Article Retrospective analysis of 1217 IVF cycles in women aged 40 years and older

RBMOnline - Vol 14. No 3. 2007 348-355 Reproductive BioMedicine Online; www.rbmonline.com/article/2623 on web 5 February 2007 Article Retrospective analysis of 1217 IVF cycles in women aged 40 years and older Avi Tsafrir received his MD in 1998 from the Hebrew University in Jerusalem. He completed his residency in Obstetrics and Gynaecology at Hadassah Medical Center, Israel, in 2005. He is currently a specialist registrar in the IVF unit of Shaare-Zedek Medical Center, Jerusalem. His research is in the field of infertility with emphasis on treatments for advanced maternal age women and health surveillance of children conceived through assisted reproduction treatment. In collaboration with a team from the Hebrew University in Jerusalem, he studies endometrial gene expression of patients with repeated IVF failure. Dr Avi Tsafrir A Tsafrir 1, A Simon, A Revel, B Reubinoff, A Lewin, N Laufer IVF Unit, Department of Obstetrics and Gynecology, Hadassah Hebrew University Medical Centres Ein-Kerem, PO Box 12000, Jerusalem 91120, Israel 1 Correspondence: Fax. +972 2 6433337; e-mail: avits@netvision.net.il Abstract Older women comprise an increasing portion of patients entering assisted reproduction programmes. This study is a retrospective summary of the files of all patients aged 40 years and older at advent of IVF, between 1995 and 2004, in the authors centre. In all, 381 women underwent 1217 initiated treatment cycles. Embryo transfer was performed in 62.6% of initiated cycles. Success rates declined with each year after age 40; pregnancy and delivery rates were 13.9 and 9.1% at age 40 and 2.8 and 0.7% at age 45. There were no deliveries at an older age. Logistic regression analysis showed the following factors were independently and significantly related to higher pregnancy rates: younger age, lower dose of gonadotrophins, greater number of mature follicles, endometrial thickness, and number of embryos transferred; prior pregnancy did not influence success. Retrieving more than four oocytes increased pregnancy rates in all women over 40. Transferring 3 embryos or more increased pregnancy rates in all ages, but reached statistical significance only in women aged 40 41 (P < 0.000). It is concluded that in women between 40 and 41 years of age, ovarian response is a major determinant of success, but not in women older than that. Unrealistic expectations may be avoided if accurate data are provided regarding delivery rates per year after age 40. Keywords: advanced maternal age, delivery rate, IVF, pregnancy rate Introduction 348 Women in their late 30s are considered to be of advanced age when attempting to conceive. Both fecundity for spontaneous conceptions and the success rate of assisted conception are markedly reduced (te Velde and Pearson, 2002; Baird et al., 2005). Ovarian and oocyte senescence that may begin as early as the 3rd decade of a woman s life are the predominant causes of embryo and pregnancy loss in older women (Armstrong, 2001). Maternal age has impacts on pregnancy and delivery rates even after multiple IVF cycles (Check et al., 1994). In a large retrospective study of consecutive single, double and triple transfers, where all embryos were of identical quality score and identical cleavage stage, a highly significant decrease in oocyte recovery and in implantation rate with increasing age was found to be almost linear, decreasing by approximately 1.5% per year (Ziebe et al., 2001). Moreover, the statistics of the outcome measure, a live healthy baby at term, are poorer than those of younger women using assisted technologies. In a very large series of women older than 40 years with low basal FSH at initiation of IVF, the cumulative live birth rate at age 40 years was 28%, but 0% at age 46 years (Klipstein et al., 2005). Thus outcomes of pregnancies for older women, a growing number with each decade as lifestyle decisions are increasingly divergent from those of previous decades, need to be more fully documented in order to obtain the best possible clinical perspective on an often long and costly endeavour (Legro et al., 1997; Chambers et al., 2006). When pregnant, however, the perinatal outcome of older and even post-menopausal women is generally reassuring (Antinori et al., 2003). 2007 Published by Reproductive Healthcare Ltd, Duck End Farm, Dry Drayton, Cambridge CB3 8DB, UK

