Myxoid Neoplasms of the Adrenal Cortex A Rare Histologic Variant
Felix M. Brown, M.D., Thomas A. Gaffey, M.D., Lester E. Wold, M.D., and Ricardo V. Lloyd, M.D., Ph.D.
The myxoid variant of adrenocortical carcinoma is a rare neo- plasm described previously in only two case reports. Because of the rarity of these lesions, the presence of myxoid changes in adrenal cortical neoplasms usually raises the possibility of ma- lignancy. We studied the histopathologic features of 14 cases of myxoid adrenocortical neoplasms, including six adenomas and eight carcinomas. All patients with adenomas with sufficient follow-up (n = 5) were alive with no recurrence of their tu- mors or evidence of metastatic disease. Four patients with car- cinomas died of their disease, two were alive with metastatic disease, and one was alive with no evidence of recurrence or metastatic disease. Histologically, the 14 tumors varied in their myxoid composition, ranging from 10% to 95%. The myxoid foci stained positively with Alcian blue and were usually nega- tive with periodic acid-Schiff and mucicarmine stains. As a group, the immunophenotype of the lesions was typical of other adrenal cortical neoplasms, with positive immunostaining for vimentin, synaptophysin, and «-inhibin. One tumor was focally positive for keratin. Myxoid adrenal cortical neoplasms should be included in the differential diagnosis of myxoid retroperito- neal neoplasms. Myxoid changes in adrenal cortical neoplasms may be present in both adenomas and carcinomas, and the usual clinical and histopathologic features for adrenocortical neo- plasms should be used to diagnose these neoplasms.
Key Words: Adrenal cortex-Adrenocortical carcinoma- Adrenocortical adenoma-Myxoid-Inhibin-Synaptophysin.
Am J Surg Pathol 24(3): 396-401, 2000.
Adrenocortical carcinomas are rare malignant neo- plasms of the adrenal cortex that are sometimes active hormonally. Patients are generally affected during the fourth and fifth decades of life, although adrenocortical carcinomas may occur at any age.4 Grossly, the lesions range in weight from a few grams to 5 kg, and most are
larger than 100 g.6 They are variably encapsulated, and gray to brown, pink, or yellow with areas of necrosis and hemorrhage.6 Histologically, the tumors have architec- tural patterns ranging from diffuse to alveolar,2,4,6,14 with a trabecular pattern being most common.4 A rare histo- logic variant, of which there have been two case reports in the English literature,2,12 is the myxoid adrenocortical carcinoma. These lesions are composed histologically of polygonal or elongate cells with moderate amounts of lightly basophilic cytoplasm. The cells are usually arranged in delicate arborizing cords, one to two cells thick, surrounding relatively large acellular clear spaces resembling the background of a myxoma.2,12 In one case, the myxoid area was only focal,2 but the other case was described grossly as gelatinous,12 suggesting that a larger percentage of the tumor had this myxoid change. As a consequence of these two reports, myxoid foci in adrenal cortical neoplasms has often raised the possibil- ity of malignancy.4 In this study we report the clinico- pathologic features of six cases of myxoid adrenocortical adenomas, and eight cases of myxoid adrenocortical carcinomas.
METHODS
The patients were acquired from the pathology files of the Mayo Clinic from 1965 to 1999 and from the con- sultation files of the authors (R.V.L., L.E.W.). A total of 300 adrenocortical tumors from the files of the Mayo Clinic from 1965 to 1984 were reviewed, and tumor with 10% or more myxoid areas were included in the study. The pathology reports, hematoxylin-eosin-stained slides, and paraffin-embedded blocks or unstained slides were available for all patients. Follow-up data were available for all patients except one with an adenoma and one with a carcinoma. Paraffin-embedded tissue sections were stained with Alcian blue (pH, 2.5), periodic acid-Schiff (PAS), and mucicarmine stains. Immunohistochemical
From the Department of Pathology (T.A.G., L.E.W., R.V.L.), Mayo Clinic and Mayo Foundation, Rochester, MN; and the Department of Pathology (F.M.B.), Brigham and Women’s Hospital and Harvard Medical School, Boston, MA, U.S.A.
Address correspondence and reprint requests to Ricardo V. Lloyd, MD, PHD, Mayo Clinic, Department of Laboratory Medicine and Pa- thology, 200 First Street SW, Rochester, MN 55905, U.S.A.
staining for keratin (CAM5.2; 1:800; Dako, Carpinteria, CA, USA), a-inhibin (1:50; Serotec, Toronto, Canada), synaptophysin (1:40; ICN, Los Angeles, CA, USA), vi- mentin (1:500; Dako), and desmin (1:100; Dako) was performed with the avidin-biotin-peroxidase complex method with appropriate positive and negative controls as reported previously.7 Ultrastructural studies were per- formed using tissues fixed previously in buffered forma- lin. Statistical analyses were done on 2 × 2 tables using Fisher’s exact test.
