GAS-LIQUID CHROMATOGRAPHIC STUDIES OF URINARY 17-KETOSTEROIDS IN THREE PATIENTS WITH ADRENOCORTICOTROPIN-PRODUCING BRONCHOGENIC CARCINOMA
LEO P. CAWLEY, M.D., HARVEY A. TRETBAR, M.D., AND BILLY O. MUSSER, B.S., C (ASCP) Divisions of Hormone Chemistry and Endocrinology, Wesley Medical Research Foundation, Wesley Medical Center, Wichita, Kansas 67214
During the course of a gas-liquid chro- matographic (GLC) investigation of the effects of HCI hydrolysis on urinary 17- ketosteroids (17-KS),4 one urine sample gave a GLC pattern distinct from any other. The patient (L. M.) had a small cell carcinoma of the right lung and clini- cally had Cushing’s syndrome. The tumor was subsequently shown to contain adreno- corticotropin (ACTH).22 The original GLC studies were done without benefit of de- rivatives on SE-30 columns. By this method androsterone (A) and dehydroepiandros- terone (DHA) cannot be separated and are recorded as a single peak on the chro- matogram.10, 22 The relationship of etio- cholanolone (E) to the A and DHA peak in the patient with an ACTH-producing tumor deviated from the normal in that very little A and DHA were present.
Urine from two additional cases has been examined and shows essentially the same GLC pattern. To determine more completely the relationship of the individual urinary 17-KS to the normal metabolic steroid pattern, the steroids as trimethylsilyl (TMSi) ethers were separated and quanti- tated by GLC. In a separate study the urinary 17-KS were separated by thin- layer chromatography (TLC) and then each 17-KS zone was analyzed by GLC. Pre- liminary results of this work have been reported.6
MATERIALS AND METHODS
Urinary 17-ketogenic steroids (17-KGS) and 17-KS were determined by the method of Sobel and associates.2 The urine samples were analyzed for 17-KS by GLC accord- ing to a previously described method.5
Supported in part by a grant from the Kansas Division of the American Cancer Society.
Briefly, a 10-ml. aliquot of 24-hr. urine sample in a 40-ml. glass-stoppered, round bottom test tube was buffered to pH 4.8 by adding 12 ml. of 0.15 M acetate buffer (pH 4.8). The steroid conjugates were hydrolyzed by sulfatase and 8-glucuronidase. The urine was incubated with the enzymes with constant agitation for 16 hr. at 37 C. and extracted with dichloromethane. The extract was washed to neutrality with two aliquots of 0.1 N NaOH. An aliquot of the extract equivalent to S ml. of urine was dried to dryness under nitrogen. The TMSi ethers were made directly from this material for GLC or the extract was used for TLC.
Thin-layer chromatography. The dried extracts were dissolved in chloroform and applied across 1 to 2 cm. on a 250-mu silica gel plate. A mixture of 17-KS (seven 17-KS, see under “Gas-Liquid Chromatog- raphy”) was applied adjacent to the urine extract. The chromatogram was developed for 45 min. in a solvent system of benzene- ethyl acetate (40:60). The steroids were localized with dichlorofluorescein (0.1% in ethanol) and ultraviolet light. Three fractions were selected as outlined by Kirsch- ner and Lipsett.13 Fraction I contained A, E, and DHA; Fraction II, the 11-oxy-17-KS; and Fraction III, pregnanetriol. Each zone was scraped off the plates into 15-ml. centrifuge tubes and extracted two times with acetone. The pooled extracts were dried under a stream of nitrogen.
Trimethylsilyl ethers. The dried extracts were dissolved in 2.0 ml. of acetone to which were added 0.25 ml. of hexamethyldisilazane and 3 drops of trimethylchlorosilane. The vials were capped and placed in a 55 C. block bath for 30 min. The white precipitate was spun down by centrifugation, after which the supernatant was removed and
dried to dryness in a separate test tube. The TMSi ethers were taken in acetone.
Gas-liquid chromatography. A Model 5000 Barber-Coleman chromatograph with a flame ionization detector was used. The 6-ft. glass U-shaped column with 4-mm. i.d. was packed with 2% neopentyl glycol succinate on Gas Chrom P. All glassware and supports were siliconized. The flash heater was oper- ated at 250 C. and the column at 215 C. TMSi derivatives of standards removed from TLC plates were analyzed by GLC. The standards were A, DHA, E, and 11-keto- etiocholanolone (11-KE), 11-ketoandros- terone (11-KA), 11-hydroxyetiocholanolone (11-OHE), and 11-hydroxyandrosterone (11- OHA). Area was determined with an auto- matic integrator .* Calibration curves with each standard showed linear relationship to concentration from 0.01 ug. to 2 ug. per ml. Pregnanediol and pregnanetriol were also analyzed and the calibration curve was linear from 0.01 ug. to 2 ug. per ml.
