ADRENOCORTICAL NEOPLASMS Prognosis and Morphology
CHIK KWUN TANG, M.D. GEORGE F. GRAY, M.D.
From the Department of Pathology, The New York Hospital- Cornell Medical Center, New York, New York
ABSTRACT - The clinical data and morphologic findings in 16 cases of adrenocortical carcinoma were compared with those in 11 cases of surgically removed functional adenomas and 12 cases of nonfunctional adenomas found at autopsy. Histopathologic changes of architectural disarray, pleomorphism, increased mitotic activity, vascular invasion, hemorrhage, or necrosis were generally reliable criteria for diagnosis of malignancy. However, weight was the parameter that most consistently correlated with outcome, since all patients with tumors under 50 Gm. survived and all lesions of 95 Gm. or over proved to be malignant.
Differentiating an adrenocortical adenoma from its malignant counterpart may be difficult because changes such as increased numbers of mitotic figures, altered architecture of cells, pleomorphic nuclei, hemorrhage, and necrosis, which are usually helpful for diagnosis of malignant tumors, have been regarded as less reliable criteria of malignancy in neoplasms of the adrenal gland. 1-3 We have been concerned about the problem of diagnosis of 2 groups of adrenocortical tumors: small encapsulated lesions exhibiting marked pleomorphism and mitotic activity, and very large but histologically innocuous-appearing tumors. To study the value of morphologic findings in estimation of prognosis of adrenocortical tumors, we have reviewed the histopathologic material, clinical findings, and outcome of patients with adrenocortical tumors seen at The New York Hospital-Cornell Medical Center between 1948 and 1971.
Clinical Material
This study includes 39 adrenocortical tumors. Only those cases with histologic slides and clinical records available for review and at least a three- year follow-up are included. The diagnoses were established on surgical specimens in 22 instances and by necropsy in 17 cases (Table I).
Thirty-four of the patients were adults and 5 were children (under fifteen years of age). Car- cinomas appeared among all age groups but were relatively more frequent among younger pa- tients. Functioning adenomas removed at surgery were found among all age groups up to
| Diagnosis | Surgical | Autopsy | Total |
|---|---|---|---|
| Adenoma | 11 | 12 | 23 |
| Carcinoma | 11 | 5 | 16 |
| Distribution | Adenoma | Carcinoma |
|---|---|---|
| AGE (years) | ||
| Under 16 | 2 | 3 |
| 16 to 39 | 4 | 4 |
| 40 to 49 | 4 | 3 |
| 50 to 59 | 5 | 3 |
| Over 59 | 8 | 3 |
| SEX | ||
| Male | 8 | 3 |
| Female | 15 | 13 |
| Syndrome | Adenoma | Carcinoma | ||
|---|---|---|---|---|
| Child | Adult | Child | Adult | |
| Cushing | 0 | 6 | 2 | 7 |
| Aldosteronism | 0 | 3 | 0 | 0 |
| Adrenogenital | 2 | 0 | 1 | 0 |
| Virilism | 0 | 0 | 0 | 3 |
| Feminization | 0 | 0 | 0 | 0 |
| Nonfunctioning | 0 | 12* | 0 | 3 |
*“Incidental” autopsy findings.
age sixty, but nonfunctioning adenomas discov- ered at autopsy were found mostly in patients above age sixty and not seen in patients less than forty years old (Table II). Symptoms due to the neoplasm were observed in all of the tumors diagnosed antemortem and in the 5 carcinomas diagnosed at autopsy (Table III). The 12 adenomas found at autopsy were assumed to be nonfunctional because none of the patients had an obvious clinical syndrome of adrenocortical hyperactivity. However, hormonal assays which would unequivocally exclude a hyperfunctioning state had not been done in any of these patients antemortem.
Symptoms were due to hyperfunction in 24 of 27 surgical cases. The syndromes caused by the tumors are summarized in Table III.
Nine of the 16 patients with carcinoma had a Cushing syndrome and in these, the twenty- four-hour urinary 17 ketosteroids ranged from 3 to 107.5 mg. and the twenty-four-hour urinary 17 hydroxycorticosteroids from 8 to 348 mg. The adenomas associated with Cushing’s syndrome had twenty-four-hour urinary 17 ketosteroid val- ues of 2 to 30 mg. and twenty-four-hour urinary 17 hydroxycorticosteroid of 7.5 to 39 mg. Elevated urinary aldosterone levels were demonstrated in 3 hypertensive patients with adenomas.
