Prevalence and Associated Risk Factors for Venous Thromboembolism in a Large Cohort of Patients With Cushing Disease

Posted on November 25, 2025 by MaryO

Abstract

Objective

Endogenous Cushing syndrome is associated with an intrinsic hypercoagulable state and an increased risk of venous thromboembolism (VTE). This study aimed to determine the prevalence and risk factors for VTE in a large cohort of patients with Cushing disease (CD).

Methods

A retrospective study was conducted at a tertiary care center, including 408 patients diagnosed with CD. Clinical, laboratory, hormonal, imaging, and outcome data were analyzed and compared based on the occurrence of VTE events. A control group of 323 patients with clinically nonfunctioning pituitary adenomas, all macroadenomas, who underwent similar surgical procedures, was used for comparison.

Results

VTE events were observed in 35 patients with CD (8.6%) and in 1 patient from the nonfunctioning pituitary adenoma group (0.3%; P < .001). The slight majority of VTE events (54%) occurred in the preoperative period. Logistic regression analysis identified obesity, mood disorders, supraclavicular fossa fullness, leukopenia or leukocytosis, elevated cortisol levels (both serum and 24-hour urinary cortisol), and the presence of postoperative complications (such as infections, cerebrospinal fluid leak, and vasopressin deficiency) as significant risk factors for VTE.

Conclusion

The findings of this study confirm a high prevalence of VTE events in patients with CD, irrespective of the surgical period. Risk factors associated with a higher likelihood of VTE include obesity, severity of hypercortisolism, and the occurrence of postoperative complications. In this patient population, thromboprophylaxis should be considered.

Introduction

Patients with endogenous Cushing syndrome (CS), including those with Cushing disease (CD), have a mortality rate that is 3 times higher than the general population.1, 2, 3, 4, 5 This increased mortality is primarily attributed to cardiovascular conditions (acute myocardial infarction, stroke, congestive heart failure, and venous thromboembolic [VTE] events), hyperglycemia, and infections.4
It is well-established that endogenous CS is intrinsically associated with VTE events,6, 7, 8, 9, 10, 11 independent of surgical procedures and metabolic disturbances. Previous studies have reported significant rates of VTE events in patients with CS, ranging from 2.6% to 18.2% (predominantly deep vein thrombosis [DVT] and pulmonary embolism [PE]),6, 7, 8, 9, 10, 11, 12 which is approximately 10 times higher than general population (DVT 0.53 to 1.62 per 1000 person-years and PE 0.39 to 1.15 per 1000 person-years).13
The pathophysiological mechanism underlying hypercortisolism as a thrombogenic condition is not fully understood. However, it is primarily attributed to the genomic action of cortisol, which leads to the upregulation of mRNA transcription for hemostatic factors, resulting in the activation of the coagulation cascade and impaired fibrinolytic capacity.6,14 Altered hemostatic parameters are observed even when compared to high-risk groups, such as those with metabolic syndrome.15, 16, 17 The studies reported increased levels of factor VIII, factor IX, von Willebrand factor, and fibrinogen; a shortened activated partial thromboplastin time (APTT); and elevated levels of factors that reduce fibrinolysis, such as plasminogen activator inhibitor-1, thrombin-activatable fibrinolysis inhibitor, and alpha-2-antiplasmin. Some studies also describe an increase in anticoagulant factors, such as protein C, protein S, and antithrombin III, likely through a compensatory mechanism.18
VTE prevalence in CD varies widely across studies, likely due to differences in populations, CS etiologies, inclusion of other events (eg, stroke), and timing (preoperative vs postoperative).6, 7, 8, 9, 10, 11, 12
These factors, along with variability in evaluated hemostatic parameters and use of thromboprophylaxis, hinder consensus on prophylaxis.18, 19, 20
The present study aimed to identify the prevalence and risk profile of VTE events in a large cohort of patients with CD.

Access through your organization

Check access to the full text by signing in through your organization.

Access through your organization

Section snippets

Patients, Controls, and Study Design

A retrospective observational cohort study was conducted at a single center, including 408 patients with CD who were evaluated between 1990 and 2020. Inclusion criteria consisted of patients with a confirmed CD, defined by pituitary adenoma with immunohistochemistry positive, remission after neurosurgery, a central-to-peripheral ACTH gradient in inferior petrosal sinus sampling (IPSS), macroadenoma, or Nelson syndrome after adrenalectomy. Exclusion criteria included lack of CD confirmation,

Description of Patients and Controls

A total of 408 patients with CD were included in the study, with a predominance of females (n = 324, 79%). The median age was 32 years (range: 8-71). Most patients presented with microadenomas (n = 207, 50.7%), while 27.0% (n = 110) had pituitary macroadenomas (≥10 mm on magnetic resonance imaging [MRI]; mean diameter 17.0 ± 9.1 mm, range 10-64 mm), including 4 giant tumors (≥4 cm). Ninety-one patients (22.3%) exhibited no visible or undefined lesions on sellar MRI. IPSS was performed in 152

Discussion

Strategies for preventing VTE events in CS have been researched in several reference centers.6,9,10,17,19
European surveys reported a VTE incidence of 14.6 per 1000 person-years in CS, about 10 times higher than in the general population. In patients on prophylaxis, the incidence dropped to 10.2 versus 25.6 in those without. Events were more common with greater disease severity, but the diversity of CS types and retrospective designs has limited standardized strategies.6
A Pituitary Society

Data Availability Statement

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.

