Case Reports
Estrogen Hyperexcretion in a Patient With Nonendocrine Cancer
MELVIN J. KRANT, MD, BOSTON
D URING recent years, the association of adrenocortical hyperfunction and nonendocrine tumors has been noted. Elevation of serum hy- droxycorticoids and of urinary 17-hydroxy- and/or 17-ketogenic steroids has been found associated with carcinomatous states, principally in patients with bronchogenic cancer of the oat- cell type, thymomas, and pancreatic cancer.1-3 In- creased corticotropin-like activity has been demonstrated in plasma and in tumor tissue ex- tracts obtained from patients with these malig- nancies and may account for the adrenocortical hyperfunction.4 Widening of the adrenal cortex at postmortem examination has been noted in patients with malignancy, especially with bron- chogenic carcinoma.3,6
In the case to be reported, an excessive amount of urinary estrogen was found in a patient with a highly anaplastic carcinoma. At postmortem examination, widening of the adrenal cortex was demonstrated. This patient developed strik- ing spider angiomata and pulmonary osteoar- thropathy during the late stage of his rapid illness, suggesting a causal relationship with the elevated estrogen level.
Report of Case
A 47-year-old white man was in excellent health until November 1962, when he developed a severe, steady midback pain. A visit to a local hospital re- sulted in a course of diathermy and physiotherapy, without improvement. In early December he noted Received for publication Nov 9, 1964; accepted Nov 16.
Director, Oncology Division, Medical Services, Lemuel Shattuck Hospital, Department of Public Health, Commonwealth of Massachusetts; Assistant Professor of Medicine, Tufts University School of Medicine.
Reprint requests to Boston, Mass 02130.
dyspnea on exertion, cough, and right lower chest pain. He was admitted to a nearby hospital, where a chest x-ray revealed a right lower lobe infiltrate with effusion. Extensive investigation failed to reveal the nature of this infiltrate with the result that an ex- ploratory thoracotomy was performed on Jan 3, 1963. Carcinoma, with extensive involvement of the right lung and with pleural seeding, was uncovered. Patho- logically, the process was interpreted as a metastatic malignancy, possibly originating in the stomach or the pancreas.
The patient was discharged after surgery, but con- tinued to do poorly, suffering from anorexia, nausea, and pain near the right scapula and in the mid- abdomen. Loose, and frequently watery stools devel- oped, but neither melena nor foul odor was noted. His cough increased in severity and became productive of clear mucus. He continued to have dyspnea on exer- tion. In the month prior to his entry to the Lemuel Shattuck Hospital, he lost 20 1b (9,072 gm) of weight.
During the week prior to his Shattuck admission, he noted the sudden onset of a skin rash on his face and upper trunk. Swelling of fingertips and pain in the wrists, ankles, and knees occurred simultaneously. The knees became red and swollen, and bilateral foot swell- ing appeared. The joint pain spontaneously subsided on the day of admission to the hospital. Pain became more prominent in the abdomen and back.
He was admitted on Feb 19, 1963, to the Lemuel Shattuck Hospital, at which time he presented as an acutely and chronically ill man. The oral temperature was 100.4 F (38.0℃), his respirations were 24 per minute, and blood pressure was 130/65 without pulsus paradoxus. The pulse was 120 and regular. The skin was warm and moist. There were hundreds of spider angiomata on the trunk, the back, and on the face. His eyes were prominent but without stare or lid-lag. Several enlarged cervical and axillary nodes were pres- ent. The neck veins were not distended. The thyroid gland was not palpable. Examination of the lungs re- vealed dullness over the entire right chest. Cardiac examination was not abnormal except for sinus tachy- cardia. On abdominal examination a large, ill-defined, firm, nontender mass was present in the midepigas- trium, extending into the right upper quadrant. The
Arch Intern Med-Vol 115, April, 1965
liver edge was not palpable through this mass. Ascites was not present. Examination of the extremities re- vealed marked clubbing of fingers and toes. There was no joint swelling or redness, although tenderness was present over the radius and tibia and on motion of the wrists and knees. There was 2++ pretibial edema. The testicles were considered small. There was no voice change, loss of hair, or gynecomastia. Neuro- logic examination revealed no abnormalities.