In Israel, the national health insurance covers most infertility treatments, including IVF, for the first two children. The age limit for performing IVF with the woman s own oocytes is 45 years, but there is no limit to the number of treatment cycles. On the other hand, the number of donor oocytes available is severely limited. Taken together, and in light of the trend in Israel as elsewhere of women delaying childbirth for non-medical reasons, there are a very large number of older women undergoing IVF. Indeed, 31% of patients undergoing oocyte retrieval in the authors unit in 2004 were older than 40 years of age. The purpose of this study is to summarize the experience of treating infertile women between 40 and 46 years of age, with a year-by-year analysis of outcomes in order to provide specific information for counselling. Materials and methods All patient data from women who were 40 years of age at the time of IVF treatment during the years 1995 to 2004 were retrospectively analysed. All initiated cycles during this period were evaluated; frozen thawed cycles and egg donation treatments were not included. The same clinical and embryology team was involved in all treatments. All patients had undergone diagnostic hysteroscopy or hysterosalpingography prior to treatment initiation to document a normal uterine cavity. Serum FSH was analysed once before commencement of IVF treatment. Patients who had repeated FSH > 20 IU/ml were not treated. Ovarian stimulation was carried out by one of the following protocols depending on previous or estimated ovarian response: (i) the long desensitizing protocol using the gonadotrophin-releasing hormone agonist (GnRHa) D-Trp-6-LHRH (Decapeptyl; Ferring, Kiel, Germany) 0.1 mg/day given subcutaneously from the mid-luteal phase of the preceding cycle (days 21 23); (ii) the short protocol in which the same dose of GnRHa is administered from day 1 of the cycle; (iii) the ultra-short protocol, in which premature ovulation was avoided by the administration of decapeptyl 0.05 mg/day for the first 3 days of the cycle only; (iv) GnRH antagonist protocol using gonadotrophin stimulation starting from day 2 with addition of a GnRH antagonist (Cetrotide; Serono International, Switzerland) 0.25 mg/day when the leading follicle achieved a diameter of 14 mm; (v) unprotected treatment cycles, in which no pituitary desensitization was attempted; and (vi) natural IVF cycles, in which no ovarian stimulation was used (Fasouliotis et al., 2000). Ovulation induction protocols involved human menopausal gonadotrophin (HMG, Pergonal; Teva, Petah Tiqua, Israel), or Menogon (Ferring Pharmaceuticals, Malmo, Sweden), urinary FSH (Metrodin; Serono), recombinant FSH (Gonal- F; Serono), or a combination of both. Ovarian response was monitored at 2- to 3-day intervals by vaginal sonography and serum oestradiol concentration measurements. Oestradiol was measured by a solid-phase ligand-labelled competitive chemiluminescence immunoassay (Immulite; DPC, Los Angeles, CA, USA). Retrieved oocytes were, according to sperm quality, either inseminated by conventional IVF or by intracytoplasmic sperm injection (ICSI) as previously described (Simon et al. 1998). Fertilized oocytes were cultured in either human tubal fluid or P1 medium (Irvine Scientific, Santa Ana, CA, USA) each enriched with 10% synthetic serum substitute (SSS, Irvine Scientific) until day 3 of fertilization when embryo transfer took place. The luteal phase was supported by intravaginal micronized progesterone (Uterogestan, Besins-Iscovesco, Paris, France) in three divided doses up to a total amount of 900 mg/day, or by progesterone in oil (Gestone; Paines and Byrne Limited, West Byfleet, Surrey, UK) starting 1 day after oocyte retrieval. Women were tested for serum β-human chorionic gonadotrophin (β-hcg) concentrations 14 days after embryo transfer. If the pregnancy test was positive, the pregnancy was monitored by serial ultrasounds to determine fetal viability. Clinical pregnancy was defined as the presence of an intrauterine gestational sac using transvaginal ultrasound. When pregnancy occurred, luteal support was continued until 10 weeks of gestation. Statistical analysis Pregnancy rates were analysed with the two-tailed