Statement of Ethics

All procedures performed in this study that involved human participants were in accordance with the Ethical Standards of the Institutional National Research Committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The Ethical and Research Committees of the University of Sao Paulo Medical School approved the study, number 44044320.4.0000.0068.

Consent to Participate Statement

All participants or their legal guardians signed a written informed consent form.

Disclosure

The authors have no conflicts of interest to disclose.

Author Contributions

All authors contributed to the study conception/design and realization (A.J.G.P., R.L.B., M.B.C.C.-N., V.A.S.C., G.O.S., M.C.B.V.F., I.N.N., A.G., and M.C.M.). The first draft of the manuscript was written by A.J.G.P. and M.C.M. and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript (A.J.G.P., R.L.B., M.B.C.C.-N., V.A.S.C., G.O.S., M.C.B.V.F., I.N.N., A.G., and M.C.M.).

References (35)

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

Delirium Induced by Rapid Titration of Osilodrostat in a Patient With Cushing’s Disease

Posted on November 21, 2025 by MaryO

Abstract

Cushing’s disease frequently presents with psychiatric symptoms such as depression, anxiety, and cognitive impairment. Osilodrostat, an 11β-hydroxylase inhibitor, is used for persistent or recurrent cases, but rapid titration may precipitate adrenal insufficiency and psychiatric complications.

We report a woman in her early 40s with a history of major depressive disorder treated with clomipramine. After transsphenoidal surgery for Cushing’s disease, she remained hypercortisolemic, and hydrocortisone replacement was continued postoperatively for safety due to unstable cortisol secretion. Cortisol secretion was unstable, with day-to-day fluctuations. Osilodrostat was initiated at 2 mg/day. Shortly thereafter, urinary free cortisol (UFC) increased, and between days 3 and 5, she developed depressive symptoms, depersonalization, and suicidal ideation. These were judged to be related to cortisol elevation, and osilodrostat was rapidly titrated, reaching 40 mg/day by day 9. Depressive symptoms improved as UFC decreased. However, from day 9, she developed delirium with fluctuating consciousness, disorientation, purposeless hyperactivity, and stereotyped speech, peaking on days 10-12. During this period, blood pressure decreased, accompanied by tachycardia and fever. Infection and metabolic abnormalities were clinically excluded. Symptoms resolved spontaneously by day 14, with amnesia for the episode, and she was discharged on day 20 without recurrence.

This case illustrates a rare clinical course where depressive symptoms during cortisol elevation and delirium during cortisol reduction occurred sequentially in the same patient following rapid osilodrostat titration. The episode suggests that abrupt cortisol fluctuations may induce psychiatric symptoms even under hydrocortisone supplementation. Clinicians should avoid rapid titration and ensure close collaboration between endocrinology and psychiatry when psychiatric symptoms arise during treatment.

Introduction

Cushing’s disease is caused by an adrenocorticotropic hormone (ACTH) secreting pituitary adenoma, leading to chronic hypercortisolism. In addition to physical features such as central obesity, moon face, and hypertension, psychiatric symptoms including depression, anxiety, and cognitive impairment are frequently observed [1-3]. Depression occurs in 40-60% of patients and is associated with increased suicide risk. Anxiety and cognitive impairment are also common, and psychiatric symptoms may even precede the physical manifestations. Thus, psychiatrists may encounter such patients at an early stage, and it is clinically important to consider underlying endocrine disorders [1,3]. The first-line treatment is transsphenoidal surgery, but remission is not always achieved [4].

Osilodrostat, an oral 11β-hydroxylase inhibitor, is primarily used for the treatment of persistent or recurrent Cushing’s disease. By inhibiting cortisol synthesis, it effectively lowers circulating cortisol levels, thereby improving the clinical manifestations of hypercortisolism. The phase III LINC 3 trial demonstrated its efficacy [5], but adverse events such as adrenal insufficiency and psychiatric symptoms have been reported [6-8]. Acute adrenal insufficiency can present with hypotension, tachycardia, fever, and gastrointestinal symptoms, and in severe cases with impaired consciousness or delirium [9]. To minimize these risks, gradual titration in 2-mg increments at intervals of at least two weeks is recommended [6].

For monitoring treatment efficacy, urinary free cortisol (UFC) is widely used as a reliable marker reflecting total cortisol secretion over 24 hours and serves as a standard index of disease activity and treatment response [1,2].

Case Presentation

The patient was a 43-year-old woman with a history of major depressive disorder since her early twenties, treated mainly with clomipramine. Although she experienced recurrent episodes, she was able to continue working as a clinical psychologist, with occasional sick leave. Her past history included papillary thyroid carcinoma treated surgically, followed by hypothyroidism managed with levothyroxine 75 µg/day.

In her thirties, she developed treatment-resistant hypertension. In March 2024, inferior petrosal sinus sampling confirmed Cushing’s disease. In April 2024, she underwent transsphenoidal surgery and started hydrocortisone replacement at 30 mg/day. However, hypercortisolism and elevated ACTH persisted. Cortisol levels showed marked day-to-day fluctuations rather than being consistently elevated, and replacement therapy was continued for safety.

