THE PATHOLOGY OF ADRENOCORTICAL NEOPLASIA: A CORRELATED STRUCTURAL AND FUNCTIONAL APPROACH TO THE DIAGNOSIS OF MALIGNANT DISEASE
Michael J. O’Hare, M.A., Ph.D.,* Paul Monaghan, B.Sc., Ph.D., t and A. Munro Neville, M.D., Ph.D., M.R.C. Path.#
Abstract
The structural and functional characteristics of nine functioning adreno- cortical tumors (four adenomas and five carcinomas) causing Cushing’s syn- drome or virilization were studied. All tumors that we considered to show histologic evidence of malignant disease and that subsequently,metastasized or recurred also showed in cell culture at least one significant functional or be- havioral difference from benign tumors. No single defect was common to all carcinomas, but predominant changes included secretion of precursor steroids, such as Il-deoxycortisol (S) and a blunted or absent response to ACTH. All adenomas examined were normal in these respects in comparison with nondiseased cortical cells in culture. In carcinomas whose functions deviated only minimally from normal the presence of highly differentiated ultrastructural characteristics did not, however, confer a better prognosis.
Integrated research during the past 20 years by clinical endocrinologists, bio- chemists, and histopathologists has result- ed in major advances in our understand- ing of the human adrenal cortex and its
diseases. Appreciation of functional zona- tion in the gland, its pathways of steroi- dogenesis, and the trophic stimuli by which its functions are regulated have to- gether enabled the various appearance’s
*Scientific Staff Member, Ludwig Institute for Cancer Research, Royal Marsden Hospital, Sutton, Surrey, England.
Scientific Staff Member, Ludwig Institute for Cancer Research, Royal Marsden Hospital, Sutton, Surrey, England.
Professor of Pathology, Ludwig Institute for Cancer Research, Sutton, Surrey, England.
of bilateral hyperplasia in Cushing’s syn- drome, hyperaldosteronism with low plasma renin levels (Conn’s syndrome), and the adrenogenital syndrome to be predicted and explained.1-4 Nodular hy- perplasia is now also recognized as a clin- icopathologic entity. Although debate re- mains as to its etiology, its distinction from single or multiple adrenocortical adenomas may be readily made, thereby contributing significantly to improved pa- tient care.5
Tissue culture methods have in re- cent years added a further dimension to endocrine research.6 Structural and func- tional studies have shown that monolayer cell culture techniques can be readily ap- plied to adult adrenocortical cells. Cul- tures prepared from different zones have illustrated their functional interrelation- ships and suggested mechanisms whereby zonation may be controlled.7-9 These con- cepts, in turn, have contributed to an un- derstanding of the protean histologic pat- terns of adrenocortical tumors causing Conn’s syndrome. 1, 6
Nonetheless many problems of adre- nal pathology remain to be solved, not the least of which is the definition of reli- able pathologic criteria to assess the ma- lignant potential of adrenocortical
tumors. At present, despite several valu- able guidelines, the only unequivocal cri- terion is the demonstration of distant me- tastases.3
This particular problem has been highlighted by some of the consultative material that we received during a three year period from patients with hypercor- ticalism in the form of Cushing’s syn- drome or virilization. We propose there- fore to devote this article to the topic of the relative value of structural and func- tional indices as aids to the diagnosis of potential malignant disease, using in ad- dition to conventional histologic study both ultrastructural observations and mono- layer tumor cell cultures.
PATIENTS
A résumé of data pertaining to nine patients with adrenocortical tumors caus- ing hypercorticalism, which were studied both histologically and in culture, is given in Table 1. Preoperatively the clinical diagnosis of a cortical tumor was con- firmed by a wide variety of functional tests. All patients were considered to have tumors localized to the adrenal gland at the time of operation.
| Patient Number | Sex | Age (Years) | Presenting Features | Tumor Weight (Grams) |
|---|---|---|---|---|
| 1 | F | 26 | Cushing's syndrome | 15 |
| 2 | F | 48 | Cushing's syndrome (mild) plus virilization (severe) | 92 |
| 3 | F | 55 | Cushing's syndrome with severe hypertension | 125 |
| 4 | F | 59 | Cushing's syndrome (severe) plus virilization (mild) | 420 |
| 5 | F | 66 | Virilization | 880 |
| 6 | F | 52 | Virilization plus hypertension | 1850 |
| 7 | M | 39 | Cushing's syndrome | 20 |
| 8 | F | 34 | Cushing's syndrome | 15 |
| 9 | F | 31 | Cushing's syndrome | 10 |
METHODS
All tumors were received in the labo- ratory within 18 hours after operation and most were obtained between two and four hours, having been despatched or collected from the operating theatre in a sterile container over ice. On receipt the lesions were examined, freed from ad- herent fat or connective tissue, weighed, and sliced under sterile conditions.
