Cushing's Disease in Children: A Review
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.287677
Source of Support: None, Conflict of Interest: None
Keywords: Children, Cushing's disease, endoscopic surgery, paediatric age group, protocol of management
Cushing's syndrome (CS) is a constellation of symptoms and signs due to prolonged exposure to excess glucocorticoid hormones. The commonest cause in all age groups is the inadvertent use of steroids, i.e., exogenous CS. Endogenous hypercortisolism can be adrenocorticotrophic hormone (ACTH)-dependent or ACTH-independent. ACTH-dependent CS in the paediatric age group is most often due to an ACTH-secreting pituitary corticotrophic adenoma and is referred to as Cushing's disease (CD). Careful investigations as per protocols, largely adopted from adult studies, help in differentiating the various entities.
Childhood is marked by growth and puberty. As hypercortisolism can adversely affect both the paediatric and adult stages of life, its prompt diagnosis and treatment become absolutely essential.
There are major differences in the childhood CD as compared to the disease in adults. These differences include the variations in gender predisposition, clinical presentation, hormonal and radiological findings, cure rates and long-term outcomes. The past decade has seen immense changes in the diagnostic and therapeutic strategies in the management of paediatric CD, which have been highlighted in this review.
CD is the commonest cause of endogenous CS in children over the age of 7 years and accounts for 75-80% of paediatric CS until the age of 18 years., In children less than 7 years of age, primary adrenal causes (adenoma, adrenocortical carcinoma or bilateral adrenal hyperplasia [primary pigmented nodular adrenocortical disease, macronodular adrenal hyperplasia and Mc-Cune Albright syndrome]) are common. ACTH secreting corticotrophic adenomas comprise about 43% of paediatric pituitary adenomas. CD usually occurs in the pre-adolescent or adolescent years, with the mean age of presentation reported in the age range of 12-14.8 years. A male preponderance (63%) is reported in prepubertal children in contrast to adults, where female patients have a higher prevalence (79%)., The cause for this gender difference in the pre- and post-pubertal age-groups is not clear.
The most common and striking feature in childhood CS is weight gain associated with growth failure, with or without the presence of short stature. These features should alert the paediatrician about the possibility of CS as an underlying pathology. The World Health Organisation Global Database on Child Growth and Malnutrition [https://www.who.int/nutgrowthdb/about/introduction/en/inde × 4.html] has proposed the standard deviation score (SDS) or Z-score to help in interpreting the weight-for-height, height-for-age and weight-for-age indices. For population-based assessment, such as surveys and nutritional surveillance, the Z-score is appropriate for the analysis of anthropometric data. At the individual level, the Z-score is a suitable predictor of malnutrition and health. The Z-score system presents the anthropometric value of height as the number of standard deviations or Z-scores below or above the reference mean or median value of the population under consideration. A fixed Z-score interval implies a fixed height or weight difference for children of a given age.
Z-score (or SD-score) = (observed value - median value of the reference population)/standard deviation value of reference population).
The height SDS is below the mean value in almost all children of CS. The body mass index (BMI) SDS in these children is always above the mean value. This is in contrast to the findings in children with simple obesity, who have an accelerated increase in height. The other manifestations in these children with CS include a moon-like facies, plethora, striae, emotional lability and depression. The skin manifestations include the development of acne, violaceous striae, an easy bruisability and acanthosis nigricans [Figure 1].
The clinical features may be subtle or absent, especially in young children. Due to the insidious onset of the clinical presentation, the parents may not regard these body changes as being pathological. The mean duration between the symptom-onset and the establishment of diagnosis is around 2.5 years. Virilisation with pubic hair development, without pubertal features, like testicular enlargement in boys or breast development in girls, occur due to androgen overproduction. A delayed puberty or an arrested puberty occurs due to gonadotropin suppression. Rarely, prepubertal boys present with gynecomastia. In contrast to the adult patients with CS, the occasional presentations in children may include sleep disruption, muscular weakness, and memory impairment. Due to the usual presence of a microadenoma in children, visual manifestations are rare.
Libuit et al., demonstrated gender differences in clinical features, with a higher mean BMI Z- score and a lower height Z-score in male subjects compared to the female ones; however, no differences in the biochemical parameters like the degree of hypercortisolemia and ACTH levels were found. These observations are similar to the finding made by Storr et al.
CD in children causes long-term morbidity with severe impairment of growth, increase in total body fat, metabolic derangements, osteoporosis and impaired quality of life that may persist after the disease remission. An increased incidence of diabetes mellitus, hypertension, stroke, pituitary apoplexy, steroid psychosis, and osteoporosis leading to vertebral fractures may be seen in paediatric patients with CD, who have a long-term persistence of high serum cortisol levels.
