Cumulative delivery rates after intracytoplasmic sperm injection: 5 year follow-up of 498 patients

Kaan Osmanagaoglu1, Herman Tournaye, Michel Camus, Mark Vandervorst, André 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
 References
 
The use of life-table analysis for infertility data has the advantages of clarity and ease of application. Success rates per cycle have been reported, but not cumulative delivery rates for intracytoplasmic sperm injection (ICSI). We selected retrospectively 498 Belgian patients <37 years old, who had their first ICSI cycle between July 1992 and December 1993. Follow-up was till the end of October 1997. Outcome measure was any delivery >25 weeks. These couples underwent 963 ICSI cycles using fresh ejaculated spermatozoa. The indications for ICSI were long-standing severe male infertility or fertilization failure after conventional in-vitro fertilization (IVF). Cumulative delivery rates were calculated by life-table analysis and compared according to age groups and sperm quality. There were 298 deliveries within a mean rate per cycle of 31%. The average number of cycles required for a delivery was 3.15 (CI 2.88; 3.43). Twenty-three (4.6%) spontaneous pregnancies occurred after the patients had finished therapy. There was no significant difference between the sperm quality groups but delivery rates decreased significantly with increasing female age. The real delivery rate after six cycles was 60%, while the expected cumulative delivery rate was 86%. This life-table analysis may provide a means by which to counsel couples on the likelihood of a delivery following ICSI.

Key words: cumulative delivery rates/ICSI/pregnancy/spermatozoa


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Since the first birth after in-vitro fertilization (IVF) and embryo transfer in 1978 (Steptoe and Edwards, 1978Go), many infertile couples have been treated successfully by IVF while others could not be helped because of fertilization failure or low sperm concentrations. The introduction of intracytoplasmic sperm injection (ICSI) in 1991 opened new perspectives for these couples (Palermo et al., 1992Go; Van Steirteghem et al., 1993Go). However, many patients have to undergo repeated ICSI cycles before a pregnancy ensues. Success rates for the technique have been reported per ICSI cycle, and so far no cumulative success rates are available, making counselling rather difficult.

The present study aimed to analyse cumulative delivery rates after ICSI by life-table analysis using a computerized ICSI database.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Study group
Four hundred and ninety-eight infertile couples who had their first ICSI treatment between July 1992 and December 1993 were included in the study, since July 1992 has been defined as the end of our ICSI learning curve (Tournaye et al., 1995Go). These patients underwent a total of 963 ICSI cycles up to the end of October 1997. The indications for ICSI were long-standing severe male infertility or fertilization failure after conventional IVF. The maximum number of cycles per patient included in this study was eight. In order to ensure optimal follow-up, only residents in Belgium were selected for this study.

Only those cycles which reached the stage of oocyte retrieval and in which fresh ejaculated spermatogen were used were included in the analysis. Frozen–thawed embryo transfers were not included. The main outcome measure was any delivery after 25 weeks of gestation. Patients with a successful delivery were not re-enrolled. The female patients were younger than 37 years at the time of their first ICSI cycle, the mean age being 30.5 years. The female patients were divided into three groups according to age: 20–29, 30–34 and 35–37 years. The age of each patient was calculated on the first day of the menstrual cycle of her first ICSI treatment cycle.

In a second analysis, the delivery rates were analysed according to the husband's semen profile. Sperm quality was expressed as total normal motile count [TNMC = volume (ml)xconcentration (106/ml)xpercentage progressive motility (a + b)/100xpercentage normal morphology/100]. Sperm morphology was assessed according to Kruger's strict criteria (Kruger et al., 1988Go). Patients were divided into three subgroups according to TNMC: <=0.8x106, 0.8x106 to 2.8x106, >2.8x106.

Follow-up
Follow-up was documented by our prospective follow-up programme for ICSI (Bonduelle et al., 1995). All the couples were asked to adhere to the follow-up conditions before starting ICSI treatment. The conditions required genetic counselling, agreement to prenatal karyotype analysis and a prospective clinical follow-up study of the children at 2, 12 and 24 months of age. Completion of a standardized questionnaire was requested where possible, as well as visits to the Centre for Medical Genetics (Dutch-speaking Brussels Free University, Brussels, Belgium) with the child after birth.

