Serum dehydroepiandrosterone sulfate and the use of clomiphene citrate in anovulatory women

Vol. 43, No.2, February 1985 Printed in U.8A.

FERTILITY AND STERILITY Copyright < 1985 The American Fertility Society

Serum dehydroepiandrosterone sulfate and the use of clomiphene citrate in anovulatory women

David Hoffman, M.D. Rogerio A. Lobo, M.D.* University of Southern California School of Medicine, Los Angeles, California

Serum dehydroepiandrosterone sulfate (DHEA-S) is frequently elevated in anovulatory women. This study was carried out to determine whethe.r the ovulatory response with clomiphene citrate (CC) in patients with elevated levels of serum DHEA-S is influenced by the pretreatment level. Also evaluated was whether this response rate was Rimilar to or different from that of anovulatory patients who had normal levels of DHEA-S. CC was administered to 40 anovulatory patients who had elevated levels of DHEA-S. Rankit analysis of these 40 elevated DHEA-S levels indicated that two populations existed. These patients were, therefore, divided into two groups of 29 and 11 with DHEA-S levels of < 5 and> 5 tJ..glml, respectively. Fiftynine anovulatory patients with normal DHEA-S levels were also treated with CC. Patients with elevated and normal DHEA-S levels had similar rates of ovulation with CC (75% and 78%). Among patients with elevated levels of DHEA-S, ovulation occurred in 55% of patients with levels> 5 tJ..glml and 83% with levels < 5 Mimi. The dose of CC at which ovulation occurred was unrelated to the level of DHEA-S. Pregnancies occurred in 15 of the 40 patients after at least four ovulatory cycles and were not influenced by the level of DHEA-S. It is concluded that CC is effective in inducing ovulation in patients with elevated levels of adrenal androgens. However, in patients with DHEA-S levels> 5 tJ..glml, the ovulatory response rate may be decreased. Fertil SteriI43:196, 1985

Serum dehydroepiandrosterone sulfate (DHEA-S) has been used as an indicator of adrenal androgen secretion. I, 2 We have recently shown that up to 50% of anovulatory women have elevated levels of DHEA-S, suggesting that increased adrenal androgen secretion occurs frequently in anovulatory women. 3 It has been reported that glucocorticoids may be effective in the treatment of anovulatory women. 4 , 5 Some patients with elevated levels of DHEA-S who fail to respond with clomiphene citReceived August 1, 1984; revised and accepted October 24, 1984. *Reprint requests: Rogerio A. Lobo, M.D., Women's Hospital, Room L913, 1240 North Mission Road, Los Angeles, California 90033. 196

Hoffman and Lobo DHEA-S and clomiphene

rate (CC) may respond with the adjunctive administration of dexamethasone (DEX).6 A recent report also has suggested that DEX (0.5 mg daily) prescribed in conjunction with CC results in higher ovulation and pregnancy rates than the use of CC alone. 7 However, CC alone has been extremely effective in the treatment of anovulation even in women having elevated levels of serum androgens. Successful ovulation has occurred in up to 95% of patients. 8 , 9 In addition, among patients ovulating with CC, the level of DHEA-S does not correlate with the dose of CC required for ovulation to occur.9 Even after DEX suppression in some patients with increased adrenal androgen levels, CC may be necessary for ovulation to occur.lO For these reasons, we wished to assess Fertility and Sterility

whether the pretreatment level of serum DHEAS influences the ovulation rate with CC. In addition, we compared the CC response and pregnancy rates of anovulatory patients who have elevated levels of DHEA-S with the responses of other anovulatory patients who have normal levels of DHEA-S. MATERIALS AND METHODS