The molecular basis of CD is largely unknown. With newer genetic tools, increasing number of genetic defects are being identified.,, Somatic mutations in the ubiquitin-specific protease-8 (USP8) gene were identified in about 31% of the paediatric patients with corticotropinomas, with a possible association with a high rate of tumour recurrence. Recently, a germline USP8 mutation causing CD along with a constellation of symptoms and signs has been reported. CD can be an initial manifestation of the multiple endocrine neoplasia (MEN1) syndrome due to the MEN1 gene mutation, as seen in 6/238 CD patients, who were screened because of the suggestive family history or other manifestations of MEN1 the syndrome. Only one out of 74 patients with isolated CD had a mutation in the Aryl hydrocarbon receptor Interacting Protein (AIP) gene (1.4%). Loss of function mutations in CABLES1 (Cdk5 and ABL enzyme substrate 1) gene, a negative cell cycle regulator protein, have been identified in a few patients suffering from CD. Apart from these genetic mutations, CD has been reported in association with MEN-2B, with Carney complex due to an inactivating PRKAR1A germline mutation, with McCune Albright syndrome (MAS) due to a gain of function mutation in the guanine nucleotide-binding protein, alpha stimulating (GNAS) gene, with tuberous sclerosis complex (that occurs due to heterozygous mutations in the TSC1 and TSC2 genes), and with DICER1 syndrome due to loss of function mutations in the DICER1 gene leading to ACTH secreting pituitary tumours.,,
Before commencing the biochemical evaluation for the confirmation of CS, a thorough history to exclude the prolonged use of oral/topical steroid therapy is mandatory. Testing for endogenous hypercortisolism should be done in children with an increasing weight and a decreasing height percentile, as this combination of signs has a high sensitivity and specificity. The algorithm of evaluation is similar to the one that is followed in adults [Figure 2]. The first step is to confirm the presence of endogenous hypercortisolism (by 8.00 AM serum cortisol estimation), followed by determination of the aetiology.
Establishment of CS
The Endocrine Society Clinical Practice Guidelines for the diagnosis of CS recommends one of the following tests. An abnormal initial test must be followed by another of these recommended tests [Figure 2] and [Table 1].
1. The ONDST is done on an outpatient basis and often used as the initial test to exclude CS. A dose of 10 mcg/kg of dexamethasone for children weighing <40 kg, and a standard dose of 1 mg of dexamethasone for children >40 kg, is given within the time range of 11 PM-to-midnight, with the measurement of serum cortisol being done between 8-9AM the next morning. As it is a screening test, a cut-off serum cortisol value of 50 nmol/l, with a high sensitivity of 95%, is used. However, no data exists pertaining to the specificity of ONDST in the paediatric population. An unsuppressed cortisol after an ONDST is followed by a LDDST for confirmation
2. LDDST is based on the principle that dexamethasone will suppress ACTH, leading to suppressed serum cortisol levels in normal subjects; whereas, in patients with autonomous cortisol production, cortisol will not be suppressed. This test involves the measurement of serum cortisol or UFC before and after oral dexamethasone (30 μg/kg/day, in divided doses every 6th hourly for 2 days for children weighing <40 kg, or 0.5 mg every 6th hourly for 2 days in children weighing >40 kg) administration. The attainment of serum cortisol level greater than 50 nmol/l is diagnostic of CS. In a study of 48 paediatric patients less than 20 years of age, a sensitivity of 100% was seen when a cut-off value of serum cortisol level of 50 nmol/l was used, which decreased to 94% when the cut-off value of serum cortisol level was increased to 88 nmol/l. Another study using a cut-off value of 38 nmol/l, obtained a sensitivity and specificity of 91%, respectively, to diagnose the presence of CD in obese children with CS and pseudo-Cushing syndrome states
During the performance of dexamethasone suppression test, care must be taken to ensure that the child is not on any anticonvulsants like phenytoin, phenobarbitone or carbamazepine. These medications may alter the metabolism and levels of dexamethasone, and thus, the test results
3. The 24-hour UFC is an easy, non-invasive investigation, that can be carried out at home and the values are not affected by the cortisol binding globulin (CBG) levels. The adequacy of collection must be ensured by the 24-hour urinary creatinine excretion. A study by Batista et al., showed a high sensitivity of 88% and a specificity of 90% of this test in paediatric patients using the cut-off value of UFC of 193 nmol/day. This was confirmed by a recent study with 47 CS patients and 19 control subjects, where the 24-hour UFC adjusted for body surface area showed a sensitivity of 89% with a 100% specificity. The cut-off levels varied according to the assay used for the assessment (normal value <240 nmol/l for radioimmunoassay, 40-340 nmol/l for immunoassay and <124 nmol/l for the liquid chromatography-mass spectrometry method). The Endocrine Society guideline recommends at least two 24-hour collections of urinary samples. This is because serial measurements enhance the accuracy of the test, as an increased cortisol excretion may be an inconsistent finding in children. However, ensuring a proper 24-hour urine collection may be challenging in younger children
4. a. Loss of diurnal rhythm of variation in the serum cortisol level is the earliest biochemical marker of endogenous hypercortisolism. Normally, cortisol reaches the highest level between 7 to 9 AM (140–690 nmol/L)) and is at the lowest level at midnight, being half of the morning value. A single midnight serum cortisol level of 121 nmol/l in children has a sensitivity of 99%, a specificity of 100% and a positive predictive value of 100% for establishing the diagnosis of CS. Catheterization prior to the sample collection is essential due to the sleeping state of children at that time
b. The late night salivary cortisol estimation is another simple, non-invasive test, based on the loss of diurnal rhythm, for differentiating CS from primary obesity. This test is widely used in adults. Either the passive drooling of saliva in children is collected into a tube, or a salivette is placed in the mouth and chewing is encouraged for 1-2 minutes to enhance the salivary flow. The sensitivity and specificity pertaining to the test of 100% and 95.2%, respectively, have been demonstrated in children.
