in vitro fertilization

FERTILITY AND STERILITY VOL 69, NO. 6, JUNE 1998 Copyright (#1998 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A. Elevated levels of basal estradiol-17,6 predict poor response in patients with normal basal levels of folliclestimulating hormone undergoing in vitro fertilization Johannes L. H. Evers, M.D., Peronneke Slaats, M.S., Jolande A. Land, M.D., John C. M. Dumoulin, Ph.D., and Gerard A. J. Dunselman, M.D. Division of Reproductive Endocrinology and Fertility, Department of Obstetrics and Gynecology, Academisch Ziekenhuis Maastricht and The University of Maastricht, Maastricht, The Netherlands Objective: To evaluate whether the predictive ability of a normal FSH level on cycle day 3 can be enhanced by levels of estradiol-17/3 (E2) on cycle day 3. Design: Prospective cohort study. Setting: University hospital-based, tertiary care infertility center. Patient(s): Two hundred thirty-one consecutively seen patients who attended the center for their first IVF attempt. Intervention(s): Blood samples were collected on day 3 of the cycle preceding IVF; IVF was performed in all patients. Main Outcome Measure(s): Patient's age, number of ampules of hmg, cancellation rate, number of oocytes, fertilization rate, and clinical pregnancy rate. Result(s): In patients with elevated FSH levels on cycle day 3, a low oocyte yield was achieved (7 versus 11) and a high number of ampules of hmg was necessary (56 versus 33). Their cancellation rate was high (67% versus 16%). In patients with normal basal FSH levels, high E 2 levels predicted a high cancellation rate (56%, versus 13% in patients with low E 2 levels) and a low oocyte yield (9, versus l 1 in patients with low E 2 levels). Patients with both normal FSH levels and low E 2 levels on cycle day 3 fared best. Conclusion(s): The basal E 2 level on cycle day 3 is a useful prognosticator of response to stimulation in IVF patients with normal basal FSH levels. (Fertil Steril 1998;69:1010-4. 9 by American Society for Reproductive Medicine.) Key Words: Poor response, IVF prognosis, basal estradiol-17/3 (E2), CD a Ea, basal FSH, CD 3 follicle-stimulating hormone (FSH) Received September 8, 1997; revised and accepted January 5, 1998. Reprint requests: Johannes L. H. Evers, M.D., Department of Obstetrics and Gynecology, Academisch Ziekenhuis Maastricht, P.O. Box 5800, NL 6202 AZ Maastricht, The Netherlands (FAX: 31-43-3874765). 0015-0282/98/$19.00 PII S0015-0282(98)00080-6 The basal FSH level has proved to be a reliable prognosticator of IVF outcome (1-3). Muasher et al. (1) concluded that on the third day of the menstrual cycle, FSH levels could differentiate populations with poor stimulation characteristics from those with good stimulation characteristics. Scott et al. (2) observed that if patients had an FSH level of ->15 miu/ml (Second International Reference Preparation [IRP] hmg standard) on cycle day 3, fewer pregnancies occurred, and if the FSH level was ->25 miu/ml, no pregnancies occurred. Toner et al. (3) determined that the FSH level on cycle day 3 was a better indicator of IVF outcome than age. Licciardi et al. (4) found that in cycles in which a GnRH agonist (GnRH-a) was not given, high levels of estradiol-17/3 (E2) on cycle day 3 were associated with low oocyte yields and low pregnancy rates (PRs). They obtained significantly fewer oocytes in patients with elevated E 2 levels (>60 pg/ml) on cycle day 3. No pregnancies occurred in patients with basal FSH levels of > 17 miu/ml if their basal E 2 level was >45 pg/ml. Smotrich et al. (5) confirmed that patients with elevated E 2 levels on cycle day 3 (in a cycle preceding IVF) had a higher cancellation rate and a lower PR, independent of FSH levels, than patients with low E 2 levels on cycle day 3. 1010