One hundred nineteen consecutive euprolactinemic anovulatory patients requiring CC for the induction of ovulation in our Endocrine-Infertility Clinic were studied. All these patients had withdrawal bleeding in. response to the administration of intramuscular progesterone (PHn-oil, suggesting normal status. It was previously determined that 60 of these patients had elevated levels of DHEA-S.3 Forty of these patients were followed for 1 year to determine their responses to CC and their pregnancy rates. These 40 patients, 19 to 39 years of age, did not have thyroid disease, cortisol excess, hyperprolactinemia, or evidence of virilization. None of these patients had evidence of an adrenal neoplasm and the diagnosis of congenital adrenal hyperplasia had been excluded by the measurement of 17-hydroxyprogesterone. l l Twenty of these 40 patients met our clinical criteria for the diagnosis of polycystic ovary syndrome. These patients had perimenarchial onset of chronic anovulation, increased body weight, hirsutism, and elevated levels of luteinizing hormone (LH) (> 20 mIUlml) and an LH/follicle-stimulating hormone (FSH) ratio of> 3. Blood was obtained from each patient prior to treatment for the measurements of testosterone (T) and DHEA-S. The values of two samples from each patient were averaged for a baseline value. These levels were compared with those of 20 ovulating, non hirsute women studied during the early follicular phase, days 2 to 7. CC was administered in a standardized regimen of increasing doses from 50 to 250 mg daily for 5 days with the addition of human chorionic gonadotropin (hCG).8, 9 Presumptive ovulation was defined by a serum P level of > 10 ng/ml obtained 2 weeks after the last CC dose. Women who received all doses ofCC up to 250 mg daily for 5 days, with the addition ofhCG, and who had insufficient P levels were considered resistant to CC. Once ovulatory, all patients were followed for at least four cycles to determine pregnancy rates. Semen analyses, Vol. 43, No.2, February 1985

postcoital tests, and hysterosalpingography were normal in these patients who were evaluated. Sera were separated and stored at - 200G until analyzed for T, DHEA-S, and P by established radioimmunoassays.12-14 Statistical analyses were carried out with Student's t-test, chi-square analysis with Fisher's exact correction, rankit analysis,15 as well as regression analysis by the method of least squares.

RESULTS

The mean ± standard error (SE) T and DHEAS values of the 40 patients were significantly higher than those of ovulatory control subjects (Table 1). The upper limit of the normal range for serum DHEA-S is 2.8 fJ;g/m1.1 The elevated DHEA-S values in these 40 women fell into two populations as determined by rankit analysis (Fig. 1). It was determined from these data that serum values of DHEA-S of < 5 fJ;g/ml and;;. 5 fJ;g/ml constituted two different populations. For this reason, these 40 patients were divided into 11 women who had DHEA-S levels of ;;. 5 fJ;g/ml and 29 women who had DHEA-S levels of < 5 fJ;g/ml. The body weights of, the patients in these two groups (63 ± 3 and 64 ± . 3 kg) were not different from the weights of our matched ovulatory control subjects (63 ± 2 kg). The ages of the patients in each group were comparable (27 ± 6 versus 26 ± 5 years), and all patients had oligomenorrhea since menarche and an average of 6.4 ± 1.8 years of infertility. The serum T value in patients with DHEA-S levels of> 5 fJ;g/ml was 67 ± 10 ng/dl, which was similar to levels in patients with DHEA-S levels of < 5 fJ;g/ml (62 ± 4 ng/dl). Serum DHEA-S in these two groups averaged 5.8 ± 0.2 fJ;g/ml (range, 0.5 to 8.2 fJ;g/ml) and 3.6 ± 0.4 fJ;g/ml (range, 2.9 to 4.2 fJ;g/ml) (Table 1). Table 1. Mean ± SE SerumLHIFSH Raiios,DHEA-S, and T Levels LH/FSH ratio

DHEA-S jJ.glml

1.5 ± 0.3 1.7 ± 0.1 Ovulatory controls Anovulatory patients: 3.6 ± O.la 3.6 ± O.4a DHEA-S < 5 J-Lg/ml Anovulatory patients: 3.3 ± 0.3 a 5.8 ± 0.2 a.b DHEA-S ~ 5 J-Lg/ml

T ngldl

32 ± 3 62 ± 4a 67 ± lOa

aDifferent from control subjects (P < 0.05). bDifferent from other anovulatory patients (P < 0.05).

Hoffman and Lobo DHEA-S and clomiphene

197

Among these 15 pregnancies, 13 occurred in patients with DHEA-S levels of < 5 f.Lg/ml (54%), and 2 pregnancies occurred in the group with levels of> 5 f.Lg/ml (33%).

2

DISCUSSION

... ... ~

~

... w z

0

c

•

-2 I

1

30 DHEA-S IN .,./1111 N=40

Figure 1 Rankit analysis of baseline DHEA-S values in 40 anovulatory patients with elevated levels of DHEA-S.