REPORT OF CASES
Case 1. L. M., a 69-year-old female, was admitted with polydipsia, polyuria, dyspnea, edema, recent 10-lb. weight gain, acneiform rash on arms and chest, and profound weakness. Seven years earlier she had had a left radical mastectomy for carcinoma. Laboratory tests were normal except for blood sugar (364 mg. per 100 ml.), potassium (2.4 mEq. per l.), CO2 (49.4 mEq. per 1.), chloride (89 mEq. per 1.), and sodium (144 mEq. per l.). The 17-KS value was 23.4 mg. per 24 hr. A 17-KS value of 35 mg. per 24 hr. was ob- tained several days later by GLC and this urine was used in this study. Dexa- methasone failed to suppress the adrenal glands. Shortly before death the 17-KGS value was 152 and 17-KS value was 110 mg. per 24 hr. At autopsy, a small cell carcinoma was found in the upper pole of the right lung, with metastasis to mediastinal lymph nodes. Together the adrenal glands weighed 40 Gm. and demon- strated hyperplasia. The pituitary gland was small, with a small cystic necrotic
area with recent hemorrhage in the pos- terior pituitary. The ACTH assay of the lung tumor was performed by Dr. Grant W. Liddlet who found 1.7 mu of ACTH per Gm. of tissue. The assay value for lymph node metastasis was 0.58 mu of ACTH per Gm. of tissue.22
Case 2. Urine from Patient L. S. was furnished by Dr. Grant W. Liddlet and is Case 2 reported by Meador and associates.19 The history and endocrinologic analysis are repeated here for continuity. According to Meador and associates, the patient, a 53-year-old male, was admitted with right chest pain, hemoptysis of 4 months’ duration, and a 20-lb. weight loss. He was found to have urinary steroid values for 17-hydroxycorticosteroids (17-OHCS), 17 to 20 mg. per 24 hr., and 17-KS of 13 mg. per 24 hr. (We found 17-KS of 15.5 mg. per 24 hr. by GLC.) Later the 17-OHCS rose to 40 to 91 mg. per 24 hr. A small cell tumor was found in the lung, with distant meta- stases to bone, liver, mediastinal lymph nodes, and right adrenal gland. The ACTH content of the tumor was 21 mu per Gm. of tissue and that of the metastatic tumor was 1.6 mu per Gm. of tissue.
Case 3 (B. D.). The urine in this case was furnished by Dr. Robert Brown of Kansas University Medical School. The patient had an oat cell carcinoma of the lung with Cushing’s syndrome. The 17-KS value was 400 mg. per 24 hr. The adrenal glands weighed 29 and 34 Gm. respectively, were free of metastases, and were hyper- plastic. Although the tumor was not as- sayed for ACTH-like substance, the clini- cal course and autopsy findings are con- sistent with an ACTH-producing tumor.
RESULTS
All three cases showed a decrease of A and DHA peak on SE-30 columns (Fig. 1). Compared with the normal, L. M. had no detectable A and DHA peak. Both B. D. and L. S. did show a small amount of A and DHA. In two of the cases, L. M. and L. S., where sufficient urine was available, the
* CRS-20, Infratronics Corp., Houston, Tex.
t Department of Medicine, Vanderbilt Univer- sity School of Medicine, Nashville, Tenn.
E
B.D.
II-OHE
11-KE
II-OHA
11-KA
11-OHE
E
11-KE
L.M.
11-OHA
11-KA
II-OHA
E
II-OHE
N
L. S.
11-KE
DHA
II-KA
E
L.M.
A
II-KE
11-OHE
DHA
II-OHA
A
E
L.S.
Il-OHA
DHA
11-OHE
11-KA
II-KE
A
E
DHA
STD.
II-KE.
II-KA
II-OHE
II-OHA
TLC FRACTIONS II
I
E
L.M.
II-KE
TI-OHA
HI-OHE
A A
DHA
JIKA
TRIOL
L.S.
E
A i
HI-OHA II-ONE
DHA
II-KE
TRIOL
STD.
