Cumulative delivery rates after ICSI in women aged >37 years

Kaan Osmanagaoglu,1, Herman Tournaye, Efstratios Kolibianakis, Michel Camus, Andre Van Steirteghem and Paul Devroey

Centre for Reproductive Medicine, Dutch-speaking Brussels Free University, Laarbeeklaan 101, B 1090 Brussels, Belgium


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
BACKGROUND: Female patients aged >37 years have a poor prognosis after ICSI. To determine the cumulative delivery rates in these women by life-table analysis, 228 patients aged >37 years who had undergone a total of 437 ICSI cycles were analysed retrospectively. METHODS: Only cycles in which fresh ejaculated sperm was used, and in which at least one oocyte was micro-injected, were analysed. The main outcome measure was cumulative rate of deliveries. Any delivery after 25 weeks gestation was included in the study. RESULTS: In women aged 38–39 years, the real cumulative delivery rate after two cycles was 21%, while the expected delivery rate was 26%. In patients aged 40–43 years, the real and expected cumulative delivery rates were 12 and 17% respectively after three cycles, when they reached a plateau. There was only one delivery in the age group >43 years, which consisted of 26 patients with 66 cycles. In women aged >37 years, an expected cumulative delivery rate of 30% may be obtained at the end of the fourth cycle. Women aged >43 years do not have a realistic chance of achieving a delivery with their own oocytes. CONCLUSIONS: This life-table analysis provides a means by which to counsel couples about their chances of achieving a delivery by ICSI at an age >37 years.

Key words: age/cumulative pregnancy rates/delivery/ICSI/life-table analysis


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Following the introduction of IVF and embryo transfer in 1978, many couples were able to achieve pregnancies and deliveries using their own gametes, with the exception of those suffering from severe andrological infertility. In the early 1990s, the introduction of ICSI changed the perspectives for those men with very poor sperm characteristics (Van Steirteghem et al., 1993Go). Many couples in which the husband had a low sperm count were able to benefit from the new technique. Among them, there were patients aged >37 years. Fortunately, these patients would nowadays be helped before reaching the borderline age of 37 years, which is associated with sharply decreased fecundity (Padilla and Garcia, 1989Go). The group of female patients >37 years of age is currently the one with poor prognosis. Delivery rates after ICSI treatment in women <37 years were reported earlier (Osmanagaoglu et al., 1999Go). Here, attention is focused on the cumulative delivery rates in patients aged >37 years. These data may also be helpful in counselling such patients.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
In total, 228 patients who were aged >37 years at the time of their first ICSI cycle in our ICSI programme were included in this retrospective study. The mean female age was 40.1 years, the age of each patient being recorded as per the first day of the menstrual cycle of her first ICSI treatment. All women included in the study had an FSH level <15 IU/ml. All patients had their first ICSI treatment between July 1992 and December 1994, and follow-up was continued until the end of December 1998. Only those cycles in which fresh ejaculated sperm was used, and in which the stage of oocyte retrieval was reached, were included in this analysis. Cryopreserved embryo transfers were excluded from the life-table analysis. The 228 couples underwent a total of 437 ICSI cycles. In all the cycles included in the study, at least one metaphase II oocyte was injected with ejaculated sperm.

The rank of trial ranged from 1–8 per couple; the average rank of trial was 1.9. The main outcome measure was any delivery after 25 weeks gestation. Patients were not re-enrolled after a first delivery. The indications for ICSI were long-standing male infertility or fertilization failure after conventional IVF. Cumulative delivery rates were calculated by life-table analysis (Hull et al., 1992Go) in three age groups, namely 38–39, 40–43 and >43 years.

Pregnancy follow-up was documented by a prospective follow-up programme for ICSI (Bonduelle et al., 1999Go).

Contact was possible with 24 out of 188 patients who had stopped their treatment in our centre; the other 164 couples could not be reached either by telephone or letter for further follow-up. All of these couples were residents of another country. A complete follow-up was obtained for 64 patients (including 40 who delivered) out of 228 patients (28%).