In June 2024, she was admitted to our endocrinology department because of persistent disease activity. Psychiatry was consulted due to her psychiatric history. At admission, she was alert, cooperative, and exhibited neither depressive nor psychotic symptoms. Clomipramine was continued. Physical examination revealed a BMI of 27.5, central obesity, moon face, and violaceous striae. Blood pressure was 155/105 mmHg. Routine chemistry and thyroid function were within normal limits. Endocrinological work-up confirmed persistent hypercortisolism: the 24-hour UFC was markedly elevated (409.2 µg/day; normal < 50 µg/day), midnight serum cortisol was inappropriately high (14.3 µg/dL; normally suppressed at night), and dexamethasone suppression testing failed to suppress morning cortisol (9.7 µg/dL after 0.5 mg dexamethasone). Corticotropin-releasing hormone stimulation testing demonstrated an exaggerated ACTH response (63.6 → 105.0 pg/mL), consistent with pituitary-dependent Cushing’s disease. Postoperative brain MRI showed only expected surgical changes without new lesions.

Figure 1 illustrates the clinical course in this case. Osilodrostat was initiated at 2 mg/day on day 1. UFC unexpectedly rose thereafter, and between days 3 and 5, she developed depressed mood, depersonalization, and suicidal ideation. These psychiatric symptoms were judged to be associated with increased cortisol secretion. Antidepressant adjustment was not attempted. Instead, priority was given to endocrine control, and osilodrostat was rapidly up-titrated. Although the risk of adrenal insufficiency was considered, treatment was deemed safe under hydrocortisone supplementation. By day 9, the dose of osilodrostat reached 40 mg/day, UFC decreased, and depressive symptoms improved.

Timeline-of-clinical-events-and-interventions-in-the-present-case.
Figure 1: Timeline of clinical events and interventions in the present case.

Panel (A) shows the osilodrostat dosage and 24-hour urinary free cortisol (UFC) levels; panel (B) depicts vital signs (sBP, systolic blood pressure; BT, body temperature); and panel (C) illustrates psychiatric symptoms and the dosages of antipsychotic medications, all plotted against treatment days.

However, from day 9 onward, delirium and psychomotor agitation emerged, peaking on days 10-12. She displayed fluctuating consciousness, global disorientation, impaired attention, purposeless hyperactivity, stereotyped behaviors, and repetitive utterances of meaningless numbers. She wandered barefoot and occasionally shouted fragmented phrases such as “Say you love me.” Anxiety and insomnia were prominent, but hallucinations and self-disturbances were absent.

At that time, her vital signs showed a decline in blood pressure from 155/105 mmHg to 125/59 mmHg, a pulse rate of 110/min, and a temperature of 38.3°C. Electrolytes and glucose were normal, and no inflammatory response or other signs of infection were detected. Because of marked psychomotor agitation, imaging and EEG were not performed. Risperidone and haloperidol were given but were ineffective.

At onset, the delirium was interpreted as a manifestation of hypercortisolism, partly because it occurred during a holiday when comprehensive evaluation was not feasible. Osilodrostat was therefore not reduced. As her symptoms improved spontaneously and she remained stable under hydrocortisone supplementation, the dose was maintained. Since the delirium resolved completely and did not recur, additional imaging or EEG was not performed.

By day 14, delirium had resolved, and the patient reported amnesia for the episode. No recurrence occurred, and she was discharged on day 20 at her and her family’s request. Outpatient follow-up confirmed stable status without recurrence of delirium.

Discussion

This case illustrates an unusual clinical course in which qualitatively distinct psychiatric symptoms appeared sequentially during rapid titration of osilodrostat. The initial depressive phase coincided with a transient rise in UFC and may have been related to unstable cortisol secretion that had already been observed prior to admission. Although not sufficient for a formal diagnosis, such variability is reminiscent of cyclical Cushing’s disease [10], which has also been associated with mood fluctuations [1,3]. Previous studies have demonstrated the link between hypercortisolism and depression [1,3], and our case is consistent with these findings during the early phase of treatment.

In contrast, the subsequent delirium phase was accompanied by hypotension, tachycardia, and fever, resembling adrenal insufficiency. Similar neuropsychiatric manifestations, including delirium and psychosis, have been described in previous reports of adrenal insufficiency [8,9]. However, unlike those cases, where symptoms typically emerged after drug withdrawal, delirium in our patient developed rapidly following dose escalation. This temporal pattern suggests that even transient cortisol reductions may precipitate acute neuropsychiatric symptoms.

Taken together, these observations both align with and extend prior findings linking cortisol dysregulation to psychiatric manifestations. Our case supports previous evidence that hypercortisolism is associated with depressive symptoms [1,3], whereas hypocortisolism predisposes to delirium or psychosis [8,9]. Importantly, it also highlights a dynamic aspect of this relationship: abrupt cortisol fluctuations themselves, regardless of direction, may transiently disrupt neuroendocrine homeostasis and trigger psychiatric symptoms. This interpretation is consistent with reports of cyclical Cushing’s disease showing alternating mood states [10], but it differs in that the fluctuation here was iatrogenic and temporally linked to rapid pharmacologic titration.

Pharmacological factors may have further amplified these effects. Clomipramine and antipsychotics such as haloperidol and risperidone are known to cause confusion or agitation, particularly under hormonal stress. It is therefore plausible that psychotropic drug interactions and cortisol fluctuations acted synergistically to produce the observed neuropsychiatric manifestations.

