, followed by impaired peripheral vision. Imaging showed a 2.1-by-1-cm mass with an associated cyst within the brain’s right posterior temporal lobe, as well as a 1.8-by-1.2-cm mass at the cervicomedullary junction, which is the region where the brainstem continues as the spinal cord. His right temporal cystic mass was then removed by craniotomy.

A histopathologic analysis was consistent with pituitary carcinoma. Cell morphology was generally similar to the primary pituitary tumor, but cell proliferation was higher. Physical exams showed no recurrence of Cushing’s disease and 24-hour free urinary cortisol was within the normal range.

His cervicomedullary metastasis was treated with radiation therapy in July 2010. He took the oral chemotherapy temozolomide until August 2011, and Avastin (bevacizumab, by Genentech) was administered from September 2010 to November 2012.

At present, the patient continues to undergo annual imaging and laboratory draws. He receives treatment with hydrocortisone, levothyroxine — synthetic thyroid hormone — and testosterone replacement with androgel.

His most recent exam showed no progression over eight years of a small residual right temporal cyst, a residual mass along the pituitary stalk — the connection between the hypothalamus and the pituitary gland — and a small residual mass at the cervicomedullary junction. Lab results continue to show no Cushing’s recurrence.

“Our case is the first to document a patient who initially presented with an endocrinologically active ACTH secreting pituitary adenoma and Cushing’s disease who later developed cranial and spinal metastases without recurrence of Cushing’s disease and transformation to a silent corticotroph pituitary carcinoma,” the scientists wrote.

They added that the report is also the first documenting “8 years of progression-free survival in a patient with pituitary carcinoma treated with radiotherapy, [temozolomide] and bevacizumab.”

Adapted from https://cushingsdiseasenews.com/2019/01/03/successful-treatment-pituitary-carcinoma-radiation-chemo-case-report/

Filed under: Cancer, Cushing's, pituitary | Tagged: , , , , , , , , , | Leave a comment »

Consecutive Resections of Double Pituitary Adenoma for Resolution of Cushing Disease

Posted on December 20, 2025 by MaryO

BACKGROUND

Double pituitary adenomas are rare presentations of two distinct adenohypophyseal lesions seen in <1% of surgical cases. Increased rates of recurrence or persistence are reported in the resection of Cushing microadenomas and are attributed to the small tumor size and localization difficulties. The authors report a case of surgical treatment failure of Cushing disease because of the presence of a secondary pituitary adenoma.

OBSERVATIONS

A 32-year-old woman with a history of prolactin excess and pituitary lesion presented with oligomenorrhea, weight gain, facial fullness, and hirsutism. Urinary and nighttime salivary cortisol elevation were elevated. Magnetic resonance imaging confirmed a 4-mm3 pituitary lesion. Inferior petrosal sinus sampling was diagnostic for Cushing disease. Primary endoscopic endonasal transsphenoidal resection was performed to remove what was determined to be a lactotroph-secreting tumor on immunohistochemistry with persistent hypercortisolism. Repeat resection yielded a corticotroph-secreting tumor and postoperative hypoadrenalism followed by long-term normalization of the hypothalamic-pituitary-adrenal axis.

LESSONS

This case demonstrates the importance of multidisciplinary management and postoperative hormonal follow-up in patients with Cushing disease. Improved strategies for localization of the active tumor in double pituitary adenomas are essential for primary surgical success and resolution of endocrinopathies.

ABBREVIATIONS

ACTH = adrenocorticotrophic hormone;  BMI = body mass index;  DHEA-S = dehydroepiandrosterone sulfate;  FSH = follicle-stimulating hormone;  GH = growth hormone;  IHC = immunohistochemical;  IPSS = inferior petrosal sinus sampling;  LH = luteinizing hormone;  MRI = magnetic resonance imaging;  POD = postoperative day;  T4 = thyroxine;  TF = transcription factor;  TSH = thyroid-stimulating hormone;  UFC = urinary free cortisol

Pituitary adenomas are adenohypophyseal tumors that can cause endocrinopathies, such as pituitary hormone hypersecretion or anterior hypopituitarism. Cell lineages are used to classify these tumors on the basis of immunohistochemical (IHC) staining of transcription factors, hormones, and other biomarkers.1 Pituitary adenomas differentiate from pluripotent stem cells along one of three lineage pathways, depending on the following active transcription factors (TFs): pituitary transcription factor 1 (PIT-1), T-box transcription factor (TPIT), or steroidogenic factor-1 (SF-1). Rarely, two or more discrete pituitary adenomas from different lineages are identified in patients; however, the etiology remains unclear.2 The incidence of multiple pituitary adenomas has been reported to be 1%–2% of all resected pituitary adenomas but is likely underestimated based on data from large autopsy series.1–4 Pluri-hormonal adenomas are also rare entities in which a single tumor contains multiple TF lineages with one or more hormonal excesses.1–3 Preoperative recognition of multiple or pluri-hormonal pituitary adenomas is rare, and most tumors are discovered incidentally upon autopsy, intraoperatively, or on histological analysis.2,3,5

In cases of multiple synchronous pituitary adenomas, only one hormone excess syndrome is most frequently evident on clinical presentation and endocrine workup. Silent pituitary tumors positive for prolactin on immunohistochemistry are the most prevalent additional, incidentally found tumor in cases of multiple pituitary adenomas.5 This is particularly true in Cushing disease.6,7 It is important to recognize the presence of multiple pituitary adenomas especially in the setting of hormonally active pituitary adenomas to provide optimal management for this subset of patients. Complete resection is curative for Cushing disease with the standard of care achieved through a transsphenoidal approach. Localization of the tumor presents a challenge because of suboptimal sensitivity of magnetic resonance imaging (MRI) in demonstrating microadenomas, the inconsistency of lateralization with inferior petrosal sinus sampling (IPSS), and delays in pathological analysis.1,8,9 Additionally, the presence of an additional pituitary adenoma can obscure the microtumor through its large size and mass effect and can act as a “decoy lesion” during MRI, IPSS, and resection.6

Consideration of multiple pituitary tumors is necessary for successful resection. In a patient with a biochemical picture of Cushing disease, the demonstration of an adenoma with negative adrenocorticotrophic hormone (ACTH) immunostaining and the absence of postoperative hypoadrenalism may indicate the existence of a double adenoma. Few cases have described a lack of remission of an endocrinopathy after transsphenoidal resection due to the presence of an additional adenoma,2,6,10 and even less so in the instance of the persistence of Cushing disease.6 We present a rare case of double pituitary adenomas in a patient presenting with Cushing disease who underwent two endoscopic endonasal transsphenoidal resections and immunostaining for prolactin and ACTH, respectively, with long-term normalization of her hypothalamic-pituitary-adrenal (HPA) axis.

Illustrative Case

History and Presentation

A 32-year-old female, gravida 0 para 0, with a history of a pituitary lesion and hyperprolactinemia presented to our institution for the evaluation for Cushing disease. Ten years earlier, the patient had presented to a gynecologist with hirsutism, galactorrhea, and oligomenorrhea. Her endocrine workup was remarkable for an elevated prolactin at 33.8 ng/mL (2.3–23.3 ng/mL), while follicle-stimulating hormone (FSH), luteinizing hormone (LH), and thyroid-stimulating hormone (TSH) levels were normal. No ACTH or cortisol levels were available. MRI demonstrated a 5 × 6 × 5–mm T1-weighted isointense pituitary lesion protruding into the suprasellar cistern due to a small sella size. She was treated with bromocriptine 2.5 mg daily for 5 years, with normalization of her prolactin level. Subsequent MRI demonstrated a stable lesion size and T1 and T2 hyperintensity in the region of the known pituitary lesion, considered to be posttreatment cystic change with proteinaceous contents and blood. After the normalization of her prolactin levels, she continued to have oligomenorrhea and abnormal hair growth. Polycystic ovaries were not visualized on ultrasound. She was started on oral contraceptives and then switched to the etonorgestrel implant.

A decade after initial presentation, she presented to endocrinology at our institution with 3 years of weight gain, hirsutism, and potential oligomenorrhea. Vital signs were stable (blood pressure: 122/86; heart rate: 72 beats/min), and facial fullness and striae on her bilateral breasts were appreciated on physical examination. She was taking isoniazid and pyridoxine for a recent diagnosis of latent tuberculosis and had discontinued bromocriptine 5 years earlier. Her weight was 66.3 kg and body mass index (BMI) was 23.9 kg/m2. She reported that her maternal uncle had a pituitary tumor. Laboratory analysis was positive for elevated urinary free cortisol (UFC) of 109 µg per 24 hours (2.5–45 µg/24 h; Table 1) and nighttime salivary cortisol of 142 ng/mL (<100 ng/dL) with high-normal prolactin of 22.8 ng/mL (2.3–23.3 ng/dL) and normal FSH, LH, TSH, and thyroxine (T4). Dehydroepiandrosterone sulfate (DHEA-S) was 128 µg/dL (98.8–340.0 µg/dL). Imaging demonstrated a 4 × 4 × 4–mm pituitary lesion with decreased T1-weighted and increased central T2-weighted signal intensity in the left lateral pituitary (Fig. 1A–C). Desmopressin (Ferring Pharmaceuticals DDAVP) stimulation increased a basal ACTH of 49.9 pg/mL to ACTH of 91.2 pg/mL, and cortisol increased from 13.7 µg/dL to 21.2 µg/dL, consistent with neoplastic hypercortisolism. IPSS was performed, which showed a right-sided, central-to-peripheral ACTH gradient (Table 2). The patient elected to undergo endoscopic endonasal resection with the initial target as the left-lateral pituitary mass to achieve a cure for Cushing disease.

