Jeffrey H. Newhouse, MD Clara S. Heffess, MD Brent J. Wagner, MD2 Thomas J. Imray, MD3 Carol F. Adair, MD Alan J. Davidson, MD
Index terms: Adrenal gland, CT, 86.12111, 86.12112
Adrenal gland, MR, 86.121411, 86.12143 Adrenal gland, neoplasms, 86.317, 86.324
Radiology 1999; 210:385-391
Abbreviation: AFIP = Armed Forces Institute of Pathology
1 From the Departments of Radiology (J.H.N., B.J.W., T.J.I., A.J.D.) and Endo- crine Pathology (C.S.H., C.F.A.), Armed Forces Institute of Pathology, Depart- ment of Radiologic Pathology, Wash- ington, DC. Received January 7, 1998; revision requested March 17; revision received June 8; accepted September 28. Address reprint requests to J.H.N., Department of Radiology, Co- lumbia-Presbyterian Medical Center, 177 Ft Washington Ave, New York, NY 10032.
2 Current address: West Reading Radi- ology Associates, West Reading, Pa.
3 Current address: Department of Radi- ology, University of Nebraska Medical Center, Omaha.
The opinions expressed herein are not to be construed as those of the Depart- ments of Defense, Army, or Navy. · RSNA, 1999
Author contributions:
Guarantor of integrity of entire study, J.H.N .; study design, J.H.N., B.J.W., A.J.D., T.J.I .; definition of intellectual content, J.H.N., C.S.H., B.J.W., T.J.I., C.F.A., A.J.D .; literature research, J.H.N .; data acquisition and analysis, J.H.N., C.S.H., B.J.W., T.J.I., C.F.A., A.J.D .; manu- script preparation, J.H.N. manuscript editing and review, J.H.N., B.J.W., T.J.I., A.J.D., C.S.H.
Large Degenerated Adrenal Adenomas: Radiologic-Pathologic Correlation1
PURPOSE: To correlate the radiologic and pathologic findings and differential diagnosis of large, degenerated adrenal adenomas.
MATERIALS AND METHODS: The authors reviewed the radiologic and pathologic characteristics of 30 large adenomas with cystic regions or areas of heterogeneity that were either intrinsic or demonstrated at contrast material-enhanced computed tomography (CT) or magnetic resonance (MR) imaging. Images of 24 adrenocortical carcinomas were also reviewed to determine whether differentiating characteristics existed.
RESULTS: Most of the adrenocortical adenomas were in asymptomatic women. Ten adenomas contained calcification. Pathologic examination revealed good correla- tion between heterogeneity and liquefied regions. Histologic examination confirmed regions of adenomatous tissue with areas of hemorrhage, amorphous degenerated material, calcification, and fibrosis. Some tumors contained myelolipomatous foci. Although some clinical and imaging findings differed between the groups, no features could be found that enabled the radiologic differentiation of adenomas from carcinomas.
CONCLUSION: A subgroup of adrenal adenomas are larger, more heterogeneous, and more frequently calcified than those with the usual imaging findings. Central necrosis, hemorrhage, or both are responsible for many of the imaging features. Differentiation of these lesions from other large adrenal masses, including adrenal carcinoma, cannot be made by means of imaging alone; resection is required for the definitive diagnosis.
Most adrenal adenomas have a limited range of radiologic findings (1-16): They are almost always smaller than 5 cm, are relatively homogeneous, and display rapid and relatively uniform enhancement with rapid washout of iodinated or gadolinium-based contrast material. They often reveal computed tomographic (CT) and magnetic resonance (MR) imaging findings associated with intracellular lipid and almost never liquefy; calcification is very rare. These lesions, however, may occasionally undergo intratumoral hemorrhagic degeneration with development of avascular and cystic internal regions and subsequent fibrosis. They then become much larger and display focal liquefaction, central or peripheral calcification and fibrosis, internal soft-tissue nodules, and patchy heterogeneous contrast enhancement, features that are commonly associated with adrenocortical carcinomas and other lesions. We performed this study to correlate the radiologic and pathologic findings in a series of such lesions, demonstrate the difficulty of distinguishing them from primary adrenal carcinomas, and discuss the implications of these features for their management.
