- Email: [email protected]
Native gonadotropin-releasing hormone for triggering follicular maturation in polycystic ovary syndrome patients undergoing human menopausal gonadotropin ovulation induction
FERTILITY AND STERILITY
Vol. 62, No. 3, September 1994
Copyright e 1994 The American Fertility Society
Printed on acid-free paper in U. S. A.
Native gonadotropin-releasing hormone for triggering follicular maturation in polycystic ovary syndrome patients undergoing human menopausal gonadotropin ovulation induction Zeev Blumenfeld, M.D. *t Nessia Lang, M.D.* Amnon Amit, M.D.*
Luna Kahana, Ph.D.:j: Nehemia Yoffe, M.D.*
Rambam Medical Center, The Bruce Rapport Faculty of Medicine, Technion-Israel Institute of Technology, and the Linn Medical Center, Haifa, Israel
Objective: To evaluate the role of GnRH administration instead of hCG for triggering follicular maturation in patients with polycystic ovaries (PCO) undergoing hMG ovulation induction when the late follicular 17-{1-E2 levels are > 1,600 pg/mL (>6,000 pmol/L). Design: Prospective study. Setting: Infertility outpatient clinic of Rambam Medical Center (general hospital), Haifa, Israel. Patients and Interventions: High serum E 2 concentrations from 1,600 to >3,600 pg/mL (2,800 ± 68, mean ± SD [6,000 to > 13,000 pmol/L, 10,279 ± 2,500]) were experienced in 44 hMG cycles. The number of preovulatory follicles visualized by transvaginal sonography was between 8 and 25. An IV injection of 200 J.Lg GnRH was administered for triggering final follicular maturation and ovulation, instead of 10,000 IU IM hCG, usually injected for this purpose, when the E 2 levels are s1,600 pg/mL (6,000 pmol/L). Serum E 2 and P levels were monitored in the luteal phase. In cycles where E 2 decreased to s1,360 pg/mL (5,000 pmol/L), 2,500 IU hCG was administered once or twice at 3-day intervals for luteal support. Main Outcome Measures: Pregnancy and abortion rates and the rate of ovarian hyperstimulation syndrome (OHSS). Results: Ten pregnancies were generated by the hMG and GnRH coctreatment in 32 patients (31.2%), in 44 cycles (23%). Two pregnancies aborted (20%), and eight generated eight healthy neonates. Ovarian hyperstimulation syndrome occurred in two cycles of patients who were both pregnant. All but two of these PCO patients also have undergone 69 hMG and hCG cycles. Only 7 patients conceived (23%) 10 times (10/69, 14.5%); 5 of these pregnancies (50%) were multiple gestations (3 twins, 1 sextuplet, and 1 heptuplet gestation). The pregnancy wastage rate was 30% (3/10). Conclusion: The use of native GnRH to trigger ovulation in PCO patients with late follicular E 2 levels >1,600 pg/mL (6,000 pmol/L) appears to be comparable with prior hMG and hCG cycles in terms of pregnancy rate, pregnancy wastage, risk of multiple gestation, and incidence of severe ovarian hyperstimulation. Unlike hMG and GnRH-agonist, which is associated with luteal phase dysfunction, hMG and GnRH offers a preferable alternative due to the ability of hCG luteal support and rescue, providing the E 2 levels are not dangerously increased. Fertil Steril 1994;62:456-60 Key Words: Gonadotropin-releasing hormone, follicular maturation, ovulation induction, PCOS, ovarian hyperstimulation syndrome (OHSS)
Received May 5, 1993; revised and accepted April 13, 1994. * Reproductive Endocrinology and Infertility Section, Department of Obstetrics and Gynecology, Rambam Medical Center, The Bruce Rappaport Faculty of Medicine. t Reprint requests: Zeev Blumenfeld, M.D., Reproductive Endocrinology and Infertility Section, Department of Obstetrics and Gynecology, Rambam Medical Center, Faculty of Medicine,
456
Blumenfeld et al.
Ovarian hyperstimulation syndrome (OHSS) is a major and life-threatening complication of hMG
Technion-Israel Institute of Technology, Haifa, 31096, Israel (FAX: 972-4-515710). tCentral Endocrine Laboratory, Linn Medical Center.
GnRH follicular maturation in hMG cycles
Fertility and Sterility
and hCG ovulation induction. Patients with polycystic ovary syndrome (PCOS), undergoing hMG ovulation induction, are prone to develop OHSS, when the late follicular serum concentration of 17,B-E2 is high (>1,600 pg/mL, >6,000 pmol/L), especially when multiple ovarian follicles are sonographically detected (1-4). Several studies (5-8) recently have documented that GnRH agonist (GnRH-a) analogue is an effective alternative to hCG for use in IVF cycles in which controlled ovarian hyperstimulation using menotropins was attempted. In these studies, luteal phase E 2 levels were significantly lower in the GnRH-a group than in the hCG group (5-7). A potential advantage of GnRH-a use, especially in patients at high risk for developing OHSS (1-4), may be a decrease in the incidence of this complication, as previously suggested (5, 6). However, an apparent short luteal phase was present in approximately 18% ofhMG and GnRH-a cycles, despite the use of natural P for luteal support (5). Segal and Casper (5), have suggested that down regulation of pituitary GnRH receptors and reduced LH support for the corpus luteum (CL) may occur, even after a single administration of GnRH -a. Alternatively, inhibitory effects of GnRH -a on luteal P production have been reported (9-12), suggesting a direct detrimental, luteolytic effect of GnRH-a (6). We attempted, therefore, to trigger final follicular maturation in PCOS patients undergoing hMG superovulation with high late follicular E 2 levels with an intravenous GnRH bolus instead of hCG GnRH -a in hope of preventing luteal dysfunction without dangerously increasing the risk of OHSS.
