- Email: [email protected]
Poor success of gonadotropin-induced controlled ovarian hyperstimulation and intrauterine insemination for older women
Poor success of gonadotropin-induced controlled ovarian hyperstimulation and intrauterine insemination for older women Isiah D. Harris, M.D.,a Stacey A. Missmer, Sc.D.,a,b,c and Mark D. Hornstein, M.D.a a Department of Obstetrics and Gynecology; b Channing Laboratory, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School; and c Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts
Objective: To determine the success rates and utility of controlled ovarian hyperstimulation in conjunction with intrauterine insemination (COH/IUI) cycles in women aged 38–39 years versus women R40 years old. Design: Retrospective cohort study. Setting: Tertiary-care academic medical center. Patient(s): There were 130 women, 57 aged 38–39 years (42.6%) and 73 aged R40 years (57.4%), who underwent 262 IUI cycles (range 1–3 cycles per woman). Intervention(s): Infertility treatments with gonadotropins and IUI. Main Outcome Measure(s): Clinical pregnancy rates and live birth rates stratified by age. Result(s): The most common infertility diagnosis was diminished ovarian reserve, which was found more frequently among women aged R40 years than among the slightly younger group. The age-specific groups were similar in their baseline characteristics and cycle parameters. Women who were 38–39 years old had an overall live birth rate of 6.1% per cycle, with no live births occurring after the second cycle, and women R40 years old had an overall live birth rate of 2.0% per cycle, with all births occurring in the first cycle. Conclusion(s): The efficacy of COH/IUI cycles significantly decreases with age, but women aged 38–39 years had reasonable success during the first two cycles. However, for women aged R40 years, no benefit after a single cycle of COH/IUI was observed. Women aged R40 years should be considered for in vitro fertilization after one failed COH/IUI cycle. (Fertil Steril 2010;94:144–8. 2010 by American Society for Reproductive Medicine.) Key Words: Infertility, age, intrauterine insemination, live birth rate, ovulation induction, gonadotropin
Controlled ovarian hyperstimulation (COH) in conjunction with intrauterine insemination (IUI) has been an important and successful tool in infertility therapy for many years. This is particularly true for patients with oligovulation and unexplained infertility, as well as for patients with surgically treated endometriosis and those who fail less invasive approaches (1–6). The use of gonadotropins for COH is more expensive, but also more effective, than chlomiphene citrate for COH with IUI (7). There are recent data showing that letrozole has efficacy similar to FSH, with substantially less cost and patient inconvenience (8), but this is not confirmed by the most recent Cochrane review (9). Female fecundity declines with age, as does the efficacy of assisted reproductive techniques (10, 11). Although COH/IUI is a widely used tool, there is limited evidence to suggest its the usefulness in women over the age of 40 years. A few studies have addressed this topic with conflicting results. Corsan et al. (12) showed some utility of COH/IUI for women aged 40–42 years (n ¼ 168), although no pregnancies occurred in women aged R43 years. However, Frederick et al. (13) Received August 28, 2008; revised and accepted February 12, 2009; published online April 25, 2009. I.H. has nothing to disclose. S.M. has nothing to disclose. M.H. has nothing to disclose. Reprint requests: Dr. Mark D. Hornstein, 75 Francis Street, Boston, MA 02115 (FAX: 617-566-7752; E-mail: [email protected]).
144
reported only a 5% fecundity rate per cycle with a 73% rate of miscarriage in COH/IUI cycles in women >40 years old (n ¼ 77). Currently, in states such as Massachusetts, where insurance covers assisted reproductive technologies, most patients are required to undergo one to three cycles of COH/IUI before moving on to in vitro fertilization (IVF). The best evidence to support this practice comes from Plosker et al. (14), who determined that 90% of pregnancies from COH/IUI occurred in the first three cycles, with a substantial drop in fecundity thereafter. However, in that analysis of COH/IUI cycles, women >40 years old had very limited efficacy (n ¼ 25). We therefore conducted a retrospective cohort study among patients aged R38 years undergoing IUI at a tertiarycare medical center to determine the efficacy of COH/IUI cycles in women aged 38–39 versus women aged R40 years.
MATERIALS AND METHODS The medical records of 130 infertile couples who had undergone a total of 242 treatment cycles with gonadotropin-induced COH/IUI from January 2005 to July 2007 at the Center for Reproductive Medicine at Brigham and Women’s Hospital were reviewed retrospectively. The study was approved by the Institutional Review Board of Brigham and Women’s Hospital, and the requirement for individual
Fertility and Sterility Vol. 94, No. 1, June 2010 Copyright ª2010 American Society for Reproductive Medicine, Published by Elsevier Inc.