TABLE 1Urinary free cortisol at baseline and 3, 5, and 7 months after the primary resection

Variable Baseline 3 Mos 5 Mos 7 Mos on Osilodrostat
Urinary free cortisol (4–50 µg/24 hrs) 109 134.2 125.4 40.3
Urinary creatinine (0.5–2.5 g/24 hrs) 0.995 1.17 1.42 1.11
Urinary vol (mL) 1950 2300 2100 2125
FIG. 1
FIG. 1

Preoperative coronal precontrast (A) and postcontrast (B) T1-weighted magnetic resonance imaging (MRI) and T2-weighted MRI (C) demonstrated a 4-mm3 lesion (arrows) with decreased T1 and increased central T2 signal intensity in the left lateral pituitary. Two days after surgery, coronal precontrast (D) and postcontrast T1-weighted (E) and T2-weighted (F) MRI demonstrated the unchanged adenoma.

TABLE 2Preoperative inferior petrosal sinus sampling with corticorelin ovine triflutate 68 µg

Time (mins) ACTH (pg/mL) Prolactin (ng/mL)
Peripheral Petrosal Sinus ACTH Ratio Peripheral Petrosal Sinus Prolactin Ratio
Rt Lt Rt Lt Rt Lt Rt Lt
−5 50.6 225 1586 4.45 31.34 21 124 295 5.90 14.05
0 48.8 389 1376 7.97 28.20 22.2 185 198 8.33 8.92
3 69.8 4680 1333 67.05 19.1 22.1 396 32.5 17.92 1.47
5 80.9 4590 1623 56.74 20.06 22.1 436 32.2 19.73 1.46
10 112 4160 1660 37.14 14.82 20.2 367 42 17.90 2.05

ACTH or prolactin ratio = inferior petrosal sinus ACTH or prolactin/peripheral blood ACTH or prolactin.

Primary Resection and Outcomes

During the primary resection, abnormal tissue was immediately visible after a linear incision along the bottom of the dura, with an excellent plane of dissection. The inferomedial adenoma was distinct from the known left lateral lesion, and the resection was considered complete by the primary neurosurgeon. Subsequently, the left-sided adenoma was not pursued because of the historical prolactinoma diagnosis and an assumption that the newly discovered adenoma was the cause of ACTH hypersecretion. However, pathology of the inferomedial tumor was strongly and diffusely positive for prolactin (Fig. 2B), synaptophysin, and cytokeratin, with an Mindbomb Homolog-1 (MIB-1) proliferative index of 2.4%. ACTH, growth hormone (GH), FSH, LH, and TSH immunostaining were negative. TF immunohistochemistry was not available. On postoperative day (POD) 1, pituitary MRI was performed and demonstrated the unchanged 4-mm3 T1-weighted hypointense lesion with small central T2-weighted hyperintensity in the left lateral gland (Fig. 1D–F). Cortisol levels ranged from 9.7 to 76.2 µg/dL (4.8–19.5 µg/dL), and ACTH was 19.5 pg/mL (7.2–63.3 pg/mL) on POD 1.

FIG. 2
FIG. 2

Histological examination of surgical specimens from the inferomedial (A–C) and left lateral (D–F) lesions. The initial resection (hematoxylin and eosin [H&E], A) was strongly and diffusely positive for prolactin (B) with normal reticulin levels (C) indicating a lactotrophic pituitary adenoma. The second operation (H&E, D) was diagnostic for a corticotropic pituitary adenoma with diffusely positive adrenocorticotrophic hormone (ACTH) (E) and decreased reticulin (F). Original magnification ×100.

Early reoperation was discussed with the patient based on the pathology and persistent hypercortisolism; however, she elected to pursue conservative management with close follow-up. Postoperative cortisol nadir was 4.8 µg/dL (4.8–19.5 µg/dL) on POD 2 during her 4-day hospital stay. DHEA-S was significantly decreased from baseline at 22.3 µg/dL (98.8–340.0 µg/dL) and a prolactin level of 3.4 ng/mL (2.3–23.3 ng/dL) was low-normal. No glucocorticoids were administered during her hospital course. There was no clinical evidence of vasopressin deficiency while she was an inpatient.

Three months postoperatively, the patient reported insomnia, poor hair quality, fatigue, nocturnal sweating, and continued increasing weight gain with fat accumulation in the supraclavicular and dorsal cervical area. She had one spontaneous menstrual period despite the use of etonogestrel implant. UFC was increased at 134.2 µg/24 hours (4–50 µg/24 h; Table 1). The 8:00 am serum cortisol was 10.2 µg/dL (5.0–25.0 µg/dL). She was started on osilodrostat 2 mg twice daily for her persistent hypercortisolism, and she reported some clinical improvement; however, she had continued elevation in her late-night salivary cortisol levels up to 7.0 nmol/L. Other endocrine lab work was normal, with a prolactin of 13.5 ng/mL (2.8–23.3 ng/mL) and TSH of 3.67 µIU/mL (0.4–4.0 µIU/mL). Her weight had increased by 4.9 kg to 71.2 kg with a BMI of 25.3 kg/m2. Approximately 6 months postoperatively, she was amenable to a secondary resection targeting the remaining left lateral pituitary adenoma.

Secondary Resection and Outcomes

After obtaining adequate exposure for the secondary resection, the lesion in the left lateral aspect of the pituitary was targeted. The tumor was clearly identified and completely resected without intraoperative complication. IHC staining was diffusely positive for ACTH (Fig. 2E), synaptophysin, and cytokeratin with decreased reticulin and an MIB-1 index of 3.3%. Prolactin, GH, TSH, LH, and FSH immunostaining were negative. Postoperative cortisol monitoring demonstrated decreased levels, with a nadir of 2.0 µg/dL on POD 0. Levels of ACTH and DHEA-S were decreased at 4.4 pg/mL (7.2–63.3 pg/mL) and 13.3 µg/dL (98.8–340 µg/dL), respectively, on POD 1. Prolactin remained within the normal range at 8.2 ng/mL (2.8–23.3 ng/mL). The patient was started on intravenous hydrocortisone 50 mg every 8 hours for adrenal insufficiency. Postoperative symptoms of nausea, headache, and muscle weakness resolved with hydrocortisone administration. She was discharged on hydrocortisone 60 mg daily in divided doses for adrenal insufficiency and had no signs of vasopressin deficiency during her 2-day hospital course.

By 3 months, the patient reported decreased fatigue, myalgia, and insomnia and improved overall well-being and physical appearance. She was weaned down to a total daily dose of 20 mg of hydrocortisone and had lost 5.2 kg. Her menstruation returned while having an etonogestrel implant. Rapid ACTH stimulation was abnormal, with decreased cortisol at 30 minutes of 4.1 µg/dL (7.2–63.3 pg/mL) demonstrating continued adrenal insufficiency. Follow-up MRI demonstrated miniscule remaining left pituitary adenoma (Fig. 3). Seven months after her second surgery, she was started on 50 µg levothyroxine for primary hypothyroidism in the setting of slightly elevated TSH of 4.1 µIU/mL (0.4–4.0 µIU/mL) and a low-normal T4 of 0.8 ng/dL (0.7–1.5 ng/dL).

FIG. 3
FIG. 3

Postoperative imaging 3 months after the second operation demonstrates near gross-total resection (yellow arrows: surgical cavity) of the left lateral pituitary adenoma on coronal precontrast (A) and postcontrast T1-weighted (B) and T2-weighted (C) MRI.

Two years after the second resection, the patient lost 10.1 kg (weight, 61.1 kg; BMI, 21.76 kg/m2). Her ACTH stimulation test became normal, and hydrocortisone therapy was discontinued. At the 2-year time point, the patient and her husband successfully conceived a child.

Patient Informed Consent

The necessary patient informed consent was obtained in this study.