MATERIALS AND METHODS
The cases were drawn from the files of the Department of Radiologic Pathology at the Armed Forces Institute of Pathology (AFIP), Washington, DC. Lesions were included in this
a.
b.
study if they were (a) definitively diagnosed as adenomas with degenerative changes af- ter independent review by two staff patholo- gists; (b) examined with CT, MR imaging, or both; and (c) larger than 5 cm in diameter with coarse heterogeneity (ie, irregularly shaped regions of variable attenuation or signal intensity that were at least half a centimeter in size) or cystic change at imag- ing examinations. All cases meeting these criteria were included. Images were re- viewed by four genitourinary radiologists (J.H.N., B.J.W., T.J.I., A.J.D.) together. The size, degree of heterogeneity and cystic change; presence of solid tissue nodules within cystic regions; presence and charac- teristics of calcification; and evidence of invasion of adjacent structures and veins were assessed. Twenty-four patients were found in the AFIP archives to have histologi- cally confirmed adrenocortical carcinoma; their images were studied in identical ways to determine whether features enabling the differentiation of the two groups of lesions could be found.
RESULTS
Adenoma Group
Patients .- Thirty patients with adreno- cortical adenomas with atypical radio- logic features were found. There were 20 women and 10 men. The average patient age was 63 years (range, 21-87 years); all but three patients were older than 53 years. Four patients had clinical or bio- chemical evidence of Cushing disease, two of whom also had elevated catechol- amine levels. The remainder of the pa- tients had no endocrine dysfunction.
Radiologic findings .- Twenty-five of the 30 patients underwent only CT, one pa- tient underwent only MR imaging, and four patients underwent both examina- tions. CT was performed without intrave- nous contrast material in 19 patients, with contrast material in seven, and both with and without contrast material in three. The dose of contrast material and timing of contrast material administra- tion were not available. In most patients, measurement of the exact CT attenuation was not performed. All contrast material- enhanced CT examinations were per- formed at a time at which contrast mate- rial was visible in the renal collecting systems. MR imaging was performed with spin-echo or fast spin-echo sequences; each patient underwent T1-weighted (rep- etition time, 34-600 msec; echo time, 11-21 msec [34-600/11-21]) and T2- weighted (2,500-6,000/104-134) imag- ing; no contrast agents were used, nor was fat-saturation or gradient-echo imag- ing performed. None of the lesions was preoperatively sampled for biopsy.
The adenomas (Figs 1-4) ranged in diameter from 5 to 20 cm (average, 11.1 cm; SD, 3.1 cm). Ten adenomas had calci- fication visible at CT. In one adenoma, the calcification was confined to the rim. Calcification was visible as irregularly shaped regions of various sizes scattered throughout the tumors in the remainder of adenomas. Two adenomas appeared completely cystic, 12 appeared to be al- most entirely cystic, and three had focal cystic regions. The walls of the cystic lesions were irregular, and six cases had soft-tissue nodules that appeared to pro- trude into the cystic regions. “Cystic”
regions were assumed to contain fluid on the basis of CT findings of homogeneous low attenuation (the attenuation of the cystic region was lower than that of muscle and blood) and by having very low signal intensity on T1-weighted MR images and very high signal intensity on T2-weighted MR images; these regions showed good correlation between MR and CT images in those cases in which both were available. The noncystic le- sions and the solid portions of the par- tially cystic lesions were heterogeneous on all images and enhanced heteroge- neously in those cases in which pre- and postcontrast CT images had been ob- tained.
Pathologic examination .- The diameter of the excised lesion (Figs 1-4) varied from the diameter measured on the im- ages by 10% or less. All lesions that appeared to be primarily cystic at CT or MR imaging had central regions that were filled with fluid, old thrombus, or gelati- nous material. The walls varied in thick- ness from 3 to 20 mm and tended to be irregular. In most cases, a portion of the wall consisted of the limbs of the adrenal gland splayed over the surface of the mass. The lesions that appeared to be heterogeneous, rather than cystic, con- tained focal regions of hemorrhage, necro- sis, or both; in one case, regions of fibrosis were found in the center of the lesion.
