Bulent Urman, M.D.* Margo R. Fluker, M.D. Basil Ho Yuen, M.B., Ch.B.t

FERTILITY AND STERILITY Copyright c 1992 The American Fertility Society Vol. 57, No.6, June 1992 Printed on acid-free paper in U.S.A. The outcome of in vitro fertilization and embryo transfer in women with polycystic ovary syndrome failing to conceive after ovulation induction with exogenous gonadotropins Bulent Urman, M.D.* Margo R. Fluker, M.D. Basil Ho Yuen, M.B., Ch.B.t Bettina G. Fleige-Zahradka, M.D. Christo G. Zouves, M.B., Ch.B. Young S. Moon, Ph.D. Division of Reproductive Endocrinology and Infertility, Department of Obstetrics and Gynecology, University of British Columbia, Vancouver, British Columbia, Canada Objective: To assess the outcome of in vitro fertilization and embryo transfer (IVF-ET) in women with refractory polycystic ovarian syndrome (PCOS). Design: Retrospective case series with an age-matched control group. Setting: Ovulation induction and IVF programs in a tertiary referral center. Patients and Interventions: Nine patients with PCOS who failed standard ovulation induction treatment (clomiphene citrate plus ~6 ovulatory human menopausal gonadotropin [hmg] cycles) underwent 19 cycles of IVF-ET. Forty age-matched tubal factor patients who completed 40 cycles of IVF-ET served as a control group. Outcome Measures: Demographic features and IVF -ET cycle characteristics were compared using Student's t-test and Fisher's exact test. Results: Cycles of IVF-ET in patients with PCOS were associated with higher estradiol levels (5,222 versus 4,009 pmoljl), lower hmg requirements (15.8 versus 19.6 vials), greater numbers of oocytes (7.6 versus 5.6), and lower fertilization rates (56% versus 75%) compared with tubal factor cycles (P < 0.05). However, the number of embryos transferred (3.9 versus 4.0) and the clinical pregnancy rate per embryo transfer (24% versus 25%) did not differ significantly between the two groups. Conclusion: These results suggest that conception failure after six or more ovulatory hmg cycles in patients with PCOS does not adversely affect subsequent IVF performance. Fertil SterilI992;57:1269-73 Key Words: Polycystic ovarian syndrome, gonadotropins, in vitro fertilization In vitro fertilization and embryo transfer (lvf ET) is an accepted form of treatment for polycystic ovarian syndrome (PCaS)-associated anovulatory infertility refractory to standard methods of ovu- Received April 16, 1991; revised and accepted February 11, 1992. * Current address: Department of Obstetrics and Gynecology, Hacettepe Universitesi, Kadin Hastaliklari ve Dogum Ana Bilim Dali, Sihhiye, Ankara, Turkey. t Reprint requests: Basil Ho Yuen, M.B., Ch.B., Department of Obstetrics and Gynecology, Grace Hospital, Room 2H30, 4490 Oak Street, Vancouver, British Columbia, Canada V6H 3V5. lation induction. Although IVF -ET in such patients yielded satisfactory pregnancy rates (PR) (1-3), the definition of failed standard ovulation induction therapy is unclear. This information is needed for critical evaluation of the efficacy of IVF -ET in pcas. An inadequate trial of prior ovulation induction therapy can artificially inflate the success ascribed to IVF-ET. In addition, data on the outcome of IVF -ET cycles after multiple failed exogenous gonadotropin cycles in patients with pcas is required for realistic counseling. Such information is currently lacking. In this study, we examined the results of IVF -ET in clomiphene citrate (CC)-resis- Vol. 57, No.6, June 1992 Urman et al. IVF-ET in PCOS refractory to CC and hmg 1269

tant patients with PCOS failing to conceive after at least six ovulatory cycles of exogenous gonadotropin treatment and compared their cycle outcome parameters with those of an age-matched group of pure tubal factor patients who would traditionally be considered good prognosis candidates for IVF. Patients MATERIALS AND METHODS Nine anovulatory infertile women with PCOS who underwent 19 cycles of IVF-ET after failing to conceive with standard ovulation induction therapy using CC and exogenous gonadotropins were studied. Forty age-matched pure tubal factor patients who completed 40 IVF-ET cycles using the same ovulation induction protocol served as a control group. The diagnosis