Browse the corpus

Cushing syndrome (CS) describes the exposure of the body to supraphysiologic levels of glucocorticoids. Although exogenous (iatrogenic) CS is common, endogenous pediatric CS, is a rare pediatric endocrine condition. Population studies of the incidence of the disease have shown that endogenous CS occurs in about 3-50 cases per million people per year, depending on the population studied; pediatric patients in these studies represent 6-7% of all cases (-).

Endogenous CS can be classified as ACTH-dependent (pituitary or ectopic) or ACTH-independent CS (adrenal-related, ) (). The etiology of pediatric CS differs based on the age group of the patient (). In patients younger than 5 years of age, ACTH-independent CS is more common compared to older children and adolescents who usually present with ACTH-dependent CS. Ectopic CS (ECS) is rare at any age group ().

ACTH-dependent CS is most commonly due to a corticotroph pituitary neuroendocrine tumor (PitNET, also called pituitary adenoma or Cushing disease, (CD). These are monoclonal lesions that continue to express some of the characteristics of the normal corticotroph cell which can be useful in the diagnostic evaluation of patients (, ). Corticotroph-secreting PitNETs are usually microadenomas with median diameter of 5mm and do not often show signs of invasion to the cavernous sinus or other parasellar structures (). Rare cases of aggressive PitNETs have been reported in the pediatric population with either resistance to treatment or distant metastasis (metastatic PitNETs) (). These are associated with specific histologic subtypes, such as Crooke cell adenomas (). Infantile onset of ACTH-dependent CS with a pituitary lesion is often due to a pituitary blastoma. In 2014 de Kock et al , collected tissues from several infants who had been diagnosed with very young onset CD and reported that the tumors were consistent with pituitary blastomas as they had histologic findings of undifferentiated epithelium Rathke-like cells, mixed with hormone producing cells (). They were able to identify germline and/or somatic DICER1 gene defects in these patients, suggesting that pituitary blastoma is a rare but almost pathognomonic presentation of DICER1 syndrome ().

ECS is due to neuroendocrine tumors secreting ACTH and/or CRH outside the hypothalamic-pituitary axis. In older children and adolescents, the most common source of ECS are bronchial carcinoids, thymic carcinoids, and gastro-entero-pancreatic NETs (-). By contrast, in children younger than 5-10 years of age, ECS often presents in the context of pediatric specific tumors such as Wilm’s tumors, neuroblastomas, and others (, ).

ACTH-independent CS is commonly caused by unilateral adrenocortical tumors, cortisol-producing adenomas or carcinomas (). Cortisol-producing adenomas are benign lesions with isolated cortisol secretion, while adrenocortical carcinomas are aggressive tumors and may commonly co-secrete cortisol and androgens in up to 80% of all cases (, ). Bilateral adrenocortical disorders account for <2% of all cases of CS but some subtypes may be more prevalent in children compared to adults given their association with germline genetic defects (). Micronodular adrenocortical disease is the most common type of bilateral adrenocortical disorder in pediatric patients. This category may be further divided in primary pigmented micronodular adrenocortical disease (PPNAD) where the adrenals present with multiple dark brown pigmented micronodules (due to lipofuscin deposition with most with diameter of <1cm) with internodular cortical atrophy, or the absence of these findings referred to as isolated micronodular adrenocortical disease (i-MAD) (). PPNAD may be identified in the context of Carney complex (CNC) and less often as isolated PPNAD (). Bilateral macronodular adrenocortical disease presenting with bilateral macronodules (most with diameter of ≥1cm) is rare in the pediatric population.

Genetic causes are found in less than half of the patients with pediatric CD and more commonly in adrenal-related CS. For patients presenting with pediatric onset CS, it is recommended to obtain genetic testing directed to the source of hypercortisolemia, i.e. adrenal vs. pituitary causes. Although the yield in CD may be low, in cases of pituitary blastomas or bilateral micronodular disease genetic testing has higher chance of identifying the genetic cause and lead to screening for other related manifestations that may be important, such as cardiac myxomas in patients with CNC.

