Obstetric and neonatal complications in pregnancies conceived after oocyte donation: a systematic review and meta-analysis

Transcription

1 DOI: / Systematic review Obstetric and neonatal complications in pregnancies conceived after oocyte donation: a systematic review and meta-analysis M Storgaard, a A Loft, b C Bergh, c UB Wennerholm, d VSöderström-Anttila, e LB Romundstad, f,g K Aittomaki, h N Oldereid, i J Forman, j A Pinborg k a Fertility Clinic, Copenhagen University Hospital, Hvidovre Hospital, Copenhagen, Denmark b Fertility Clinic, Section 4071, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark c Reproductive Medicine, Departments of Obstetrics and Gynaecology, Institute of Clinical Sciences, Sahlgrenska Academy, Gothenburg University, Sahlgrenska University Hospital, Gothenburg, Sweden d Departments of Obstetrics and Gynaecology, Institute of Clinical Sciences, Sahlgrenska Academy, Gothenburg University, Sahlgrenska University Hospital/East, Gothenburg, Sweden e Väestöliitto Fertility Clinic, Helsinki, Finland f Department of Obstetrics and Gynaecology, IVF Unit, St Olav s University Hospital, Trondheim, Norway g Department of Public Health, Norwegian University of Science and Technology, Trondheim, Norway h Department of Medical Genetics, Helsinki University Central Hospital (HUCH) and University of Helsinki, Helsinki, Finland i Section for Reproductive Medicine, Department of Gynecology, Oslo University Hospital, Oslo, Norway j Section of Biostatistics, Department of Public Health, University of Copenhagen, Copenhagen, Denmark k Fertility Clinic, Copenhagen University Hospital, Hvidovre, Denmark Correspondence: A Pinborg, Fertility Clinic, Copenhagen University Hospital, Hvidovre, Kettegård alle 30, DK-2650 Hvidovre, Denmark. anja.bisgaard.pinborg.01@regionh.dk Accepted 27 June Published Online 5 September This article includes Author Insights, a video abstract available at Background Approximately oocyte donation (OD) treatment cycles are now performed annually in Europe and the US. Objectives To ascertain whether the risk of adverse obstetric and perinatal/neonatal outcomes is higher in pregnancies conceived by OD than in pregnancies conceived by conventional in-vitro fertilisation (IVF)/intracytoplasmic sperm injection (ICSI) or spontaneously. Search Strategy A systematic search was performed in the PubMed, Cochrane and Embase databases from Primary outcomes were hypertensive disorders of pregnancy, preeclampsia (PE), gestational diabetes mellitus, postpartum haemorrhage, caesarean section, preterm birth, low birthweight and small for gestational age. Selection criteria Inclusion criteria were original studies including at least five OD pregnancies with a control group of pregnancies conceived by conventional IVF/ICSI or spontaneous conception, and case series with >500 cases reporting one or more of the selected complications. Studies not adjusting for plurality were excluded. Data collection and analysis Thirty-five studies met the inclusion criteria. A random-effects model was used for the meta-analyses. Main results For OD pregnancies versus conventional IVF/ICSI pregnancies the risk of PE was adjusted odds ratio (AOR) 2.11 (95% CI, ) in singleton and AOR 3.31 (95% CI, ) in multiple pregnancies. The risks of preterm birth and low birthweight in singletons were AOR 1.75 (95% CI, ) and 1.53 (95% CI, ), respectively. Conclusions OD conceptions are associated with adverse obstetric and neonatal outcomes. To avoid the additional increase in risk from multiplicity, single-embryo transfer should be the choice of option in OD cycles. Keywords Caesarean section, low birthweight, oocyte donation, pre-eclampsia, preterm birth. Tweetable abstract Oocyte donation pregnancies have increased risk of a range of obstetric and neonatal complications. Please cite this paper as: Storgaard M, Loft A, Bergh C, Wennerholm UB, S oderstr om-anttila V, Romundstad LB, Aittomaki K, Oldereid N, Forman J, Pinborg A. Obstetric and neonatal complications in pregnancies conceived after oocyte donation: a systematic review and meta-analysis. BJOG 2017;124: Introduction The first successful oocyte donation (OD) pregnancy was reported in Although initially used for overcoming infertility due to premature ovarian insufficiency such as occurs in Turner syndrome, the technique was soon used to overcome natural peri- or postmenopausal infertility as well, thus making motherhood possible for women in their ª 2016 Royal College of Obstetricians and Gynaecologists 561

