Primary Aldosteronism Caused by Adrenal Cortical Carcinoma
HARRIS C. TAYLOR1, JANICE G. DOUGLAS?, GREGORY J. BERG3, AND EMMANUEL L. BRAVO4
1 Asst. Clinical Professor of Medicine Case Western Reserve University, Cleveland Ohio
2 Assoc. Professor of Medicine Case Western Reserve University, Cleveland Ohio
3 Director, Steriod Hormone Section Endocrine Sciences, Tarzana, California
+ Research Division, The Cleveland Clinic Foundation, Cleveland Ohio
Abstract
Adrenocortical carcinoma rarely produces pure primary aldosteronism. We document the occurrence of such a case. Through tumor steroid analysis, we show that the content of aldosterone per gram of tumor tissue is diminished compared to aldosterone producing adenomas. We also demonstrate the occurrence of angiotensin II receptors on the tumor, a finding hitherto noted only on aldosterone producing adenomas.
Primary aldosteronism ordinarily signifies autonomous production of aldosterone by an adrenal adenoma or bilateral hyperplasia of the adrenal glands. Rarely, it may be caused by a carcinoma. When produced by a tumor, the neoplasm is usually benign, less than four centimeters in diameter and associated with the classical features of mineralocorticoid excess, hypokalemia, hyporeninemia and hy- peraldosteronism (Weinberger et al., 1979; Bennett et al., 1971). In contrast, carcinomas producing primary aldosteronism are larger and exhibit increased secretion of cortisol and/ or androgens (Alterman et al., 1969; Brooks et al., 1972; Filipecki et al., 1972). Nonetheless, a smaller group of eight patients (Crane et al., 1965; Revach et al., 1977; Salassa et al., 1974) has been reported in which primary aldo- steronism appears to have occurred as the only
Received August 2, 1982.
Report of a Case with Measurement of Tumor Aldosterone Content and Angiotensin II Receptors
endocrine manifestation of an adrenocortical carcinoma. We report a similar patient with chemical and clinical primary aldosteronism, due to adrenocortical carcinoma. Results of the tumor analysis of steroid content indicate diminished aldosterone content per gram of tumor tissue compared to benign aldosterone producing adenomas (APA). We demonstrate the presence of angiotensin II receptors on the tumor, a finding hitherto noted only on APA (Brown et al., 1980).
Methods
Tumor tissue aldosterone determination
Tissue was prepared in the following manner: eight grams of frozen tumor tissue were pulverized with a tissue pulverizer. Eight ml of phosphate saline buffer was then added, and the tissue was homogenized with a polytron. Methanol (32 ml) containing 0.1% triethylamine was added to the homogenate. The mix- ture was shaken for 30 minutes and then centrifuged for 15 minutes at 3000 r.p.m. Multiple aliquots of the supernatant were transferred to extraction tubes, and
the methanol was evaporated with a stream of nitrogen in a water bath. Recovery tracers, approximately 20,000 dpm, were added for the steroid measured. Final tissue concentrations are adjusted for the per- centage recovery. Distilled water was added to each tube to adjust the volume to the amount used in each assay. Control cadaver adrenal tissue was processed in a similar manner at a later date. Only the tail por- tion of the adrenal glands was used, therebyl eminat- ing most of the medulla. Three of the four glands were frozen within 6 hours of death and the fourth within 12 hours. All cadavers were stored in a cooler at a temperature of 3.5-6.8℃ prior to autopsy. Previ- ous study (Kaplan, 1967) shows less than 15% loss of steroid activity in adrenal tissues processed in this fashion.
Aldosterone was measured by RIA using a pro- cedure previously described (Mayes et al., 1970). Aliquots of tumor supernatant were extracted with methylene chloride, evaporated under nitrogen and purified by paper chromatography using a Bush B-5 system. The aldosterone content of the purified sam- ples was determined by RIA using an antiserum ·developed against an aldosterone 3-oxime conjugate.
Angiotensin II receptor binding procedure.