At histologic examination, the neo- plasms were well circumscribed and par- tially surrounded by connective tissue capsules. Occasionally, a thick capsule separated the adenoma from the remain- ing noninvolved cortex. Each had large, well-differentiated adrenocortical cells re-
a.
b.
a.
b.
sembling those seen in the adrenal cor- tex. The abnormal cortical cell prolifera- tion was of variable thickness; the irregular inner surface was contiguous to the degen- erative areas (Fig 5). Some of the cyst walls contained adipocytes and myeloid elements, which comprised small lipoma- tous and myelolipomatous foci. Some also contained small calcifications. The amount of connective tissue was also variable; occa-
sional bands of dense connective tissue de- limited the adenomatous tissue from the cystic or degenerated areas. The degenera- tive changes consisted mainly of hemor- rhage, amorphous material, calcification, fi- brosis, small and large cavernous blood vessels (“neovascularization”), and viable and nonviable islands of cortical cells (Fig 5). No mitoses or tumor necrosis of the kind seen in cortical carcinomas were found, nor
were any of the histologic patterns that are seen in carcinomas found.
Carcinoma Group
Patients .- Adrenocortical carcinomas were found in 11 male and 13 female patients ranging in age from 13 to 61 years (mean, 43 years). Twelve patients had endocrine dysfunction, including
a.
b.
*
C.
d.
Cushing syndrome, virilization, feminiza- tion, hyperaldosteronism, and elevated metanephrine and normetanephrine lev- els.
Twenty patients underwent only CT, one patient underwent only MR imaging, and three patients underwent both exam- inations. CT was performed without intra- venous contrast material in 17 patients, with contrast material in four, and with and without contrast material in two. Attenuation measurements were usually not available. MR imaging was performed
with spin-echo or fast spin-echo se- quences; each patient underwent T1- weighted (120-600/11-30) and T2- weighted (2,400-5,400/84-120) imaging. Gadolinium enhancement was used in one patient; gradient-echo techniques were not used.
Radiologic findings .- Carcinomas were on the left in 11 patients and on the right in 13. The carcinomas were slightly larger than the adenomas, ranging in diameter from 5 to 50 cm (average, 15 cm). Eight carcinomas (33%) contained calcifica-
tion; one had rim calcification and seven had irregular calcific deposits of varying sizes. None of the carcinomas had a uni- locular cystic configuration; one had fo- cal cystic degeneration. Two carcinomas were homogeneous; in the remainder, the attenuation, signal intensity, or both had irregular heterogeneity. None of the carcinomas had nodules in cystic regions. Four patients had evidence of metastatic or invasive disease: One patient had mul- tiple lung metastases and three patients had tumor thrombus that extended from
a.
C.
the primary neoplasm into the inferior vena cava.
Radiologic features that permitted con- fident differentiation between carcino- mas and adenomas were found in only a minority of cases. Metastases and tumor thrombi were present only in patients with carcinomas; however, 20 (83%) of the patients with carcinomas did not have these findings. Completely cystic lesions and lesions containing soft-tissue nodules were found only in patients with adenomas; however, 22 (73%) of the pa- tients with adenomas did not have these findings.
DISCUSSION
Adrenal adenomas are relatively com- mon in adults. They appear in autopsy series with reported prevalences of 2%-9% (17,18) and are frequently encountered incidentally at CT; literally hundreds of them have been described in the litera-
ture (1-16). The features of the adenomas in this series (large size, heterogeneity, cystic change, and calcification), how- ever, appear to be very rare among the reported lesions (1,19-22). The fact that 30 such adenomas could be found in the AFIP archives is almost certainly a reflec- tion of the requirement that accession into the radiologic pathology archives generally requires resection and examina- tion of the entire lesion, which very rarely occur with lesions that have typical radiologic appearances. Selection bias not- withstanding, the number of cases in this series suggests that radiologists should be aware that adenomas may have this atypi- cal appearance.
Our tumors were considerably larger than most that have been previously described. The growth rate of adrenal adenomas is not known, but is probably slow-indeed, a characteristic frequently invoked to indicate that an adrenal mass is an adenoma is its failure to increase in size during a lengthy observation. The
increasing prevalence of adrenal adeno- mas with increasing age, however, sug- gests that adrenocortical adenomas must grow, so it is not surprising that a few ultimately become large.
