Clinical characteristics of adrenocortical tumors in children
Qiu-Li Chen, Zhe Su, Yan-Hong Li, Hua-Mei Ma, Hong-Shan Chen and Min-Lian Du*
Department of Pediatrics, The First Affiliated Hospital of Sun Yat-Sen University, China
Abstract
Background: Adrenocortical tumors (ACTs) are rare in children. Because of the rarity and various manifestations of ACTs, patients of ACTs are not easily diagnosed. Some patients were misdiagnosed before surgery.
Objective: Identify the clinical, laboratorial, imaging and histopathological characteristics of adrenocortical tumors in children. Compare adrenalcortical adenoma with carcinoma.
Methods: A retrospective review of 34 identified patients who were younger than 15 years old with histologic confir- mation of adrenocortical carcinoma (ACC) or adenomas from 1991 to 2010.
Results: In these 34 patients, 19 were adrenocortical ade- noma (ACA) and 15 were ACC. The median age at diagnosis was 3.33 years (range, 0-16 years), and 70.6% of the patients were younger than five years. Girls slightly predominated over boys (1.4:1). For endocrine abnormality, 14 patients had isolated precocious puberty, five patients had isolated Cushing syndrome, 10 patients had precocious puberty plus Cushing syndrome, and five patients did not have any symptoms. The most frequent findings in laboratory tests were disturbance of the normal circadian rhythm of cortisol secretion (93.8%), followed by elevated serum level of testosterone (89.7%). Only 3.8% of ultrasound diagnosis and 12.1% of computed tomography (CT) diagnosis were consistent with pathologic diagnosis.
Conclusion: Different from those in adult, the most frequent presentation in children with ACTs is peripheral precocious puberty with or without Cushing syndrome, and isolated Cushing syndrome. Few present with non-functional local mass. Laboratory tests usually reveal the discordantly ele- vated serum levels of sexual corticosteroid hormones, change of diurnal rhythm of cortisol or increase of morning cortisol. The differentiation of malignant from benign tumor cannot merely depend on imaging. Final diagnosis relies on compre- hensive evaluation of clinical manifestations, laboratory data, imaging and pathology.
*Corresponding author: Min-Lian Du, Department of Pediatrics, The First Affiliated Hospital of Sun Yat-Sen University, No.58, Zhongshan 2nd Road, Guangzhou, 510080, China Phone: +86-020-87755766-8310, Fax: +86-020-87330736, E-mail: szzxsums@163.com
Keywords: adrenocortical adenoma; adrenocortical carci- noma; adrenocortical tumor; children; clinical characteristic.
Introduction
Adrenocortical tumor (ACTs) is a rare disease in children. The annual worldwide incidence of childhood ACTs is only 0.3-0.38 per million children below the age of 15 years (1). It accounts for 0.2% of all tumors during childhood. The peak incidence is in those younger than five years (2).
The adrenal gland consists of cortex and medulla. Three zones can be identified within the adrenal cortex: zona glomerulosa, zona fasciculate and zona reticularis. The main products of the peripheral zone of the cortex, zona glom- erulosa, are mineral corticoids (mainly aldosterone), while secretion of glucocorticoids (mainly cortisol) predominates in zona fasciculate and sex steroids (androgens, estrogens and progesterone) are secreted by the inner most zone, zona reticularis. Accordingly, the symptoms of ACTs depend on where the tumor is located (3) and whether it causes the corresponding steroid hormone hypersecretion syndromes. In general, the characteristics of children’s adrenocortical tumors, which most commonly present with precocious puberty and androgen elevation, are not the same with those of adults’ (4, 5). In addition, children with functional adre- nocortical tumors may present in sequence of peripheral precocious puberty presenting in virilization or feminization (isosexual or heterosexual precocious puberty), hypercorti- solism, alone or in combination (6); but hyperaldosteronism is rare.
The etiology of ACTs is not very clear yet. Mutations in P53 are the inherited genetic abnormalities most commonly associated with increased ACTs frequency in familial cancer syndromes (7). Other important factors in childhood ACTs pathogenesis are overexpression of the steroidogenic factor-1 transcription factor, causing overexpression of insulin-like growth factor-2 (8). Transcriptome analysis suggests also that the Wnt/b-catenin signaling pathway is activated in ACTs (9).
Because of the rarity of ACTs, pediatricians are not so familiar with the characteristics of children’s ACTs. Some patients were misdiagnosed before surgery, especially by radiologist and surgeons. Peripheral precocious puberty can be seen in both ACTs and congenital adrenal hyperpla- sia (CAH). Therefore, it is not easy to differentiate these two diseases. In this article, symptoms, laboratory features, imaging and pathology of ACTs in children were studied; the differences between adrenocortical adenomas (ACA) and adrenocortical carcinomas (ACC) were analyzed; the clinical features of ACTs were compared with those in CAH also.
