全文

Abstract

Objectives

Women delivering singletons after in vitro fertilization (IVF) procedures have a greater risk of preterm delivery (PD). The aim of our study was to analyze PD risk factors and to identify those that could possibly be prevented.

 

Study design

In our matched controlled study we analyzed 1127 singleton deliveries after IVF and transfer of fresh embryos performed at the University Medical Center Ljubljana between 1 January 2002 and 31 December 2010. For every delivery included in the study group we chose three consecutive controls matched by maternal age, parity and maternity hospital. The main outcome measure was PD (<37 weeks). Investigated variables were: previous PD (PPD), conization, pregestational diabetes mellitus, uterine anomaly, operation on the uterus, chronic renal disease, maternal age and parity, and body mass index (BMI). Variables investigated within the IVF group were: stimulation protocol, laboratory procedure, number of retrieved oocytes and number and quality of transferred embryos.

 

Results

The PD rate after IVF was 1.5 times higher than after natural conception (11.5% in the IVF group and 7.7% in the control group, p < 0.001). Conization and chronic renal disease were shown to be significant risk factors for PD in both the IVF group and the naturally conceiving controls. BMI > 30 was an important risk factor only in the IVF group (OR 1.86 (1.06–3.27) vs. 1.10 (0.67–1.80)) and PPD only in the controls (OR 1.83 (0.78–4.28) vs. 3.22 (1.55–6.67)). Among the investigated PD risk factors, an IVF procedure was shown to be the fifth most important one. On analyzing parameters of the ovarian stimulation and IVF procedure, no PD risk factor was identified.

 

Conclusions

IVF was shown to be a significant risk factor for PD. In the IVF population, BMI plays a far more important role in PD than in the fertile population. In our research PD reoccurrence in IVF group was less than expected, which could perhaps be explained by the surgical correction of gynecological pathology and, where necessary, its being combined with cerclage. The investigation of parameters related to the IVF procedure did not identify any risk factors for PD.

 

Keywords

Preterm delivery; IVF; Singletons; Risk factors; BMI
 

Introduction

Women delivering after an in vitro fertilization (IVF) procedure are at increased risk of preterm delivery (PD) [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13] and [14]. Previous authors have reported up to two times higher incidence of PD in women conceiving singletons in an IVF procedure compared to naturally conceiving controls [1], [2], [3], [4] and [5]. The PD rate is higher than in a fertile population, even a subfertile population needing 2 years for conception [9]. It is not clear whether the higher PD rate is a consequence of the IVF procedure itself or the characteristics of the infertile population [9]. Since there is a lower PD rate after a frozen-thawed embryo transfer, the hyperestrogenism after ovarian stimulation was considered to be one of the risk factors [4], [11] and [12]. In previous reports, some pregnancy-related complications, such as placenta previa and placental abruption, are significantly more common among the IVF population and consequently contribute to PD [2].

 

Considering maternal factors for PD, we do know some risk factors from the general population, like advanced age, chronic renal disease, previous conization or previous PD. It is possible that these factors are more common in the IVF population or, on the other hand, that they perhaps act under different rules. It is known that women delivering after an IVF procedure are usually older than women conceiving naturally, but the influence of age is not enough to explain their higher PD rate [10]. Infertility per se also plays an important role in poor pregnancy outcome [15].

 

Considering the findings of previous reports, our hypothesis was that factors from the IVF procedure as well as women's own risk factors for PD contributed to an increased rate of PD after an IVF procedure. This study was designed to identify the risk factors, especially those that could perhaps be minimized or prevented in the future.

 

The aim of our study was to examine PD risk factors connected to IVF procedures and to analyze woman's risk factors of PD in the IVF population and in naturally conceiving controls.

 

Materials and methods

Study design

In our matched controlled study we analyzed 1,127 singleton deliveries after IVF and transfer of fresh embryos performed at the University Medical Center Ljubljana between 1 January 2002 and 31 December 2010. For every delivery included in the study group we identified three consecutive controls matched by age, parity and maternity hospital.

 

In Slovenia all 14 maternity hospitals systematically collect data on maternal demographic characteristics, medical, gynecological and reproductive history, prenatal care, pregnancy, delivery, the postpartum period and neonates for each mother-infant pair, using the same definitions of variables and same form of medical record. Data are by default sent to the National Institute of Public Health of the Republic of Slovenia.

 

The study was performed in accordance with the Personal Data Protection Act of Slovenia. Considering that this is a retrospective analysis of anonymized data from the National Register, additional approval from the Medical Ethics Committee was not necessary.

