Intercycle variability of day 3 follicle-stimulating hormone levels and its effect on stimulation quality in in vitro fertilization*

Transcription

1 FERTILITY AND STERILITY Copyright C> 1990 The American Fertility Society Printed on acid-free paper in U.S.A. Inter variability of day 3 follicle-stimulating hormone levels and its effect on stimulation quality in in vitro fertilization* Richard T. Scott, Jr., M.D.t Glen E. Hofmann, M.D., Ph.D.:!: Sergio Oehninger, M.D. Suheil J. Muasher, M.D. Jones Institute for Reproductive Medicine, Eastern Virginia Medical Sclwol, Norfolk, Virginia Prior studies have demonstrated that gonadotropin stimulation quality and pregnancy rates are better in in vitro fertilization (IVF) patients with low basal day 3 folliclestimulating hormone (FSH) levels. The records of 81 patients who had undergone three or more IVF attempts during a 2-year period were studied to determine the degree and potential impact of inter variability in basal FSH concentrations. The mean of the individual standard deviations for all 81 patients was 4.2 ± 0.4 miu/ml. However, the patients with a mean basal FSH of <15 miu/ml had a mean deviation of only 2.6 ± 0.2 miu/ml, whereas those with a mean basal FSH of 2:15 miu/ml had a mean deviation of 7.3 ± 0.7 miu/ml. Inter variability in basal FSH values did not predict changes in ovarian response to gonadotropin stimulation and thus may not be used to select an optimal in which to stimulate an individual patient. Furthermore, patients with large inter variation responded poorly to gonadotropin stimulation independent of their basal FSH concentration. This information allows more precise counseling of patients regarding their appropriateness for assisted reproduction. Fertil Steril54:297, 1990 Basal day 3 gonadotropin levels have been shown to be predictive of stimulation quality and pregnancy rates in patients undergoing in vitro fertilization (IVF).I,2 Patients with basal follicle-stimulating hormone (FSH) levels on day 3 of <15 miu /ml have ongoing pregnancy rates twice as high as those with basal values of 15 to 24.9 miu /ml and 6 times higher than those patients with concentrations ~ 25 miu/ml.2 Since some patients with elevated basal values do conceive, Received May 23, 1989; revised and accepted April 11, * Presented at the 44th Annual Meeting of The American Fertility Society, Atlanta, Georgia, October 10 to 13, 1988_ t Reprint requests and present address: Richard T. Scott, M.D., Division of Reproductive Endocrinology, Department Obstetrics and Gynecology, Wilford Hall USAF Medical Center, San Antonio, Texas :j: Division of Reproductive Endocrinology, Department of Obstetrics and Gynecology, Mt. Sinai School of Medicine, New York, New York basal gonadotropin values have not been used as exclusion criteria in the Norfolk program. Nevertheless, this information has been useful in counseling patients regarding their chances for achieving pregnancy through IVF and in determining optimal gonadotropin stimulation regimens. l,2 Patients who have had multiple IVF attempts have occasionally had wide fluctuations in their basal FSH levels. Currently, no information is available as to what degree of inter variation in basal day 3 FSH concentrations is normal. However, large fluctuations in gonadotropin levels have been observed in women with altered ovarian function such as perimenopausal women and women with autoimmune disorders that interfere with ovarian metabolic activity.3-7 These women frequently have only intermittent ovarian function, and the fluctuations in gonadotropin levels probably reflect the activity of the ovary at any single Scott et al. Inter variability in basal FSH levels 297

2 point in time. The altered gonadotropin secretion probably results from changes in the secretion of gonadal steroids and proteins and their subsequent feedback relationships with the hypothalamic-pituitary axis. Although the fluctuations in basal FSH concentrations seen in most IVF patients are less than those seen in women with intermittent ovarian function, it seems possible that these fluctuations could represent differences in ovarian activity, or more specifically, the activity of the developing cohort of follicles. The current literature provides no information regarding the correlation of inter variation in basal FSH concentrations with alterations in the quality of the developing cohort of follicles as gauged by response to gonadotropin stimulation or pregnancy rates. Specifically, the information and conclusions derived from cross-sectional population studies cannot be applied to differences in FSH concentrations that result from the inter variation of an individual patient. The purpose of this study was twofold. First, to determine the degree of inter variability in basal FSH concentrations in patients attempting pregnancy through IVF, and second, to determine if alterations in basal FSH concentrations are predictive of alterations