Of the entire group of 40 patients, 30 patients (75%) ovulated, based on serum P determinations. Six of the 11 patients (55%) with DHEA-S levels of ~ 5 f.Lg/ml ovulated, and 24 of 29 patients (83%) with DHEA-S levels of < 5 f.Lg/mlovulated (Table 2). This difference approached statistical significance (P = 0.079). Among the 59 anovulatory patients who had normal levels of DHEA-S and who were studied during this same time in: terval, 46 patients ovulated (78%). This ovulation rate was similar to that of patients with elevated levels of DHEA-S. The P levels in the ovulatory CC cycles were similar whether DHEA-S levels were normal, < 5 f.Lg/ml, or > 5 f.Lg/ml. There was no correlation in this study between the ovulatory dose of CC and the level of serum DHEA-S (r = - 0.21). Of the six patients ovulating with DHEA-S levels of> 511g/ml, three (50%) required only 50 mg, one required 100 mg, and. two required 250 mg of CC. In the 24 ovulatory patients with DHEA-S levels of < 5 f.Lg/ml, 14 (58%) ovulated with 50mg, 2 with 100 mg, 2 with 150 mg, 1 with 200 mg, 4 with 250 mg, and 1 with 250 mg ofhCG. Among the 40 treated ovulatory patients followed for 1 year, 15 pregnancies occurred (38%). 198

Hoffman and Lobo DHEA-S and clomiphene

The ovulatory response with CC in patients with elevated levels of DHEA-S (75%) was virtually identical to the response of patients with normal levels ofDHEA-S (78%). These figures are comparable, but a little lower than those previously reported by our clinic for a mixed group of patients (85%).8 This difference may be related to the fact that in this study we only considered P levels of > 10 ng/ml to be ovulatory, whereas values of> 3 ng/ml had been used previously.8, 9 Recent data from conception cycles have suggested that a midluteal P level should be > 10 ng/ml. 16, 17 We had reported that the adjunctive administration of DE X allows some CC-resistant patients to ovulate if they have elevated levels of serum DHEA-S and T.6 It has been well documented that DEX, used alone, results in ovulation in many patients who have elevated levels of androgens. However, it appears from this study that the overall ovulatory response in patients with increased adrenal androgen secretion is similar to that of patients with normal androgen levels. Whether this ovulatory response rate with CC is similar to the use of DEX alone cannot be addressed in this report, but the ovulatory rates are generally comparable to other reports from the literature. 4 , 5 We have observed, as have others, 10 that normalization of serum DHEA-S with DEX does not always result in an ovulatory response without the addition of CC. It appears clear, therefore, that CC has the ability to overcome or circumvent the feedback alterations which occur in patients who have increased adrenal androgen secretion. Rankit analysis of the elevated levels ofDHEAS in our anovulatory patients indicated that two Table 2. Ovulatory Responses with CC in Patients with Normal Serum DHEA-S and DHEA-S Levels < 5 fJ-glml and > 5 fJ-g1m1a

Normal DHEA-S (n = 59) DHEA-S < 5 fJ-g/ml (n = 29) DHEA-S ;;, 5 fJ-g/ml (n = 11)

Ovulatory onCC

Anovulatory onCC

46 (78%) 24 (83%)

13

6 (55%)