Discussion

Double or multiple pituitary adenomas are discovered in 0.37%–2.6% of resected pituitary lesions.3,4,6,11,12 A majority of multiple pituitary adenomas are not suspected before surgery with an inconclusive clinical presentation or endocrine laboratory workup.6 The presentation of multiple synchronous neoplasms is thought to be more common than having a single neoplasm with multiple lineages.1 Studies have shown that additional pituitary adenomas are seen at a rate of 1.6%–3.3% in Cushing disease in studies including both contiguous and noncontiguous double pituitary adenomas.6 Additional pituitary adenomas that are hormonally active make up 40% of resected double pituitary adenomas, with most staining for gonadotroph adenoma.13 Overall, the most common incidental pituitary adenoma is prolactinoma,6 which occurs most frequently with GH or ACTH adenomas.5 In very rare instances, Cushing cases can present with hyperprolactinemia and Cushing synchronously.6 Hormonal secretion and clinical presentation are variable, with the pathology most often attributed to only one component of double pituitary adenoma.3,14 The multiple-hit theory is the most common hypothesis for double pituitary adenoma etiology with coincidental monoclonal expansion of two or more lineages, which present with separate pseudo-capsules for each lesion.15

Observations

On presenting with Cushing disease, the differential diagnosis before the initial operation considered that the known left lateral pituitary adenoma could be a mixed tumor with both prolactin and ACTH lineages. Therefore, it was the initial target of the resection until discovering the second adenoma intraoperatively. With two distinct adenomas, the inferomedial adenoma was presumed to be the source of the ACTH hypersecretion and was subsequently resected. The left lesion was thought to be a prolactinoma and hormonally inactive after historical dopaminergic therapy and thus was not pursued during the initial surgery. However, pathology confirmed that the opposite was true. Few cases have also involved incidental pituitary tumors that look like the hormonally active adenoma and encourage resection of it, leaving the primary pituitary adenoma behind.6,7 It has been reported that these “decoy lesions” can cause surgical failure and require secondary operations.6,7,10,16 Intraoperative localization and confirmation of the adenoma classification may have also been helpful during the case, including tissue-based ACTH antibody assay,9 plasma ACTH measurements with a immunochemiluminometric method,17 or intraoperative ultrasound.5,6

The inferomedial second tumor was not appreciated or reported throughout her serial MRI studies from 2010 to 2020. Interestingly, imaging did demonstrate the left pituitary adenoma that was medically treated as a prolactinoma, although it was later diagnosed as an ACTH-secreting lesion on IHC staining. Preoperative visualization of a pituitary adenoma in Cushing disease is reported to be limited, with a reported 50% incidence with negative MRI with standard 1.5 T.1,18,19 MRI technical refinements in magnet strength, slice thickness, or enhanced spin sequences have increased sensitivity, but one-third of patients with Cushing disease still have negative scans.20 Small prolactinomas, especially those near the cavernous sinus, are also notoriously difficult to visualize on MRI, although recent advances using co-registration of 11C-methionine positron emission tomography–computed tomography with MRI (Met-PET/MRICR) may prove useful.21 Difficulty with preoperative visualization complicates a diagnosis of multiple adenomas, with or without multiple endocrinopathies, and negatively affects surgical planning. In a single-institution retrospective review of MRI in all cases of double pituitary tumors, only one of eight patients (12.5%) over 16 years of age had a positive MRI for double pituitary tumors and was diagnosed preoperatively.2

The patient’s preoperative IPSS demonstrated a right central-to-peripheral gradient. This was incongruent with the MRI demonstrating the single left-sided tumor. While IPSS is useful in confirming Cushing disease, its sensitivity for lateralization has been reported at only 59%–71%.9 With this in mind and a known left-sided adenoma on MRI, exploration of the right side of the pituitary was not originally planned. Ultimately, the left-sided adenoma was the source of ACTH hypersecretion, which remains incongruent with preoperative IPSS. It has been suggested that multiple pituitary adenomas in Cushing disease could further decrease its accuracy.1,6

The patient’s initial historical prolactin levels (33.8 ng/dL) were lower than reported levels of 100–250 ng/dL for microadenoma and >250 ng/dL in cases of macroadenoma. Normally, in active single prolactinoma, prolactin secretion is correlated to size. We do not suspect that the presence of more than one pituitary adenoma would affect the level of prolactin hypersecretion.6 Slight elevations in prolactin can be attributed to causes such as pituitary stalk effect, medications, and physiological stimulation. During the 5 years of bromocriptine therapy, the effect on the inferomedial prolactinoma was unknown, as it was not appreciated on MRI. There are reports of prolactinomas being less responsive to dopaminergic agonist therapy in cases of double adenomas.14,22 Upon resection of the inferomedial prolactinoma during the initial operation, there was no further change in the patient’s prolactin levels, which could most likely be attributed to prior dopaminergic therapy. Unfortunately, the initial endocrine laboratory workup did not include levels of ACTH or cortisol. In addition to hyperprolactinemia, Cushing disease can also present with changes in menstruation. After the secondary resection and removal of the ACTH-secreting pituitary adenoma, the patient’s oligomenorrhea resolved and she achieved pregnancy. Retrospectively, it remains unclear if the prolactinoma was once truly active hormonally.

Lessons

The rare presence of two pituitary adenomas can complicate the diagnosis, medical and surgical management, and long-term outcomes for patients. A complete endocrine workup is essential when a pituitary adenoma is suspected and can help screen for pluri-hormonal and multiple pituitary adenomas. In our patient, it is unknown when the onset of hypercortisolism was with the limited initial hormonal workup.

Currently, localizing and resecting the hormonally active adenoma in double or multiple pituitary adenomas remain a challenge, with limitations in preoperative imaging and intraoperative measures. After encountering the additional inferomedial lesion during surgery, resection of both adenomas during the initial surgery may have been prudent to ensure the resolution of Cushing disease. Although exploration for additional pituitary adenomas is not usually recommended, it could be considered in cases of multiple pituitary adenomas and uncertainty of the culprit of Cushing disease.

The current characterization of pituitary tumors by the World Health Organization includes immunohistochemistry for both transcription factors and pituitary hormones, with clinical usefulness to be determined by future studies. Multiple lineages can occur mixed in a single pituitary adenoma or across different noncontiguous adenomas and can only be determined by TF immunostaining. The left ACTH-staining lesion in our patient had some shrinkage and MRI changes, which may have been a response to dopaminergic therapy. Full characterization of the tumor cell lineages in this case remains undetermined without staining for TFs.

In conclusion, we report a rare case of Cushing disease concurrent with a prolactinoma leading to the need for repeat resection. This is one of the few reported cases of a double pituitary adenoma leading to a lack of biochemical remission of hypercortisolism after the initial surgery. Strategies for localization of the active tumor in double pituitary adenomas are essential for primary surgical success and the resolution of endocrinopathies.

Author Contributions

Conception and design: Zwagerman, Tavakoli, Shah, Findling. Acquisition of data: Zwagerman, Armstrong, Tavakoli, Shah, Ioachimescu, Findling. Analysis and interpretation of data: Zwagerman, Armstrong, Tavakoli, Shah, Coss, Ioachimescu, Findling. Drafting of the article: Zwagerman, Armstrong, Shah. Critically revising the article: Zwagerman, Armstrong, Tavakoli, Shah, Ioachimescu, Findling. Reviewed submitted version of the manuscript: Zwagerman, Armstrong, Tavakoli, Shah, Laing, Ioachimescu, Findling. Approved the final version of the manuscript on behalf of all authors: Zwagerman. Statistical analysis: Armstrong, Shah. Administrative/technical/material support: Zwagerman, Armstrong, Shah. Study supervision: Zwagerman, Tavakoli, Shah, Laing.

References

Filed under: Cushing's, pituitary, Rare Diseases, symptoms, Treatments | Tagged: , , , , , , , | Leave a comment »

The CuPeR Model: A Dynamic Online Tool for Predicting Cushing’s Disease Persistence and Recurrence After Pituitary Surgery

Posted on October 30, 2025 by MaryO

Abstract

Objective

Predicting postoperative persistence and recurrence of Cushing’s disease (CD) remains a clinical challenge, with no universally reliable models available. This study introduces the CuPeR model, an online dynamic nomogram developed to address these gaps by predicting postoperative outcomes in patients with CD undergoing pituitary surgery.

Methods

A retrospective cohort of 211 patients treated for CD between 2010 and 2024 was analyzed. Key patient and tumor characteristics, imaging findings, and treatment details were evaluated. Multivariate logistic regression identified independent predictors of postoperative persistence or recurrence of CD (PoRP-CD), which were then incorporated into the CuPeR model using stepwise selection based on Akaike Information Criterion. Internal validation was performed using a testing dataset, and a user-friendly online nomogram was developed to facilitate immediate, patient-specific risk estimation in clinical practice.

Results

The final predictive model identified four key factors: symptom duration, MRI Hardy’s grade, tumor site, and prior pituitary surgery. Longer symptom duration and a history of prior surgery significantly increased the risk of recurrence, while bilateral tumor location reduced this risk. The model demonstrated an area under the receiver operating characteristic curve (AUC-ROC) of 0.70, with 83% accuracy, specificity of 96%, and sensitivity of 33%.

Conclusions

The CuPeR model may offer a practical tool for predicting PoRP-CD, enhancing preoperative decision-making by providing personalized risk assessments.