Frozen adrenal tumors (aldosterone-producing carcinoma and cortisol producing adenoma) were thawed, minced and dispersed using a teflon-glass homogenizer in 20 mm sodium bicarbonate solution as previously described in detail for the rat adrenal (Douglas and Catt, 1976). The homogenates were filtered through nylon fabric and the 10,000-100,000 x g fractions were used for studies of angiotensin II bind- ing. The pellets were resuspended in 50 mM Tris buffer pH 7.4 to a final concentration of 2.1 mg/ml with the aldosterone producing carcinoma and 5 mg/ml with the cortisol producing adenoma. Adrenal tissue was incubated with 50 mm Tris-HCI (pH 7.4) with 120 mM NaCl, 5 mm dithiothreitol, 0.8% BSA, 20 fmoles of 125I-angiotensin II and 100-250 µg of protein. Binding studies were performed in triplicate at 20℃ for 2 to 20 minutes. All binding was corrected for non-specific binding in the presence of excess unlabeled angiotensin II (1 µg). Non-specific binding represented less than 25% of total binding with the aldosterone-producing carcinoma and approximately 50% of total binding with the cortisol-producing adenoma. Separation of bound and free hormone was achieved by Millipore filtration using HAWP (0.45 u) nitrocellulose filters. Filters were counted in a Packard gamma spectrometer. Comparisons are made to binding by fresh and frozen human and rat adrenal glomerulosa. A Statie Riggo microtone was used to obtain the outer 0.1 mm cortical slices from a patient who underwent bilateral adrenal- ectomy for Cushing’s Disease. This procedure yields tissue with significant increments in aldosterone pro- duction to low concentrations of angiotensin II in vitro (Brown et al., 1980). Rat tissue was the decapsu-
lated portion of adrenals from female Sprague Dawley rats (225-250 g) as previously described (Douglas and Catt, 1976). Identical procedures were utilized for preparation of all tissues. Binding constants were determined by Scatchard analyses with the aid of a computer for linear regression analysis. All human tissue used for these studies were obtained with the approval of the Human Experimentation Committee, University Hospitals of Cleveland.
Urinary aldosterone, plasma aldosterone (PA) and plasma renin activity (PRA) were determined in ac- cordance with previously published procedures (Bravo et al., 1975a; Bravo et al., 1975b; Bravo et al., 1973). Each PA determination is the mean of duplicate deter- minations at two separate dilutions.
Case Report
A 53 year-old white male was evaluated in July 1978, because of hypertension. The pati- ent was aware of his elevated blood pressure for at least ten years and noted it had always been poorly controlled. Since November 1977, he had experienced muscle weakness in both shoulders and arms. Because of an elevated CPK determination of 626 IU (nl 0-83 IU), his previous physician diagnosed polymyositis and began him on prednisone, 20 mg each day. No other diagnostic procedure was performed. Although the patient believed he experienced some benefit from treatment, his symptoms persisted. His past history revealed chronic alcoholism with a recent exacerbation mani- fested by the consumption of a half gallon of gin every week for the preceeding four months.