Our tumors were heterogeneous, in contrast to most of those described previ- ously. Heterogeneity may be linked with size: Larger tumors are more likely to become ischemic centrally, and ischemic regions may become fibrotic or liquefy, either of which may produce the hetero- geneity seen at CT and MR imaging. Whether the small amount of fatty tissue found in the tumors was in part respon- sible for the heterogeneity is not clear. Similar findings have been reported in other adrenal adenomas (3,23-25); the reason for the association of tissue types is not known.
A large lesion may also experience inter- nal hemorrhage (15,26-28) and ulti- mately evolve to become a “pseudocyst.” The cystic region may be so much more voluminous than the solid tumor that the adenoma might be overlooked in radiologic examinations, during surgical inspection, and even at pathologic exami- nation. Several of the lesions in our series were originally diagnosed as cysts, pseudo- cysts, or hemangiomas and were subse-
quently found to be adenomas only after careful microscopic examination revealed residual neoplastic cortical tissue (29-31). Neoplastic regions, which contain cells that have more cytoplasm and are larger than normal cells, must be distinguished from islands of normal cortical tissue, which have been described in hemor- rhagic pseudocysts arising in adrenal glands without tumors (29).
The appearances of our lesions pose problems for making the differential diag- nosis. As we have shown, these adenomas share many radiologic features with adre- nocortical carcinomas, which are usually large and heterogeneous and frequently contain necrotic regions. About one- third of adenomas calcify (32,33). Al- though adrenocortical carcinomas may display specific features of malignancy, including invasion of adjacent structures, metastases, and venous tumor thrombi, the absence of these features by no means is proof of benignity. The functional sta- tus of a tumor usually does not permit differentiation between carcinoma and adenoma. Many of the tumors in our series were resected because their size and heterogeneity led to a preoperative diag- nosis of carcinoma. These lesions can be misdiagnosed histologically if the criteria for malignancy (34), which include mito- ses, a trabecular growth pattern, and an increased nuclear-cytoplasmic ratio, are not applied correctly, so the mistake may be perpetuated after pathologic examina- tion (35,36). It is also easy to miss the neoplasm completely at histologic exami- nation and to call the lesion a pseudocyst (37) or hemangioma (38) due to the marked cystic degeneration and vascular proliferation.
Metastases to the adrenal glands have a wide range of sizes; large ones may be heterogeneous. Unlike the tumors in this series, metastases are unlikely to calcify and may be bilateral. They are usually discovered in patients already known to have primary malignancies elsewhere; oc- casionally, they may be the primary mani- festation of the disease (39). None of the tumors in this series was believed preop- eratively to represent a metastasis.
Pheochromocytomas may share fea- tures with tumors in this series (40,41), including large size, cystic regions, and occasional calcification. Most pheochro- mocytomas, however, produce endocrine abnormalities and characteristic symp- toms that permit a specific diagnosis, as do the features they sometimes display on T2-weighted MR images and scinti- grams obtained with metaiodobenzylgua- nidine. None of our tumors was believed
preoperatively to be a pheochromocy- toma.
Calcified cystic lesions may also be produced by hydatid disease (15,42,43), and epithelial cysts without neoplasm have also been described (44,45).
The tumors in our series were selected so that none demonstrated the homoge- neous low-attenuation appearance that permits specific diagnosis at CT (9), but other techniques for diagnosing adeno- mas, such as needle biopsy and out-of- phase gradient-echo MR imaging, were not applied to any of our tumors. We suspect these techniques would not have been effective. Loss of signal has not been shown to be diagnostic of adenoma if it occurs in only a small portion of a large, heterogeneous lesion imaged with gradi- ent-echo techniques with fat and water out of phase (5). Biopsy of the adenomas in this series would frequently have recov- ered necrotic and hemorrhagic tissue and, even if a small sample were to contain viable adenoma, it would be difficult to be confident that no other part of a heterogeneous tumor contained malig- nant tissue. We conclude, therefore, that these radiologically atypical adenomas will continue to require resection.
Acknowledgment: The illustrations contained herein come from the cases in the archives of the department of radiologic pathology. The authors are indebted to all of the radiologists and pathologists whose contributions of out- standing cases to the AFIP have made this publication possible.