2 Storgaard et al. forties, fifties and even sixties. 2 4 OD also assists women with functioning ovaries but with repeated in-vitro fertilisation (IVF) failure or inherited diseases to achieve pregnancy. 2,5,6 In Europe the number of reported OD treatment cycles now exceeds annually, almost half of these performed in Spain, which offers highly developed oocyte donation programmes and liberal legal regulation. 7 9 In the US around donor oocyte cycles were performed in 2013, accounting for 10.5% of all assisted reproductive technology (ART) cycles. 10 Soon after the first experiences with OD pregnancies, studies on obstetric outcomes revealed an increased incidence of pre-eclampsia (PE) and pregnancy-induced hypertension (PIH) in them Further, several studies have found a higher incidence of low birth weight (LBW), preterm birth (PTB) and caesarean section (CS) in OD than in autologous IVF pregnancies Hypertensive disorders of pregnancy (HDP), which include chronic hypertension, PIH, PE/eclampsia and the Hemolysis, Elevated Liver enzymes and Low Platelet count syndrome (HELLP), is one of the major causes of maternal morbidity and mortality worldwide and may lead to stillbirth, preterm or small-forgestational-age (SGA) babies In high-income countries 16% of maternal deaths are caused by HDP. 23 CS is also associated with poor maternal outcomes even when performed electively in low-risk pregnancies, whereas LBW and PTB are associated with increased morbidity and mortality of the infants In this systematic review and meta-analysis we compare obstetric complications and neonatal outcomes in OD pregnancies with those of conventional IVF/intracytoplasmic sperm injection (ICSI) pregnancies and pregnancies after spontaneous conception (SC). Meta-analysis was performed on the following eight outcomes: HDP, PE, gestational diabetes mellitus, postpartum haemorrhage, CS, PTB, LBW and SGA. Methods We searched the PubMed, Cochrane and Embase databases from 1982 to January The main outcome measures were HDP, PE, gestational diabetes mellitus, postpartum haemorrhage, CS, PTB, LBW and SGA. Meta-analyses were performed for these measures to calculate pooled adjusted risk estimates for OD versus IVF/ICSI and OD versus SC. Secondary outcome measures were placenta praevia, placental abruption, blood transfusion, very preterm birth, very low birth weight, birth defects, stillbirth and perinatal and neonatal death, which are presented in Tables S8, S9 and S10. Adjusted odds ratios (AORs) or adjusted relative risks are presented when available. Otherwise crude odds ratios (ORs)/relative risks (RRs) or absolute values with or without P-values are presented. Definitions HDP included diagnosis of PIH or PE; some studies also included chronic hypertension. PIH was defined as the onset after the 20th week of pregnancy of hypertension [blood pressure (BP) 140/90 mmhg] without proteinuria. PE was defined as the onset after the 20th week of pregnancy of hypertension (BP 140/90 mmhg with proteinuria 0.3 g/24 hours) and included the subdiagnoses eclampsia and the HELLP syndrome. Gestational diabetes mellitus was defined according to the American College of Obstetricians and Gynecologists criteria in two studies but in most other studies it was not defined. Postpartum haemorrhage was defined as >1000 ml blood loss after delivery, and PTB as delivery before 37 weeks gestation; three included studies defined PTB as delivery before 34 weeks. LBW was defined as a birthweight of <2500 g and SGA as birthweight below the 10th centile. If studies had other definitions than those mentioned here, it is stated in Table S1. By conventional IVF/ICSI we mean IVF or ICSI using autologous oocytes, and we will refer to this as just IVF. ART basically includes any method involving handling ova in vitro including more advanced methods such as ICSI and OD. Inclusion and exclusion of studies Original articles published in the English language reporting rates of HDP, PE, gestational diabetes mellitus, placenta praevia, placental abruption, postpartum haemorrhage, blood transfusion, PTB, very PTB, LBW, very LBW, SGA, birth defects, stillbirth, perinatal and neonatal death were included. We included controlled studies including a minimum of five OD pregnancies and where the control groups consisted of either IVF or SC pregnancies. We also included case series of >500 OD pregnancies. In case of double publications the later study was included. Case reports and studies published only