Singleton pregnancy outcomes after assisted and non‐assisted reproductive technology in infertile patients

Abstract

Purpose

Singleton pregnancy after assisted reproductive technology (ART) has been associated with higher risks of adverse pregnancy outcome than naturally conceived singleton pregnancy. This study was to elucidate whether the ART procedure is responsible for abnormal pregnancy outcome comparing those after ART and non‐ART in infertile patients.

Methods

We compare the singleton pregnancy outcome of infertile patients in our university hospital between 2000 and 2008 following ART (351 pregnancies) and non‐ART (213 pregnancies) procedures. Pregnancy outcome parameters were incidence of pregnancy induced hypertension, placenta previa, placental abruption, cesarean delivery, preterm birth, very preterm birth, stillbirth, low birth weight and very low birth weight.

Results

Most of the pregnancy outcome parameters were not significantly different between the ART group and the non‐ART group. Only placenta previa was significantly higher in the ART group than in the non‐ART group (odds ratio 4.0; 95 % CI 1.2–13.7).

Conclusions

ART procedure may itself be a risk factor for the development of placenta previa. Some of the abnormal perinatal outcomes that had been previously attributed to ART, however, may be due to the baseline characteristics of infertile patients.

Introduction

ART has become a major treatment option in the treatment of infertility. With more than 4 million babies born worldwide every year, ART‐induced births are still increasing. Particularly in Japan, approximately 2.5 % of the births in 2009 were conceived by in vitro fertilization (IVF), intracytoplasmic sperm injection (ICSI), or frozen embryo replacement (FER) [1]. The increased chance of multiple pregnancies is one of the most serious problems of ART pregnancy. It causes not only a rise of prematurity and low birth weight, but also of pregnancy induced hypertension and placenta previa which may affect maternal mortality. To prevent adverse outcomes of multiple pregnancies, single embryo transfer has been strongly suggested by both the Japan Society of Obstetrics and Gynecology and the Japan Society of Reproductive Medicine. Yet, many studies indicate that compared with naturally conceived singleton pregnancy, singleton pregnancy after ART may have a higher risk of adverse pregnancy outcome: low birth weight, preterm birth, placenta previa, stillbirth, etc. [2, 3, 4, 5, 6].

An essential question remains: Is the worse pregnancy outcome after ART due to the use of the ART technique itself or to the specific characteristics of infertile patients? Wang et al. [7] have suggested that ART procedures such as in vitro manipulations or ovarian stimulatory drugs might cause worse pregnancy outcomes. On the other hand, Lambert reviewed several articles and suggested that it is not the ART procedure, but infertility itself that is at the origin of the increased risk of health problems in singleton conception [8]. However, there are not necessarily sufficient rationale data on this issue. Therefore, it is important to obtain further data comparing ART pregnancies with pregnancies in non‐ART infertility treatment rather than with naturally conceived pregnancies. Then, it is possible to evaluate the contribution of ART technique itself to the outcomes, eliminating the influence of infertile patients’ characteristics. The purpose of this study was to elucidate whether the ART procedure is responsible for abnormal pregnancy outcome comparing those after ART and non‐ART infertility cases at our university hospital.

Methods

ART and non‐ART patients

Infertility treatments, including ART and other conventional treatments, were performed in our infertility clinic. Some patients were followed up in our hospital until delivery. The others were seen in another hospital easily accessible to the patients. We obtained the data on delivery by addressing questionnaires to the hospital in which the patients had delivered. Data for 351 pregnancies ending in singleton birth (>22 weeks of gestation) were obtained after ART treatment between 2000 and 2008 at the infertility clinic of the University of Tokyo Hospital. Data for 213 pregnancies ending in singleton birth (>22 weeks of gestation) were obtained after non‐ART treatment (infertility medication without ART) between 2000 and 2008 at the infertility clinic of the University of Tokyo Hospital.