Keywords

Cushing’s disease
Transsphenoidal surgery
Nomogram
Recurrence
Disease Persistence

Abbreviations

ACTH

Adrenocorticotropic Hormone

AIC

Akaike Information Criterion

AUC

Area Under the Curve

BMI

Body Mass Index

CD

Cushing’s Disease

CI

Confidence Interval

CRH

Corticotropin-Releasing Hormone

DFS

Disease-Free Survival

DL

Deep Learning

eTSS

Endoscopic Transsphenoidal Surgery

HR

Hazard Ratio

IPSS

Inferior Petrosal Sinus Sampling

ML

Machine Learning

MRI

Magnetic Resonance Imaging

OS

Overall Survival

PoRP-CD

Persistent or Recurrent Cushing’s Disease

SIADH

Syndrome of Inappropriate Antidiuretic Hormone Secretion

TSS

Transsphenoidal Surgery

UFC

Urinary Free Cortisol

Introduction

Cushing’s disease (CD) is a rare endocrine disorder, with an annual incidence rate of approximately 0.24 cases per 100,000 individuals [1]. Transsphenoidal surgery (TSS), performed using either endoscopic or microscopic approaches, remains the cornerstone of treatment for CD. Notably, meta-analytical studies have reported that TSS achieves remission and provides long-term disease control in 71–80 % of patients [[2], [3], [4]]. The remaining cases may experience persistent disease despite surgery, while others may face disease recurrence despite initial remission. In such cases, additional treatment options include second pituitary surgery, pituitary irradiation, targeted medical therapies, and bilateral adrenalectomy, each with varying success rates ranging from 25 % for medical therapy to 100 % for bilateral adrenalectomy [5].
To date, no single predictive factor has proven effective in reliably forecasting treatment outcomes in patients with CD [6]. This underscores the critical need for developing predictive models to assess the likelihood of postoperative recurrence or persistence of Cushing’s disease (PoRP-CD). However, only a limited number of studies have addressed this gap. Notably, two studies from Peking Union Medical College Hospital attempted to tackle this issue using machine learning (ML) and deep learning (DL) approaches [6,7]. These studies utilized demographic, clinical, and paraclinical variables to construct predictive models, with DL approaches showing potential to enhance predictive accuracy [7]. While the results of these models were promising, their applicability in routine clinical practice remains limited. Both studies focused exclusively on patients undergoing their initial transsphenoidal surgery, making them less applicable for cases involving patients with a prior history of pituitary surgery or radiotherapy. Furthermore, these models incorporated both preoperative and postoperative parameters, such as changes in cortisol and adrenocorticotropic hormone (ACTH) levels. However, serum cortisol, ACTH, and comprehensive endocrine testing should be available before any treatment decisions are made, and each patient should ideally be reviewed by a multidisciplinary tumor board, including neurosurgery, radiology, endocrinology, and oncology, prior to pituitary surgery. As such, more comprehensive and practical predictive tool that can support timely clinical decision-making and accommodate a broader range of patient scenarios in the management of CD.
The current study was designed to address these critical limitations and provide a more practical solution for predicting postoperative outcomes in CD. Applying one of the largest available CD cohorts, we incorporated a wide array of patient and tumor characteristics, imaging findings, and treatment details to develop a robust and comprehensive predictive model. This model offers treating surgeons reliable insights into the likelihood of tumor recurrence or persistence. By providing individualized risk predictions, the model is intended to assist clinicians in considering different therapeutic options before pituitary surgery, complementing—but not replacing—standard multidisciplinary decision-making. To further enhance its utility in clinical practice, we also developed an interactive online dynamic nomogram, allowing individualized predictions of postoperative persistence or recurrence.

Methods

Study design, patients, and endpoints

The experimental protocol was approved by the Institutional Review Board of Shahid Beheshti University of Medical Sciences (Tehran, Iran). This retrospective study investigates the clinical outcomes of pituitary surgery in patients with CD underwent pituitary surgery between 2010 and 2024 in the neurosurgery department at Loghman Hakim Hospital. Surgeries were conducted by a group of experienced neurosurgeons under the supervision of the first author (G.S). The primary objective of this study was to develop and validate a predictive model for assessing the risk of PoRP-CD. The secondary objectives were (a) to summarize patient and tumor characteristics; (b) to report surgical outcomes and remission rates following surgery; and (c) to analyze patient survival. This study was performed in accordance with the Declaration of Helsinki, and adheres to the reporting guidelines outlined in the STROBE Statement. Due to retrospective nature of the study informed consent was waived by Shahid Beheshti University of Medical Sciences Ethics Committee. All methods were performed in accordance with the relevant guidelines and regulations.

Preoperative assessments

The “index surgery” was set to the most recent pituitary surgery. Before the index surgery, patients underwent comprehensive clinical evaluations, including biochemical and neurological assessments as well as visual field examinations. This research utilized the Endocrine Society Clinical Practice Guideline to establish the diagnosis of CD [8]. Three main steps were involved in the diagnostic process: in the first step, the focus was on detecting hypercortisolemia, which was determined by examining 24-hour urinary free cortisol levels (normal: <60 mcg/24 h), as well as plasma and salivary cortisol profiles. Low-dose dexamethasone suppression testing was performed using the 2 mg/48 h protocol, which was the standard practice in our institution during the study period (2010 onward) [8]. The second step aimed to confirm ACTH-dependent cause of hypercortisolemia, through measuring plasma ACTH levels. The final step aimed to distinguish Cushing’s disease from ectopic sources of ACTH. This was performed using a high-dose dexamethasone suppression test (8  mg overnight), with a plasma cortisol suppression exceeding 50 % typically considered indicative of a pituitary origin [9].
Next, the patients were subjected to thin-slice (3 mm) 1.5-tesla dynamic pituitary gland magnetic resonance imaging (MRI) with gadolinium contrast. The MR evaluation adhered to a strict protocol, requiring an independent agreement of treating neurosurgeon and radiologist to confirm the diagnosis. MR scans were categorized according to the Hardy and Knosp classifications [10]. Normal scans required to demonstrate the absence of direct signs, including inhomogeneity in the pituitary, as well as indirect signs such as a deviation of the pituitary stalk, bulging or erosion of the Sella contour. In cases where the CD was confirmed but pituitary MRI was inconclusive, bilateral inferior petrosal sinus sampling (IPSS) was performed per standard protocol under corticotropin-releasing hormone (CRH) stimulation [11]. Patients with macroadenoma or signs of elevating the optic chiasm were candidates for Humphrey visual field examination.

Surgical approach

Patients underwent endoscopic transsphenoidal approach using conventional “Two Nostrils–Four Hands” technique [12]. Given the diminutive size and deep-seated location of most adenomas, locating the adenoma emerged as a formidable challenge, particularly when the tumor remained not visualized in pre-operative imaging studies. The surgical procedure entailed extensive drilling of the Sellar floor laterally up to the carotid artery on both sides, providing a comprehensive view of the medial wall of the cavernous sinus and exposure of the anterior and posterior intercavernous sinuses. The exploration of the entire Sella commenced in the region where the original tumor had been localized. Upon identification of a tumor, a selective adenomectomy was performed, accompanied by a thorough inspection of the pituitary gland to detect and eliminate any potential tumor remnants. The removal of any pseudo capsule was executed meticulously.
The primary surgical objective was selective adenomectomy, with further exploration guided by the side recommended by IPSS in cases where no adenoma was initially observed. The exploration involved making a plus-like incision on the corresponding half of the gland, enabling deep exploration to leave no part unexplored. In instances where creamy material suggestive of a tumor was drained after a pituitary incision, a tissue biopsy was obtained, although it was not conclusively considered a tumor. Exploration continued on the opposite side in such cases.
When no distinct adenoma was found, a peri-glandular inspection was conducted to visualize the medial wall of the cavernous sinus, diaphragm, and Sellar floor, aiming to detect an ectopic microadenoma. If an apparent tumor remained undetected, the procedure was repeated on the contralateral side, and a vertical medial incision on the pituitary gland adjacent to the pituitary stalk and neurohypophysis was made as a final effort for tumor detection. In the absence of identified pathology during the surgical procedure, hemi-hypophysectomy was considered on the side where IPSS had detected the gradient or on the side with an apparent or suspicious MRI finding. Considering the typical central location of corticotroph cells in the pituitary gland, microadenoma exploration extended posteriorly and medially to confirm extirpation.

Postoperative assessments

In this study, the patients were closely monitored for signs of diabetes insipidus and syndrome of inappropriate antidiuretic hormone secretion (SIADH). Serum sodium levels, urine-specific gravity, and volume were checked regularly. Following surgery, morning cortisol levels were measured on the first day, and other anterior pituitary hormones were evaluated on day 3. Hydrocortisone therapy was initiated based on the patient’s symptoms, signs of adrenal insufficiency, and low cortisol levels. The first postoperative check-up occurred two weeks after surgery, followed by another at three months, which included a comprehensive assessment of pituitary hormones. This evaluation was repeated every three months for two years and then annually. Additionally, patients underwent a dynamic 1.5-Tesla pituitary MRI at six months post-surgery and annually thereafter, with a minimum follow-up period of 12 months.
Remission was defined as having low cortisol levels, indicated by early morning serum cortisol level ≤ 5 μg/dL within two days post-surgery [13]. Persistent CD was characterized by ongoing hypercortisolism, and postoperative recurrence refers to the reappearance of CD symptoms despite initial remission marked by hypercortisolemia. In case of persistence or recurrence, patients were candidates for second-line treatment options selected by their physicians, including revision surgery, targeted medical therapy, pituitary radiotherapy, or bilateral adrenalectomy. Disease-free survival (DFS) was defined as the time from the index surgery to the first occurrence of disease recurrence or death from any cause, while overall survival (OS) was defined as the time from the index surgery to death from any cause.