in the ovarian response to gonadotropin stimulation. MATERIALS AND METHODS The records of all patients participating in the Norfolk IVF program over a continuous 24-month period (January 1987 through December 1988) were reviewed. All s where the patients were receiving gonadotropin-releasing hormone agonist suppression were excluded from the study. The records of all patients who had three or more IVF s during the study interval were selected for further review. Eighty-one women had three or more IVF attempts during the study interval. Their stimulations were reviewed, and the following information was recorded: (1) the date of initiation of gonadotropin stimulation for the IVF attempt; (2) the patient's age at the time of the IVF ; (3) the cause of infertility, including any history of ovulatory dysfunction; (4) the basal FSH concentration; (5) the basal estradiol (E2) concentration; (6) the peak E2 concentration before oocyte retrieval; (7) the quantity of gonadotropins administered during stimulation; (8) the duration of gonadotropin stimulation; (9) the number of follicles as- pirated; and (10) the number of preovulatory oocytes (metaphase I or II) recovered. All patients had undergone pure gonadotropin stimulations using established protocols.s Oocyte retrieval was performed either transvaginally or laparoscopically using previously published techniques. 9 Preincubation of oocytes, sperm preparation, concentration of sperm for insemination, culturing conditions and techniques, and general embryology laboratory procedures were employed as previously described. lo,11 Commercially available radioimmunoassay (RIA) kits were used to determine basal day 3 FSH concentrations (Leeco Diagnostics, Southfield, MI). The inter- and intra-assay coefficients of variation (CVs) were 6.2% and 4.8% for FSH, respectively. This creates a 95% confidence interval of ±3.3 mlu/ml (range of 6.6 miu/ml) for a concentration of 15 mlu /ml. Estradiol levels were also determined by a commercially available RIA (Pantex, Santa Monica, CA) and had interassay and intra-assay CVs of 5.8% and 5.7%, respectively. The mean and standard deviation of the basal FSH concentrations were calculated individually for each of the 81 patients. The range in basal FSH concentrations was also calculated for each of the 81 patients by subtracting their lowest from their highest basal value. The mean deviation of the FSH concentrations was defined as the mean ofthe individual standard deviations and the mean range as the mean of the individual ranges. The data in this study were then analyzed from three distinct perspectives. The first was based on the study population as a whole and involved calculation of a mean deviation and a mean range. A paired analysis of the s in which the patients had their highest and lowest basal FSH concentrations was also done to determine if inter variability in basal FSH values prognosticated differences in ovarian response to gonadotropin stimulation for the entire population. The second portion of the analysis involved dividing the patients into different categories based on their mean basal FSH concentrations and determining the mean deviation and the ranges of basal FSH concentrations. The three groups were defined as having mean basal FSH concentrations of <15,15 to 24.9, and ~25 miu/ml, since these values have previously been shown to be predictive of ovarian response to gonadotropin stimulation and to correlate with pregnancy rates in IVF patients.2 Finally, the third area of analysis looked specifically at the patients with the largest ranges in their 298 Scott et al. Inter variability in bm;al FSH levels Fertility and Sterility

3 :: '" 70 1:: '" OJ OJ Q.. 50 Q ~ 20 ~ n_ , ,20 FSH miu/ml FSH miu/ml Figure 1 Distribution of mean deviation in basal FSH concentrations (A) and ranges of basal FSH concentrations (B) for 81 IVF patients with three or more attempts within a 24-month interval. 80 basal FSH concentrations, i.e., those patients with the largest inter variability. A range of > 10 miu/ml was empirically defined as increased inter variability. Evaluations in this third group included determination of mean deviations and paired analyses of stimulation characteristics from the s in which they had their highest and lowest basal FSH values. The last comparison was done since the patients with the largest ranges in basal FSH concentrations might be the most likely to show inter differences in their responses to gonadotropin stimulations. Statistical analyses were performed using paired t-tests, two tailed t-tests, and x 2 analyses as indicated. Corrections for multiple comparisons were employed where appropriate, and a confidence limit of 5% or better was considered significant. The mean deviation and ranges in FSH concentrations and the characteristics are presented as means with associated standard errors. RESULTS Eighty-one patients had 3 or more IVF s during the 24-month study