Germline mutations are identified in less than 10% of patients with pediatric CD (). Of the most common causes are MEN1 (causing multiple endocrine neoplasia type 1 syndrome, MEN1), CDKN1B (causing MEN4), and CABLES1 gene defects (). Genes associated with familiar isolated pituitary adenoma (FIPA) syndrome, such as AIP , SDHx , and MAX , or syndromes associated with pituitary tumors amongst other manifestations, such as CNC due to PRKAR1A gene defects, do not commonly cause corticotropinomas and have only been reported in few case reports (). As mentioned above, young children (<2 years old) presenting with pituitary blastomas should be screened for DICER1 gene defects (). DICER1 codes for an endoribonuclease that processes miRNAs (). Patients with DICER1 or pleuropulmonary syndrome present with multiple tumors in lungs, kidneys, multinodular goiter, and other manifestations. Pituitary blastomas are present in less than 10% of all patients and always within the first years of life ().

Somatic mutations are more likely to be identified in corticotropinomas. USP8 mutations in the 14-3-3 binding motif hotspot region of the gene have been reported as the cause of 40-60% of adults with CD (, ). Pediatric data suggest that USP8 mutations are less common and identified in up to 30% of cases (). USP8 is a deubiquitinase involved in recycling of the epidermal growth factor receptor (EGFR) and mutations in the hotspot region led to increased catalytic activity, activation of the EGF pathway, and increased POMC expression. In children, USP8 mutant tumors presented with larger size and higher risk for persistent disease after surgery or recurrence after initial remission (). Data in adult patients did not confirm this finding, and the prognostic value of identifying a USP8 mutation is still unclear (). Other somatic mutations identified in corticotropinomas include USP48 , TP53 , and BRAF , but the incidence in pediatric patients is unknown (). Finally, in a subset of patients with pediatric corticotropinomas large genomic chromosomal deletions/gains are identified and are associated with larger tumor and higher risk of invasion of the cavernous sinus ().

ECS may present in various neuroendocrine tumors and the genetic background is associated with the primary tumor. MEN1, MEN2 ( RET gene mutations), and some gene fusions have been described according to the tissue involved in ectopic ACTH secretion (, ).

Pediatric cortisol producing adrenocortical carcinomas may present in the context of TP53 mutations (). In the Brazilian South and Southeast population, high prevalence of a germline founder TP53 mutation (p.R337H) is associated with high incidence of pediatric adrenocortical carcinomas (, ). Germline TP53 mutations may also present as Li-Fraumeni syndrome where patients have high risk for breast, central nervous system, bone, and other tumors (). Cortisol-producing adrenocortical adenomas may be associated with gene defects leading to activation of the cyclic AMP (cAMP) protein kinase A (PKA) pathway, such as somatic mutations in PRKACA, PRKAR1A , and PRKACB genes (, ). Finally, somatic gene defects in the Wnt signaling pathway have also been identified in adrenocortical tumors ().

ACTH-independent CS due to PPNAD presents commonly in the context of CNC (). CNC is an autosomal dominant multiple neoplasia syndrome caused by inactivating mutations of the gene PRKAR1A , coding for the regulatory subunit 1 alpha of PKA, or less often linked to a second locus at chromosome 2p16 (-). Inactivating mutations in PRKAR1A lead to constitutive activation of PKA and downstream pathways (). Patients with CNC present with several manifestations including PPNAD, pituitary abnormalities most often presenting as growth hormone dysregulation or acromegaly, thyroid nodules or carcinomas, testicular tumors, cardiac and skin myxomas, characteristic skin lesions, breast myxomatosis or adenomas, osteo-chondro-myxomas and psammomatous melanotic schwannomas (). PPNAD in CNC is often diagnosed in the third decade of life but patients as young as in the first decade of life have been reported (). Additional information about CNC can be found in the chapter entitled “Carney Complex” of Endotext ().

Additional genetic defects associated with bilateral adrenocortical disease include PRKACA genomic gains, PDE11A , and PDE8A gene defects identified in patients with macronodular adrenocortical disease or isolated micronodular disease (-). PRKACA codes for the catalytic subunit of PKA, and chromosomal gains lead to increased PKA signaling (). PDE11A and PDE8A codes for phosphor-diesterases that catalyze and decrease cAMP levels. Inactivating mutations in these genes lead to increased circulation of cAMP and increased PKA activity (, ). Macronodular adrenocortical disease due to ARMC5 gene defects often seen in adults is rare in the pediatric population (). Neonatal ACTH-independent CS may be seen in the context of McCune-Albright syndrome (MAS) (). In these cases, CS presents within the first year of life and may have detrimental and rapidly developing symptoms which may even lead to death. However, if managed medically, neonatal CS in MAS may resolve on its own (). Rare cases of neonatal onset adrenocortical disease have also been reported in the context of Beckwith-Wiedemann syndrome (, ).