On physical examination, the blood pres- sure was 180/100 mm Hg while on hydrochloro- thiazide, 50 mg each day, and propranolol, 40 mg BID. Grade II vascular changes were noted in the fundi. Examination of the abdo- men revealed a large mass in the area of the left kidney. The remainder of the physical examination was normal except for generalized muscle weakness. In particular, there were no stigmata of Cushing’s syndrome. The com- plete blood count was within normal limits except for a WBC of 13,100, with a normal differential. Serum potassium was 1.9 mEq/L. The urinalysis showed a trace of protein but
no other abnormality. The chest X-ray was within normal limits. A previous ECG from April 1978 revealed T-U fusion characteristic of hypokalemia when the serum potassium was 1.8 mEq/L. Neurological consultation failed to substantiate a diagnosis of polymyositis. The prednisone was gradually tapered and the hydrochlorothiazide discontinued. Oral po- tassium supplementation was begun and on August 4th, the blood pressure was 200/100 mm Hg, the serum potassium 3.0 mEq/L, sodium 148 mEq/L and CO2 29 mEq/L. Serum creatinine was 1.5 mg/dl. A twenty four hour urine protein was 607 mg. An IVP showed a large mass above the left kidney. The patient developed alcoholic hepatitis and was hospi- talized beginning August 18th. During this period, KCI supplementation initially 40 mEq TID and later 40 mEq BID was given. Follow- ing resolution of the alcoholic hepatitis, evalua- tion of his adrenal function was undertaken. Table 1 shows the diurnal variation of serum cortisol, urinary excretion of glucocorticoid metabolites and the response of serum and urinary glucocorticoids to low dose dexa- methasone suppression. Table 2 shows the baseline level of PA, PRA, postural changes and the 24-hour aldosterone excretion rate (AER). The effect on these values of an abortive attempt to salt load the patient is likewise noted. Unfortunately, left ventricular failure followed the initial 2000 ml saline load and precluded further attempts to suppress aldosterone production. A 24-hour urine for metanephrines was 0.8 mg/24 h (nl 0.3-0.9 mg/ 24 h). Following a left renal angiogram, the patient was operated on September 19th, 1978. A 320 gram encapsulated tumor meas- uring 9x9x8 cm was removed. There was no evidence of local invasion or lymph node metastases. Limited exploration of the liver failed to reveal any gross metastases. No liver biopsy was performed. Pathological interpretation of the tumor revealed a plemor- phic neoplasm of the adrenal cortex with mitoses, capsular invasion, broad fibrous bands and areas of vascular invasion (Fig. 1). In
addition, several of the slides showed a pat- tern of anastomosing vascular sinusoids with parallel arrays of necrosis following the pat- tern of these sinusoids in the tumor. On October 12, 1978, off potassium supplementa- tion, the serum potassium was 5.5 mEq/L and the blood pressure 190/110 mm Hg. Hydro- chlorothiazide was begun and on October 24th, while taking 50 mg BID, the serum potassium was 4.7 mEq/L. On the same day, an afternoon ACTH stimulation test using Cortrosyn(R) 0.25 mg I.M. was performed. The patient had been up ad lib during the day. The baseline serum cortisol of 4.4 ug/dl and aldosterone of 10.6 ng/dl increased to 9.1 ug/ dl and 10.8 ng/dl at 30 minutes, 10.9 ug/dl and 16.9 ng/dl at 60 minutes and 11.3 ug/dl and 16.2 ng/dl at 90 minutes, respectively. When last seen on February 1, 1979, the blood pres- sure was 160/100 mm Hg, while on pro- pranolol, 80 mg QID, but no hydrochloro- thiazide. The serum potassium was 5.5 mEq/ L. The patient was known to be alive and well as of July 1981. He continued to take pro- pranolol 80 mg QID and hydrochlorothiazide 100 mg QD from another physician because of continued hypertension.
Results
As shown in Table 1, diurnal variation of serum cortisol was exhibited while baseline excretion of urinary 17-hydroxycorticosteroids (17-OHCS), 17-ketosteroids and urinary free cortisol were all within normal limits. Urinary free cortisol and plasma cortisol were suppres- sed normally on dexamethasone, 0.5 mg every 6 hours, although urinary 17-OHCS failed to show normal suppression. Table 2 shows markedly elevated values of both PA and urinary AER and suppressed PRA. Assuming the upright posture for four hours produced a slight increment in PRA on two occasions. At the same time, PA increased significantly. An attempt to demonstrate autonomy of aldosterone production by three days of saline
a
AT
FC
A
b
4
| 17-OHCS ** (mg/24 hrs) | 17-Ketosteroids (mg/24 hrs) | Urinary Free Cortisol (µg/24 hrs) | TV (ml) | Serum Cortisol (µg/dl) | |
|---|---|---|---|---|---|
| 8/27/78-8/28/78 | 10.7 (nl 4-12) | 5.5 (nl 10-20) | 43 (nl 20-90) | 2650 | 0.800 h=18.0 (nl 8-19) 16.00 h=7.0 (nl 50% of 8 am value) |
| 8/31/78-9/1/78 | 8.9 | ||||
| 6.7 | 8 | 1600 | 0.800 h= 2.9 (9/1) |
* Patient received 0.5 mg dexamethasone q6h from 8 a.m. on 8/30 through 2 a.m. 9/1 and chlordiazepoxide 25 mg hs QD.