References
1. Ichikawa T, Ohtomo K, Uchiyama G, et al. Contrast-enhanced dynamic MRI of adre- nal masses: classification of characteristic enhancement patterns. Clin Radiol 1995; 50:295-300.
2. Korobkin M, Lombardi TJ, Aisen AM, et al. Characterization of adrenal masses with chemical shift and gadolinium-enhanced MR imaging. Radiology 1995; 197:411- 418.
3. . Sato N, Watanabe Y, Saga T, et al. Adreno- cortical adenoma containing a fat compo- nent: CT and MR image evaluation. Ab- dom Imaging 1995; 20:489-490.
4. Korobkin M. Overview of adrenal imaging/ adrenal CT. Urol Radiol 1989; 11:221- 226.
5. Outwater EK, Siegelman ES, Huang AB, et al. Adrenal masses: correlation between CT attenuation value and chemical shift ratio at MR imaging with in-phase and opposed-phase sequences. Radiology 1996; 200:749-752.
6. Korobkin M, Brodeur FJ, Francis IR, et al. Delayed enhanced CT for differentiation of benign from malignant adrenal masses. Radiology 1996; 200:737-742.
7. Korobkin M, Lombardi TJ, Aisen AM, et al. Characterization of adrenal masses with chemical shift and gadolinium-enhanced
MR imaging. Radiology 1995; 197:411- 418.
8. Reinig JW, Stutley JE, Leohardt CM, et al. Differentiation of adrenal masses with MR imaging: comparison of techniques. Radiology 1994; 192:41-46.
9. Lee MJ, Hahn PF, Papanicolaou N, et al. Benign and malignant adrenal masses: CT distinction with attenuation coefficients, size, and observer analysis. Radiology 1991; 179:415-418.
10. Krestin GP, Steinbrich W, Friedmann G, et al. Adrenal masses: evaluation with fast gradient-echo MR imaging and Gd-DTPA- enhanced dynamic studies. Radiology 1989; 171:675-680.
11. 1. Chezmar, JL, Robbins SM, Nelson RC, et al. Adrenal masses: characterization with T1-weighted MR imaging. Radiology 1988; 166:357-359.
12. Reinig JW, Doppman JL, Dwyer AJ, et al. Adrenal masses differentiated by MR. Ra- diology 1986; 158:81-84.
3. Glazer GM, Woolsey EJ, Borrello J, et al. Adrenal tissue characterization using MR imaging. Radiology 1986; 158:73-79.
14. Tsushima Y, Ishizaka H, Kato T, et al. Differential diagnosis of adrenal masses using out-of-phase FLASH imaging: a pre- liminary report. Acta Radiol 1992; 33:262- 265.
15. Sroujieh AS, Farah GR, Haddad MJ, et al. Adrenal cysts: diagnosis and manage- ment. Br J Urol 1990; 65:570-575.
16. Epstein S, Goldin AR, McLaren EH. Anomalous responses to stimulation and suppression tests in Cushing’s syndrome due to a calcified adrenal adenoma. Post- grad Med J 1973; 49:923-926.
17. . Kokko JP, Brown TC, Berman MM. Adre- nal adenoma and hypertension. Cancer 1967; 1:468-470.
18. Shamma AH, Goddard JW, Sommers SC. A study of the adrenal status in hyperten- sion. J Chron Dis 1958; 8:587-595.
19. Paster SB, Rosen RT. Calcification in be- nign adrenal adenoma. Urology 1974; 3:646-648.
20. Jelic TM, Triest W, Sheils J, et al. Calcified aldosterone-producing adrenocortical ad- enoma. Clin Imaging 1996; 20:50-54.
21. Paster SB, Rosen RT. Calcification in be- nign adrenal adenoma. Urology 1974; 3:646-648.
22. Rosenblit A, Morehouse H, Amis ES Jr. Cystic adrenal lesions: CT features. Radiol- ogy 1996; 201:541-548.
23. Weiner SN, Bernstein RG, Lowy S, et al. Case report: combined adrenal adenoma and myelolipoma. J Comput Assist To- mogr 1981; 5:440-442.