Representative portions were then processed in three ways. Material was placed in 4 per cent neutral buffered for- malin for wax embedding and prepara- tion of routine 5 p histologic sections. Adjacent tissue was diced into 1 cu. mm. fragments and fixed for one hour in 2 per cent glutaraldehyde in 0.05 M phos- phate buffer (pH.7.2 to 7.4), the osmotic pressure of which was adjusted to 350 mosm. per liter with sucrose, post- fixed in 1 per cent phosphate buffered osmium tetroxide, dehydrated in ethanol, and embedded in Epon-Araldite resin. Ultrathin sections after staining with uranyl acetate and lead citrate were ex- amined using a Philips EM 400 electron microscope.
Material for tissue culture was select- ed from regions devoid of fibrous or ob- viously necrotic tissue, and cell sus- pensions were prepared by enzymatic (collagenase-hyaluronidase) disaggrega- tion exactly as described elsewhere for the preparation of monolayer cultures from normal adult human adrenocortical cells.10
Cultures in 25 cm.2 flasks were main- tained at 37°C., and the medium (Dul- becco’s Eagle medium plus 15 per cent [w/v] fetal calf serum) was changed daily and retained for measurement of secret- ed steroids. The number of tumor cells in individual cultures was determined at intervals, when they could be identified unambiguously, by direct counting in randomly chosen fields of known area under an inverted phase contrast micro- scope, to enable an estimate of their ca- pacity for growth in vitro to be made.
All tumors were maintained for at least one month as primary cultures and were tested consecutively with 1 ug. per ml. (100 mU. per ml.) of ACTH 1-24 (Syn- acthen, CIBA) and 0.5 mmole per liter
of monobutyryl cyclic AMP (mbCAMP; Sigma Chemical Co.) after at least ten days’ prior culture in the absence of added hormones. After these dynamic functional tests had been completed, rep- resentative cultures were fixed with phos- phate buffered glutaraldehyde and pro- cessed as already described for electron microscopy but embedded in Epon.
Corticosteroids (cortisol plus cortico- sterone) synthesized by the tumor cultures were measured by acid-fluorescence,7 and the complete spectrum of secreted steroids was examined by high pressure liquid chromatographic and radioactive precursor incorporation methods that have been described in detail elsewhere.11
RESULTS
That there is a real problem in mak- ing a histologic diagnosis of adrenocorti- cal malignant disease was highlighted re- cently by the presentation of material from three of the patients studied here to a panel of 15 senior qualified histo- pathologists. All participants in this con- sultative exercise had had extensive expe- rience in the diagnosis of neoplasia and were aware of the difficulties that endo- crine tumors pose in this respect, al- though none had specifically specialized in this type of lesion. The results are shown in Table 2.
Two facts emerged from this limited survey. First was the importance of com- plete clinical and biochemical data being made available. These were deliberately withheld from the other histopathologists in the present study. It is apparent that some cortical and medullary tumors may not be reliably distinguished solely on morphologic grounds at the light micro- scopic level. Second, difficulty in distin- guishing benign from malignant adreno- cortical tumors is all too apparent.