Methods
The clinical data of 34 cases of childhood ACTs, confirmed by pathology and treated in The First Affiliated Hospital of Sun Yet-Sen University between 1991 and 2010, were reviewed in detail. The data including gender, age at diagnosis, interval between first symptoms and diagnosis, clinical symptoms and signs, laboratory data, imag- ing and pathology were collected and were used to discuss various aspects of the disease.
Data are reported as means±SD t-Test and ×2-test were used to compare the data.
Results
Demographics
Thirty-four patients had confirmed adrenocortical tumors. The demographics of ACTs are shown in Table 1. Girls slightly predominated over boys (1.4:1). The median age at diagno- sis was 3.33 years (range, 0 day to 16 years). Twenty-four patients (70.6%) were younger than five years. No positive tumor family history was found except for one ACC patient whose mother had breast cancer.
Between carcinoma and adenoma, there were no significant difference in gender, age at diagnosis, and interval between first symptom and diagnosis.
Clinical manifestation
Approximately 90% of the patients had clinical evidence of an endocrine syndrome. The most common clinical presenta- tion was peripheral precocious puberty, which was presented in 24 children (70.6%), followed by Cushing syndrome [clinical suspected and confirmed by biochemical diagno- sis of hypercortisolaemia (10)], which was presented in 15 children (44.1%). The number and ratio of clinical manifes- tations in ACTs, ACA and ACC were presented in Table 2. Signs and symptoms of ACTs were presented in Table 3. The five patients without endocrine abnormality were discovered by abdominal pain, abdominal distention, and/or abdomi- nal mass, or found by accident. In these five non-functional tumors, four cases were carcinoma.
Peripheral precocious puberty included premature pub- arche (13 cases), heterosexual peripheral precocious puberty in female (virilization, 12 cases), isosexeual peripheral pre- cocious puberty in female (one case), isosexual peripheral
precocious puberty in male (eight cases), and heterosexual peripheral precocious puberty in male (gynecomastia, two cases, one had penis enlargement at the same time).
Hypertension was noted in nine children of 15 patients with Cushing syndrome. In one patient with ACT, the reason of the first visit was stroke and hemiparalysis at her blood pressure of 225/150 mm Hg, though the history collected later revealed that she had her Cushing syndrome for four years and preco- cious puberty for two years. Her motor function and blood pressure completely recovered two years after operation.
Five cases were congenital adrenocortical tumors, of which two patients were found by prenatal sonogram. Clinical man- ifestations were presented at birth in three cases. One boy presented with penis enlargement; one girl presented with clitoral enlargement; and one girl presented with acne within a week. In these five patients, three were adenoma and two were carcinoma.
In one patient of ACA was found a retroperitoneal tumor by incident in neonatal. She had no clinical manifestation at that time. Four years later, she developed clitorimegaly. An ACT was found by computed tomography (CT) and conformed by pathology.
Laboratory
Elevated serum levels of testosterone (T) were found in 26 patients (89.7%) of the 29 patients tested. The testosterone levels were at 2.33-187.5 times of the upper limits of normal value for corresponding ages and pubertal stage. In eight pre- pubertal children, T was even >52 nmol/L, which is the upper limit of the laboratory test. Dehydroepiandrosterone sulfate (DHEAS) were tested in 13 cases (it was carried out in the hospital a few years previously). Serum level of DHEAS was elevated in 11 of 13 patients (84.6%). In nine patients (three ACC and six ACA) DHEAS were higher than 40.71 µmol/L (six patients had serum T level >52 nmol/L at the same time). The DHEAS levels were in accordance with the T level in these 11 patients. There were 15 patients (68.2%) with high estrodiol (E2) levels in 22 patients tested whose testoster- one levels were all elevated at the same time. Seven of nine patients had elevated serum level of DHEAS accompany with elevated E2. The E2 levels were 1.13-12.93 times of the upper limits of normal children with corresponding ages and puber- tal stage. In the patients with peripheral precocious puberty, all (23/23) had elevated serum levels of T, 90% (9/10) had elevated DHEAS, and 76.5% (13/17) had elevated E2. Even in
| Number, % | Boys | Girls | Boys/girls | Median age at diagnosis, year | Median interval between first symptom and diagnosis, months | Diameter of tumor, cm | |
|---|---|---|---|---|---|---|---|
| ACT | 34 | 14 | 20 | 1:1.4 | 3.33 | 5.5 (0-54) | 7.27±4.22 (0-16) |
| ACA | 19 (55.9) | 6 | 13 | 1:2.2 | 2.83 | 6 (0-54) | 4.92±2.67(0-11) |
| ACC | 15 (44.1) | 8 | 7 | 1:0.9 | 4.33 | 4 (0.25-36) | 10.24±3.98 (4.5-16) |
| p-Value | 0.296 | 0.803 | 0.142 | <0.001* |
ACA, adrenocortical adenoma; ACC, adrenocortical carcinoma; ACT, adrenocortical tumor. * p<0.05: Adenoma vs. carcinoma.