0015-0282/$36.00 doi:10.1016/j.fertnstert.2009.02.040
informed consent was waived owing to the nature of the study and limited use of personal information. All patients aged R38 years at the start of the treatment cycle with infertility of at least 12 months’ duration and who underwent COH/ IUI were eligible for inclusion. The minimum evaluation included a semen analysis; ovulation assessment by basal body temperature recording, measuring midluteal P level, urinary LH testing, or an endometrial biopsy in the luteal phase; hysterosalpingography or hysteroscopy to assess the uterine cavity; and demonstration of tubal patency by hysterosalpingography or laparoscopy. The patients were divided into two groups based on the age of the female partner: 38–39 and R40 years old. The couples were also grouped according to their infertility diagnosis: male factor (defined as a semen analysis showing a concentration of <20 106/mL, motility of <50%, or morphology of <20% normal forms by World Health Organization criteria and the absence of any pathology in the female partner; n ¼ 21); infertility related to ovulatory dysfunction (n ¼ 8) or endometriosis (n ¼ 18); tubal factor infertility (defined as any abnormality of one or both fallopian tubes or a history of any tubal surgery; n ¼ 6); infertility related to diminished ovarian reserve (defined as a day 3 FSH level >10 ng/mL) or abnormal chlomiphine citrate challenge test (n ¼ 55); infertility related to uterine abnormalities (most frequently uterine leiomyoma or asherman’s syndrome; n ¼ 10); and unexplained infertility, defined as the absence of identifiable pathology (i.e., a normal semen analysis, documentation of ovulation, a normal uterine cavity, patent fallopian tubes, and the absence of peritoneal pathology) (n ¼ 30). Starting on cycle day 3, patients received FSH, with or without human menopausal gonadotropin (hMG) at a dose of 75–600 IU, based on patient baseline characteristics or knowledge of prior responses. Beginning on cycle day 7 or 8, ultrasound evaluation of the pelvis and serum E2 measurements were performed and repeated every 1–3 days depending on the stimulation protocol. The gonadotropin dose was adjusted accordingly. Human chorionic gonadotropin (5,000 or 10,000 IU) was given when the leading follicle reached preovulatory size (R16 mm in diameter). Intrauterine insemination generally was performed twice, at 12 hours and 36 hours after the injection of hCG, and patients were instructed to restrain from intercourse during the cycle month. A chemical pregnancy was defined by a delay in menses combined with the presence of a serum hCG level >5 IU, without ultrasound confirmation of pregnancy. Ultrasounds were conducted between 5 and 6 weeks’ gestational age, and a clinical pregnancy was defined by ultrasonographic findings consistent with an intrauterine gestation. Fecundity was defined as any pregnancy including both chemical and clinical pregnancies. Differences in patient characteristics between the two age strata (women aged 38–39 compared to women aged R40 years) were quantified through the calculation of a two-sided t test with the Fisher exact test for categoric variables and with unpaired t tests for continuous variables. The likelihood Fertility and Sterility
of clinical pregnancy and of live birth per cycle, comparing the older with the younger women, was estimated through unconditional logistic regression calculation of odds ratios (relative risk [RR]) and 95% confidence intervals (CI) as well as two-sided Wald P values. Additionally, to determine the independent effect of age on cycle outcomes, a multivariable logistic regression analysis was performed. The model included potential confounders that were found to be statistically significant in univariate analysis and those that were generally considered to be clinically significant. Statistical significance was assumed at P%.05. RESULTS The 130 infertile couples underwent 262 treatment cycles with gonadotropin-induced COH/IUI (range 1–3 cycles per couple). There were 57 women aged 38–39 years at first cycle start who contributed 115 cycles, 73 women aged R40 years at first cycle start who contributed 147 cycles. Among the 57 women aged 38–39 years, the mean age was 39.1 0.5 years, and among the 73 women aged R40 years, the mean age was 41.9 1.4 years (range 40–45 years). Of the 262 cycles, 255 were completed, and 7 were canceled because of poor response, a high risk of ovarian hyperstimulation syndrome, or inadequate sperm specimen. Table 1 compares the age-specific groups of patients by diagnosis, number of prior pregnancies and deliveries, partner’s age, and partner’s semen analysis. Women in the 38–39-yearold group had a lower mean peak E2 level of 739.2 pg/mL compared with 794.0 pg/mL in the R40-year-old group, but this was not a statistically significant difference (P¼.17), and overall the stimulation parameters were similar between the two groups. The day 3 FSH levels (P¼.64), numbers of prior pregnancies and deliveries (P¼.56 and P¼.46, respectively), numbers of follicles R12 mm and >16 mm (P¼.50 and P¼.54, respectively), and the endometrial thicknesses on the day of hCG administration were also similar (P¼.99). There were some differences noted in the characteristics of the male partners. The mean age of the partners was significantly higher among women R40 years old (43.1 years vs. 40.6 years; P¼.01). Interestingly, the partners of these women were found to have a higher percentage of normal forms on semen analysis than partners of slightly younger women (30.8% vs. 23.7%; P¼.01), although all other parameters were similar between the two groups: sperm count (80.8 106 vs. 64.8 106; P¼.14), volume (2.7 mL vs. 2.9 mL; P¼.56), and motility (59.1% vs. 52.6%; P¼.15). There were a total of 17 clinical pregnancies and 10 live births in the 262 cycles evaluated (Fig. 1), and both groups used a second or third cycle at a similar rate, with an average of two cycles per patient in each group. In the group 38–39 years old, the overall fecundity rate was 15.8% with an overall live birth rate of 12.3%. The fecundity per cycle was 9.0%, and the live birth per cycle was 5.2%. In the R40-year-old group, the overall fecundity rate was 12.3% with a live birth 145
TABLE 1 Baseline characteristics and cycle parameters.