as abstracts were excluded. Studies with pooled data for singletons and twins were excluded unless adjustments had been made. Systematic reviews and meta-analyses were excluded but used as background material. Reference lists of identified reviews and meta-analyses from 2005 and later were searched manually for additional references without yielding further studies. One librarian and all the co-authors participated in performing the literature searches and abstract screening, and any uncertainty or difference of opinion was solved by discussion and consensus. Appendix S1 contains the full search strategy. Quality appraisal of the evidence Methodological quality, in terms of risk of bias, was assessed by five reviewers (CB, UBW, AL, AP and MS) using the tools developed by the Swedish Agency for Health Technology Assessment and Assessment of Social 562 ª 2016 Royal College of Obstetricians and Gynaecologists

3 Complications in oocyte-donation pregnancies Services for original articles ( These tools grade articles as low-, moderate- and high-quality articles after evaluating the following variables: selection bias, treatment bias, assessment of outcome bias, loss of follow-up bias, reporting bias, directness and precision. Only the 21 cohort studies that were designed comparing an OD group with a control group were assessed for methodological quality (Table S3). The GRADE system was used for assessing the quality of evidence of the results of our meta-analyses. 26 This system evaluates design, study limitations, consistency, directness, precision, publication bias, magnitude of effect, relative effect and absolute effect for all studies combined and per outcome and results in quality levels divided into high, moderate, low or very low quality of evidence. Data synthesis and meta-analyses Meta-analyses were performed to compare the AORs for the risk of HDP, PE, gestational diabetes mellitus, postpartum haemorrhage, CS, PTB, LBW and SGA. Owing to the clinical heterogeneity from the different study cohorts a random-effects model was chosen for the analyses. Tests of heterogeneity were not performed, as they are not valid when based on only a few studies. 27 Meta-analyses were performed using the rmeta package in R statistical software version ,29 Results The search identified 1195 abstracts of which 35 were included in the systematic review (Figure S1). Of these, 22 were cohort studies, including four national cohort register studies, and 13 were annual reports from The American Society for Reproductive Medicine/Society for Assisted Reproductive Technology Registry or Morbidity and Mortality Weekly Reports. Included studies are described in Table S1. Excluded studies with the reasons for exclusion are presented in Table S2, and quality assessment of the cohort studies are presented in Table S3. Results for primary and secondary outcomes are presented in Tables S4 S7 and S8 S10. Hypertensive disorders of pregnancy and preeclampsia The incidence of HDP in singleton pregnancies was % in the OD group, % in the IVF group and % in the SC group (Table S4). Five studies reported AORs and the meta-analysis resulted in a pooled risk AOR of 2.30 (95% CI, ) for OD singletons versus IVF singletons (Figure 1A). Five studies reported incidences of HDP ranging from 23.8 to 62.5% in OD multiple pregnancies versus 7.0 to 25.0% in IVF multiple pregnancies. Our meta-analysis was based on the AORs from two studies, resulting in a pooled estimate AOR of 2.45 (95% CI, ) (Figure 1B). There is moderate-quality evidence that the risk of HDP is higher in OD singleton and multiple pregnancies than in IVF pregnancies (GRADE+++). The incidence of PE in singleton pregnancies was reported in 10 studies and ranged from 9.3 to 16.9% in OD pregnancies compared with % in IVF controls and % in SC controls. For singleton pregnancies the AORs for PE in OD versus IVF pregnancies are presented in six cohort studies (AOR ), and the meta-analysis showed a pooled estimate AOR of 2.11 (95% CI, ) (Figure 1C). Another meta-analysis was conducted of three studies comparing OD pregnancies with SC pregnancies, yielding a pooled estimate AOR of 2.94 (95% CI, ) (Figure 1D). Considering multiple pregnancies, PE was present in % in OD pregnancies, 7.4 to 13.0% in IVF pregnancies and 7.1% in SC pregnancies, as reported by four studies. Only two studies presented AORs of PE in OD multiple versus IVF multiple pregnancies, and our metaanalysis revealed a pooled estimate AOR of 3.31 (95% CI, ) (Figure 