University of Pennsylvania School of Medicine, Childrens Hospital of Philadelphia, Philadelphia, Pennsylvania 19104
Address all correspondence and requests for reprints to: Thomas Moshang, Jr., M.D., Professor of Pediatrics, University of Pennsylvania School of Medicine, Childrens Hospital of Philadelphia, Philadelphia, Pennsylvania 19104.
It was 70 yr ago, before most of the present endocrinologists were even born, when Harvey Cushing first described the finding of basophilic pituitary adenoma as a cause of the disorder that now bears his name (1). If one does a literature search, using the search words "Cushings syndrome,"over 8000 citations are listed, and close to 300 in the year 2002. So, after 70 yr, we are still publishing studies about Cushings disease because, clearly, our diagnostic, treatment, and posttreatment modalities can and need to be refined. This is especially true in the pediatric age range because the incidence of pituitary Cushings disease in children is rare as compared with the occurrence rate in adults. It is necessary after learning from studies related to adult patients to evaluate the extrapolated information to similarly affected children. Estrada et al. (2) reported several years ago the successful outcome using pituitary irradiation to treat adult patients with persistent or recurrent Cushings disease after transsphenoidal surgery. The success in inducing remission of Cushings disease after pituitary irradiation was time dependent, but perhaps the use of ketoconazole to treat the patients until biochemical and clinical cure was, as well, an important factor in terms of management. The time to remission was at least 6 months but was as long as 60 months. Subsequently, following pituitary irradiation of these adult patients, approximately 57% developed GH deficiency, 40% developed gonadotropin deficiency, 16% had TSH deficiency, and one patient demonstrated ACTH deficiency.
In this issue of JCEM, the group from St. Bartholomews Hospital in London, United Kingdom, reports their experience with the treatment of Cushings disease in children from 1983 through 2000 (3). Eighteen children were all treated by transsphenoidal surgery initially. Successful cure of the hypercortisolism after surgery was 66%. The remaining seven patients were subsequently cured with pituitary irradiation, using 45 Gy, a radiation dose similar to, although slightly less than, the dose (50 Gy) used for adult patients in the study by Estrada et al. (2). In the interim, after pituitary irradiation until biochemical and clinical cure (which took up to 2.86 yr), the patients were maintained on medical adrenal therapy: Mitotane [1,1-dichloro-2-(0-chlorophenyl)-2-(p-chlorophenyl)-ethane, opDDD] or ketoconazole and other adjunctive therapy, including metyrapone. After pituitary radiation, all seven children were cured of their Cushings disease. After pituitary irradiation, the children were followed for a mean of 6.9 yr, and 86% (six of seven) demonstrated GH deficiency but no other pituitary hormone deficiency. This groups experience documents not only the successful outcome using pituitary irradiation for retreatment of Cushings disease following surgical failure but also emphasizes the need for experience in managing pediatric Cushings disease. In the United Kingdom and Europe, in general, pediatric subspecialists, including neurosurgeons, are truly localized in a few referral centers. In the United States, there may be as many as 5 or 10 major pediatric centers located within a 50-mile radius, and, therefore, our individual experience is certainly not comparable with this reported study. In fact, the relative infrequent pediatric patient with Cushings disease presents unique problems for the pediatric endocrinologist. These problems include appropriate screening studies, technical procedures, as well as treatment modalities.
Unlike the adult population, in whom the most common cause of Cushings syndrome is due to ACTH excess (whether because of pituitary tumor or hypothalamic dysfunction or ectopic ACTH production), in children the clinical manifestations of Cushings syndrome are frequently due to an adrenal tumor, especially in the young child (4). The clinical manifestations of obesity, plethoric moon-facies, hirsutism, muscle weakness, but especially finding hypertension and/or growth arrest clinically suggest Cushings syndrome. Such patients will normally be sent initially for adrenal imaging (as Willie Sutton said, "I rob banks because thats where the money is."). Adrenal tumors causing Cushings syndrome are not small and easily visualized by ultrasound. The finding of an adrenal tumor by ultrasound obviates the need for the diagnostic tests, including suppression tests, CRH testing, and urine collections. It is useful nevertheless to measure serum concentrations of cortisol, androgens, and steroid precursors, such as 17-OH progesterone and 11-deoxycortisol, before surgical removal of the tumor because these steroids may be useful tumor markers as an indicator of cure or recurrence.