Statistical analysis

Categorical variables were expressed as numbers and percentages, and continuous variables as mean, range, and standard deviation. The distribution of variables was checked using the Shapiro-Wilk test, which showed a deviation from normal distribution. Contingency tables were used for categorical variables with Pearson’s Chi-squared or Fisher’s Exact test used to examine their association with outcomes for univariate analyses. For continuous variables, the unpaired t-test was applied to compare means between two independent groups when the data met the assumption of normality. Analyses were conducted with R Statistical Software v4.4.0 (“Puppy Cup”). All statistical inferences were two-sided, and P < 0.05 were considered statistically significant.

Model development and internal validation

The dataset was split by “caret package” into a training set (70 %) and a testing set (30 %) using stratified sampling to ensure representative proportions of outcomes. Binary logistic regression was used to identify predictors of PoRP-CD. Patients with adequate follow-up data were included in the analysis. The variables with a marginal level of association (P < 0.15) in the univariate analysis were further included in the multivariate logistic regression analysis to identify the independent predictors of PoRP-CD. Imported factors included demographic, medical history, imaging and pathology results, and treatment details. To identify predictors of PoRP-CD, a multivariable logistic regression model was developed using stepwise selection based on Akaike Information Criterion (AIC). Model performance, including sensitivity, specificity, and area under the receiver operating characteristic curve (AUC-ROC), was evaluated using internal validation on the test dataset.

Nomogram creation and deployment

A nomogram was constructed using the validated logistic regression model. The nomogram was then integrated into a web-based application using the “Shiny package” in R program. The dynamic nomogram allows clinicians to input patient data and obtain individualized risk predictions for PoRP-CD.

Survival analysis

Survival analysis was conducted to evaluate DFS across various patient subgroups. The log-rank test was applied to assess statistical differences in survival distributions between subgroups. Cox proportional hazard regression was used to estimate hazard ratios (HR) and 95 % confidence intervals (CI). The “survival” and “survminer” R packages were applied in this section.

Results

Patients and tumors characteristics

A total of 211 patients with CD had been treated by a group of experienced neurosurgeons under the supervision of the first author (G.S) between March 2010 and January 2024 in the neurosurgery department at Loghman Hakim Hospital. Table 1 summarizes the baseline characteristics of patients at the timepoint of index surgery. The patients had a mean age of 35.9 ± 12.1 years (range: 11–67), among which 21 patients (9.9 %) were in the pediatric age range, and 165 (78.1 %) were female. Obesity was the most common patients’ symptoms (45.9 %), and physical examination reported centripetal obesity (84.3 %), moon face (75.8 %), and striae (64.4 %) as the most common clinical manifestations. Compared to the adult patients, pediatrics had less common hypertension on physical examination (35.2 vs. 5.9 %) and medical history of diabetes mellitus (36.8 vs. 4.7 %) (P < 0.05). The majority of patients (63.9 %, 135/211) had not received any prior treatment. Among those who had, surgery alone was the most common approach (n = 57, 27.0 %), performed once in 50 patients (23.6 %), twice in 6 patients (2.8 %), and three times in a single patient.

Table 1. Baseline characteristics of adult and pediatric patients with Cushing’s disease.

Demographics Total
n = 211		 Adults
n = 190		 Pediatrics
n = 21		 P Medical Hx Total
n = 211		 Adults
n = 190		 Ped.
n = 21		 P Drug-Family Hx Total
n = 211		 Adults
n = 190		 Ped.
n = 21		 P
Age; mean-SD (y) 35.9–12.1 38.3–10.2 14.8–1.7 0<.001 Hypertension 97 (45.9) 92 (48.4) 5 (23.8) 0.31 Cabergoline 3 (1.4) 3 (1.5) 0 1.0
Sex; female 165 (78.1) a 149 (78.4) 16 (76.1) 0.78 Diabetes mellitus 71 (33.6) 70 (36.8) 1 (4.7) 0<.001 Ketoconazole 12 (5.6) 12 (6.3) 0 0.61
Marital status; married 105 (70.0) b 103 (76.8) b 2 (12.5) b 0<.001 Dyslipidemia 45 (21.3) 42 (22.1) 3 (14.2) 0.56 Metyrapone 0 − − −
Smoking status; active–passive-non 17 (10)-27(17)-113(72) b 17 (11)–23(15)-101(70) b 0–4(25)-12(75) b 0.70 Prior pituitary surgery 57 (27.0)) 51 (26.8) 6 (28.5) 1.0 Pasireotide 0 − − −
Height; mean-SD (cm) 163.9–8.7 163.8–8.9 165.1–6.6 0.59 Fatty liver 37 (17.5) 32 (16.8) 5 (23.8) 0.36 Somatostatin 0 − − −
Weight; mean-SD (Kg) 74.1–22.5 74.6–22.5 69.3–23.1 0.58 Thromboembolism 6 (2.8) 6 (3.1) 0 1.0
BMI; mean-SD (Kg/m2) 28.8–6.1 29.0–6.2 27.6–5.5 0.72 DVT 3 (1.4) 3 (1.5) 0 1.0 FH of Cushing 5 (2.3) 4 (2.1) 1 (4.7) 0.43
Symptom duration; mean-SD (m) 30.7–41.2 32.0–43.2 20.0–14.2 0.78 MEN 1 (0.4) 1 (0.5) 0 1.0 FH of MEN 1 (0.4) 1 (0.5) 0 1.0
Presenting Symptoms
Obesity 75 (45.9) b 66 (45.2) b 9 (52.9) b 0.61 Striae 10 (6.1) b 8 (5.4) b 2 (11.7) b 0.27 Headache 4 (2.4) b 3 (2.0) b 1 (5.8) b 0.35
Menstrual disorders 16 (9.8) b 13 (8.9) b 3 (17.6) b 0.22 Edema 7 (4.2) b 7 (4.7) b 0 1.0 Diabetes mellitus 3 (1.8) b 3 (2.0) b 0 1.0
Hypertension 12 (7.3) b 12 (8.2) b 0 0.61 Muscular weakness 7 (4.2) b 6 (4.1) b 1 (5.8) b 0.54 Bone fracture 3 (1.8) b 3 (2.0) b 0 1.0
Blurred vision 10 (6.1) b 9 (6.1) b 1 (5.8) b 1.0 Moon face 6 (3.6) b 6 (4.1) b 0 1.0 Other 10 (6.1) b 10 (6.8) b 0 0.60
Clinical Manifestations
Acanthosis nigricans 35 (16.5) 34 (17.8) 1 (4.7) 0.12 Easy bruising 103 (48.8) 92 (48.4) 11 (52.3) 0.91 Male pat. hair loss 111 (52.6) 100 (52.6) 11 (52.3) 1.0
Acne 68 (32.2) 58 (30.5) 10 (47.6) 0.16 Ecchymosis 58 (27.5) 50 (26.3) 8 (38.0) 0.37 dysmenorrhea 96 (45.4) 84 (44.2) 12 (57.1) 0.49
Ankle edema 105 (49.7) 96 (50.5) 9 (42.8) 0.57 Exophthalmia 50 (23.7) 47 (24.7) 3 (14.2) 0.27 Moon face 160 (75.8) 141 (74.2) 19 (90.4) 0.69
Backache 66 (31.2) 60 (31.5) 6 (28.5) 0.88 Facial plethora 97 (45.9) 85 (44.7) 12 (57.1) 0.33 Osteoporosis 25 (11.8) 25 (13.1) 0 0.14
Blurred vision 70 (33.2) 67 (35.2) 3 (14.2) 0.27 Fatigue 146 (69.2) 130 (68) 16 (76.1) 0.76 Prox. myopathy 94 (44.5) 86 (45.2) 8 (38.0) 0.63
Buffalo hump 123 (58.3) 107 (56.3) 16 (76.1) 0.43 Fracture 12 (5.6) 12 (6.3) 0 0.61 Skin atrophy 81 (38.4) 73 (38.4) 8 (38.0) 1.0
Centripetal obesity 178 (84.3) 159 (83.6) 19 (90.4) 0.50 Headache 109 (51.6) 97 (51.0) 12 (57.1) 1.0 Striae 136 (64.4) 119 (62.6) 17 (80.9) 0.55
Cerebrospinal fluid leakage 5 (2.3) 4 (2.1) 1 (4.7) 0.41 Hirsutism 104 (49.3) 92 (48.4) 12 (57.1) 0.72 Supraclav. fat pad 38 (18.0) 33 (17.3) 5 (23.8) 0.67
Cranial nerve palsy 3 (1.4) 3 (1.5) 0 1.0 Hyperpigmentation 38 (18.0) 37 (19.4) 1 (4.7) 0.12 Visual field defect 24 (11.3) 22 (11.5) 2 (9.5) 1.0
Diplopia 18 (8.5) 15 (7.8) 3 (14.2) 0.41 Hypertension 69 (32.7) 67 (35.2) 2 (9.5) 0.009 Weight gain 108 (51.1) 95 (50.0) 13 (61.9) 0.39
Prior Treatments
Treatment naïve 135 (63.9) 122 (64.2) 13 (61.9) 1.0 Pituitary surgery alone 39 (18.4) 33 (17.3) 6 (28.5) 0.23 Radiotherapy alone 6 (2.8) 5 (2.6) 1 (4.7) 0.47
Medication alone 5 (2.3) 5 (2.6) 0 1.0 Combination therapy 17 (8.1) 17 (8.9) 0 0.22 Adrenalectomy alone 11 (5.2) 10 (5.2) 1 (4.7) 1.0
Hormonal Assessments
Hypothyroidism 24 (31.1) b 24 (31.1) b 0 0.09 GH deficiency 6 (8.8) b 6 (8.8) b 0 1.0 Hypogonadism 7 (9.8) b 7 (9.8) b 0 1.0
Panhypopituitarism 2 (2.5) b 2 (2.5) b 0 1.0
Imaging Features
Hardy’s grading
(sphenoid bone invasion)
0
1
2
3
4		 37 (21.1) b
102 (58.2) b
27 (15.4) b
4 (2.2) b
5 (2.8) b		 35 (22.7) b
88 (57.1) b
23 (14.9) b
3 (1.9) b
5 (3.2) b		 2 (9.5)
14 (66.7)
4 (19.0)
1 (4.7)
0		 0.45 Hardy’s staging
(suprasellar extension)
A
B
C
D
E		 36 (20.4) b
86 (48.8) b
14 (7.9) b
4 (2.2) b
36 (20.4) b		 34 (21.9) b
73 (47.1) b
14 (9.0) b
2 (1.2) b
32 (20.6) b		 2 (9.5)
13 (62)
0
2 (9.5)
4 (19.0)		 0.07 Knosp grading