Day 3 FSH,9, V/L Peak E2 levels, pg/ml Peak endometrial thickness, mm Follicles R12 mm, n Follicles R16 mm, n Prior pregnancies, n Prior deliveries, n Partner parameter Age, y Sperm count (106) Seminal volume, mL Motility, % motile Morphology, % normal Infertility diagnosis, n (%) Dimished ovarian reserve Male factor Unknown Ovulatory dysfunction Tubal disease Endometriosis Other
38–39 y old
R40 y old
P value
9.0 (3.80) 690.9 (304.1) 9.0 (2.23) 4.3 (2.54) 2.26 (1.47) 1.08 (1.66) 0.27 (0.45)
9.3 (2.7) 815.9 (339.2) 9.0 (2.36) 4.62 (2.51) 2.41 (1.2) 1.23 (1.22) 0.34 (0.61)
.643 .174 .988 .501 .538 .560 .464
40.6 (5.0) 64.8 (46.4) 2.86 (1.39) 52.6 (23.1) 23.7 (11.9)
43.1 (5.7) 80.8 (54.8) 2.70 (1.34) 59.1 (20.2) 30.8 (14.2)
.0120 .141 .564 .152 .0124
17 (30.4) 9 (16.1) 13 (23.2) 4 (7.1) 1 (1.8) 9 (16.1) 3 (5.4)
38 (41.3) 12 (13.0) 17 (18.5) 4 (4.3) 5 (5.4) 9 (9.8) 7 (7.6)
.0268 .999 .999 .722 .238 .607 .514
Note: Except as noted for infertility diagnosis, values are mean (SD). Harris. Poor birth rates with IUI after 40. Fertil Steril 2010.
rate of 4.1%. The fecundity rate per cycle was 7.8%, and the live birth rate per cycle rate was 2.0%. In the multivariable analysis (Table 2), women aged R40 years had substantially lower live birth rates and fecundity rates. The women R40 years old were 71% (P¼.19), 78% (P¼.08), and 78% (P¼.08) less likely to have a live birth in each of the three cycles, respectively, and were 43% (P¼.40), 42% (P¼.29), and 40% (P¼.29) less likely to conceive in each of the three cycles, respectively. Although these figures did not reach statistical significance, the large magnitude of difference suggests insufficient statistical power, which was confirmed by a retrospective power analysis showing only a 30% statistical power with respect to fecundity.
R40 years of age who underwent COH/IUI cycles. For these women, neither fertility diagnosis nor day 3 FSH levels affected outcomes. Frederick et al. (13) reported a live birth rate of 3.9% per patient and 1.4% per cycle in women >40
FIGURE 1 Reverse Kaplan-Meier curve showing the live birth rates and fecundity rates by cycle stratified by age. 25%
When stratified by diagnosis, the live birth rate ranged from 16.7% for couples with unexplained infertility or tubal disease to 0% for couples with infertility related to ovulatory dysfunction (Fig. 2). Regardless of diagnosis, all live births in the group 38–39 years old were achieved in the first two cycles, and all live births in the R40-year-old group were achieved in the first cycle.
20%
DISCUSSION The results of this retrospective study confirm a substantial age-related decline in fecundity and cumulative pregnancy rates (15), with particularly low live birth rates for women
0%
146
Harris et al.
Poor birth rates with IUI after 40
Fecundity, Age 38-39
15%
Fecundity, Age >40 Live Birth Rate, Age 38-39
10%
Live Birth Rate, Age >40
5%
Cycle 1
Cycle 2
Cycle 3
Harris. Poor birth rates with IUI after 40. Fertil Steril 2010.
Vol. 94, No. 1, June 2010
TABLE 2
FIGURE 2
Results of univariate and multivariate analysis, comparing woman R40 years old with women aged 38–39 years.
Live birth rates stratified by diagnosis.
Odds P ratio 95% CI value Results after first cycle Live births, univariate Live births, multivariate Fecundity, univariate Fecundity, multivariate Results after 2 cycles Live births, univariate Live births, multivariate Fecundity, univariate Fecundity, multivariate Results after 3 cycles Live births, univariate Live births, multivariate Fecundity, univariate Fecundity, multivariate
0.45 0.29 0.64 0.57
0.10–1.95 0.05–1.86 0.20–2.02 0.15–2.13
.28 .19 .45 .40
0.31 0.22 0.61 0.58
0.08–1.26 0.04–1.18 0.24–1.53 0.21–1.59
.12 .08 .52 .29
0.31 0.22 0.62 0.60
0.08–1.26 0.04–1.18 0.26–1.50 0.23–1.56
.24 .08 .47 .29
Note: Multivariate model included peak E2, number of follicles >16 mm, number of prior pregnancies, sperm motility, and tubal infertility diagnosis. CI ¼ confidence interval. Harris. Poor birth rates with IUI after 40. Fertil Steril 2010.
years old, with each woman undergoing an average of 2.7 cycles (n ¼ 77). These are very similar to the results in the present study, in which the live birth rate was 4.1% per patient and 2.0% per cycle. Although a similar analysis by Corsan et al. (12) found slightly better results (n ¼ 168; 4.5% delivery rate per cycle), there were no deliveries with COH/IUI for women >42 years old. The present study goes beyond confirming these previous results by showing that all women who had a live birth with COH/IUI did so in the first cycle. Furthermore, after adjusting for potential and observed confounders of the age-cycle outcome relation, we observed that the two groups in this study had very minimal differences in fecundity rates per cycle but a substantial difference in live birth rates. The consistently small odds ratios suggest a clinically significant difference in live birth rates that may be masked in clinical practice by the similarities in fecundity rates. Most of the earlier literature has used fecundity or clinical pregnancy rates as a success measure for COH/IUI; however, the true benchmark of assisted reproductive technologies is the ‘‘take-home baby’’ rate or live birth rate. Using this yardstick, we can better evaluate the cost-effectiveness of our interventions, as has previously been done to evaluate male factor infertility (16, 17), and may find that a policy that seems to be optimally cost-effective among women >40 years old would require only one COH/IUI cycle before proceeding to IVF. Fertility and Sterility
Harris. Poor birth rates with IUI after 40. Fertil Steril 2010.