Determining the etiology
The establishment of hypercortisolism is followed by tests to identify the etiology based on plasma ACTH levels, provided all requisite precautions were taken during the sample collection and processing. A suppressed basal (8 AM) plasma ACTH level of less than <1.1 pmol/l suggests ACTH-independent CS, most likely an adrenal pathology. A value of 6 pmol/l gave the test a specificity of 100% for CD, with a sensitivity of 70%. An ACTH level of 1.1 pmol/l or more indicates ACTH dependent CS, either from the pituitary gland, or rarely, from an ectopic source like a carcinoid. Children with ACTH levels between 2 pmol/l and 6 pmol/l are further evaluated by the CRH (corticotrophin releasing hormone) stimulation test, to differentiate a pituitary lesion from an ectopic ACTH source. A corticotrophin increase of 35% from the baseline at 15 and 30 minutes after the intravenous administration of CRH (1 μg/kg, maximum 100 μg, sampling at the basal level and after 15, 30 and 45 minutes), gave a sensitivity of 81%; and, a cortisol increase of >20% from the baseline level at 30 and 45 minutes after ovine CRH (oCRH) administration gave a sensitivity of 74% in identifying patients with CD. An ectopic CS is exceedingly rare in children; and also, due to the non-availability of CRH, the latter test is not routinely performed in India.
Localisation of the adenoma
Once the ACTH-dependent CS is proven, the next step is to localise the lesion in the pituitary gland by imaging techniques. Thin-section, high-resolution, contrast-enhanced, magnetic resonance imaging (MRI) of the sella and suprasellar region is the standard imaging modality used. The corticotroph adenomas in children, unlike other pituitary adenomas, are most often (>90%) microadenomas. It is, therefore, important to identify the gland as being separate from the lesion, if possible. Initially, a pre-contrast, thin slice, high resolution T1- and T2-weighted spin echo coronal and sagittal sections are acquired. Both the dynamic and routine post-contrast images and delayed scanning after 30-60 minutes may be combined in one study for performing an optimum imaging of a pituitary microadenoma [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8], [Figure 9].
The post-contrast spoiled gradient recalled acquisition MRI scans are superior to the conventional dynamic spin echo MRI images for detection of pituitary microadenomas. A corticotroph adenoma shows up as a hypointense lesion in the early phase, which becomes less hypointense in subsequent dynamic cycles.
The dynamic contrast MRI has been proven to also be a useful imaging tool in the evaluation of pituitary adenomas. A 3-dimensional Fourier transformation gradient echo or fast turbo spin echo sequence (TSE) may be used for the dynamic MRI study. After a bolus injection of intravenous gadolinium, four-to-six consecutive sets of three images are obtained in the coronal plane every 10 seconds. Sequential enhancement occurs in the pituitary stalk, in the pituitary tuft (the junction point of the stalk and gland), and the entire anterior lobe of the pituitary gland. Within 30-60 seconds, the entire gland shows a homogenous enhancement. The maximum image contrast between the normal pituitary tissue and the microadenoma is attained about 30-60 seconds after the bolus injection of the intravenous contrast, with microadenoma appearing as relatively non-enhancing lesion within an intensely enhancing pituitary gland. The peak enhancement of the pituitary adenoma occurs at 60-200 seconds, persisting for a longer duration. This usually follows the uniform enhancement of the normal pituitary gland. A delayed scan (30-60 minutes after contrast injection) may demonstrate a reversal of the image contrast initially obtained on the dynamic scanning. This is because the contrast from the normal pituitary gland fades, but diffuses into the microadenoma, which stands out as a hyperintense focus. Dynamic sagittal images may be incorporated with the routine coronal images to increase the overall detection rate of pituitary microadenomas.
