Primary Adrenocortical Carcinoma: CT Evaluation with Clinical Correlation
Elliot K. Fishman1 Beth M. Deutch1 David S. Hartman2 Stanford M. Goldman1,3 Elias A. Zerhouni1 Stanley S. Siegelman1
Clinical histories and CT findings were reviewed in 38 patients with primary adreno- cortical carcinomas. The primary tumors exhibited central areas of low attenuation representing tumor necrosis (n = 26), irregular contrast enhancement (n = 16), detect- able calcification (n = 9), and a thin, capsulelike rim surrounding the tumor (n = 7). Tumors metastasized to liver (n = 9), lung (n = 5), and lymph nodes (n = 5). In eight of nine cases of liver metastasis the primary tumor arose in the left adrenal gland. Evidence of endocrinopathy was present in each of nine patients with lesions 6 cm or less in diameter, but in only two of seven adults with lesions exceeding 15 cm in diameter. We conclude that, contrary to established concepts, adrenocortical carcinoma may present as a smooth, homogeneous, functioning mass 6 cm or less in diameter on CT.
The size of an adrenal mass as an indicator of malignancy or benignancy has been the subject of controversy [1, 2]. It is generally held that tumors in the range of 10-15 cm in diameter, in asymptomatic patients, and associated with hormone production, regardless of size, are likely to be malignant and are properly managed by excision [3]. On the other hand, some studies have suggested that adrenal masses under 5 cm in diameter that are not associated with symptoms should be treated conservatively [1, 2]. Still others have expressed concern about the malignant potential of any adrenal mass. This study assesses the relationship between tumor size, hormonal activity, and CT findings in 38 patients with primary adrenal cortical carcinoma.
Received July 14, 1986; accepted after revision September 26, 1986.
The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the view of the Department of the Navy or the Department of De- fense.
Presented at the annual meeting of the American Roentgen Ray Society, Washington, DC, April 1986.
1 Department of Radiology, Johns Hopkins Uni- versity School of Medicine, 600 N. Wolfe St., Balti- more, MD 21205. Address reprint requests to E. K. Fishman.
2 Department of Radiologic Pathology, Armed Forces Institute of Pathology, Washington, DC 20030, and Department of Radiology, Uniformed Services University of the Health Sciences, Be- thesda, MA 20814.
3 Department of Radiology, Francis Scott Key Medical Center, Baltimore, MD 21205.
AJR 148:531-535, March 1987 0361-803X/87/1483-0531 American Roentgen Ray Society
Materials and Methods
A computer search of the files of the Johns Hopkins Oncology Center and the Armed Forces Institute of Pathology (AFIP) identified all cases of primary adrenal cortical carcinoma accessioned between January 1978 and September 1985. Nine cases from The Johns Hopkins Hospital and 29 cases from the AFIP were found in which a CT scan had been obtained at the time of initial clinical presentation. All CT scans were reviewed to determine lesion size, appearance both before and after contrast enhancement, presence of calcification, extent of metastases, and presence of a “capsulelike rim.” Available plain films, sonograms, and nuclear medicine studies were also reviewed in all patients to determine the extent of metastatic disease. The patients’ charts and/or clinical records were reviewed to determine the clinical presentation in each case.
Patients scanned at Johns Hopkins Hospital were examined on either a Pfizer/AS&E 0500 or a Siemens Somatom DR-3. Scanning parameters were 10 sec, 20 mAs, 125 kVp, and 5- or 10-mm collimation; or 3 sec, 230 mAs, 125 kVp, and 4- or 8-mm collimation, respectively. Patients were examined after administration of 720 ml of oral Hypaque (3% solution) over a 90-min period before the study. All patients were examined both before and after a bolus administration of 50-100 ml of 60% Hypaque. Scans were obtained at 1- to 1.5-cm intervals from the diaphragm to the symphysis with additional scans (4-mm intervals) through the adrenal glands.
In those cases with material submitted to the AFIP, various scanners were used with techniques differing from institution to institution. In all cases, tissue proof of diagnosis was
obtained. All images were reviewed by three of the authors and were correlated with the available pathologic specimens.
Results
The patients’ ages ranged from 8 months to 79 years (mean, 39 years). Seventeen of the patients were men and 21 were women. Thirty-four were white, three were black, and one was Hispanic.
The tumor originated in the left adrenal gland in 21 cases and in the right adrenal gland in 17; no tumors were bilateral. Tumor size varied from 3 to 25 cm and all tumors were readily visible on CT (Table 1). In the group of seven patients under 15 years of age, the smallest tumor was 3.5 cm in diameter (Fig. 1) and the largest measured 20 cm. The average tumor diameter in this population was 9 cm. Of the 31 patients 15 years of age or older, the smallest lesion was 3 cm in diameter (Fig. 2) and the largest was 25 cm. The mean tumor diameter in this age group was 11.5 cm.
