Does transient hyperprolactinemia during ovarian hyperstimulation interfere with conception or pregnancy outcome?*

Vol.

FERTILITY AND STERILITY Copyright <>

1989 The American Fertility Society

51, No.6, June 1989 Printed in U.S.A.

Does transient hyperprolactinemia during ovarian hyperstimulation interfere with conception or pregnancy outcome?*

Yael Gonen, M.D.t Robert F. Casper, M.D.:j: Department of Obstetrics and Gynecology, University of Toronto, Toronto, Ontario, Canada

The significance of transiently increased serum prolactin (PRL) levels on pregnancy rates in in vitro fertilization (IVF) is unknown. The aim of this study was to evaluate PRL levels in IVF patients who conceived and in matched controls who did not. Thirtyseven IVF cycles resulting in pregnancy and 74 nonpregnant cycles were compared. Prolactin was measured before ovarian stimulation with clomiphene citrate, and human menopausal gonadotropin and estradiol (E2 ) and PRL were measured 8 hours after human chorionic gonadotropin (hCG) administration at midcycle. Before ovarian stimulation, serum PRL levels were not different in the pregnant and nonpregnant women (11.1 ± 0.6 [mean ± standard error] ~g/1 and 10.1 ± 0.4 ~g/1, respectively). After hCG, PRL levels were significantly higher in the pregnant women than in the nonpregnant women (20.8 ± 1.6 and 16.0 ± 0.9 ~g/1, respectively; P < 0.005) and more pregnant than nonpregnant women had elevated PRL levels (49% versus 28%, respectively; P < 0.05). There was no correlation between PRL and E 2 in either group. The abortion rate was not different between the women with elevated PRL (22.2%) and the normoprolactinemic women (31.6% ). These results do not support treatment of transient hyperprolactinemia with dopamine agonists in IVF patients. Fertil Steril51:1007, 1989

The normal production of progesterone (P) by the corpus luteum appears to be dependent on the presence of prolactin (PRL).1·2 However, high levels of plasma PRL have been reported to interfere with follicular steroidogenesis and oocyte development.1·3-5 Transient hyperprolactinemia has been observed during follicular maturation in spontaneous2 and stimulated cycles.6 •7 Although the signifiReceived November 21, 1988; revised and accepted February 14,1989. * Supported by the Medical Research Council of Canada and by a Fellowship Training grant to Yael Gonen, M.D., from Wyeth Pharmaceutical Company, Toronto, Ontario, Canada. t Visiting clinical research fellow, Department of Obstetrics and Gynecology, Carmel Hospital, Haifa, Israel. +Reprint requests: Robert F. Casper, M.D., Division of Reproductive Science, Department of Obstetrics and Gynecology, 6-240 EN, Toronto General Hospital, 200 Elizabeth Street, Toronto, Ontario, Canada M5G 2C4.

Vol. 51, No.6, June 1989

cance of mildly increased serum PRL levels as a result of ovarian hyperstimulation is unknown, the empiric normalization of PRL concentrations with a dopamine agonist has been used by some authors8 in an attempt to improve pregnancy rates in women undergoing in vitro fertilization (IVF). Similarly, in patients with unexplained infertility with transient elevations of serum PRL, bromocriptine has been used in an attempt to improve fertility. 9 The benefit of such therapy is not supported by well-controlled clinical studies. 10•11 The objective of this study was to evaluate PRL levels during ovarian stimulation in patients who conceived in our IVF program, and in matched controls who failed to conceive. If transient PRL elevation with ovarian hyperstimulation has a detrimental effect on conception, we reasoned that the pregnant group should have a lower incidence of hyperprolactinemia.

Gonen and Casper Hyperprolactinemia in IVF

1007

MATERIALS AND METHODS Subjects

From January 1986 to May 1988 inclusive, 39 IVF cycles resulted in pregnancies. Thirty-seven of these patients had normal baseline PRL levels ( <17 Jtg/1). Two others had baseline hyperprolactinemia treated by bromocriptine during the cycle, and therefore were excluded from the study. A control group was selected by choosing 2 patients who did not conceive during an IVF cycle in the same week as a conception cycle, and as close to the same day as possible. The only other selection criterion was a normal baseline PRL before beginning IVF therapy, as required in the pregnant women. Age, gravidity, parity, cause of infertility, and stimulation protocols were similar in both groups. Protocol