hCG administration was witheld and an IV bolus of 200 J.Lg GnRH (Relisorm; Serono, Aubonne, Switzerland) was administered for triggering final follicular maturation and ovulation, instead of 10,000 IU hCG, usually injected IM for this purpose to our patients when late follicular concentration is <1,600 pg/mL (6,000 pmoljL) (13). Ovulation was triggered by 10,000 IU hCG when the late follicular E 2 concentration was between 300 and 1,600 pgjmL (1,100 to 6,000 pmol/L) ir the presence of one to five follicles with a mean diameter > 16 rom, measured by transvaginal ultrasound (13-15). However, when these sonographic criteria were met but the late follicular E 2 levels were >1,600 pg/mL (6,000 pmoljL) ovulation was triggered with native GnRH instead of hCG. In all these cycles the number of preovulatory follicles was between 8 and 25. Serum E 2 and P concentrations were monitored every 3 to 7 days throughout the luteal phase. In cycles where E 2 concentration decreased to ::::;;1,360 pgjmL (::::;;5,000 pmoljL), 2,500 IU IM hCG was administered once or twice at 3-day intervals, for luteal support. In cases where the postcoital test (PCT) was normal, the couples were advised to have intercourse on the day of GnRH administration and on the two following days. In cases where the PCT was negative, lUI with washed husband spermatozoa, after swim-up preparation using Ham's F-10 medium (Biological Industries, Beth Haemek, Israel) was accomplished, as previously described (14). The criteria for definition of OHSS, which we employed in this study, were published previously (1, 4).
RESULTS
MATERIALS AND METHODS
High serum E 2 concentration, from 1,600 pg/mL (6,000 pmoljL) to over 3,600 pg/mL (13,000 pmolj L) was measured in 44 ovulation induction cycles in 32 PCOS patients using hMG (Pergonal; Teva, Petah Tikva, Israel) or FSH (Metrodin; Teva), two ampules per day starting from the 3rd menstrual day, as previously described (13). In brief, ovarian response to ovulation induction was monitored by measurements of serum E 2 and P concentrations by RIA and by measurement of the number of developing follicles and their diameter by transvaginal ultrasound (US). The mean E 2 concentration before GnRH administration was 2,800 ± 681 (±SD) pg/ mL (10,279 ± 2,500 pmol/L). There were between 8 and 25 preovulatory follicles visualized by US. Because of the high risk ofOHSS in these cycles (1, 3), Vol. 62, No.3, September 1994
Ten pregnancies were generated by 44 hMG and GnRH co-treatment cycles (22. 7%) in these 32 PCOS patients (31.2%). Two of these pregnancies were multiple gestations (2/10, 20%), one was a twin pregnancy that ended in missed abortion at 10 weeks gestation, and the other, a triplet gestation, ended by cesarean section at 30 weeks gestation due to placental abrubtion. Two of the three premature newborns, weighing 1,250 to 1,455 g suffered early neonatal death, and one survived after 1 month in the neonatal intensive care unit. The pregnancy wastage was 20% (2/10), due to two spontaneous first trimester abortions. The remaining eight pregnancies ended in seven spontaneous vaginal deliveries and one cesarean section (due to pre-eclamptic toxemia), of eight healthy, term neonates.
Blumenfeld et al.
GnRH follicular maturation in hMG cycles
457
60
•
bMG/GnRH
121 bMG!bCG
[E2) nmol/L
PR/C
PR/P
MPR
PW
OHSS/P OHSS/C
Figure 1 The mean (±SD) late follicular E 2 concentrations [E nmol/L] in hMG and GnRH (•) or hMG and hCG (~) and the comparative incidence of pregnancy rate per cycle (PR/C), pregnancy rate per patient (PR/P), multiple pregnancy rate (MPR), pregnancy wastage (PW), and ovarian hyperstimulation rates per patient (OHSS/P) or per cycle (OHSS/C).
the conception cycles. In those cycles where the E 2 remained dangerously increased, > 1,360 pg/mL (>5,000 pmol/L) after the GnRH injection, no luteal hCG was administered and a daily IM injection of 50 mg Pin oil (Gestone; Paines & Byrne, Greenford, United Kingdom) for 14 days was administered instead. Severe OHSS necessitating hospitalization, parenteral fluid treatment, and abdominal paracentesis (of 4 to 6 L) of clear abdominal fluid (17) occurred in two cycles (4.5%). Both these patients conceived (one was delivered of a normal singleton and the other of premature triplets). Six patients experienced severe ovarian hyperstimulation in the previous hMG and hCG cycles, all of them in conception cycles. DISCUSSION
All but 1 of these 32 patients have been treated also by hMG and hCG ovulation induction cycles. Of the 69 previous hMG and hCG cycles in these 31 patients, 10 pregnancies were achieved (10/69, 14.5%) in 7 patients (7 /31, 22.5% ). The difference between the cycle fecundity [pregnancy rate (PR)] of the hMG and GnRH cycles (22. 7%) and that generated by the previous hMG and hCG cycles in the same patients (14.5%) was not statistically significant (P > 0.05, x2 test). Five of the 10 gestations (50%) achieved in the PCO patients by previous hMG and hCG treatments were multiple pregnancies: one heptuplet pregnancy ending in abortion, one sextuplet gestation spontaneously turning into a triplet gestation (16), three twin pregnancies, and five singletons. The pregnancy wastage rate was 30% (3/10) in these hMG- and hCG-generated gestations (Fig. 1). The mean E 2 serum concentrations on the day of IV GnRH bolus administration was 2,800 ± 681 pg/ mL (10,279 ± 2,500 pmol/L) (n = 44), decreasing to 1,057 ± 620 pg/mL (3,880 ± 2,275 pmol/L) (n = 40) within 3 to 5 days afterward. An IM injection of 2,500 IU hCG was administered once or twice, 3 and 6 days after the IV GnRH injection for luteal support, as previously described for hMG and hCG ovulation induction (16), in those cycles where the E 2 level decreased to <1,360 pg/mL (5,000 pmol/L). The midluteal E 2 (7 to 10 days after the IV GnRH bolus) was 2,297 ± 888 pg/mL (8,433 ± 3,260 pmol/ L) (n = 44). The mean midluteal E 2 concentration was 2,083 ± 882 pg/mL (7,647 ± 3,240 pmoljL) (n = 34) for those who did not conceive, compared with 2,976 ± 530 pg/mL (10,925 ± 1,950 pmol/L) in 458
Blumenfeld et a!.