RESULTS
The clinical findings in the 14 cases of myxoid adre- nocortical neoplasms are summarized in Table 1. A re- view of 300 adrenocortical tumors in the files at the Mayo Clinic from 1965 to 1984 showed that less than 3% of the carcinomas and 1% of the adenomas had sub- stantial (>10%) myxoid areas. The patients ranged in age from 16 to 73 years. Patients with a diagnosis of carci- noma were older than those diagnosed with adenoma (age, >40 years; p = 0.01). The overall female-to-male ratio was 2.5:1 (5:1 for the adenomas and 1.7:1 for the carcinomas). Data on the hormonal status of patients at the time of presentation were available in 12 patients. Eight patients (one adenoma and seven carcinomas) pre- sented with evidence of glucocorticoid excess. Six of the carcinoma patients manifested clinically with Cushing’s syndrome, including one male patient with concomitant feminization and one female patient with concomitant elevated aldosterone levels. Two additional patients (one
with an adenoma and one with a carcinoma) presented with primary hyperaldosteronism.
The follow-up period extended from the time of pri- mary surgical excision to 33 years, with a mean of 12 years for patients with adenomas and 1.6 years for pa- tients with carcinomas (Table 1). All five patients with adenoma for whom there was follow-up data were alive, and none had any local recurrence or metastatic disease. Two of the patients with carcinoma had locally recurrent disease, and six had metastatic disease, including four at the time of presentation. The liver was the most common site of metastasis (n = 4) followed by the lungs (n = 2). Four patients with carcinomas died of disease: one with extensive local disease and three of metastatic disease.
Pathologic Findings
The pathologic findings in the 14 cases are summa- rized in Table 2. The adenomas were generally smaller than the carcinomas, most weighing less than 100 g and measuring 7.5 cm or less in greatest diameter. Necrosis and vascular invasion were absent in the adenomas, whereas five cases of carcinomas had areas of confluent necrosis and three cases showed vascular invasion. Four cases of adenomas and six cases of carcinomas had areas of capsular invasion, but extensive extracapsular in- volvement was present only in carcinomas. Mitoses numbered three or less per 10 high power fields (HPF) among the adenomas, whereas most carcinomas had more than three mitoses per 10 HPF (range, three to 13 mitoses per 10 HPF; p = 0.002).
| Case | Age (yrs) | Sex | Hormonal status | Time to local recurrence | Time to metastasis | Location of metastasis | Total follow-up period | Current status |
|---|---|---|---|---|---|---|---|---|
| Adenomas | ||||||||
| 1 | 21 | F | Slightly elevated blood cortisol levels | None | None | None | 3 yrs | Alive |
| 2 | 35 | F | — | None | None | None | 6 mos | Alive |
| 3 | 31 | F | Normal laboratory values | None | None | None | 5 yrs | Alive |
| 4 | 73 | M | — | — | — | — | — | — |
| 5 | 16 | F | Conn's syndrome | None | None | None | 33 yrs 8 mos | Alive |
| 6* | 62 | F | Normal laboratory values | None | None | None | 18 yrs 6 mos | Alive |
| Carcinomas | ||||||||
| 7 | 44 | F | Markedly elevated blood cortisol levels | None | None | None | 2 yrs 3 mos | Alive |
| 8 | 48 | M | Conn's syndrome | — | — | — | — | — |
| 9 | 43 | F | Cushing's syndrome | 7 mos | 7 mos | Liver | 1 yr 1 mo | AWD |
| 10 | 48 | M | Cushing's syndrome | 1 yr and 3 yrs | 0 mos | Omentum | 3 yrs 4 mos | DOD: 3 yrs 4 mos |
| 11 | 53 | F | Cushing's syndrome | None | 0 mos | Liver | 2 yrs 5 mos | DOD 2 yrs 5 mos |
| 12 | 51 | M | Cushing's syndrome | None | 1 yr | Liver, eye, lung | 1 yr 2 mos | DOD 1 yr 2 mos |
| 13 | 63 | F | Cushing's syndrome | None | 0 mos | Brain | 5 mos | DOD 5 mos |
| Elevated aldosterone levels | 4 mos | Liver | ||||||
| 14 | 56 | F | Cushing's syndrome | None | 0 mos | Lung | 3 mos | AWD |
0 mos, at presentation; AWD, alive with disease; DOD, died with disease.