Ovarian follicular stimulation was routinely induced with clomiphene citrate (Serophene, Serono, Randolph, MA) (50 to 100 mg/day, cycle days 5 to 9) and human menopausal gonadotropins (Pergonal, Serono) 75 to 150 IU intramuscularly (IM) daily starting on cycle day 7. Follicular development was monitored by daily determination of serum levels of estradiol (E 2 ) and luteinizing hormone (LH) and serial ultrasound evaluation. Human chorionic gonadotropin (hCG, Profasi, Serono) 5,000 IU IM was routinely used to trigger the final stage of follicular maturation when at least two follicles were 1.8 to 2.0 em in the greatest diameter and E 2 levels per follicle were 800 to 1,000 pmol/1. Serum PRL concentrations were measured before ovarian stimulation began, and both Ez and PRL concentrations were measured approximately 8 hours after hCG administration. Oocyte retrieval was performed 34 to 36 hours after hCG administration by laparoscopy or transvaginal ultrasoundguided follicle aspiration (after October 1987). In vitro fertilization and embryo transfer were performed using standard techniques. 12 The serum E 2 and PRL levels after hCG administration were compared in the pregnancy and control groups, and the number of patients in each group with transient hyperprolactinemia (>20 Jtg/ 1) was determined. The correlation coefficient of serum E 2 and PRL levels also was determined for both groups. E 2 and PRL Assay

Prolactin was measured by a solid-phase monoclonal antibody immuno-enzymetric assay kit

1008

Gonen and Casper Hyperprolactinemia in IVF

(Tandem-E Prolactin) purchased from Hybritech Inc. (San Diego, CA). Intra-assay and interassay coefficients of variation were determined from serum pools containing low and high levels of PRL. In the low pool (12 Jtg/1 PRL), the intra-assay and interassay coefficients of variation were 9.3% and 9.5%, respectively. In the high pool (78 Jtg/1 PRL), the intra-assay and interassay coefficients of variation were 4.3% and 7.0%, respectively. Estradiol was measured by a solid-phase radioimmunoassay using P25 I]-labeled estradiol (CoatA-Count Estradiol) purchased from Diagnostic Products Corporation (Los Angeles, CA). Estradiol values were expressed in picomoles per liter (pg/ml X 0.003671). Intra-assay and interassay coefficients of variation also were determined directly in serum pools containing two or three different concentrations of E 2 • The intra-assay coefficients of variation at 290 pmol/1 and 1,000 pmol/1 Ez were 5.6% and 4.0%, respectively. Interassay coefficients of variation at 280 pmol/1, at 500 pmol/1, and at 2,500 pmol/1 were 12%, 9.5%, and 7.5%, respectively. Statistical Evaluation

The data obtained were compared by Student's unpair~d results and the chi-square test. Correlation was checked by linear regression analysis. P values <0.05 were regarded as statistically significant.

t-test for

RESULTS

The serum PRL levels before ovarian hyperstimulation (Table 1) were not significantly different between the pregnant and nonpregnant groups (11.1 ± 0.6 [mean ± standard error] Jtg/1 and 10.1 ± 0.4 Jtg/1, respectively). None of the women had a baseline serum PRL concentration > 17 Jtg/1. Mean serum PRL levels after hCG administration were significantly higher in the group that subsequently conceived than in the nonpregnant group (Table 1) (20.8 ± 1.6 and 16 ± 0.9, respectively; P < 0.005). Of the patients who conceived, 18 of 37 (49%) had increased (>20 Jtg/1) PRL (mean, 28.3 ± 1.9 Jtg/1), and 19 women (51%) remained n01:moprolactinemic (mean, 13.7 ± 0.8 Jtg/ l) after hCG. Of the patients who did not conceive, only 21 of 74 women (28%) had increased serum PRL (mean, 25.4 ± 1.8 Jtg/1), and 53 women (72%) remained normoprolactinemic (mean, 12.3 ± 0.5

Fertility and Sterility

Table 1 Prolactin and Estradiol Serum Concentrations in IVF Patients Who Conceived (n or in Matched Controls (n = 74) Who Did Not Conceive

Pregnant (n = 37) Nonpregnant (n = 74) a

=

37)

Baseline PRL

PRL (post-hCG)

Delta PRL

Ez (post-hCG)

E 2 -PRL

1'{1/l

p.g(l

1'{1/1

pmol/l

r'

11.1 ± 0.6

20.8 ± 1.6"

9.7 ± 1.6b

4,744 ± 263

o.o4c

10.1 ± 0.4

16.0 ± 0.9

5.9 ± 1.0

4,318 ± 197

0.01 c

P<0.005.

c

NS, not significant.

b P<0.02.