To clarify the issues involved in dis- criminating between benign and malig- nant adrenocortical tumors, we propose to review the histologic features that we consider characteristic of adenomas and carcinomas on the basis of previous expe- rience, as illustrated by the nine cases presented in this study.1-4 The cellular features, endogenous steroidogenic abil- 130
| Patient Number | Tumor Weight (Grams) | Histologic Assessment (A.M.N.) | Histologic Assessment of 15 Other Pathologists Cortical Tumor Pheochromo- Adenoma Carcinoma Cytoma | Functional Assessment in Vitro* | Outcome (Time from Surgery) | ||
|---|---|---|---|---|---|---|---|
| 1 | 15 | Adenoma | - | Not submitted | - | Adenoma | Well (12 months)t |
| 2 | 92 | Possible carcinoma | 2 | 8 | 5 | Carcinoma | Recurrence plus metastases (6 months) |
| 3 | 125 | Probable carcinoma | 6 | 8 | 1 | Carcinoma | Recurrence plus metastases, dead (8 months) |
| 4 | 420 | Carcinoma | - | Not submitted | - | Carcinoma | Dead (15 months) |
| 5 | 880 | Probable carcinoma | 6 | 7 | 2 | Carcinoma | Recurrence (15 months) |
| 6 | 1850 | Carcinoma | - | Not submitted | - | Carcinoma | Metastases at surgery, dead (3 days) |
*See Table 3. tPatients with three similar adenomas, weighing 20, 15, and. 10 grams causing Cushing’s syndrome and possessing normal structural and functional responses in culture, were well at 21/2 years, 18 months, and 15 months, respectively (cases 7, 8, and 9 in Table 1).
ity, and trophic responses of these tumors in culture will also be described to illustrate the contribution of this dy- namic functional approach to the accu- rate categorization of cortical neoplasms. At all times the responses of the cultured tumor cells were compared with those of an extensive series of cultures of nondis- eased (“normal”) adult human adreno- cortical cells derived from adrenal glands removed for therapeutic reasons from patients with metastatic mammary carci- noma. The behavior of these cells has been described in detail elsewhere.6, 10
Adenocortical Adenomas
The tumor from patient 1 is a typical and classic example of an adrenocortical adenoma causing Cushing’s syndrome. Such tumors are generally although not exclusively associated with a “pure” form of Cushing’s syndrome, i.e., without sig-
nificant virilization (Table 1) and they are usually detected when small (<50 gm.). Such tumors are encapsulated and are composed of two types of cells, clear lipid laden cells similar to those of the normal zona fasciculata, and lipid sparse compact cells typical of the cells of the zona reti- cularis (Fig. 1). Both types form small al- veoli or nests surrounded by a delicate fi- brovascular stroma. Mitoses are rare. The adrenal glands attached to or contralater- al to all these tumors invariably show marked atrophy.2
At the ultrastructural level the aden- oma cells are very similar to normal adrenocortical cells. Characteristic fea- tures such as mitochondria with tubulo- vesicular (fasciculata-like) or tubulolamel- lar (reticularis-like) cristae can be seen. There is an extensive smooth endoplas- mic reticulum in most cells, and the rough endoplasmic reticulum forms ei- ther discrete stacks or characteristic short
strands (Fig. 2). At the boundary between adenoma cells, desmosome-like forma- tions may be seen as well as complex focal interdigitations, usually at the junc- tion between several cells (Fig. 2). Base- ment membrane material is not generally observed between them.
All adenomas that we cultured were morphologically indistinguishable from cultures of normal adult adrenocortical cells (Fig. 3).6,10 The tumor from pa- tient I was no exception in this regard and is presented here as being typical of this group. The cultured cortical tumor cells showed a typical morphologic re- sponse to ACTH1.24. After attaching and spreading in the absence of hormone to form epithelial-like groups of lipid laden cells with a granular cytoplasm (Fig. 3), the addition of ACTH caused the characteristic retraction noted with normal cortical cells. No proliferation of the adenoma cells was observed in prima- ry culture.
Cultures of this adenoma were also examined electron microscopically after several weeks in culture. Some apparent loss of, or diminution in, specific dif- ferentiated structures such as mitochon- drial cristae and smooth endoplasmic re- ticulum was noted (Fig. 4), but similar changes have been seen in cultured normal adrenocortical cells maintained without added hormones. The cul- tured adenoma cells were therefore indistinguishable from the latter at the electron optical as well as the light micro- scopic level. Adrenal features were, how- ever, still evident, allowing the tumor cells to be distinguished from the stromal cells also present in the cultures (fibro- blasts, macrophages, and endothelial cells).