* This patient also had MEN1.
| Case | Weight (g) | Diameter (cm) | Myxoid area (%) | Necrosis | Mitoses | Vascular invasion | Capsular invasion |
|---|---|---|---|---|---|---|---|
| Adenomas | |||||||
| 1 | 72.2 | 7.5 | 95 | None | 3/10 HPF | None | None |
| 2 | 13 | 3.5 | 40 | None | 0/10 HPF | None | Focal |
| 3 | 37.22 | 6 | 70 | None | 2/10 HPF | None | Focal |
| 4 | 43 | 4 | 50 | None | 0/10 HPF | None | Focal |
| 5 | — | 7 | 80 | None | 1/10 HPF | None | Focal |
| 6 | 106 | 6 | 40 | None | 1/10 HPF | None | None |
| Carcinomas | |||||||
| 7 | 180 | 9 | 60 | None | 4/10 HPF | None | None |
| 8 | 151.9 | 7.7 | 20 | Present | 7/10 HPF | None | Present |
| 9 | 65 | 8 | 10 | None | 6/10 HPF | None | Present |
| 10 | 195 | 9 | 50 | None | 5/10 HPF | Present | Present |
| 11 | 138 | 8.4 | 60 | Present | 13/10 HPF | None | Present |
| 12 | 700 | 16 | 60 | Present | 5/10 HPF | Present | Focal |
| 13 | — | 11 | 20 | Present | 9/10 HPF | None | Present |
| 14 | 470 | 10 | 30 | Present | 3/10 HPF | Present | None |
HPF, high-power fields.
A common feature in all 14 cases was the presence of prominent, often grossly evident, myxoid matrix mate- rial, ranging from as little as 10% of the tumor to more than 70% in some cases (Fig. 1). In areas that resembled more closely conventional adrenal cortical neoplasia his- tologically, the tumor cells formed trabeculae or nests separated by delicate, fibrovascular stromal septa. The cells were polygonal in shape with distinct cell borders and had moderate amounts of clear, finely granular, eo- sinophilic, or amphophilic cytoplasm. The nuclei were enlarged, round to oval, hyperchromatic, and mildly to moderately pleomorphic with distinct nucleoli.
The myxoid areas exhibited a number of architectural patterns (Figs. 1B, C). Some tumors had thin, one to two-cell-thick, delicate, anastomosing cords whereas others formed broad, irregularly bordered clusters. The cells appeared to float loosely in a copious acellular myxoid background divided into small lobules by thin, fibrovascular stroma. Other tumors had a nested pattern, with the constituent cells separated from their fibrous stromal frameworks by variable amounts of myxoid ma- trix. Particularly striking in some cases was a well- formed pseudoglandular architecture in which the cell nests acquired distinct central lumens with myxoid changes in the lumen (Fig. 1B). A pseudoglandular pat- tern was present in four adenomas and two carcinomas. Individual tumor cells in the myxoid foci were often smaller than those in the more conventional areas, with polygonal to elongated and rectangular shapes. In most cases, they retained ample cytoplasm that was predomi- nantly eosinophilic, with only occasional clear cells. Nu- clei were slightly enlarged, round to oval, and hyperchro- matic, with distinct nucleoli. Nuclear pleomorphism was generally minimal. Of note, mitoses, when present, were generally less frequent in myxoid areas. Myxoid changes were more common in carcinomas than in adenomas.
Histochemical, Immunohistochemical, and Ultrastructural Findings
Table 3 summarizes the results of the histochemical and immunohistochemical staining of paraffin- embedded sections from the 14 tumors. Alcian blue staining showed the presence of extracellular acidic mu- cosubstances in the acellular myxoid areas in all 14 cases (Fig. 1D). Focal weak staining with PAS was present in three cases. Focal staining for mucicarmine positivity was present in three tumors. None of the histochemical stains demonstrated intracellular mucosubstances, sug- gesting that the myxoid material is a stromal product.
All 14 tumors stained for synaptophysin, as did the normal medulla (Fig. 1E). Staining for «-inhibin was positive focally to diffusely in all cases (Fig. 1F) except for one adenoma. In cases with residual normal adrenal, inhibin positivity was demonstrable in the normal cortex, and the most intense staining was present in the zona reticularis. Twelve of 14 tumors were vimentin positive, and all tumors were negative for desmin. Only one case demonstrated focal dotlike positivity in tumor cells for cytokeratins. Histochemical and immunohistochemi- cal staining revealed no differences between adenomas and carcinomas or among cortisol-producing tumors, aldosterone-producing tumors, and nonfunctioning tumors.