Those 193 patients who had stopped treatment in our centre were telephoned and information pertaining to the following was requested: (i) pregnancies occurring with or without infertility treatment outside our department; (ii) outcome of the pregnancies: delivery, spontaneous abortion, medical termination of the pregnancy or ectopic pregnancy; (iii) mode of conception: spontaneous (same partner or another partner), with other infertility treatment protocols or with treatment in another department; (iv) their status with regard to treatment: no further treatment or still continuing treatment. If infertility treatment was stopped, inquiries were made as to the main reasons for this: age, pregnancy, psychological aspects, fertilization or cleavage problems, other medical problems, divorce, death or the adoption of a child. A complete follow-up was obtained for 454 out of 498 patients (91%) according to the study inclusion criteria. The remaining 44 couples could not be contacted.

Clinical and laboratory procedures
A combination of long-term desensitizing gonadotrophin-releasing hormone agonist in association with human menopausal gonadotrophin was used in most cases. The details of sperm and oocyte assessment and handling have been extensively described elsewhere (Van Steirteghem et al., 1996Go). Up to three embryos, or in exceptional cases four, were transferred into the uterine cavity 48 h after the sperm injection procedure. Our embryo transfer policy has been described in detail previously (Staessen et al., 1995Go). Micronized progesterone (600 mg per day) was administered intravaginally in three separate doses for luteal-phase supplementation with or without human chorionic gonadotrophin (Smitz et al., 1988Go).

Statistical analysis
Cumulative delivery rates were estimated by life-table analysis using the Kaplan–Meier product limit procedure and differences between groups were assessed by the log–rank test. The Kaplan–Meier method was developed to calculate the estimated survival rates of cancer patients in different therapy models, and thus provides only expected delivery or pregnancy rates. A similar technique has been used elsewhere (Hull et al., 1992Go). A P level of 0.05 was used to test for significance of differences in results. Cumulative delivery rates were expressed as cumulative percentage probabilities with 95% confidence intervals (95% CI). These computational procedures were run on SPSS for Windows version 6.1 (SPSS Inc., Chicago, IL, USA). The intervals were calculated only where the number of patients was at least 20, since otherwise the CI would be too imprecise.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Life table analysis
Four hundred and ninety-eight patients underwent a total of 963 ICSI treatment cycles. The average number of cycles per patient was 1.93. Drop-out rates were approximately 20% per cycle until the fifth cycle, after which the rate increased to more than 30%. The proportion of patients quitting treatment per cycle increased with age. There were 298 deliveries. The mean delivery rate per cycle was 31%. The average number of cycles required for a delivery was 3.15 (2.88; 3.43) for all patients. The delivery rate per cycle remained stable until the fourth cycle and thereafter showed a drop of 10% per cycle. The cumulative delivery rates increased in the first three cycles. After the fourth cycle the increase in rates became less pronounced and reached a plateau at the sixth cycle. The real delivery rate after six cycles was 60% while the expected cumulative delivery rate was 86%.

Life-table analysis of the real and expected cumulative delivery rates is given in Figure 1Go, together with the dropout rates due to fertilization and/or cleavage problems. Cumulative delivery rates according to age group are shown in Table IGo and in Figure 2Go. The mean number of cycles needed to achieve a delivery in age groups 20–29, 30–34 and 35–37 years were respectively 2.80 (2.48; 3.11), 3.11 (2.72; 3.50) and 3.65 (2.95; 4.35). Pregnancy rates decreased significantly with increasing female age (P = 0.017). There was no significant difference in this respect between the sperm-quality groups (P = 0.101). The cumulative delivery rates according to the sperm-quality groups (TNMC) are shown in Table IIGo. All patients with fertilization failure following conventional IVF (n = 56) were in group 3 (Table IIGo). The 44 (9%) couples who could not be contacted were assumed not to be pregnant. There were 74 deliveries below 37 weeks including six triplets and 48 twin pregnancies. There were also five stillbirths at term.



View larger version (17K):
[in this window]
[in a new window]
 
Figure 1. Cumulative delivery rates after ICSI, showing the curves of expected deliveries (according to life-table analysis), real deliveries, total drop-outs and drop-outs due to fertilization and/or cleavage problems.