Routinely, a long protocol was used, but exceptionally (in <1% of cycles) a short protocol was used in combination with a GnRH agonist, buserelin (Suprefact®; Hoechst, Frankfurt, Germany), and HMG (Humegon®; Organon, Oss, The Netherlands or Pergonal®; Serono, Geneva, Switzerland). The details of sperm and oocyte assessment and handling have been described extensively elsewhere (Van Steirteghem et al., 1996Go). The embryo transfer policy used has been described in detail previously (Adonakis et al., 1997Go). However, in patients aged >40 years, up to six embryos were transferred into the uterine cavity 48 h after the sperm injection procedure. For luteal phase supplementation, micronized progesterone (600 mg/day; Utrogestan; Piette, Brussels, Belgium) was administered intravaginally in three separate doses with or without additional HCG (3x1500 or 5000 IU as a single dose; Pregnyl; Organon or Profasi; Serono) (Staessen et al., 1995Go).

Cumulative delivery rates were calculated by life-table analysis using the Kaplan–Meier product limit procedure (Kaplan and Meier, 1958Go) and differences between groups were assessed by the log-rank test. A P-value of 0.05 was assumed to test for significance of differences in results. Cumulative delivery rates were expressed as cumulative percentage probabilities with their 95% confidence intervals (CI). These computational procedures were performed using Statistics Package for Social Sciences (SPSS) for Windows version 6.1 (SPSS Inc., Chicago, IL, USA).


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Cumulative delivery rates
Among 228 patients, the average number of cycles per patient was 1.91. Overall, there were 40 deliveries; four deliveries were <37 weeks gestation, including one twin pregnancy, giving a total of three twin pregnancies. One stillbirth at term was also noted. The mean delivery rate per cycle was 9%. The average numbers of cycles required for a delivery were 3.91 (95% CI 3.45, 4.38), 5.50 (95% CI 4.87, 6.14) and 7.55 (95% CI 6.70, 8.39) for the patients in the respective age categories of 38–39, 40–43 and >43years.

In the 38–39 years age group, the real cumulative delivery rate was 21%, and the expected delivery rate was 26% after two cycles, following which a plateau was reached.

In the 40–43 years age group, the real cumulative delivery rate was 12%, and the expected delivery rate was 17% after three cycles, following which a plateau was reached. There was only one delivery in the >43 years age group, which comprised 26 patients who underwent a total of 66 cycles.

When all age groups were analysed together, the expected cumulative delivery rate was seen to increase up to the sixth cycle but, owing to the small numbers of patients, figures should not be considered as clinically relevant above four cycles.

Cumulative delivery rates according to age groups are shown in Table IGo. Life-table analysis of the real and expected cumulative delivery rates is shown in Figure 1Go, together with the drop-out rates. The drop-out rate was ~40% per treatment cycle, and 60% of patients stopped treatment after their first two cycles. Reasons for withdrawal included: psychologically too heavy a treatment; too-advanced maternal age; and financial burden too high. An analysis of the computerized database showed that 36 of 228 patients (16%) stopped their treatment due to fertilization or cleavage problems. Increasing age did not influence the drop-out rate of patients. The tendency towards decreasing pregnancy rates with increasing female age was statistically significant (P = 0.0039). The 164 couples (72%) who were not contacted during follow-up were assumed not to be pregnant.


View this table:
[in this window]
[in a new window]
 
Table I. Cumulative delivery rates according to age group in patients >37 years old
 


View larger version (12K):
[in this window]
[in a new window]
 
Figure 1. Cumulative delivery rates (CDR) after ICSI. Curves for expected deliveries (according to life-table analysis), real deliveries and drop-outs in all age groups >37 years.

 
There were three deliveries after ending ICSI therapy. There were no deliveries after donor insemination, after oocyte donation, after assisted reproductive treatment in another centre, or as the result of a spontaneous pregnancy in a new relationship. Spontaneous conception within the same relationship occurred in three couples.


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Although cumulative delivery rates after ICSI treatment in women aged <37 years have been previously reported (Osmanagaoglu et al., 1999Go), the cumulative efficiency of ICSI in women aged >37 years has not yet been identified. Cumulative delivery rates are particularly important with a view to counselling couples.

It is well known that increasing age is associated with a decrease in pregnancy rate, either in natural conception or after assisted reproductive techniques (Federation CECOS, 1982Go; Van Noord-Zaadstra et al., 1991Go; Scott et al., 1995Go; Devroey et al., 1996Go; Grimbizis et al., 1998Go). Ovarian reserve decreases with advancing age, leading to a decrease in the numbers of oocytes obtained (Faddy and Gosden, 1996Go), an increase in the incidence of aneuploidy in oocytes (Munné et al., 1995Go), and poor oocyte quality (Navot et al., 1991Go). Age is therefore the most important limiting factor for the success of assisted reproduction in women aged >37 years. The onset of this phenomenon is individually variable (Grimbizis et al., 1998Go), and there may therefore be a subgroup of patients with an adequate follicular response to stimulation yielding higher success rates after assisted reproduction (Vandervorst et al., 1997Go; Grimbizis et al., 1998Go).