** Urinary 17-hydroxycorticosteroids=17-OHCS.
| Plasma Aldosterone (ng/dl) | Plasma Cortisol (µg/dl) | Plasma Renin Activity (ng/ml/hrs) | Serum Na/K (mEq/L) | Aldosterone Excretion rate (µg/24 hrs) | ||
|---|---|---|---|---|---|---|
| 9/5/78 | 0.800 h | 105 | 18.0 | 0.1 | 139/ | 105.6 |
| (recumbent) | (nl 3-16) | (nl 0.6-1.2) | 4.9 | (nl 3-12) | ||
| 12.00 h | 196.8 | 12.5 | 0.35 | |||
| (upright) | (nl 2-3 fold in- | (nl 2-3 fold | ||||
| crease) | increase) | |||||
| 9/7/78 | 11.30 h | 139 | 0.1 | 145/ | ||
| (10 minutes post 40 mg | 4.2 | |||||
| of furosemide) | ||||||
| 15.30 h | 90 | 6.9 | 0.08 | |||
| 9/8/78 | 141/ | |||||
| 3.6 | ||||||
| 9/9/78 | 0.800 h | 124.8 | 18.4 | 0.1 | ||
| (recumbent) | ||||||
| 12.00 h | 167.5 | 7.2 | 0.18 | |||
| (4 hrs upright) |
* From 9/3 through 9/9, the patient was maintained on a 5 Gm NaCl diet with KCL, 40 mEq BID. Two liters of normal saline were administered IV on 9/6. Pulmonary edema developed the night of 9/6.
loading was aborted by the patient developing left ventricular failure after the first day (Table 2). At least partial refractoriness of PRA to stimulation was confirmed by obtaining a value of 0.07 ng/ml/h on 8/28/78 three and a half hours after administration of 20 mg furosemide IV and assuming the upright pos- ture at 8 a.m. Five weeks following surgery the upright PA level was approximately 6% of the average pre-operative values and demon- strated impaired responsiveness to ACTH.
Results of tumor and normal adrenal tis- sue analysis for aldosterone are as follows.
The aldosterone concentration in the present tumor, 0.106 ug/gm, is markedly lower than the previously reported average value of 10.3 µg/gm (Kaplan, 1967), 0.82 µg/gm (Biglieri et al., 1963) and 1.89 µg/gm (Louis and Conn, 1961) in APA. Control adrenal tissue (n=4) showed an average aldosterone concentration of .041 µg/gm, similar to that of .06 ug/gm (Louis and Conn, 1961) but less than 0.27 µg/ gm (Biglieri et al., 1963; Kaplan, 1967). How- ever, the weight of the present tumor is 320 grams and the total aldosterone content, therefore, is significantly higher than usually
ANGIOTENSIN II BOUND (fmoles/gm protein)
600
400
ALDOSTERONE PRODUCING CARCINOMA
200
O
9
T
I
CORTISOL PRODUCING ADENOMA
₹
5
10
15
20
TIME
(MINUTES)
found in APA. Figure 2 demonstrates specific angiotensin II bound by the present aldo- sterone-producing carcinoma and, in contrast, a cortisol-producing adenoma, utilizing a single concentration of ligand (20 fmoles). Both tissues were stored frozen prior to use. Bind- ing was highest between ten and fifteen minutes with both tumors, but was 3.6 fold higher with the aldosterone-producing carcinoma. Use of tissues for receptor binding that were previously frozen results in loss of receptor number and lower affinity as compared with use of the same tissues when fresh. Examples of two such experiments, employing normal rat adre- nal and human adrenal glomerulosa are pre- sented in Table 3. Each represents a significant loss in the number of receptor sites, whereas only the rat demonstrated a significant increase in Ka·
Discussion
Several interesting features in this patient deserve comment. Initially, he demonstrates the occurrence of an unusually “pure” form of primary aldosteronism in association with a large adrenal cortical neoplasm. Normal
| Human | Rat | ||
|---|---|---|---|
| Receptor Number | FRESH | 736 | 2678 |
| (fmoles/mg protein) | FROZEN | 602 | 1693 |
| Ka (nM) | FRESH | 2.6 | 0.71 |
| FROZEN | 3.2 | 2.1 | |