Pregnancy outcome

Pregnancy outcome measures were incidence of pregnancy induced hypertension, placenta previa, placental abruption, cesarean delivery, preterm birth, very preterm birth, stillbirth, low birth weight and very low birth weight. In this study placenta previa included four types of placenta previa: complete placenta previa, incomplete placenta previa, marginal placenta previa and low‐lying placenta as described previously [9, 10, 11].

Preterm birth was defined as birth before 37 weeks of gestation. Very preterm birth was defined as birth before 32 weeks of gestation. Low birth weight was defined as birth weight of less than 2,500 g. Very low birth weight was defined as birth weight of less than 1,500 g.

IVF/ICSI and FER

Pregnancies after both fresh and frozen transfers were included. There were 121 conventional IVF cases, 124 ICSI cases, 27 split cases and 79 FER cases. Our primary protocol of ovarian stimulation was the long gonadotropin‐releasing hormone (GnRH) agonist protocol as previously reported [12]. Briefly, pituitary down‐regulation was achieved by daily use of a GnRH agonist [nafarelin acetate (Nasanyl); Astellas, Tokyo, Japan] administered at the midluteal phase of the previous cycle. Ovulation induction was performed by daily injections of hMG (hMG‐Kowa; Kowa Pharmaceutical Co, Tokyo, Japan) starting on day 3 of the menstrual cycle. Some of the recent patients received recombinant FSH [13]. Oocyte maturation was triggered by injecting 10,000 IU hCG (Mochida Pharmaceutical Co., Osaka, Japan). Transvaginal oocyte retrieval was performed under sonographic guidance 34 h after hCG administration. Luteal support was given by administration of 1.44 mg transdermal estradiol patch (Estraderm M; Novartis, Basel, Switzerland) every other day and transvaginal suppository of 200 mg progesterone every day.

Non‐ART treatment

Infertile patients who conceived at the infertility clinic in Tokyo University without ART (non‐ART group) were included in the study. Our clinic provides not only ART treatment but also non‐ART treatment proactively [14]. There were 85 intrauterine insemination (IUI) cases and 128 timed intercourse cases. Ovarian induction was performed in 107 cases in total; there were 54 clomifene cases and 53 gonadotropin cases. One hundred six patients did not undergo ovarian induction.

Statistics

Differences in categorical outcomes between ART and non‐ART groups were assessed by Fisher's exact test, unpaired t test, and Mann–Whitney U test as appropriate. P < 0.05 was considered statistically significant. Statistical analyses were conducted using Excel‐Toukei 2008 (SSRI, Tokyo, Japan).

Results

The maternal age, parity and rate of previous cesarian section were shown in Table 1. There were no significant differences in age between the ART group (34.9 ± 3.5 years) and non‐ART groups (34.2 ± 4.3 years). There were no differences in parity or the rate of previous cesarean section, both of which may affect placental position, between the two groups either.

Table 1.

Characteristics of the patients

	ART	Non‐ART	P value
Age in years [mean (SD)]	34.9 (3.5)	34.2 (4.3)	0.066
Parity [mean (SD)]	0.16 (0.38)	0.18 (0.42)	0.862
Previous cesarean section (%)	2.6	5.2	0.157

ART assisted reproductive technology, SD standard deviation

Comparisons between outcome measures of pregnancy are shown in Table 2. Rates of low birth weight and very low birth weight in ART were 14.2 and 3.1 %, respectively. Those of non‐ART were 16.2 and 1.4 %, respectively; these rates were not significantly different between the two groups. The mean birth weight was 2,894.0 ± 559.2 g [(mean ± standard deviation (SD)] and 2,935.2 ± 519.3 g in the ART and non‐ART groups, respectively. There was no significant difference in birth weight between the two groups either. Rates of preterm and very preterm births in ART were 12.0 and 3.4 %, respectively. Those of non‐ART were 10.8 and 2.4 %, respectively, with no significant difference between the two groups. Birth weight, stillbirth rate, cesarean section rate, placental abruption rate and pregnancy induced hypertension rate were similar between the two groups.