This report has several limitations. The onset of delirium occurred during a holiday, and severe agitation precluded blood sampling for serum cortisol, ACTH testing, or therapeutic steroid administration. Thus, strict diagnostic criteria for adrenal insufficiency could not be fulfilled. Nonetheless, the clinical presentation, with hypotension, tachycardia, fever, and altered consciousness, was consistent with an adrenal insufficiency-like state. Electrolytes, glucose, and inflammatory markers remained within normal limits, making infection or metabolic causes unlikely. However, structural or neurological contributors could not be completely excluded because imaging and EEG were not performed. Although the clinical picture resembled adrenal insufficiency, true adrenal crisis was unlikely given the normal electrolyte levels, spontaneous recovery, and maintained oral intake. Therefore, this episode may be better characterized as a state of functional adrenal dysregulation rather than frank adrenal insufficiency.

Conclusions

This case highlights a rare course in which depressive symptoms during cortisol elevation and delirium during cortisol reduction occurred sequentially in the same patient following rapid titration of osilodrostat. The episode suggests that even under hydrocortisone supplementation, abrupt cortisol fluctuations can induce psychiatric symptoms. However, because some observations were paradoxical and certain assessments could not be performed during the acute phase, these interpretations should be made with caution. The episode may represent a state of functional adrenal dysregulation rather than distinct phases of hyper- or hypocortisolism.

This case offers two clinical lessons. First, osilodrostat should be titrated gradually according to established guidelines. Second, if psychiatric symptoms arise during treatment, they are best managed through close collaboration between endocrinology and psychiatry.

References

  1. Pivonello R, Simeoli C, De Martino MC, et al.: Neuropsychiatric disorders in Cushing’s syndrome. Front Neurosci. 2015, 9:129. 10.3389/fnins.2015.00129
  2. Sharma ST, Nieman LK, Feelders RA: Cushing’s syndrome: epidemiology and developments in disease management. Clin Epidemiol. 2015, 7:281-93. 10.2147/CLEP.S44336
  3. Sonino N, Fava GA, Raffi AR, Boscaro M, Fallo F: Clinical correlates of major depression in Cushing’s disease. Psychopathology. 1998, 31:302-6. 10.1159/000029054
  4. Fleseriu M, Auchus R, Bancos I, et al.: Consensus on diagnosis and management of Cushing’s disease: a guideline update. Lancet Diabetes Endocrinol. 2021, 9:847-75. 10.1016/S2213-8587(21)00235-7
  5. Pivonello R, Fleseriy M, Newell-Price J, et al.: Efficacy and safety of osilodrostat in patients with Cushing’s disease (LINC 3): a multicentre phase 3 study with a double-blind, randomised withdrawal phase. Lancet Diabetes Endocrinol. 2020, 8:748-61. 10.1016/S2213-8587(20)30240-0
  6. U.S. Food and Drug Administration. Osilodrostat prescribing information. (2020). Accessed: October 18, 2025: https://www.accessdata.fda.gov/drugsatfda_docs/label/2020/212801s000lbl.pdf.
  7. Gadelha M, Bex M, Feelders RA, et al.: Randomized trial of osilodrostat for the treatment of Cushing disease. J Clin Endocrinol Metab. 2022, 107:e2882-95. 10.1210/clinem/dgac178
  8. Ekladios C, Khoury J, Mehr S, Feghali K: Osilodrostat-induced adrenal insufficiency in a patient with Cushing’s disease. Clin Case Rep. 2022, 10:e6607. 10.1002/ccr3.6607
  9. Arlt W: Society for Endocrinology endocrine emergency guidance: Emergency management of acute adrenal insufficiency (adrenal crisis) in adult patients. Endocr Connect. 2016, 5:G1-3. 10.1530/EC-16-0054
  10. Meinardi JR, Wolffenbuttel BH, Dullaart RP: Cyclic Cushing’s syndrome: a clinical challenge. Eur J Endocrinol. 2007, 157:245-54. 10.1530/EJE-07-0262

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

Rapid Endocrine Remission After ZAP-X Gyroscopic Radiosurgery for Cushing’s Disease

Posted on November 19, 2025 by MaryO

Abstract

Cushing’s disease is a rare but potentially life-threatening disorder caused by excessive adrenocorticotropic hormone (ACTH) secretion from a pituitary adenoma. Although transsphenoidal surgery remains the first-line treatment, radiotherapy (RT) can provide effective local and hormonal control in patients with persistent or recurrent disease; however, endocrine remission typically occurs only after several months or even years. To our knowledge, we report the first documented case of an exceptionally rapid hormonal remission following gyroscopic stereotactic radiosurgery (SRS) using the self-shielding ZAP-X system (ZAP Surgical Inc., San Carlos, CA, USA) in a patient with recurrent Cushing’s disease. The patient received a single-fraction dose of 25 Gy prescribed to the 57% isodose line. Remarkably, ACTH and cortisol levels normalized within one month after SRS, accompanied by a striking improvement in clinical symptoms and no treatment-related toxicity. This case highlights the potential of the ZAP-X gyroscopic radiosurgery platform to achieve rapid biochemical control in ACTH-secreting pituitary adenomas and suggests that the unique dose distribution characteristics of this novel technology may contribute to accelerated endocrine responses.