The GnRH analogues have gained much popularity in the last decade for treatment of various indications including PCOS and as an adjunct to ovulation induction for IVF and assisted reproduction (18). Several recent studies (5-8) have documented that GnRH-a are an effective alternative to hCG for use in IVF cycles in which controlled ovarian hyperstimulation using hMG was attempted, especially in PCOS patients with high late-follicular levels of E 2 , who are more prone to develop the lifethreatening OHSS complication. However, a short luteal phase and apparently defective CL may be associated with the hMG and GnRH-a co-treatment, despite the use of P for luteal support (5). This shortcome may be expected in light of the reported inhibitory effect of even a single GnRH -a administration on the luteal P production, suggesting a direct detrimentalluteolytic effect ofGnRH-a (5, 6, 9-12). Native GnRH has been used successfully to trigger follicular maturation in patients pretreated with hMG for almost two decades, either in the form of nasal spray (19) or IM (20). However, an apparent discrepancy between the ovulations obtained and the lack of achieved pregnancies was observed associated with short luteal phase in most of the co-treatment cycles (20). We attempted, therefore, a modification of the previously reported hMG and GnRH -a or hMG and GnRH combinations in PCOS patients with dangerously increased high late-follicular E 2 levels :2:1,600 pg/mL (:2:6,000 pmol/L) by using an IV GnRH bolus and by supporting the luteal phase with one or two doses of2,500 IU hCG (16), every 3 days. We
GnRH follicular maturation in hMG cycles
Fertility and Sterility
hoped to prevent the previously reported high incidence ofluteal dysfunction (5) without dangerously increasing the risk of OHSS. Our results of 22.7% cycle fecundity and 31.2% PR (per patient) with a 20% (2/10) pregnancy wastage rate are at least as good as the respective parameters in the same group of PCOS patients in 69 previous hMG and hCG cycles (14.5% PR per cycle; 23% PR/patient; 30% pregnancy wastage, Fig. 1). However, the 20% multiple PR (2/10) in the hMG and GnRH cotreatment is <50% multiple gestations (5/10) in the previous hMG and hCG treatment cycles in the same group of patients. Although the difference is not statistically significant, probably due to the small numbers in each group, the difference is encouraging, and further experience (by increasing the group numbers) may achieve statistical significance. One may critically argue against the safety of supporting the luteal phase with one or two doses of 2,500 IU IM hCG. We attempted this luteal support, as previously described (16), when the E 2 concentrations decreased to :::;;1,360 pg/mL (5,000 pmol/L), and administered daily Pin oil instead in those cycles where E 2 remained persistently high. Not unexpectedly, the two cases of severe OHSS necessitating hospitalization and abdominal paracentesis occurred both in pregnant patients, one of whom received two luteal hCG injections, while the other did not. This preliminary experience suggests that OHSS is associated mainly with achievement of gestation and not necessarily with exogenous luteal hCG, versus P supplementation when no subsequent pregnancy occurs. Because luteal exogenous hCG administration may prevent the treatment associated luteolysis, the resulting cycle fecundity and PRs may be therefore higher, with a higher associated risk of subsequent OHSS. Because of the life-threatening danger of severe OHSS, we suggest to reserve the luteal hCG supplementation after hMG and GnRH co-treatment only to those cycles where E 2 concentrations decrease to <1,360 pg/mL (5,000 pmol/L) and to support the luteal phase with P whenever the E 2 levels remain higher. In the previous 69 hMG and hCG cycles in these PCO patients, the late follicular E 2 concentrations were significantly lower, :::;;1,600 pg/mL (:::;;6,000 pmol/L), and in spite of the fact that the PR was lower, the rate of multiple pregnancies was higher (50% versus 20%) and the rate ofOHSS necessitating hospitalization was higher (19.4% [6/31] versus 6.2% [2/ 32]). Vol. 62, No.3, September 1994
One would have anticipated that the patients in the study group with the extremely high late-follicular E 2 levels and a greater number of large- and medium-sized preovulatory follicles would have had a greater risk of severe ovarian hyperstimulation than the same PCO patients in previous hMG and hCG cycles, where the late follicular E 2 concentration was <1,600 pg/mL ( <6,000 pmoljL) (1-4). The results are surprisingly quite opposite: the OHSS was less common in the hMG and GnRH co-treatment than in the hMG and hCG cycles. Although not statistically different, because of the small number in each group, these results suggest encouragingly that triggering ovulation with GnRH and providing hCG (in low doses) for luteal support in cycles with dangerously high E 2 levels does not appear to increase the incidence of OHSS. These results are in keeping with the common experience of high risk of OHSS in the PCOS subgroup of patients (1-4). The 31% PR in our series is also as good as the 20% to 33% PR in the hMG and GnRHa combination reported by other investigators (57), with an apparently lower pregnancy wastage rate (20% versus 33% to 50%), suggesting a smaller incidence of luteal dysfunction. Recently, the popularity of the controlled ovarian stimulation protocols using GnRH -a during the follicular phase to prevent premature luteinization for IVF and assisted reproduction has increased (5). In these protocols, the induced gonadotroph desensitization to GnRH may make the triggering of follicular maturation with GnRH-a unpractical (5). However, Filicori et al. (21) have been successful in inducing follicular maturation, growth, and even pregnancies in PCOS patients previously treated with GnRH -a, by pulsatile intravenous administration of native GnRH. This previous experience may make the hMG and GnRH co-treatment possible, and an optional alternative to hCG in PCOS patients undergoing hMG ovarian stimulation after previous GnRH -a desensitization, in cycles where the late follicular E 2 is dangerously increased, ~1,600 pg/mL (~6,000 pmol/L). As we have hypothesized several years ago (16), the hCG luteal support may be superior to P administration due to the optional increase in the secretion of relaxin, inhibin, and possible other growth factors playing a crucial role in the establishment of gestation, by the CL in response to hCG and not toP. Recent data seem to apparently support this assumption (22). Prospective comparative studies between hMG and GnRH versus hMG and GnRH-a and between hMG and GnRH with hCG or P luteal support are
Blumenfeld et al.