Unenhanced CT studies revealed an inhomogeneous but well-defined oblong, spherical, or lobulated mass in the area of the adrenal gland. Posterior compression of the ipsilateral kidney was common, as was anterior displacement of the stomach and pancreas in those left-sided lesions in which the diameter of the mass exceeded 8 cm. Areas of low density, proven by pathologic correlation to be tumor necrosis, were recognized in 26 cases and were usually centrally located (Fig. 3). The CT studies of the six lesions 5 cm or less in diameter demonstrated homogeneous masses of attenuation similar to muscle without evidence of central or peripheral necrosis. Tumors greater than 6 cm in diameter invariably
| Group: Characteristic | No. of Tumors by Size (in cm) | |||
|---|---|---|---|---|
| 0-6 | >6-12 | >12-14 | ≥15 | |
| Adults: | ||||
| Total | 5 | 12 | 6 | 7 |
| Hormonally active | 5 | 6 | 3 | 2 |
| Cushing's syndrome | 3 | 6 | 2 | 0 |
| Mixed | 0 | 0 | 0 | 0 |
| Evidence of metastasis | ||||
| (location): | ||||
| Liver | 2 | 2 | 2 | 2 |
| Lung | 0 | 4 | 0 | 1 |
| Nodes | 1 | 1 | 1 | 1 |
| Children: | ||||
| Total | 4 | 1 | 0 | 2 |
| Hormonally active | 4 | 1 | 0 | 2 |
| Cushing's syndrome | 1 | 0 | 0 | 0 |
| Adrenogenital syndrome | 2 | 0 | 0 | 0 |
| Precocious puberty | 1 | 1 | 0 | 1 |
| Mixed | 1 | 0 | 0 | 0 |
| Evidence of metastasis | ||||
| (location): | ||||
| Liver | 0 | 1 | 0 | 0 |
| Lung | 0 | 0 | 0 | 0 |
| Nodes | 0 | 0 | 0 | 1 |
Note .- Adults were older than 15 years of age; children were 15 years of age or younger.
exhibited some degree of inhomogeneity or central necrosis. Tumor calcifications, including microcalcifications (n = 6) and dense accretions of calcium (n = 3), were present in nine cases (Fig. 3). In two cases the contralateral adrenal glands appeared atrophic. In both of these cases the primary tumor was hormonally active. The other contralateral glands were normal in size, shape, and configuration.
After IV contrast enhancement via either a bolus or infusion technique, inhomogeneous enhancement of the tumor was noted with an increase in attenuation of the more peripheral portions of the tumor and relatively little enhancement in the central low-density portion. In seven cases with contrast enhancement a thin, capsulelike rim was demonstrated around the tumor (Fig. 4). These capsules were continuous, and distinct from both high-density tumor tissue and sur- rounding organs. They were seen in both small and large tumors. There was minimal central enhancement of those tumors that appeared to be homogeneous on unenhanced studies.
At the time of initial examination, tumor dissemination was seen in 21 cases, nine of which had hepatic metastases (Figs. 1 and 3). Of these nine cases, eight originated from left adrenal tumors. One right adrenal tumor directly invaded the liver. Five tumors had metastasized to the lungs and five to the lymph nodes: paraaortic (four), pararenal (one), and subcrural (one). Two tumors directly invaded the ipsilateral renal pelvis, and one left adrenal tumor was metastatic to the contralateral right renal pelvis. Three other tumors extended to adjacent Gerota’s fascia and the paraadrenal soft tissues. A thrombus in the inferior vena cava was visible in one case and was confirmed at surgery and by pathology studies to represent direct tumor extension. Three other cases were found to have left renal vein involvement; however, only one of these was shown by CT.
Endocrine dysfunction was present in 21 patients and included Cushing’s syndrome (12); precocious puberty
A
B
in liver.
Fig. 3 .- 23-year-old woman with history of generalized weakness, hy- pertension, and cushingoid features. Left adrenal carcinoma measures 13 cm. Note dense calcification within mass (long arrows). Liver metastases also seen (short arrow).
(three); adrenogenital syndrome (two); mixed syndrome (one); and/or the complaints of weight gain, hirsutism, amenorrhea, or recent onset of hypertension in a young patient. In a number of patients, several of these clinical symptoms were present simultaneously. Sixteen patients presented with pain and an abdominal or flank mass. Three patients complained of long-standing fatigue and malaise, while one patient pre- sented with a fever of unknown origin.