In general, lower rates of microadenoma detection have been reported in children. MRI alone may fail to correctly localise the laterality of the adenoma, or in some cases, even in diagnosing CD [Table 2].,,,, A low detection rate in children is due to the smaller size of the adenoma (usually <5 mm) within a small pituitary gland. A sagittal reconstructed image of the computed tomographic (CT) scan is useful to assess the degree of pneumatization of the sella and in the assessment of the dimensions of the surgical trajectory through the anterior nares, the nasal cavity, the sphenoid sinus and the sella. The postoperative assessment of the extent of tumour excision and any evidence of surgical site hematoma formation may also be detected by a CT scan image. The CT scan of the adrenal glands is useful to distinguish between CD and adrenal causes of Cushing syndrome. This mainly refers to unilateral adrenal tumours. In the case of bilateral adrenal adenomas or hyperplasia, unequivocally establishing the diagnosis with a CT scan may become difficult. A CT or MRI scan of the neck, chest, abdomen, and pelvis may detect an ectopic source of ACTH production. Labelled octreotide scanning, positron-emission tomography (PET), and venous sampling help in locating an ectopic ACTH source.18 Fluorodeoxyglucose positron emission tomography (FDG PET) has been shown to be inferior to a CT and MRI scan in the detection of ectopic ACTH sources.
Bilateral simultaneous inferior petrosal sinus sampling (BIPSS) is a highly specialised technique, and is routinely used in adults, to distinguish CD from ectopic ACTH syndrome, and for lateralisation of the pituitary microadenoma. Since an ectopic ACTH syndrome is very rare in children, the primary utility of BIPSS in the paediatric age group is for the localisation of the pituitary microadenoma. It is done only in cases of ACTH-dependent CS where MRI does not reveal a lesion in the pituitary gland. Very young children may require general anaesthesia during the procedure. The diagnosis of CD depends upon the demonstration of the ratio of inferior petrosal sinus-to-peripheral (IPS/P) ACTH level >2 at baseline, and >3 after stimulation with CRH. Sampling lateralization, defined by an interpetrosal sampling gradient (IPSG) >1.4 predicts the side of the tumour. Desmopressin is commonly administered in adult patients as a cheaper and an easily available alternative to CRH. Its use in children has been validated recently in a cohort of 16 children after intravenous injection of 10 mcg of desmopressin. The best sensitivity was obtained after 3 minutes of stimulation and did not increase at further intervals. On an overall basis, although a variable sensitivity of BIPSS has been reported in literature, studies have shown a better localisation of the ACTH-producing adenoma with inferior petrosal sinus sampling as compared to the pituitary imaging [Table 2].
Assessment of the extent of spread of pituitary adenoma in paediatric CD
A. Hardy's classification modified by Wilson (assesses infra- and suprasellar extension of the tumour; [Figure 10])Grade I: Sella normal; tumour less than 10 mm; Grade II: Sella enlarged, intact but with a bulging margin, tumour greater than or equal to 10 mm; Grade III: Local perforation of sellar floor; Grade IV: Diffuse sellar floor destruction
Stage 0: No suprasellar extension; Stage A: Extension to the suprasellar cistern; Stage B: Obliteration of the third ventricle recesses; Stage C: Third ventricle grossly displaced; Stage D: Intracranial extension; Stage E: Extension into or beneath the cavernous sinus
B. Knosp grade (assesses tumour encroachment or invasion into the cavernous sinus) [Figure 11]
Grade 0: The tumour is restricted to the tangential line drawn along the medial margin of the supra- and intra-cavernous internal carotid artery (ICA) ; Grade 1: The tumour extends beyond the medial margin of ICA, but not beyond the midpoint (also known as the intercarotid line) of the supra- and intra-cavernous ICA; Grade 2: The tumour extends beyond the intercarotid line, but not beyond the lateral margins of the supra- and intracavernous ICA; Grade 3: The tumour extends beyond the lateral margin of the supra- and intra-cavernous ICA; and, Grade 4: Total encasement of ICA is present.
C. SIPAP classification (assesses the five directions of tumour extension beyond the sella or its penetration into adjacent structures of the sellar region [suprasellar, infrasellar, parasellar, anterior and posterior]).
Suprasellar: Grade 0: No bulging of adenoma into the suprasellar space is seen; Grade 1: The adenoma bulges upwards into the suprasellar cistern but without reaching the optic chiasma; Grade 2: The tumour reaches the optic chiasma but without displacing it. Grade 3: The tumour displaces and stretches the chiasma; Grade 4: Obstructive hydrocephalus caused by tumour extension is present [Figure 12]
Infrasellar: Grade 0: Intact floor of sella; Grade 1: Focal tumour bulging with extension beyond the floor of the sella; Grade 2: Tumour extension beyond the sphenoid sinus [Figure 13]
Parasellar: 5 grades, as per the Knosp grading [Figure 11]
Anterior: Grade 0: Tumour within the sella but is not having an anterior extension beyond a sagittal line perpendicular to the tuberculum sellae; Grade 1: Tumour having an anterior cranial fossa extension beyond the line perpendicular to the tuberculum sellae [Figure 14]
Posterior: Grade 0: The tumour extension may be suprasellar but without postero-inferior growth behind the clivus, when evaluated in the sagittal plane. Grade I : The tumour growth is behind and inferior to the dorsum sellae or clivus into the subarachnoid space in front of the pons. There may be a complete destruction of the dorsum sellae [Figure 15].