urinary free cortisol, plasma cortisol, diurnal variation of plasma cortisol and response of both to low dose dexamethasone suppression adequately eliminate concurrent excessive cortisol production. The failure of urinary 17-OHCS to normally decrease on low dose dexamethasone may reflect the presence in the urine of chlordiazepoxide metabolites, which are known to interfere with the colorimetric determination (AMA Drug Evaluations 1977). Salassa reported six similar patients with large adrenal cortical tumors and primary aldo- steronism culled from a group of 105 patients with primary aldosteronism operated on at the Mayo Clinic over a seventeen year period (Salassa et al., 1974). Unfortunately, neither urinary free cortisol determinations nor dexa- methasone suppression testing were performed
in this study although diurnal plasma cortisol variations appeared to exist. Occasional patients with similar presentation, likewise, had normal urinary excretion of 17-ketosteroids and 17-OHCS (Revach et al., 1977); Crane et al., 1965). These patients differed from pre- viously reported cases of adrenal cortical carcinoma associated with primary aldostero- nism where elevation in either 17-ketosteroids, 17-OHCS or urinary free cortisol was noted (Alterman et al., 1969; Brooks et al., 1972; Filipecki et al., 1972).
Secondly, the aldosterone concentration of the present tumor is diminished compared with APA. The striking clinical picture, therefore, is consequent to the large weight of the tumor. In this regard, it is of interest that recalcula- tion of previous data (Louis and Conn, 1961) shows that the mean aldosterone content of 12 APA is 5.3 ug/APA. Allowing for differences in methodology, this value is less than the total content of aldosterone in our patient’s tumor (33.9 µg).
Thirdly, studies of angiotensin II binding to the tumor confirm the presence of angio- tensin II receptors on this carcinoma indicat- ing that there is potential for some degree of regulation of hormonal output by angiotensin II. Similar studies of benign aldosterone-pro- ducing adenomas have linked the presence of angiotensin II receptors with a sensitive aldo- sterone response to angiotensin II in vitro (Brown et al., 1980). While the present studies were limited and performed at a single ligand concentration, there was specific binding by the carcinoma which would likely have been higher if the study had been performed using fresh rather than frozen tissue (Table 3). At an equivalent ligand concentration fresh normal human adrenal glomerulosa bound 1060 fm/ gm protein and a benign APA bound 685 fm/gm protein at steady state. Hence, the plateau of binding at 470 fm/gm protein by the aldosterone-producing carcinoma was not significantly different from the APA and would have probably been higher had the tissue been studied while fresh. The cortisol-producing
adenoma, however, bound significantly less angiotensin II with a plateau of only 130 fm/ gm protein.
On the basis of the preceding data, we believe the classical syndrome of primary aldosteronism may be produced by an adrenal cortical carcinoma. The significance of angio- tensin II receptors on this as well as benign aldosterone-producing tumors has as yet not been determined (Brown et al., 1980). We suggest the hypothesis that perhaps their pre- sence may explain why some of these neo- plasms are associated with a parodoxical rise in plasma aldosterone on assuming the up- right posture.
Acknowledgements
We are indebted to Dr. Vinod Sutaria for referring this patient, Drs. Manuel Velasco, Aubrey Hough and Jerome Berner for reviewing the pathology slides and Nancy Coleman for typing the manuscript. These studies were supported in part by Grant HL-22990 from the USPHS. Dr. Douglas is a Research Career Development Awardee (HL-00399).