Laboratory tests revealed: urine essentially nega- tive; hemoglobin, 9.5 gm%-11.5 gm%; hematocrit, 34 gm%-38 gm%; erythrocyte sedimentation rate, 56 mm per hour ; white blood cell count, 10,400 with a normal differential; blood urea nitrogen, 7.6 mg%; two-hour postprandial blood sugar, 186 mg% and 200 mg%; car- bon dioxide, 27 mEq per liter; chlorides, 90-99 mEq per liter ; sodium, 135 mEq per liter ; potassium, 4.2 mEq per liter ; calcium, 9.3 mg% ; phosphorus, 3.1 mg%; al- kaline phosphatase, 5.3 Bessie-Lourie units; bilirubin, 0.2 mg%; cephalin flocculation, negative; thymol tur- bidity, 0.7 units; prothrombin time, 15 seconds (76%) ; total protein, 6.1 gm%, with an albumin of 2.8 gm and globulin of 3.3 gm, cholesterol, 218 mg%; serum glutamic oxaloacetic transaminase, 22 units; lactic acid dehydrogenase, 90 Wacker units. Protein-bound iodine was 5.4 mg% (repeated twice) ; radiactive iodine uptake, 20% and 15.8% on two occasions. Twenty-four hour urine showed: 17-ketosteroid level, 6.8 mg, 17-hydroxycorticoids, 11.1 mg; catecholamines, 105µg. A stool specimen gave a negative guaiac test, and blood cultures were negative. Urinary estrogen deter- mination performed by the method of Barlow7 in the laboratory of Dr. Kendall Emerson at the Harvard Medical School revealed total esterone, 83.35µg; es- tradiol-178, 5.49ug; and estriol, 43.90µg per 24-hour urine. (Approximate normal male values are : estrone, 4.5μg; estradiol-17β, 2.5μg; estriol, 5.5μg per 24-hour urine.)
Electrocardiogram showed nonspecific ST-T wave changes with sinus tachycardia. Roentgenographic studies of the chest revealed the right hemithorax to be smaller than the left. Roentgenograms of the bones showed no osseous metastasis. Long bone films re- vealed that the periosteum of the femoral shafts and distal one inch of the radius and ulna was thickened and elevated, consistent with hypertrophic pulmonary osteoarthropathy; the hands and the wrists showed unremarkable joints. Intravenous pyelographic study showed deflection of left kidney laterally. Gastrointes- tinal series revealed increasing extrinsic pressure on duodenal bulb with derangement of mucosal pattern. A mass was present, continuous with the liver, press- ing the stomach interiorly. Skull films demonstrated a normal appearing sella turcica.
The patient did poorly during the next month. Bronchopneumonia then developed, and death occurred.
Pathology .*- Postmortem examination revealed a widespread cancer involving the right and left lungs, the visceral and parietal pleura, the pericardium,
* Autopsy performed by George W. Curtis, MD, Lemuel Shattuck Hospital.
mediastinal lymph nodes, peripancreatic nodes, and trachea. The primary was felt to reside in the liver. Sections revealed a hemangioma in which there was gradual transition of the epithelial lining cells to large tumor cells filling the vascular spaces, suggesting a primary angiosarcoma. The multiple metastatic lesions were characterized by markedly pleomorphic large cells, surrounded by fibrous tissue, and gave the suggestion of lining vascular spaces, although no evidence of contained blood elements was found.
In addition, there were foci of microabscesses in the liver with acute inflammatory cells. More than 50% of the liver was free from tumor, fibrosis, or in- flammation.