All of the patients with functioning tumors diagnosed as adenomas were alive and without
clinical evidence of an adrenocortical neoplasm for periods of four to seventeen years after excision of the tumor. Twelve of the patients with carcinoma died of the malignant neoplasm. One patient died three years after diagnosis of causes unrelated to the adrenal carcinoma, but residual tumor was found at autopsy. Three patients with tumors interpreted as carcinoma on the basis of histologic findings survived and had no further evidence of tumor for periods of two to eleven years after surgical removal of the neoplasm.
Pathologic Findings
The carcinomas were strikingly larger than their benign counterparts. The sizes of the surgi- cally removed adenomas and those incidentally discovered at autopsy were similar. The weights of the fatal carcinomas did not overlap the weights of the adenomas, but the dimensions of the 3 nonfatal carcinomas did correspond to the adenomas (Table IV).
The cut surface of the adenomas displayed uniform color, usually bright yellow, but less commonly reddish brown or brown. Zones of necrosis, hemorrhage, or grossly obvious cal- cification were apparent in none of the adenomas found at autopsy, and these changes were seen in only one of the surgically excised adenomas. In the 13 carcinomas that led to a fatal outcome, 6 had obvious local invasion, 4 were focally necrot- ic, and 7 were hemorrhagic. The color of the cut surfaces of the carcinomas tended to be varie- gated, including yellow-brown, white, or gray areas. All 3 tumors classified as carcinomas but associated with long-term survival were cir- cumscribed, and only 1 tumor had grossly obvious zones of necrosis and calcification.
Microscopically, the contrast between adenomas and carcinomas was striking. All of the adenomas found at autopsy had well-preserved architecture in cords and nests closely resembling normal adrenal cortex (Fig. 1A). The absence of atrophy of the adjacent cortex in these cases may
| Adenoma | Carcinoma | |||
|---|---|---|---|---|
| Incidental (12 Cases) | Surgical (11 Cases) | Fatal (13 Cases) | Survivors (3 Cases) | |
| Size | 1 to 5 | 0.8 to 6 | 7 to 28 | 0.7 to 4 |
| (Centimeters) (mean 2.4) | (mean 3.3) | (mean 13) | (mean 2.7) | |
| Weight | 12 to 35* | 7 to 40 | 96 to 2,460 | 2 to 30 |
| (Grams) | (mean 18.5) | (mean 705) | (mean 15) | |
*Only 3 adenomas recorded separately from total adrenal weight.
A
B
G
B
D
E
F
be a further indication that the tumors were not producing excessive glycocorticoids, since corti- cal atrophy was found in the functional tumors associated with Cushing’s syndrome. In the 12
tumors discovered incidentally no mitotic activity was seen in 11 cases, and the mitotic activity was 5 figures in 20 HPF (high-power field) in the twelfth case. None of these tumors had capsular or
vascular invasion, pleomorphism or bizarre giant cells, or any microscopic foci of necrosis, hemorrhage, or calcification. The 11 adenomas which were surgically excised were all well dif- ferentiated architecturally with the exception of 1 tumor which had an exaggerated trabecular pattern. One to 5 mitotic figures per 20 HPF were counted in 4 instances and no mitoses were seen in 7. Capsular and vascular invasion were not seen in any instance. Microscopic foci of necrosis, hemorrhage, or calcification were seen in only 1 case. Multinucleated cells and nuclear pleo- morphism of moderate degree were seen focal- ly in 3 tumors but not in the other 8 (Fig. 1B).
All but 1 of the 13 ultimately fatal carcinomas had definite architectural disarray. These pat- terns included exaggerated trabecular growth, random sheet-like growth, and marked increased vascularity (Fig. 1C). Only in 1 carcinoma did the over-all pattern of cellular growth closely simu- late normal adrenal. Mitotic activity was not observed in that lesion. In the other 12 car- cinomas the mitotic count was 1 to 5 per 20 HPF in 6, 6 to 10 per 20 HPF in 4, 11 to 16 per 20 HPF in 1, and more than 16 per HPF in 1 (Fig. 1D). In the slides available for review, definite capsu- lar invasion was observed in 4 cases and vascular invasion in 8. Since these tumors were quite large, it is possible that capsular invasion might have been demonstrated more often if more sections had been available for review. Moderate to marked pleomorphism was apparent in 11 instances, microscopic foci of necrosis in 11, hemorrhage in 12, calcification in 4 (Fig. 1E).