References
Alterman, S. I., C. Dominguez, A. Lopez-Gomez, and A. L. Lieber (1969). Primary adrenocortical carci- noma causing aldosteronism. Cancer 24, 602-609. AMA Drug Evaluations, Third Edition, (1977), PSG Publishing Company, Littleton, Massachusetts. p. 1214.
Bennett A. H., J. H. Harrison, and G. W. Thorn (1971). Neoplasms of the adrenal gland. J. Urol. 106, 607-614.
Biglieri E. G., S. Hane, P. E. Slaton, Jr., P. H. Forsham, M. A. Herron, and S. Horita (1963). In Vivo and In Vitro studies of adrenal secretions in Cushing’s syndrome and primary aldosteronism. J. Clin. Invest. 42, 516-524.
Bravo E. L., H. P. Dustan, and R. C. Tarazi (1975a). Selective hypoaldosteronism despite prolonged pre- and post operative hyperreninemia in primary aldo- steronism. J. Clin. Endocrinol. Metab. 41, 611-617. Bravo E. L., M. C. Khosla, and F. M. Bumpus (1975b). Vascular and adrenocortical responses to a specific antagonist of angiotensin II. Am. J. Physiol. 228, 110-114.
Bravo, E. L., H. P. Dustan, and R. C. Tarazi (1973). Spironolactone as a nonspecific treatment for pri-
mary aldosteronism. Circulation 48, 491-498.
Brooks R. V., D. Felix-Davies, L. M. Radcliffe, and P. W. Robertson (1972). Hyperaldosteronism from adrenal carcinoma. Br. Med. J. 1, 220-221.
Brown, G. P., J. G. Douglas, and E. L. Bravo (1980). Angiotensin II receptors and in vitro aldosterone responses of aldosterone-producing adenomas, adjacent non-tumorous tissue, and normal human adrenal glomerulosa. J. Clin. Endocrino !. Metab. 51, 718-723.
Crane M. G., J. J. Harris, and R. Herber (1965). Primary aldosteronism due to an adrenal carcinoma. Ann. Intern. Med. 63, 494-503.
Douglas, J. G., and K. J. Catt (1976). Regulation of Angiotensin II receptors in the rat adrenal cortex by dietary electrolytes. J. Clin. Invest. 58, 834-843. Filipecki S., T. Feltynowski, W. Poplawska, K. Lapin- ska, S. Krus, B. Wocial, and W. Januszewicz (1972). Carcinoma of the adrenal cortex with hyperaldo- steronism. J. Clin. Endocrinol. Metab. 35, 225-229. Kaplan N. M. (1967). The steroid content of adrenal adenomas and measurements of aldosterone produc-
tion in patients with essential hypertension and primary aldosteronism. J. Clin. Invest. 46, 728-734. Louis L. H., and J. W. Conn (1961). Primary aldo- steronism: content of adrenocortical steroids in adrenal tissue. Recent. Prog. Horm. Res. 17, 415- 436.
Mayes D., S. Furuyama, D. C. Kem, and C. A. Nugent (1970). A radioimmunoassay for plasma aldo- sterone. J. Clin. Endocrinol. Metab. 30, 682-685.
Revach M., S. Shilo, S. Cabili, Z. Rubinstein, and G. Selzer (1977). Hyperaldosteronism caused by adre- nal cortical carcinoma. Isr. J. Med. Sci. 13, 1123- 1128.
Salassa R. M., R. E. Weeks, R. C. Northcott, and J. A. Carney (1974). Primary aldosteronism and malignant adrenocortical neoplasia. Trans. Am. Clin. Climatol. Assoc. 86, 163-172.
Weinberger M. H., C. E. Grim, J. W. Hollifield, D. C. Kem, A. Ganguly, N. J. Kramer, H. Y. Yune, H. Wellman, and J. P. Donohue (1979). Primary Aldosteronism. Diagnosis, localization and treat- ment. Ann. Intern. Med. 90, 386-395.