The functional activity of the cul- tured adenoma cells closely resembled that of normal cortical cells.10 Thus high pressure liquid chromatography of se- creted products showed that the 118-
M
R
1H
L
F
0:5p
A
50 ₽
AD
AD
M
F
B
50H
5Qu
T
N
F-
2p
hydroxysteroids, cortisol (F) and cortico- sterone (B), predominated (Fig. 5). On stimulation with ACTH there was a 10- to 15-fold increment in the total cortico- steroid output (Table 3). Subsequent ad- dition of monobutyryl cyclic AMP gave a response of similar magnitude. The only apparent distinction from normal adrenal cells was the secretion of greater amounts of corticosterone relative to cortisol.6 In absolute terms the total output of steroids per cell was comparable with that of normal cortical cells, as was the pat- tern of conjugated steroid secretion. The change in the ultrastructural appearance of the cultured adenoma cells thus in no way precludes continued hormone syn- thesis and secretion in vitro for extended periods of time (Fig. 4).
These functional features of the tumor from patient 1 were identical to those of three other adenomas (including two “black” adenomas) causing Cushing’s syndrome that we cultured over a three year period (cases 7 to 9).
Adrenocortical Carcinomas
The generally accepted histologic cri- teria of malignancy, viz., frequent mi- toses, pleomorphism, capsular invasion, and the presence of tumor cells in vascu- lar and lymphatic spaces, are unfortu- nately of limited value in the assessment of adrenocortical neoplasia.2-4 They may in our experience occur in tumors that subsequently pursue a benign course and can be absent in those that subsequently metastasize.3 Past experience with proven carcinomas would indicate, however, that the following morphologic indices are of particular value in assessing malignant potential: extensive arcas of necrosis, marked nuclear pleomorphism, and en- larged vesicular nuclei with one or more prominent nucleoli. These features are prominent in some tumors, mainly those larger than 500 grams, and the diagnosis in these cases is readily apparent. Howev- er, in others some areas may show mini- mal cellular and nuclear pleomorphism,
(Cholesterol)
Progesterone (P)
17a-Hydroxyprogesterone (17aOH-P)
Androstenedione (AD)
Deoxycorticosterone (DOC)
Deoxycortisol (S)
11
11
11
Corticosterone (B)
Cortisol (F)
118-Hydroxyandrostenedione (11BOH-AD)
although contiguous zones may contain bizarre, pleomorphic, and multinucleated cells (Fig. 6).
The attached glands in cases of func- tional adrenocortical carcinoma show varying degrees of atrophy, in contrast to
the normal zonation seen with “nonfunc- tioning” carcinomas.12 The most difficulty in making a histologic diagnosis of malig- nant disease is generally experienced with tumors of 100 to 300 grams, when the aforementioned cytologic indices, to-
gether with the formation of large alveoli or sheets of cells, become valuable indica- tors of malignant change.
In the present series of cases, which were selected only on the basis of the presence of a syndrome of steroid excess (excepting hyperaldosteronism), obvious features of malignant disease were pres- ent in tumors from patients 4 and 6, both of which metastasized within a short time (Fig. 7). Both patients are now dead (Table 2).
The functional responses of cultured cells from these patients were clearly ab- normal in comparison with both adeno- ma and normal adult adrenocortical cells (Table 3). Both failed to secrete signifi- cant amounts of 118-hydroxysteroids, with compounds such as 11-deoxycortisol (S), deoxycorticosterone (DOC), and an- drostenedione (AD) predominating (Fig. 5). There was no observable change in the morphology of the cultured cells on the addition of ACTH. In neither case
was extensive proliferation of tumor cells noted in vitro (Table 3).
Case 5 presented problems (Tables 2, 3). This 880 gram tumor consisted of sheets of relatively uniform, compact cells with minimal nuclear pleomorphism (Fig. 8). There was little necrosis and mitotic activity was inconspicuous. Many of our panel of histopathologists believed that this was probably a benign lesion (Table 2). We, however, have never seen a prov- en benign tumor of this size, and func- tional studies in culture indicated distinct abnormalities (see later discussion). The virilizing syndrome associated with this tumor recurred 15 months after resec- tion, and although metastases had not been detected at the time of this writing, the prognosis must be considered poor.
This tumor from patient 5 also showed clear functional abnormalities in culture (Table 3). The major steroid se- creted was 11-deoxycortisol, and al- though a small response to cyclic AMP
was elicited, no response to ACTH was obtained. No significant proliferative ac- tivity of these tumor cells was noted, and cultures were eventually overgrown by non-neoplastic nonsteroidogenic fibro- blasts.