The adrenal glands were hypertrophied and to- gether weighed 18 gm. On section, there was a widened zona fasciculata. The testicles were small, measuring 1.5 cm in diameter.
Immediate cause of death appeared to be an embolus in the main pulmonary artery. Bilateral bronchopneu- monia was present, and Staphylococcus aureus, coagu- lase positive, was grown post mortem.
Comment
Although excessive serum and urine levels of adrenocortical hormones have been reported as- sociated with nonhormonal tumors, estrogens have been infrequently searched for. Ginsburg and Brown 8 investigated urinary estrogen ex- cretion in ten patients with bronchogenic car- cinoma demonstrating hypertrophic pulmonary osteoarthropathy with radiographic evidence of periostitis and found that total estrogens were elevated. In one case, the level was approxi- mately five times normal. Patients with bron- chogenic carcinoma, but without periostitis, had normal excretory levels. Elevated estrogens were not found to be associated with gyneco- mastia, nor with digital clubbing in the absence of periostitis. Hardy, in 1960, described three cases of gynecomastia associated with lung can- cer, in one of whom “elevated” urinary estro- gens were found.9 However, neither laboratory techniques nor the level of estrogen is given, and therefore the term “elevated” is unclear. Men- tion of periostitis is not made. Periostitis with clinical arthropathy was a feature of the case reported in this paper; gynecomastia was not present.
Utilizing the bioassay method of Allen and Doisy (changes in vaginal smear of castrated mice), Rupp et al measured urinary estrogen in four patients with metastatic cancer to the liver and found elevated levels in two.10
The relationship of spider angiomata to estro- gen metabolism has been the subject of much investigation and speculation.12 The frequent finding of spider nevi in normal pregnancy and the association of these vascular lesions with exogenously administered estrogen have been given as evidence of a causal relationship. Pa- tients with angiomata and cirrhosis have been studied for evidence of deranged estrogen me- tabolism. Cameron 11 studied the urinary excre- tion of estrogenic substances in 12 cirrhotics by the Brown biochemical technique and found ele- vated estrogens in two. No correlation between the presence of spider angiomata and elevated estrogens was noted. Previous investigators, however, using bioassay techniques, concluded that estrogen excretion was raised in about half of the cirrhotic population studied,13,14 although a clear-cut relation to spider angiomata was not derived. Shaver and colleagues 15 observed that clearance of isotopically labeled estradiol from the plasma in patients with cirrhosis was de- layed, implicating faulty liver metabolism for estrogens, and, subsequently, a possible etiology for phenomenon as spider angiomata. A recent paper by Brown and colleagues 21 demonstrated increased urinary estrogen excretion in 16 men and women with liver disease. These authors concluded that the augmented output was due to increased secretion rates rather than to impaired metabolism. No direct connection between spider nevi and elevated estrogen levels could be dem- onstrated.
In the present case, a sudden eruption of mul- tiple spider angiomata occurred simultaneously to the appearance of joint pains. The latter were in association with periostitis, as demonstrated by x-rays. These two events would appear causally related to the subsequently discovered elevated urinary estrogens. If so, two physio- logic aberrations, namely spider angiomata and hypertrophic osteoarthropathy, seen conjoined to the abnormal estrogen metabolism in the same patient.
The elevated urinary estrogen level most likely represents excess production of the hor- mones. The presence of a normal blood urea nitrogen, serum electrolytes, and pyelographic function of the kidneys imply the absence of a renal abnormality to account for increased clear- ance of estrogen. The possibility of impaired hepatic metabolism seems diminished by the
pathologic demonstration of more than 50% normal liver tissue, as well as virtually normal liver function tests during life. Abnormal utili- zation of the estrogen by body tissue may have played a role, but there is no evidence that this occurs in the body.