1E). There is moderate-quality evidence that the risk of PE is higher in singleton and multiple OD pregnancies than in IVF pregnancies and SC pregnancies (GRADE+++). Gestational diabetes mellitus For singleton pregnancies seven studies provided data on gestational diabetes mellitus (Table S5). The incidence ranged from 0 to 13.0% in OD pregnancies and 1.0 to 23.3% in IVF controls, and it was 1.1% in the large SC group studied by Nejdet et al. 19 None of the studies found a significant difference between the OD and IVF or SC groups. A meta-analysis of OD versus IVF singleton pregnancies found a pooled estimate AOR of 1.33 (95% CI, ) (Figure 1F). Four studies presented data on gestational diabetes in multiple pregnancies. The incidence ranged from 0 to 31% in OD pregnancies and from 1.7 to 8.0% in three studies with IVF controls and one study with both IVF and SC as controls. In two of these studies the risk of gestational diabetes was significantly higher in OD pregnancies than in IVF or IVF + SC pregnancies. There is low-quality evidence that the risk of gestational diabetes is not higher in singleton OD than in singleton IVF pregnancies (GRADE++). Caesarean section CS rates were presented in 12 studies (Table S5). CS was carried out in % of OD singleton pregnancies versus % of IVF pregnancies and % of SC ª 2016 Royal College of Obstetricians and Gynaecologists 563

4 Storgaard et al. pregnancies. Only the study by Malchau et al. 17 presented AORs for emergency CS and planned CS separately, and found the rates similar. In the pooled AORs we included only the AOR for emergency CS from this study. The meta-analyses showed pooled estimate AORs of 2.20 (95% CI, ) for OD versus IVF singleton pregnancies (Figure 1G) and 2.38 (95% CI, ) for OD versus SC singleton pregnancies (Figure 1H). Considering multiple pregnancies, the incidence of CS ranged from % in OD pregnancies and % in IVF pregnancies. In the large SC group described by Malchau et al. 17 the CS rate was 54.4%. There is moderate-quality evidence that the risk of CS is higher in singleton OD than in singleton IVF and SC pregnancies (GRADE+++). Postpartum haemorrhage The incidence of postpartum haemorrhage in singleton pregnancies was reported by three studies (Table S6), with a range of % after OD and 0 9.4% after IVF. In the large SC group in the study by Nejdet et al., 19 5% suffered from postpartum haemorrhage. Our meta-analysis generated a pooled estimate AOR of 2.40 (95% CI, ) for OD versus IVF singleton pregnancies (Figure 1I). For multiple pregnancies, van Dorp et al. 30 reported an incidence of postpartum haemorrhage of 37.5% after OD and 10.7% after IVF; AOR, 4.91 (95% CI, ). There is low-quality evidence that the risk of postpartum haemorrhage is higher in OD singleton than in IVF singleton pregnancies (GRADE++). Preterm birth US ART surveillances during the years reported the incidence of PTB in both OD and IVF singleton pregnancies with fresh embryos, and found that PTB occurred in % of OD pregnancies and % of IVF pregnancies, with P < 0.01 for all four years (Table S6). PTB (<37 weeks) in singleton pregnancies was reported by 11 other studies, with a range of % in OD, % in IVF and % in SC pregnancies. A metaanalysis including four studies for PTB in OD singleton versus IVF singleton pregnancies resulted in a pooled estimate AOR of 1.75 (95% CI, ) (Figure 1J), whereas the meta-analysis comparing OD singleton with SC singleton pregnancies yielded a pooled estimate AOR of 2.30 (95% CI, ) (Figure 1K). The incidence of PTB in multiple pregnancies was reported in eight studies. PTB occurred in % of OD pregnancies, % of IVF and in 37.7% of SC pregnancies. The risk of PTB was found to be significantly higher in OD pregnancies than in IVF controls by four of the studies, while three others found no difference. Only the study by Malchau et al. 17 presented adjusted results, with an AOR of 1.5 (95% CI, ). There is moderate-quality evidence that the risk of PTB is higher in singleton OD pregnancies than in singleton IVF and SC pregnancies (GRADE+++). Low birth weight LBW in singletons was reported in US ART surveillances during the years in children born after OD and IVF with fresh embryo transfer (Table S7). The incidence of LBW ranged from 10.7 to 11.2% after OD and 9.3 to 9.5% after IVF, and the difference was statistically significant every year. Incidences of LBW in singletons were reported by seven other studies; % after OD, % after IVF and % after SC. AORs for LBW in OD versus