of PCOS was established by the presence of hyperandrogenism (total testosterone > 2.43 nmol/l) and anovulation (monophasic basal body temperature charts and luteal progesterone [P] concentrations < 16 nmol/l). One patient with PCOS had a solitary patent fallopian tube, whereas the remainder demonstrated bilateral tubal patency at hysterosalpingography. Significant tubal or peritoneal factors in the PCOS group were ruled out by laparoscopy. The male partners of all patients with PCOS and tubal factor had normal semen analyses. Standard Ovulation Induction Cycles Ovulation induction was initiated with CC, 50 mg/d, for 5 days. Dexamethasone (0.25 mg nightly) was added to CC in unresponsive patients. Ovulation was documented by midluteal P concentrations > 16 nmol/l. Lack of ovulation despite adequate dose increases (up to 200 mg/d of CC for 5 days, n = 3) or lack of pregnancy after six or more ovulatory cycles (n = 6) were indications for exogenous gonadotropins. Human menopausal gonadotropin (hmg, Pergonal; Serono Laboratories Inc., Mississauga, Ontario, Canada) was administered in an individualized manner (75 to 450 IU/d) as previously described (4). All patients with PCOS underwent 6 or 7 ovulatory cycles of exogenous gonadotropin therapy. Human menopausal gonadotropin cycles were monitored using daily serum estradiol (E2) assays and periodic ultrasonographic assessment of follicular growth with an abdominal or vaginal sector scanner. Human chorionic gonadotropin (hcg), 10,000 U (Profasi; Serono), was administered once E2 levels reached 2,000 to 7,000 pmol/l (550 to 1,900 pg/ml) in the presence of at least one leading follicle ~ 14 mm in diameter. Ten cycles were performed before the widespread availability of follicular ultrasonography and were monitored with E2 levels alone. Luteal phase P concentrations were monitored 3,7, and 10 days after the hcg injection. Cycles of IVF-ET Ovulation induction in patients with PCOS was achieved using one of two protocols: (1) hmg (75 to 450 IU) from day 3 onward (n = 16) or (2) buserelin acetate (Suprefact; Hoechst, Montreal, Quebec, Canada) 500 Jlg subcutaneously from day 21 of the previous cycle plus hmg (75 to 450 IU) once down regulation was achieved (E2 < 110 pmol/l in the absence of follicles > 10 mm in size). One patient with PCOS received two cycles of combined hmg/ buserelin acetate after cancellation of her previous IVF cycle because of a premature luteinizing hormone (LH) surge. All patients in the tubal factor group were stimulated with straight hmg as described above. The dose of hmg was individually adjusted according to daily E2 levels and follicular ultrasound (US) measurements. Human chorionic gonadotropin was administered intramuscularly as a single dose of 10,000 U when there were at least two follicles ~ 14 mm in diameter and an E2 concentration ~ 2,200 pmol/l. Human chorionic gonadotropin was administered 30 to 32 hours (n = 10) or 54 to 58 hours (n = 47) after the last dose of hmg. Laparoscopic (n = 45) or transvaginal (n = 12) oocyte recovery was scheduled for 36 hours after hcg. Oocyte retrievals performed before the advent of cryopreservation (n = 11 PCOS; n = 40 tubal) were terminated after eight mature oocytes were obtained. In procedures performed after the establishment of the cryopreservation program (n = 6 PCOS), all accessible follicles were aspirated with the intent of collecting as many oocytes as possible. In vitro fertilization was achieved and ETs were performed using previously described standard techniques (5). The luteal phase was supported with P vaginal suppositories that were continued until 10 weeks' gestation in the event of pregnancy. Clinical pregnancies included those demonstrated by US at 6 to 7 weeks' gestation, plus histologically confirmed spontaneous abortions and ectopic gestations. Estradiol and P were measured by radioimmunoassay (Coat-A-Count; Diagnostic Products Corporation, Los Angeles, CA). The within- and be- 1270 Urman et al. IVF-ET in pcas refractory to CC and hmg Fertility and Sterility