0
1
2
3
4

 152 (82.6) b
13 (7.0) b
7 (3.8) b
4 (2.1) b
8 (4.3) b		 135 (82.8) b
10 (6.1) b
6 (3.6) b
4 (2.4) b
8 (4.9) b		 17 (80.9)
3 (14.2)
1 (4.7)
0
0		 0.46
Tumor size
Microadenoma
Macroadenoma
MR-negative		 122 (58.6) b
50 (24.0) b
36 17.3) b		 111 (59.3) b
42 (22.4) b
34 (18.1) b		 11 (52.3)
8 (38.0)
2 (9.5)		 0.28 Sphenoid shape
Sellar
Presellar
Conchal		 205 (97.6) b
3 (1.4) b
2 (0.9) b		 184 (97.3) b
3 (1.5) b
2 (1.0) b		 21 (100)
0
0		 1.0 Multifocality
Unifocal
Multifocal		 113 (80.1)
28 (19.8)		 97 (79.5)
25 (20.4)		 16(84.2)
3 (15.7)		 0.79
Invasion c
No invasion
Cavernous sinus
Carotid
Dura
Clivus		 185 (88.5) b
12 (5.7) b
3 (1.4) b
6 (2.8) b
3 (1.4) b		 165 (87.7) b
11 (5.8) b
3 (1.5) b
6 (3.1) b
3 (1.5) b		 20 (95.2) b
1 (4.8) b
0
0
0		 1.0 Tumor site
Right lobe
Left lobe
Bilateral
Central
Stalk		 22 (15.6)) b
16 (11.3)) b
51 (36.1) b
49 (34.7) b
3 (2.1) b		 20 (16.2) b
13 (10.5) b
43 (34.9) b
45 (36.5) b
2 (1.6) b		 2 (11.1) b
3 (16.6) b
8 (44.4) b
4 (22.2) b
1 (5.5) b		 0.38 Empty sella
No
Yes		 207 (98.1)
4 (1.8)		 187 (98.4)
3 (1.5)		 20(95.2)
1 (4.7)		 0.34
Pituitary apoplexy
No
Yes		 185 (97.3) b
5 (2.6) b		 167 (98.2) b
3 (1.7) b		 18 (90.0) b
2 (10.0) b		 0.08 Kissing carotids
No
Yes		 209 (99.0)
2 (0.9)		 188 (98.9)
2 (1.0)		 21 (100)
0		 1.0
a
the numbers in parentheses represent the percentage for each patient group.
b
percentage after ruling out missing data.
c
one patient had invasion to cavernous sinus and carotid and another one had clivus and dural invasion.
A comprehensive preoperative hormonal assessment was conducted on 77 patients (36.4 %), revealing hormonal dysregulation in 28 patients (36.3 %). Hypothyroidism was the most common abnormality, affecting 35 % of those assessed (24 out of 77). On MRI scans, most tumors were microadenomas (58.6 %), with fewer macroadenomas (24.0 %) and some cases with no detectable tumor (17.3 %). Tumors were commonly localized bilaterally (36.1 %) or centrally (34.7 %), and most were unifocal (80.1 %). Knosp grading indicated no cavernous sinus invasion in the majority (82.6 %), with only 6.4 % showing grades 3–4. According to Hardy’s grading, most patients had mild sphenoid bone invasion, predominantly grade 1 (58.2 %). For Hardy’s staging of suprasellar extension, nearly half were at stage B (48.8 %), with smaller groups in stages A and E (20.4 % each), and fewer in stages C and D. Other MRI findings are summarized in Table 1. There was no significant difference between adult and pediatric patients in terms of hormonal and imaging findings (P > 0.05). Pathology reports were available for 36 patients. The most common finding was sparse cellularity, observed in 11 patients (30.6 %) followed by dense cellularity identified in 9 patients (25 %). Crooke cell changes were the least common, present in 7 patients (19.4 %). Nine specimens (25 %) had no tumor identified in the sample submitted to pathology.

Treatment details and outcomes

A total of 36 patients (17.1 %) underwent preoperative IPSS, among which 13 had right lateralization, 13 left, 4 bilateral, 3 central, 2 central-right, and 1 central-left. Pituitary surgery was predominantly performed using the endoscopic transsphenoidal (eTSS) approach (98.5 %, 208/211), while the transplanum approach was used in 3 patients (1.5 %). Adenomectomy was the most common surgical procedure (n = 187, 88.6 %), followed by total hypophysectomy in 17 patients (8.1 %) and hemi-hypophysectomy in 7 patients (3.3 %). In addition, four patients in the total hypophysectomy group and one patient in the adenomectomy group also underwent hypophyseal stalk resection. Information on disease persistence or recurrence was available for 204 patients. Median follow-up of patients was 58.4 months (range: 4.5–170.4 months) after index surgery. In total, 23 patients (11.2 %) experienced persistent disease following the index surgery, while 10 patients (4.9 %) had disease recurrence, with a median time to recurrence of 7 months (range: 1–78 months). The median recurrence-free interval for the entire cohort was 37 months.
The surgical complication rates were as follows (Fig. 1A): cerebrospinal fluid leaks were observed in 22 patients (10.4 %), followed by cranial nerve injury in 7 patients (3.3 %) and meningitis in 5 patients (2.3 %). Carotid injury and intracerebral bleeding each occurred in 3 patients (1.4 %). Nasal bleeding, the need for a ventriculoperitoneal shunt, and embolic events were each reported in 1 patient (0.4 %). Perioperative mortality was observed in one female patient (0.4 %) due to an iatrogenic carotid injury. This patient had previously undergone three pituitary surgeries and received radiotherapy at the pituitary site. Hormonal dysregulation following surgery included hypothyroidism in 99 patients (46.9 %), diabetes insipidus in 76 patients (36 %), hypogonadism in 28 patients (13.2 %), growth hormone deficiency in 10 patients (4.7 %), and panhypopituitarism in 7 patients (3.3 %) (Fig. 1B).

  1. Download: Download high-res image (108KB)
  2. Download: Download full-size image

Fig. 1. Rates of surgical complications. (a) Intraoperative complications; (b) hormonal dysregulation rates following surgery.

Multivariate analysis on the predictors of Persistent/Recurrent Cushing’s disease

To identify potential predictive factors for PoRP-CD, we conducted a comprehensive binary logistic regression analysis, examining key clinical and imaging variables (Table 2). In the univariate analysis, factors including symptom duration (OR [odds ratio] 1.01, 95 % CI [confidence interval] 1.00–1.02, P = 0.04), MRI Hardy’s grade (OR 1.62, 95 % CI 0.98–2.69, P = 0.05), and previous pituitary surgery (OR 3.56, 95 % CI 1.39–9.07, P = 0.007) demonstrated significant association with PoRP-CD. MR-reported tumor size showed increased odds of recurrence with an increased tumor size (OR for microadenoma vs. no tumor: 2.41, 95 % CI: 0.50–11.53; OR for macroadenoma vs. no tumor: 4.15, 95 % CI 0.80–21.42), though the effect was not statistically significant (P > 0.05). To impede missing the marginal significant factors, three factors with P values between 0.05 and 0.15 were also included in the multivariate analysis, including “MRI Knosp grading”, “MR-reported tumor site”, and “previous pituitary radiotherapy”. In the multivariate analysis, “symptom duration” was positively correlated with recurrence, with an odds ratio (OR) of 1.03 (95 % CI: 1.01–1.06, P = 0.01), indicating a higher risk of recurrence with prolonged symptoms. Additionally, a history of “previous pituitary surgery” was significantly associated with recurrence, with an OR of 4.67 (95 % CI: 1.04–20.89, P = 0.04). Other factors, including tumor grading, tumor site, and previous radiotherapy, did not reach statistical significance.

