CASE REPORTS
Unusual Adrenal Cortical Tumor of Unknown Biologic Potential: A Nodule in a Nodule in a Nodule
SARANGARAJAN RANGANATHAN, 1,3* KECHA LYNSHUE,2 JENNIFER L. HUNT,3 TIMOTHY KANE,4 AND RONALD JAFFE1, 1,3
1Department of Pathology, Children’s Hospital of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA 15213, USA
2Department of Pediatric Endocrinology, Children’s Hospital of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA 15213, USA
3Department of Pathology, University of Pittsburgh School of Medicine, Pittsburgh, PA 15261, USA
4Department of Pediatric Surgery, Children’s Hospital of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA 15213, USA
Received November 5, 2004; accepted April 8, 2005; published online July 14, 2005.
ABSTRACT
Adrenocortical tumors are uncommon neoplasms in childhood. Most pediatric adrenal tumors are virilizing and carcinomas are more common than adenomas. Recent molecular data suggest an adenoma-to-carci- noma progression sequence in adrenal cortical neo- plasms. We report a case of a 5-year-old boy who presented with virilizing symptoms secondary to an adrenal tumor that was resected laparoscopically. The bulk of the tumor was a large, yellow mass with typical features of an adrenal cortical adenoma. In the center was a well-circumscribed tan-brown nodule that was distinct from the adenoma and had oncocytic features. A third minute focus (3.0 mm) was noted that was not circumscribed or encapsulated, but showed marked pleomorphism and abundant mitoses, including atypi- cal forms and increased Ki67 compared with the outer 2 nodules. Molecular analysis to assess the clonality and mutation rates of the 3 distinct areas showed only 2 genetic loci with allelic imbalances.
Key words: adrenal adenoma, oncocytic, loss of het- erozygosity
INTRODUCTION
Benign and malignant adrenocortical tumors (ACTs) in children are rare, with about 25 cases being reported each year; 75% of these are adrenal cortical carcinomas (ACCs) [1]. The estimated annual incidence in the United States of ACCs is 3 per million in patients younger than 20 years [2]. There are 2 peaks in the age distribution, with 1 peak occurring between ages 1 and 12 years (80% are younger than 5 years), and the other peak occurring between the fourth and fifth decades [3]. Although the histopathologic criteria for malignancy are fairly well established in adults, the same features may not apply to children. Wieneke et al. [1] described a group of features that may predict malignancy of adrenal cortical neoplasms in children. These include tumor weight greater than 400 g, tumor larger than 10.5 cm, extension into periadrenal soft tissues and inferior vena cava, presence of vascular invasion, capsular invasion, tumor necrosis, increased mi- totic activity greater than 15 of 20 high-power fields, and atypical mitotic figures.
*Corresponding author, e-mail: rangsx@chp.edu
ACTs are usually detected incidentally by imaging scans performed for unrelated reasons. Although virilization occurs in 20% to 30% of adults with functional ACC, it is the most common manifestation (72%) in children with adrenocor- tical cancer [4]. Most pediatric patients with vir- ilization have pubic/body hair (up to 95%), clitoral or penile enlargement (75%), and adult voice (50%). Up to 74% of patients with ACC will have Cushing syndrome [3,5]. We report a case of a 5- year-old boy who presented with virilization and Cushing syndrome and was diagnosed with an adrenal cortical adenoma with unusual histo- pathologic findings.