Table 2.

Pregnancy outcomes in ART and non‐ART cases

	ART (%)	Non‐ART (%)	Odds ratio (95 % CI)	P value
Low birth weight	50/351 (14.2)	34/210 (16.2)	0.9 (0.5–1.4)	0.543
Very low birth weight	11/351 (3.1)	3/210 (1.4)	2.2 (0.6–8.1)	0.270
Birth weight [g; mean (SD)]	2,894.0 (559.2)	2,935.2 (519.3)	–	0.377
Preterm birth	42/349 (12.0)	23/212 (10.8)	1.1 (0.6–1.9)	0.786
Very preterm birth	12/349 (3.4)	5/212 (2.4)	1.5 (0.5–4.2)	0.614
Stillbirth	5/351 (1.4)	5/213 (2.3)	0.6 (0.2–2.1)	0.515
Cesarean section	115/347 (33.1)	73/210 (34.7)	0.9 (0.6–1.3)	0.712
Placenta previa	19/351 (5.4)	3/213 (1.4)	4.0 (1.2–13.7)	0.023*
Placental abruption	2/351 (0.6)	2/213 (0.5)	1.2 (0.1–13.4)	1.0
Pregnancy‐induced hypertension	19/351 (5.4)	12/213 (5.6)	1.0 (0.5–2.0)	1.0

ART assisted reproductive technology, CI confidence interval, SD standard deviation

*P < 0.05

The incidence of placenta previa has been reported to be increased by ART [3, 15, 16]. As shown in Table 2, the rate of placenta previa in the ART group was 5.4 %, which was significantly higher (by four‐fold) than in the non‐ART group (1.4 %).

The incidence of placenta previa has also been described to be increased by parity [15, 16]. Thus, we analyzed the rate of placenta previa in reference to parity of the patients. In the ART group, the rate of placenta previa in nullipara was 5.1 % (15/294) and that of paras was 7.0 % (4/57). This difference, however, was not statistically significant. In the non‐ART group, three out of 178 cases were placenta previa in nulliparous patients (1.7 %), yet there were no placenta previa cases in parous patients. Thus, we could not confirm the positive relationship between parity and the occurrence of placenta previa.

Discussion

This study revealed that the ART procedure might cause abnormal placental position of the embryos transferred compared with those fertilized and implanted naturally in vivo after infertility treatment. Indeed, the rate of placenta previa was increased in the ART group as compared with the non‐ART group. Several other pregnancy outcomes, such as low‐birth weight and preterm delivery which had been shown to be affected by ART compared with normal population, were also investigated [2, 3, 4, 5, 6]. However, there were no differences in those factors between the ART and non‐ART groups (Table 2).

Low birth weight babies have been a serious problem in maternal and child health in Japan. According to the Ministry of Health Labour and Welfare Japan, the rate of low birth weight in singleton babies was 7.4 % in 2000 and gradually increased to 8.3 % in 2009 when the present study was performed [17]. The overall percentage of low birth weight in our infertile patients was 15.0 %. As age, parity, plurality, and many other factors are different between the general Japanese population data and our data, we are unable to establish any comparisons. However these incidences were approximately double as many as the general Japanese population data. It is in accord with a previous report from Belgium [18]. In the present study, the rate of low birth weight was 14.2 and 16.2 % in the ART and non‐ART groups respectively. There was no significant difference in the rate of low birth weight between ART and non‐ART groups which suggests that ART itself may not cause low birth weight. Preterm birth babies have also been a serious problem in maternal and child health in Japan. The rate of preterm birth, including singleton and multiple births, was 5.4–5.8 % in 2000–2009 [19]. The overall percentage of preterm birth babies in our infertile patients was 11.6 %. The other studied parameters were not significantly different between the ART and non‐ART groups except for placenta previa.