Table 1 Characteristics of Previous HMG Cycles in Patients With PCOS Patients HMG cycles Ovulatory cycles No. of ampules of hmg Length of treatment (d) Final E2 (pmoljl) Largest follicle (mm) Follicles> 10 mm Peak P (nmoljl) Characteristic 9 58 55 (95)' 15.3 ± 6.9t 7.1 ± 1.6 4,006 ± 1,289 17.9 ± 5.3 6.3 ± 4.4 157 ± 82 Table 2 Demographic and Ovulation Induction Characteristics in PCOS Versus Tubal Factor Patients Undergoing IVF PCOS Tubal factor Patients 9 40 Total cycles 19 40 Canceled cycles 2 (11)' 0 Age (y) 32.6 ± 3.2t 32.7 ± 2.1 Length of infertility (y) 5.2 ± 2.4 6.1 ± 3.9 Ampules of hmg 15.8 ± 4.9 19.6 ± 6.9 Peak E, (pmol/l) 5,222 ± 2,538 4,009 ± 1,548 Severe OHSS 2 (11) 0 Significance NSt NS P = 0.036 P = 0.027 NS, Value in parentheses is percent. t Values are means ± SD per ovulatory cycle. tween-assay coefficients of variation were 4% and 5%, respectively. Statistical analyses of the results were computed using unpaired Student's t-tests, X2 analyses, and Fisher's exact test. All data were expressed as means ± SD. RESULTS Table 1 summarizes the characteristics of 58 previous hmg cycles in the nine patients with PCOS. The results of their subsequent IVF cycles and those of the tubal factor comparison group are summarized in Tables 2 and 3. Cycles of PCOS were associated with higher peak E2 levels on the day of hcg and lower hmg requirements. Although more oocytes were retrieved in the PCOS group, an equivalent number of embryos were transferred in the two groups. This conclusion remains unaltered by exclusion of the 6 PCOS cycles in which cryopreservation was available. There was one case of failed fertilization in each group, plus one patient with PCOS who had only polyspermic fertilization in three of six oocytes. The mean percentage of oocytes fertilized was significantly lower in the PCOS group; this difference persisted after excluding the case of failed fertilization from each group. Reduced fertilization rates also persisted among patients with PCOS, regardless of whether all accessible oocytes or only a limited number were retrieved. The clinical PR per ET (24% versus 25%) and the spontaneous abortion rate (25% versus 10%) did not differ significantly between groups. There was a nonsignificant trend toward more cases of ovarian hyperstimulation syndrome (6) within the PCOS group. The development of ascites necessitating conservative in-hospital management occurred in two (11 %) PCOS cycles. Peak E2 levels, Both cycles canceled because of a premature LH surge. Values in parentheses are percents. t Values are means ± SD. t NS, not significant. Ovarian hyperstimulation syndrome with ascites. on the day of hcg were 5,644 and 9,644 pmoljl, respectively. In both instances, ovarian US showed >20 follicles < 10 mm and> 10 follicles :2: 10 mm in diameter. Both cycles were initiated before the introduction of cryopreservation, and all accessible follicles had not been aspirated. In one cycle, twin sacs with fetal cardiac activity were noted at 6 weeks' gestation followed by spontaneous reduction to a singleton pregnancy by 9 weeks' gestation. DISCUSSION Exogenous gonadotropins are the treatment of choice in patients with PCOS who fail to ovulate or conceive after first line therapy with CC, alone or in combination with glucocorticoids. In this center, hmg therapy in hyperandrogenic anovulatory pa- Table 3 In Vitro Fertilization Cycle Characteristics and Outcome in PCOS Versus Tubal Factor Patients Tubal PCOS factor Significance Oocytes retrieved 7.6 ± 1.8' 5.6 ± 1.6 P < 0.001 Fertilization (%) 55 ± 26t 75 ± 20t P = 0.006 Embryos transferred 3.9 ± 1.9 4.0 ± 1.6 NS:j: Clinical pregnancies 4/17 (24) II 10/40 (25)~ NS Deliveries 3/17 (18) 8/40 (20)1J NS Spontaneous abortions 1/4 (25) 1/10 (10) NS Ectopics 0 1/10 (10) NS, Values are means ± SD. t Including one case of failed fertilization in each group. :j: NS, not significant. Per ET. II Including one twin pregnancy that aborted and 1 twin pregnancy in which one sac aborted spontaneously. Values in parentheses are percents. ~ Including two twin and one triplet pregnancies. Vol. 57, No.6, June 1992 Urman et al. IVF-ET in pcas refractory to CC and hmg 1271