CASE REPORT
The patient was a 5-year-old male who presented to the endocrinology clinic for evaluation of pre- mature pubic hair. His parents had first noticed the pubic hair 1 year previously. However, during the preceding 3 to 4 months, they noticed that his phallus was enlarging and the pubic hair was increasing. They also reported an adult body odor and acne. Serum dehydroepiandrostenedione sulfate level was 104 µg/dL (60 to 190 µg/dL, laboratory values standardized for sex and Tan- ner stage), 17-hydroxyprogesterone level was 134 ng/dL (5 to 115 ng/dL), testosterone level was 77.8 ng/dL (20 to 250 ng/dL), and androstenedi- one level was 155 ng/dL (18 to 89 ng/dL). A bone age obtained 2 weeks before diagnosis was interpreted as appropriate. Medical history was unremarkable. On physical examination, the pa- tient’s height and weight were at the 75th and 97th percentiles, respectively. His blood pressure was 113/72 mmHg. The abdomen was soft, non- tender, and nondistended with no appreciable hepatosplenomegaly or masses. Testes were pre- pubertal in size (2 to 3 mL, orchidometer). Stretched phallus size was approximately 5 cm in length and 2.5 cm in width. He had Tanner stage 2 to 3 for pubic hair.
An adrenocorticotropic hormone stimulation test was performed to rule out congenital adrenal hyperplasia, the results of which were not sug- gestive of congenital adrenal hyperplasia (17-hy- droxyprogesterone levels were 159 ng/dl at baseline and 149 ng/ml after stimulation). Be-
cause of the high androgen levels and the fairly rapid progression of clinical findings, a tumor was suspected and abdominal computed tomogram showed a 3.7- x 3.0- x 3.5-cm left suprarenal mass. The right adrenal was normal. Laparoscopic resection of the adrenal mass was performed. Be- cause of high preoperative serum cortisol level, the patient was placed on steroid replacement therapy postoperatively with the assumption that the functional tumor might have suppressed the con- tralateral adrenal gland. His hydrocortisone dose was slowly tapered over a 2-month period. Morn- ing cortisol obtained while off hydrocortisone several weeks later was 6.5 µg/dL with an adre- nocorticotropic hormone level of 34 pg/mL (6 to 48 pg/mL).
Postoperatively, chest computed tomogram and bone scan were negative, and androgen levels returned to baseline (androstenedione level of 6 ng/dL, dehydroepiandrostenedione sulfate level < 1 µg/dL, 17-hydroxyprogesterone level of 8 ng/dL, and testosterone level < 11.5 ng/dL). On follow-up endocrinology visit 3 months later, the patient had lost 3.2 kg and had grown 3.5 cm. Clinically, his symptoms resolved with no progression of pubic hair, axillary hair, or acne.
PATHOLOGY
Gross pathology
A 23-g, 4.2- x 3.9- x 2.6-cm adrenalectomy speci- men that was replaced by a rubbery-firm, pink- tan, very well-circumscribed, encapsulated mass was received in pathology. The capsular surface was smooth. The cut surface showed a tan, faintly lobulated area surrounding a more circumscribed brown nodule 2.3 cm in diameter that had several, irregular orange foci. A minute 3-mm yellow- brown nodular area was seen within the second nodule (Fig. 1). No rim of adrenal tissue was grossly appreciated.
Microscopic features
Sections showed 3 concentric nodules. The larg- est nodule was rimmed by a thin fibrous capsule with no extension beyond the capsule. This nod- ule was composed mainly of sheets of interme- diate-size cells that were characterized by amphophilic cytoplasm and vesicular nuclei with
finely granular chromatin and inconspicuous nucleoli (Fig 2A,B). Rare mitoses (0 to 2 per 50 high power fields) were present and no atypical forms were identified. A Ki-67 (Immunotech; monoclonal, 1:200) immunostain showed a low proliferation index of about 5% (Fig. 2C). Deli- cate fibrous septae were seen coursing between the cells forming small poorly defined aggregates, with a vague nested appearance. There was no evidence of capsular or vascular invasion. A small rim of normal adrenal cortex was present be- tween the outer 2 nodules and focally beneath the capsule.