The primary limitation of the present study is the potential bias inherent to all records-based retrospective cohort studies. It is possible that the outcomes of the second and third cycles may have been positively biased if couples with poor response were discouraged from continuing with COH/IUI. Also, patients who were able to self-pay may have elected to move on to IVF after a single failed COH/IUI cycle or may have forgone this intervention altogether in favor of IVF, which would further bias the results. Finally, the small numbers of clinical pregnancies and live births in the second and third cycles suggests the need for a larger sample size and greater statistical power, although the findings appear to be consistent with those of other studies. In conclusion, the findings of this retrospective cohort study suggest that women of advanced reproductive age demonstrate a poor prognosis with COH/IUI. Additionally, women R40 years old should undergo no more than one COH/IUI cycle before proceeding to IVF. Further study is needed to evaluate the cost-effectiveness of IUI versus IVF in this population of older women.
REFERENCES 1. Melis GB, Paoletti AM, Ajossa S, Guerriero S, Depau GF, Mais V. Ovulation induction with gonadotropins as sole treatment in infertile couples with open tubes: a randomized prospective comparison between intrauterine insemination and timed vaginal intercourse. Fertil Steril 1995;64:1088–93. 2. Vlahos NF, Coker L, Lawler C, Zhao Y, Bankowski B, Wallach EE. Women with ovulatory dysfunction undergoing ovarian stimulation with
147
3. 4.
5.
6.
7.
8.
9.
clomiphene citrate for intrauterine insemination may benefit from administration of human chorionic gonadotropin. Fertil Steril 2005;83:1510–6. Dodson WC, Haney AF. Controlled ovarian hyperstimulation and intrauterine insemination for treatment of infertility. Fertil Steril 1991;55:457–67. Martinez AR, Bernardus RE, Vermeiden JP, Schoemaker J. Basic questions on intrauterine insemination: an update. Obstet Gynecol Surv 1993;48:811–28. Deaton JL, Gibson M, Blackmer KM, Nakajima ST, Badger GJ, Brumsted JR. A randomized, controlled trial of clomiphene citrate and intrauterine insemination in couples with unexplained infertility or surgically corrected endometriosis. Fertil Steril 1990;54:1083–8. Hughes EG. Stimulated intra-uterine insemination is not a natural choice for the treatment of unexplained subfertility. ’’Effective treatment’’ or ‘‘not a natural choice’’? Hum Reprod 2003;18:912–4. Dankert T, Kremer JA, Cohlen BJ, Hamilton CJ, Pasker-de Jong PC, Straatman H, et al. A randomized clinical trial of clomiphene citrate versus low dose recombinant FSH for ovarian hyperstimulation in intrauterine insemination cycles for unexplained and male subfertility. Hum Reprod 2007;22:792–7. Gregoriou O, Vlahos NF, Konidaris S, Papadias K, Botsis D, Creatsas GK. Randomized controlled trial comparing superovulation with letrozole versus recombinant follicle-stimulating hormone combined to intrauterine insemination for couples with unexplained infertility who had failed clomiphene citrate stimulation and intrauterine insemination. Fertil Steril 2008;90:678–83. Cantineau AE, Cohlen BJ, Heineman MJ. Ovarian stimulation protocols (antioestrogens, gonadotrophins with and without GnRH agonists/antag-
148
Harris et al.
Poor birth rates with IUI after 40
10. 11.
12.
13.
14. 15.
16.
17.
onists) for intrauterine insemination (IUI) in women with subfertility. Cochrane Database Syst Rev 2007;(2):CD005356. Gindoff PR, Jewelewicz R. Reproductive potential in the older woman. Fertil Steril 1986;46:989–1001. Jacobs SL, Metzger DA, Dodson WC, Haney AF. Effect of age on response to human menopausal gonadotropin stimulation. J Clin Endocrinol Metab 1990;71:1525–30. Corsan G, Trias A, Trout S, Kemmann E. Ovulation induction combined with intrauterine insemination in women 40 years of age and older: is it worthwhile? Hum Reprod 1996;11:1109–12. Frederick JL, Denker MS, Rojas A, Horta I, Stone SC, Asch RH, et al. Is there a role for ovarian stimulation and intra-uterine insemination after age 40? Hum Reprod 1994;9:2284–6. Plosker SM, Jacobson W, Amato P. Predicting and optimizing success in an intra-uterine insemination programme. Hum Reprod 1994;9:2014–21. Sahakyan M, Harlow BL, Hornstein MD. Influence of age, diagnosis, and cycle number on pregnancy rates with gonadotropin-induced controlled ovarian hyperstimulation and intrauterine insemination. Fertil Steril 1999;72:500–4. Goverde AJ, McDonnell J, Vermeiden JP, Schats R, Rutten FF, Schoemaker J. Intrauterine insemination or in-vitro fertilisation in idiopathic subfertility and male subfertility: a randomised trial and cost-effectiveness analysis. Lancet 2000;355:13–8. Van Voorhis BJ, Barnett M, Sparks AE, Syrop CH, Rosenthal G, Dawson J. Effect of the total motile sperm count on the efficacy and cost-effectiveness of intrauterine insemination and in vitro fertilization. Fertil Steril 2001;75:661–8.