IVF singleton children were presented by three studies, and the meta-analysis resulted in a pooled estimate AOR of 1.53 (95% CI, ) (Figure 1L). The meta-analysis of LBW for OD singleton versus SC singleton children gave a pooled estimate AOR of 1.94 (95% CI, ) (Figure 1M). In six studies LBW in multiples was present in % of OD infants, in % of IVF infants and in 42.4% of SC infants. One study did not present a P-value, two smaller studies found no statistically significant Figure 1. Forest plots of the risk of obstetric and neonatal complications in oocyte donation groups compared with control groups. (A) Pooled estimate on the risk of HDP in OD versus IVF/ICSI singleton pregnancies. (B) Pooled estimate on the risk of HDP in OD versus IVF/ICSI multiple pregnancies. (C) Pooled estimate on the risk of PE in OD versus IVF/ICSI singleton pregnancies. (D) Pooled estimate on the risk of PE in OD versus SC singleton pregnancies. (E) Pooled estimate on the risk of PE in OD versus IVF/ICSI multiple pregnancies. (F) Pooled estimate on the risk of gestational diabetes mellitus in OD versus IVF/ICSI singleton pregnancies. (G) Pooled estimate on the risk of CS in OD versus IVF/ICSI singleton pregnancies. (H) Pooled estimate on the risk of CS in OD versus SC singleton pregnancies. (I) Pooled estimate on the risk of postpartum haemorrhage in OD versus IVF/ICSI singleton pregnancies. (J) Pooled estimate on the risk of PTB in OD versus IVF/ICSI singleton pregnancies. (K) Pooled estimate on the risk of PTB in OD versus SC singleton pregnancies. (L) Pooled estimate on the risk of LBW in OD versus IVF/ICSI singleton pregnancies. (M) Pooled estimate on the risk of LBW in OD versus SC singleton pregnancies. (N) Pooled estimate on the risk of SGA in OD versus IVF/ICSI singleton pregnancies. (O) Pooled estimate on the risk of SGA in OD versus SC singleton pregnancies. (P) Pooled estimate on the risk of SGA in OD versus IVF/ICSI multiple pregnancies. CI, confidence interval; CS, caesarean section; HDP, hypertensive disorders of pregnancy; IVF/ICSI, in-vitro fertilisation/intracytoplasmic sperm injection; LBW, low birth weight; OD, oocyte donation; OR, odds ratio; PE, pre-eclampsia; PTB, preterm birth; SC, spontaneous conception; SGA, small-for-gestational-age. 564 ª 2016 Royal College of Obstetricians and Gynaecologists

5 Complications in oocyte-donation pregnancies (A) Pooled estimate on the risk of HDP in OD versus IVF/ICSI singleton pregnancies. (B) Pooled estimate on the risk of HDP in OD versus IVF/ICSI multiple pregnancieṣ (C) Pooled estimate on the risk of PE in OD versus IVF/ICSI singleton pregnancies. (D) Pooled estimate on the risk of PE in OD versus SC singleton pregnancies. difference, but one small study found the risk of LBW to be significantly higher in OD than in IVF pregnancies. One large study found 5.3% more OD children than IVF children with LBW (P < 0.001), and another large study presented an AOR of LBW in OD versus IVF babies of 1.8 (95% CI, ). ª 2016 Royal College of Obstetricians and Gynaecologists 565

6 Storgaard et al. (E) Pooled estimate on the risk of PE in OD versus IVF/ICSI multiple pregnancies. (F) Pooled estimate on the risk of gestational diabetes mellitus in OD versus IVF/ICSI singleton pregnancies. (G) Pooled estimate on the risk of CS in OD versus IVF/ICSI singleton pregnancies. (H) Pooled estimate on the risk of CS in OD versus SC singleton pregnancies. Figure 1. (Continued). There is moderate-quality evidence that the risk of LBW is higher in singleton OD pregnancies than in both IVF and SC pregnancies (GRADE+++). Small for gestational age The rate of SGA in singleton pregnancies was reported by seven studies and ranged from 0 to 9.3% in OD, 3.3 to 566 ª 2016 Royal College of Obstetricians and Gynaecologists

7 Complications in oocyte-donation pregnancies (I) Pooled estimate on the risk of post partum haemorrhage in OD versus IVF/ICSI singleton pregnancies. (J) Pooled estimate on the risk of PTB in OD versus IVF/ICSI singleton pregnancies. (K) Pooled estimate on the risk of PTB in OD versus SC singleton pregnancies. (L) Pooled estimate on the risk of LBW in OD versus IVF/ICSI singleton pregnancies. Figure 1. (Continued). 13.5% in IVF and 2.3 to 3.2% in SC pregnancies (Table S7). No studies found a significant difference. For singleton pregnancies AORs were presented by four studies, with a pooled estimate AOR of 1.14 (95% CI, ) for OD versus IVF (Figure 1N). The meta-analysis comparing OD singleton with SC singleton pregnancies resulted in ª 2016 Royal College of Obstetricians and Gynaecologists 567