GnRH follicular maturation in hMG cycles
459
needed to more accurately answer some of the still unanswered clinical questions. REFERENCES 1. Golan A, Ron-El R, Herman A, Soffer Y, W einraub Z, Caspi E. Ovarian hyperstimulation syndrome: an update review. Obstet Gynecol Surv 1989;44:430-8. 2. Kemmann E, Tavakou F, Schoen RM, Janes JR. Induction of ovulation with menotropins in women with polycystic ovary syndrome. Am J Obstet Gynecol 1981;141:58-64. 3. Blankstein J, Shalev J, Saadon T, Kukia EE, Rabinovici J, Pariente C, et a!. Ovarian hyperstimulation syndrome: prediction by number and size of preovulatory ovarian follicles. Fertil Steril 1987;4 7:597-602. 4. Pride SM, James CS, Ho Yuen B. The ovarian hyperstimulation syndrome. Sem Reprod Endocrinol1990;8:247-60. 5. Segal S, Casper RF. Gonadotropin-releasing hormone agonist versus human chorionic gonadotropin for triggering follicular maturation in in vitro fertilization. Fertil Steril 1992;57:1254-8. 6. ltskovitz J, Boldes R, Levron J, Erlik Y, Kahana L, Brandes JM. Induction of preovulatory luteinizing hormone surge and prevention of ovarian hyperstimulation syndrome by gonadotropin-releasing hormone agonist. Fertil Steril 1991;56:213-20. 7. Gonen Y, Balakier H, Powell W, Casper RF. Use of gonadotropin-releasing hormone agonist to trigger follicular maturation for in vitro fertilization. J Clin Endocrinol Metab 1990;71:918-22. 8. Imoedemhe DAG, Sigue AB, Pacpaco ELA, Olazo AB. Stimulation of endogenous surge of luteinizing hormone with gonadotropin-releasing hormone analog after ovarian stimulation for in vitro fertilization. Fertil Steril 1991;55:328-32. 9. Parinaud J, Beaur A, Bourreau E, Vieitez G, Pontonnier G. Effect of a luteinizing hormone-releasing hormone agonist (buserelin) on steroidogenesis of cultured human preovulatory granulosa cells. Fertil Steril1988;50:597-602. 10. Maruo T, Otani T, Mochizuki M. Antigonadotropic actions of GnRH agonist on ovarian cells in vivo and in vitro. J Steroid Biochem 1985;23:765-802. 11. Tureck RW, Mastroianni L, Blasco L, Strauss JF III. Inhibition of human granulosa cell progesterone secretion by a gonadotropin-releasing hormone agonist. J Clin Endocrinol Metab 1982;54:1078-85.
460
Blumenfeld et al.
12. Reddy PV, Azhar S, Menon KNJ. Multiple inhibitory actions of luteinizing hormone-releasing hormone agonist on luteinizing hormone/human chorionic gonadotropin receptor mediated ovarian responses. Endocrinology 1980; 107:930-6. 13. Blumenfeld Z, Barkey RJ, Youdim MBH, Brandes JM, Amit T. Growth hormone-binding protein regulation by estrogen, progesterone, and gonadotropins in human: The effect of ovulation induction with menopausal gonadotropins, GH and gestation. J Clin Endocrinol Metabol 1992; 75:1242-49. 14. Blumenfeld Z, Nahhas F. Pretreatment of sperm with human follicular fluid for borderline male infertility. Fertil Steril1989;51:863-8. 15. Blumenfeld Z, Nahhas F. Luteal dysfunction in ovulation induction: the role of repetitive human chorionic gonadotropin supplementation during the luteal phase. Fertil Steril 1988;50:403-7. 16. Blumenfeld Z, Dirnfeld M, Abramovici H, Amit A, Bronshtein M, Brandes JM. Spontaneous fetal reduction in multiple gestations assessed by transvaginal ultrasound. Br J Obstet Gynaecol 1992;99:333-7. 17. Thaler I, Yoffe N, Kaftory JK, Brandes JM. Treatment of ovarian hyperstimulation syndrome: the physiologic basis for a modified approach. Fertil Steril 1981;36:110-3. 18. Blumenfeld Z, Enat R, Brandes JM, Baruch Y. Gonadotropin-releasing hormone analogues for dysfunctional bleeding in women after liver transplantation: a new application. Fertil Steril1992;57:1121-3. 19. Potashnik G, Homburg R, Lunenfeld B. Ovulation induction in an amenorrheic woman by using GnRH in nasal spray. Harefuah 1975;88:453-5. 20. Crosignani PG, Trojsi L, Attanasio A, Tonani E, Donini P. Hormonal profiles in anovulatory patients treated with gonadotropins and synthetic luteinizing hormone releasing hormone. Obstet Gynecol 1975;46:15-22. 21. Filicori M, Flamigni C, Meriggiola MC, Ferrari P, Michelacci, Campaniello E, et a!. Endocrine response determines the clinical outcome of pulsatile gonadotropin-releasing hormone ovulation induction in different ovulatory disorders. J Clin Endocrinol Metab 1991;72:965-72. 22. Norman RJ, McLoughlin JW, Borthwick GM, Yohkaichiya T, Matthews CD, MacLennan AH, eta!. lnhibin and relaxin concentrations in early singleton, multiple, and failing pregnancy: relationship to gonadotropin and steroid profiles. Fertil Steril 1993;59: 130-7.