II-KE
MER
E DHA
II-KA
11-OHE
II-OHA
TRIOL
N
| Steroid | Normalt | L. S. | L. M. | |||
|---|---|---|---|---|---|---|
| % | mg./24 hr. | % | mg./24 hr. | % | mg./24 lır. | |
| A | 30.6 | (3.09) | 11.3 | (1.75) | 3.6 | (1.26) |
| E | 30.6 | (3.09) | 18.0 | (2.79) | 19.3 | (6.76) |
| DHA | 7.7 | (0.78) | 0.2 | (0.03) | 0.1 | (0.04) |
| 11-KA | 4.8 | (0.49) | 2.0 | (0.31) | 1.0 | (0.35) |
| 11-KE | 8.8 | (0.89) | 5.0 | (0.78) | 15.0 | (5.27) |
| 11-OHA | 10.9 | (1.10) | 35.5 | (5.50) | 21.0 | (7.35) |
| 11-OHE | 6.6 | (0.67) | 28.0 | (4.34) | 40.0 | (14.0) |
| Total 17-KS | (10.1) | (15.5) | (35.0) | |||
| 11-Desoxy: 11-oxy ratio (%) | 79:21 | 30:70 | 22:78 | |||
* The percentage of distribution is the first value given, with the amount of individual 17-ketosteroids being shown in parentheses.
* Average of values for 20 males and 20 females. 5
combined analytic approach of TLC and GLC on the TMSi ethers gave clear evidence that both DHA and A were reduced (Figs. 2 and 3). The changes noted in the 11-oxy- 17-KS were variable but increased in all cases. In L. M. and B. D. 11-KE and 11- OHE were considerably elevated. In L. S. 11-OHA was markedly elevated (Fig. 1). This is brought out more clearly for L. S. and L. M. in Figures 2 and 3. DHA is de- creased to very low levels in both L. S. and L. M., particularly in L. M. The percentage of distribution and amount of individual 17-KS for normal, L. S., and B. D. is shown in Table 1. Pregnanediol and pregnanetriol were not elevated.
DISCUSSION
The urinary 17-KS pattern in all three cases showed a decreased concentration of A and DHA, although we cannot be certain of the amount of A or DHA in the case of B. D. because TMSi studies were not done. The E: A ratio was elevated above normal value of 1 in L. M. and L. S. In L. S. the concentration of A was slightly greater than in L. M. In both instances the amount of DHA was below normal (Table 1). In all cases the 11-oxy-17-KS were elevated. In L. M., 11-KE, 11-OHA, and 11-OHE were the major components. In L. S. the major components were 11-OHE and 11- OHA, and in B. D. 11-KE and 11-OHE were highest of the 11-oxy-17-KS.
Table 1 also shows a computation of the ratio of the percentage of 11-deoxy: 11-oxy for normal persons, L. S., and L. M. The normal is a composite of values from a previous study of 20 healthy females and 20 healthy males.5 In both L. S. and L. M. the amount of 11-deoxy steroids is not increased, but the 11-oxy steroids are markedly increased. The 11-deoxy: 11-oxy per cent ratios are virtually the reverse of normal (79:21) for L. S. (30:70) and L. M. (22:78). Even though L. M. had more than twice the total urinary 17-KS value of L. S., the amount of 11-deoxy steroids did not differ significantly. These patients were excreting essentially the usual amount of 11-deoxy-17-KS except for the extremely low amount of DHA. The pre- dominant abnormality, therefore, is an increase in the 11-oxy fraction of the 17-KS, and a decrease of DHA.
The amount of A secreted in the urine of L. M. and L. S., and probably B. D., is slightly less than normal; the amount of E excreted is within the normal range5 except for L. M. who did have a slight excess (Table 1). Alterations of the E:A ratio are known to occur in malignancies involving the adrenal cortex,11, 23 and re- cent evidence suggests that a decrease of A is not an unusual situation in certain other types of carcinoma, particularly of the pros- tate and breast.21 The precursors of the urinary 11-deoxy-17-KS’s are DHA, an-
A
E
P
D
T
KA
KE
HA
HE
A
P
E
D
T
KA
HE
KE
HA
drostenedione, testosterone, and desoxy- cortisol.7, 9 Both DHA and testosterone are converted to androstenedione, which is converted into equal portions of A and E by 44-reductase (5x and 58) of the liver.1, 2 These enzymes are susceptible to various precursors and hormonal influences and the E: A ratio may be shifted as a result of these influences. Exogenous DHA sup- presses 5a-44-reductase, and a rise in E develops much as in malignancy of adrenals with excess DHA formation.14 Triiodo- thyronine stimulates 5a-44-reductase, re- sulting in an increase of A.15 Deoxycortisol and other corticoids are metabolized to E.17, 23 The increase of E:A ratio in the cases of this report suggests that some metabolites are influencing the activity of the 5a or 53-44-reductases, and in addi- tion some corticoids may be undergoing conversion.