Vol. 94, No. 1, June 2010
Objective: To determine the success rates and utility of controlled ovarian hyperstimulation in conjunction with intrauterine insemination (COH/IUI) cycles in women aged 38–39 years versus women R40 years old. Design: Retrospective cohort study. Setting: Tertiary-care academic medical center. Patient(s): There were 130 women, 57 aged 38–39 years (42.6%) and 73 aged R40 years (57.4%), who underwent 262 IUI cycles (range 1–3 cycles per woman). Intervention(s): Infertility treatments with gonadotropins and IUI. Main Outcome Measure(s): Clinical pregnancy rates and live birth rates stratified by age. Result(s): The most common infertility diagnosis was diminished ovarian reserve, which was found more frequently among women aged R40 years than among the slightly younger group. The age-specific groups were similar in their baseline characteristics and cycle parameters. Women who were 38–39 years old had an overall live birth rate of 6.1% per cycle, with no live births occurring after the second cycle, and women R40 years old had an overall live birth rate of 2.0% per cycle, with all births occurring in the first cycle. Conclusion(s): The efficacy of COH/IUI cycles significantly decreases with age, but women aged 38–39 years had reasonable success during the first two cycles. However, for women aged R40 years, no benefit after a single cycle of COH/IUI was observed. Women aged R40 years should be considered for in vitro fertilization after one failed COH/IUI cycle. (Fertil Steril 2010;94:144–8. 2010 by American Society for Reproductive Medicine.) Key Words: Infertility, age, intrauterine insemination, live birth rate, ovulation induction, gonadotropin
Controlled ovarian hyperstimulation (COH) in conjunction with intrauterine insemination (IUI) has been an important and successful tool in infertility therapy for many years. This is particularly true for patients with oligovulation and unexplained infertility, as well as for patients with surgically treated endometriosis and those who fail less invasive approaches (1–6). The use of gonadotropins for COH is more expensive, but also more effective, than chlomiphene citrate for COH with IUI (7). There are recent data showing that letrozole has efficacy similar to FSH, with substantially less cost and patient inconvenience (8), but this is not confirmed by the most recent Cochrane review (9). Female fecundity declines with age, as does the efficacy of assisted reproductive techniques (10, 11). Although COH/IUI is a widely used tool, there is limited evidence to suggest its the usefulness in women over the age of 40 years. A few studies have addressed this topic with conflicting results. Corsan et al. (12) showed some utility of COH/IUI for women aged 40–42 years (n ¼ 168), although no pregnancies occurred in women aged R43 years. However, Frederick et al. (13) Received August 28, 2008; revised and accepted February 12, 2009; published online April 25, 2009. I.H. has nothing to disclose. S.M. has nothing to disclose. M.H. has nothing to disclose. Reprint requests: Dr. Mark D. Hornstein, 75 Francis Street, Boston, MA 02115 (FAX: 617-566-7752; E-mail: [email protected]).
144
reported only a 5% fecundity rate per cycle with a 73% rate of miscarriage in COH/IUI cycles in women >40 years old (n ¼ 77). Currently, in states such as Massachusetts, where insurance covers assisted reproductive technologies, most patients are required to undergo one to three cycles of COH/IUI before moving on to in vitro fertilization (IVF). The best evidence to support this practice comes from Plosker et al. (14), who determined that 90% of pregnancies from COH/IUI occurred in the first three cycles, with a substantial drop in fecundity thereafter. However, in that analysis of COH/IUI cycles, women >40 years old had very limited efficacy (n ¼ 25). We therefore conducted a retrospective cohort study among patients aged R38 years undergoing IUI at a tertiarycare medical center to determine the efficacy of COH/IUI cycles in women aged 38–39 versus women aged R40 years.
MATERIALS AND METHODS The medical records of 130 infertile couples who had undergone a total of 242 treatment cycles with gonadotropin-induced COH/IUI from January 2005 to July 2007 at the Center for Reproductive Medicine at Brigham and Women’s Hospital were reviewed retrospectively. The study was approved by the Institutional Review Board of Brigham and Women’s Hospital, and the requirement for individual
Fertility and Sterility Vol. 94, No. 1, June 2010 Copyright ª2010 American Society for Reproductive Medicine, Published by Elsevier Inc.