Table 2. Regression analysis of patient and tumor’s factors related to postoperative persistence or recurrence in Cushing disease.

Parameters Univariate Analysis Multivariate Analysis
OR (95 % CI) P OR (95 % CI) P
Age 0.97 (0.94–1.01) 0.23
Sex (male vs. female) 1.17 (0.39–3.50) 0.77
Smoking (active smoker vs. non) 0.78 (0.65–10.28) 0.77
Family history of CD (positive vs. negative) 0.01 (0–Inf) 0.99
Family history of MEN (positive vs. negative) 0.01 (0–Inf) 0.99
Preoperative BMI 1.03 (0.94–1.13) 0.43
Symptom duration 1.01 (1.00–1.02) 0.04 ** 1.03 (1.01–1.06) 0.01 **
Preop serum ACTH (high vs. normal) 0.88 (0.13–6.00) 0.90
Preop free serum cortisol (high vs. normal) 1.18 (0.40–3.45) 0.74
Preop urine free cortisol (high vs. normal) 0.15 (0.01–2.98) 0.21
Knosp grading (ref: grade 0) 1.41 (0.93–2.15) 0.10 * 1.56 (0.61–3.97) 0.34
Hardy’s grading (ref: grade 0) 1.62 (0.98–2.69) 0.05 ** 1.98 (0.54–7.21) 0.29
Hardy’s staging (ref: stage A) 2.97 (0.61–14.38) 0.17
Tumor size
Macro vs. no tumor
Micro vs. no tumor		 4.15 (0.80–21.42)
2.41 (0.50–11.53)		 0.17
Multifocality (multifocal vs. unifocal) 1.68 (0.44–6.42) 0.44
MR-based tumor sitea
Bilateral vs. central
Left vs. central
Right vs. central
Stalk vs. central		 0.16 (0.01–1.53)
0.82 (0.18–4.40)
0.49 (0.09–2.82)
5.33 (0.37–144.16)		 0.14 * 0.34 (0.02–3.95)
0.23 (0.01–3.12)
5.36 (0.19–146.38)

  • (0.0002–0.67)
 0.03 **
0.39
0.27
0.31
Invasion (pos. vs. neg.) 1.18 (0.31–4.51) 0.80
Surgical approach (transplanum vs. eTSS) 6.21 (0.37–103.55) 0.20
Surgical type (adenomectomy vs. hypophysectomy) 1.55 (0.46–5.22) 0.47
Histopathology
Dense type vs. Crooke’s cell adenoma
Normal appearing vs. Crooke’s cell adenoma
Sparse type vs. Crooke’s cell adenoma		 2.00 (0.09–69.06)
0.80 (0.04–23.23)
0.28 (0.01–9.45)		 0.56
Ki-67 (>3% vs. ≤ 3 %) 1.34 (0.14–12.64) 0.79
Previous pituitary surgery (yes vs. no) 3.56 (1.39–9.07) 0.007 ** 4.67 (1.04–20.89) 0.04 **
Previous pituitary radiotherapy (yes vs. no) 3.36 (0.89–12.62) 0.07 * 3.63 (0.28–46.07) 0.31
Postop decrease in BMI 0.90 (0.73–1.03) 0.22
Abbreviations: ACTH − Adrenocorticotropic Hormone; BMI − Body Mass Index; CD − Cushing’s Disease; CI − Confidence Interval; eTSS − Endoscopic Transsphenoidal Surgery; Inf − Infinity; MEN − Multiple Endocrine Neoplasia; MR − Magnetic Resonance; OR − Odds Ratio; PoRP-CD − Persistent or Recurrent Cushing’s Disease; Preop − Preoperative; Postop − Postoperative.
aMR-reported.
* Significant at the level of 0.15.
** Significant at the level of 0.05.
The stepwise selection–in both forward and backward directions–retained four predictors— symptom duration, Hardy’s grading, tumor site, and prior surgery —for the final model. The final multivariate model with four predictors of “symptom duration”, “MRI Hardy’s grading”, “tumor site”, and “previous pituitary surgery” demonstrated significant associations for “symptom duration” (OR 1.03, 95 % CI 1.005–1.05, P = 0.02), previous pituitary surgery (OR 4.61, 95 % CI 1.12–22.0, P = 0.03), and a certain tumor site; tumors located bilaterally had significantly lower odds of recurrence compared to central tumors (OR 0.01, 95 % CI 0.0002–0.45, P = 0.02). On the testing dataset, the four-factor model achieved an AUC of 0.70, specificity of 96 %, and sensitivity of 33 %. The model’s accuracy in predicting PoRP-CD is 83 %.

Predicting persistent or recurrent Cushing’s disease–The CuPeR nomogram

A nomogram was developed based on the multivariate model comprising four key predictors: “Symptom duration”, “MRI Hardy’s grading”, “Previous pituitary surgery”, and “MRI-reported tumor site” (Fig. 2). This nomogram visually represents the impact of each predictor on the likelihood of PoRP-CD. The total score derived from the nomogram aligns with the probability scales, allowing for estimation of the risk of PoRP-CD. Higher cumulative points correspond to an increased likelihood of persistent or recurrent disease. To facilitate individualized predictions of postoperative persistence or recurrence, we developed an online dynamic nomogram (link: https://cushing.shinyapps.io/cuper/).

  1. Download: Download high-res image (97KB)
  2. Download: Download full-size image

Fig. 2. Nomogram for predicting postoperative persistence or recurrence of Cushing’s disease (PoRP-CD). This nomogram visually represents the predictive model for assessing the risk of recurrence or persistence of Cushing’s disease following surgery. Each predictor variable—Symptom duration (months), Knosp grading, Hardy’s grading, previous pituitary surgery, and tumor site— contributes a point value that aligns with the “Linear Predictor” scale, which maps to the “Probability of Persistence” scale, allowing estimation of recurrence likelihood.

Survival analysis

Survival analysis demonstrated a steady, gradual decline in DFS across the entire cohort, with the median DFS not reached despite substantial follow-up (Fig. 3A). Among the predefined variables, Hardy’s Grade 3 was associated with a significantly worse DFS compared with Grade 0 (HR = 6.02, 95 % CI: 1.09–33.02, P = 0.03) (Fig. 3B), whereas other Hardy’s Grades did not reach statistical significance (P > 0.05). Regarding tumor site, no site was a statistically significant risk factor for DFS; stalk tumors showed a trend toward poorer DFS but did not reach significance (HR = 5.09, 95 % CI: 0.84–30.63, P = 0.07) (Fig. 3C). Patients with a history of previous pituitary surgery had significantly worse DFS (HR = 4.72, 95 % CI: 2.29–9.75, P < 0.01) (Fig. 3D). In contrast, symptom duration was not associated with poor DFS (HR = 1.26, 95 % CI: 0.56–2.81, P = 0.57) (Fig. 3E). A similar analysis on OS was not performed, as only five events were recorded among the 211 patients (2.36 %), rendering meaningful statistical analysis infeasible.

  1. Download: Download high-res image (250KB)
  2. Download: Download full-size image

Fig. 3. Disease-free survival (DFS) analysis. (A) Kaplan-Meier curve of DFS for the entire cohort, showing a gradual decline over time; (B) DFS stratified by Hardy’s Grade, demonstrating significant impact of grade 3 on survival outcomes (P = 0.03); (C) DFS by tumor site, highlighting no significant association between tumor site and survival care (P > 0.05); (D) DFS based on previous surgery status, indicating a higher risk of recurrence or death in patients with prior surgical interventions (P < 0.01); (E) DFS by symptom duration, highlighting no significant association (P = 0.57).

Discussion

In this large cohort study, we developed the CuPeR model, a comprehensive predictive tool for PoRP-CD, by analyzing diverse patient and tumor characteristics, imaging findings, and treatment details. This model identified four key predictors—symptom duration, MRI Hardy’s grade, tumor site, and previous pituitary surgery. Multivariate analysis revealed that longer symptom duration and a history of prior surgery significantly increased recurrence risk, while bilateral tumor location was associated with a reduced risk. Validated with an AUC of 0.70 and 83 % accuracy on the testing dataset, the model offers significant clinical utility by providing treating surgeons with valuable insights into postoperative outcomes.
This study is among the few to develop a predictive model for estimating PoRP-CD (Table 3). Previous efforts, such as those by Liu et al. [6] and Fan et al. [7], employed machine learning and deep learning methodologies, respectively, demonstrating promising results (AUCs of 0.78 and 0.86). However, both studies were limited in their applicability to many clinical settings, as they focused solely on patients undergoing initial surgeries and incorporated postoperative parameters, which are unavailable for preoperative decision-making. By addressing these gaps, our study contributes a more practical tool for use in diverse clinical scenarios. Moreover, the findings of this study align with predictors identified in prior research. For instance, factors such duration of symptoms and history of previous pituitary surgery have been highlighted as critical for recurrence [6,14]. Importantly, our inclusion of MRI-based predictors and preoperative variables ensures the model’s relevance during preoperative planning, distinguishing it from previous approaches.