Introduction

Cushing’s disease stems from an excess production of adrenocorticotropic hormone (ACTH) by a pituitary adenoma, leading to elevated cortisol levels and symptoms resembling Cushing’s syndrome [1]. Untreated or inadequately managed hypercortisolism is associated with substantial morbidity and elevated mortality rates for patients with Cushing’s syndrome. While transsphenoidal surgery is frequently considered the preferred initial treatment approach, radiotherapy (RT) can also be contemplated, either as a standalone option for patients ineligible for surgery or as part of a multidisciplinary approach in instances where an adequate response is not attained following surgery. Conventional fractionated RT (CFRT), fractionated stereotactic radiosurgery (F-SRS), and stereotactic radiosurgery (SRS) have all been employed in the treatment of Cushing’s disease, yielding comparable disease control rates ranging from 46% to 100% [2-4]. However, unlike surgery, the response to RT may require time to manifest, and in certain instances, this duration can extend over months or years [4]. Sheehan et al. [5] indicated that the cure rate after RT in patients with Cushing’s disease was 34% in the first year, increasing to 78% by the fifth year. Although CyberKnife (Accuray Inc., Sunnyvale, CA, USA), Gamma Knife (Elekta AB, Stockholm, Sweden), and Linear Accelerator (LINAC)-based systems are all available for F-SRS and SRS, technological advancements are increasing the options. We have integrated the relatively new vault-free, frameless, gyroscopic radiosurgery system, ZAP-X (ZAP Surgical Inc., San Carlos, CA, USA), into our department and have begun using it for cranial F-SRS and SRS treatments [6,7]. To our knowledge, this report introduces the initial case of a patient diagnosed with Cushing’s disease who underwent gyroscopic SRS with the ZAP-X system, experiencing an unexpectedly rapid endocrine response following SRS.

Case Presentation

A 48-year-old female with an unremarkable medical history except for hyperlipidemia underwent investigation in 2018 due to symptoms and findings consistent with Cushing’s syndrome, leading to the detection of hypercortisolism. Her 24-hour urinary free cortisol and ACTH levels were 75 µg/day and 32 pg/mL, respectively. Serum cortisol remained unsuppressed following the 1 mg dexamethasone suppression test (DST), measuring 15.7 mcg/dL. On the magnetic resonance imaging (MRI) of the pituitary gland, a 4.5×3 mm microadenoma was detected in the left half of the adenohypophysis. Following the referral to the neurosurgery department, the patient underwent tumor resection via transsphenoidal endoscopic surgery in December 2018. The pathology resulted in a corticotroph pituitary adenoma. All her symptoms and signs related to Cushing’s disease resolved after surgery, and postoperative MRI showed no residual tumor. The patient was placed under observation without additional treatment. During follow-up visits, the patient remained asymptomatic for approximately five years.

In September 2023, the patient presented with complaints of proximal muscle weakness, irregular menstruation, and Cushingoid appearance. Upon evaluation, hypercortisolism was detected once again. On the pituitary MRI, no residual or recurrent lesion was observed. Subsequently, the patient underwent a second transsphenoidal surgery, but the pathology result did not reveal tissue consistent with a pituitary adenoma. In January 2024, upon initial presentation to our center, the patient’s ACTH level was 29.8 pg/mL, 24-hour urinary free cortisol was 442 µg/day, and serum cortisol following a 1 mg DST was 19 mcg/dL. The levels of the remaining anterior pituitary hormones were within normal ranges. The patient, who continued to exhibit symptoms consistent with Cushing’s syndrome, underwent another pituitary MRI. At this point, it was discovered that there was a recurrent lesion measuring 2×1 mm on the left half of the adenohypophysis. Pasireotide (0.6 mg once daily) was initiated for persistent hypercortisolism but was discontinued due to frequent diarrhea and a widespread allergic skin reaction. The patient was then evaluated by the multidisciplinary neuro-oncology tumor board at our hospital, which recommended RT as the next step. SRS was selected as the RT technique due to the tumor’s small size and its lack of proximity to critical structures such as the optic chiasm. A simulation computed tomography (CT) scan with a 1 mm axial slice thickness was conducted with the patient in the supine position. Intravenous contrast and a thermoplastic mask were utilized to ensure better visualization and precise immobilization. Gross tumor volume (GTV) was delineated as the macroscopic tumor volume according to the MRI, which was performed a few days before SRS. A planning target volume (PTV) was not generated for this case. The prescription dose was 2500 cGy in a single fraction to the 57% isodose (Figure 1). Dose-volume histogram (DVH) was presented in Figure 2. The SRS plan was generated with the integrated ZAP-X treatment planning software (version 1.8.58.12369), and detailed parameters of the plan were presented in Table 1.

Dose-distribution-of-the-gyroscopic-stereotactic-radiosurgery-plan-for-pituitary-adenoma
Figure 1: Dose distribution of the gyroscopic stereotactic radiosurgery plan for pituitary adenoma

The image shows the three-dimensional dose distribution generated using the ZAP-X system (ZAP Surgical Inc., San Carlos, CA, USA). The prescription dose of 25 Gy to the 57% isodose line is illustrated. A: planning computed tomography (CT) scan showing the isodose distribution around the target; B: planning magnetic resonance imaging (MRI) fused with CT for target delineation.

Dose-volume-histogram-(DVH)-of-the-gyroscopic-stereotactic-radiosurgery-plan
Figure 2: Dose-volume histogram (DVH) of the gyroscopic stereotactic radiosurgery plan

The DVH demonstrates a steep dose fall-off beyond the target margins, with minimal exposure to the optic nerves, optic chiasm, and brainstem, confirming optimal dose conformity and effective sparing of organs at risk. Within the gross tumor volume (GTV), the dose distribution is intentionally inhomogeneous, with a hot spot centrally located to ensure adequate tumor coverage and biological effectiveness.