JLg/1}. Significantly more patients had transient elevation of PRL in the pregnant group than in the group that did not conceive (49% and 28%, respectively; P < 0.05). The mean serum E 2 levels after hCG administration were not significantly different (4, 744 ± 263 pmol/1 and 4,318 ± 197 pmol/1 in the pregnant and nonpregnant women, respectively) (Table 1). Serum PRL levels were correlated with serum E 2 concentration by linear regression analysis in both groups (Fig. 1). There was no correlation between serum PRL and E 2 levels after hCG administration in the pregnant women (r = 0.04) or in the nonpregnant women (r = 0.01). In the pregnant women, PRL levels during ovarian hyperstimulation did not appear to be associated with pregnancy outcome. No differences were observed in the spontaneous abortion rate between the women with elevated PRL before oocyte retrieval (4 of 18; 22.2%) and the normoprolactinemic women (6 of 19; 31.6%).

•

".

•

• •

ESTRADIOL {pmol/1.)

Figure 1 Serum PRL concentrations plotted against serum E 2 concentrations in 37 patients who conceived (e) and in 74 women who failed to conceived (A) with IVF. No significant correlation was observed in either the pregnant (r = 0.04) or in the nonpregnant (r = 0.01) women.

Vol. 51, No.6, June 1989

DISCUSSION

The mechanism by which PRL participates in the regulation of ovarian function is only partly understood. Overt hyperprolactinemia may disturb the development and function of ovarian follicles, leading to infertility because of anovulation or luteal phase inadequacy13•14; however, the clinical significance of short-term elevation of serum PRL at midcycle is unclear. This study examined the relationship of transient hyperprolactinemia to conception and pregnancy outcome in patients undergoing IVF. The results indicated a significantly higher mean serum PRL concentration and a higher proportion of women with hyperprolactinemia in the group who conceived than in the group who did not. Therefore, it is unlikely that transient hyperprolactinemia is detrimental to the occurrence of pregnancy in IVF, and the results support the concept that midcycle elevation of serum PRL is likely physiologic rather than pathologic. The elevation of PRL at midcycle has been hypothesized to result from elevated serum E 2 concentrations in spontaneous and especially in stimulated cycles. 6 However, this study demonstrated no significant correlation between serum PRL and E 2 concentrations, suggesting that another mechanism may be responsible for the midcycle rise in serum PRL. A possible, but at present, speculative, hypothesis is that the transient hyperprolactinemia is linked to the midcycle surge of LH. Clinically, administration of gonadotropin-releasing hormone (GnRH) or GnRH agonists to women results in simultaneous pulsatile release of PRL as well as LH, 15 •16 and spontaneous pulses of LH and PRL are synchronous during the menstrual cycle. 17 Denef and Andries 18 using co-cultured ganadotropes and lactotropes, have demonstrated that the simultaneous release of LH and PRL in response

Gonen and Casper Hyperprolactinemia in IVF

1009

to GnRH in vitro may be a result of a paracrine response involving the endogenous opioid peptide dynorphin. These data are supportive of a physiologic midcycle elevation of serum PRL. Recent observations that endogenous LH surges occur in more than half of IVF cycles in which hCG is given 19 are consistent also with the hypothesis that PRL may be released together with LH in a surge at midcycle. Reinthaller et al. 8 reported that treatment of transient hyperprolactinemia with bromocriptine during IVF cycles significantly increased fertilization and cleavage rates of oocytes and luteal phase serum P levels, although no improvement in pregnancy rates was seen. In contrast, it has been observed that pharmacologic lowering of serum PRL may interfere with follicular steroidogenesis. Kauppila et al. 20 and Schulz et al. 2 reported that suppression of PRL secretion with a dopamine agonist in normal women decreased luteal phase progesterone and significantly shortened the length of the luteal phase. McNatty et al. 1 demonstrated that low levels of PRL are required for P production by human granulosa/luteal cells in culture. In summary, our data show that transient elevation of PRL as a result of ovarian hyperstimulation does not appear to affect pregnancy occurrence or outcome adversely. These results do not support the treatment of transient hyperprolactinemia with dopamine agonists in IVF patients. The implications of this study for patients with transient hyperprolactinemia and idiopathic infertility remain to be determined. Acknowledgments. We wish to thank Dr. Ido Katz for his assistance in statistical evaluation of the data and Mary Coates for her help with the preparation of the manuscript.