| Clinical manifestations | ACT, % | ACA, % | ACC, % |
|---|---|---|---|
| Isolated precocious puberty | 14 (41.2) | 11 (57.9) | 3 (20.0) |
| Isolated Cushing syndrome | 5 (14.7) | 3 (15.8) | 2 (13.3) |
| Mixed type | 10 (29.4) | 4 (21.1) | 6 (40.0) |
| No endocrine abnormalities | 5 (14.7) | 1 (5.3) | 4 (26.7) |
| Total | 34 | 19 | 15 |
ACA, adrenocortical adenoma; ACC, adrenocortical carcinoma; ACT, adrenocortical tumor.
patients without precocious puberty, 50.0% (3/6) T were ele- vated, 66.7% (2/3) DHEAS were elevated, and 40.0% (2/5) E2 were elevated. Seven patients were performed high-dose dexamethasone suppression test and all the elevated T levels could not be suppressed.
Abnormal circadian rhythm of cortisol secretion was the most common finding in the group, which could be seen in 15 of 16 patients who took the test (93.8%). Ten of 15 patients with Cushing syndrome (66.7%) had elevated morning cortisol level (883.5-1486.5 nmol/L). Except for one case, the awake midnight concentrations of cortisol in plasma were 292.2-1418.3 nmol/L [more than 207 nmol/L (10)], and the mean ratio of 0 Am/8 Am cortisol level were 0.90 (range 0.65-1). In patients with Cushing syndrome, all (15/15) had disturbance of the normal circadian rhythm of cortisol secretion; and all (7/7) had high 24-h urinary free cortisol (UFC). On the other hand, among patients without Cushing syndrome, the results of serum cortisol and UFC were all normal.
In the group, the lactate dehydrogenase (LDH) levels were above the upper limits of normal (range 114-240 U/L) in 86.7% (26/30) patients. The mean serum LDH level was 739.5±306.3 U/L (range 294 to 1466 U/L) in ACC and 309.6±114.2 U/L (range 129 to 594 U/L) in ACA. Serum LDH level was the only laboratory test which was signifi- cant difference between carcinoma and adenoma (p<0.01). Other mentioned laboratory tests had no significant difference between carcinoma and adenoma either in positive rate or in numerical value. The positive rates of laboratory tests in ACT, ACA and ACC were presented in Table 4.
Imaging
Each patient accepted ultrasound and/or CT-scan. Ultrasound and CT-scans were both accepted in 22 patients, while ultra- sound alone in four and CT-scans alone in eight.
Ultrasonography was performed in 26 patients. Ultra- sonography was able to find tumor in 24 patients (92.3%), give correct location in 20 patients (76.9%) yet with precise diagnosis (in accord with pathologic results) in only one patient (3.8%). Tumor could be located in 16 cases (61.5%) without qualitation. The ultrasound diagnoses were incorrect in nine cases (34.6%), including two cases (7.7%) of nega- tive mass finding and four cases (15.4%) of dislocation (e.g., nephroblastoma).
CT was performed in 30 patients and the result was no better than ultrasonography. CT was able to find tumor in 28 patients (93.3%), yet with precise diagnosis (in accord with pathologic results) in only four patients (12.1%). Eleven cases (36.7%) located the tumor without qualitation. The incorrect diagnoses were in 14 cases (46.7%), includ- ing diagnosed as adrenal hyperplasia in two cases, adrenal neuroblastomas in eight cases, nephroblastoma in one case, and three cases of ACA diagnosed as ACC (because they were relatively large and had ischemia and necrosis in the middle).
The mean diameter of adrenocortical carcinomas (10.24±3.98 cm) was significantly larger than that of ade- nomas (4.92±2.67 cm). Local invasion or metastases were found in 10 cases (66.7%) with ACC: six cases at the time of diagnosis and four cases at the time of relapse. Invasion or metastases were located in liver (three cases), lung (three cases), retroperitoneal lymph nodes (two cases), inferior vena cava, omentum, and abdominal cavity.
Pathology
Nineteen patients were diagnosed as ACA. Fifteen patients met Weiss criteria (11, 12) and were diagnosed as ACC.