To address the question of whether the worse pregnancy outcome after ART is due to the specific patient characteristics or to the use of the ART technique, it would be ideal to perform a randomized control study comparing ART pregnancies and naturally conceived pregnancies. However, such a study is not allowed for ethical reasons, but there are some studies comparing pregnancy outcomes between treated and untreated infertile women. Draper et al. [20] conducted a case–control study that compared with women without infertility, women with infertility treatment were at increased risk of perinatal death by three‐fold. This study suggests that infertility itself is a risk factor of adverse perinatal outcome. There is some information comparing different treatment groups of infertile women. Wang et al. compared the risk of preterm birth with IUI, ART, and spontaneous conceptions. They found 50 % increased risk for preterm birth in the IUI group and more than twice the increase in risk in the ART group, implying that both infertility itself and ART may be related to increased preterm birth [7]. Yet, that article focused on preterm birth only. De Sutter et al. [18] compared pregnancy outcomes, including preterm birth, birth weight, neonatal intensive care unit admission, and Apgar score between IVF and IUI in a matched case–control study and found no statistically significant differences. They said it indicated that the worse pregnancy outcome after IVF as compared with spontaneous conceptions is due to the specific patient characteristics rather than to the use of IVF itself. Lambert reviewed several articles comparing the outcome of singleton pregnancies with different infertility treatment and implied that health problems in singleton ART babies appear to originate from infertility itself [8].

In the present study, rate of placenta previa (including complete placenta previa, incomplete placenta previa, marginal placenta previa or low‐lying placenta [9, 10, 11]) was 5.4 and 1.4 % in the ART and non‐ART groups, respectively, which represent a significant 4‐fold increase in the ART group. The prevalence of placenta previa is estimated at approximately 0.28–2.0 % in the general population [9, 21]. It increases with the increment of cesarean section, which is related to the occurrence of placenta previa. Several studies demonstrated that the risk factors for placenta previa are not only previous cesarean section but also termination of pregnancy, uterine surgery, increased age, multiparity, multiple pregnancy, cocaine use, smoking and ART itself [15, 16]. These factors are related to placenta previa with endometrial atrophy, a defect caused by previous scarring or inflammation in the endometrium, resulting in abnormally low implantation of the placenta [22].

Meta‐analysis of perinatal outcome comparing ART pregnancies and spontaneous pregnancies has shown a three‐fold higher risk of placenta previa in ART pregnancies [3]. In the present study, we found that ART pregnancies have four‐fold higher risk of placenta previa than non‐ART pregnancies. These findings suggest that the ART procedure itself increases the risk of placenta previa. Romundstad et al. [23] reported that among mothers who had conceived both naturally and after assisted fertilization, the risk of placenta previa was three‐fold higher in the ART pregnancy compared with that in the naturally conceived pregnancy. However, the mechanism underlying this phenomenon is still to be elucidated. One possible explanation is that ART technology or the endocrinological environment under ART might cause placenta previa. Romundstad et al. assume that mechanically induced uterine contractions by transcervical catheter could lead to higher frequencies of implantation in the lower uterine segment. Farhi et al. [24] have demonstrated another possible explanation for abnormal placental position in ART. They suggested that high serum estradiol concentrations in ART cycles might increase the risk of abnormal placentation because estradiol affects endometrial growth and preparation for decidualization.

In the present study, most of the pregnancy outcome parameters, including low birth weight, were not significantly different between the ART and non‐ART groups. Only placenta previa was higher in the ART group than in the non‐ART group. Some of the abnormal perinatal outcomes that had been attributed to ART may be due to characteristics of infertile patients. Since the present study was a single‐institution study and number of cases was limited, a multicenter study with a larger number of cases is necessary to clarify the relationship between ART and perinatal outcomes such as abnormal placental position. Finally, it is important for the physicians in charge to inform their patients of the pregnancy risk of infertility and ART. The obstetrician responsible for those patients should also be aware of the pregnancy risk, and especially of placenta previa.