Fisher s Exact test. For larger contingency tables, the chi-squared test was used, including the test for linearity. Quantitative variables were compared between groups by a two-sample t-test. A logistic regression model was applied to outcome (pregnancy in a treatment cycle) in order to assess the simultaneous effect of variables which were significant in the univariate comparison of cycles ending with and without pregnancy. The adjusted odds ratio was calculated for each of the independent variables in the model. A P-value of <0.05 was considered statistically significant. Results Demographic data During the study period, 1217 treatment cycles were initiated in 381 women in this age group. As shown in Table 1, the majority of women had unexplained, male factor, or mechanical infertility. About 15% of women had combined diagnoses, mostly including a mechanical factor. Mean age at treatment was 41.9 ± 1.8 (range 40 46) years. Mean FSH concentration at the beginning of IVF was 8.6 ± 6.6 IU/ml. Mean number of treatment cycles was 3.3 ± 2.8 (range 1 15). Only 32% of these women had delivered previously. IVF treatments The long protocol was used in 189 (15.5%) cycles; the short protocol was used in 683 cycles (56.1%); and in 183 cycles (15.0%), GnRH antagonist was given. The mean number of ampoules (75 IU each) of various gonadotrophins/cycle was 51 ± 25 (range: 4 174) ampoules/cycle. Additional details regarding ovarian and endometrial response are summarized in Table 2. 349

Table 1. Cause of infertility, reproductive history and FSH concentrations of 381 IVF patients aged 40 years and above. Parameter Per cent of all patients Cause of infertility Unexplained 31 Male factor 29 Mechanical factor 19 Anovulation 1.8 Endometriosis 2.7 Combined factors 15.3 IVF treatment before age 40 None 53.5 Successful 16.5 Unsuccessful 30 Number of previous pregnancies 0 39 1 25 2 16 3 20 Number of previous deliveries 0 68 1 21.5 2 7.6 >2 2.9 Table 2. Treatment details of 1217 IVF cycles in 381 women aged 40 and above. GnRH = gonadotrophin-releasing hormone; HCG = human chorionic gonadotrophin; ICSI = intracytoplasmic sperm injection. Treatment details Per cent of Mean ± SD all cycles (range) 350 Type of gonadotrophin Urinary menotropin 77.0 Recombinant FSH 9.5 Combined urinary and recombinant 8.5 None (spontaneous cycle) 5.0 No. of 75 IU gonadotrophin ampoules per cycle All cycles 51 ± 25 Long protocol (n = 189) 43 ± 21 Short protocol (n = 683) 50 ± 24 Ultrashort protocol (n = 12) 55 ± 21 GnRH antagonist protocol (n = 183) 55 ± 25 Gonadotrophins only (n = 81) 61 ± 32 No. of follicles 18 mm on HCG day 1.9 ± 2.4 Serum oestradiol on HCG day (pmol/l) 3670 ± 4240 Endometrium thickness on HCG day (mm) 9.0 ± 2.4 Method of fertilization IVF 29.5 ICSI 67 Both 3.5 No. of oocytes retrieved 4.3 ± 4.7 (0 35) Embryos transferred per cycle 0 37.4 1 16 2 14.5 3 12 4 20.5

Cancellation, oocyte retrieval, and embryo transfer rates Of 1217 initiated cycles, 202 (16.6%) did not reach oocyte retrieval. The main reason for cancellation was low response to ovarian stimulation. Other reasons for cancellation included premature luteinization, ovarian cysts, and unfavourable or inadequate endometrium for embryo transfer. Additionally, in 52 treatment cycles (4.3%) in which oocyte retrieval was performed, no oocytes were retrieved. The mean number of oocytes/cycle when retrieval was performed was 5.1 ± 4.5. Embryos were transferred in only 762 treatment cycles (62.6%), but this rate rose to 79% in cycles in which oocyte retrieval was performed. The mean number of embryos transferred was 2.7 ± 1.4 (range 1 6). Notably, in 20% of cycles more than three embryos were transferred (Table 2). Additional details are summarized in Table 3. Outcome There were 88 pregnancies. Overall pregnancy rates (PR) and delivery rates (DR) per initiated cycle were 7.3 and 4.7% respectively. No woman had more than one delivery. Treatment outcome, stratified according to age at treatment, is presented in Table 3. There was one additional extrauterine pregnancy not included in the summary. Pregnancy loss rate was 33%, i.e. 57 women delivered successfully (15% of all patients). There were a total of 62 babies. A statistically significant decline in PR and DR was noted with age (Figure 1), starting at 13.9 and 9.1% respectively for age 40 years and