Parameters Values
Volume (GTV) 0.13 cm3
Prescription dose & isodose 2500 cGy & 57.6%
Coverage 95.68%
Homogeneity index 1.74
New conformity index 1.48
Gradient index 3.58
GTV Dmean 3249 cGy
GTV Dmax 4340 cGy
GTV Dmin 2364 cGy
Optic chiasm (Dmax) 452 cGy
Left optic nerve (Dmax) 480 cGy
Right optic nerve (Dmax) 212 cGy
Brainstem (Dmax) 233 cGy
Number of beams 128
Number of isocenters 3
Monitor units 16.121
Collimator thicknesses 4 & 4 & 5 mm
Treatment delivery time 33 min
Table 1: Detailed parameters of the gyroscopic radiosurgery plan

Dmax: maximum dose; Dmean: mean dose; Dmin: minimum dose; GTV: gross tumor volume; mm: millimeter; min: minute

The treatment was well tolerated, and a marked biochemical response was observed one month after SRS, with ACTH and 24-hour urinary free cortisol levels decreasing to 14.2 pg/mL and 116 µg/day, respectively. Serum cortisol following a 1 mg DST was suppressed to 1.6 µg/dL. Concurrently, there was a noticeable improvement in the clinical signs and symptoms of Cushing’s disease. The patient was subsequently followed with regular clinical assessments at three-month intervals for one year. Throughout the follow-up period, ACTH, 24-hour urinary free cortisol, and post-DST serum cortisol levels remained near-normal (Figure 3). Levels of other anterior pituitary hormones were within normal limits. The patient’s biochemical parameters, including ACTH, urinary free cortisol, and serum cortisol levels before and after SRS, are summarized in Table 2. At the three-month post-SRS MRI, the lesion was found to be radiologically stable. However, the patient reported a subjective improvement in proximal muscle weakness beginning one month after treatment. No SRS-related toxicity was observed during the follow-up period, and partial regression of the Cushingoid phenotype was documented (Figure 4).

Temporal-changes-in-hormonal-parameters-following-gyroscopic-stereotactic-radiosurgery
Figure 3: Temporal changes in hormonal parameters following gyroscopic stereotactic radiosurgery

A: adrenocorticotropic hormone (ACTH) levels showed a rapid decline within the first month after treatment, remaining suppressed throughout follow-up; B: twenty-four-hour urinary free cortisol (UFC) demonstrated a similar sharp reduction after radiosurgery, indicating early biochemical response; C: serum cortisol levels after dexamethasone suppression normalized by the first month and remained within the physiological range during subsequent evaluations, consistent with sustained hormonal remission.

Parameter Unit Reference Range At Initial Diagnosis (2018) Recurrence (Jan 2024, before SRS) 1 Month After SRS 3 Months After SRS 6 Months After SRS 12 Months After SRS
ACTH pg/mL 7.2 – 63.3 32 29.8 14.2 11.0 12.5 15.0
24-hour UFC µg/day 20 – 90 75 442 116 65 55 45
Serum cortisol after 1 mg DST µg/dL < 1.8 (suppressed) 15.7 19.0 1.6 9.0 2.5 3.0
Table 2: Summary of laboratory findings before and after gyroscopic stereotactic radiosurgery

ACTH: adrenocorticotropic hormone; UFC: urinary free cortisol; DST: dexamethasone suppression test; SRS: stereotactic radiosurgery

Facial-appearance-before-and-after-gyroscopic-stereotactic-radiosurgery-(SRS)
Figure 4: Facial appearance before and after gyroscopic stereotactic radiosurgery (SRS)

The images illustrate the patient’s appearance at the time of initial diagnosis (A), before SRS (B), and after the procedure (C).

Discussion

To our knowledge, we report a rapid endocrine response observed in the first patient with Cushing’s disease treated using the ZAP-X gyroscopic radiosurgery system. Despite the patient having a significantly high 24-hour urinary free cortisol level before SRS, there was a considerable decrease within a short period after SRS.

Both surgical and medical treatments, along with RT, are viable approaches for managing pituitary adenomas. Approximately 70% of pituitary adenomas are associated with syndromes characterized by excessive hormone secretion, with the most common types producing prolactin, growth hormone, and ACTH [8]. Unlike non-secreting adenomas, the treatment goal for secreting adenomas extends beyond local tumor control to include the management of endocrinopathies. Although transsphenoidal surgery is commonly regarded as the first-line treatment, RT may also be considered, either as a primary modality in patients who are not surgical candidates or as part of a multimodal strategy when surgical outcomes are suboptimal. CFRT, F-SRS, and SRS have all been employed in treating patients with Cushing’s disease, yielding comparable disease control rates [9]. CFRT may be preferred, particularly for larger tumors or those located near organs at risk (OAR). In appropriately selected cases, advanced techniques such as F-SRS and SRS can shorten treatment duration and enable dose escalation within the tumor while providing a rapid dose fall-off outside the target volume. While RT can effectively control local tumor growth, its success in addressing endocrinopathies is typically more limited. In a systematic review, the rates of local tumor control and endocrine control for Cushing’s disease were reported as 92% and 48%, respectively [9]. Additionally, the radiation doses required for tumor control and endocrine response vary from each other [10]. While SRS doses ranging from 12 to 20 Gy typically achieve adequate local tumor control, especially in non-secreting adenomas, it has been observed that endocrine response rates improve at marginal doses around 30 Gy [11,12]. However, administering high doses can be challenging due to the presence of OAR, such as the optic apparatus, which is located in close proximity to the target volumes. It is recommended that the volume of the OAR receiving a dose of 8 Gy for the optic apparatus in SRS plans should be <0.2 cm³, and the volume receiving a dose of 10 Gy should be <0.035 cm³ [13]. Therefore, modern SRS platforms, which enable the delivery of high doses within the target volume while ensuring steep dose fall-off beyond it, offer the potential to widen the therapeutic window. In our patient, the ZAP-X gyroscopic SRS system enabled the delivery of 2500 cGy to the GTV at the 57% isodose line, while maintaining doses to OAR below recommended thresholds. Additionally, intratumoral hotspots allowed for the desired dose inhomogeneity, aligning with the core principles of SRS.