We performed this study to validate the findings of Licciardi et al. (4) in an unstimulated cycle preceding IVF. If the findings of these investigators were also proven valid for patients in whom blood samples were obtained in a spontaneous cycle before IVF, this would allow more appropriate selection and counseling of patients for IVF. We studied an independent set of patients but used the same cutoff levels for E 2 (60 pg/ml) and FSH (17 miu/ml) on cycle day 3. Our hypothesis was that on cycle day 3, FSH levels of >17 miu/ml (Second IRP standard; equal to > 10 miu/ml using the World Health Organization 78/549 calibration standard [6]) identify patients who respond poorly in IVF and that E 2 levels of >60 pg/ml on cycle day 3 help to differentiate between good and intermediate responders in a group with normal basal FSH levels (---17 miu/ml). MATERIALS AND METHODS The study population consisted of 231 consecutive IVF patients who had not previously undergone a stimulation cycle. Only first IVF stimulation cycles were included in this study. Patients who received donor oocytes or micromanipulated oocytes were excluded. At the time of this investigation, the license to perform IVF in our country depended on achieving an ongoing PR of --> 10% per cycle started. Therefore, a need existed for early (pretreatment) identification of potential poor responders. So far, only age (<40 years) and basal FSH level (<34 miu/ml) have been applied as inclusion criteria for admission into our IVF program. No further selection was performed in the patient group of this study. Our stimulation regimen has been published before (6). In short, all cycles were stimulated using hmg (Humegon; Organon, Oss, The Netherlands; or Pergonal; Serono, The Hague, The Netherlands) in combination with intranasal administration of a GnRH-a (Suprefact, buserelin, 300 /~g t.i.d.; Hoechst, Frankfurt, Germany; or Synarel, nafarelin, 400 ~g b.i.d.; Searle, Maarssen, the Netherlands). After suppression of ovarian activity with the GnRH-a for ->12 days (long protocol), hmg therapy was started. In their first cycle, all patients received a fixed-dose regimen of three ampules (225 IU) of hmg as a single daily injection. Follicular growth was assessed by transvaginal ultrasonography, with follicular diameter representing the arithmetic mean of three perpendicular dimensions. No hormone levels were monitored. According to National Health Service regulations, patients in The Netherlands are entitled to receive three IVF cycles free of charge. In our clinic, in first IVF attempts, the administration of hmg is discontinued if insufficient follicles develop (<4 follicles with a diameter of -> 16 mm after 12 days of stimulation; this was adopted as the definition of poor response in the present study). As soon as four or more follicles reached a mean diameter of -->18 mm each, hcg (5,000 IU, Pregnyl; Organon; or Profasi; Serono) was administered. At this time, hmg and GnRH-a were discontinued. Oocyte retrieval was performed under vaginal ultrasonographic guidance 34-36 hours after hcg administration. The transfer of two or sometimes three embryos to the uterus was scheduled for 40-70 hours after insemination. The luteal phase was supported by vaginal progesterone suppositories (2 100 mg t.i.d., Progestan, micronized progesterone; Organon). A clinical pregnancy was defined as ultrasonographic evidence of intrauterine fetal heart activity 5 weeks after ET. Levels of FSH and E 2 on cycle day 3 were determined in a spontaneous cycle preceding the cycle of stimulation. We recorded the patient's age, number of days of stimulation, total number of ampules administered, number of oocytes retrieved, cancellation rate (number of cycles abandoned divided by number of cycles started), fertilization rate (number of oocytes fertilized divided by number of oocytes retrieved), and clinical PR (number of pregnancies with fetal heart activity 5 weeks after ET divided by number of cycles started). We measured FSH with a commercially available RIA (Delfia; Wallac Oy, Turku, Finland). The sensitivity of the FSH assay was 0.085 miu/ml, calibrated to the Second IRP standard. The intra-assay and interassay coefficients of variation were <4%. A special problem occurs in comparing the results of FSH determinations because of the use of different reference preparations to calibrate the respective assays. Our cutoff level of 10 miu/ml of FSH, calibrated to World Health Organization standard 78/549 and widely used in Europe, equals a level of 17 mlu/ml if the assay is calibrated against the Second IRP, widely used in the United States (6). In the present study, FSH values are expressed as calibrated to the Second IRP standard. Estradiol levels also were measured by a commercially available double-antibody RIA (Coat-A-Count; Diagnostic Products Corporation, Los Angeles, CA). The sensitivity of the E 2 assay was 9.5 pg/ml. The intra-assay and interassay coefficients of variation were < 14%. Statistical analysis was performed