Within this large nodule was a second, smaller, circumscribed nodule, measuring 2.3 cm, that was composed of a distinct population of larger oncocytic cells (Fig. 2D,E). A thin, delicate fibrous capsule surrounded this nodule. These cells were larger than the cells in the outer nodule and had abundant eosinophilic cytoplasm, large oval pleomorphic nuclei with prominent nucleoli, nuclear pseudo-inclusions, and rare mitotic fig- ures (1 to 2 per 50 high-power fields). The prolif- eration index was also low (3%; Fig. 2F). Within the second nodule was a third minute microscopic nodular focus that was poorly defined and focally rimmed by a delicate fibrovascular capsule (Fig. 2G,H). The cells in this smallest nodule were large, pleomorphic, and characterized by a high nuclear-to-cytoplasmic ratio, inconspicuous nucleoli, and nuclear pseudo-inclusions. Multinu-
cleated giant cells and atypical mitotic figures were present. A Ki-67 immunostain showed high proliferative activity (60%) in this small nodule (Fig. 2I). An immunohistochemical panel was performed to characterize the tumor cells. The cells in all the nodules were strongly positive for inhibin (Dako; polyclonal,1:100) and Melan A (Dako; monoclonal, 1:100) in a cytoplasmic staining pattern. Chromogranin (Biogenex; monoclonal, 1:200) and S100 protein (Novocastra; polyclonal, 1:250) stains were negative in the en- tire lesion. A p53 stain (Dako; monoclonal, 1:1000) showed scattered nuclear staining in the larger 2 nodules, with the smallest nodule having been cut through entirely in the section.
The final diagnosis was that of adrenal cor- tical adenoma containing an oncocytic nodule with a central, highly proliferative focus. The possibility that this third minute proliferative fo- cus represented a form of in situ malignant change or incipient malignancy was suggested.
MOLECULAR ANALYSIS
To examine clonality and to elucidate the mutation rates in each of the morphologically distinct areas of this adrenal tumor, a molecular analysis was performed. Serial unstained 4-mm-thick sections were used for visually guided microdissection under a stereomicroscope. The microdissection targets included normal adjacent nontumor tissue and 6 different areas from the tumor nodules. Polymerase chain reaction was performed using primers for a panel of short tandem repeats near known tumor suppressor genes to assess for loss of heterozygosity, including 1p34 (D1S1193, D1S407, MYCL, probably p21 gene), 3p26 (D3S1539, D3S2303, VHL gene), 5q21 (D5S592, D5S615, probably APC gene), 9p21 (D9S254, D9S251, p16 gene), 9q (D9S252, PTCH gene), 10q23 (D10S520, D10S1173, pten gene), 17p13 (D17S974, D17S1289, p53 gene), and 22q13 (D22s532, NF2 gene). Polymerase chain reaction products were analyzed semiquantitatively using capillary elec- trophoresis. For heterozygous loci, the peak heights were compared between tumor and nor- mal. When the ratio of (tumor allele 1/tumor al- lele2)/(normal allele 1/normal allele 2) was less than 0.7, the result was interpreted as allelic
A
B
C
B
E
F
G
H
I
imbalance. Overall, allelic imbalance was not widespread in any of the microdissection targets. Four allelic losses were identified (3p26, 5q33, 17p13, and 21q22), but none of these showed evidence of clonal expansion throughout the mul- tiple different tumor targets sampled.
DISCUSSION
We have described a 3-cm functional adrenal tu- mor that was laparoscopically excised. The tumor had an unusual histologic and grossly a nesting or target-like appearance.
morphism but no increased mitoses (hematoxylin and eosin, 400x). F. Immunohistochemistry shows low pro- liferation in tumor cells within this nodule (Ki-67, 100x). G. Third small nodule with no definite capsule, showing large cells with pleomorphism and eosino- philic cytoplasm (hematoxylin and eosin, 100x). H. Higher magnification showing the marked nuclear pleomorphism and increased mitotic rate with atypical mitoses (hematoxylin and eosin, 400x). I. Im- munohistochemistry shows high proliferation index in this nodule (approximately 35% to 40%; Ki-67, 200x).