Ultrastructural studies of two adenomas and one car- cinoma showed cells with well-developed rough and smooth endoplasmic reticulum. Some of the mitochon- dria had blunt cristae. Neurosecretory granules were not present. The clear spaces seen on light microscopy con- sisted of acellular myxoid material (Fig. 2).
DISCUSSION
Myxoid adrenocortical carcinoma was first reported by Tang et al.12 in 1979 and a second case was added to
METRIC 1
A
B
C
D
E
F
| Stain | Positive cases | |
|---|---|---|
| Adenomas (n = 6) | Carcinomas (n = 8) | |
| Histochemical | ||
| Alcian Blue | 6 | 8 |
| Periodic acid-Schiff | 2 | 1 |
| Mucicarmine | 1 | 1 |
| Immunohistochemical | ||
| Inhibin | 5 | 8 |
| Synaptophysin | 6 | 8 |
| Vimentin | 5 | 7 |
| Keratin (CAM5.2) | 1 | 0 |
| Desmin | 0 | 0 |
the literature by Forsthoefel2 in 1994. However, these lesions are rare and may not come readily to mind when confronted with a myxoid neoplasm, especially in a needle biopsy. Our study of 14 myxoid adrenocortical neoplasms indicate that these changes are uncommon in
A
B
tumors of the adrenal cortex and that they can be seen in adenomas as well as carcinomas.
Despite the description by Tang et al.12 of a grossly gelatinous lesion, subsequent reports have suggested that the myxoid histology is usually only a focal finding.2,4 In our series, more than half the cases demonstrated at least 50% myxoid histology, and one tumor had more than 90% myxoid areas. The histochemical profile of the myxoid material in our series was consistent with that in previous case reports,2,12 with prominent Alcian blue positivity of the extracellular mucosubstances and nega- tive or focal weak staining with PAS stain and for mucicarmine.
Vimentin is expressed commonly in cells of the adre- nal cortex, as well as adrenal cortical neoplasms, and was present in most of our cases. All 14 adrenocortical car- cinomas and adenomas were positive for synaptophy- sin.3,5,9 One case showed focal dot-like cytokeratin posi- tivity, but the remaining 13 cases were negative.
Inhibin is a glycoprotein that acts to inhibit secretion of follicle-stimulating hormone by the anterior pituitary gland. It is composed of an «-subunit and one of two B-subunits and is produced by ovarian granulosa and theca cells, testicular Sertoli cells, and adrenocortical cells.1 Inhibin «-subunit expression has been detected in Leydig cells1,14 and granulosa cells of the ovary,13 and extragonadally in the adrenal gland, placenta, pituitary gland,8,13 liver,8 and central nervous system.13 Nor- mal human adrenal cortex and adrenal cortical tumors, especially cortisol-producing adenomas, produce and secrete inhibinlike immunoreactivity into the peripheral blood.10,11 Granular, cytoplasmic inhibin immunoreac- tivity has been detected in normal human adrenal cor- tex,1,8 the intensity increasing progressively from the glomerular to the reticular layer,1,8 whereas the adrenal medulla is typically negative.1,8 The vast majority of adrenocortical hyperplasias,8 adenomas, and carcino- mas1,8,11 stain positively for inhibin, and some studies suggests that tumors with glucocorticoid1,11 or andro- gen11 overproduction stain more intensely than other types. Like the adrenal medulla,1,8 most pheochromocy- tomas are negative.1,11 Renal cell carcinomas1,11 and other types of metastases1 are also negative. Most of our tumors (13 of 14) showed inhibin immunoreactivity. In sections containing normal adrenal cortex, there was more intense staining of the zona reticularis than in the two other zones.
Both previously described myxoid adrenocortical neo- plasms were carcinomas, with both patients developing distant metastases.2,12 Because of the rarity of these le- sions, these two reports suggested that all of these tumors might be malignant. Our study shows that some adrenal cortical tumors with myxoid change are also benign based on histologic features and absence of metastatic disease after a mean follow-up period of 12 years for the
benign tumors. Although myxoid foci were slightly more common among adrenocortical carcinomas, the number of cases was too small to show any significant difference from the myxoid adrenocortical adenomas.