In older women, up to six embryos were replaced back in the uterus, as the chance of chromosomally abnormal embryos increases with the patient's age. This in turn can result in a poor reproductive outcome, as most chromosomally abnormal embryos will not implant or will lead to miscarriage. Although increasing the number of embryos transferred may not have been standard practice in other centres at the time this study was performed, it is now known that at least 60% of embryos in women >37 years of age are chromosomally abnormal (Staessen et al., 2001Go). Therefore, in older women in whom no pre-implantation diagnosis is performed, it is justifiable to transfer more embryos after consulting the patient. This can increase the chance of conception without a proportional increase in the chance of multiple pregnancy.

It is widely accepted that non-ovarian factors have a limited effect on reproductive potential (Scott and Hofmann, 1995Go); for example, sperm parameters hardly influence the results of ICSI outcome (Nagy et al., 1995Go; Osmanagaoglu et al., 1999Go).

The findings of this study with regard to ICSI treatment in women aged >37 years are comparable with those from studies dealing with IVF. Most pregnancies were observed during the first four treatment cycles (Guzick et al., 1986Go; Haan et al., 1991Go; Tan et al., 1992Go). In a comparable study, only 14 patients aged >41 years among a total of 337 patients aged >36 years, had shown an ongoing pregnancy rate of 24% after two IVF/ICSI treatment cycles (Stolwijk et al., 2000Go). These figures corroborate the findings of the present study, namely a 21% delivery rate in the group of women aged 38–39 years. A delivery rate of 8.7% per embryo transfer in women aged 40–44 years has also been reported by others (Lass et al., 1998Go), and a comparable birth rate was observed in the present study in women aged 40–43 years. Women aged >43 years appear not to have a realistic chance of delivery of a child, even after repeated cycles; indeed, only one delivery was observed out of 66 cycles in 26 patients aged >=44 years. Some of these patients went through up to eight cycles, despite being aware of their limited chances. The findings of the present study corroborate those of others (Ron-El et al., 2000Go), who did not record any deliveries in women aged >43 years among a group of 56 patients.

In a previous study published by our group on cumulative delivery rates in women aged <37 years, a decrease of 10% in real cumulative delivery rate was observed between the age groups of 20–29, 30–34 and 35–37 years. While in the 35–37 years age group the real cumulative delivery rate was 45%, it fell to 24% in the present study in women aged 38–39 years (Osmanagaoglu et al., 1999Go).

Three twin pregnancies occurred among patients aged >=40 years. In a previous study, six triplet and 48 twin pregnancies were reported out of 298 deliveries in women aged <37 years (Osmanagaoglu et al., 1999Go).

It has been suggested (Engmann et al., 1999Go) that when calculating cumulative delivery rates by life-table analysis, the age of each woman should be recorded at the start of her first treatment cycle, though this causes an underestimation of age-related cumulative pregnancy rates. For example, when a woman starts her first treatment at 34 years, her second at 37 years and her third at 40 years, calculation of the cumulative pregnancy rate will consider her to be 34 years. This calculation method introduces a biased decrease due to advancing female age (Engmann et al., 1999Go). In order to minimize this pitfall, the reproductive period was divided into six age groups. The first three groups, which represent the most fertile period of reproductive life, have already been described (Osmanagaoglu et al., 1999Go; Figure 2Go). Discontinuation of treatment because of poor outcome was seen in only 2.6% of patients aged <37 years (Osmanagaoglu et al., 1999Go). As expected in patients aged >37 years, the drop-out rate was much higher; moreover, 16% (36/228) of the patients discontinued their treatment due to fertilization or cleavage problems. Consequently, in this group of patients the application of life-table analysis may indeed overestimate the expected results.



View larger version (16K):
[in this window]
[in a new window]
 
Figure 2. Comparison of real cumulative delivery rates in women aged less and more than 37 years. The solid curves represents data points where outcome involved adequate numbers of couples (~30); the dotted curves indicate where fewer couples renders the cumulative calculation less reliable. Data points are not shown where the couple number is <10. This would result in the 37–40 years curve having three data points, the 40–43 years curve four data points, and the >43 years curve three data points.