8 Storgaard et al. (M) Pooled estimate on the risk of LBW in OD versus SC singleton pregnancies. (N) Pooled estimate on the risk of SGA in OD versus IVF/ICSI singleton pregnancies. (O) Pooled estimate on the risk of SGA in OD versus SC singleton pregnancies. (P) Pooled estimate on the risk of SGA in OD versus IVF/ICSI multiple pregnancies. Figure 1. (Continued). a pooled estimate AOR of 1.29 (95% CI, ) (Figure 1O). Considering multiple pregnancies SGA was present in % of OD pregnancies, % of IVF pregnancies and 22.4% of SC pregnancies, as reported by five studies. A meta-analysis including two studies resulted in a pooled estimate AOR of 0.97 (95% CI, ) (Figure 1P). 568 ª 2016 Royal College of Obstetricians and Gynaecologists

9 Complications in oocyte-donation pregnancies There is moderate-quality evidence that the risk of SGA is not higher in singleton and multiple OD pregnancies than in IVF pregnancies and also in singletons when compared with SC pregnancies (GRADE+++). Discussion Main findings Our meta-analyses showed that the risk of HDP, PE, LBW, PTB and CS was higher in OD than in IVF singleton pregnancies. For OD versus SC singleton pregnancies the risk of PE, PTB, LBW and CS was increased, and for OD versus IVF multiple pregnancies the risk of HDP and PE was increased. The AORs ranged from 1.55 to 3.31, the highest comparative risk being that of PE in OD versus PE in IVF multiple pregnancies and the lowest comparative risk being that of LBW in OD versus LBW in IVF singleton pregnancies. The prevalence of SGA was similar in OD versus IVF and SC singleton infants and also in OD versus IVF multiple infants. Further meta-analyses showed that the risk of postpartum haemorrhage in singleton pregnancies was higher after OD than after IVF, while there was no statistically significant difference in the prevalence of gestational diabetes mellitus. Strengths and limitations of the study The strength of this systematic review is the large sample size, with more than OD, IVF/ICSI and SC pregnancies, which enabled us to assess the incidence of eight different maternal and neonatal outcomes in OD versus IVF and/or SC pregnancies. PRISMA guidelines for meta-analyses were strictly followed. Since the risk of adverse outcomes is significantly higher in multiple than in singleton pregnancies, studies comparing outcomes in OD with those in ART or SC pregnancies should adjust or stratify for multiples to avoid confounding Hence, we excluded studies not stratifying or adjusting for multiples. The 16 meta-analyses comprised 13 meta-analyses with studies that either adjusted for multiples or included only singletons and of three meta-analyses of studies that included only multiples. The limitations of the review are the quality and heterogeneity of the included studies. Of the 21 included cohort studies comparing an OD group to a control group only two were of high quality, 11 were of medium quality and eight were of low quality. To ensure reliable results only studies of high and medium quality were included in the meta-analyses. In 13 of the studies less than one hundred OD pregnancies were included. Moreover some outcomes were poorly defined, e.g. only three out of 11 studies included a strict definition of gestational diabetes. Furthermore the medical indications for OD treatment were not reported in several of the studies. Also, avoidance of age confounding was attempted by different means among studies, including adjustment, stratifying or matching by age. The meta-analyses are based on studies that included a mix of both fresh and frozen oocytes and fresh and frozen embryos. Only one study distinguished between fresh and frozen embryos and found that only the incidence of largefor-gestational-age babies was influenced by this. 19 The indications for OD are heterogeneous, including primary and secondary ovarian insufficiency, previously radio/chemotherapy, autoimmune diseases, inadequate ovarian response to stimulation, repeated IVF treatment failure, advanced maternal age and genetic diseases. Hence OD patients are very heterogeneous regarding age, inheritance and infertility history. Obviously patient characteristics and general health influence obstetric outcomes, therefore large cohort studies including specific data on the OD recipients are urgently required. Oocyte donors also constitute a heterogeneous group including both healthy young women donating eggs