Discussion
Adrenocortical tumor is an uncommon disease. There were only 34 cases in 20 years in the hospital and they accounted for 0.06% of all pediatric in-patients. Similar to the prior reports (3), our data showed that ACTs are more likely to occur in girls than in boys; the peak incidence is under five years of age.
Clinical features
The clinical manifestations of adrenocortical tumors in children include isosexual or heterosexual periphery
| Signs and symptoms | n | % |
|---|---|---|
| Central obesity | 15 | 44.1 |
| Hirsutism | 15 | 44.1 |
| Pubic hair | 13 | 38.2 |
| Acne | 13 | 38.2 |
| Abdominal mass | 12 | 35.3 |
| Clitorimegaly | 12 | 35.3 |
| Penile enlargement | 8 | 23.5 |
| Hypertension | 9 | 26.5 |
| Abdominal pain or distension | 6 | 17.6 |
| Deepen voice | 6 | 17.6 |
| Accelerated growth velocity | 5 | 14.7 |
| Breast development | 3 (2 boys, 1 girl) | 8.8 |
| Found by incident | 3 | 8.8 |
| Fever or symptosis | 2 | 5.9 |
| Growth retardation | 2 | 5.9 |
| ACT, % | ACA, % | ACC, % | p-Value | |
|---|---|---|---|---|
| Disturbance of the normal circadian rhythm of cortisol secretion | 15/16 (93.8) | 7/8 (87.5) | 8/8 (100) | 1.000 |
| TÎ | 26/29 (89.7) | 15/18 (83.3) | 11/11 (100) | 0.268 |
| LDHÎ | 25/29 (86.2) | 11/15 (73.3) | 15/15 (100) | 0.100 |
| Urinary-free cortisol (UFC)Î | 7/8 (87.5) | 4/5 (80) | 3/3 (100) | 1.000 |
| DHEAST | 11/13(84.6) | 7/9 (77.8) | 4/4 (100) | 0.497 |
| E,Î | 15/22 (68.2) | 10/13 (76.9) | 5/9 (55.6) | 0.376 |
| 17-hydroxyl progesterone (17-OHP)Î | 2/4 (50) | 1/3 (33.3) | 1/1 (100) | 1.0 |
| 8 Am CorticolÎ | 14/30 (46.7) | 6/17 (35.3) | 8/13 (61.5) | 0.269 |
| ACTHI | 7/18 (38.9) | 5/13 (38.5) | 2/5 (40) | 1.0 |
| AldosteroneÎ | 3/8 (37.5) | 0/3 (0) | 3/5 (60) | 0.196 |
ACA, adrenocortical adenoma; ACC, adrenocortical carcinoma; ACT, adrenocortical tumor. p-Value: Adenoma vs. carcinoma.
precocious puberty (such as virilization in female and femi- nization in male), Cushing syndrome, alone or in combi- nation, and non-functional tumor. Most of adrenocortical tumors are functional and secrete excessive androgens and/or glucocorticoids that lead to clinical symptoms (13, 14). In this series, 85.3% tumors were functional. Unlike in adults, presentation of precocious puberty (70.1%) were more common than Cushing syndrome (44.6%), which was in accordance with those previously reported (2, 12, 15-19). The sequences of presenting frequency were iso- lated precocious puberty (41.2%), mixed type (precocious puberty plus Cushing syndrome, 29.4%), isolated Cushing syndrome (14.7%), and non-functional tumor (14.7%). There were 125 cases of ACTs reported in Brazil, similarly, that the most common forms of presentation were isolated virilization (51.2%) and virilization plus Cushing syndrome (42%); non-functioning tumors and isolated Cushing syn- drome comprised 4.8% and 0.8% of the cases, respectively (20). Therefore, periphery precocious puberty in children, with or without Cushing syndrome, should be an alert for adrenocotical tumor. Except for the exogenous use of steroid hormones in medical therapy, more than 80% of Cushing syndromes in children are caused by adrenocortical tumors (21), while in adults it is mainly caused by Cushing dis- ease. If children suddenly get obese, particularly those with peripheral precocious puberty, they should be screened for adrenocortical tumors.
The patients with adrenal incidentalomas have the risks of progression towards subclinical autonomous glucocorticoid overproduction and even overt Cushing syndrome (22), and so other hormones and symptoms. It was proved by one of the ACA patients who was mentioned above. The five non- functional tumors, except for one ACA, four cases were carcinoma. Non-functional reasons probably were that the carcinomas grew so fast that they presented with abdominal masses only; even they secrete some hormones, the corre- sponding endocrine syndromes being target organs response would yet not present. Another reason may be that the degree
of cell differentiation of carcinomas was low and they could not secrete hormones. Once adrenal incidentalomas was found in children, active treatment should be taken due to the high- risk of malignant adrenocortical tumor, and relative hormonal tests should be performed.