The last two patients (2 and 3) pre- sented different problems. The tumor from case 3, despite its small size, showed histologic features highly suggestive to us of malignancy (Fig. 9). Our histologic im- pression, that the tumor from patient 3 was potentially malignant, was also sug- gested by functional studies. Although this tumor retained responses to both ACTH and cyclic AMP, both were signi- ficantly blunted in quantitative terms (Table 3). Like the larger tumors (cases 5 and 6; Table 3), its major secreted steroid was not cortisol (F) but its precursor 11- deoxycortisol (S), although stimulation with ACTH (and cAMP) did in this case cause a reversion to cortisol secretion. Its most significant deviation from the be- havior of normal and benign neoplastic adrenocortical cells was that it initially
underwent rapid proliferation in culture, with increasing steroid levels, although this replicative activity was not sustained. Patient 3 died with metastases eight months after operation.
The tumor from patient 2 was even smaller (92 grams) than that from patient 3. It consisted of small islands of com- pact, relatively uniform cells between which numerous bizarre pleomorphic, often giant cells were present, many of which contained aggregates of PAS posi- tive material (Fig. 10). This uncharac- teristic appearance in fact led a signifi- cant proportion of our panel to suggest, not unreasonably in the absence of clini- cal data, that it might be a small alveolar type of pheochromocytoma.13 The tumor from patient 2, however, was composed of cells with recognizable and indeed prominent ultrastructural features of adrenocortical cells (Fig. 11). Many cells contained large numbers of highly dif- ferentiated mitochondria with tubulola- mellar and tubulovesicular cristae, exten- sive areas of smooth endoplasmic 147
reticulum, and characteristic short strands of rough endoplasmic reticulum (Fig. 11). Careful examination revealed somewhat fewer desmosome-like junc- tions between adjacent cells than in case 1, but there were notably many complex interdigitations between apposed cell membranes (Fig. 11). Thus the degree of ultrastructural adrenocortical differentia- tion observed in this tumor was if any- thing even greater than in the benign tumors. Nonetheless the malignant po- tential of this tumor was demonstrated by recurrent hypercorticalism and metas- tases within six months of surgery.
Unlike the other carcinomas studied, that from patient 2 gave an essentially normal pattern of steroidogenesis in cul- ture, with cortisol as the major product in both the presence and the absence of ACTH, although it retained a much greater response to cyclic AMP than to the hormone (Table 3). The most signifi- cant deviation from normal behavior in
this case was that the tumor cells did not form a monolayer at all, but remained as loosely attached or floating aggregates of spherical, lipid laden cells that underwent slow proliferation for several months (Fig. 3). Indeed it was only by ultrastruc- tural examination of these aggregates that we were able to confirm that they were composed of cortical tumor cells (Fig. 12).
DISCUSSION
The present studies have demon- strated two facts. First, diagnosis of the malignant potential of adrenocortical neoplasms by light microscopic examina- tion poses difficulties with a significant number of functionally active lesions. Nonetheless, the criteria of extensive areas of necrosis, marked nuclear pleo- morphism, and enlarged vesicular nuclei with one or more prominent nucleoli, af-
| Case . No. | Morphology | Morphologic Response to ACTH | Growth | Functional Response | Major Secreted UV-absorbing Steroids | |
|---|---|---|---|---|---|---|
| ACTH | mbCAMP | |||||
| 1 | Epithelioid* | ++++ | - | +++ | 118-hydroxysteroids (F + B) | |
| 2 | Unattached spherical aggregates | - | ++ | ++++ | 118-hydroxysteroids (F) | |
| 3 | Fibroblast-like | = | +++ ** | + | ++ | 11-deoxysteroids (S + DOC) *** |
| 4 | Fibroblast-like | - | - | (+) **** ☒ | (+) **** | 11-deoxysteroids (17aOH-P + S) |
| 5 | Fibroepithelioid | - | - | - | ++ | 11-deoxysteroids (S) |
| 6 | Attached spherical aggregates | - | + | - | - | I 1-deoxysteroids (S) |
*The same functional and structural responses were seen in tumors from cases 7, 8, and 9.