The increase of the 11-oxy-17-KS steroid is an indication of adrenal hyperfunction. This is of some interest since DHA, also of adrenal origin, is deficient in all three cases. The very low DHA values in these cases are in contrast to what one might expect from ACTH stimulation or in the usual case of Cushing’s syndrome caused by bilateral adrenal hyperplasia (Fig. 4). In ACTH stimulation in a healthy adult, GLC studies showed a 5-fold increase of DHA.5 In some cases of Cushing’s syndrome produced by adrenal carcinoma, DHA is excreted in large amounts. The increase of DHA in adrenal carcinoma is probably due to a deficiency of adrenal 45-3a-hy- droxysteroid dehydrogenase.3, 8
Low excretion of DHA, despite stimula- tion of adrenal cortex by ACTH from the lung carcinoma in these patients, ap- pears paradoxical. The conversion of 45- pregnenolone through 17a-OH pregnenolone to DHA appears not to be as active as the pathway from 45-pregnenolone through progesterone to cortisol. The urinary pat- tern of 17-KS found in these three patients can best be explained on the basis that the great majority of the substrate, 45-preg- nenolone, was converted to corticoids. The corticoids in turn were metabolized to 11-oxy-17-KS, A, and E. Little, if any,
DHA would be formed by this mechanism.7 Proof that this type of metabolic alteration occurs in patients with ACTH-producing tumors12, 16, 18, 19 could be established by GLC examination of the products formed from the action of the patients’ adrenal enzymes on the substrate, 45-pregnenolone.
SUMMARY
A gas-liquid chromatographic study of the urinary 17-ketosteroids (17-KS) of three cases of ACTH producing small cell carcinoma of the lung showed an increase of 11-oxy-17-KS, an elevation of etiocholano- lone (E), and a decrease of androsterone (A) or dehydroepiandrosterone (DHA), or both. Two of the cases were studied with trimethylsilyl ether derivatives of steroids and revealed that both androsterone and dehydroepiandrosterone were decreased. The 17-KS excreted in these cases is different from reported studies of Cushing’s syndrome due to adrenal hyperplasia and adrenal neoplasia. The predominance of etiocholano- lone over androsterone suggests a disturb- ance of liver A4-reductase in these cases; however, the best explanation of the findings of the urinary 17-KS pattern is that 45- pregnenolone was preferentially converted to corticoids with little or no conversion to DHA. The corticoids in turn were me- tabolized to 11-oxy-17-KS, A, and E. Proof that patients with ACTH-producing tumors have a unique alteration of steroid metabo- lism will require specific enzymatic studies of their adrenal glands.
Additionally the ACTH from the tumor may be functionally distinct from the natu- ral ACTH and thereby capable of “turning on” an alternate metabolic pathway for the synthesis of corticoids.
REFERENCES
1. Baulieu, E .- E., and Mauvais-Jarvis, P .: Studies on testosterone metabolism. I. Conversion of testosterone-17a-3H to 5a- and 56-androstane-3x, 178-diol-17a-3H: a new “178-hydroxyl pathway.” J. Biol. Chem., 239: 1569-1577, 1964.
2. Baulieu, E .- E., and Mauvais-Jarvis, P .: Studies on testosterone metabolism. II. Metabolism of testosterone-4-14C and androst-4-ene-3, 17-dione-1,2-3H. J. Biol. Chem., 239: 1578-1584, 1964.
3. Bongiovanni, A. M .: Digitonin precipitable
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
steroids in normal human urine. J. Clin. Endocrinol., 26: 1240-1244, 1966.
4. Cawley, L. P., Musser, B. O., Faucette, W., Beckloff, S., and Learned, H .: Evaluation of hydrolytic products of 17-ketosteroids by means of gas-liquid chromatography. Clin. Chem., 11: 1009-1018, 1965.
5. Cawley, L. P., Musser, B. O., and Tretbar, H. A .: Gas-liquid chromatography of uri- nary 17-ketosteroids, pregnanediol, and pregnanetriol in normal individuals. Am. J. Clin. Path., 48: 216-224, 1967.