Laboratory characteristics
The laboratory characteristics of ACTs are the discordantly elevated serum levels of sexual corticosteroids, and distur- bance of the normal circadian rhythm of cortisol secretion.
The major adrenal androgens are DHEA, DHEA-S and androstenedione (44-A) (23). DHEA and DHEA-S are mainly secreted by zona reticularis. 44-A and T are secreted by both zona reticularis and zona fasciculate (24). Ninety-five percent DHEAS are secreted by adrenal in any age. So the elevation of DHEAS level may suggest adrenal disorders. DHEAS were elevated in 84.6% patients who were tested in this group. However, normal level of DHEAS could not exclude ACTs. From our data, no significant differences in positive rate or degree of DHEAS elevation were found between ACC and ACA. DHEAS is a weak androgen and also a precursor of other form of androgens (T, 44-A) and estrogens. Other sexual corticosteroids, such as T, E, and E2, may also be increased by either direct secretion or peripheral conversion of DHEA and DHEA-S. We discovered that there were ele- vated serum levels of sexual corticosteroids, but the elevated degree was discordantly in each other. It was proved that the human adrenocortical carcinoma (HAC) cells expressed mRNA for ACTH receptor (MC2R) and Ang II receptor and MC2R protein. HAC clone 15 (HAC15) cells responded to treatment with ACTH, Ang II, with increased cortisol and aldosterone production, and also increased expression of mRNA, encoding all steroidogenic enzymes needed for vari- ant steroid hormone (such as cytochrome P450 enzymes sys- tem) and aldosterone biosynthesis. However, the expression of different P450 enzymes was discordance. Cortisol and adrenal androgen were produced at significantly higher levels
than aldosterone, particularly for androgen (25). It may imply that the biochemical characteristic of HAC cells was different from normal adrenolcotical cells. Elevated T level was found in 89.7% of the patients and it was one of the most common findings, which was reported before (15). Elevated E2 levels were found in 68.2%. The clinical presentations were not always identical to the levels of T and E2. One patient pre- sented with fermilization but did not have elevated E2. On the contrary, three patients had the T levels as high as adult- hood; but no virilization. The aromatase enzyme catalyzes the final stage of estrogen (E, and E2) biosynthesis pathway from androgens (DHEA-S, T and 44-A). Aromatase enzyme express in the fat tissue, gonadal tissue, normal adrenal and ACTs. It was proved that aromatase is expressed at the similar levels in normal adrenal and in adrenocortical tumors, but at variably high levels in feminizing adrenal tumors. Different promoter utilization patterns are found among tumor subtypes (26). If a boy has ACT, which has high aromatase activity, he may present with gynecomastia. In the group, two boys with gynecomastia had high T levels, while only one patient had consistently elevated E2 levels. Unfortunately E, was not tested. Another boy had gynecomastia (not ACT case) with normal E2 level but significantly elevated E, level.
Therefore, if one of the sexual corticosteroids hormones is extremely high, it may be a characteristic of ACTs. The individual difference could present in the discordance of dif- ferent sexual corticosteroid hormones levels or discordance between levels of sexual corticosteroids and sexual charac- ters. This complex relationship may relate to discordance in activity of different cytochrome P450 enzymes system in dif- ferent tumor subtypes cells.
In this group, 93.8% of patients had disturbance of the nor- mal circadian rhythm of cortisol secretion while only 46.7% patients had cortisol elevation in 8 Am. Therefore, children with normal morning cortisol level cannot be excluded from adrenocortical disease. Disturbance of the normal circadian rhythm of cortisol secretion with abnormal levels in T, E2, DHEAS are more sensitive than 8 Am cortisol for diagnos- ing of ACTs. Disturbance of the normal circadian rhythm of cortisol secretion and elevated testosterone were found in most of the ACTs patients, even in those patients without corresponding endocrine symptoms but who presented with abdominal mass and went directly to the surgeon. So it is nec- essary to measure the 8 Am, 0 Am cortisol and T, E2, DHEAS for the patients with periphenal precocious puberty, Cushing syndrome and adrenal tumor or nearby tumor is found by imaging.
Imaging
Patients should undergo ultrasonography and CT-scan of adrenal, whenever periphery precocious puberty (especially virilization and feminization) and (or) Cushing syndrome are presented.