with use of the Student's t-test and the X 2 test or Fisher's exact probability test. P<0.05 was defined as statistical significance. Data are expressed as means _+ standard deviations, medians and ranges, or means and 95% confidence interval (CI), whichever is appropriate. Diagnostic test evaluation included risk ratio and likelihood ratio estimations for positive and negative test results, considering FSH and E 2 determinations as independent diagnostic tests. Written informed consent was obtained from each patient for the anonymous use of data from their IVF cycles. Institutional Review Board approval is not required for chart reviews in The Netherlands. FERTILITY & STERILITY 1011

Stimulation and IVF outcome for all 231 patients according to FSH concentrations on cycle day 3. Group 1: Group 2: FSH <--17 miu/ml FSH >17 mlu/ml Variable (n = 213) (n = 18) P value Mean (_+SD) patient age (y) Mean percentage of poor responders (95% CI)* Mean ( total no. of ampules Mean ( no. of oocytes retrieved Mean (_+SD) fertilization rate (%)? Mean percentage of pregnancies/opu (95% CI):~ Mean clinical PR (%) (95% CI)w 33 _+ 4 35 + 4 <0.005 16 (11-21) 67 (41-87) <0.00002 33-+ 6 56 24 <0.05 11 5 7 4 <0.05 63 _+ 30 61 _+ 24 NS 20 (14-26) 0 (046) NS 15 (10-22) 0 (0-19) NS Note: NS = not significant. * Fewer than four follicles with a diameter of -->16 mm after 12 days of stimulation.? Fertilization rate = number of oocytes with two pronnclei divided by number of oocytes retrieved, multiplied by 100%. Pregnancies/OPU = number of pregnancies with fetal heart activity 5 weeks after ET divided by number of ovum pickup procedures. w Clinical PR = number of pregnancies with fetal heart activity 5 weeks after ET divided by number of cycles started, multiplied by 100%. RESULTS Table 1 shows the stimulation and cycle outcome data according to FSH levels on cycle day 3. In all, 213 patients (group 1) had FSH levels of -<17 miu/ml (median, 10.1; range, 3.4-17.0 miu/ml), and 18 patients (group 2) had FSH levels of > 17 miu/ml (median, 20.7; range, 17.2-33.5 miu/ml). Patients with FSH levels of -----34 miu/ml on cycle day 3 were not accepted into the program. In group 1 (normal FSH levels), 35 cycles were canceled because of poor response (16%) and 32 patients became pregnant (15%). In group 2 (high FSH levels), 12 cycles were canceled (67%) and no patient became pregnant. The mean patient age differed significantly (P<0.005) between the groups. Patients in group 1 (normal FSH levels) needed significantly less stimulation (lower total ampules of hmg), had significantly more oocytes retrieved, and had significantly fewer cycles canceled than group 2 patients (high FSH levels). To study the potential contribution of the E 2 level on cycle day 3 to the prediction of IVF outcome, we subdivided group 1 patients (normal FSH levels) according to their E 2 level on cycle day 3: group 1A (low E 2 level, i.e., -<60 pg/ml) and group 1B (high E 2 level, i.e., >60 pg/ml). The stimulation and outcome data for these patients are shown in Table 2. The 197 patients with a low E 2 level (group 1A) had a median E 2 level of 30 pg/ml (range, <9.5-60 pg/ml), with 26 cancellations (13%) and 32 pregnancies (16%). Group 1B (high E 2 level) consisted of 16 patients with a median E 2 level of 74 pg/ml (range, 63-218 pg/ml), with 9 cancellations (56%) and no pregnancies. The mean patient age did not differ between these groups. Group 1A (normal FSH levels, low E 2 levels) had significantly more oocytes retrieved and significantly fewer cycles canceled than group 1B (normal FSH levels, high E 2 levels) (Table 2). The differences in total number of ampules administered, fertilization rates, and PRs did not reach statistical significance. Because no pregnancies occurred in the 18 patients in group 2 (high FSH levels), no further breakdown according to their basal E 2 levels was performed. All patients in this group had E 2 levels of --<60 pg/ml. Considering FSH and E 2 determinations as independent tests for the prediction of poor response in an IVF treatment cycle, patients with FSH levels of >17 miu/ml were 4.1 times more likely (relative risk) to have their first IVF treatment cycle abandoned (95% CI, 2.4-6.1) than patients with FSH levels of -< 17 miu/ml. The likelihood ratio of an abnormal test was 7.8 (95% CI, 3.2-19.1), indicating that an FSH level of >17 miu/ml was 7.8 times more likely