The differential diagnosis of a myxoid neoplasm in the retroperitoneum includes chordomas4 and carcinomas with myxoid areas, myxoma, lipoma, liposarcoma (myxofibrosarcoma-like foci rather than true myxoid li- posarcoma), benign or malignant nerve sheath tumors,2 2 myxoid leiomyomas, and leiomyosarcomas.2,12 Careful histopathologic examination of the tumors and a battery of immunohistochemical stains and/or ultrastructural studies can help to make the distinction between these lesions.
The presence of clear to eosinophilic cytoplasm with variable amounts of lipid vacuoles along with positive immunostaining for synaptophysin and inhibin are help- ful diagnostic features of myxoid adrenocortical tu- mors. The ultrastructural features of abundant smooth endoplasmic reticulum, lipid droplets, and prominent, stacked, rough endoplasmic reticulum helps to support the diagnosis. Among the tumors in the differential di- agnosis, chordomas may be the most difficult to distin- guish from myxoid adrenocortical tumors. However, the presence of strong positive staining for keratin in chor- domas, unlike myxoid adrenocortical tumors and the ab- sence of abundant smooth endoplasmic reticulum on ul- trastructural examination, allows one to make this dis- tinction. Most myxoid carcinomas are also strongly positive for keratin, unlike the adrenocortical tumors in this series.
Among the mesenchymal tumors in the differential diagnosis, myxomas are much more sparsely cellular than myxoid adrenocortical tumors. Lipomas and lipo- sarcomas have distinct adipocytes and are S-100 protein positive, whereas this marker is negative in myxoid ad- renocortical tumors. Nerve sheath tumors are usually S-100 protein positive and inhibin negative, unlike myxoid adrenocortical tumors. Myxoid leiomyomas and leiomyosarcomas consist of spindle and epithelial cells, which usually express muscle markers such as
desmin and smooth muscle actin and are negative for synaptophysin.
In conclusion, myxoid changes can be seen in both benign and malignant adrenocortical neoplasms, and the usual histopathologic features4,14 must be used to predict the biologic behavior of myxoid adrenocortical tumors. ☐
REFERENCES
1. Chivite A, Matias-Guiu X, Pons C, Algaba F, Prat J. Inhibin A expression in adrenal neoplasms: a new immunohistochemical marker for adrenocortical tumors. Appl Immunohistochem 1998;6: 42-9.
2. Forsthoefel KF. Myxoid adrenal cortical carcinoma: a case report with differential diagnostic considerations. Arch Pathol Lab Med 1994;118:1151-3.
3. Haak HR, Fleuren GJ. Neuroendocrine differentiation of adreno- cortical tumors. Cancer 1995;75:860-4.
4. Lack EE. Tumors of the adrenal gland and extra-adrenal para- ganglia. Washington, DC: Armed Forces Institute of Pathology, 1997.
5. Li Q, Johansson H, Kjellman M, Grimelius L. Neuroendocrine differentiation and nerves in human adrenal cortex and cortical lesions. APMIS 1998;106:807-17.
6. Lloyd RV. Endocrine pathology. New York: Springer-Verlag, 1990.
7. Lloyd RV, Scheithauer BW, Kovacs K, Roche PC. The immuno- phenotype of pituitary adenomas. Endocr Pathol 1996;7:145-50.
8. McCluggage WG, Burton J, Maxwell P, Sloan JM. Immunohisto- chemical staining of normal, hyperplastic, and neoplastic adrenal cortex with a monoclonal antibody against alpha inhibin. J Clin Pathol 1998;51:114-6.
9. Miettinen M. Neuroendocrine differentiation in adrenocortical car- cinoma. New immunohistochemical findings supported by electron microscopy. Lab Invest 1992;66:169-74.
10. Nishi Y, Haji M, Takayanagi R, Yanase T, Ikuyama S, Nawata H. In vivo and in vitro evidence for the production of inhibin-like immunoreactivity in human adrenocortical adenomas and normal adrenal glands: relatively high secretion from adenomas manifest- ing Cushing’s syndrome. Eur J Endocrinol 1995;132:292-9.
11. Pelkey TJ, Frierson HF Jr, Mills SE, Stoler MH. The alpha subunit of inhibin in adrenal cortical neoplasia. Mod Pathol 1998;11:516- 24.
12. Tang CK, Harriman BB, Toker C. Myxoid adrenal cortical carci- noma: a light and electron microscopic study. Arch Pathol Lab Med 1979;103:635-8.
13. Voutilainen R. What is the function of adrenal inhibins? Eur J Endocrinol 1995;132:290-1.
14. Weiss LM. Comparative histologic study of 43 metastasizing and non-metastasizing adrenocortical tumors. Am J Surg Pathol 1984; 8:163-9.