** Rapid proliferation of tumor cells for three to four weeks only.
*** Reverting to 118-hydroxysteroids (F) with ACTH.
**** Qualitative but no quantitative changes.6
ford a good guide to malignancy. Sec- ond, dynamic functional cell culture stud- ies give additional parameters that are of significance in the distinction of benign and malignant tumors, although no sin- gle criterion has emerged as diagnostic in this respect. In the cases that we have ex- amined there has nonetheless been a good correlation between, on the one hand, the presence of morphologic fea- tures we consider indicative of malignan- cy and abnormalities of function and be- havior in the cultured cells and the final outcome of the disease on the other (Tables 2, 3).
One question of prime importance in assessing the role of cell culture in the diagnosis of adrenal malignant disease is whether the tumor cells retain the func- tional attributes they express in vivo. We have shown that factors such as cell den- sity in monolayer cultures can influence
the functional behavior of normal adult adrenocortical cells.10 Nevertheless, care- ful comparison of tumor cultures with equivalent normal adrenal cultures has convinced us that the functional attri- butes of the tumors are characteristic and pathognomonic. Thus, all benign tumors in culture behave structurally and func- tionally as if they were normal cells. On the contrary five of six carcinomas showed a relative deficiency of 116- hydroxysteroid secretion (Table 3). There have been only a few other isolat- ed reports of functioning adrenocortical tumors maintained in culture, but they substantiate our findings that adenomas behave like normal cells, whereas carcino- mas may show distinct functional abnor- malities. 14, 15
In one of the largest functional stud- ies of adrenocortical carcinomas in vivo Lipsett and Wilson16 noted elevated levels
S
M
0.5p
R
0.5p
of urinary 11-deoxysteroid metabolites, such as tetrahydro-S, in nine of 10 tumors examined. Subsequent studies in vivo and short term in vitro experiments have confirmed the presence of an 118- hydroxylase “block” in many individual functioning adrenocortical carcinomas.17- 22 However, this has not invariably been found, as is illustrated by our case 2, which we have illustrated here as a “mini- mal deviation” case from a functional point of view. A deficiency of 118- hydroxysteroids is not necessarily, there- fore, a sine qua non of malignancy or metastasis.23 We suggest rather that it is a probable consequence of abnormalities, possibly relating to cell contact and com- munication, that are common in malig- nant, but not benign, cortical tumor cells.6
Is the failure of response to ACTH pathognomonic of cortical malignant change? In previous studies in vivo it was noted that approximately half of all adenomas but only a small minority of
carcinomas responded to ACTH with in- creased steroid secretion.24, 25
In our cultures we have observed that all adenomas (four cases of Cush- ing’s syndrome reported here and 16 al- dosteronomas) have invariably responded to long term treatment with ACTH with an increase in total steroid output compa- rable to that observed with normal corti- cal cells.1º The failure of about half of the adenomas to respond to ACTH in vivo may well lie in the duration of the stimulus, i.e., acute versus trophic activa- tion.23 Short term in vitro preparations of benign tumors, such as cell suspensions or tissue slices, may or may not respond to ACTH.26, 27 A failure to respond
usually occurs with lesions composed pre- dominantly of compact (reticularis-like) cells, including’ some “black” adeno- mas.27, 28 This behavior mirrors the limit- ed response of normal human reticularis cells in similar short term experiments.29 Separated human adult fasciculata and 151
T
T
T
N
1μ
reticularis cells in culture, however, show an equal trophic response to ACTH when treated for several days with the hormone (unpublished observations). The value of long term culture prepara- tions in enhancing this discriminant be- tween benign and malignant cortical tumors is thus evident.
The present studies have also shown that more carcinomas are stimulated by ACTH (three of five) than might be pre- dicted from the studies in vivo.23 In all cases, however, the response was markedly blunted, in comparison with benign le- sions and thus may not have been detect- able in vivo (Table 3). Four of five of these tumors did nevertheless respond to monobutyryl cyclic AMP (mbcAMP) sometimes to a much greater extent than to ACTH, indicating that only part of the mechanism of hormone activated steroi- dogenesis may be deficient in some “au- tonomous” functional carcinomas (Table 3).30 This response to mbcAMP affords another discriminant, as all benign le-
sions cultured have responded, like nor- mal cortical cells, equally to hormone and cyclic nucleotide.