Treatment of CD
Treatment of CD includes the management of comorbidities apart from attainment of a definitive cure by surgery or radiation therapy, or a combination of these two modalities. Attention should be paid towards optimisation of blood pressure, blood glucose, electrolytes, calcium and vitamin D status and bone health. A multidisciplinary team approach consisting of paediatric endocrinologists, interventional radiologists, neurosurgeons and radiotherapists is recommended.
The various surgical approaches to the functioning cortisol producing pituitary adenoma are summarized in [Table 3]. In order to normalize the increased serum cortisol levels due to the hyperfunctioning pituitary gland, a radical excision of the adenoma should be the norm. In case that is not possible, adjunctive radiotherapy or chemotherapy may be utilized to address the residual tumour tissue.
Transsphenoidal surgery (TSS) with selective adenomectomy is the first-line therapy for paediatric CD. TSS is safe and effective, leaves behind the normal pituitary gland in situ, and decreases the risk of post-surgical hypopituitarism. There are 2 different techniques of TSS: microscopic and endoscopic [Figure 16] and [Figure 17]. Endonasal transsphenoidal endoscopic pituitary surgery (ETES) is less invasive and more advantageous, being increasingly used for pituitary tumours in the recent years as an alternative to the traditional microscopic sublabial approach. A study of 6 paediatric patients showed a remission in 5/6 (83%) patient after ETES with lower complication rates. It provides a better anatomical access, avoids the painful incision in the upper gum and damage to the developing tooth buds and causes significantly fewer nasal symptoms, achieving equivalent rates of tumour resection with a shorter hospital stay and less cost. The transcranial approach is now used only when TSS is not possible or the tumour size and location demands this approach. Regardless of the surgical approach employed, all patients should be started on hydrocortisone at 100 mg/m2 on the day of surgery, which is tapered in the early postoperative period. However, TSS in children is fraught with technical difficulties due to the small size of the adenoma and the pituitary fossa, along with lack of pneumatization of the sphenoid sinus. Early postoperative complications include cerebrospinal fluid (CSF) leak, transient diabetes insipidus (DI), hypopituitarism, and rarely, syndrome of inappropriate anti diuretic hormone (SIADH) and visual loss.
Special considerations during surgery in pediatric patients
Children have smaller nasal apertures. The sublabial route may provide a wider corridor to access the pituitary adenoma than the direct transnasal approach and permits the use of the same nasal speculum, as is used in adult patients. The blades of Hardy's nasal speculum should not be opened too wide within the sphenoid sinus, due to the higher chances of injuring the carotid artery in children, as the surrounding bones are thin and the sphenoid sinus is narrower. During the sellar dural opening phase, especially in microadenomas, the intercavernous sinuses at the anterior and posterior limits or the cavernous sinuses laterally should be meticulously preserved, since a breach in any of these venous sinuses may cause excessive bleeding. In younger children, any significant bleeding may lead to hypotension during surgery. The paranasal air sinuses in paediatric patients may not be sufficiently pneumatised during surgery. This may lead to a non-pneumatised conchal type of sphenoid sinus that may have to be drilled under neuronavigational guidance during surgery to access the sellar dura. In children, pituitary hyperplasia or multiple adenomas may be responsible for the lack of recovery after transsphenoidal surgery. If the side of the pituitary gland harbouring the adenoma may be localised using the inferior petrosal sinus sampling, then a hemihypophysectomy may be performed during repeat surgery to bring about a more lasting normalisation of the hormonal levels. Completeness of excision may be ensured by cauterizing the tumour bed with absolute alcohol or excising a part of the normal gland surrounding the microadenoma. A histological confirmation of the extent of adenoma removal and of the absence of multiple adenomas or foci of hyperplasia, may be obtained during surgery. Despite the presence of a small sella in the presence of a microadenoma, the risk of cerebrospinal leak following a suprasellar arachnoidal breach exists because of the arachnoidal outpouching into the sella during surgery. Thus, packing of the sella and sphenoid sinus with fat is mandatory in every case of TSS, irrespective of whether a micro- or macroadenoma is present.,,,,
Outcome of surgery
The outcome of TSS depends upon the definition of cure or remission used. There are no defined criteria for remission, and different centres utilise different parameters. The most widely used criterion is the postoperative 8 AM cortisol level <50 nmol/l, done between postoperative day 5 and 2 weeks, assessed after discontinuing steroid use for at least 12 hours.