It is important to note that many patients with secreting pituitary adenomas suffer from symptoms caused by excessive hormone secretion, significantly impacting their quality of life and requiring consideration in treatment decisions [14]. In cases where patients experience severe symptoms due to elevated hormone levels, surgery may be prioritized, although various medical treatments are also viable options to consider. Pasireotide, a somatostatin analogue with multireceptor targeting, serves as an effective treatment for patients with persistent or recurring hypercortisolism post-surgery or when surgery isn’t viable. However, its tolerability is debatable due to various adverse effects such as hyperglycemia and diarrhea [15]. In our patient, despite initiating pasireotide due to persistently elevated hormone levels post-surgery, it was discontinued within less than two weeks due to intolerable adverse effects.

The biological effect of SRS on hormonal hypersecretion is believed to unfold gradually, and in some cases, this process may span months or even years. Sheehan et al. [5] reported outcomes for patients with Cushing’s disease, revealing a response rate of 34% at one year, 54% at two years, 72% at three years, and 78% at five years following SRS. In our patient, a significant decrease in 24-hour urinary free cortisol level was observed within only one month after SRS. To the best of our knowledge, this represents the most rapid endocrine response after SRS reported to date. It is important to consider, however, that the rapid hormonal normalization observed in our case may have been influenced by prior medical and surgical interventions. The patient underwent two transsphenoidal surgeries and briefly received pasireotide before radiosurgery, which could have altered tumor biology or hormonal responsiveness. Nevertheless, the close temporal relationship between ZAP-X treatment and biochemical remission strongly suggests a causal association. Potential factors contributing to this rapid endocrine response may include the administration of an effective radiation dose, such as 2500 cGy, utilization of a precise SRS technique like gyroscopic radiosurgery, and the presence of hotspots within the tumor, reaching up to 4000 cGy in a single fraction. During RT, the presence of hotspots within target volumes has been shown to be associated with increased local control for various tumor types [16,17]. Therefore, while it may not be directly attributable to hotspots, it seems possible that the underlying mechanism of the rapid endocrine response observed in our patient could be related to them. However, the short follow-up duration is the main limitation of this report.

Conclusions

To our knowledge, we report the first case of a refractory ACTH-secreting pituitary adenoma successfully treated using the vault-free ZAP-X gyroscopic SRS system. This case demonstrates that the unique design of the ZAP-X platform enables the safe delivery of a highly effective dose to the target while minimizing exposure to surrounding structures. In patients with Cushing’s disease, SRS can achieve rapid endocrine remission, although prospective studies are warranted to define the optimal dose and treatment parameters.