It does not appear, on the basis of the present tumors studied, that ultra- structural examination aids significantly in assessing the malignant potential of functional lesions. Proven malignant change may in some cases go hand in hand with highly differentiated features (thus case 2).23, 31, 32 Criteria such as amorphous as opposed to differentiated mitochondrial structure and fragmented as opposed to membranous rough endo- plasmic reticulum probably serve only to distinguish “nonfunctional” and weakly active lesions from those causing severe hypercorticalism.12, 28 The latter can in- clude both benign and malignant lesions. Structural criteria such as nuclear pleo- morphism and enlarged nucleoli can be observed as well if not better at the light microscopic level. Disruption of basement membrane material between cortical and endothelial cells may possibly indicate
malignancy but may be difficult to ob- serve in areas of imperfect preserva- tion.28, 31
We conclude, therefore, that any neoplasm with detectable functional ab- normalities in vitro must be viewed as po- tentially malignant and would re- emphasize that a search for significantly elevated levels of 11-deoxysteroids rela- tive to 118-hydroxysteroids may prove the most effective means of distinguish- ing the majority of functionally active carcinomas of the adrenal cortex in vivo. The assay of 11-deoxysteroids in plasma or their metabolites in urine might also afford a more effective method of moni- toring the postoperative course of malig- nant disease and its response to therapy.
ACKNOWLEDGMENTS
We are grateful to our surgical col- leagues for supplying the material re- ported here and to E. C. Nice and R. Magee-Brown for excellent technical as- sistance.
REFERENCES
1. Neville, A. M., and Symington, T. S .: Pathology of primary aldosteronism. Cancer, 19:1854, 1966.
2. Neville, A. M., and Symington, T. S .: The pa- thology of the adrenal gland in Cushing’s syndrome. J. Path. Bact., 93:19, 1967.
3. Symington, T. S .: Functional Pathology of the Human Adrenal Gland. Edinburgh, E. & S. Livingstone, 1969.
4. Neville, A. M., and Mackay, A. M .: The struc- ture of the human adrenal cortex in health and disease. Clin. Endocr. Metab., 1:361, 1972.
5. Neville, A. M .: The nodular adrenal. Invest. Cell. Path., 1:99, 1978.
6. O’Hare, M. J., Ellison, M. L., and Neville, A. M .: Tissue culture methods in endocrine re- search. Curr. Top. Exp. Endocr., 3:1, 1978.
7. O’Hare, M. J., and Neville, A. M .: Steroidogen- ic response of adult rat adrenocortical cells in monolayer culture. J. Endocr., 56:537, 1973.
8. Hornsby, P. J., O’Hare, M. J., and Neville, A. M .: Functional and morphological observa- tions on rat adrenal glomerulosa cells in mon- olayer culture. Endocrinology, 90:1240, 1974.
9. Hornsby, P. J., and O’Hare, M. J .: The roles of potassium and corticosteroids in determining the pattern of metabolism of (3H) deoxycorti- costerone by monolayer cultures of rat adre- nal glomerulosa cells. Endocrinology, 101:997, 1977.
10. Neville, A. M., and O’Hare, M. J .: Cell culture and histopathology of the adrenal cortex in relation to hypercorticalism. In James, V. H. T., et al. (Editors): The Endocrine Function of the Human Adrenal Cortex. New York, Academic Press, Inc., 1978, p. 229.
11. O’Hare, M. J., Nice, E. C., Magee-Brown, R., and Bullman, J .: High pressure liquid chro- matography of steroids secreted by human adrenal and testis cells in monolayer culture. J. Chrom., 125:357, 1976.
12. Lewinsky, B. S., Grigor, K. M., Symington, T., and Neville, A. M .: The clinical and patho- logical features of “non-hormonal” adreno- cortical tumors. Cancer, 33:778, 1974.
13. Neville, A. M .: The adrenal medulla. In Sy- mington, T. S. (Editor): Functional Pathology of the Human Adrenal Gland. Edinburgh, E. & S. Livingstone, 1969, p. 219.
14. Israeli, E., Levy, J., Rosenthal, E., Auslaender, L., Peleg, I., and Barzilai, D .: A human adrenocortical adenoma in tissue culture: morphology and hormone secretion. Israeli J. Med. Sci., 11:1106, 1975.
15. Szabo, D., Gyevai, A., Glaz, E., Stark, E., Peteri, M., and Alant, O .: Changes in the fine struc- ture and function of a hormone-secreting adrenocortical tumour investigated in tissue culture. Virchows Arch. A .Path. Anat., 367:273, 1975.