6. Cawley, L. P., Tretbar, H. A., and Musser, B. O .: Gas-liquid chromatographic studies of urinary 17-ketosteroids in three patients with bronchogenic carcinoma producing adrenocorticotrophic hormone. Am. J. Clin. Path., 42: 532, 1964 (abstract).
7. Dorfman, R. I., and Sharma, D. C .: An out- line of the biosynthesis of corticosteroids and androgens. Steroids, 6: 229-236, 1965.
8. Goldman, A. S., Bongiovanni, A. M., Yako- vac, W. C., and Prader, A .: Study of 4538- hydroxysteroid dehydrogenase in normal, hyperplastic and neoplastic adrenal cortical tissue. J. Clin. Endocrinol., 24: 894-909, 1964.
9. Heftmann, E., and Mosettig, E .: Biochemistry of Steroids. New York: Reinhold Publish- ing Corp., 1960, pp. 117-124, 138-146.
10. Horning, E. C., Van den Heuvel, W. J. A., and Creech, B. G .: Separation and determi- nation of steroids by gas chromatography. Methods Biochem. Anal., 11: 69-147, 1963.
11. Jailer, J. W., Vande Wiele, R., Christy, N. P., and Lieberman, S .: Studies in Cushing’s syndrome. III. Urinary 17-ketosteroids in patients with bilateral adrenal cortical hyperplasia. J. Clin. Invest., 38: 357-365, 1959.
12. Jarett, L., Lacy, P. E., and Kipnis, D. M .: Characterization by immunofluorescence of an ACTH-like substance in nonpituitary tumors from patients with hyperadrenocor- ticism. J. Clin. Endocrinol., 24: 543-549, 1964.
13. Kirschner, M. A., and Lipsett, M. B .: The analysis of urinary steroids using gas-liquid chromatography. Steroids, 8: 277-294, 1964.
14. Kirschner, M. A., and Lipsett, M. B .: The effect of exogenous precursors on etiocho-
Janolone/androsterone ratios in man. Acta endocrinol., 46: 207-216, 1964 (abstract).
15. Kirschner, M. A., Lipsett, M. B., and Wilson, H .: Metabolism of exogenous dehydroepi- androsterone in man. Acta endocrinol., 43: 387-398, 1963.
16. Liddle, G. W., Island, D. P., Ney, R. L., Nicholson, W. E., and Shimizu, N .: Non- pituitary neoplasms and Cushing’s syn- drome. Ectopic “adrenocorticotropin” pro- duced by nonpituitary neoplasms as a cause of Cushing’s syndrome. Arch. Int. Med., 111: 471-475, 1963.
17. MacDonald, P. C., Vande Wiele, R. L., and Lieberman, S .: Precursors of the urinary 11-desoxy-17-ketosteroids: influence of the level of secretion of dehydroisoandrosterone upon its metabolism. J. Clin. Endocrinol., 22: 1222-1228, 1962.
18. Marks, L. J., Rosenbaum, D. L., and Russ- field, A. B .: Cushing’s syndrome and cortico- tropin-secreting carcinoma of the lung. Ann. Int. Med., 58: 143-149, 1963.
19. Meador, C. K., Liddle, G. W., Island, D. P., Nicholson, W. E., Lucas, C. P., Nuckton, J. G., and Luetscher, J. A .: Cause of Cush- ing’s syndrome in patients with tumors arising from “nonendocrine” tissue. J. Clin. Endocrinol., 22: 693-703, 1962.
20. Sobel, C., Golub, O. J., Henry, R. J., Jacobs, S. L., and Basu, G. K .: Study of the Norym- berski methods for determination of 17-keto- genic steroids (17-hydroxycorticosteroids) in urine. J. Clin. Endocrinol., 18: 208-221, 1958.
21. Stern, E., Hopkins, C. E., Weiner, J. M., and Marmorston, J .: Hormone excretion patterns in breast and prostate cancer are abnormal. Science, 145: 716-719, 1964.
22. Tretbar, H. A., and Cawley, L. P .: Cushing’s syndrome: gas-liquid chromatographic studies of urinary 17-ketosteroids of a patient with probable ACTH producing bronchogenic carcinoma: a preliminary report. J. Kansas Med. Soc., 64: 487-490, 1963.
23. Wilson, H., and Schenker, S .: Effect of corti- costeroids on urinary 58 and 5& C19 steroids in man. Acta endocrinol., 46: 197-206, 1964.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016