REFERENCES 1. McNatty KP, Sawers RS, McNeilly AS: A possible role for prolactin in control of steroid secretion by the human Graafian follicle. Nature 250:653, 1974 2. Schulz KD, Geiger W, del Pozo E, Kunzig HJ: Pattern of sexual steroids, prolactin and gonadotropic hormones during prolactin inhibition in normally cycling women. Am J Obstet Gynecol132:561, 1978 3. Darrington J, Gore-Langton RE: Prolactin inhibits oestrogen synthesis in the ovary. Nature 290:600, 1981

1010

Gonen and Casper Hyperprolactinemia in IVF

4. Demura R, Ono M, Demura H, Shizume K, Oouchi H: Prolactin directly inhibits basal as well as gonadotropin-stimulated secretion of progesterone and 17 /3-estradiol in the human ovary. J Clin Endocrinol Metab 54:1,246, 1982 5. Kauppila A, Leinonen P, Vihko R, Ylostalo P: Metoclopramide-induced hyperprolactinemia impairs ovarian follicle maturation and corpus luteum function in women. J Clin Endocrinol Metab 54:955, 1982 6. Ho Yuen BH, McComb P, Sy L, Lewis J, Cannon W: Plasma prolactin, human chorionic gonadotropin, estradiol, testosterone, and progesterone in the ovarian hyperstimulation syndrome. Am J Obstet Gynecol133:316, 1979 7. Healy DL, Burger GH: Serum-follicle-stimulating hormone and prolactin during the induction of ovulation with exogenous gonadotropin. J Clin Endocrinol Metab 56:474, 1983 8. Reinthaller A, Bieglmayer C, Deutinger J, Csaicsich P: Transient hyperprolactinemia during cycle stimulation: influence on the endocrine response and fertilization rate of human oocytes and effects of bromocriptine treatment. Fertil Steril49:432, 1988 9. DeVane GW, Guzick DS: Bromocriptine therapy in normoprolactinemic women with unexplained infertility and galactorrhea. Fertil Steril46:1,026, 1986 10. Wright CS, Steele SJ, Jacobs HS: Value ofbromocriptine in unexplained primary infertility: a double-blind controlled trial. BrMedJ 1:1,037,1979 11. McBain JC, Pepperell RJ: Use of bromocriptine in unexplained infertility. Clin Reprod Fertil1:145, 1982 12. Wood C, Trounson A: Clinical in vitro fertilization. Berlin, Springer-Verlag, 1984 13. Seppala M, Ranta T, Hirvonen E: Hyperprolactinemia and luteal insufficiency. Lancet 1:229, 1976 14. Bohnet HG, Dahler HG, Wuttke W, Schneider HPG: Hyperprolactinemic anovulatory syndrome. J Clin Endocrinol Metab 42:132, 1975 15. Yen SSC, Hoff JD, Lasley BL, Casper RF, Sheehan KL: Induction of prolactin release by LRF and LRF -agonist. Life Sci 26:1,963, 1980 16. Casper RF, Yen SSC: Simultaneous pulsatile release of PRL and LH induced by an LRF agonist. J Clin Endocrinol Metab 52:934, 1981 17. Cetel NS, Quigley ME, Yen SSC: Naloxone induced prolactin secretion in women: evidence against a direct prolactin stimulatory effect of endogenous opioids. J Clin Endocrinol Metab 60:191, 1985 18. Denef C, Andries M: Evidence for paracrine interaction between gonadotrophs and lactotrophs in pituitary cell aggregates. Endocrinology 112:813, 1983 19. Maroulis GB: Endogenous human chorionic gonadotrophin (hCG) does not inhibit the endogenous luteinizing hormone (LH) surge. (Abstr. 074). Presented at the Annual Meeting of The American Fertility Society, Atlanta, GA, October 10 to 13, 1988. Published by The American Fertility Society in the Program Supplement, 1988, p S26 20. Kauppila A, Marikainen H, Puistola U, Reinila M, Ronnberg L: Hypoprolactinemia and ovarian function. Fertil Steril49:437, 1988

Fertility and Sterility