GnRH follicular maturation in hMG cycles
Fertility and Sterility
Vol. 62, No. 3, September 1994
Copyright e 1994 The American Fertility Society
Printed on acid-free paper in U. S. A.
Native gonadotropin-releasing hormone for triggering follicular maturation in polycystic ovary syndrome patients undergoing human menopausal gonadotropin ovulation induction Zeev Blumenfeld, M.D. *t Nessia Lang, M.D.* Amnon Amit, M.D.*
Luna Kahana, Ph.D.:j: Nehemia Yoffe, M.D.*
Rambam Medical Center, The Bruce Rapport Faculty of Medicine, Technion-Israel Institute of Technology, and the Linn Medical Center, Haifa, Israel
Objective: To evaluate the role of GnRH administration instead of hCG for triggering follicular maturation in patients with polycystic ovaries (PCO) undergoing hMG ovulation induction when the late follicular 17-{1-E2 levels are > 1,600 pg/mL (>6,000 pmol/L). Design: Prospective study. Setting: Infertility outpatient clinic of Rambam Medical Center (general hospital), Haifa, Israel. Patients and Interventions: High serum E 2 concentrations from 1,600 to >3,600 pg/mL (2,800 ± 68, mean ± SD [6,000 to > 13,000 pmol/L, 10,279 ± 2,500]) were experienced in 44 hMG cycles. The number of preovulatory follicles visualized by transvaginal sonography was between 8 and 25. An IV injection of 200 J.Lg GnRH was administered for triggering final follicular maturation and ovulation, instead of 10,000 IU IM hCG, usually injected for this purpose, when the E 2 levels are s1,600 pg/mL (6,000 pmol/L). Serum E 2 and P levels were monitored in the luteal phase. In cycles where E 2 decreased to s1,360 pg/mL (5,000 pmol/L), 2,500 IU hCG was administered once or twice at 3-day intervals for luteal support. Main Outcome Measures: Pregnancy and abortion rates and the rate of ovarian hyperstimulation syndrome (OHSS). Results: Ten pregnancies were generated by the hMG and GnRH coctreatment in 32 patients (31.2%), in 44 cycles (23%). Two pregnancies aborted (20%), and eight generated eight healthy neonates. Ovarian hyperstimulation syndrome occurred in two cycles of patients who were both pregnant. All but two of these PCO patients also have undergone 69 hMG and hCG cycles. Only 7 patients conceived (23%) 10 times (10/69, 14.5%); 5 of these pregnancies (50%) were multiple gestations (3 twins, 1 sextuplet, and 1 heptuplet gestation). The pregnancy wastage rate was 30% (3/10). Conclusion: The use of native GnRH to trigger ovulation in PCO patients with late follicular E 2 levels >1,600 pg/mL (6,000 pmol/L) appears to be comparable with prior hMG and hCG cycles in terms of pregnancy rate, pregnancy wastage, risk of multiple gestation, and incidence of severe ovarian hyperstimulation. Unlike hMG and GnRH-agonist, which is associated with luteal phase dysfunction, hMG and GnRH offers a preferable alternative due to the ability of hCG luteal support and rescue, providing the E 2 levels are not dangerously increased. Fertil Steril 1994;62:456-60 Key Words: Gonadotropin-releasing hormone, follicular maturation, ovulation induction, PCOS, ovarian hyperstimulation syndrome (OHSS)
Received May 5, 1993; revised and accepted April 13, 1994. * Reproductive Endocrinology and Infertility Section, Department of Obstetrics and Gynecology, Rambam Medical Center, The Bruce Rappaport Faculty of Medicine. t Reprint requests: Zeev Blumenfeld, M.D., Reproductive Endocrinology and Infertility Section, Department of Obstetrics and Gynecology, Rambam Medical Center, Faculty of Medicine,
456
Blumenfeld et al.
Ovarian hyperstimulation syndrome (OHSS) is a major and life-threatening complication of hMG
Technion-Israel Institute of Technology, Haifa, 31096, Israel (FAX: 972-4-515710). tCentral Endocrine Laboratory, Linn Medical Center.
GnRH follicular maturation in hMG cycles
Fertility and Sterility
and hCG ovulation induction. Patients with polycystic ovary syndrome (PCOS), undergoing hMG ovulation induction, are prone to develop OHSS, when the late follicular serum concentration of 17,B-E2 is high (>1,600 pg/mL, >6,000 pmol/L), especially when multiple ovarian follicles are sonographically detected (1-4). Several studies (5-8) recently have documented that GnRH agonist (GnRH-a) analogue is an effective alternative to hCG for use in IVF cycles in which controlled ovarian hyperstimulation using menotropins was attempted. In these studies, luteal phase E 2 levels were significantly lower in the GnRH-a group than in the hCG group (5-7). A potential advantage of GnRH-a use, especially in patients at high risk for developing OHSS (1-4), may be a decrease in the incidence of this complication, as previously suggested (5, 6). However, an apparent short luteal phase was present in approximately 18% ofhMG and GnRH-a cycles, despite the use of natural P for luteal support (5). Segal and Casper (5), have suggested that down regulation of pituitary GnRH receptors and reduced LH support for the corpus luteum (CL) may occur, even after a single administration of GnRH -a. Alternatively, inhibitory effects of GnRH -a on luteal P production have been reported (9-12), suggesting a direct detrimental, luteolytic effect of GnRH-a (6). We attempted, therefore, to trigger final follicular maturation in PCOS patients undergoing hMG superovulation with high late follicular E 2 levels with an intravenous GnRH bolus instead of hCG GnRH -a in hope of preventing luteal dysfunction without dangerously increasing the risk of OHSS.