 

Definitions

PD was defined as delivery before completed 37 weeks of gestation. Gestational age in the control group was determined in weeks since the last menstrual period, confirmed or corrected by early ultrasonography. In the IVF group it was determined as the number of weeks between the oocyte retrieval and the delivery plus two weeks.

 

Nulliparous women were women who had never given birth to a fetus of more than 22 weeks’ gestation. Body mass index (BMI) was calculated from the pre-pregnancy weight.

 

Prior premature delivery (PPD) was regarded as spontaneous or induced vaginal delivery or elective cesarean section before 37 completed weeks of gestation. “Old primipara” was defined as a woman aged 33 and over at the time of her first delivery and “old multipara” as a woman aged 36 and over at the time of the observed delivery. Cold knife cone biopsy or loop electrosurgical excision procedure (LEEP) was regarded as conization. “Adnexal surgery” included all ovarian surgical procedures regardless of indication. The definition of uterine surgery included removal of uterine myomas and correction of uterine anomalies. Good quality embryos were all embryos with 8 or more cells transferred on day 3 or blastocysts without significant fragmentation transferred on day 5, whereas embryos in earlier development stages were classified as low quality embryos.

 

Multivariate logistic regression was used to identify independent risk factors for PD. Odds ratios (OR) and 95% confidence intervals (95% CI) were calculated with two-sided probability (p) values, and a p-value of <0.05 was considered as significant. Statistical analysis was performed using IBM SPSS Statistics, version 19 (IBM Corp, Armonk, NY).

 

Results

The analysis included 1127 singleton deliveries in the IVF group and 3,381 singleton deliveries in the control group. In both groups the average maternal age was 33.43 ± 3.94 years and women were predominantly primiparous (72.3%).

 

The PD rate in the IVF group was significantly higher than in the control group, 11.5% and 7.7% respectively (p < 0.001).

 

Multivariate logistic regression of the study and control groups, showing the significance of PD risk factors, is presented in Table 1. The IVF procedure is shown to be a less important risk factor than chronic renal disease, previous conization, PPD and uterine anomaly.

 

         

 

Multivariate logistic regression for PD risk factors in the IVF and control groups is presented in Table 2. In the IVF group the significant risk factors for PD were: conization, chronic renal disease and BMI >30. In the control group the significant risk factors for PD were PPD, conization, uterine anomaly and chronic renal disease. The influence of an uncorrected uterine anomaly could not be evaluated, because there were five women with uterine anomalies (four women with a unicornuate uterus and one with uterus duplex) in the IVF group, of whom none delivered prematurely. In the control group there were 23 women with uterine anomalies (four with a unicornuate uterus, 16 with a uterine septum and three with a bicornuate uterus) of whom seven delivered prematurely. There were no women with pregestational diabetes mellitus who delivered preterm.

 

         

 

Subgroups of women who were subject to PPD preventive measures (surgical correction of gynecological pathology and cerclage) were also evaluated. In the IVF group more surgical corrections of gynecological pathology had been performed (p = 0.005), while there was no difference in cerclage as a preventive measure ( Table 3).

 

        

 

Multivariate logistic regression for PD risk factors in the IVF group is presented in Table 4. Investigated parameters included the number of oocytes retrieved, the protocol of stimulation, the quality and number of embryos transferred and the fertilization procedure (IVF or ICSI), none of which was shown to be a significant PD risk factor.

 

         
         

 

In both groups in this 9-year observation period, no statistically significant trend in PD prevalence was found using logistic regression (IVF group p = 0.403, OR = 0.970 and control group p = 0.086, OR = 1.045).

 

Discussion

A previously observed and reported finding is that in comparison to natural conception the PD rate is higher after an IVF procedure, even in singleton pregnancies [1], [2], [5], [10] and [13]. This was confirmed by the results obtained from our study, where the PD rate was 11.5% in the IVF group and 7.7% in the control group (p < 0.001), even though controls were matched for age and parity.

 

Some previous studies were conducted to clarify this phenomenon and to determine risk factors that could be avoided in order to ensure better pregnancy outcomes following IVF. What we do know until now is that subfertility per se probably plays an important role in the poorer perinatal outcome in singleton deliveries after IVF [9], [15], [16] and [17].

 

We investigated some potential PD risk factors in both groups of women. The IVF procedure proved to be the fifth among the investigated PD risk factors in our study (Table 1). Conization and chronic renal disease, as expected [18] and [19], were shown to be an important PD risk factors in both groups, but there were also some differences between the two groups (Table 2). The first is BMI, which seems to be quite irrelevant in women who conceive naturally but is, on the other hand, an important risk factor in the IVF group. Women with a BMI above 30 in the IVF group had a higher incidence of PD, as shown in previous studies [20] and [21]. It seems that the negative impact of a higher BMI is stronger in the IVF group than in than in the group of women who conceived naturally. Therefore women undergoing IVF should be encouraged to normalize their BMI, not only in order to improve their chances of conceiving and to decrease spontaneous abortion rate, which has been shown in previous studies, but also to avoid PD [22], [23], [24] and [25].