0015-0282/$36.00 doi:10.1016/j.fertnstert.2009.02.040
informed consent was waived owing to the nature of the study and limited use of personal information. All patients aged R38 years at the start of the treatment cycle with infertility of at least 12 months’ duration and who underwent COH/ IUI were eligible for inclusion. The minimum evaluation included a semen analysis; ovulation assessment by basal body temperature recording, measuring midluteal P level, urinary LH testing, or an endometrial biopsy in the luteal phase; hysterosalpingography or hysteroscopy to assess the uterine cavity; and demonstration of tubal patency by hysterosalpingography or laparoscopy. The patients were divided into two groups based on the age of the female partner: 38–39 and R40 years old. The couples were also grouped according to their infertility diagnosis: male factor (defined as a semen analysis showing a concentration of <20 106/mL, motility of <50%, or morphology of <20% normal forms by World Health Organization criteria and the absence of any pathology in the female partner; n ¼ 21); infertility related to ovulatory dysfunction (n ¼ 8) or endometriosis (n ¼ 18); tubal factor infertility (defined as any abnormality of one or both fallopian tubes or a history of any tubal surgery; n ¼ 6); infertility related to diminished ovarian reserve (defined as a day 3 FSH level >10 ng/mL) or abnormal chlomiphine citrate challenge test (n ¼ 55); infertility related to uterine abnormalities (most frequently uterine leiomyoma or asherman’s syndrome; n ¼ 10); and unexplained infertility, defined as the absence of identifiable pathology (i.e., a normal semen analysis, documentation of ovulation, a normal uterine cavity, patent fallopian tubes, and the absence of peritoneal pathology) (n ¼ 30). Starting on cycle day 3, patients received FSH, with or without human menopausal gonadotropin (hMG) at a dose of 75–600 IU, based on patient baseline characteristics or knowledge of prior responses. Beginning on cycle day 7 or 8, ultrasound evaluation of the pelvis and serum E2 measurements were performed and repeated every 1–3 days depending on the stimulation protocol. The gonadotropin dose was adjusted accordingly. Human chorionic gonadotropin (5,000 or 10,000 IU) was given when the leading follicle reached preovulatory size (R16 mm in diameter). Intrauterine insemination generally was performed twice, at 12 hours and 36 hours after the injection of hCG, and patients were instructed to restrain from intercourse during the cycle month. A chemical pregnancy was defined by a delay in menses combined with the presence of a serum hCG level >5 IU, without ultrasound confirmation of pregnancy. Ultrasounds were conducted between 5 and 6 weeks’ gestational age, and a clinical pregnancy was defined by ultrasonographic findings consistent with an intrauterine gestation. Fecundity was defined as any pregnancy including both chemical and clinical pregnancies. Differences in patient characteristics between the two age strata (women aged 38–39 compared to women aged R40 years) were quantified through the calculation of a two-sided t test with the Fisher exact test for categoric variables and with unpaired t tests for continuous variables. The likelihood Fertility and Sterility
of clinical pregnancy and of live birth per cycle, comparing the older with the younger women, was estimated through unconditional logistic regression calculation of odds ratios (relative risk [RR]) and 95% confidence intervals (CI) as well as two-sided Wald P values. Additionally, to determine the independent effect of age on cycle outcomes, a multivariable logistic regression analysis was performed. The model included potential confounders that were found to be statistically significant in univariate analysis and those that were generally considered to be clinically significant. Statistical significance was assumed at P%.05. RESULTS The 130 infertile couples underwent 262 treatment cycles with gonadotropin-induced COH/IUI (range 1–3 cycles per couple). There were 57 women aged 38–39 years at first cycle start who contributed 115 cycles, 73 women aged R40 years at first cycle start who contributed 147 cycles. Among the 57 women aged 38–39 years, the mean age was 39.1 0.5 years, and among the 73 women aged R40 years, the mean age was 41.9 1.4 years (range 40–45 years). Of the 262 cycles, 255 were completed, and 7 were canceled because of poor response, a high risk of ovarian hyperstimulation syndrome, or inadequate sperm specimen. Table 1 compares the age-specific groups of patients by diagnosis, number of prior pregnancies and deliveries, partner’s age, and partner’s semen analysis. Women in the 38–39-yearold group had a lower mean peak E2 level of 739.2 pg/mL compared with 794.0 pg/mL in the R40-year-old group, but this was not a statistically significant difference (P¼.17), and overall the stimulation parameters were similar between the two groups. The day 3 FSH levels (P¼.64), numbers of prior pregnancies and deliveries (P¼.56 and P¼.46, respectively), numbers of follicles R12 mm and >16 mm (P¼.50 and P¼.54, respectively), and the endometrial thicknesses on the day of hCG administration were also similar (P¼.99). There were some differences noted in the characteristics of the male partners. The mean age of the partners was significantly higher among women R40 years old (43.1 years vs. 40.6 years; P¼.01). Interestingly, the partners of these women were found to have a higher percentage of normal forms on semen analysis than partners of slightly younger women (30.8% vs. 23.7%; P¼.01), although all other parameters were similar between the two groups: sperm count (80.8 106 vs. 64.8 106; P¼.14), volume (2.7 mL vs. 2.9 mL; P¼.56), and motility (59.1% vs. 52.6%; P¼.15). There were a total of 17 clinical pregnancies and 10 live births in the 262 cycles evaluated (Fig. 1), and both groups used a second or third cycle at a similar rate, with an average of two cycles per patient in each group. In the group 38–39 years old, the overall fecundity rate was 15.8% with an overall live birth rate of 12.3%. The fecundity per cycle was 9.0%, and the live birth per cycle was 5.2%. In the R40-year-old group, the overall fecundity rate was 12.3% with a live birth 145
TABLE 1 Baseline characteristics and cycle parameters.