hCG administration was witheld and an IV bolus of 200 J.Lg GnRH (Relisorm; Serono, Aubonne, Switzerland) was administered for triggering final follicular maturation and ovulation, instead of 10,000 IU hCG, usually injected IM for this purpose to our patients when late follicular concentration is <1,600 pg/mL (6,000 pmoljL) (13). Ovulation was triggered by 10,000 IU hCG when the late follicular E 2 concentration was between 300 and 1,600 pgjmL (1,100 to 6,000 pmol/L) ir the presence of one to five follicles with a mean diameter > 16 rom, measured by transvaginal ultrasound (13-15). However, when these sonographic criteria were met but the late follicular E 2 levels were >1,600 pg/mL (6,000 pmoljL) ovulation was triggered with native GnRH instead of hCG. In all these cycles the number of preovulatory follicles was between 8 and 25. Serum E 2 and P concentrations were monitored every 3 to 7 days throughout the luteal phase. In cycles where E 2 concentration decreased to ::::;;1,360 pgjmL (::::;;5,000 pmoljL), 2,500 IU IM hCG was administered once or twice at 3-day intervals, for luteal support. In cases where the postcoital test (PCT) was normal, the couples were advised to have intercourse on the day of GnRH administration and on the two following days. In cases where the PCT was negative, lUI with washed husband spermatozoa, after swim-up preparation using Ham's F-10 medium (Biological Industries, Beth Haemek, Israel) was accomplished, as previously described (14). The criteria for definition of OHSS, which we employed in this study, were published previously (1, 4).
RESULTS
MATERIALS AND METHODS
High serum E 2 concentration, from 1,600 pg/mL (6,000 pmoljL) to over 3,600 pg/mL (13,000 pmolj L) was measured in 44 ovulation induction cycles in 32 PCOS patients using hMG (Pergonal; Teva, Petah Tikva, Israel) or FSH (Metrodin; Teva), two ampules per day starting from the 3rd menstrual day, as previously described (13). In brief, ovarian response to ovulation induction was monitored by measurements of serum E 2 and P concentrations by RIA and by measurement of the number of developing follicles and their diameter by transvaginal ultrasound (US). The mean E 2 concentration before GnRH administration was 2,800 ± 681 (±SD) pg/ mL (10,279 ± 2,500 pmol/L). There were between 8 and 25 preovulatory follicles visualized by US. Because of the high risk ofOHSS in these cycles (1, 3), Vol. 62, No.3, September 1994
Ten pregnancies were generated by 44 hMG and GnRH co-treatment cycles (22. 7%) in these 32 PCOS patients (31.2%). Two of these pregnancies were multiple gestations (2/10, 20%), one was a twin pregnancy that ended in missed abortion at 10 weeks gestation, and the other, a triplet gestation, ended by cesarean section at 30 weeks gestation due to placental abrubtion. Two of the three premature newborns, weighing 1,250 to 1,455 g suffered early neonatal death, and one survived after 1 month in the neonatal intensive care unit. The pregnancy wastage was 20% (2/10), due to two spontaneous first trimester abortions. The remaining eight pregnancies ended in seven spontaneous vaginal deliveries and one cesarean section (due to pre-eclamptic toxemia), of eight healthy, term neonates.
Blumenfeld et al.
GnRH follicular maturation in hMG cycles
457
60
•
bMG/GnRH
121 bMG!bCG
[E2) nmol/L
PR/C
PR/P
MPR
PW
OHSS/P OHSS/C
Figure 1 The mean (±SD) late follicular E 2 concentrations [E nmol/L] in hMG and GnRH (•) or hMG and hCG (~) and the comparative incidence of pregnancy rate per cycle (PR/C), pregnancy rate per patient (PR/P), multiple pregnancy rate (MPR), pregnancy wastage (PW), and ovarian hyperstimulation rates per patient (OHSS/P) or per cycle (OHSS/C).
the conception cycles. In those cycles where the E 2 remained dangerously increased, > 1,360 pg/mL (>5,000 pmol/L) after the GnRH injection, no luteal hCG was administered and a daily IM injection of 50 mg Pin oil (Gestone; Paines & Byrne, Greenford, United Kingdom) for 14 days was administered instead. Severe OHSS necessitating hospitalization, parenteral fluid treatment, and abdominal paracentesis (of 4 to 6 L) of clear abdominal fluid (17) occurred in two cycles (4.5%). Both these patients conceived (one was delivered of a normal singleton and the other of premature triplets). Six patients experienced severe ovarian hyperstimulation in the previous hMG and hCG cycles, all of them in conception cycles. DISCUSSION
All but 1 of these 32 patients have been treated also by hMG and hCG ovulation induction cycles. Of the 69 previous hMG and hCG cycles in these 31 patients, 10 pregnancies were achieved (10/69, 14.5%) in 7 patients (7 /31, 22.5% ). The difference between the cycle fecundity [pregnancy rate (PR)] of the hMG and GnRH cycles (22. 7%) and that generated by the previous hMG and hCG cycles in the same patients (14.5%) was not statistically significant (P > 0.05, x2 test). Five of the 10 gestations (50%) achieved in the PCO patients by previous hMG and hCG treatments were multiple pregnancies: one heptuplet pregnancy ending in abortion, one sextuplet gestation spontaneously turning into a triplet gestation (16), three twin pregnancies, and five singletons. The pregnancy wastage rate was 30% (3/10) in these hMG- and hCG-generated gestations (Fig. 1). The mean E 2 serum concentrations on the day of IV GnRH bolus administration was 2,800 ± 681 pg/ mL (10,279 ± 2,500 pmol/L) (n = 44), decreasing to 1,057 ± 620 pg/mL (3,880 ± 2,275 pmol/L) (n = 40) within 3 to 5 days afterward. An IM injection of 2,500 IU hCG was administered once or twice, 3 and 6 days after the IV GnRH injection for luteal support, as previously described for hMG and hCG ovulation induction (16), in those cycles where the E 2 level decreased to <1,360 pg/mL (5,000 pmol/L). The midluteal E 2 (7 to 10 days after the IV GnRH bolus) was 2,297 ± 888 pg/mL (8,433 ± 3,260 pmol/ L) (n = 44). The mean midluteal E 2 concentration was 2,083 ± 882 pg/mL (7,647 ± 3,240 pmoljL) (n = 34) for those who did not conceive, compared with 2,976 ± 530 pg/mL (10,925 ± 1,950 pmol/L) in 458
Blumenfeld et a!.