 

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



View larger version (32K):
[in this window]
[in a new window]
 
Figure 2. (A) Expected cumulative delivery rates after ICSI in three age categories, i.e. 20–29 years, 30–34 years, 35–37 years and in all age groups between 20 and 37 years. (B) Real cumulative delivery rates after ICSI in three age categories, i.e. 20–29 years, 30–34 years, 35–37 years and in all age groups between 20 and 37 years.

 

View this table:
[in this window]
[in a new window]
 
Table II. Cumulative delivery rate according to sperm quality, i.e. total normal motile count (TNMC)
 
Reasons for discontinuation of therapy
The 149 couples who stopped ICSI treatment without achieving a delivery were as follows: restricted numbers of cycles planned (n = 40), psychologically too stressful (n = 27), financial burden too high (n = 15), fertilization or cleavage problems (n = 13), other medical problems (n = 9), went to another centre (n = 9), too advanced maternal age (n = 6), divorce (n = 2), female partner died (n = 1), male partner died (n = 1), other reasons connected with work, relations, transport, etc. (n = 26). Twenty couples adopted a total of 28 babies.

Deliveries after discontinuation of therapy
There were 34 deliveries after cessation of ICSI therapy: five after donor insemination, one after oocyte donation, one spontaneous conception from another partner and four after infertility treatment in another centre. Spontaneous conception from the same partner occurred in 23 couples (4.6%).


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
It is important for the candidate couples and for the counselling physicians to know the probability of delivering a baby following ICSI treatment in a defined number of cycles. Life-table analyses have been used to estimate the success rates of assisted reproduction techniques (Hull et al., 1992Go). Life-table analysis assumes that those who continue and those who quit have the same probability of achieving the defined event (i.e. a clinical pregnancy or a delivery). Some authors argue that patients who discontinue the treatment may have done so because of poor treatment prognosis (Doody, 1993Go; Walters, 1994Go). However, another study (Haan et al., 1991Go) found no over-representation of patients with a poor prognosis in the group of drop-outs, and lack of any statistically significant difference in in-vitro fertilization for a poor treatment prognosis between drop-outs and those who continued has also been reported (Roest et al., 1998Go). In our study only 13 patients stopped ICSI because of poor treatment prognosis, i.e. 2.6% of the patient population. This small percentage is unlikely to cause a fundamental change in outcome figures. The curve of the expected cumulative delivery rate increased until the sixth cycle, after which it reached a plateau at 86%. The curve of the real cumulative delivery rate plateaued at the sixth cycle at 60%. This suggests that patients not older than 37 years may be counselled that 86% of them may deliver within six ICSI cycles (Figure 1Go).

As expected (Devroey et al., 1996Go), age had a strong effect on the cumulative delivery rates especially during the first cycle; 37% of the patients between 20–29 years old achieved a pregnancy compared with 31% in the age group of 30–34 years and only 23% between 35–37 years. A similar but a less pronounced trend was also observed for the subsequent cycles (Figure 2Go).

Sperm quality did not appear to influence cumulative delivery rates. This finding was not surprising, since a lack of effect of the three basic sperm parameters (total sperm count, sperm motility and morphology) on ICSI results has already been shown. This finding may be explained by positive selection: the spermatozoa which are actually injected into the oocytes are not necessarily representative of the whole sperm population as analysed in the initial sperm sample (Nagy et al., 1995Go). The real influence of sperm morphology on ICSI requires assessment of the specific spermatozoon that is injected.

In the follow-up, 34 deliveries were recorded in drop-out patients, 23 of which were pregnancies with the same partner and 14 of which were in group 1, (TNMC <=0.8x106) for semen characteristics. Twenty-two women out of 23 who became spontaneously pregnant from the same partner were younger than 34 years old. This low spontaneous conception rate of 4.6% is in accordance with a previous study reporting a 4% spontaneous pregnancy rate in azoospermic patients (Stanwell-Smith and Hendry, 1984Go). However a 6.6% cumulative spontaneous pregnancy rate after 12 months has been reported in couples with male subfertility waiting for IVF or ICSI (Evers et al., 1998Go). This confirms that the criteria for including oligoasthenoteratospermic patients in our ICSI programme were very strict.

The cumulative discontinuation rate was 34% after three cycles. The majority of couples who discontinued treatment had decided at the beginning of treatment to undergo a fixed number of cycles. Various other problems such as jobs, moving and organizational problems were the second main reason for discontinuing ICSI treatment, while psychological reasons were the third in rank.