tients with PCOS yields a fecundity rate of 9.5% per ovulatory cycle and a cumulative conception rate of 42% over six treatment cycles (Ho Yuen, B, unpublished data). In the current study, only patients failing to conceive after six ovulatory cycles of hmg therapy were offered IVF -ET. We believe this represents an adequate trial of standard ovulation induction therapy before embarking on IVF -ET because the majority of the patients destined to conceive on hmg therapy will do so during the initial six treatment cycles (7). Although the success of IVF-ET in treating infertile women with anovulation because of PCOS has been previously reported (1-3), the number and nature of standard ovulation induction cycles have been variably defined. In such a setting, women who conceived with IVF may also have conceived before IVF if standard ovulation induction therapy had been given an adequate trial. This possibility was minimized in the current study by the provision of at least six ovulatory hmg cycles before IVF. Compared with their IVF cycles, the prior hmg cycles were associated with lower peak E2 levels, reflecting a more conservative approach to ovulation induction outside an IVF program. However, the proportion of ovulatory cycles (95%), mean preovulatory E2 values, and peak luteal phase P levels underscore the adequacy of the previous ovulation induction attempts. Ashkenazi and co-workers (1) reported six pregnancies (33%) after a single IVF cycle in 18 women with refractory PCOS. The 18 subjects had failed to conceive during 6 to 12 previous ovulation induction cycles; however, the proportion of cycles involving exogenous gonadotropin therapy was not defined. Although they were randomly assigned to one of the three ovarian stimulation protocols during IVF-ET, the small sample size precluded adequate evaluation of protocol efficacy. In contrast, all but 2 of the cycles in the current study used the same hmg protocol. None of the cycles reported by Ashkenazi et al. (1) were complicated by ovarian hyperstimulation syndrome, a finding that was attributed to the aspiration of all accessible follicles and cystic structures. The two cases of severe ovarian hyperstimulation syndrome noted in the current study occurred before the introduction of a cryopreservation program, so all accessible follicles had not been aspirated. Salat-Baroux and co-workers also treated 16 infertile patients with refractory PCOS using IVF -ET (2). However, 11 of their 16 patients (69%) also had absolute tubal infertility, so they had not received a prior trial of standard ovulation induction therapy. The remainder underwent at least four previous unsuccessful attempts at ovulation induction with either CC or exogenous gonadotropins. The fertilization rate, number of embryos replaced, and PRs did not differ between the patients with PCOS and a control group of pure tubal factor patients. One severe ovarian hyperstimulation syndrome was recorded in the PCOS group. In the current study, among a carefully selected group of nine patients with PCOS resistant to standard ovulation induction therapy, including 6 ovulatory cycles of hmg, 19 cycles of IVF-ET yielded a clinical PR of 24% per ET. This rate was almost identical to that achieved in a control group of tubal factor patients. Although there was no difference in the spontaneous abortion rates, statistical significance may have been achieved with a larger sample size. The patients with PC OS required less hmg to develop higher peak E2 levels and were more prone to the development of ovarian hyperstimulation syndrome. Although more oocytes were retrieved in the PCOS group, both before and after the institution of a cryopreservation program, true differences between the two groups could not be accurately assessed because of arbitrary limitations on number of oocytes retrieved before the introduction of cryopreservation. Failure to conceive despite six previous cycles of superovulation with gonadotropins in patients with PCOS raises the possibility of other undiagnosed infertility factors such as defective tubal transport or in vivo fertilization failure. Although it was not possible to adequately assess the former factor, the ability of IVF to bypass tubal transport may have been responsible for some of the PCOS pregnancies that followed failed ovulation induction. In addition, the lower fertilization rates among the patients with PCOS in the current study