declining to 2.8 and 0.7% respectively at age 45 years (P < 0.001). There were 52 singleton deliveries, 50 of which were delivered after 37 weeks of gestation. The rate of Caesarean sections was 50%. The mean birth weight was 2890 g, but five babies weighed <2500 g. All the babies had Apgar scores at 1 and 5 min of >7; no major complications were recorded during pregnancy or at birth. Two singleton pregnancies ended with a preterm delivery before 37 weeks. One patient had placental abruption requiring emergency delivery by Caesarean section at 25 weeks; the baby weighed 890 g and died of complications of prematurity at 3 weeks. The second patient had a spontaneous preterm delivery at 26 weeks; the baby weighed 715 g and was discharged from the neonatal unit after several weeks. Five of the 88 pregnancies were twins: all five delivered after 35 weeks of gestation, with the mean birth weight 2400 g (range: 1880 2680). No neonatal complications were recorded at birth. Stratified by age, the rate of successful pregnancy outcome declined from 16.7% at 40 years to 3.3% at 45 years (Table 3). None of 10 patients treated after age 45 years delivered a baby. Cumulative pregnancy rates Calculated pregnancy rates per treatment cycle (for all women) were 3.2, 5.4 and 5.7% in the first, second, and third treatment cycle respectively (Figure 2). Thus, 75% of all pregnancies were achieved in the first two treatment cycles; only three of 88 (3.4%) pregnancies occurred after the sixth cycle of treatment. Factors influencing outcome In order to identify possible modifiers, women who became pregnant, and treatment cycles in which pregnancy was achieved, were compared with women and treatment cycles that were not successful. Women who became pregnant were younger, and had a smaller number of treatment cycles prior to conception. Treatment cycles resulting in pregnancy were characterized by the following: smaller dose of gonadotrophins, endometrial thickness > 8.5 mm, and a greater number of follicles, oocytes, and embryos transferred. Ovarian stimulation protocol, type of gonadotrophins, and fertilization technique (conventional IVF or ICSI) were not significantly different between successful pregnancy cycles and failed ones. Other clinical variables including IVF treatment before age 40 years, pregnancies before age 40 years, and cause of infertility, were also not significantly different between women who became pregnant and those who did not. In analysis of FSH assays, no correlation between elevated FSH and pregnancy or delivery rates was observed, whether a cut-off of either 10 or 15 IU/ml was employed or not. In multivariate logistic regression analysis of patient and treatment variables, the following factors were independently and significantly related to pregnancy: age at treatment (adjusted OR for clinical pregnancy 0.76, 95% CI 0.64 0.91), mean number of gonadotrophin ampoules (OR 0.98, 95% CI 0.97 0.99), number of follicles 18 mm (OR 1.11, 95% CI 1.01 1.22), endometrial thickness > 8.5 mm (OR 1.9, 95% CI 1.02 3.52), and number of embryos transferred (OR 1.32, 95% CI 1.08 1.61). The number of collected oocytes (as a reflection of ovarian response) was positively related to treatment success. When a cut-off limit of four oocytes was chosen as an indicator of a good ovarian response, it was found that while a significantly higher pregnancy rate was achieved by the younger women (ages 40 41 years, P > 0.0001) who attained this criterion, the effect was blunted in older patients (Table 4). Similarly, for all the women, both pregnancy and delivery rates increased when more embryos were transferred. The mean number of embryos transferred was 2.7, but in 32% of cycles 3 embryos were replaced. Transfer of 3 embryos correlated with a higher pregnancy rate in younger women (ages 40 41 years, P < 0.00001), but did not reach statistic significance in older women (Table 5). 351