Differentiation between benign and malig- nant ACTs can be difficult. Grossly, malignant tu- mor are usually larger (>10.5 cm) and weigh more (>400 g) than benign tumors. The histopathologic criteria used to differentiate the 2 include capsular invasion, vascular invasion, increased mitosis including atypical forms (>15 mitoses per 20 high- power fields), and presence of tumor necrosis. A recent review of ACTs in children suggested a more frequent occurrence of malignant tumors in this age group. In that series, benign and malignant tumors showed evidence of hormonal production
[1]. Our patient also showed evidence of hormonal production that led to the discovery of the tumor. ACT can occur in some hereditary syndromes including Li-Fraumeni, Beckwith-Wiedemann, and multiple endocrine neoplasia syndrome type 1 [6]. There was no clinical evidence of any genetic association in this patient.
Recent studies have suggested that there may be molecular differences between adrenal cortical adenomas and carcinomas [7,8]. Molec- ular studies have demonstrated p53 mutations in familial and sporadic adrenal tumors [7]. Other defects such as loss of heterozygosity of 11p, 13q, and 17p have also been documented. None of these defects, however, help differentiate between adenomas and carcinomas. Little is known regarding mechanisms of tumorigenesis or if there is step-wise progression from normal adre- nal to adenoma to carcinoma. There is, however, a suggestion that key tumor suppressor genes play a role in tumorigenesis. One early event appears to be mutations or deletions in chromo- some 17 (p53), which can be present in adeno- mas and carcinomas. Additional multiple mutations occur in other chromosome loci including 11q (MEN1), 1q (frequent loss of het- erozygosity), and 9 (p15/p16) in carcinomas [8].
Our patient had a histologically and clinically unusual tumor that contained several morpho- logically distinct areas. Features of a typical adrenal cortical adenoma were clearly present at the periphery, but a second well-circumscribed nested nodule showed oncocytic differentiation. A third small indistinct nested nodule showed marked atypia, numerous mitoses, some atypical, and a high proliferative index. This raised a diag- nostic dilemma because this minute nodule had histologic, cytologic, and proliferative criteria for carcinoma but was entirely contained within the capsule of the larger tumor [1].
Although cases of adrenal tumors are com- mon in the literature, typically, most adrenal adenomas are solitary and unilateral. Although few cases of multiple adrenal adenomas exist, and multinodularity of adrenal cortical tumors have been described, we were unable to find a case that showed such concentric arrangement of nodules, with possible different biologic potentials. Also, there are no reports of any such highly prolifera-
tive foci in adenomas. Proliferation markers such as Ki-67 in combination with morphologic fea- tures are also considered predictive features for adrenocortical carcinomas [9]. Lin et al. [10] de- scribed oncocytic tumors with a high proliferative index that they have called oncocytic neoplasms of unknown biologic potential. However, their 2 tu- mors were large masses that showed increased proliferation throughout, which was not seen in our case.
In an attempt to further understand the biology of the 3 morphologically distinct nodules, we performed microdissection with sampling of 7 different regions in the tumor. Polymerase chain reaction was performed to assess for loss of het- erozygosity at different loci. Interestingly, the number of allelic losses was low in all targets examined. None of the areas sampled showed more than 2 allelic losses by loss of heterozy- gosity analysis, including the proliferative nodule. This paucity of allelic imbalance in all 3 nodules and the absence of any capsular or vascular invasion suggested that overall this lesion is best classified as a benign tumor [8,11,12]. The crite- ria laid down so far to diagnose adrenal cortical malignancies require the presence of multiple morphologic features in addition to evidence of virilization and multiple genetic defects [13-15]. Weiss et al. [14] suggested that at least 4 of their list of 9 pathologic criteria are required for a diagnosis of malignancy in ACT. Having said that, it is also clear that pathologic criteria ap- plied to adult ACT do not always apply to pedi- atric tumors and even adrenal adenomas in children show high proliferation or large size. The presence of only 2 criteria, high proliferation and increased mitoses including atypical forms, prevented us from making a definitive diagnosis of malignancy in our patient. However, because of the possibility of the central nodule behaving aggressively, long-term clinical follow-up would be essential. A detailed workup of the child ex- cluded any evidence of extra-adrenal disease. The child is currently being followed up with serial computed tomograms.