When the criterion of postoperative cortisol level <50 nmol/l is used to define cure, the recurrence rate is very low. Batista et al., (in 72 children, with a follow-up duration of 24-120 months) studied the role of estimating postoperative morning cortisol, plasma ACTH and UFC levels, starting after the third postoperative day; and, performing an ovine corticotropin releasing hormone (oCRH) stimulated cortisol and ACTH assessment on day 10, in predicting the cure following the TSS. The study concluded that UFC is a poor predictor and that dynamic testing with the combined use of postoperative oCRH test and the assessment of morning ACTH levels are superior to the evaluation of morning serum cortisol level alone for predicting recurrence.
Other factors that determine the surgical outcome include an older age at the time of development of symptoms, a younger age at surgery, no tumour identification at surgery, lack of histological confirmation, a larger tumour diameter, dural or cavernous sinus invasion, higher postoperative basal and oCRH stimulated cortisol and ACTH levels, an early recovery of the hypothalamo-pituitary axis (HPA) after the TSS, and mutations in the USP8 gene in the tumour tissue. Apart from these factors, the availability of better quality MRI scans, the presence of a neuro-endovascular set up for conduction of BIPSS, and the ready access to neurosurgical expertise, significantly affect the success rates. The use of intraoperative direct micro-cytology and micro-biopsy have been shown to improve the detection and selective removal of minute ACTH adenomas, and the achieving of remission rates of >90%. The remission rate with TSS is variable from 50% to >90%. Recurrence rates in the range of 8-16% have been reported [Table 4],,,, Therefore, a lifelong follow-up is required, as recurrence has been reported in adults even up to 15 years after an initial clinical and endocrinal cure has been achieved with undetectable post-surgical cortisol levels.
Second line therapy
A failed TSS, or the tumour recurrence after an initial remission, may be treated with a repeat TSS, pituitary radiotherapy (RT), long-term medical treatment, and/or bilateral adrenalectomy. The choice of treatment is individualised for each case.
RT is an effective second-line treatment for paediatric CD, with a failed TSS, though limited studies have reported on the efficacy and long-term outcome in children [Table 5].,, RT is available through external beam RT, stereotactic RT or radiosurgical approaches. Conventional fractionated RT have better cure rates in children (50-100%) against 56-83% in adults. There is a more rapid onset of action (mean time to cure: 0.75-2.86 years) than in adults (range: 1.5-5 years). Another recent study reported on better efficacy and lower adverse effect profile of stereotactic radiosurgery (SRS) in 24 paediatric CD patients, who were administered this modality either as first-line treatment (7 patients) or after failed TSS. An endocrinal remission rate of 80% at a median time of 12 months, and a tumour control in 87.5% cases was obtained at a median follow up of 46 months. New-onset multiple pituitary hormonal deficiencies occurred in 20.8% patients at a median interval of 18 months, and recurrence of the tumour occurred in 21% (4/24) patients after a median interval of 9.5 months.
The potential complications in children and adults receiving cranial irradiation include the development of hypopituitarism, cranial neuropathies, cerebrovascular events, second intracranial tumour formation and cognitive effects. Patients may develop growth hormone deficiency as well as hypogonadism with or without other anterior pituitary hormonal deficiencies, and require a close and regular assessment. However, recovery of the somatotrophic axis has been documented, and therefore, a sequential reassessment is essential, as shown by Chan et al.
Medical therapy is used to control the hypercortisolism in very sick children who are not fit for surgery/RT, in patients with life threatening complications of CD, while preparing for surgery, or as a bridging therapy until the effects of RT are seen.
The drugs are categorised as adrenal directed therapy (steroidogenesis inhibitors like ketoconazole, metyrapone, mitotane, etomidate) and pituitary directed therapy (dopamine or somatostatin analogs like cabergoline and pasireotide, glucocorticoids or progesterone antagonists like mifepristone).
The mechanism of action of these drugs has been well summarised by Stratakis. Mitotane inhibits 11-β-hydroxylase and cholesterol side chain cleavage enzymes, thereby impairing corticosteroids synthesis and destroying adrenocortical cells that secrete cortisol. Other adrenal enzyme inhibitors include aminoglutethimide, metyrapone, trilostane, and ketoconazole, which may be used alone or in combination to control hypercortisolism. Aminoglutethimide inhibits the conversion of cholesterol to pregnenolone in the adrenal cortex, inhibiting the synthesis of cortisol, aldosterone, and androgens. Metyrapone acts by inhibiting the conversion of 11-deoxycortisol to cortisol. Trilostane inhibits the conversion of pregnenolone to progesterone. Ketoconazole is also helpful in inhibiting adrenal steroidogenesis.
Studies in adult patients report a success rate of approximately 25-30% with pituitary directed therapies. Data regarding the success of these therapies in a large group of children is lacking. Medical therapies are less effective, do not allow for cure of the disease, are associated with various side effects limiting their long-term usage, and recurrence of hypercortisolism shortly after the drug withdrawal is common.