Day 3 FSH,9, V/L Peak E2 levels, pg/ml Peak endometrial thickness, mm Follicles R12 mm, n Follicles R16 mm, n Prior pregnancies, n Prior deliveries, n Partner parameter Age, y Sperm count (106) Seminal volume, mL Motility, % motile Morphology, % normal Infertility diagnosis, n (%) Dimished ovarian reserve Male factor Unknown Ovulatory dysfunction Tubal disease Endometriosis Other
38–39 y old
R40 y old
P value
9.0 (3.80) 690.9 (304.1) 9.0 (2.23) 4.3 (2.54) 2.26 (1.47) 1.08 (1.66) 0.27 (0.45)
9.3 (2.7) 815.9 (339.2) 9.0 (2.36) 4.62 (2.51) 2.41 (1.2) 1.23 (1.22) 0.34 (0.61)
.643 .174 .988 .501 .538 .560 .464
40.6 (5.0) 64.8 (46.4) 2.86 (1.39) 52.6 (23.1) 23.7 (11.9)
43.1 (5.7) 80.8 (54.8) 2.70 (1.34) 59.1 (20.2) 30.8 (14.2)
.0120 .141 .564 .152 .0124
17 (30.4) 9 (16.1) 13 (23.2) 4 (7.1) 1 (1.8) 9 (16.1) 3 (5.4)
38 (41.3) 12 (13.0) 17 (18.5) 4 (4.3) 5 (5.4) 9 (9.8) 7 (7.6)
.0268 .999 .999 .722 .238 .607 .514
Note: Except as noted for infertility diagnosis, values are mean (SD). Harris. Poor birth rates with IUI after 40. Fertil Steril 2010.
rate of 4.1%. The fecundity rate per cycle was 7.8%, and the live birth rate per cycle rate was 2.0%. In the multivariable analysis (Table 2), women aged R40 years had substantially lower live birth rates and fecundity rates. The women R40 years old were 71% (P¼.19), 78% (P¼.08), and 78% (P¼.08) less likely to have a live birth in each of the three cycles, respectively, and were 43% (P¼.40), 42% (P¼.29), and 40% (P¼.29) less likely to conceive in each of the three cycles, respectively. Although these figures did not reach statistical significance, the large magnitude of difference suggests insufficient statistical power, which was confirmed by a retrospective power analysis showing only a 30% statistical power with respect to fecundity.
R40 years of age who underwent COH/IUI cycles. For these women, neither fertility diagnosis nor day 3 FSH levels affected outcomes. Frederick et al. (13) reported a live birth rate of 3.9% per patient and 1.4% per cycle in women >40
FIGURE 1 Reverse Kaplan-Meier curve showing the live birth rates and fecundity rates by cycle stratified by age. 25%
When stratified by diagnosis, the live birth rate ranged from 16.7% for couples with unexplained infertility or tubal disease to 0% for couples with infertility related to ovulatory dysfunction (Fig. 2). Regardless of diagnosis, all live births in the group 38–39 years old were achieved in the first two cycles, and all live births in the R40-year-old group were achieved in the first cycle.
20%
DISCUSSION The results of this retrospective study confirm a substantial age-related decline in fecundity and cumulative pregnancy rates (15), with particularly low live birth rates for women
0%
146
Harris et al.
Poor birth rates with IUI after 40
Fecundity, Age 38-39
15%
Fecundity, Age >40 Live Birth Rate, Age 38-39
10%
Live Birth Rate, Age >40
5%
Cycle 1
Cycle 2
Cycle 3
Harris. Poor birth rates with IUI after 40. Fertil Steril 2010.
Vol. 94, No. 1, June 2010
TABLE 2
FIGURE 2
Results of univariate and multivariate analysis, comparing woman R40 years old with women aged 38–39 years.
Live birth rates stratified by diagnosis.
Odds P ratio 95% CI value Results after first cycle Live births, univariate Live births, multivariate Fecundity, univariate Fecundity, multivariate Results after 2 cycles Live births, univariate Live births, multivariate Fecundity, univariate Fecundity, multivariate Results after 3 cycles Live births, univariate Live births, multivariate Fecundity, univariate Fecundity, multivariate
0.45 0.29 0.64 0.57
0.10–1.95 0.05–1.86 0.20–2.02 0.15–2.13
.28 .19 .45 .40
0.31 0.22 0.61 0.58
0.08–1.26 0.04–1.18 0.24–1.53 0.21–1.59
.12 .08 .52 .29
0.31 0.22 0.62 0.60
0.08–1.26 0.04–1.18 0.26–1.50 0.23–1.56
.24 .08 .47 .29
Note: Multivariate model included peak E2, number of follicles >16 mm, number of prior pregnancies, sperm motility, and tubal infertility diagnosis. CI ¼ confidence interval. Harris. Poor birth rates with IUI after 40. Fertil Steril 2010.
years old, with each woman undergoing an average of 2.7 cycles (n ¼ 77). These are very similar to the results in the present study, in which the live birth rate was 4.1% per patient and 2.0% per cycle. Although a similar analysis by Corsan et al. (12) found slightly better results (n ¼ 168; 4.5% delivery rate per cycle), there were no deliveries with COH/IUI for women >42 years old. The present study goes beyond confirming these previous results by showing that all women who had a live birth with COH/IUI did so in the first cycle. Furthermore, after adjusting for potential and observed confounders of the age-cycle outcome relation, we observed that the two groups in this study had very minimal differences in fecundity rates per cycle but a substantial difference in live birth rates. The consistently small odds ratios suggest a clinically significant difference in live birth rates that may be masked in clinical practice by the similarities in fecundity rates. Most of the earlier literature has used fecundity or clinical pregnancy rates as a success measure for COH/IUI; however, the true benchmark of assisted reproductive technologies is the ‘‘take-home baby’’ rate or live birth rate. Using this yardstick, we can better evaluate the cost-effectiveness of our interventions, as has previously been done to evaluate male factor infertility (16, 17), and may find that a policy that seems to be optimally cost-effective among women >40 years old would require only one COH/IUI cycle before proceeding to IVF. Fertility and Sterility
Harris. Poor birth rates with IUI after 40. Fertil Steril 2010.