24. Miyake O, Hara T, Matsumiya K, et al. Adrenal myelolipoma associated with Cushing’s syndrome: a case report. Hin- yokikia Kiyo 1992; 38:681-684.
25. Hedeland H, Ostbery G, Hokfelt B. On the prevalence of adrenocortical adenomas in an autopsy material in relation to hyper- tension and diabetes. Acta Med Scand 1968; 184:211-214.
26. Rhodes RE, Gaede JT, Meyer GA, et al. Hemorrhagic adrenal adenoma simulat- ing myelolipoma: CT evaluation. Com- put Tomogr 1991; 16:301-304.
27. Saito T, Kurumada S, Kawakami Y, et al. Spontaneous hemorrhage of an adrenal cortical adenoma causing Cushing’s syn- drome. Urol Int 1996; 56:105-105.
28. Ghandur-Mnaymneh L, Slim M, Muakassa
K. Adrenal cysts: pathogenesis and histo- logical identification with a report of 6 cases. J Urol 1979; 122:87-91.
29. Gaffey MJ, Mills SE, Fechner RE, et al. Vascular adrenal cysts: a clinicopatho- logic and immunohistochemical study of endothelial and hemorrhagic (pseudocys- tic) variants. Am J Surg Pathol 1989; 13: 740-747.
30. Medeiros LJ, Lewandrowski KB, Vickery AL. Adrenal pseudocyst: a clinical and pathologic study of eight cases. Hum Pathol 1989; 20:660-665.
31. Granger JK, Houn HY, Collins C. Massive hemorrhagic functional adrenal adenoma histologically mimicking angiosarcoma: re- port of a case with immunohistochemical study. Am J Surg Pathol 1991; 15:699-704.
32. Kenney PJ, Stanley RJ. Calcified adrenal masses. Urol Radiol 1987; 9:9-15.
33. Fishman EK, Deutch BM, Hartman DS, et al. Primary adrenocortical carcinoma. AJR 1987; 148:531-535.
34. Weis LM. Comparative histologic study of 43 metastasizing and non-metastasizing
adrenocortical tumors. Am J Surg Pathol 1984; 8:163-169.
35.
Desai MB, Kapadia SN. Feminizing adreno- cortical tumors in male patients: ad- enoma vs carcinoma. J Urol 1988; 139: 101-103.
36. Yamakita N, Saitoh M, Mercado-Asis LB, et al. Asymptomatic adrenal tumor: 386 cases in Japan including our 7 cases. En- docr J 1990; 37:671-684.
37. Griffing GT, Chobainian AV, Egdahl R, et al. An adrenal cyst associated with 19-nor- deoxycorticosterone excess and low renin hypertension. A. Theory and practice. Clin Exp Hypertens 1989; 11:317-321.
38. Rothberg M, Bastidas J, Mattey WE, et al. Adrenal hemangiomas: angiographic ap- pearance of a rare tumor. Radiology 1978; 126:341-344.
39. Kasperlik-Zeluska AA, Roslonowska E, Slowinska-Srzednicka J, et al. Incidentally discovered adrenal mass (incidentaloma): investigation and management of 208 patients. Clin Endocr 1997; 46:29-37.
40. Bush WH, Elder JS, Crane RE, et al. Cystic
pheochromocytoma. Urology 1985; 25: 332-334.
41. Krebs TL, Wagner BL, Kenney PJ. MR appearance of pheochromocytoma of the adrenal gland with pathologic correlation (abstr). Radiology 1997; 205(P):342.
42. Bastounis E, Pikoulis E, Leppaniemi A, et al. Hydatid disease: a rare cause of adrenal cyst. Am Surg 1996; 62:383-385.
43. Bartsch G, Bodner E, Buchsteiner R, et al. Echinococcal cyst of the adrenal gland. Eur Urol 1975; 1:240-242.
44. Tung GA, Pfister RC, Papanicolaou N, et al. Adrenal cysts: imaging and percutane- ous aspiration. Radiology 1989; 173:107- 110.
45. Medeiros LJ, Weiss LM, Vickery AL Jr. Epithelial-lined (true) cyst of the adrenal gland: a case report. Br J Urol 1989; 20: 491-492.