This option is used in life threatening situations or when TSS or RT are not possible or available. Apart from the lifelong need for glucocorticoid and mineralocorticoid replacement, Nelson syndrome (NS), due to the growth of a corticotropic adenoma may occur months or years after adrenal surgery in some patients with CD as a result of removing both adrenal glands, resulting in the lowering of serum cortisol levels. In Nelson's syndrome, the pituitary tumour enlarges and releases ACTH by a feedback mechanism. This often causes visual loss, hypopituitarism and headache. A dark skin pigmentation (caused by high levels of circulating ACTH that bind to the melanocortin 1 receptor on the surface of dermal melanocytes to produce melanin) may also result due to the excessive release of ACTH.
This syndrome occurs more commonly in children (25-75%) than in adults (8-43%). A long-term annual monitoring of the patient with MRI brain and the assessment of plasma ACTH levels is required to diagnose NS. NS often requires treatment with pituitary surgery or RT.
After surgery, the resultant adrenal insufficiency due to impairment of the hypothalamo-pituitary-adrenal axis requires the administration of stress doses of cortisol. This is subsequently tapered to a physiologic replacement dose. At follow-up visit after 3-6 months, the adrenocortical function should be assessed with a 1-hour ACTH test.
A single prospective study in 40 children (34 with CD) showed an impaired quality of life (QoL) pre- and post-one-year cure of CS. Despite successful treatment with remission, residual morbidities lead to an impaired QoL in children, as occurs in adults. These include the negative effects of CD on attainment of the final height, body-mass index (BMI), body composition, bone health, and on neuropsychiatric and cognitive functions. Growth failure occurs due to the inhibitory effects of hypercortisolism on growth hormone dynamics and growth plate physiology, bone age acceleration occurs due to increased androgens, and osteoporosis and vertebral fractures, may also be seen. Further, many children fail to attain catch-up in growth with their age-matched normal cohort of subjects after their remission, due to the growth hormone (GH) deficiency resulting from the TSS or RT. GH therapy, along with gonadotropin releasing hormone (GnRH) analogues or with aromatase inhibitors, has been tried with success to enable an enhancement in the catch-up growth, and to attain the adult height within the range of the target height.
The body mass index remains elevated with an increase in total body and visceral fat, with an increased risk of obesity and the presence of the metabolic syndrome. This scenario is the same as seen in adults, despite the attainment of eucortisolism in these paediatric patients.
Osteoporosis results from the direct effects of hypercortisolism on the bone turnover, as well as due to hypogonadism, GH deficiency, decreased calcium absorption, and renal calcium reabsorption. With treatment, the bone mass has been shown to recover, with improvement in the lumbar spine, total hip and femoral neck bone mineral density and bone mineral apparent density standard deviation scores, when compared to the preoperative scores. This phenomenon is demonstrable within 1-3 years of remission of hypercortisolism., Partial or complete pituitary hormonal deficiencies can occur variably as a result of TSS or radiation. Cerebral atrophy, with a decrease in cognitive function that persists after a cure has been obtained, has been reported.
Our published surgical experience in paediatric patients with CD
In a previously published study, twenty-seven children presented with endogenous Cushing's syndrome in our institution between 1990 and 2004. Fourteen of these paediatric patients had CD, of which 4 either refused surgery or were lost to follow-up. A total of 10 children with CD (age range, 12-17 years; median age, 15 years; male-female ratio, 7:3; 2.5% of total pituitary tumours operated during that period) operated between January 1991 and July 2004, of the 406 patients operated for pituitary adenomas, were included in this study. During the same period, overall, 37 (adult; 27 excluded) patients with CD were operated. Dynamic magnetic resonance (where the adenoma had a delayed contrast uptake when compared to the rest of the pituitary gland) was helpful in diagnosing 2 patients having a microadenoma where the routine MRI sequences were equivocal. BIPSS was performed in 2 patients in whom the MRI revealed a normal gland. Transsphenoidal surgery was performed in 9 patients, whereas in one patient with a macroadenoma with a conchal sphenoid sinus, a pterional, transsylvian approach was adapted. The serum cortisol level was estimated every 3 to 6 months by an ACTH stimulation test to determine whether or not pituitary reserve was adequate, that is, the hypothalamic-pituitary axis had recovered. ACTH (synacthen 0.25 mg) was administered intramuscularly, and serum cortisol levels were estimated at 0 and 60 minutes. A 60-minute level greater than 500 to 540 nmol/L was considered as significant and indicative of an adequate pituitary reserve when maintenance steroids could be stopped.