The GnRH analogues have gained much popularity in the last decade for treatment of various indications including PCOS and as an adjunct to ovulation induction for IVF and assisted reproduction (18). Several recent studies (5-8) have documented that GnRH-a are an effective alternative to hCG for use in IVF cycles in which controlled ovarian hyperstimulation using hMG was attempted, especially in PCOS patients with high late-follicular levels of E 2 , who are more prone to develop the lifethreatening OHSS complication. However, a short luteal phase and apparently defective CL may be associated with the hMG and GnRH-a co-treatment, despite the use of P for luteal support (5). This shortcome may be expected in light of the reported inhibitory effect of even a single GnRH -a administration on the luteal P production, suggesting a direct detrimentalluteolytic effect ofGnRH-a (5, 6, 9-12). Native GnRH has been used successfully to trigger follicular maturation in patients pretreated with hMG for almost two decades, either in the form of nasal spray (19) or IM (20). However, an apparent discrepancy between the ovulations obtained and the lack of achieved pregnancies was observed associated with short luteal phase in most of the co-treatment cycles (20). We attempted, therefore, a modification of the previously reported hMG and GnRH -a or hMG and GnRH combinations in PCOS patients with dangerously increased high late-follicular E 2 levels :2:1,600 pg/mL (:2:6,000 pmol/L) by using an IV GnRH bolus and by supporting the luteal phase with one or two doses of2,500 IU hCG (16), every 3 days. We
GnRH follicular maturation in hMG cycles
Fertility and Sterility
hoped to prevent the previously reported high incidence ofluteal dysfunction (5) without dangerously increasing the risk of OHSS. Our results of 22.7% cycle fecundity and 31.2% PR (per patient) with a 20% (2/10) pregnancy wastage rate are at least as good as the respective parameters in the same group of PCOS patients in 69 previous hMG and hCG cycles (14.5% PR per cycle; 23% PR/patient; 30% pregnancy wastage, Fig. 1). However, the 20% multiple PR (2/10) in the hMG and GnRH cotreatment is <50% multiple gestations (5/10) in the previous hMG and hCG treatment cycles in the same group of patients. Although the difference is not statistically significant, probably due to the small numbers in each group, the difference is encouraging, and further experience (by increasing the group numbers) may achieve statistical significance. One may critically argue against the safety of supporting the luteal phase with one or two doses of 2,500 IU IM hCG. We attempted this luteal support, as previously described (16), when the E 2 concentrations decreased to :::;;1,360 pg/mL (5,000 pmol/L), and administered daily Pin oil instead in those cycles where E 2 remained persistently high. Not unexpectedly, the two cases of severe OHSS necessitating hospitalization and abdominal paracentesis occurred both in pregnant patients, one of whom received two luteal hCG injections, while the other did not. This preliminary experience suggests that OHSS is associated mainly with achievement of gestation and not necessarily with exogenous luteal hCG, versus P supplementation when no subsequent pregnancy occurs. Because luteal exogenous hCG administration may prevent the treatment associated luteolysis, the resulting cycle fecundity and PRs may be therefore higher, with a higher associated risk of subsequent OHSS. Because of the life-threatening danger of severe OHSS, we suggest to reserve the luteal hCG supplementation after hMG and GnRH co-treatment only to those cycles where E 2 concentrations decrease to <1,360 pg/mL (5,000 pmol/L) and to support the luteal phase with P whenever the E 2 levels remain higher. In the previous 69 hMG and hCG cycles in these PCO patients, the late follicular E 2 concentrations were significantly lower, :::;;1,600 pg/mL (:::;;6,000 pmol/L), and in spite of the fact that the PR was lower, the rate of multiple pregnancies was higher (50% versus 20%) and the rate ofOHSS necessitating hospitalization was higher (19.4% [6/31] versus 6.2% [2/ 32]). Vol. 62, No.3, September 1994
One would have anticipated that the patients in the study group with the extremely high late-follicular E 2 levels and a greater number of large- and medium-sized preovulatory follicles would have had a greater risk of severe ovarian hyperstimulation than the same PCO patients in previous hMG and hCG cycles, where the late follicular E 2 concentration was <1,600 pg/mL ( <6,000 pmoljL) (1-4). The results are surprisingly quite opposite: the OHSS was less common in the hMG and GnRH co-treatment than in the hMG and hCG cycles. Although not statistically different, because of the small number in each group, these results suggest encouragingly that triggering ovulation with GnRH and providing hCG (in low doses) for luteal support in cycles with dangerously high E 2 levels does not appear to increase the incidence of OHSS. These results are in keeping with the common experience of high risk of OHSS in the PCOS subgroup of patients (1-4). The 31% PR in our series is also as good as the 20% to 33% PR in the hMG and GnRHa combination reported by other investigators (57), with an apparently lower pregnancy wastage rate (20% versus 33% to 50%), suggesting a smaller incidence of luteal dysfunction. Recently, the popularity of the controlled ovarian stimulation protocols using GnRH -a during the follicular phase to prevent premature luteinization for IVF and assisted reproduction has increased (5). In these protocols, the induced gonadotroph desensitization to GnRH may make the triggering of follicular maturation with GnRH-a unpractical (5). However, Filicori et al. (21) have been successful in inducing follicular maturation, growth, and even pregnancies in PCOS patients previously treated with GnRH -a, by pulsatile intravenous administration of native GnRH. This previous experience may make the hMG and GnRH co-treatment possible, and an optional alternative to hCG in PCOS patients undergoing hMG ovarian stimulation after previous GnRH -a desensitization, in cycles where the late follicular E 2 is dangerously increased, ~1,600 pg/mL (~6,000 pmol/L). As we have hypothesized several years ago (16), the hCG luteal support may be superior to P administration due to the optional increase in the secretion of relaxin, inhibin, and possible other growth factors playing a crucial role in the establishment of gestation, by the CL in response to hCG and not toP. Recent data seem to apparently support this assumption (22). Prospective comparative studies between hMG and GnRH versus hMG and GnRH-a and between hMG and GnRH with hCG or P luteal support are
Blumenfeld et al.