Table 3. Outcome of 1217 IVF cycles in 381 women aged 40 years and older, stratified according to age. Parameter Patient age during IVF cycle All 40 41 42 43 44 45 46 7 ages No. of patients (%) 114 (30) 69 (18) 71 (19) 44 (12) 43 (11) 30 (8) 10 (3) 381 No. of treatment cycles (%) 209 (17) 230 (19) 237 (19) 199 (16) 150 (12) 143 (12) 49 (4) 1217 Cancellation prior to oocyte 14.4 13.5 17.7 17.0 16.7 19.6 26.1 16.6 retrieval (%) Mean no. of oocytes retrieved 6.5 5.4 5.8 4.8 5.5 4.7 3.2 5.5 Mean no. of embryos transferred 3.0 2.7 2.7 2.6 3.1 2.5 2.0 2.7 No. of clinical pregnancies 29 14 21 14 6 4 0 88 No. of live births 19 10 15 8 4 1 0 57 Pregnancy rate per initiated 13.9 6.1 8.9 7.0 4.0 2.8 0.0 7.3 cycle (%) a Miscarriage rate (%) 35 29 28 43 33 75 33 Delivery rate per initiated 9.1 4.3 6.3 4.0 2.7 0.7 0.0 4.7 cycle (%) a Patients having a live birth (%) 16.7 14.5 21.1 18.2 9.3 3.3 0.0 15 a Significant association between age and outcome (P < 0.01), with a significant linear trend (P < 0.001). 16 14 Pregnancy rate Delivery rate 12 10 10 % 8 6 4 2 Pregnancy rate (% of initiated cycles) 8 6 4 2 0 40 41 42 43 Age (years) 44 45 46 0 1 2 3 4 No. of IVF cycles 5 6 7 8 Figure 1. Pregnancy and delivery rates per initiated cycle according to age. Figure 2. Cumulative pregnancy rates in 1217 IVF treatment cycles in 318 women aged 40 years and above. 352

Table 4. Effect of number of oocytes retrieved on pregnancy rates at different age subgroups. Age (years) Cycles with 1 4 oocytes Cycles with 5 oocytes P-value Total no. No. of Total no. No. of of cycles pregnancies of cycles pregnancies (%) (%) 40 41 172 8 (4.6) 189 35 (18.5) <0.0001 42 43 195 16 (8.0) 148 22 (15.0) 0.04 44 45 127 2 (1.6) 100 8 (8.0) 0.016 Table 5. The effect of the number of embryos transferred at different age subgroups on pregnancy rate. NS = not statistically significant. Age (years) 1 2 embryos transferred 3 embryos transferred P-value Total no. No. of Total no. No. of of cycles pregnancies of cycles pregnancies (%) (%) 40 41 127 8 (0.7) 163 35 (21.0) <0.0001 42 43 136 15 (11.0) 129 20 (15.0) NS 44 45 86 2 (2.3) 94 8 (8.5) NS Discussion Most data on IVF outcome in older patients derive from multicentre experience and national registries, which, by their nature, include disparate data as a single cohort (American Society for Reproductive Medicine, 2004; Andersen et al., 2005). In addition, there is rarely a breakdown of rates of pregnancy and delivery by year from 40 to 46 years of age. The results reported herein comprise a single centre experience in older women, reported according to age at treatment. So far as is known, this is the largest report on IVF in women aged 40 years and older. In a broad sense, the present experience underscores that of Klipstein et al. from a very large infertility centre with 1263 women undergoing 2705 cycles: there was a gradual decrease in both pregnancy and delivery rates with each year over 40 years and there were no successful pregnancies when IVF treatment was begun after the age of 44 years (Klipstein et al., 2005). Klipstein reported a pregnancy rate of 28% if IVF was started after age 40 years, but 53% of cycles in that cohort were performed in the younger women (40 41 years) relative to only 36% of cycles in the present group. In an earlier report (Lass et al., 1998) in a total of 1087 cycles in 471 women between the ages of 40 44 years, the pregnancy rate was 11.3%, with a sharp decline in success after the age of 42 years. The present experience also implies that the age of the woman at advent of treatment is a critical variable, but in this series the majority of women (64%) were aged >42 years, relative to earlier reports. A review of the literature confirms that the international experience is better in women 40 41 years of age: the results in women 40 41 years are different from those in women 42 44 years of age, which in turn, are different from those above the age of 45 at the start of treatments. In a series of 163 women 40 years who had undergone 713 IVF cycles, the pregnancy and delivery rates were 16.6 and 10% respectively; of these, 60% were in women aged 40 41 years of age (Widra et al., 1996). Similar results were seen in 431 IVF cycles in women aged 41 years and older (Ron-El et al., 2000), and in 736 ICSI cycles in women aged 40 years and older (Grimbizis et al., 1998). In the latter study, for example, 53% of women were aged 40 41 years: embryos were transferred in 78% of initiated cycles, with clinical pregnancy and viable pregnancy rates per started cycle 11 and 7% respectively; a significant trend for fewer pregnancies and deliveries with each year of age was noted. Thus, in a report of oldest women, 708 cycles initiated in women aged 43 45 years of age, pregnancy rate per initiated cycles was only 6.6% (Orvieto et al., 2004). Women who conceived had a better ovarian response, expressed by higher serum oestradiol concentrations on HCG day, number of oocytes and embryos and lower doses of gonadotrophins used. This observation corroborates similar previous reports demonstrating the importance of a good ovarian response (Lass et al., 1998; Çiray et al., 2006). The ovarian response was invariably better in women aged 40 41 years relative to the older women. 353