5 5

aDHEA-S < 5 fJ-g/ml versus> 5 fJ-g/ml: p. = 0.079. Fertility and Sterility

populations exist with a separation between populations occurring at a level of 5 f,Lg/ml. Thus, it was appropriate to analyze our data dividing patients into two groups with higher and lower levels of DHEA-S. Our data suggest that patients with DHEA-S levels of> 5 f,Lg/ml have decreased ovulatory responses with CC. That this difference only approached statistical significance may be because of sample size. Patients who had DHEAS levels < 5 f,Lg/ml and those with levels> 5 f,Lg/ml had similar levels of T and were comparable endocrinologically. When serum P values were compared in those patients who ovulated who had higher (> 5 f,Lg/ml) and lower « 5 f,Lg/ml) levels ofDHEA-S, there was no difference. Serum P values were all > 10 ng/ml. Although patients with higher levels of DHEAS may have lower ovulatory rates with CC, the CC dose required for ovulation did not correlate with the level of serum DHEA-S, confirming our previous report. 9 Our study size does not allow us to make definitive statements regarding pregnancy rates. However, it would appear that the pregnancy rate after at least four ovulatory cycles is similar to that reported by us previously for all anovulatory women (43%) and comparable to other data using life-table analysis; 18 This report reaffirms the clinical utility of using CC for the treatment of anovulation even when there is an excess of adrenal androgen secretion. However, because the response rate with CC may be lower in patients who have serum DHEA-S levels of> 5 f,Lg/ml, initial adrenal suppression with corticoids in this instance may be indicated. REFERENCES 1. Lobo RA, Paul W, Goebelsmann U: Dehydroepiandrosterone sulfate (DHEA-S) as an indication of adrenal androgen function. Obstet Gynecol 57:69, 1981 2. Lobo RA, Paul W, Goebelsmann U: Serum levels of DHEA-S in gynecologic endocrinopathy and infertility. Obstet Gynecol 57:607, 1981 3. Hoffman DI, Klove K, Lobo RA: The prevalence and significance of elevated dehydroepiandrosterone sulfate levels in anovulatory women. Fertil Steril 42:76, 1984

Vol. 43, No.2, February 1985

4. Greenblatt RE, Barfield W, Lampros CP: Cortisone in the treatment of infertility. Fertil Steril 7:203,1956 5. Perloff WH, Smith KD, Steinberger E: Effect of prednisone on female infertility. Int J Fertil10:31, 1965 6. Lobo RA, Paul W, March CM, Granger L, Kletzky OA: Clomiphene and dexamethasone in women unresponsive to clomiphene alone. Obstet Gynecol 60:497, 1982 7: Daly DC, Walters CA, Soto-Albors CE, Tohan N, Riddick DH: A randomized study of dexamethasone in ovulation induction with clomiphene citrate. Fertil Steril 41:844, 1984 8. Gysler M, March CM, Mishell DR Jr, Bailey EJ: A decade's experience with an individualized clomiphene treatment regimen including its effect on the postcoital test. Fertil Steril 37:161, 1982 9. Lobo RA, Gysler M, March CM, Goebelsmann U, Mishell DR Jr: Clinical and laboratory predictors of clomiphene response. Fertil Steril 37:168, 1982 10. Chang RJ, Abraham GE: Effect of dexamethasone and clomiphene citrate on peripheral steroid levels and ovarian function in a hirsute amenorrheic patient. Fertil Steril 27:640, 1976 11. Lobo RA, Goebelsmann U: Adult manifestation of con gen· ital adrenal hyperplasia due to incomplete 21-hydroxylase deficiency mimicking polycystic ovarian disease. Am J Obstet Gynecol 138:720, 1980 12. Goebelsmann U, Arce JJ, Thorneycroft IH, Mishell DR J r: Serum testosterone concentrations in women throughout the menstrual cycle and following hCG administration. Am J Obstet GynecoI119:445, 1974 13. Lobo RA, Kletzky OA, Kaptein EM, Goebelsmann U: Prolactin modulation of dehydroepiandrosterone sulfate secretion. Am J Obstet Gynecol 138:632, 1980 14. Thorneycroft IH, Stone SC: Radioimmunoassay of serum progesterone in women receiving oral contraceptive ste. roids. Contraception 5:129, 1972 15. Kletzky OA, Nakamura RM, Thorneycroft IH, Mishell DR Jr: Log normal distribution of gonadotropins and ovarian steroid values in the normal menstrual cycle. Am J Obstet Gynecol 121:688, 1975 16. Hull MGR, Savage PE, Bromham DR, Ismail AAA, Mor. ris AF: The value of a single serum progesterone measurement in the midluteal phase as a criterion of a potentially fertile cycle ("ovulation") derived from treated and untreated conception cycles. Fertil Steril 37:355, 1982 17. Hammond MG, Talbert LM: Clomiphene citrate therapy of infertile women with low luteal phase progesterone levels. Obstet Gynecol 59:275, 1982 18. Hammond MG, Halme JK, Talbert LM: Factors affecting the pregnancy rate in clomiphene citrate induction of ovulat~on. Obstet Gynecol 62:196, 1983

Hoffman and Lobo DHEA-S and clomiphene

199