suggest that disordered fertilization may have contributed to conception failure during ovulation induction therapy. Indeed, five of the nine patients with PCOS in the current study experienced ::;;50% fertilization during one or more IVF cycles. Only one patient in the PCOS group had complete failure of fertilization in vitro; however, this was not repeated in her other IVF cycle. Interestingly, another patient who had only polyspermic fertilization of her oocytes had a past history of molar pregnancy. Oocyte immaturity may also contribute to reduced fertilization but was unlikely to account for all of the observed reduction in fertilization because 86% of the oocytes in the PCOS group were morphologically matur(~. Factors such as hyperandrogenemia and tonically elevated 1272 Urman et al. IVF-ET in pcas refractory to CC and hmg Fertility and Sterility

LH levels may contribute to a reduction in oocyte quality and fertilization potential in patients with PCOS. The roles of excess androgens and LH in the pathogenesis of poor oocyte quality resulting from induced ovulation in PCOS were discussed elsewhere (4). Also pertinent to the current findings are the recent clinical observations (8) showing an association between high basal LH, infertility, and miscarriage. We cannot exclude the possibility of spontaneous conceptions among the patients with PCOS undergoing IVF -ET, particularly during the 11 cycles in which all accessible follicles had not been aspirated. However, we believe this possibility was minimized by sequential aspiration of the largest follicles until six to eight oocytes were obtained and by their history of failure to conceive in response to CC and in at least 6 ovulatory cycles of hmg. In summary, IVF-ET in patients with PCOS failing to conceive with standard methods of ovulation induction, including at least six ovulatory hmg cycles, resulted in PRs comparable with those of women with pure tubal factor infertility. The patients with PCOS had lower hmg requirements, higher E2 concentrations, and a greater tendency to develop ovarian hyperstimulation syndrome when stimulated for IVF-ET. In addition, their spontaneous abortion rates may have been higher than those women with tubal disease, although the current sample size was too small to adequately evaluate pregnancy wastage. Reduced fertilization was evident among the PC OS group and could be implicated in their failure to conceive during standard ovulation induction therapy. These data suggest that conception failure after six or more ovulatory hmg cycles does not adversely affect subsequent IVF outcome. Acknowledgments. We thank Mrs. Sandi Barrow and Mrs. Patti Paul for assistance in compiling and analyzing the data. REFERENCES 1. Ashkenazi J, Feldberg D, Dicker D, Yeshaya A, Ayalon D, Goldman JA. IVF-ET in women with refractory polycystic ovarian disease. Eur J Obstet Gynecol 1989;30:157-61. 2. Salat-Baroux J, Alvarez S, Antoine JM, Cornet D, Tibi C, Plachot M, et al. Results of IVF in the treatment of polycystic ovary disease. Hum Reprod 1988;3:331-5. 3. Salat-Baroux J, Alvarez S, Antoine JM, Tibi C, Cornet D, Mandelbaum J, et al. Comparison between long and short protocols of LHRH agonist in the treatment of polycystic ovary disease by in-vitro fertilization. Hum Reprod 1988;3: 535-9. 4. Ho Yuen B, Pride SM. Induction of ovulation with exogenous gonadotropins in anovulatory infertile women. Sem Reprod Endocrinol 1990;8: 186-97. 5. Ho Yuen B, Pride SM, Rowe TC, Moon YS, McComb PF, Poland BJ, et al. Comparison of the outcome of ovulation induction therapy in an in vitro fertilization program employing a low-dose and an individually adjusted high-dose schedule of human menopausal gonadotropins. Am J Obstet Gynecol 1985;151:172-5. 6. Pride SM, James CS, Ho Yuen B. The ovarian hyperstimulation syndrome. Sem Reprod EndocrinoI1990;8:247-60. 7. Dor J, Itzkowich DJ, Mashiach S, Lunenfeld B, Serr DM. Cumulative conception rates following gonadotropin therapy. Am J Obstet Gynecol 1980;136:102-5. 8. Regan L, Owen EJ, Jacobs HS. Hypersecretion ofiuteinising hormone, infertility, and miscarriage. Lancet 1990;336: 1141-4. Vol. 57, No.6, June 1992 Urman et al. IVF-ET in pcas refractory to CC and hmg 1273