 

The second difference between the two groups is the presence of a previous PD. PPD was shown to be a very important risk factor in women who conceived naturally, but quite insignificant in women undergoing IVF. Trying to explain this, we examined what preventive measures had been taken after PPD. We found that infertile women had significantly more gynecological surgery (correction of uterine anomalies, or removal of myomas or adnexal masses) aimed at restoring their fertility. Fertile women were probably not systematically investigated and possible uterine or adnexal pathologies were not eliminated. There was no difference in cerclage as a preventive measure. A limitation of our study perhaps lies in the fact that we do not have data on other potential preventive measures, like progesterone intake. Considering our data we can expect that systematic treatment of uterine or adnexal pathology could prevent PD recurrence in fertile women. At first glance a 20.5% rate of PD recurrence in this high-risk subgroup of women could lead us to assume that preventive measures were ineffective. Nevertheless, we have to keep in mind that the rate of PD among women conceiving after IVF in our research is 1.5 times higher than in naturally conceiving controls. Therefore, the expected recurrence rate of PD in this subgroup was approximately 30%.

 

The presence of a uterine anomaly could not be evaluated in this analysis, since all five women with an uncorrected uterine anomaly in the IVF group delivered after the 37th week. In the control group the PD rate in women with uterine anomalies was 26%, which was expected according to the literature [27].

 

The analysis of IVF procedure parameters failed to identify any PD risk factor.

 

In previous studies, controlled ovarian stimulation was investigated from a different perspective. A hyperestrogenic state at stimulation and multiple corpora lutea after ovarian puncture were suggested as having a negative effect on implantation and early pregnancy [24]. Considering this hypothesis, we could expect a higher ratio of preterm deliveries in women from whom a higher number of oocytes had been retrieved. Our research failed to confirm this hypothesis, however, as there was no established connection between the PD rate and the number of oocytes. Some earlier reports found the same [14] and [22].

 

A comparison of the IVF versus ICSI laboratory procedures in our research showed no difference in the PD rate, as was also shown in previous studies [5], though some other authors reported a lower rate of preterm deliveries after ICSI [4] and [26]. This could be explained by the fact that ICSI is used predominately in the case of male factor infertility and the reproductive health of women undergoing ICSI is supposed to be normal (or better than other IVF patients) [4] and [28]. Since the indications for ICSI are getting wider, however, and there are often factors influencing fertility in both partners, it is necessary that this should be verified in a subgroup of healthy women.

 

Considering the number of embryos transferred, it has been reported that in approximately 10% of singleton deliveries after double embryo transfer, there is a vanishing twin phenomenon. This phenomenon increases the risk of PD compared to deliveries after single embryo transfer [29]. A limitation of our study is that we do not have data on vanishing twins. The number and quality of transferred embryos was not shown to be a significant important risk factor of PD. The point is that even slower developing embryo with no genetic abnormalities that would lead to spontaneous abortion can implant normally and has the same probability to be carried to term as the fast developing embryo. The same was shown in previous studies [14] and [24].

 

The PD rate was not affected by the use of different stimulation protocols, like GnRH agonists or GnRH antagonists compared to IVF in minimally stimulated cycles.

 

The IVF procedure per se was shown to be the fifth among the investigated PD risk factors, after chronic renal disease, conization, PPD and uterine anomaly (Table 1). We believe that efforts should be made to optimize the IVF procedure in order to lower the PD rate. A contributing factor toward lowering the PD rate, however, could also lie in careful consideration of expectant management vs. conization and investigating the cause of PPD, especially in the case of uterine anomalies that can be surgically corrected.

 

Conclusion

Our study showed that women conceiving as a result of an IVF procedure are at 1.5 times greater risk of PD compared to naturally conceiving women. In the IVF population, BMI over 30 plays a far more important role in PD than in the fertile population. In women who had experienced PPD, surgical correction of gynecological pathology, combined with cerclage, where necessary, led to lower PD recurrence. The investigation of parameters related to the IVF procedure failed to identify any risk factors for PD.

 

Acknowledgments

The authors thank the nurses at the Department of Human Reproduction at the University Medical Centre Ljubljana for maintaining the databases of in vitro fertilization procedures, and the obstetricians in Slovenia for reporting to the National Perinatal Information System.