The primary limitation of the present study is the potential bias inherent to all records-based retrospective cohort studies. It is possible that the outcomes of the second and third cycles may have been positively biased if couples with poor response were discouraged from continuing with COH/IUI. Also, patients who were able to self-pay may have elected to move on to IVF after a single failed COH/IUI cycle or may have forgone this intervention altogether in favor of IVF, which would further bias the results. Finally, the small numbers of clinical pregnancies and live births in the second and third cycles suggests the need for a larger sample size and greater statistical power, although the findings appear to be consistent with those of other studies. In conclusion, the findings of this retrospective cohort study suggest that women of advanced reproductive age demonstrate a poor prognosis with COH/IUI. Additionally, women R40 years old should undergo no more than one COH/IUI cycle before proceeding to IVF. Further study is needed to evaluate the cost-effectiveness of IUI versus IVF in this population of older women.
REFERENCES 1. Melis GB, Paoletti AM, Ajossa S, Guerriero S, Depau GF, Mais V. Ovulation induction with gonadotropins as sole treatment in infertile couples with open tubes: a randomized prospective comparison between intrauterine insemination and timed vaginal intercourse. Fertil Steril 1995;64:1088–93. 2. Vlahos NF, Coker L, Lawler C, Zhao Y, Bankowski B, Wallach EE. Women with ovulatory dysfunction undergoing ovarian stimulation with
147
3. 4.
5.
6.
7.
8.
9.
clomiphene citrate for intrauterine insemination may benefit from administration of human chorionic gonadotropin. Fertil Steril 2005;83:1510–6. Dodson WC, Haney AF. Controlled ovarian hyperstimulation and intrauterine insemination for treatment of infertility. Fertil Steril 1991;55:457–67. Martinez AR, Bernardus RE, Vermeiden JP, Schoemaker J. Basic questions on intrauterine insemination: an update. Obstet Gynecol Surv 1993;48:811–28. Deaton JL, Gibson M, Blackmer KM, Nakajima ST, Badger GJ, Brumsted JR. A randomized, controlled trial of clomiphene citrate and intrauterine insemination in couples with unexplained infertility or surgically corrected endometriosis. Fertil Steril 1990;54:1083–8. Hughes EG. Stimulated intra-uterine insemination is not a natural choice for the treatment of unexplained subfertility. ’’Effective treatment’’ or ‘‘not a natural choice’’? Hum Reprod 2003;18:912–4. Dankert T, Kremer JA, Cohlen BJ, Hamilton CJ, Pasker-de Jong PC, Straatman H, et al. A randomized clinical trial of clomiphene citrate versus low dose recombinant FSH for ovarian hyperstimulation in intrauterine insemination cycles for unexplained and male subfertility. Hum Reprod 2007;22:792–7. Gregoriou O, Vlahos NF, Konidaris S, Papadias K, Botsis D, Creatsas GK. Randomized controlled trial comparing superovulation with letrozole versus recombinant follicle-stimulating hormone combined to intrauterine insemination for couples with unexplained infertility who had failed clomiphene citrate stimulation and intrauterine insemination. Fertil Steril 2008;90:678–83. Cantineau AE, Cohlen BJ, Heineman MJ. Ovarian stimulation protocols (antioestrogens, gonadotrophins with and without GnRH agonists/antag-
148
Harris et al.
Poor birth rates with IUI after 40
10. 11.
12.
13.
14. 15.
16.
17.
onists) for intrauterine insemination (IUI) in women with subfertility. Cochrane Database Syst Rev 2007;(2):CD005356. Gindoff PR, Jewelewicz R. Reproductive potential in the older woman. Fertil Steril 1986;46:989–1001. Jacobs SL, Metzger DA, Dodson WC, Haney AF. Effect of age on response to human menopausal gonadotropin stimulation. J Clin Endocrinol Metab 1990;71:1525–30. Corsan G, Trias A, Trout S, Kemmann E. Ovulation induction combined with intrauterine insemination in women 40 years of age and older: is it worthwhile? Hum Reprod 1996;11:1109–12. Frederick JL, Denker MS, Rojas A, Horta I, Stone SC, Asch RH, et al. Is there a role for ovarian stimulation and intra-uterine insemination after age 40? Hum Reprod 1994;9:2284–6. Plosker SM, Jacobson W, Amato P. Predicting and optimizing success in an intra-uterine insemination programme. Hum Reprod 1994;9:2014–21. Sahakyan M, Harlow BL, Hornstein MD. Influence of age, diagnosis, and cycle number on pregnancy rates with gonadotropin-induced controlled ovarian hyperstimulation and intrauterine insemination. Fertil Steril 1999;72:500–4. Goverde AJ, McDonnell J, Vermeiden JP, Schats R, Rutten FF, Schoemaker J. Intrauterine insemination or in-vitro fertilisation in idiopathic subfertility and male subfertility: a randomised trial and cost-effectiveness analysis. Lancet 2000;355:13–8. Van Voorhis BJ, Barnett M, Sparks AE, Syrop CH, Rosenthal G, Dawson J. Effect of the total motile sperm count on the efficacy and cost-effectiveness of intrauterine insemination and in vitro fertilization. Fertil Steril 2001;75:661–8.
Vol. 94, No. 1, June 2010