However, the only findings common to all of the fatal tumors was large size. All of these were greater than 7 cm. in largest diameter and weighed more than 96 Gm. One large histologi- cally well-differentiated tumor had no demon- strable metastases or invasion at the time of sur- gery but was subsequently shown to be malignant on the basis of radiographic evidence of pulmo- nary and bony metastases which were success- fully treated by radiation, and residual tumor in the retroperitoneum found at autopsy.
In 3 cases a diagnosis of carcinoma was made on the basis of histologic appearance, but the pa- tients appear to have been treated successfully by surgical excision. In each of these cases the lesions were characterized by definite architectural dis- array, marked cellular pleomorphism with numerous bizarre nuclei and giant cells, abun- dant mitotic figures, and invasion of vessels within the tumor, findings indistinguishable from those characterizing 12 of 13 fatal tumors (Fig.
1F). Definite capsular invasion was not seen in any of these 3 lesions and all were small. Two of the patients were children, aged eighteen months and three years, and the third was an adult thirty years of age.
Comment
Generally, diagnosis of benignancy or malig- nancy in an adrenal tumor can be made on the basis of the histopathologic findings. However, the benign histologic appearance of 1 of 16 carcinomas and the favorable clinical outcome of 3 of 16 carcinomas suggest that other parameters are essential in estimation of prognosis. Hein- becker, O’Neil, and Ackerman2 on the basis of a review of 16 adrenal tumors regarded frequent mitoses, venous invasion, and capsular invasion as certain criteria of malignancy, and necrosis, hemorrhage, calcification, pleomorphism, and nuclear atypia as helpful accessory findings. They had 2 long-term survivors who had encapsu- lated tumors with giant nuclei and pleomorphism but without venous invasion. Schteingart et al.3 on the basis of a study of 12 tumors producing Cushing’s syndrome believed that accurate as- sessment of malignancy could be made by micro- scopic examination of representative tissue on the basis of capsular invasion, and that increased and abnormal mitotic activity and severe nu- clear pleomorphism were also reliable criteria.
The tumors in patients with a benign course were less than 4 cm. in diameter and the malig- nant tumors were over 7 cm. in diameter. Lipsett, Hertz, and Ross4 in an analysis of 38 carcinomas found the size to be over 500 Gm. in 20 cases and emphasized that weight and steroid excretion were of more value than histologic appearance in differentiation of adenomas from carcinomas. They suggested that nuclear variations and mitot- ic figures were prominent in both benign and malignant adrenal tumors. Similarly, Macfarlane1 stated that giant and bizarre nuclei, numerous mitoses and marked pleomorphism do not con- clusively indicate a malignant tumor. On the other hand, Huvos et al.5 found microscopic characteristics of malignancy such as cellular pleomorphism, nuclear hyperchromasia, and prominent mitotic figures in all of their 34 adrenal carcinomas.
The morphologic feature that we found most consistently in tumors with favorable outcome was small size. Although very large adenomas have been reported, no benign tumor exceeded 44 Gm. in our series.6 Likewise, the largest recorded weight of a benign tumor in the series of
adenomas in Schteingart et al.3 was 27 Gm. This suggests that benign tumors infrequently obtain a size of more than 50 mg. Furthermore, 12 of 13 patients with tumors over 95 Gm. died of neoplas- tic disease. The thirteenth patient died of other causes but had residual tumor at autopsy.
The tumors regarded as malignant by his- topathologic criteria which had a good clinical outcome were all small, the largest weighing only 30 Gm. These tumors were undoubtedly discov- ered at a curable stage because they were functional, and the symptoms of hyperactivity led to diagnosis and treatment before extra-adrenal invasion or metastasis developed. The demon- stration of malignant-appearing small tumors suggests that most adrenocortical carcinomas arise as such and not from transformation of pre- existing adenomas.