for altruistic or commercial reasons, and egg sharing, in which infertile women with different infertility diagnoses donate surplus oocytes from IVF treatment. This affects pregnancy rates, but it is not known whether it influences obstetric and neonatal outcomes. Interpretation Three meta-analyses have been published considering complications of OD pregnancies. Pecks et al. 34 conducted a case-series and meta-analyses of PIH and PE (i.e. hypertensive disorders of pregnancy) in OD pregnancies versus those achieved by ART, such as IVF or insemination, and SC pregnancies. They found the risk of HDP to be significantly higher in OD than in both ART [AOR, 2.57 (95% CI, )] and SC [AOR, 6.60 (95% CI, )] pregnancies, but two out of five included studies in the OD versus SC analyses did not adjust for plurality, 34 which may explain the high AOR when comparing OD with SC pregnancies. Masoudian et al. 35 conducted meta-analyses of PE and PIH in OD versus ART and SC pregnancies. For PE, the OR was 4.34 (95% CI, ) for OD versus SC, which is higher than our result, probably because two out of four of their included studies did not adjust for multiples. Considering OD versus ART, the OR for PE was 2.24 (95% CI, ) when including only singletons, which is close to our result. AORs were not presented. 35 Another recent meta-analysis from 2015 on neonatal outcomes in OD versus ART pregnancies found significantly increased risks of LBW [RR, 1.18 (95% CI, )] and PTB [RR, 1.26 (95% CI, )], which changed only slightly when restricted to singletons. 18 Our results, based mainly on Danish and Swedish national cohorts, yielded AORs of 1.53 (95% CI, ) for LBW and 1.75 (95% CI, ) for PTB in OD versus IVF singletons. ª 2016 Royal College of Obstetricians and Gynaecologists 569

10 Storgaard et al. Although direct comparisons are hindered, as there is a mix of RRs and ORs, our meta-analyses confirm previous findings showing that the risk of HDP and PE in both singleton and multiple pregnancies is two- to three-fold higher in OD than in IVF pregnancies, which is also the case when adjusting for maternal age, and that risks for LBW and PTB are also significantly higher. The higher rates of LBW may be at least partly due to the higher rates of CS, since elective CS is often performed before week 40. Our meta-analyses showed that CS is performed more than twice as often in OD as in IVF and SC pregnancies. This probably reflects a considerable proportion of planned CS, though the study of Malchau et al. 17 was the only one that reported CS rates for planned and emergency CS separately, and found the rates equally increased. Other studies reported neither reasons for CS nor proportions of planned versus emergency CS. The included studies adjusted for maternal age a known confounder, since the incidence of elective CS increases significantly with increasing maternal age are presented in Table S1. 36 A large national register cohort from Canada showed a three-fold increased risk [AOR, 3.1 (95% CI, )] in overall severe maternal morbidity including haemorrhage requiring hysterectomy, cardiac arrest, venous thromboembolism and major infection in almost planned low-risk CS versus more than 2.3 million planned low-risk vaginal deliveries. 24 This elevated risk of serious adverse maternal outcomes should be taken into account when planning the mode of delivery for pregnant women after OD. It has been hypothesised that the abnormal placental development in HDP pregnancies is caused by the genetically unknown fetus that induces immunological reactions in the oocyte recipients. 34,37,38 This is confirmed by the fact that studies on histological and immuno-histochemical abnormalities in human placentae have shown that signs of placental pathology such as villitis, chronic deciduitis, maternal-floor infarction and ischaemic changes are more likely to occur in OD than in IVF pregnancies. 39,40 In has been suggested that human leukocyte antigen (HLA) matching of oocyte donors and their recipients may hinder the immunologic mismatch between mother and fetus, thus lowering the risk of placental pathology and HDP. 40 This is supported by a recent finding of a significantly higher level of HLA matching between mother and child than would be expected by chance in successful, uncomplicated OD pregnancies with unrelated donors. 