In several of the cases with symptoms caused by hormone production of the tumor, the clinical symptoms preceded the diagnosis by several years. In one case, the patient presented with mediastinal widening on routine chest studies, at which time excessive fat in the mediastinum was seen on a chest CT scan. Two years later, when he reappeared with fatigue
and malaise, the patient was found to have a 4-cm adrenal cancer and hepatic metastases.
The study included a group of small adrenal neoplasms: Nine lesions were 6 cm or less in diameter; six of these were 5 cm or less in diameter. All were well defined, sharply outlined, and nonadherent to adjacent tissues. Seven were homogeneous; two (5-6 cm in diameter) had central areas of lower attenuation. Two of these tumors had metastasized to the liver; in a third case, there was evidence of enlarged paraaortic and subcrural lymph nodes on CT.
There was an inverse relationship between tumor size and endocrine function (Table 1). All nine patients with lesions 6 cm or less in diameter had symptoms of hormone overpro- duction, whereas only four of nine patients with lesions 15 cm or greater had endocrinopathy.
Discussion
While primary adrenocortical carcinoma is a very rare can- cer [4], the incidence of benign, nonfunctioning adrenal ade- noma is estimated at 2-8% of the general population [5, 6]. As such, an adrenal lesion imaged by CT is much more likely to be a benign adenoma or an adrenal metastasis than to be a primary carcinoma. Nonetheless, the distinction is an im- portant one since adrenocortical carcinoma has a median survival of only 6 months if left untreated, but is potentially curable by surgery if detected early [7].
Several adrenal lesions have CT findings that permit an accurate pathologic diagnosis. Cysts and myelolipomas are discernible by their respective water and fat attenuation val- ues on CT [8-11]. When adrenal hemorrhage is suspected clinically by an appropriate history of trauma, meningococ- cemia, anticoagulation, or other bleeding diatheses, it can be confirmed by CT studies showing increased attenuation [12]. Adrenal hyperplasia classically leads to glandular en-
A
B
largement with maintenance of the Y-shaped configuration of the adrenal glands and occasional minimal diffuse nodularity [11, 13]. The consideration of metastasis becomes para- mount when bilateral, homogeneous adrenal lesions are dis- cernible, when there is evidence of other metastases, or when the patient is known to have a primary neoplasm elsewhere.
Distinguishing the very common benign adrenal adenomas from adrenocortical carcinoma may be difficult, even when tissue is available for examination. Diagnosis in these cases has rested largely on the pattern of growth, behavior, and size of the lesion, that is, the larger the lesion, the more need for concern. Mitnick et al. [2] monitored a series of patients for 1 year in whom small, nonfunctional adrenal lesions were found incidentally, and recommended CT criteria for estab- lishing benignancy of adrenal lesions: Tumors with a smooth contour, a well-defined margin, no growth on serial CT scans over 1 year, and a diameter less than 5 cm need not be followed further for possible malignancy. Previous studies of primary adrenocortical carcinoma lend support to these cri- teria, in that all have reported the carcinomas to be large at presentation [1, 3, 10]. Dunnick et al. [3] reported eight cases in 1982: The smallest tumor in the study measured 9 cm and the median size was 12 cm. Adrenocortical carcinomas 20 cm or larger have often been reported, whereas there has been only a single case report of a carcinoma less than 3 cm.
In addition to size, several CT characteristics have been proposed as typical of primary adrenocortical carcinoma. Dunnick et al. [3] reported central diminished attenuation, representing central tumor necrosis, and irregular contrast enhancement in all eight of their cases; three had tumor calcifications. Similarly, we identified central necrosis in 26 cases and tumor calcification in nine. As Dunnick et al. [3] noted, these findings have also been seen in pheochromocy- tomas and large metastases [3]. In addition, Mitnick et al. [2] reported centrally diminished attenuation in three benign, nonfunctioning adrenal adenomas smaller than 5 cm. Review of the pathology reports in two of these cases revealed
hemorrhagic necrosis in one and hyalinization in the other. Furthermore, two adenomas contained calcifications seen on CT. Thus, calcification and decreased central attenuation can be identified in both benign and malignant processes and are of little help in making the differential diagnosis.
A new finding identified in seven of our cases was a thin, enhancing, capsulelike rim surrounding the neoplasms (Fig. 4). This may add to the specificity of CT diagnosis. All seven “capsules” were continuous, smooth, and nonadherent to adjacent structures. The capsule was present in both small and large tumors of 3.5-15 cm, suggesting that it is not simply a rim of adjacent tissue compressed by a large, ex- panding tumor mass. Furthermore, all seven capsules en- hanced with IV contrast material, suggesting that they rep- resent well-vascularized portions of the tumor. There have been no reports in the literature of this finding in either benign adrenal adenomas or metastases. In several cases that had the capsule, no normal compressed adrenal gland was pres- ent on pathologic examination of the specimen. Rather, the enhancing rim was part of the tumor. The cause of the enhancement is not known.