This case illustrates a scenario of an adrenal adenoma with an unusual nested morphology and a central, small, highly proliferative focus. This suggests the possibility of an adenoma-to-carci-
noma tumorigenesis sequence in ACTs. With early detection of adrenal lesions using radiology and early resection via laparoscopic surgery, the pos- sibility of diagnosing incipient carcinoma becomes more likely. Hence, pathologists will need to be viligant about sampling these tumors and clinicians will need to commit to long-term follow-up of the patients for us to truly understand more about the biology of ACTs.
REFERENCES
1. Wieneke JA, Thompson LDR, Heffess CS. Adrenal cortical neoplasms in the pediatric population. A clinicopathologic and immunophenotypic analysis of 83 patients. Am J Surg Pathol 2003;27:867-881.
2. Lack EE, Mulvihill JJ, Travis WD, Kozakewich HP. Adrenal cortical neoplasms in the pediatric and adolescent age group. Clinicopathologic study of 30 cases with emphasis on epidemiological and prognostic factors. Pathol Annu 1992;27:1-53.
3. Sabbaga CC, Avilla SG, Schulz C, Garbers JC, Blucher D. Adrenocortical carcinoma in children: clinical aspects and prognosis [see comment]. J Pediatr Surg 1993;28:841-843.
4. Latronico AC, Chrousos GP. Extensive personal experience: adrenocortical tumors. J Clin Endocrinol Metab 1997;82:1317-1324.
5. Wolthers OD, Cameron FJ, Scheimberg I, et al. Androgen secreting adrenocortical tumors [see comment]. Arch Dis Child 1999;80:46-50.
6. Gortz B, Roth J, Speel EJM, et al. MEN1 gene mutation analysis of sporadic adrenocortical lesions. Int J Cancer 1999;80:373-379.
7. McNicol AM, Nolan CE, Struthers AJ, Farquharson MA, Hermans J, Haak HR. Expression of p53 in adrenocortical tumors: clinicopathological correlations. J Pathol 1997;181: 146-152.
8. Zhao J, Speel EJM, Muletta-Feurer S, et al. Analysis of genomic alterations in sporadic adrenocortical lesions. Am J Pathol 1993;155:1039-1045.
9. Stojadinovic A, Brennan MF, Hoos A, et al. Adrenocortical adenoma and carcinoma: histopathological and molecular comparative analysis. Mod Pathol 2003;16:742-751.
10. Lin BTY, Bonsib S, Mierau GW, Weiss LM, Medeiros JL. Oncocytic adrenocortical neoplasms: a report of seven cases and review of the literature. Am J Surg Pathol 1998;22:603-614.
11. Sidhu S, Gicquel C, Bambach CP, et al. Clinical and molecular aspects of adrenocortical tumorigenesis. . Aust N Z J Surg 2003;73:727-738.
12. Sidhu S, Sywak M, Robinson B, Delbridge L. Adrenocorti- cal cancer: recent clinical and molecular advance. Curr Opin Oncol 2003;16:13-18.
13. Van Slooten H, Schaberg A, Smeenk D, Moolenaar AJ. Morphologic characteristics of benign and malignant adrenocortical tumors. Cancer 1985;55:766-773.
14. Weiss LM, Medeiros LJ, Vickery AL. Pathologic features of prognostic significance in adrenocortical carcinoma. Am J Surg Pathol 1989;13:202-206.
15. Bugg MF, Ribeiro RC, Roberson PK, et al. Correlation of pathologic features with clinical outcome in pediatric adrenocortical neoplasia. Am J Clin Pathol 1994;101:625- 629.