The site of estrogen production in this case probably was the adrenals. The hyperplasia of these glands, with thickened cortices, suggests biochemical productivity. The adrenals appear to be quite capable of elaborating specific estro- genic substances in abundance.23,24 Estrogens have been identified in adrenal tissue by isolation techniques.26 Increased excretion of estrogens has been reported in patients suffering from adrenocortical hyperactivity.16,17 Large quan- tities of urinary estradiol-170 and estriol have been isolated from the urine of patients with adrenal cancer.18,19 In addition, the administra- tion of corticotropin (adrenocorticotropin) to oophorectomized women results in increased urinary estrogens, implicating the adrenal gland as the source.17,20.22,25 Several reports have noted, however, that the urinary estrogen re- sponse after corticotropin (ACTH) adminis- tration to oophorectomized subjects consists principally of estriol.22,25
In this case, the estrogen production may have been a singular event of adrenocortical overac- tivity, inasmuch as urinary levels of 17-keto- steroids and 17-hydroxycorticoids were normal. This finding parallels the report of Ginsburg and Brown,8 who reported normal urinary levels of the latter two steroids in patients with ele- vated estrogens. Although estrogens may be produced by the testes, the presence of small testicles would tend to disfavor these organs as the source of significant estrogen contribution in this patient.
Conceivably, the estrogens in the present case could have been produced directly by the tumor. Unfortunately, a postmortem examination of the tumor (and of the adrenals) for estrogens was not performed. (Such examination may not have distinguished between production and sequestration). More likely a specific estrogen- stimulating protein was elaborated by the tumor. That this theoretical protein was not a classical corticotropin is in evidence from the normal urinary levels of 17-ketosteroids and 17-hy- droxycorticoids. The pituitary gland was not available for examination, but skull films showed
Arch Intern Med-Vol 115, April, 1965
a normal sella turcica on several occasions in the month prior to death.
The distribution of the urinary estrogens is of interest in this case. Estrone levels were ap- proximately twice as high as the estriol. The latter compound is felt to be the least active bio- logically of the three classical estrogens and to represent an irreversible end product of metabo- lism of estrone and estradiol-173.27 The high ratio of estrone (second in biological activity by the Allan and Doisy assay) may account for the formation of the spider angiomata and the periostitis. However, it is certainly possible that an unrecognized estrogen, produced in abundance but not recognized by the testing method, was the responsible steroid.
Summary
A significant elevation of urinary total estro- gen level was found in a 47-year-old man with a highly anaplastic, widespread cancer, sugges- tive of angiosarcoma. The levels of the three estrogens were: estrone, 83.35µg; estradiol, 5.49µg; and estriol, 43.90µg. The adrenal gland is implicated as the source of these hormones. The patient’s clinical course was marked by the sudden appearance of a multitude of spider an- giomata, marked periostitis of long bones, digi- tal clubbing, and arthropathy. An association between the clinical findings and estrogen hyper- production appears likely.
Supported in part by Grant CA 07190-01 from the National Cancer Institute of National Institutes of Health, Public Health Service, Bethesda, Md.
Generic and Trade Names of Drugs Corticotropin-Acth, Acthar, Corticotropin.
REFERENCES
1. McDaniel, H. G., et al : Adrenal Cortical Hyper- function Associated With Bronchogenic Carcinoma, Amer J Med 35:427, 1963.
2. Prunty, F. T., et al: Adrenocortical Hyperfunc- tion and Potassium Metabolism in Patients With “Non- endocrine” Tumors and Cushing’s Syndrome, J Clin Endocr 23:737, 1963.
3. Werk, E. E., Jr .; Sholiton, L. J .; and Marnell, R. T .: Further Studies of Adrenocortical Function in Patients With Carcinoma of Lung, Amer J Med 34: 192, 1963.
4. Meador, C. K., et al: Cause of Cushing’s Syn- drome in Patients With Tumors Arising From “Non- endocrine Tissue,” J Clin Endocr 22:693, 1962.