GnRH follicular maturation in hMG cycles
459
needed to more accurately answer some of the still unanswered clinical questions. REFERENCES 1. Golan A, Ron-El R, Herman A, Soffer Y, W einraub Z, Caspi E. Ovarian hyperstimulation syndrome: an update review. Obstet Gynecol Surv 1989;44:430-8. 2. Kemmann E, Tavakou F, Schoen RM, Janes JR. Induction of ovulation with menotropins in women with polycystic ovary syndrome. Am J Obstet Gynecol 1981;141:58-64. 3. Blankstein J, Shalev J, Saadon T, Kukia EE, Rabinovici J, Pariente C, et a!. Ovarian hyperstimulation syndrome: prediction by number and size of preovulatory ovarian follicles. Fertil Steril 1987;4 7:597-602. 4. Pride SM, James CS, Ho Yuen B. The ovarian hyperstimulation syndrome. Sem Reprod Endocrinol1990;8:247-60. 5. Segal S, Casper RF. Gonadotropin-releasing hormone agonist versus human chorionic gonadotropin for triggering follicular maturation in in vitro fertilization. Fertil Steril 1992;57:1254-8. 6. ltskovitz J, Boldes R, Levron J, Erlik Y, Kahana L, Brandes JM. Induction of preovulatory luteinizing hormone surge and prevention of ovarian hyperstimulation syndrome by gonadotropin-releasing hormone agonist. Fertil Steril 1991;56:213-20. 7. Gonen Y, Balakier H, Powell W, Casper RF. Use of gonadotropin-releasing hormone agonist to trigger follicular maturation for in vitro fertilization. J Clin Endocrinol Metab 1990;71:918-22. 8. Imoedemhe DAG, Sigue AB, Pacpaco ELA, Olazo AB. Stimulation of endogenous surge of luteinizing hormone with gonadotropin-releasing hormone analog after ovarian stimulation for in vitro fertilization. Fertil Steril 1991;55:328-32. 9. Parinaud J, Beaur A, Bourreau E, Vieitez G, Pontonnier G. Effect of a luteinizing hormone-releasing hormone agonist (buserelin) on steroidogenesis of cultured human preovulatory granulosa cells. Fertil Steril1988;50:597-602. 10. Maruo T, Otani T, Mochizuki M. Antigonadotropic actions of GnRH agonist on ovarian cells in vivo and in vitro. J Steroid Biochem 1985;23:765-802. 11. Tureck RW, Mastroianni L, Blasco L, Strauss JF III. Inhibition of human granulosa cell progesterone secretion by a gonadotropin-releasing hormone agonist. J Clin Endocrinol Metab 1982;54:1078-85.
460
Blumenfeld et al.
12. Reddy PV, Azhar S, Menon KNJ. Multiple inhibitory actions of luteinizing hormone-releasing hormone agonist on luteinizing hormone/human chorionic gonadotropin receptor mediated ovarian responses. Endocrinology 1980; 107:930-6. 13. Blumenfeld Z, Barkey RJ, Youdim MBH, Brandes JM, Amit T. Growth hormone-binding protein regulation by estrogen, progesterone, and gonadotropins in human: The effect of ovulation induction with menopausal gonadotropins, GH and gestation. J Clin Endocrinol Metabol 1992; 75:1242-49. 14. Blumenfeld Z, Nahhas F. Pretreatment of sperm with human follicular fluid for borderline male infertility. Fertil Steril1989;51:863-8. 15. Blumenfeld Z, Nahhas F. Luteal dysfunction in ovulation induction: the role of repetitive human chorionic gonadotropin supplementation during the luteal phase. Fertil Steril 1988;50:403-7. 16. Blumenfeld Z, Dirnfeld M, Abramovici H, Amit A, Bronshtein M, Brandes JM. Spontaneous fetal reduction in multiple gestations assessed by transvaginal ultrasound. Br J Obstet Gynaecol 1992;99:333-7. 17. Thaler I, Yoffe N, Kaftory JK, Brandes JM. Treatment of ovarian hyperstimulation syndrome: the physiologic basis for a modified approach. Fertil Steril 1981;36:110-3. 18. Blumenfeld Z, Enat R, Brandes JM, Baruch Y. Gonadotropin-releasing hormone analogues for dysfunctional bleeding in women after liver transplantation: a new application. Fertil Steril1992;57:1121-3. 19. Potashnik G, Homburg R, Lunenfeld B. Ovulation induction in an amenorrheic woman by using GnRH in nasal spray. Harefuah 1975;88:453-5. 20. Crosignani PG, Trojsi L, Attanasio A, Tonani E, Donini P. Hormonal profiles in anovulatory patients treated with gonadotropins and synthetic luteinizing hormone releasing hormone. Obstet Gynecol 1975;46:15-22. 21. Filicori M, Flamigni C, Meriggiola MC, Ferrari P, Michelacci, Campaniello E, et a!. Endocrine response determines the clinical outcome of pulsatile gonadotropin-releasing hormone ovulation induction in different ovulatory disorders. J Clin Endocrinol Metab 1991;72:965-72. 22. Norman RJ, McLoughlin JW, Borthwick GM, Yohkaichiya T, Matthews CD, MacLennan AH, eta!. lnhibin and relaxin concentrations in early singleton, multiple, and failing pregnancy: relationship to gonadotropin and steroid profiles. Fertil Steril 1993;59: 130-7.
GnRH follicular maturation in hMG cycles
Fertility and Sterility