interval, resulting in 281 IVF study s. Fifty patients had three attempts, 25 patients had four attempts, 5 patients had five attempts, and 1 patient had six attempts. The average age ofthe patients for all attempts was 35.6 ± 0.3 years. The patients' diagnoses included tubal factor (n = 60), male factor (n = 28), endometriosis (n = 14), cervical factor (n = 6), and idiopathic (n = 8). Nine of the 81 patients had histories of prior ovulatory dysfunction requiring treatment with either clomiphene citrate (CC) or gonadotropins in the past. None of these 9 patients had noted any hypoestrogenic symptoms or were noted to have any evidence of premature ovarian failure. Results for the Entire Study PopUlation Evaluation of the study population as a whole revealed a mean standard deviation of basal FSH concentrations of 4.2 ± 0.4 miu /ml (range 0.8 to 18.2 miu/ml). The mean of the ranges of basal FSH values for the 81 patients was 10.1 ± 0.8 miu/ml (range 0.8 to 42 miu/ml). The distribution of the standard deviations and ranges are dem- 0nstrated in Figure 1. Comparisons via a paired data analysis of the s in which the patients had their highest and lowest basal FSH values revealed no differences in any of the parameters of stimulation quality (Table 1). Furthermore, there were no differences in the ages of the patients at the time of their high and low basal FSH s, i.e., the passage of time did not correlate with elevations in basal FSH concentrations during the study interval in these patients. Finally, no difference was detected in cancellation rates in the high and low s (low, 14/81; high, 14/81). Premature luteinizing hormone surges occurred in 14 s. The incidence was equivalent in the high (8/14) and low (6/14) s. Results by Mean Basal FSH Concentration The second portion of the evaluation was done by dividing the patients into three distinct groups based on their mean basal FSH concentrations. Ranges of <15, 15 to 24.9, and ~25 miu /ml were Table 1 Comparison of the Stimulation Characteristics from the Low and High Basal FSH Cycles for the Entire Study Population (n = 81)a Basal FSH (miu ImL) Basal E2 (pg/ml) Peak E2 (pg/ml) Ampules of gonadotropins Duration of stimulation (d) Follicles aspirated Preovulatory oocytes retrieved a Values are means ± SD. LowFSH 9.1 ± ± ± ± ± ± ± 0.2 HighFSH 19.2 ± ± ± ± ± ± ± 0.2 Scott et al. Inter variability in basal FSH levels 299

4 Table 2 Comparison of the Stimulation Characteristics From the Low and High Basal FSH Cycles for the Patients With High (>10 miu/ml) Inter Variability in Basal FSH Concentrations (n = 28)a Basal FSH (miu/ml) Basal E2 (pg/ml) Peak E2 (pg/ml) Ampules of gonadotropins Duration of stimulation (d) Follicles aspirated Preovulatory oocytes retrieved LowFSH Cycle 9.3 ± ± ± ± ± ± O.4 b 2.3 ± 0.3 a Values are means ± SD. blow FSH > high FSH ; P < High FSH Cycle 28.2 ± ± ± ± ± ± ±0.2 used. The mean deviations for the three groups are listed in Table 2. The mean deviation in patients whose basal values were <15 miu /ml was 2.6 miu /ml. This is within the range of variation, which could be explained by interassay and intraassay variability. In contrast, patients with mean basal FSH concentrations of 15 to 24.9 and ~25 miu/ml had mean deviations of 7.1 and 7.4 miu/ml, respectively, and are greater than that which could be explained by variability in the assay. The mean deviation in the patients with mean basal FSH concentrations of <15 miu /ml was significantly smaller than either of the other two groups (P < ). The mean range in basal FSH concentrations was also smaller in the low basal FSH group (P < ) (Table 2). Results by Range in Basal FSH Concentrations The final set of evaluations involved grouping patients by the range of their basal FSH concentrations. Twenty-eight of the 81 patients had ranges in their basal FSH concentrations of> 10 miu /ml. The mean deviation of the 28 patients with high inter variability (> 10 miu /ml) was higher than the remainder of the group (7.8 ± 0.6 versus 2.3 ± 0.2 miu/ml). This difference is likely because of the selection criteria for the two groups. Comparisons between patients with small and large ranges of inter variability revealed that the patients with less variability had lower basal FSH values, higher peak E2 concentrations, required fewer ampules of gonadotropins, and had more follicles aspirated and more preovulatory 00- cytes recovered at the time of retrieval (Table 3). A paired analysis of the characteristics Table 3 Mean Deviations in Basal FSH Concentrations for 81 Patients With Three or More IVF Attempts During a 24-Mo Period Mean No. of basalfsh patients Mean deviation Mean range miu/ml miu/ml < ± 0.2a 6.4 ± 0.5 b 15 to ± ± 1.9 ~ ± ± 2.5 a<15, <15 to 24.9, and ~25 miu/ml; P < b <15, <15 to 24.9, and ~25 miu/ml; P < from the s in which these 28 patients had their lowest and highest basal FSH concentrations revealed no difference in the patient's age at the time of stimulation, peak