to be found in a patient who was going to have treatment abandoned in her first cycle than in a patient who would proceed to oocyte retrieval. For patients with normal FSH levels (-<17 miu/ml), the risk that their first treatment cycle would be abandoned if their basal E 2 level was >60 pg/ml was 4.3 times (95% CI, 2.3-7.1) the risk of patients with normal FSH levels and basal E 2 levels of --<60 pg/ml. The likelihood ratio of an abnormal test for the basal E 2 determination as a diagnostic test was 6.5 (95% CI, 2.6-15.8), indicating that an E 2 level of >60 pg/ml was 6.5 times more likely to be found in a patient who was going to have treatment abandoned in her first cycle than in a patient proceeding to oocyte retrieval. In clinical medicine, a likelihood ratio of an abnormal test of >5.0 is considered to reflect a useful diagnostic test. In the population studied, with an overall prevalence of 20% for poor response (<4 follicles with a diameter of >--16 mm after 12 days of stimulation), high FSH levels correctly predicted poor response to stimulation in 67% of the patients. High E 2 levels correctly predicted poor response in 56% of those patients who had normal FSH levels. 1012 Evers et al. Basal E 2 and response in IVF patients Vol. 69, No. 6, June 1998

Stimulation and IVF outcome accordin9 to E z concentrations on cycle day 3 for those 2]3 patients (StOUp ]) who had FSH levels of _<]7 mlu/mt_ on cycle day 3. Group IA: Group 1B: E 2 --<60 pg/ml E 2 >60 pg/ml Variable (n = 197) (n = 16) P value Mean (_+SD) patient age (y) Mean percentage of poor responders (95% CI)* Mean (_+SD) total no. of ampules Mean (+SD) no. of oocytes retrieved Mean (-+SD) fertilization rate (%)? Mean percentage of pregnancies/opu (95% CI):~ Mean clinical PR (%) (95% CI)w 33 +_ 4 33 + 4 NS 13 (8-18) 56 (30-80) <0.0002 33 +_ 5 36 -+ 17 NS 11 -- 5 9-+ 2 <0.05 63-+ 31 52_+ 30 NS 20 (14-27) 0 (0-52) NS 16 (11-21) 0 (0-21) NS Note: NS = not significant. Group 1 patients are described in Table 1. * Fewer than four follicles with a diameter of ->16 mm after 12 days of stimulation. 1- Fertilization rate = number of oocytes with two pronuclei divided by number of oocytes retrieved, multiplied by 100%. :~ Pregnancies/OPU = number of pregnancies with fetal heart activity 5 weeks after ET divided by number of ovum pickup procedures. w Clinical PR = number of pregnancies with fetal heart activity 5 weeks after ET divided by number of cycles started, multiplied by 100%. DISCUSSION In the present study, we confirmed that the FSH level on cycle day 3 in a spontaneous cycle preceding IVF is a good prognosticator of a patient's response to stimulation in a subsequent IVF cycle, in which hmg stimulation is performed during GnRH-a down-regulation (long protocol). The group with a high FSH level on cycle day 3 had an increased need for hmg (56 ampules, compared with 33 in patients with a normal FSH level), a poor response to stimulation (67% cancellations versus 16%), and a diminished oocyte yield (7 versus 11 per oocyte pickup procedure). Within the group with a normal FSH level on cycle day 3, those patients who showed a high E 2 level had a poor response to stimulation (56% cancellations) and an intermediate oocyte yield (9 per oocyte pickup procedure). The group with both a normal FSH level and a low E 2 level on cycle day 3 showed an appropriate response to stimulation (13% cancellations) and a good yield (11 oocytes). As women progress through reproductive life, ovarian reserves decline. This is reflected in increasing basal FSH levels in the early follicular phase of the menstrual cycle (7). The pituitary gland will increase FSH output in an effort to maintain a normal follicular response. In parallel to this increased demand for endogenous FSH, the response to exogenous FSH during (multiple) ovulation induction treatment is impaired (6). Estradiol, a steroid hormone produced by granulosa cells, and inhibin, a peptide hormone also of granulosa cell origin, have been implicated in the negative feedback relation between the developing follicle and the pituitary, although the role of inhibin awaits more detailed research (7). Seifer et al. (8) showed that women with low serum inhibin-b levels on cycle day 3 responded poorly to ovarian stimulation and that the outcome of assisted reproduction was poor for these women. Hughes et al. (9) found an age-related reduction of inhibin