354 As in reports by others, IVF treatment before age 40 years, pregnancies before age 40 years, basal FSH concentrations at advent of IVF, and cause of infertility did not significantly differ between women who became pregnant relative to those who did not. The fact that previous pregnancies did not improve pregnancy and delivery rates in older women undergoing assisted technology is unexpected and is in contradiction to the authors own experience (Laufer et al., 2004) in spontaneous pregnancies above age 45 years in grandmultiparous women. In the authors experience, basal FSH was not predictive of pregnancy and delivery rates. These observations should be interpreted with caution since FSH was measured prior to commencing IVF, and women who had repeated FSH over 20 IU/ml were not treated. As reported by others, the number of embryos transferred significantly influenced pregnancy and delivery rates (Widra et al., 1996; Ron-El et al., 2000; Opsahl et al., 2001; Klipstein et al., 2005); however, multi-fetal pregnancies are not common, and in this study as in others, twins were the only non-singleton births regardless of the number of embryos transferred. All twin pregnancies in this series went to term, highlighting the experience of others that multiple transfers may achieve acceptable pregnancy and delivery rates in older women without a concomitant increase in multi-fetal pregnancies and their complications. Recently, Heijnen have shown that in a slightly younger group of IVF patients, transfer of two embryos achieved similar pregnancy rates to transfer of three embryos with a significant reduction in multi-fetal pregnancies, although more treatment cycles will be needed (Heijnen et al., 2006). Another important difference between the present series and that of Klipstein et al. (2005) was that women who requested to continue beyond two cycles were allowed to carry on. The mean number of treatment cycles per patient in the present group was 3.3 versus 2.3 in the Klipstein study. Unlike most reports in the literature, Israeli women enjoy the availability of health care coverage for several cycles of IVF, even as the success rate declines. Since health policy in Israel does not strictly limit the number of treatment cycles per patient, it was possible to assess the biological potential for IVF success in older women. In conclusion, the present study suggests the following: (i) unnecessary delays in advent of treatment should be avoided; (ii) the effect of each year of age is very important in predicting rates of pregnancy and delivery; (iii) the woman s previous history, including cause of infertility and/or previous pregnancies and deliveries, does not affect the chances for conception with IVF at an older age; (iv) transfer of multiple embryos increases the chances of success without concomitant increased risk for multi-fetal pregnancies, although experience with twin pregnancies was good; (v) importantly, virtually all of the pregnancies that resulted in deliveries were uneventful and the majority of the babies born were healthy; and finally (vi) it is recommended that women should not consider starting IVF after the age of 44 years. With respect to the latter point, it may be illustrative to note that with regard to number of cycles, 75% of pregnancies in the present study occurred in the first two cycles, and 97% in the first six cycles. Among women aged 40 43 years, one in five to one in seven women had a baby, but fewer than one in 10 at age 44 years. With respect to the number of pregnancies achieved in the very oldest women, only one in 30 women undergoing a total of 143 IVF cycles at age 45 years had a baby, a rate very similar to those reported by others. No woman treated at an older age had a delivery. Thus, it is instructive to note that these results underscore the caveat that IVF