The presentation of pediatric CS has similarities and differences from that in adults (). The hallmark of pediatric CS is weight gain with concomitant height deceleration () (). This finding can help discriminate patients with CS from with simple obesity who often have preserved height percentile (). Fat deposition in pediatric patients may not be as prominently centripetal as noted in adults, and may present often as generalized obesity similar to other causes (). Although height deceleration is seen in most cases of growing children, patients may not be short at presentation, may have completed growth by the time hypercortisolemia occurred, or may be exposed to episodic hypercortisolemia which may have more limited effect on their height (, , ). Bone age is often within the expected range for the chronologic age or advanced in pediatric patients with endogenous CS, and is correlated with the levels of adrenal androgens which are often increased in ACTH-dependent CS (). Dermatologic findings present in CS include striae (which are present in 60-80% of patients and may not have the characteristic appearance of deep purple color and thickness as in adults), facial plethora, acne (more common in ACTH-dependent CS possibly due to stimulation of adrenal zona reticularis by ACTH), hirsutism in women, hypertrichosis, acanthosis nigricans, and easy bruising (, , ).

Patients with CS often present with delayed puberty in males and females and/or irregular menses and secondary amenorrhea in females (). As in patients with iatrogenic CS, pediatric patients with endogenous CS present with decreased bone mineral density, with lower scores in the spinal measurements (-). Proximal muscle weakness although reported is less frequent than adult patients (). Pediatric patients with endogenous CS, especially younger in age, often present with behavioral and neurocognitive changes. They may report behavioral changes including compulsive behaviors with overachievement goals, described as excellent students, along with increased anxiety and irritability (). They may also report mood changes, depressed mood, sleep problems, and memory issues similar to adults. Headaches are common in pediatric patients and can be noted in up to 80% of them ().

Hypercortisolemia and its related obesity lead to metabolic syndrome (). Patients often present with insulin resistance and up to 30% of patients may have impaired glucose metabolism (). Hyperlipidemia and elevated ALT as a surrogate marker of metabolic associated fatty liver disease (MAFLD) are also present in almost half of the patients (, ). Hypertension is present in almost 50% of patients with endogenous CS and cases of posterior reversible encephalopathy syndrome (PRES) due to hypertensive emergency have been reported as the initial manifestation of CS in pediatric patients (, ). Similar to adults, pediatric patients with CS present with a hypercoagulable state associated with abnormal levels of procoagulants, antifibrinolytics, and anticoagulant factors, such as factor VIII, antithrombin III, protein C and S, and prolonged partial thromboplastin time (PTT) (). Although in adult patients with CS the risk of venous thromboembolism is more studied, the exact incidence, risk, and thromboprophylaxis protocols in children are not as well delineated and depend on clinical judgement ().

Additional findings in pediatric CS include characteristic abnormalities in CBC due to glucocorticoids effects including increased WBC count, neutrophil count, low normal lymphocyte count, and increased neutrophil-to-lymphocyte ratio (NLR) (). Although immunosuppression may lead to severe infections in patients with significantly elevated cortisol levels, in most pediatric cases infections are limited to less clinically significant areas such as skin infections etc. (). However, in very young patients, especially in neonatal CS, or patients with severe hypercortisolemia, such as in ECS, opportunistic infections may lead to significant morbidity and even death and prophylaxis should be initiated (, ). Electrolyte abnormalities seen in endogenous CS include hypokalemia, uncommon overall but seen more frequently in ECS, and hypercalciuria which may lead to nephrolithiasis (, ).

Patients with hypercortisolemia also present with other hormonal defects including abnormal thyroid function test with a pattern of central hypothyroidism, abnormal GH secretion with IGF-1 levels usually preserved within the reference range, and suppressed gonadotropins (-). Tumor stalk compression effects may lead to hyperprolactinemia, although this is uncommon due to the small size of most corticotropinomas. Androgen levels are commonly elevated in ACTH-dependent CS due to adrenal zona reticularis stimulation from ACTH, or in adrenocortical carcinomas where co-secretion of cortisol and DHEAS may be seen.