In summary, while the real cumulative delivery rate reached a plateau of 60% after six ICSI cycles, the expected cumulative delivery rate may reach 86% after six cycles. Discontinuation of treatment because of poor treatment outcome was seen in only 2.6% of couples. Couples where the female partner was younger than 37 years may therefore be counselled that they have an 86% probability of delivery within six ICSI cycles with ejaculated spermatozoa. While female age had a pronounced negative effect, sperm quality did not influence the cumulative delivery rates. Life-table analysis provides a valuable means by which to counsel couples about the probability of delivery after ICSI. However, each ICSI programme should perform a similar analysis in order to offer realistic counselling. It is possible that statistically more reliable figures on cumulative delivery rate will be obtained in the future after a longer term study of a larger group.


    Acknowledgments
 
The authors wish to thank the clinical, paramedical and laboratory staff of the Centre for Reproductive Medicine. Furthermore, we are grateful to Ms Andrea Debrabanter for her help in collecting the follow-up data, to Mr Frank Winter of the Language Education Centre at our University for correcting the manuscript and to Mr Tom Loeys from the Department of Mathematics and Informatics at the University of Ghent for his advice in statistical calculations. Finally, we would like to thank Mr Hubert Joris for establishing and maintaining our computerized database and for all his technical assistance in ICSI procedures. This work was supported by grants from the Belgian Fund for Medical Research.


    Notes
 
1 To whom correspondence should be addressed Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Bonduelle, M., Aytoz, A., Van Assche, E. et al. (1998) Incidence of chromosomal aberrations in children born after assisted reproduction through intracytoplasmic sperm injection. Hum. Reprod., 13, 781–782.[Free Full Text]

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]

Doody, M. (1993) Dropout behaviour and fertility table analysis of pregnancy rates. Hum. Reprod., 8, 886–889.[Abstract]

Evers, J.L.H., de Haas, H.W., Land, J.A. et al. (1998) Treatment-independent pregnancy rate in patients with severe reproductive disorders. Hum. Reprod., 13, 1206–1209.[Abstract]

Haan, P., Bernardus, R.E., Hollanders, H.M.G. et al. (1991) Selective dropout in successive in-vitro fertilization attempts: the pendulum danger. Hum. Reprod., 6, 939–943.[Abstract]

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

Kruger, T.F., Acosta, A.A., Simmons, K.F. et al. (1988) Predictive value of abnormal sperm morphology in in vitro fertilization. Fertil. Steril., 49, 112–117.[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]

Palermo, G., Joris, H., Devroey, P. et al. (1992) Pregnancies after intracytoplasmic injection of single spermatozoon into an oocyte. Lancet, 340, 17–18.[ISI][Medline]

Roest, J., van Heusden, A.M., Zeilmaker, G.H. and Verhoeff, A. (1998) Cumulative pregnancy rates and selective dropout of patients in in-vitro fertilization treatment. Hum. Reprod., 13, 339–341.[Medline]

Smitz, J., Devroey, P., Camus, M. et al. (1988) The luteal phase and early pregnancy after combined GnRH-agonist/HMG treatment for superovulation in IVF or GIFT. Hum. Reprod., 3, 585–590.[Abstract]

Stanwell-Smith, R.E. and Hendry, W.F. (1984) The prognosis of male subfertility: a survey of 1025 men referred to a fertility clinic. Br. J. Urology, 56, 422–428.[ISI][Medline]

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 programmes. Hum. Reprod.,10, 3305–3312.[Abstract]

Steptoe, P.C. and Edwards, R.G. (1978) Birth after reimplantation of a human embryo. Lancet, 2, 366.

Tournaye H., Liu J., Nagy Z. et al. (1995) Intracytoplasmic sperm injection (ICSI): the Brussels experience. Reprod. Fertil. Dev., 7, 269–279.[ISI][Medline]

Van Steirteghem, A., Liu, J., Joris, H. et al. (1993) Higher success rate by intracytoplasmic sperm injection than by subzonal insemination. Report of a second series of 300 consecutive treatment cycles. Hum. Reprod., 8, 1055–1060.[Abstract]

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

Walters, D.E. (1994) A plea for a more fastidious and objective use of statistics. J. In Vitro Fertil. Embryo Transf., 60, 1065–1067.

Submitted on March 12, 1999; accepted on June 28, 1999.