The evaluation of children with the clinical manifestations of Cushings syndrome after exclusion of an adrenal tumor is more complex because many of the diagnostic tests are based on studies in adults, and then extrapolated for children. One major concern for pediatricians and pediatric endocrinologists is the appropriate screening test for obese children (becoming a greater problem in the United States at present) to exclude Cushings syndrome. As an example, 24-h urine collections before and after low-dose dexamethasone (2 mg over 2 d) are not easy and are based on adult studies but often still applied to children. This dose of dexamethasone may be a considerable amount of dexamethasone for a child, even an obese child. Over 25 yr ago, Streeten et al. (5) suggested that a low dose in children is a 5 µg/kg dose (or 20 µg/kg·d). In 42 children with Cushings disease treated with transsphenoidal surgery, 3 of 21 children did not suppress urinary-free cortisol excretion using 20 µg/kg·d dexamethasone over 48 h (i.e. a false negative rate of 14.3%; Ref. 6). Nevertheless, it is now generally standard that this dose be used in suppression testing in children (4). As to the 1-mg dose of dexamethasone for overnight suppression of morning cortisol, Hindmarsh and Brook (7) suggested that the dose that suppressed the morning cortisol was 0.3 mg/m2 in children, having evaluated several doses in 15 normal children.
Given the vagaries of an appropriate low dose of dexamethasone and urine collections in children, the most effective screening test with a high degree of sensitivity is probably a 2400-h serum cortisol (although also not easily obtainable as an outpatient in children). A good alternative, however, is a bedtime salivary cortisol. Using bedtime salivary cortisol determination, 83% of the patients with Cushings syndrome, using a cutoff cortisol value of >27.6 nmol/liter (1 µg/dl), were delineated from normal and obese children, whereas a 2400-h salivary cortisol delineated 93%, using a cortisol value >7.5 nmol/liter (0.27 µg/dl) (8). It is likely that obtaining bedtime saliva samples from children late at night is relatively doable (spitting seems so natural for children), suggesting that this is an easier and good test to screen for Cushings syndrome.
If the screening studies are suggestive of Cushings disease, the diagnostic test of choice of hypercortisolism is finding that repeated (several) 24-h urine samples of urinary-free cortisol excreted over 24 h are consistently elevated. But this simple diagnostic test can be a nightmare for some families, as in children with enuresis. Furthermore, collecting 24-h urine samples in children may not be easy or convenient (what with school and other activities). Nevertheless, the finding of consistently elevated levels of urinary-free cortisol, correcting for body surface (and perhaps needing correction for creatinine excreted in incomplete urine sampling), has a high degree of sensitivity (between 90% and 100%) for hypercortisolism. The cutoff for normal urinary-free cortisol excretion is suggested to be <72 µg (198.7 nmol)/m2/d (8).
Once it is clear that the child has cortisol excess, and it is not due to an adrenal tumor, the localizing of the pituitary tumor and the possible treatment create the next level of difficulty in management for children. The present study by Storr et al. (3) and other studies have indicated that computed tomography (CT) especially, but even magnetic resonance imaging (MRI), often detect less than 25% of Cushings pituitary microadenomas. An earlier series of patients treated with transsphenoidal surgery reported that MRI with gadolinium had a sensitivity of 72% (13 of 18 evaluated) and CT had a sensitivity of 23% (5 of 22) (6). However, in a large series of children studied at the National Institutes of Health (9), MRI scans with gadolinium revealed only a pituitary adenoma in 26 of 50 patients (52%). Pituitary MRI scanning with contrast, as a procedure itself, presents unique problems for children and adolescents. Sedation is generally necessary for young children. In the United States, almost all of our adolescents have dental orthodontia, making it either necessary to remove the metal dental appliances or accept a poor quality scan that even without the artifacts created by the dental braces is of low yield. Dental appliances need to be removed to evaluate the pituitary despite the additional financial cost to parents.
Bilateral inferior petrosal sinus sampling (IPSS) of ACTH following CRH is recommended by many to confirm the lateralization of ACTH overproduction, even if the MRI or CT demonstrates a pituitary adenoma. It is obvious that this procedure, even for experienced pediatric interventional radiologists, is difficult because of the rarity of Cushings disease in children. Conversely, interventional radiologists familiar with this technique attempting this procedure may not be familiar with the problems of sedation and anesthesia in young children. It is of interest that in the present report, two of seven children underwent IPSS preoperatively and both demonstrated a central etiology as well as lateralization, and still failed surgical treatment. Others have reported surgical failures following lateralization of ACTH production using IPSS, as well. In the series of 50 pediatric patients studied at the National Institutes of Health with IPSS, 95% (41 of 53) of the tests were diagnostic of Cushings disease, but only 76% following CRH correctly predicted the lateralization of the microadenoma (9). Other smaller series have reported higher discordant rates from IPSS testing and surgical findings. IPSS is an important technique to localize microadenoma, but, again, experience increases the rate of success but postsurgical failures will nevertheless occur.