E2 levels, ampules of gonadotropins administered, duration of stimulation, the number of follicles aspirated, or the number of preovulatory oocytes recovered. Significantly, the basal E2 concentrations were higher in the s in which the patients had their lowest basal FSH concentrations (Table 4). DISCUSSION Measurements of basal and stimulated FSH concentrations have been shown to correlate well with the physiological status of a given patient's ovaries as gauged by her response to gonadotropin stimulation and clinical pregnancy rates.2 A similar test, the clomiphene citrate challenge test described by Navot et al.,12 assesses both the basal FSH concentration as well as the value after the administration Table 4 Gonadotropin Stimulation Characteristics in Patients With Low (,;;;10 miu/ml) and High «10 miu/ml) Inter Variability in Basal FSH Concentrations a LowFSH HighFSH variability variability (n = 53) (n = 28) Basal FSH (miu/ml) 11.1 ± ±0.9 Basal E2 (pg/ml) 26 ± ± 1.0 Peak E2 (pg/ml) 585 ± 29b 434 ± 21 Ampules of gonadotropins 13.9 ± ± 0.6 Duration of stimulation (d) 6.2 ± ± 0.1 Follicles aspirated 7.2 ± O.4 b 4.87±0.6 Preovulatory oocytes retrieved 3.5 ± 0.2b 2.9 ± 0.2 a Values are means ± SD. b Low inter variability > high inter variability; P< Scott et al. Inter variability in basal FSH levels Fertility and Sterility

5 of 5 days of CC. 13,14 The physiological principles are similar for evaluation of either basal or stimulated FSH values. Basal FSH concentrations reflect the balance between ovarian steroids and peptides and the hypothalamic-pituitary axis during the period of follicular requirement but before selection of the dominant follicle. The more completely the developing cohort of follicles is able to suppress FSH secretion during this period, or stated another way, the less FSH stimulation that is required to allow development of the cohort of follicles, the better the quality of those developing follicles as gauged by response to gonadotropin stimulations and outcomes after IVF. Evaluation of the response to stimulation is based on a similar physiological theory. In this case, the ability of the developing cohort of follicles to produce enough gonadal steroids or proteins to suppress FSH levels, even in the presence of some persistent antiestrogen effect from the CC, reflects the quality of those follicles. Clinical application of this information has allowed more appropriate and accurate counseling of patients participating in assisted reproduction programs. However, a lack of information regarding the degree and impact of inter variability in FSH concentrations left a number of questions unanswered. In this study, the degree of inter variability and its potential impact on stimulation quality has been determined. Information regarding pregnancy outcome is not available since the selection criteria for the study required that the patients all have multiple s without pregnancies. The mean deviation in basal FSH concentrations for all 81 patients in this study was over 4 miujml, which creates a 95% confidence interval of ±8 miu jml. This degree of variation would certainly be enough potentially to alter most patient prognostic categories. However, it must be noted that patients with mean basal FSH concentrations of <15 miu /ml had a mean deviation of only 2.6 miu jml. Therefore, if a patient's mean value were to fall in this range, she could be expected consistently to have basal FSH concentrations in this range with relatively small inter variations. Conversely, patients with mean basal FSH values of ~ 15 miu /ml had mean deviations of > 7 miu /ml. Therefore, patients with values in this range might be expected to have large fluctuations in their basal FSH levels. An equally important clinical issue is the potential impact of inter variability in basal FSH concentrations on stimulation quality. Unfortunately, changes in basal FSH levels do not prognos- tic ate changes in the responsiveness of the ovary or the number of oocytes that can be obtained. Thus, this parameter is not useful in selecting the optimal in which to stimulate an individual patient. This finding was true for the study population as a whole and also for those individuals with the greatest range of inter variability. The one parameter that was different between the s with the highest and lowest basal FSH concentrations was the basal E2 level. There are at least two physiological explanations for this difference. First, the elevation in E2, although an average increase of only 10 pgjml in magnitude, may have increased negative feedback suppression of FSH secretion. Another explanation could be that these patients have progressed further into follicular development with increased suppression by both ovarian steroids and peptides. This is consistent with the fact that the follicular phase of the is shorter in women who may be becoming perimenopausal. 