Ultrasonography is easy and cheap. But because of its low spatial resolution, its sensitivity and specificity are not as good as CT and magnetic resonance imaging (MRI). Among the patients, ultrasonic diagnoses were identical to patho- logic diagnoses in only 3.8% patients and were misdiagnosed
(including 15.4% incorrect positioning) in 34.6% patients. Therefore, ultrasonography can only be used as a screening method.
CT is the major method for early positioning diagnosis and is very important to preoperative staging, formulate the operative plan and judge the prognosis. But CT also has limi- tations, especially in defining the tumor’s nature (15). Among the patients, CT diagnoses were identical to pathologic diag- noses in only 12.1% patients and were misdiagnosed in 46.7% patients.
Typical imaging findings of pediatric ACTs consist of a large, well-defined suprarenal tumor containing calcifications with a thin capsule and central necrosis or hemorrhage. But the characteristics of imaging of different adrenal tumors are very similar. It is hard to make a correct diagnosis, especially to determine the nature and differentiation of benign or malig- nant tumor by imaging alone. For example, one patient who had both precocious puberty and Cushing syndrome was mis- diagnosed as hepatoblastoma and nephroblastoma by imaging in other hospitals and adrenal neuroblastoma in our hospital, while the pathological diagnosis was adrenocortical carcinoma. Therefore, ultrasonography and CT are extremely important to discover an ACT and very helpful for clinical diagnosis, but imaging diagnosis cannot replace clinical analysis. Clinical doctors should combine imaging presentations with clinical features and laboratory test to make a clinical diagnosis, instead of following blindly to the imaging qualitation diagnosis.
Pathology
However, to define an ACT as a benign or malignant tumor sometimes is difficult even in pathology (27). The clinical behavior may be more important in determining outcome (14). For example, one of our patients was diagnosed as ACA at the first visit. Five months later, however, he had relapse and the tumor metastasis to the liver. Operation was per- formed once again and pathologic diagnosis was carcinoma. The pathologic diagnosis of another patient was adrenocorti- cal adenoma with small amounts of atypical cells. Because of huge tumor size (16×15×10 cm3) and extremely high level of LDH (five times the upper limit of normal), the patient was diagnosed as adrenocortical carcinoma despite no metastasis disease being found at that time. After four years, the patient had metastasis to the liver. Therefore, in addition to patho- logic diagnosis, clinical behavior, tumor size and laboratory test should be considered when a benign or malignant tumor is defined.
Differentiation of adrenocortical adenoma and carcinoma
Twice as many girls had ACA than boys, while girls were equal to boys in ACC. Between ACA and ACC, there were also no statistical differences in age at diagnosis and inter- val between appearance of first symptoms and diagnosis. The interval between first symptoms and diagnosis were mostly shorter in ACC than ACA (4 months vs. 6 months), yet few of them were long (1-2 years) in ACC. Benign tumors of the
adrenal cortex are generally smaller than malignant tumors (6). In these cases, the mean diameter in ACC (10.24 cm) was twice as that in ACA (4.92 cm), and there was signifi- cant statistical difference between them. In this group, 90% tumors ≤5 cm were ACA whereas 90% tumors ≥10 cm were ACC. However, there was overlap between them. Among the patients, the shortest diameter in ACC was 4.5 cm, whereas the longest diameter in ACA was 11 cm. Therefore, no cut- off value can be drawn between ACA and ACC. Small tumor (≤5 cm) tends to be ACA, while large tumor (≥10 cm) is more likely to be malignant.
ACA were more likely to present with precocious puberty (78.9%), especially isolated precocious puberty (57.9%). On the contrary, Cushing syndrome appeared to occur more fre- quently (53.3%) in ACC as previously reported (12, 19) and it was similar to precocious puberty (60%) in ACC. The mixed type (40.0%) was most common in ACC. Non-functional tumors also tend to occur more frequently in ACC (26.7%) than in ACA (6.3%). Four cases of non-functional ACC were all misdiagnosed because they lacked of specific symptoms and were misdiagnosed as neuroblastomas, nephroblastoma and pheochromocytoma, respectively.
In laboratory test, LDH was the only one that had signifi- cant statistically difference between ACA and ACC. LDH is a kind of glycolytic enzyme. Patients with tumor, especially malignant tumor, often have high level of LDH. LDH levels in ACC patients were all elevated and much higher than ACA (p<0.01). Except for one case whose LDH was 594 U/L, almost all ACA patients’ LDH were lower than 500 U/L. So LDH should be tested in all adrenal tumors and it is important to differentiate malignant tumor from benign tumor. If LDH is higher than 500 U/L which is two times the upper limit of normal, the tumor is more likely to be malignant. Of course, LDH is a non-specific biochemical mark and ACC cannot be diagnosed or excluded only by LDH. 8 Am cortisol, DHEA, T, E2 and aldosterone were all not significant for differentiat- ing ACC from ACA.