in older women should be limited to a handful of cycles at most, and that IVF in women aged older than 45 years using their own oocytes is unjustified. These caveats should be taken into consideration when assessing IVF treatment in older women, in order to avoid unrealistic expectations. References American Society for Reproductive Medicine 2004 Assisted reproductive technology in the United States 2004; 2000 results generated from the American Society for Reproductive Medicine/ Society for Assisted Reproductive Technology Registry. Fertility and Sterility 81, 1207 1220. Andersen AN, Gianaroli L, Felberbaum R et al. 2005 Assisted reproductive technology in Europe, 2001. Results generated from European registers by ESHRE. Human Reproductionu 20, 1158 1176. Antinori S, Gholami GH, Versaci C et al. 2003 Obstetric and prenatal outcome in menopausal women: a 12-year clinical study. Reproductive BioMedicine Online 6, 257 261. Armstrong DT 2001 Effects of maternal age on oocyte developmental competence. Theriogenology 55, 1303 1322. Baird DT, Collins J, Egozcue J et al. 2005 Fertility and ageing. Human Reproduction Update 11, 261 276. Chambers GM, Ho MT, Sullivan EA 2006 Assisted reproductive technology treatment costs of a live birth: an age-stratified costoutcome study of treatment in Australia. Medical Journal of Australia 184, 155 158. Check JH, Lurie D, Callan C et al. 1994 Comparison of the cumulative probability of pregnancy after in vitro fertilizationembryo transfer by infertility factor and age. Fertility and Sterility 61, 257 261. Çiray HN, Ulug U, Tosun S et al. 2006 Outcome of 1114 ICSI and embryo transfer cycles of women 40 years of age and over. Reproductive BioMedicine Online 13, 516 522. Fasouliotis SJ, Simon A, Laufer N 2000 Evaluation and treatment of low responders in assisted reproductive technology: a challenge to meet. Journal of Assisted Reproduction and Genetics 17, 357 373. Grimbizis G, Vandervorst M, Camus M et al. 1998 Intracytoplasmic sperm injection, results in women older than 39, according to age and the number of embryos replaced in selective or non-selective transfers. Human Reproduction 13, 884 889. Heijnen EM, Klinkert ER, Schmoutziguer AP et al. 2006 Prevention of multiple pregnancies after IVF in women 38 and older: a randomized study. Reproductive BioMedicine Online 13, 386 393. Klipstein S, Regan M, Ryley DA et al. 2005 One last chance for pregnancy: a review of 2,705 in vitro fertilization cycles initiated in women age 40 years and above. Fertility and Sterility 84, 435 445. Lass A, Croucher C, Duffy S et al. 1998 One thousand initiated cycles of in vitro fertilization in women 40 years of age. Fertility and Sterility 70, 1030 1034. Laufer N, Simon A, Samueloff A et al. 2004 Successful spontaneous pregnancies in women older than 45 years. Fertility and Sterility 81, 1328 1332. Legro RS, Shackleford DP, Moessner JM et al. 1997 ART in women 40 and over. Is the cost worth it? Journal of Reproductive Medicine 42, 76 82. Opsahl MS, Blauer KL, Black SH et al. 2001 The number of embryos available for transfer predicts successful pregnancy outcome in women over 39 years with normal ovarian hormonal reserve testing. Journal of Assisted Reproduction and Genetics 18, 551 556. Orvieto R, Bar-Hava I, Yoeli R et al. 2004 Results of in vitro fertilization cycles in women aged 43 45 years. Gynecologic Endocrinology 18, 75 78.

Ron-El R, Raziel A, Strassburger D et al. 2000 Outcome of assisted reproductive technology in women over the age of 41. Fertility and Sterility 74, 471 475. Simon A, Holzer H, Hurwitz A et al. 1998 Comparison of cryopreservation outcome following intracytoplasmic sperm injection and conventional in vitro fertilization. Journal of Assisted Reproduction and Genetics 15, 431 437. te Velde ER, Pearson PL 2002 The variability of female reproductive ageing. Human Reproduction Update 8, 141 154. Widra EA, Gindoff PR, Smotrich DB et al. 1996 Achieving multipleorder embryo transfer identifies women over 40 years of age with improved in vitro fertilization outcome. Fertility and Sterility 65, 103 108. Ziebe S, Loft A, Petersen JH et al. 2001 Embryo quality and developmental potential is compromised by age. Acta Obstetricia et Gynecologica Scandinavica 80, 169 174. Received 16 October 2006; refereed 22 November 2006; accepted 12 January 2007. 355