5-7 The ovarian response to gonadotropin stimulation in those patients with increased inter variability is consistent with the hypothesis of declining ovarian function. They had lower peak E2 concentrations, fewer follicles aspirated, and fewer preovulatory oocytes recovered than those patients with smaller inter variabilities. Other factors potentially affecting stimulation quality such as obesity, polycystic ovarian syndrome (PCOS), hyperandrogenic states, and the ages of the patients could possibly impact on the results of this study. In this population, there were no differences in the ages of patients who had high or low inter variability in their basal FSH concentrations. The more specific question of how hyperandrogenic states or PC OS impacts on these results cannot be addressed with these data, since the patients in this study were not specifically screened for these disorders. Anecdotally, patients undergoing IVF in our program with clinical and endocrine parameters consistent with PCOS have had low basal FSH values with minimal variability. This question awaits more detailed study. Similarly, we are currently unable to address the impact of obesity on these results. The accurate diagnosis of obesity requires estimation of percent body fat with skinfold thickness measurements or more complex techniques. The impact of obesity on a patient's basal gonadotropin levels as well as on her stimulation quality, retrieval characteristics, and ultimate success rates is currently being evaluated. Based on these data, two basal FSH values that Scott et al. Inter variability in basal FSH levels 301

6 are in agreement may be used to counsel patients regarding their performance during gonadotropin stimulation. Furthermore, if the patient has wide fluctuations in her basal FSH values, then she is more likely to respond poorly to gonadotropin stimulation. This information should be useful in counseling patients regarding their appropriateness for assisted reproduction. REFERENCES 1. Muasher SJ, Oehninger S, Simonetti S, Matta J, Ellis LM, Liu H-C, Jones GS, Rosenwaks Z: The value of basal and/ or stimulated serum gonadotropin levels in prediction of stimulation response and in vitro fertilization outcome. Fertil Steril 50:298, Scott RT, Toner JF, Muasher SJ, Oehninger SC, Robinson S, Rosenwaks Z: Follicle-stimulating hormone levels on day 3 are predictive of in vitro fertilization outcome. Fertil Steril51:651, Irvine WJ, Chan MMW, Scarth L, Kolb FO, Hartog M, Bayliss RIS, Drury MI: Immunological aspects of premature ovarian failure associated with idiopathic Addison's disease. Lancet 2:883, Vazquez AM, Kenny FM: Ovarian failure and antiovarian antibodies in association with hypoparathyroidism, moniliasis and Addison's and Hashimoto's Disease. Obstet Gynecol 41:414, Sherman BM, Korenman SG: Hormonal characteristics of the human menstrual throughout reproductive life. J Clin Invest 55:699, Sherman BM, West JH, Korenman SG: The menopausal transition: analysis of LH, FSH, estradiol and progesterone concentrations during menstrual s of older women. J Clin Endocrinol Metab 42:629, Reyes FI, Winter JS, Faiman C: Pituitary-ovarian relationships preceding the menopause. I. A cross-sectional study of serum follicle-stimulating hormone, luteinizing hormone, prolactin, estradiol, and progesterone levels. Am J Obstet GynecoI129:557, Rosenwaks Z, Muasher SJ: Recruitment offertilizable eggs. In In Vitro Fertilization-Norfolk, Edited by HW Jones, Jr, GS Jones, GD Hodgen, Z Rosenwaks. Baltimore, Williams & Wilkins, 1986, p Muasher SJ, Flood J, Simonetti S, Kreiner DK, Acosta A, Rosenwaks Z: Comparison between laparoscopic and ultrasonographically guided transvaginal follicular aspiration methods in an in vitro fertilization (IVF) program in the same patients using the same stimulation protocol. (Abstr.) Presented at the American College of Obstetrics and Gynecology Annual Clinical Meeting, Boston, MA, May 2 to 5, Veeck LL, Worthan JWE, Jr, Witmyer J, Sandow BA, Acosta AA, Garcia JE, Jones GS, Jones HW, Jr: Maturation and fertilization of morphologically immature human oocytes in a program of in vitro fertilization. Fertil Steril 39:594, Veeck LL: Atlas ofthe Human Oocyte and Early Conceptus. Baltimore, Williams & Wilkins, Navot D, Boyd C, Kliot G, Sandler B, Fox J, Grunfeld L: The clomiphene citrate challenge test (CCCT) can discriminate between favorable and unfavorable IVF outcome. (Abstr. 77) Presented at the Sixth World Congress on In Vitro Fertilization and Alternate Assisted Reproduction, Jerusalem, Israel, April 2 to 7, Loumaye E, Billion JM, Mine JM, Psalti I, Pensis M: Prognostic value of a clomiphene challenge test (CCT) in IVF. (Abstr. 210) Presented at the Sixth World Congress on In Vitro Fertilization and Alternate Assisted Reproduction, Jerusalem, Israel, April 2 to 7, Navot D, Rosenwaks Z, Margalioth EJ: Prognostic assessment offemale fecundity. Lancet 2:645, Scott et ai. Inter variability in basal FSH levels Fertility and Sterility