References

  1. Lonser RR, Nieman L, Oldfield EH: Cushing’s disease: pathobiology, diagnosis, and management. J Neurosurg. 2017, 126:404-17. 10.3171/2016.1.JNS152119
  2. Estrada J, Boronat M, Mielgo M, et al.: The long-term outcome of pituitary irradiation after unsuccessful transsphenoidal surgery in Cushing’s disease. N Engl J Med. 1997, 336:172-7. 10.1056/NEJM199701163360303
  3. Minniti G, Osti M, Jaffrain-Rea ML, Esposito V, Cantore G, Maurizi Enrici R: Long-term follow-up results of postoperative radiation therapy for Cushing’s disease. J Neurooncol. 2007, 84:79-84. 10.1007/s11060-007-9344-0
  4. Hughes JD, Young WF, Chang AY, et al.: Radiosurgical management of patients with persistent or recurrent Cushing disease after prior transsphenoidal surgery: a management algorithm based on a 25-year experience. Neurosurgery. 2020, 86:557-64. 10.1093/neuros/nyz159
  5. Sheehan J, Kondziolka 😧 Results of gamma knife surgery for Cushing’s disease. J Neurosurg. 2013, 119:1642.
  6. Weidlich GA, Bodduluri M, Achkire Y, Lee C, Adler JR Jr: Characterization of a novel 3 megavolt linear accelerator for dedicated intracranial stereotactic radiosurgery. Cureus. 2019, 11:e4275. 10.7759/cureus.4275
  7. Ehret F, Kohlhase N, Eftimova D, et al.: Self-shielding gyroscopic radiosurgery: a prospective experience and analysis of the first 100 patients. Cureus. 2024, 16:e56035. 10.7759/cureus.56035
  8. Daly AF, Beckers A: The epidemiology of pituitary adenomas. Endocrinol Metab Clin North Am. 2020, 49:347-55. 10.1016/j.ecl.2020.04.002
  9. Mathieu D, Kotecha R, Sahgal A, et al.: Stereotactic radiosurgery for secretory pituitary adenomas: systematic review and International Stereotactic Radiosurgery Society practice recommendations. J Neurosurg. 2022, 136:801-12. 10.3171/2021.2.JNS204440
  10. Minniti G, Osti MF, Niyazi M: Target delineation and optimal radiosurgical dose for pituitary tumors. Radiat Oncol. 2016, 11:135. 10.1186/s13014-016-0710-y
  11. Kotecha R, Sahgal A, Rubens M, et al.: Stereotactic radiosurgery for non-functioning pituitary adenomas: meta-analysis and International Stereotactic Radiosurgery Society practice opinion. Neuro Oncol. 2020, 22:318-32. 10.1093/neuonc/noz225
  12. Paddick I: A simple scoring ratio to index the conformity of radiosurgical treatment plans. Technical note. J Neurosurg. 2000, 93 Suppl 3:219-22. 10.3171/jns.2000.93.supplement
  13. Timmerman R: A story of hypofractionation and the table on the wall. Int J Radiat Oncol Biol Phys. 2022, 112:4-21. 10.1016/j.ijrobp.2021.09.027
  14. Johnson MD, Woodburn CJ, Vance ML: Quality of life in patients with a pituitary adenoma. Pituitary. 2003, 6:81-7. 10.1023/b:pitu.0000004798.27230.ed
  15. Manetti L, Deutschbein T, Schopohl J, et al.: Long-term safety and efficacy of subcutaneous pasireotide in patients with Cushing’s disease: interim results from a long-term real-world evidence study. Pituitary. 2019, 22:542-51. 10.1007/s11102-019-00984-6
  16. Owen D, Siva S, Salama JK, Daly M, Kruser TJ, Giuliani M: Some like it hot: the value of dose and hot spots in lung stereotactic body radiation therapy. Int J Radiat Oncol Biol Phys. 2023, 117:1-5. 10.1016/j.ijrobp.2023.03.056
  17. Abraham C, Garsa A, Badiyan SN, et al.: Internal dose escalation is associated with increased local control for non-small cell lung cancer (NSCLC) brain metastases treated with stereotactic radiosurgery (SRS). Adv Radiat Oncol. 2018, 3:146-53. 10.1016/j.adro.2017.11.003

 

From https://www.cureus.com/articles/430830-rapid-endocrine-remission-after-zap-x-gyroscopic-radiosurgery-for-cushings-disease-a-case-report?score_article=true#!/

​

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

Prospective Assessment of Mood and Quality of Life in Cushing Syndrome before and after Biochemical Control

Posted on November 17, 2025 by MaryO

Abstract

Context

Cushing syndrome (CS) impairs quality of life (QoL) and mood. Prospective real-life data on post-treatment recovery and predictors of improvement are limited.

Objectives

Evaluate changes in QoL, depression, and anxiety in patients with CS, before and after biochemical control, and identify predictors of clinically meaningful improvement.

Design and Setting

Prospective observational study at a tertiary center.

Patients

67 patients with endogenous CS (60 pituitary, 7 adrenal) were assessed with active disease and again after achieving biochemical control through surgery and/or medication.

Outcomes

Patient-reported outcomes included CushingQoL, Beck Depression Inventory-II (BDI-II), and State-Trait Anxiety Inventory (STAI).

Results

Mean and longest follow-up was 2.3 and 11.5 years, respectively. Treatment led to improvements in mean scores across all domains (QoL: +18.2±20.9, BDI: –6.8±8.6, STAI-State: –9.6±12.5, STAI-Trait: –8.6±12.6; all p < 0.001). However, minimal important difference was achieved in 64.6% for QoL, 67.9% for BDI, 53.2% and 52.8% for STAI subscales. After multivariable analysis, QoL improvements were predicted by lower baseline BMI, pre-treatment symptoms ❤ years, post-operative hydrocortisone replacement >6 months, and normal follow-up late-night salivary cortisol (LNSC). Depression improvements were predicted by symptoms ❤ years, normal follow-up LNSC, and surgical treatment. Anxiety improvements were predicted by younger age and >6 months post-operative hydrocortisone. Depression improved more gradually than QoL and anxiety.

Conclusions

Although effective treatment improves mood and QoL in CS, clinically meaningful recovery is variable and incomplete for some patients. Our findings highlight the need to limit diagnostic delay and provide comprehensive post-treatment care that includes normalization of cortisol circadian rhythm.

Accepted manuscripts
Accepted manuscripts are PDF versions of the author’s final manuscript, as accepted for publication by the journal but prior to copyediting or typesetting. They can be cited using the author(s), article title, journal title, year of online publication, and DOI. They will be replaced by the final typeset articles, which may therefore contain changes. The DOI will remain the same throughout.

This content is only available as a PDF.

© The Author(s) 2025. Published by Oxford University Press on behalf of the Endocrine Society.
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs licence (https://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial reproduction and distribution of the work, in any medium, provided the original work is not altered or transformed in any way, and that the work is properly cited. For commercial re-use, please contact reprints@oup.com for reprints and translation rights for reprints. All other permissions can be obtained through our RightsLink service via the Permissions link on the article page on our site—for further information please contact journals.permissions@oup.com. See the journal About page for additional terms.

Filed under: adrenal, Clinical trials, Cushing's, pituitary, 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 »

« Previous Page — Next Page »