Weight gain with centripetal obesity (n = 10), easy bruisability (n = 7), short stature (n = 5), abdominal stria (n = 6), skin pigmentation (n = 6), proximal muscle weakness (n = 5), hypertension (n = 4), decreased vision (n = 3), headache (n = 3), amenorrhoea (n = 2), and diabetes mellitus (n = 2) were the commonest presenting features. HDDST showed a variable suppression: 3 of 10 patients were more than 50% suppressed, 6 were suppressed between 16% and 50%, and one did not suppress at all, suggesting that the test cannot be used as an unequivocal parameter to diagnose CD of pituitary origin. The median serum ACTH level was 72 pg/mL. MRI showed a microadenoma in 5, a macroadenoma in 3, and a normal gland in 2 patients. There was no variation in the dexamethasone suppressibility with the size of the adenoma. All patients with macroadenoma had a suprasellar extension less than 10 mm. There was no operative mortality. Postoperative CSF leak occurred in 2 patients with a microadenoma with arachnoidal pouching. It was managed using a closed lumbar drain for 3 days with oral acetazolamide (250 mg QID) administration. Clinical remission was achieved in 7 of 10 operated patients during the next 6 months of follow-up. Of those 3 patients in whom surgery failed to achieve remission, one had a microadenoma, and the other 2, a normal pituitary gland on MRI at initial evaluation. Their IPSS had suggested the presence of central CD, and the tumour was seen at surgery and resected. These 3 patients had neither clinical nor biochemical remission; 2 underwent bilateral adrenalectomy and one received radiotherapy.
In 2 patients where clinical remission was achieved, postoperative basal serum cortisol (BSC) could not be done. Postoperative BSC was less than 50 nmol/L in 2 (25%) out of 8 patients and remained elevated in 6 others. Remission was achieved in both the patients with postoperative BSC less than 50 nmol/L and in 3 of 6 with elevated levels. Over the median follow-up of 82 months (range, 24-120 months), of the 7 patients who initially remitted, there was recurrence of disease in 3 patients (42.8%) after a median interval of 5 years (range, 3-8 years). Among these 3 patients, one had recurrence of disease after 8 years (where postoperative BSC could not be done). Two other patients had recurrence after 3 and 5 years (their postoperative basal cortisol levels were 27 and 90 nmol/L, respectively). All 3 patients had presented with a microadenoma initially, and only one patient had shown more than 50% suppressibility during the conduction of the HDDST. One patient with recurrence of the disease was managed with radiotherapy; the second was managed with resurgery followed by radiotherapy; and the third patient was lost to follow-up. Of the original 10 patients, 9 remain in active follow-up, the median follow-up being 64 months (range, 10- 120 months). Only a single patient had pituitary dysfunction in the form of hypothyroidism and was placed on thyroxine supplementation; none had hypogonadism. In CD, a number of recurrences have been observed over greater than 5 years. The paediatric population with CD is especially prone to a high risk of recurrence. Thus, regular follow-up visits and repeated biochemical evaluations are mandatory. An immediate (within the first or second week following surgery) postoperative serum cortisol estimation has often been used to predict late biochemical relapses.
Using this test, it has been found that the early presence of low hormonal levels are not specific predictors of success. This was well exemplified in our series where 2 patients developed relapse after an early remission, and another patient, who had high serum cortisol levels on the initial testing, showed a complete normalization of values at follow-up. High initial levels, however, have shown a higher correlation with recurrence. If clinical or biochemical recurrence is observed, the various options available are resurgery, medical suppression of serum cortisol, or radiotherapy. However, radiotherapy may cause a delayed growth retardation, cognitive dysfunction, hypopituitarism, or development of other tumours.
Three of our patients had persistently high basal cortisol levels despite surgery. One of them underwent a bilateral adrenalectomy directly. The second patient who showed multiple small adenomas and persistently high serum cortisol levels was administered ketoconazole, which blocks the synthesis of cortisol at the adrenal cortical level. He was also been recommended a bilateral adrenalectomy because none of the recent imaging had localized an adenoma. Medical adrenalectomy using mitotane was an option, but its toxicity precludes its frequent use – especially in children. The third patient was advised conventional fractional radiotherapy 45 Gy to the pituitary region. On reviewing other series of paediatric CD, it was seen that initial remission was achieved in the range of 60% to 100%. In the present series, an initial remission rate of 70% was observed. However, the recurrence rates in the various series including ours, have been rather high and have ranged from 6-28%.,,, This higher recurrence rate among paediatric patients is similar to the recurrence rate that has been noted in adult patients with CD in a study by Saini et al., and by Sarkar et al.,
CS is rare in children but causes significant lifelong morbidity, especially affecting growth and puberty in children. A set of investigations carried out according to a formal protocol helps in the confirmation of diagnosis and in the recognition of etiology. Management should be carried out in an experienced centre with access to a specialised multidisciplinary team. The use of minimally invasive surgical techniques to selectively remove the tumour by an experienced neurosurgeon helps in improving outcome, with a higher rate of biochemical and radiological remission and preservation of pituitary function.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8], [Figure 9], [Figure 10], [Figure 11], [Figure 12], [Figure 13], [Figure 14], [Figure 15], [Figure 16], [Figure 17]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]