Our findings suggest that in adrenocortical neoplasms, architectural disarray, necrosis and hemorrhage, bizarre nuclei, particularly abnor- mal mitotic figures, are as important in diagnosis of malignancy as in other organs. Capsular inva- sion or metastases were evident at the time of diagnosis only in a minority of the ultimately fatal adrenocortical tumors. Invasion of vessels within the tumor considered to be a certain sign of malignancy by some authors2 was more often seen than capsular invasion in our series but was not found in 5 of the 16 carcinomas. Since vascular invasion was never identified in any of the adenomas, but was seen in the carcinomas of patients who were cured, this finding may be useful for diagnosis of carcinoma but appears to have little bearing on development of metas- tases. Mitotic figures are indicative of carcinoma only when they are bizarre or present in large numbers.
The relationship of functional activity to his- tologic appearance was in line with previous observations. 7 Extremely high levels of urinary 17 ketosteroids, 17 hydroxycorticosteroids or of plasma cortisol and negative suppression tests were indicative of carcinoma. 8,9 Functioning car- cinomas had the highest levels of 17 ketosteroids and 17 hydroxycorticosteroid excretion and of plasma cortisol, but the functional activity of some of the carcinomas overlapped that of some of the adenomas. In the tumors studied there appeared to be virtually no relationship between degree of anaplasia and functional activity. The nonfunc- tional carcinomas were among those attaining the largest size, an expected finding, since in the absence of an adrenocortical syndrome, a tumor growing in the upper retroperitoneum would be
ideally situated to attain great bulk before produc- ing local symptoms.10 All 3 of the aldosterone- producing tumors were benign, a finding in line with the experience reported by others, 11 al- though a few malignant aldosterone-producing lesions have been reported. 12
We believe that the diagnosis of adenoma or carcinoma can generally be made with accuracy if all of the clinical and pathologic data are con- sidered. No single histopathologic finding will consistently differentiate adrenal carcinoma from adenoma. However, in most instances of hor- monal assays, size and gross extent of the lesion and microscopic appearance permits correct es- timation of prognosis. Small tumors, even though malignant appearing microscopically, are likely to have a good prognosis if still confined to the adrenal gland. This appears to apply especially in tumors in children and young adults. Large tumors over 100 Gm. should have a guarded prognosis even though the microscopic pattern appears benign.
525 East 68th Street New York, New York 10021 (DR. GRAY)
References
1. MACFARLANE, D. A .: Cancer of the adrenal cortex, Ann. R. Coll. Surg. Engl. 23: 155 (1958).
2. HEINBECKER, P., O’NEAL, L. W., and ACKERMAN, L. V .: Functioning and non-functioning adrenal cortical tumors, Surg. Gynecol. Obstet. 105: 21 (1957).
3. SCHTEINGART, D. E., OBERMAN, H. A., FRIEDMAN, B. A., and CONN, J. W .: Adrenal cortical neoplasms producing Cushing’s syndrome, Cancer 22: 1005 (1968).
4. LIPSETT, M. B., HERTZ, R., and Ross, G. T .: Clinical and pathophysiologic aspects of adrenocortical carcino- ma, Am. J. Med. 35: 374 (1963).
5. Huvos, A. G., HAJDU, S. I., BRASFIELD, R. D., and FOOTE, F. W., JR .: Adrenal cortical carcinoma, Cancer 25: 354 (1970).
6. KARSNER, H. T .: Tumors of the adrenal cortex. Atlas of Tumor Pathology, Washington, D. C., Armed Forces Institute of Pathology, 1950, section 8, part 29, p. 16.
7. BIRKE, G., et al .: Adrenal cortical tumors, Acta Chir. Scand. 117: 233 (1959).
8. SCOTT, H. W., JR., et al .: Surgical management of adrenocortical tumors with Cushing’s syndrome, Ann. Surg. 173: 892 (1971).
9. ORTH, D. N., and LIDDLE, G. W .: Results of treat- ment in 108 patients with Cushing’s syndrome, N. Engl. J. Med. 285: 243 (1971).
10. KNIGHT, C. D., TRICHEL, B. E., and MATTHEWS, W. R .: Non-functioning carcinoma of the adrenal cortex, Ann. Surg. 151: 349 (1959).
11. NEVILLE, A. M., and SYMINGTON, T .: Pathology of primary aldosteronism, Cancer 19: 1854 (1966).
12. ALTERMAN, S. L., DOMINGUEZ, C., LOPEZ-GOMEZ, A., and LIEBER, A. L .: Primary adrenocortical car- cinoma causing aldosteronism, ibid. 24: 602 (1969).