The diagnostic evaluation of pediatric CS follows the guidelines of the endocrine and pituitary society adjusted for the pediatric population (, , ). Screening for hypercortisolemia is preferably done with at least two of the following tests: 24-hour urinary free cortisol (UFC, measured on 2-3 days), midnight (or late night) cortisol measured on 2-3 days, and suppression of cortisol to low dose dexamethasone (). Specific details on these tests can be found in the chapter entitled “Endocrine Testing Protocols: Hypothalamic Pituitary Adrenal Axis” of Endotext ().

The loss of the diurnal rhythm of ACTH/cortisol secretion is the first abnormality noted in patients with CS (, ). In clinical practice, salivary cortisol has been used to measure midnight or late night cortisol levels as it is convenient and can be collected at home (, ). If this is not available, then serum midnight cortisol is an alternative accurate screening test (). A serum cortisol level of ≥4.4mcg/dL was able to distinguish almost all pediatric patients with CS with a sensitivity of 99% and a specificity of 100% (). Serum cortisol needs to be measured from an indwelling catheter that has been placed at least 2 hours prior to sampling. We instruct patients to turn off all screens by 10pm and blood should be collected without awakening the patient ().

The 24-hour urine collection should be performed on two or three days, to ensure optimal urine collection and account for the known day-to-day variability in urinary cortisol in patients with CS (, ). It is generally recommended to collect urine on days of routine physical activities and avoid days when increased stressful activities are expected, like competitive sports games etc. (). Additionally, patients are advised to consume normal amount of fluids as excessive fluid intake and urine output may lead to false positive results (). The urine samples should be measured for urine creatinine to ensure normal kidney function, but we do not routinely correct UFC levels for the urine creatinine level as this may lead to inaccurate results (). The low dose or 1mg overnight dexamethasone suppression test is performed similar to adults (). Dose adjustment has been used in several studies, though no study has been done to specifically investigate the appropriate dose in children with CS. Various protocols recommend the use of 15mcg/kg, 25mcg/kg, or 0.3mg/m2 (max 1mg) once at 11pm for the overnight test or 1200mcg/kg/day (max 2mg/day) divided Q6 hours for two days (, ). Measurement of a serum dexamethasone level at the same time as cortisol is important to ensure the desired dexamethasone level has been reached.

If screening labs suggest cortisol excess, it is important to rule out physiologic/non-neoplastic hypercortisolism (previously known as pseudo-Cushing syndrome) (). If suspicion is high, additional testing should be considered, including dexamethasone-CRH test (if available) or DDAVP stimulation test. If results remain inconsistent, close monitoring with repeat physical examination and labs within 3 months should be offered to monitor clinical and biochemical findings while at the same treating causes that may contribute to activation of the hypothalamic-pituitary-adrenal axis ().

Once endogenous CS is confirmed, the next step is to identify the source of hypercortisolemia. ACTH levels are used to guide next steps. Elevated ACTH levels of >20-29pg/mL suggest ACTH-dependent CS while suppressed ACTH is consistent with ACTH-independent CS (). Intermediate values may need further evaluation for both ACTH-dependent and ACTH-independent causes, but most often a non-suppressed ACTH level suggests ACTH-dependent CS, except in the case of mild subclinical hypercortisolemia or cyclical CS.

In cases of ACTH-dependent CS, additional biochemical and imaging studies include pituitary MRI (with and without contrast, pituitary protocol), CRH stimulation test (if available), DDAVP stimulation test and/or high dose dexamethasone suppression test. Corticotroph PitNETs are often shown as hypo-enhancing microadenomas in pituitary MRI (), but a normal/negative MRI may be seen in up to 30% of patients (). In cases of normal MRI or biochemical testing inconsistent with pituitary source, bilateral inferior petrosal sinus sampling (BIPSS) is the gold standard in diagnosing or ruling out CD. Non-invasive strategies are described if BIPSS is not feasible and/or not available (). In our pediatric patients, all patients who showed suppression to high dose dexamethasone administration and stimulation to CRH/DDAVP consistent with a pituitary source, had CD irrespective of imaging findings ().