response, but not E 2 response, to ovarian stimulation. This may indicate that pituitary FSH release in women of advancing reproductive age is less restrained than in young women by factors from the granulosa cell-oocyte complex (7). In these women, loss of inhibin-mediated feedback inhibition of the pituitary will result in increased FSH output and a secondary increase in E 2 production by the ovaries. This in turn will suppress FSH production and release. A new balance between pituitary stimulus and ovarian response will result, reflected in normal FSH levels at the expense of increased circulating E 2. Apparently this new balance is temporary. Further progression toward menopause and aging of the ovary will be accompanied by decreased E 2 levels in the early follicular phase and a concomitant rise in FSH levels, as reflected in our group of patients with elevated FSH levels, all of whom had low E 2 levels. The findings of the present study confirm, in an independent group of patients before the start of IVF, earlier findings by Licciardi et al. (4) that FSH levels of >17 mlu/ml on cycle day 3 predict an impaired ovarian response in patients with a regular menstrual cycle and that within the group with normal FSH levels, E 2 levels may help predict the response to stimulation. In this group, E 2 levels of >60 pg/ml indicated a higher chance of poor response and a lower oocyte yield after ovum pickup. Patients with high FSH levels (> 17 mlu/ml) should be counseled that their risk of having a poor response in their first treatment cycle is 4.1 times the risk in patients with normal FSH levels and that if their basal FSH level is normal but they have increased basal E 2 levels, their risk of poor response is 4.3 times the risk of patients with both normal FSH levels and low E 2 levels. FERTILITY & STERILITY 1013

In conclusion, an elevated FSH level on cycle day 3 is a useful prognosticator of poor response to stimulation in prospective IVF patients. Estradiol levels on cycle day 3 are of added value for counseling patients with normal basal FSH levels or for explaining an unexpectedly poor outcome in a patient with a normal basal FSH level who has already undergone an IVF attempt. Unfortunately, patients who re- spond poorly to a stimulation dose of three ampules of hmg per day do not fare better in a subsequent stimulation cycle if the daily dose is increased to six ampules (6). References 1. Muasher SJ, Oehninger S, Simonetti S, Matta J, Ellis LM, Liu HC, et al. The value of basal and/or stimulated serum gonadotropin levels in prediction of stimulation response and in vitro fertilization outcome. Fertil Steril 1988;50:298-307. 2. Scott RT, Toner JP, Muasher SJ, Oehninger S, Robinson S, Rosenwaks Z. Follicle-stimulating hormone levels on cycle day 3 are predictive of in vitro fertilization outcome. Fertil Steril 1989;5l:651-4. 3. Toner JP, Philput CB, Jones GS, Muasher SJ. Basal follicle-stimulating hormone level is a better predictor of in vitro fertilization performance than age. Fertil Steril 1991;55:784-91. 4. Licciardi FL, Liu HC, Rosenwaks Z. Day 3 estradiol serum concentrations as prognosticators of ovarian stimulation response and pregnancy outcome in patients undergoing in vitro fertilization. Fertil Steril 1995; 64:991-4. 5. Smotrich DB, Widra EA, Gindoff PR, Levy MJ, Hall JL, Stillman RJ. Prognostic value of day 3 estradiol on in vitro fertilization outcome. Ferfil Steril 1995;64:1136-40. 6. Land JA, Yarmolinskaya MI, Dumoulin JCM, Evers JLH. High-dose human menopausal gonadotropin stimulation in poor responders does not improve in vitro fertilization outcome. Fertil Steril 1996;65: 961-5. 7. Lenton EA, De Kretser DM, Woodward AJ, Robertson DM. Inhibin concentrations throughout the menstrual cycles of normal, infertile, and older women compared with those during spontaneous conception cycles. J Clin Endocrinol Metab 1991 ;73:1180-90. 8. Seifer DB, Gardiner AC, Lambert-Messerlian G. Day 3 serum inhibin-b is predictive of assisted reproductive technologies outcome. Fertil Steril 1997;67:110-4. 9. Hughes EG, Robertson DM, Handelsman DJ, Hayward S, Healy DL, de Kretser DM. Inhibin and estradiol responses to ovarian hyperstimulation: effects of age and predictive value for in vitro fertilization outcome. J Clin Endocrinol Metab 1990;70:358-64. 1014 Evers et al. Basal E 2 and response in IVF patients Vol. 69, No. 6, June 1998