 
Life-table analysis also tends to overestimate the expected cumulative rates when the group studied is too small. A striking example of this situation is shown in Table IGo for the 38–39 years age group. While at the third cycle the expected cumulative delivery rate was 32%, it increased to 66% at the fifth cycle by virtue of only two deliveries in a subgroup of four patients. A similar overestimation was observed in the 40–43 years age group, where the expected cumulative delivery rate rose from 17% at the third cycle to 49% at the sixth cycle on the basis of only four deliveries. Thus, when applied to small groups of patients, life-table analysis may distort results dramatically. It is suggested therefore that life-table analysis be applied to groups of at least 30 patients in order not to overestimate the outcome. Comparing results from different fertility centres can also be misleading due to differences in the fertility patients themselves, and to differing ovulation stimulation protocols, statistical methodologies, definitions and expertise of the centre (Check et al., 1994Go; Stolwijk et al., 1996Go).

In limiting our findings to subgroups of at least 30 patients, it may be concluded that after three ICSI cycles, the real cumulative delivery rate reached a plateau at 15% in women aged >37 years. However, an expected cumulative delivery rate of 21% may be obtained at the end of the third cycle, assuming that no selective drop-outs occur. Women aged >43 years do not have a realistic chance of achieving a delivery using their own oocytes. The present results show a real and expected cumulative delivery rate of 2% after three cycles. This analysis provides an important means of counselling patients about their chances of achieving a delivery by ICSI when they are aged >37 years.


    Acknowledgements
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The authors wish to thank the clinical, scientific, nursing and technical staff of the Centre for Reproductive Medicine, and especially the nurses and technicians involved in the follow-up project. Mr Frank Winter of the Language Education Centre at our University corrected the manuscript. These studies were supported by grants from the Fund for Scientific Research, Flanders.


    Notes
 
1 To whom correspondence should be addressed at: E-mail: osmanagaoglu{at}hotmail.com Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Adonakis, G., Camus, M., Joris, H. et al. (1997) The role of the number of replaced embryos on intracytoplasmic sperm injection outcome in women over the age of 40. Hum. Reprod., 11, 2542–2545.

Bonduelle, M., Camus, M., De Vos, A. et al. (1999) Seven years of intracytoplasmic sperm injection and follow-up of 1987 subsequent children. Hum. Reprod., 14 (Suppl. 1), 243–264.[ISI][Medline]

Check, J.H., Lurie, D., Callan, C. et al. (1994) Comparison of the cumulative probability of pregnancy after in vitro fertilization–embryo transfer by infertility factor and age. Fertil. Steril., 61, 257–261.[ISI][Medline]

Devroey, P., Godoy, H., Smitz, J. et al. (1996) Female age predicts embryonic implantation after ICSI: a case controlled study. Hum. Reprod., 11, 1324–1327.[Abstract]

Engmann, L., Maconochie, N., Bekir, J.S. et al. (1999) Cumulative probability of clinical pregnancy and live birth after a multiple cycle IVF package: a more realistic assessment of overall and age-specific success rates? Br. J. Obstet. Gynaecol., 106, 165–170.[ISI][Medline]

Faddy, M.J. and Gosden, R.G. (1996) A model confirming the decline in follicle numbers to the age of menopause in women. Hum. Reprod., 11, 1484–1486.[Abstract/Free Full Text]

Federation CECOS (1982) Female fecundity as a function of age. N. Engl. J. Med., 306, 404–406.[ISI][Medline]

Grimbizis, G., Vandervorst, M. and Camus, M. (1998) Intracytoplasmic sperm injection, results in women older than 39, according to age and the number of oocytes replaced in selective or non-selective transfers. Hum. Reprod., 13, 884–889.[Abstract]

Guzick, D.S., Wilkes, C., Jones, H.W. Jr (1986) Cumulative pregnancy rates for in vitro fertilization. Fertil. Steril., 46, 663–667.[ISI][Medline]

Haan, P., Bernardus, R.E., Hollanders, H.M.G. et al. (1991) Results of IVF from a prospective multicentre study. Hum. Reprod., 6, 805–810.[Abstract]

Hull, M.G.R., Eddozes, H.A., Fahy, U. et al. (1992) Expectations of assisted conception for infertility. Br. Med. J., 304, 1465–1469.[ISI][Medline]