Table 3. Studies on predictive models or patients and tumors predictive factors of post-operative remission of Cushing’s disease.

Empty Cell Year Country Study Size Methods Main Findings Ref.
Predictive Models
Comprising 8 factors:
age,
disease coarse,
morning serum ACTH (preop),
morning serum cortisol (preop),
urine free cortisol (preop),
morning serum ACTH nadir (postop),
morning serum cortisol nadir (postop),
urine free cortisol nadir (postop)		 2019 China 354 Machine-learning using Random Forest algorithm Sensitivity 87 %, specificity 58 %
AUC 0.78		 [6]
Comprising 5 factors:
age,
disease coarse,
morning serum ACTH (postop),
morning serum cortisol nadir (postop),
urine free cortisol nadir (postop)		 2021 China 354 Deep-learning using factorization‑machine based neural approach AUC 0.86 [7]
Predictive Factors
Serum cortisol < 35 nmol/L (6–12 w after surgery) 1993 UK 11 Prospective Favorable long-term remission rate [15]
Serum 11-deoxycortisol > 150 nmol/L after metyrapone test at 14 days post-surgery 1997 Netherlands 29 Retrospective Higher risk of recurrence
Sensitivity 100 %, specificity 75 %		 [16]
Serum cortisol < 2 μ/dL (3–8 d after surgery) 2001 Japan 49 Retrospective Recurrent disease in 4 % of patients [17]
MRI-based tumor size and cavernous sinus invasion 2003 Italy 26 Retrospective Unfavorable factors of persistent disease [18]
No histological evidence of adenoma 2007 US 490 Retrospective Lower remission rate [19]
Long-term hypocortisolism after surgery (≥13 m) 2017 India 230 Retrospective Favorable for remission
Sensitivity 46 %, specificity 100 %		 [20]
Greater decrease in BMI after surgery
Lower DHEAS before surgery		 2017 Taiwan 41 Retrospective Favorable factors for higher remission [21]
High serum ACTH/cortisol ratio before surgery 2018 Turkey 119 Retrospective Risk factor for disease recurrence [22]
USP8 mutation 2018 Germany 48 Retrospective Higher recurrence rate [23]
Serum cortisol < 107 nmol/L after betamethasone suppression test following surgery 2018 Sweden 28 Interventional Sensitivity 85 %, specificity 94 %
AUC 0.92		 [24]
Tumor visualization on MRI before surgery 2022 Spain 40 Retrospective Favorable factor for remission [25]
Abbreviations: ACTH − Adrenocorticotropic Hormone; AUC − Area Under the Curve; BMI − Body Mass Index; DHEAS − Dehydroepiandrosterone Sulfate; MRI − Magnetic Resonance Imaging; PoRP-CD − Persistent or Recurrent Cushing’s Disease; Preop − Preoperative; Postop − Postoperative; USP8 − Ubiquitin Specific Peptidase 8.
Several other studies aimed to explore the predictive value of single predictors. Braun et al. (2020) summarized the predictors for CD remission following TSS in a systematic review. Key predictors include pre-surgical identification of the tumor via MRI and the absence of adenoma invasion into the cavernous sinus. Postoperative hormonal levels, particularly low cortisol (< 2 µg/dL) and ACTH levels (< 3.3 pmol/L) as well as low cortisol levels (< 35 nmol/L) at 6–12 weeks post-surgery and sustained hypocortisolism requiring long-term replacement therapy, were significant indicators of remission. Additionally, post-surgical decreases in BMI contributed to favorable outcomes. Other reported predictors included a high level of surgical expertise, younger patient age, non-mutant USP8 corticotroph tumors, and swift recovery from postoperative adrenal insufficiency [5].
This study has certain limitations that should be acknowledged. The reliance on retrospective data may result in potential biases in variable selection and data completeness. While the model demonstrated good predictive accuracy, its limited sensitivity may restrict its ability to identify all high-risk patients. Moreover, the model has not been externally validated in independent cohorts, which limits its generalizability to other clinical settings. Despite these limitations, the study possesses significant strengths that underscore its contribution to the field. Applying one of the largest CD cohorts, it provides a robust statistical foundation and enhances the reliability of the findings. The comprehensive inclusion of diverse patient and tumor characteristics, imaging findings, and treatment details resulted in a clinically relevant and well-rounded predictive model. Notably, this model stands out for its applicability to a broader spectrum of patients, including those with prior surgeries or radiotherapy, addressing a gap left by earlier studies. Furthermore, the development of an online dynamic nomogram bridges the gap between research and clinical practice, allowing personalized predictions and aiding surgeons in making informed decisions before pituitary surgery.
Although this study incorporated long-term follow-up (median 58 months) to define persistence and recurrence and to internally validate the model, external validation in prospective, multi-institutional cohorts remains essential to confirm its broader applicability. Although the CuPeR model incorporates a wide array of clinical, radiological, biochemical, and demographic variables, other potential prognostic factors were not included and may warrant consideration in future studies. For instance, the presence of osteoporosis, degree of tumor invasion, and early recovery of the adrenal axis during the postoperative period have all been reported as relevant predictors of outcomes in Cushing’s disease [26]. Moreover, the role of surgical expertise is critical, as higher surgeon and institutional experience are strongly associated with improved remission and lower recurrence rates [27]. Incorporating novel parameters, such as genetic markers or advanced imaging techniques, could further enhance the predictive accuracy and clinical utility of the model. Prospective implementation of the nomogram in routine clinical workflows will provide valuable insights into its performance and its potential to improve patient outcomes.

Conclusions

This study introduced a practical, predictive model for estimating the risk of postoperative persistence and recurrence in Cushing’s disease, possibly offering a reliable tool for preoperative planning. By integrating key clinical predictors into an interactive online dynamic nomogram, the CuPeR model may provide surgeons with personalized risk assessments to aid in preoperative planning. Its focus on preoperative data ensures broader applicability, paving the way for tailored therapeutic strategies and improved patient outcomes in diverse clinical scenarios.

Funding details

None.

CRediT authorship contribution statement

Guive Sharifi: Supervision, Conceptualization. Elham Paraandavaji: Investigation, Data curation. Nader Akbari Dilmaghani: Investigation, Data curation. Tohid Emami Meybodi: Investigation, Data curation. Ibrahim Mohammadzadeh: Investigation, Data curation. Neginalsadat Sadeghi: Investigation, Data curation. Amirali Vaghari: Visualization. Behnaz Niroomand: Visualization. Seyed Mohammad Tavangar: Resources. Mohammad reza Mohajeri Tehrani: Validation. Zahra Davoudi: Resources. Marjan Mirsalehi: Writing – review & editing. Seyed Ali Mousavinejad: Validation, Resources. Farzad Taghizadeh-Hesary: Writing – review & editing, Writing – original draft.

Informed consent

Not applicable.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

None.
The data that support the findings of this study are available on request from the corresponding author.

References

https://www.sciencedirect.com/science/article/pii/S2214623725000353

Filed under: Cushing's, pituitary, symptoms, Treatments | Tagged: , , , | Leave a comment »

Changing face of Cushing’s Disease Over Three Decades in Pituitary Center

Posted on October 1, 2025 by MaryO

Abstract

Objective

Cushing Disease (CD) presents with typical clinical findings, even though, there is a wide spectrum of manifestations. Over the years, the sings and symptoms of Cushing’s syndrome (CS) have become more subtle and atypical forms of CS have emerged. In this study, we aimed to investigate the changes in the clinical presentation of CD in recent years.

Materials and methods

In this study, CD patients followed by our center were examined. A total of 258 patients with CD were included in the study. The clinical findings at the time of presentation, laboratory and imaging findings, treatment modalities and remission status in the first year after treatment were evaluated.

Results

The mean age of the patients included in the study was 41.3 ±13.28 years. CD patients diagnosed between 2013 and 2023 were older than those diagnosed between 1990 and 2012 (p < 0.001). There was no difference between the groups in terms of gender. Moon face, purple striae, hirsutism, and menstrual irregularities were statistically significantly less frequent in the last 10 years than in previous years (p < 0.001; p = 0.004; p < 0.001; p < 0.001, respectively). In addition, patients who applied after 2013 had lower baseline cortisol and adrenocorticotropic hormone (ACTH) levels, and a smaller median size of the pituitary adenoma. Limitations of the study include its retrospective design and the subjectivity of clinical data.

Conclusion

As the clinical presentation of Cushing’s disease changes over time, waiting for the typical Cushing’s clinic can delay diagnosis. It is important that clinicians take this into account when they suspect CD.

 This is a preview of subscription content, log in via an institution  to check access.

Filed under: Cushing's, pituitary, symptoms | Tagged: , , , , , , | Leave a comment »

« Previous Page — Next Page »