Both small and large tumors showed a predilection for metastasizing to liver; furthermore, of our nine cases with hepatic metastases, eight had originated from left-sided tu- mors. The only right-sided tumor with hepatic metastases abutted on the liver and most likely spread by direct contiguity to hepatic portal vessels. From this location nests of tumor cells presumably “embolized” to more distant liver sites. One could speculate that the liver metastases of purely left-sided origin arrived via adrenal-splenic collateral circulation to the portal vein. Adrenal-splenic circulation may become particu- larly significant in the setting of tumor thrombosis in the renal vein, and has been described in renal cell carcinoma [14].
Distinguishing malignant from benign adrenal lesions is difficult because no single criterion is specific. Central necrosis and tumor calcification are seen in a variety of lesions, includ- ing small benign adenomas. Adrenocortical carcinomas may
be functional or silent. They may be large, or, as in our study, less than 5 cm in diameter, a size previously believed to be safe for nonsurgical follow-up. They may have irregular mar- gins and shapes but may also be smooth, well-defined spheres, especially when small. The only definite criterion for malignancy is the presence of metastases. Certainly a 20-cm irregular, centrally necrotic, partially calcified adrenal mass merits adrenal suspicion of malignancy, just as a 3-cm, non- functional, homogeneous, well-circumscribed adrenal lesion will most likely be benign. Earlier diagnosis might be achieved by closer examination of adrenal glands in patients with even vague endocrine complaints. When a small adrenal lesion is found in this setting, suspicion of malignancy should warrant prompt removal of the lesion. Both unenhanced and IV con- trast enhancement studies should be performed to better delineate central necrosis, tumor capsule, venous extension, and hepatic metastases.
REFERENCES
1. Glazer HS, Weyman PJ, Sagel SS, Levitt RG, McClennan BL. Nonfunc- tioning adrenal masses: incidental discovery on CT. AJR 1982;139:81-85
2. Mitnick JS, Bosniak MA, Megibow AJ, Naidich DP. Non-functioning adrenal
adenomas discovered incidentally on CT. Radiology 1983;148:495-499
3. Dunnick NR, Heaston D, Halvorsen R, Moore AV, Korobkin M. CT ap- pearance of adrenal cortical carcinoma. J Comput Assist Tomogr 1982;6(5): 978-982
4. King DR, Lack EE. Adrenal cortical carcinoma: a clinical pathologic study of 49 cases. Cancer 1979;44:239-244
5. Sommers SC. Adrenal glands. In: Anderson WAD, Kissane JM, eds. Pathology, vol 27. St Louis: Mosby, 1977:1658-1679
6. Commons RR, Callaway CP. Adenomas of the adrenal cortex. Arch Intern Med 1948;81:37-41
7. Daneman A, Chan HS, Martin J. Adrenal carcinoma and adenoma in children: a review of 17 patients. Pediatr Radio/ 1983;13(1):11-18
8. Behan M, Martin EC, Moeche EC, Kazam E. Myelolipoma of the adrenal: 2. Cases with ultrasound and CT findings. AJR 1977;129:993-996
9. Kleinhaus U, Ensel A, Adler O. Computed tomography in the diagnosis of adrenal pathology. Isr J Med Sci 1983;19(6):524-529
10. Hattery RR, Sheedy PF II, Stephens DH, van Heerden JA. Computed tomography of the adrenal gland. Semin Roentgeno/ 1981;16(4):290-300
11. Hauser H, Battikha JG, Wettstein P. Pathology of the adrenal glands. Common and uncommon findings in computed tomography. Eur J Radiol 1981;1(3):215-226
12. Eghari M, McLoughlin MJ, Rosen IE, et al. The role of computed tomog- raphy of the adrenal glands. J Comput Assist Tomogr 1980;4:71-77
13. Harrison JH, Mahoney EM, Bennet AH. Tumors of the adrenal cortex. Cancer 1973;32:1227-1235
14. Saitoh H. Distant metastasis of renal adenocarcinoma in patients with tumor thrombus in the renal vein and/or vena cava. J Urol 1982;127:652- 653
Downloaded from www.ajronline.org by 128.120.194.195 on 02/11/15 from IP address 128.120.194.195. Copyright ARRS. For personal use only; all rights reserved