16. Lipsett, M. B., and Wilson, H .: Adrenocortical cancer: steroid biosynthesis and metabolism evaluated by urinary metabolites. J. Clin. En- docr., 22:906, 1962.
17. Martin, F. I. R .: Evidence of an hormonal in- fluence on the steroid output of adrenal car- cinoma. Am. J. Med., 32:795, 1962.
18. West, C. D., Kumagai, L. F., Simons, E. L., Do- minguez, O. W., and Berliner, D. L .: Adreno- cortical carcinoma with feminization and hy- pertension associated with a defect in 118-hydroxylation. J. Clin. Endocr., 24:567, 1964.
19. Bryson, M. J., Young, R. B., Reynolds, W. A., and Sweat, M. L .: Biosynthesis of steroid hor- mones in a human feminizing adrenocortical carcinoma. Cancer, 21:501, 1968.
20. Herbeuval, R., Boulangé, M., Rauber, G., Guer- ci, O., and Petitier, H .: Cortico-surrénalome malin avec hyperminéralocorticisme prédo- minant: production anormale par la tumeur, de cortexone et de cortexolone. Rev. Fr. En- docrinol. Clin., 9:489, 1968.
21. Powell-Jackson, J. D., Calvin, A., Fraser, R., Graham, R., Mason, P., Missen, G. A. K., Powell-Jackson, P. R., and Wilson, A .: Excess deoxycorticosterone secretion from adreno- cortical carcinoma. Br. Med. J., 2:32, 1974.
22. Begue, R .- J., Brun, J .- M., Desgres, J., and Pa- dieu, P .: Steroides urinaires dans un cortico- surrénalome virilisant. J. Steroid Biochem., 7:583, 1976.
23. Fraser, R., James, V. H. T., Landon, J., Peart, W. S., Rawson, A., Giles, C. A., and MacKay, A. M .: Clinical and biochemical studies of a patient with a corticosterone-secreting adren- ocortical tumour. Lancet, 2:1116, 1968.
24. Soffer, L. J., Geller, J., and Gabrilove, J. L .: Re-
sponse of the plasma 17-hydroxycorticosteroid level to gel-ACTH in tumorous and non- tumorous Cushing’s syndrome. J. Clin. Endocr., 17:878, 1957.
25. Scott, H. W., Liddle, G. W., Mulherin, J. L., McKenna, T. J., Stroup, S. L. and Rhamy, R. K .: Surgical experience with Cushing’s dis- ease. Ann. Surg., 185:524, 1977.
26. Voight, K. H., Fehm, H. L., and Pfeiffer, E. F .: Dissociation of in vivo and in vitro autonomy in a human adrenocortical tumour. Acta En- docr., 78:302, 1975.
27. Hasagawa, K., Moriwaki, K., Igarashi, T., Suga- sa, T., Kawakami, F., Itoh, Y., and Nishi- kawa, M .: Studies on the responsiveness of human adrenocortical tumours to ACTH. Folia Endocr. (Jap.), 51:749, 1975.
28. Mitschke, H., Saeger, W., and Breustedt, H .- J .: Zur Ultrastructur der Nebennierenrindentu-
moren beim Cushing-syndrome. Virchows Arch. Abt. A Path. Anat., 360:253, 1973.
29. Griffiths, K., Grant, J. K., and Symington, T .: A biochemical investigation of the functional zonation of the adrenal cortex in man. J. Clin. Endocr., 23:776, 1963.
30. Saez, J. M., Dazord, A., and Gallet, D .: ACTH and prostaglandin receptors in human adren- ocortical tumors. J. Clin. Invest., 56:536, 1975.
31. Mackay, A. M .: Atlas of human adrenal cortex ultrastructure. In Symington, T. S. (Editor): Functional Pathology of the Human Adrenal Gland. Edinburgh, E. & S. Livingstone, 1969, p. 346.
32. Thiele, J .: Feinstrukturelle Untersuchungen an einem endokrinaktiven Carcinom der Neben- nierenrinde. Virchows Arch. B Cell. Path., 17:51, 1974.
Unit of Human Cancer Biology Ludwig Institute for Cancer Research (London Branch)
Royal Marsden Hospital Sutton, Surrey SM2 5PX England (Dr. O’Hare)