Therefore, an adrenocortical tumor with large size in imag- ing (especially the one diameter more than 10 cm) and/or extremely high LDH level (especially the one LDH more than two times the upper limit of normal) may predict the malig- nant nature. Imaging and pathological evidence of metastasis should be defined as malignant tumor. Other clinical signs and symptoms, laboratory test and imaging characteristics had limited contribution to the differentiation between ACC and ACA.
Differentiation of ACTs and congenital adrenal hyperplasia (CAH)
CAH (especially 21-hydroxylase deficiency) is also pre- sented with precocious puberty and hypersecretion of sexual hormones. A few patients with ACTs were misdiagnosed as CAH because the tumors were too small to be detected by ultrasonography in early stage. One of our patients with ACC had clitoromegaly at birth. When she was 1.5 years in her first visit, ultrasonography did not find adrenal mass, but she had elevated T and high renin activity, so she was diagnosed and
treated as CAH (21-hydroxylase deficiency). One year later, she had increased clitoromegaly, gradually presented Cushing syndrome and extremely elevated testosterone that could not be controlled by high dosage of cortisol replacement. She was rechecked with ultrasonographs when she was three years old and found an adrenal neoplasm. Her pathology was adrenocortical carcinoma. Two other patients also presented with clitoromegaly and diagnosed as CAH (21-hydroxylase deficiency). Their ultrasonographs were all negative at that time. They were treated with high dosage of cortisol but tes- tosterones were persistently high. After several months, they were rechecked with ultrasonography and found ACTs which pathologies were ACA. They recovered after surgery.
The sexual corticosteroids and their precursors secreted by tumors are similar to CAH, e.g., DHEAS, T, E2, 17-OHP, which can be elevated in both tumors and CAH. In addition, high rennin is not specific to salt wasting CAH; it can be seen in ACTs. Two cases of this group were with high activity of rennin. So the high level of 17-OHP and renin activity could not be regarded as the specific mark for CAH. Low basic ACTH and high-dose dexamethasone-suppression tests are helpful to differentiate ACTs from CAH (28).
Before the diagnosis of CAH is made, ultrasonography should be done to exclude ACTs. If adrenocortical tumor is highly suspected from clinical presentations and laboratory tests while sonogram is negative, adrenal lamellar CT, which is more sensitive, should be done in order to discover tumor earlier. If CT is still negative while testosterone cannot be sup- pressed by cortisol replacement, ultrasonography or CT-scan should be rechecked periodically.
On the other hand, a few patients who were definitely diag- nosed as CAH may appear with adrenocortical adenomas after their school age. It probably is the result of long-term effect of ACTH that makes the adrenal cortical tissue over hyperplasia. One girl was definitely diagnosed as CAH when she was virilizing and salt-wasting shortly after birth. Since school age, her testosterone had remained at the level of male’s mid-puberty, even though she was taking a high dose of cortisol. She underwent CT-scan and an adrenal local nodular lobe was found which pathology was adrenocortical adenoma. The serum level of androgens had reduced after surgery. Therefore, if a patient’s condition is not under control when he/she is diagnosed as CAH, adrenal ultrasonography should be rechecked regularly to find out whether there is a hyperplasia adrenocortical adenoma.
Conclusion
Different from those in adult, the most frequent presentation in children with ACTs is peripheral precocious puberty with or without Cushing syndrome, and isolated Cushing syndrome. Few present with non-functional local mass. Laboratory tests usually reveal the discordantly elevated serum levels of sexual corticosteroid hormones, change of diurnal rhythm of cortisol or increase of morning cortisol. The differentiation of malig- nant from benign tumor cannot merely depend on imaging.
Final diagnosis relies on comprehensive evaluation of clinical manifestations, laboratory data, imaging and pathology.
References
1. Sandrini R, Ribeiro RC, DeLacerda L. Childhood adrenocortical tumors. J Clin Endoc Metab 1997;82:2027-31.
2. Sbragia L, Oliveira-Filho AG, Vassallo J, Pinto GA, Guerra- Junior G, et al. Adrenocortical tumors in Brazilian children. Arch Pathol Lab Med 2005;129:1127-31.
3. Jarolim L, Breza J, Wunderlich H. Adrenal tumours. Eur Urol 2003;43(Curric Urol I-X).
4. Wajchenberg BL, Pereira MA, Medonca BB, Latronico AC, Carneiro PC, et al. Adrenocortical carcinoma clinical and labora- tory observations. Cancer 2000;88:711-36.
5. Sarwar ZU, Ward VL, Mooney DP, Testa S, Taylor GA. Congenital adrenocortical adenoma: case report and review of literature. Pediatr Radiol 2004;34:991-4.