5. Caranasos, G., and Reubner, B. H. : Adrenal Width and Metastasis in Bronchogenic Carcinoma, Arch Path 76:263, 1963.
6. Sholiton, L. J .; Incze, J. S .; and Werk, E., Jr .: Adrenocortical Width in Carcinoma of Lung, Cancer 14:105, 1961.
7. Barlow, J. J .: Sensitive Method of High Specifi- city for Determination of Urinary Estrogens, Anal Bio- chem 6:435, 1963.
8. Ginsburg, J., and Brown, J. B. : Increased Estrogen Excretion in Hypertrophic Pulmonary Osteoarthro- pathy Lancet 2:1274, 1961.
9. Hardy, J. D .: Gynecomastia Associated With Lung Cancer, JAMA 173:1462, 1960.
10. Rupp, J., et al : Hormone Excretion in Liver Dis- ease and in Gynecomastia, J Clin Endocr 2:688, 1951.
11. Cameron, C. B .: Urinary Excretion of Oester- one, Oestradiol-173 and Oestriol in Patients With Chronic Liver Damage, J Endocr 15:199, 1957.
12. Bean, W. B .: Arterial Spider and Similar Le- sions of Skin and Mucus Membrane, Circulation 8: 117, 1953.
13. Edmondson, H. A .; Glass, S. J .; and Soll, S. N .: Gynecomastia Associated With Cirrhosis of Liver, Proc Soc Exp Biol Med 42:97, 1939.
14. Dohan, F. C., et al: Hormone Excretion in Liver Disease, J Clin Invest 31 :481, 1952.
15. Shaver, J. C .; Roginsky, M. S .; and Christy, N. P .: Clearance From Plasma of Isotopically Labelled Oestradiol in Patients With Hepatic Cirrhosis, Lancet 2:335, 1963.
16. Eberlein, W. R .; Bongiovanni, A. M .; and Fran- cis, C. M .: Simplified Method for Routine Measure- ment of Urinary Estriol, J Clin Endocr 18:1274, 1958.
17. Soffer, L. J .; Dorfman, R. I .; and Gabrilove, J. L .: Human Adrenal Gland, Philadelphia : Lea & Febiger, 1961.
18. Mason, H. L., and Kepler, E. J .: Isolation of Steroids From Urine of Patients With Adrenal Corti- cal Tumors and Adrenal Cortical Hyperplasia : New 17-Ketosteroid, Androstane-3(a), 11-Diol-17-One, J Biol Chem 161:235, 1945.
19. Sandberg, H., et al: Estrogen Excretion in Ovariectomized Women Receiving Adrenocorticotro- pin, J Clin Endocr 18:1268, 1958.
20. West, C. D .; Damast, B .; and Pearson, O. H .: Adrenal Estrogens in Patients With Metastatic Breast Cancer, J Clin Invest 37:341, 1958.
21. Brown, J. B .; Crean, G. P .; and Ginsburg, J .: Oestrogen Metabolism and Excretion in Liver Dis- ease, Gut 5:56, 1964.
22. Brown, J. B .; Falconer, C. W. A .; and Strong, J. A .: Urinary Oestrogens of Adrenal Origin in Women With Mammary Cancer : Effect of Cortisone Treatment, Proc Soc Exp Biol Med 85:264, 1954.
24. Bulbrook, R. D., et al : Adrenalectomy in Breast Cancer: Attempt to Correlate Clinical Results With Oestrogen Production, Brit Med J 2:12, 1958.
25. Barlow, J. J .: Adrenocortical Influences on Es- trogen Metabolism in Normal Females, J Clin Endocr 24:586, 1964.
26. Beall, D .: Isolation of Oestrone From Adrenal Gland, Nature 144:76, 1939.
27. Brown, J. B .: Determination and Significance of Natural Estrogens, Advances Clin Chem 3:157, 1960.
Arch Intern Med-Vol 115, April. 1965