The recommended first line of therapy for Cushings disease is transsphenoidal surgery. Transsphenoidal surgery for Cushings disease by experienced pituitary neurosurgeons, using microscopic instruments and magnifying imaging during the procedure, has shown success rates up to 90%. However, in most published reports (4, 6, 9, 10), the cure rates of children following transsphenoidal surgery with Cushings disease is similar to the results in the present study (i.e. approximately 50% to two thirds). But it is important to recognize that in this report 16 of the 18 patients were operated on by the same neurosurgeon, similar to the earlier report from California (6) in whom 41 of 42 patients were operated on by a single neurosurgeon. This latter report stressed the need for experience in the operating neurosurgeon. In the series from Mayo Clinic (10), the cure rate after transsphenoidal surgery was 50% and the discussion focused on potential differences in pediatric Cushings disease and adults that might account for the higher recurrence rate in children. However, the Mayo Clinic study did not indicate whether a single surgeon or different surgeons performed the transsphenoidal surgery during that time span between 1975 and 1990. The more recent studies would suggest that the cure rate after transsphenoidal surgery by experienced neurosurgeons is similar whether in children or adults with Cushings disease, approximately 6680% cures. The experience of the neurosurgeon in performing transsphenoidal surgery, which often requires drilling of incomplete pneumatization of the sphenoid sinus in young children, cannot be discounted. As stated in a classic textbook of endocrinology (11), discussing the cure of Cushings disease by transsphenoidal surgery, "In less experienced hands, the cure rate can fall to zero."
The use of pituitary radiation for retreatment of failed neurosurgery has shown good success in adult patients (2) and now also in this report in children. The advent of stereotactic focusing, cobalt 60, gamma knife, and other technologies are moving rapidly in the field of radiation therapy. It is important to recognize that the cure following radiation is nether immediate nor rapid. The clinical and biochemical cure may be years [as long as 2.86 yr in this study but as long as 5 yr in series by Estrada et al. (2)], requiring medical management in the interim. The investigators used a number of temporizing medications including opDDD, ketoconazole, and other medications. However, the patience with this mode of therapy (i.e. using medical therapy after pituitary radiation for a significant period of time) seems to be justified because retreatment with surgery has not only resulted in still failure to cure but also often panhypopituitarism. In the present study, only GH deficiency occurred after radiation. Still, it should be stated that pituitary hormone deficiencies following the effects of radiation may be time dependent and children must be followed closely both clinically and biochemically through childhood, adolescence, and even through adulthood for the possibility of further pituitary hormone deficiencies. Only time will tell whether hypogonadotropic hypogonadism or central hypothyroidism or ACTH deficiency will occur after pituitary irradiation, similar to the reports in adults.
A review of the recent literature indicates that transsphenoidal surgery is the first line of treatment for Cushings disease in children. The success rate after transsphenoidal surgery approximates 6680%. It now seems that pituitary irradiation is the appropriate therapy for retreatment of surgical failures, but the success rate as achieved in the present study may also be dependent on the skill of the radiation therapist as well as the endocrinologists familiarity with the use of adjunctive medical antiadrenal medication to temporize the effects of hypercortisolism until clinical and biochemical cure take place. Finally, posttreatment surveillance for the possibility of recurrence as well as pituitary failure must be life-long.
Lastly, one strong plea. Pediatric Cushings disease is rare even in the largest pediatric medical centers, especially so in the United States. This can be a deadly disease, and we should not be provincial or egotistical in caring for children with Cushings disease. Clearly, experience is an absolute necessity. Patients requiring inferior petrosal venous sampling should be referred to centers familiar with such procedures. Similarly, transsphenoidal surgery and pituitary irradiation should be performed at centers having extensive experience in treating this disorder in children. As well, both ketoconazole and opDDD have side effects, and the use of these drugs requires appropriate experience in their use and knowledge for monitoring of adverse events. The recent studies, including the present report, indicate that cure of pediatric Cushings disease is close to 100%, but probably only in experienced centers.
Acknowledgments
Footnotes
Abbreviations: CT, computed tomography; IPSS, inferior petrosal sinus sampling; MRI, magnetic resonance imaging.
Received November 8, 2002.
Accepted November 10, 2002.
References