For patients suspected to have ECS, further evaluation should include imaging of the neck, chest, abdomen, and pelvis with thin cuts as carcinoids can be small in diameter. Chest imaging is preferably done with CT due to higher accuracy in the lung parenchyma, but some centers use MRI for abdominal/pelvic imaging to reduce radiation. Nuclear imaging, preferably with Ga-68 DOTATATE or, if not available or negative, with 18F-FDG PET, may identify some of these ectopic sources (, , ). If a lesion found on imaging studies is suspicious but not convincing, one may attempt venous sampling close to the possible lesion for measurement of CRH and/or ACTH and compare the levels to a peripheral source (). If a gradient is reported then this may further support the diagnosis of ectopic tumor (). Other markers of potential interest in these cases include chromogranin A and CRH, which may be helpful in the follow-up of patients. Patients with ECS may present with pituitary hyperplasia if CRH is co-secreted, which should be considered when interpreting the imaging and biochemical results.

When ACTH-independent CS is suspected, imaging of the adrenals is the best next step. Imaging can be preferably with CT since it has good accuracy for lesions <1cm and less artifacts due to motion, but MRI may be an alternative to avoid radiation exposure. Ultrasound however is not accurate in identifying adrenal lesions other than large adrenocortical tumors (). In ACTH-independent CS, it is important to review the anatomy of both adrenals; noting a unilateral lesion with atrophy of the contralateral adrenal supports the diagnosis of unilateral disease, whereas bilateral symmetrical adrenal enlargement or bilateral normal appearing adrenals suggests bilateral disease (). In case of bilateral micronodular adrenocortical disease, adrenal anatomy is often read as normal or sometimes asymmetric appearance of the contour of the adrenals described as “beads on a string” may be apparent (). When bilateral adrenocortical disease is suspected, confirmation of the diagnosis prior to proceeding with surgical intervention involves the performance of Liddle’s test (). The paradoxical increase of urinary free cortisol or 17-hydroxy steroids with increasing doses of dexamethasone is pathognomonic for PPNAD ().

Surgical intervention is the first line of treatment in all types of CS whenever the source is identified. In patients with CD, transsphenoidal resection of the pituitary tumor is the preferred approach. Endoscopic or microscopic approaches have been attempted. A recent meta-analysis has not showed significant differences in the remission rate between the two approaches overall, but endoscopic approach may be preferrable in macroadenomas (, ). In very young patients, pneumatization of the sphenoid sinus may be incomplete and the surgical approach more be more difficult but transsphenoidal access is still possible (). In rare cases of very young children with pituitary lesions or in giant complex pituitary tumors, the transcranial approach may be considered (). Remission is defined as postoperative nadir cortisol levels of <2-5mcg/dl and early postoperative hypocortisolemia is a sensitive marker of durable remission (, ). In cases of non-remission patients may be managed with immediate reoperation and partial hypophysectomy (). In the pediatric cohorts the remission rate after surgery ranges from 62 to 98% depending on the cohort and the criteria used (, -).

In cases of ACTH-independent CS, bilateral or unilateral adrenalectomy is recommended depending on the underlying cause (). Although unilateral adrenalectomy has been suggested in cases of bilateral macronodular adrenocortical disease, data on unilateral adrenalectomy in micronodular adrenocortical disease are not clear (). ECS should primarily be managed with surgical resection.

In cases of persistent CD after surgery, medical therapy or radiation should be considered. At this time, no medical therapy for CS in the pediatric population has been approved by the FDA in the US and all treatments are considered as off-label use, but ketoconazole is approved by the European Medical Association for children >12 years of age. Medical therapies are divided in those directed to adrenal steroidogenesis, to pituitary tumor function, or to peripheral glucocorticoid action. Most commonly, steroidogenesis inhibitors are considered first line as they are more potent and faster acting. Of these, ketoconazole, metyrapone, osilodrostat, levo-ketoconazole and others have been used (). Radiation therapy could be considered as an alternative second-line treatment but requires medical management and close monitoring until the radiation effect is apparent (, ). Finally, bilateral adrenalectomy is reserved for cases of severe CS persistent despite surgical or medical intervention. This is followed by lifelong adrenal insufficiency and patients should be monitored for the risk of Nelson syndrome ().

After successful surgical management, patients experience adrenal insufficiency. In CD the median duration of adrenal insufficiency is almost 12 months (). Additionally, management of patients after remission of CS should also target symptoms of glucocorticoid withdrawal which may require supraphysiologic doses of glucocorticoids for a period of time and slow tapering to physiologic levels (). After recovery of the axis, regular screening for possible recurrence should be offered. Long term recurrence has been reported in 8-20% of pediatric patients with CD after initial postoperative remission (). Screening for recurrence should be done preferably as in adults with two midnight or late-night salivary cortisol levels or with overnight dexamethasone suppression test annually ().