Kaplan, E.L. and Meier, P. (1958) Nonparametric estimation from incomplete observations. J. Am. Statist. Assoc., 53, 457–481.[ISI]

Lass, A., Croucher, C., Duffy, S. et al. (1998) One thousand cycles of in vitro fertilization in women 40 years of age. Fertil. Steril., 70, 1030–1034.[ISI][Medline]

Munné, S., Alikani, M., Tomkin, G. et al. (1995) Embryo morphology, developmental rates, and maternal age are correlated with chromosome abnormalities. Fertil. Steril., 64, 382–391.[ISI][Medline]

Nagy, Z.P., Liu, J., Joris, H. et al. (1995) The result of intracytoplasmic sperm injection is not related to any of the three basic sperm parameters. Hum. Reprod., 10, 1123–1129.[Abstract]

Navot, D., Bergh, A.B., Williams, M.A. et al. (1991) Poor oocyte quality rather than implantation failure as a cause of age related decline in female fertility. Lancet, 337, 1375–1337.[ISI][Medline]

Osmanagaoglu, K., Tournaye, H., Camus, M. et al. (1999) Cumulative delivery rates after ICSI: a five-years follow-up of 498 patients. Hum. Reprod., 14, 2651–2655.[Abstract/Free Full Text]

Padilla, S.L. and Garcia, J.E. (1989) Effect of maternal age and number of in vitro fertilization procedures on pregnancy outcome. Fertil. Steril., 52, 270–273.[ISI][Medline]

Ron-El, R., Raziel, A., Strassburger, D. et al. (2000) Outcome of assisted reproductive technology in women over the age of 41. Fertil. Steril., 74, 471–475.[ISI][Medline]

Scott, R.T. Jr and Hofmann, G.E. (1995) Prognostic assessment of ovarian reserve. Fertil. Steril., 63, 1–11.[ISI][Medline]

Scott, R.T., Opsahl, M.S., Leonardi, M.R. et al. (1995) Life table analysis of pregnancy rates in a general infertility population relative to ovarian reserve and patient age. Hum. Reprod., 10, 1706–1710.[Abstract]

Staessen, C., Nagy, Z.P., Liu, J. et al. (1995) One year's experience with elective transfer of two good quality embryos in the human in-vitro fertilization and intracytoplasmic injection programs. Hum. Reprod., 10, 3305–3312.[Abstract]

Staessen, C., Van Assche, E., Platteau, P., Devroey, P. et al. (2001) The impact of blastocyst transfer with or without preimplantation genetic diagnosis for aneuploidy screening in couples with advanced maternal age. Hum. Reprod., 16 (Abstract Book 1), 23–24.[Free Full Text]

Stolwijk, A.M., Hamilton, C.J., Hollanders, J.M. et al. (1996) A more realistic approach to the cumulative pregnancy rate after in-vitro fertilization. Hum. Reprod., 11, 660–663.[Abstract]

Stolwijk, A.M., Wetzels, A.M.M. and Braat, D.D.M. (2000) Cumulative probability of achieving an ongoing pregnancy after in-vitro fertilization and intracytoplasmic sperm injection according to a woman's age, subfertility diagnosis and primary or secondary subfertility. Hum. Reprod., 15, 203–209.[Abstract/Free Full Text]

Tan, S.L., Royston, P., Campbell, S. et al. (1992) Cumulative conception and live birth rates after in-vitro fertilisation. Lancet, 339, 1390–1394.[ISI][Medline]

Vandervorst, M., Joris, H., Van Steirteghem, A. et al. (1997) Correlation between ongoing pregnancy rates and the number of cumulus–oocyte complexes retrieved in agonist-HMG-stimulated ICSI cycles. Hum. Reprod., 12, 168–169.

Van Noord-Zaadstra, B.M., Looman, C.W., Alsbach, H. et al. (1991) Delaying child bearing: effect of age on fecundity and outcome of pregnancy. Br. Med. J., 302, 1361–1365.[ISI][Medline]

Van Steirteghem, A.C., Nagy, Z., Joris, H. et al. (1993) High fertilization and implantation rates after intracytoplasmic sperm injection. Hum. Reprod., 8, 1061–1066.[Abstract]

Van Steirteghem, A., Nagy, P., Joris, H. et al. (1996) The development of intracytoplasmic injection. Hum. Reprod., 11 (Suppl.1), 59–72.[ISI][Medline]

Submitted on February 13, 2001; resubmitted on August 28, 2001; accepted on November 28, 2001.