6. Wolthers OD, Cameron FJ, Scheimberg I, Honour JW, Hindmarsh PC, et al. Androgen secreting adrenocortical tumors. Arch Dis Child 1999;80:46-50.
7. Ribeiro RC, Pinto EM, Zambetti GP. Familial predisposition to adre- nocortical tumors: clinical and biological features and management strategies. Best Pract Res Clin Endoc Metab 2010;24:477-90.
8. El Wakil A, Doghman M, Latre De Late P, Zambetti GP, Figueiredo BC, et al. Genetics and genomics of childhood adre- nocortical tumors. Mol Cell Endocrinol 2011;336:169-73. Epub 2010 Nov 19.
9. Bertherat J, Bertagna X. Pathogenesis of adrenocortical cancer. Best Pract Res Clin Endoc Metab 2009;23:261-71.
10. Newell-Price J, Bertagna X, Grossman AB, Nieman LK. Cushing’s syndrome. Lancet 2006;367:1605-17.
11. Weiss LM. Comparative histologic study of 43 metastasizing and nonmetastasizing adrenocortical tumors. Am J Surg Pathol 1984;8:163-9.
12. De Leon DD, Lange BJ, Walterhouse D, Moshang T. Long-term (15 Years) outcome in an infant with metastatic adrenocortical carcinoma. J Clin Endocrinol Metab 2002;87:4452-6.
13. Frederick J, Rescorla MD. Malignant adrenal tumors. Semin Pediat Surg 2006;15:48-56.
14. Mendonca BB, Lucon AM, Menezes CA, Saldanha LB, Latronico AC, et al. Clinical, hormonal and pathological findings
in a comparative study of adrenocortical neoplasms in childhood and adulthood. J Urol 1995;154:2004-9.
15. Hartley AL, Birch JM, Marsden HB, Reid H, Harris M, et al. Adrenal cortical tumours: epidemiological and familial aspects. Arch Dis Child 1987;62:683-9.
16. Michalkiewicz E, Sandrini B, Figueiredo B, Miranda EC, Caran E, et al. Clinical and outcome characteristics of children with adrenocortical tumors: a report from the international pediatric adrenocortical tumor registry. J Clin Oncol 2004;22:838-45.
17. Cordera F, Grant C, Heerden JV, Thompson G, Young W. Androgen-secreting adrenal tumors. Surgery 2003;134:874-80.
18. Ciftci AO, Senocak ME, Tanyel C, Buyukpamukcu N. Adrenocortical tumors in children. J Pediatr Surg 2001;36: 549-54.
19. Weiss LM, Medeiros LJ, Vickery AL Jr. Pathologic features of prognostic significance in adrenocortical carcinoma. Am J Surg Pathol 1989;13:202-6.
20. Pereira RM, Michalkiewicz E, Sandrini F, Figueiredo BC, Pianovski M, et al. Childhood adrenocortical tumors. Arq Bras Endocrinol Metabol 2004;48:651-8.
21. Stewart DR, Jones PH, Holleys A. Carcinoma of the adrenal gland in children. J Pediatr Surg 1974;9:59-67.
22. Barzon L, Fallo F, Sonino N, Boscaro M. Development of overt Cushing’s syndrome in patients with adrenal incidentaloma. Eur J Endocrinol 2002;146:61-6.
23. Ghizzoni L, Mastorakos G, Vottero A. Adrenal hyperandro- genism in children. J Clin Endocrinol Metab 1999;84:4431-5.
24. Pang SY, Legido A, Levine LS, Temeck JW, New MI. Adrenal androgen response to metyrapone, adrenocorticotropin, and corticotrophin-releasing hormone stimulation in children with hypopituitarism. J Clin Endocrinol Metab 1987;65:282-9.
25. Parmar J, Key RE, Rainey WE. Development of an adrenocorti- cotropin-responsive human adrenocortical carcinoma cell line. J Clin Endocrinol Metab 2008;93:4542-6.
26. Moreau F, Mittre H, Benhaim A, Bois C, Bertherat J, et al. Aromatase expression in the normal human adult adrenal and in adrenocortical tumors: biochemical, immunohistochemical and molecular studies. Eur J Endocrinol 2009;160:93-9.
27. Neblett WW, Frexes-Steed M, Scott HW. Experience with adre- nocortical neoplasms in childhood. Am Surg 1987;53:117-25.
28. Street ME, Weber A, Camacho-Hubner C, Perry LA, Brain CE, et al. Girls with virilisation in childhood: a diagnostic protocol for investigation. J Clin Pathol 1997;50:379-83.