41 There is high-quality evidence from meta-analyses that the incidence of PE is reduced by at least 10% when administering mg aspirin daily after the first trimester of pregnancy to women at increased risk of developing PE It has been speculated that aspirin adjusts an imbalance in the thromboxane A 2/ /prostacyclin ratio that is associated with increased vasoreactivity, which is part of the pathology in pre-eclamptic pregnancies. In the UK National Institute for Health and Clinical Excellence (NICE) guidelines it is recommended that aspirin should be administered to women with one high-risk factor for PE, such as chronic hypertension or pregestational diabetes mellitus, or two or more moderate-risk factors for PE, such as first pregnancy or/and age > Meanwhile, oocyte donation is not listed among the risk factors in the NICE guideline. Since OD increases the risk for PE by about two to three times compared to SC and even more than some of the other risk factors mentioned in the NICE guideline, prophylactic low-dose aspirin treatment should be considered for OD pregnancies. Since SGA is associated with HDP, which is increased two- to three-fold in OD versus IVF pregnancies, rates of SGA might be expected to be higher as well. However this was not shown in our meta-analysis, and SGA was not included in any of the other published meta-analyses. Though their sample size was small, Lashley et al. 46 found no cases of SGA in nine OD pregnancies with PE, whereas almost 30% of the infants born to pre-eclamptic mothers who conceived with their own oocytes were SGA (P < ). They suggest that PE in OD pregnancies has a different pathophysiology to that in autologous pregnancies, as they found certain differences at a microscopic level between placentae from OD and IVF pregnancies. 46 Among the included studies those providing information on birthweight showed that mean birthweight corrected for gestational age is not significantly lower in OD singletons than in IVF singletons. 2,6,14,16,17,37,47,30 It is puzzling that hypertensive disorders including PE have less impact on birthweight in OD singleton pregnancies than in IVF singletons pregnancies, and this requires further research into the pathology and time of onset of hypertensive disorders in OD pregnancies. The use of freezing techniques is associated with higher birthweight, and this may have influenced results since it is not clear from the studies to what extent oocyte and embryo freezing was used in study and control groups. 19,48 Possibly the uterine environment also plays a role, especially among mothers of advanced maternal age who would have conceived spontaneously at a younger age and may provide more beneficial growth conditions for the fetus than mothers with different kinds of reproductive disease. Pregnancy beyond age forty is in general associated with a higher risk for adverse outcomes. 49,50 Apart from a higher risk of HDP when comparing OD and IVF pregnancies in women of advanced maternal age there was no significant difference in adverse outcomes such as placental and bleeding complications, gestational diabetes, admission to the intensive care unit, gestational age and birthweight Since only four of the included studies provided such comparisons these results should be viewed with caution. 570 ª 2016 Royal College of Obstetricians and Gynaecologists

11 Complications in oocyte-donation pregnancies Conclusions This review and meta-analysis show elevated risks of HDP, PE, LBW, PTB, CS and postpartum haemorrhage in OD versus IVF and SC pregnancies, while there was no difference in the incidence of SGA and gestational diabetes mellitus. The increased risks of adverse maternal and neonatal outcomes should be taken into account when offering parental counselling prior to fertility treatment with OD and when planning ante- and perinatal care. Since some of the risks in OD pregnancies such as HDP and postpartum bleeding are further increased in multiple pregnancies, single-embryo transfer should be recommended, and couples wishing for a multiple pregnancy should be properly informed about their risk profile prior to double-embryo transfer. Conflict of interest None declared. Completed disclosure of interests form available to view online as supporting information. Contributions to authorship MS, AP, AL, CB and UBW contributed to study planning, literature search, analyses, writing and critical review of the manuscript. VSA, LBR, KA, NO and JF contributed to study planning, abstract screening, analyses and critical review of the manuscript. Funding A grant was obtained from Hvidovre Hospital Research Fund to cover scholarship salary to Marianne Bach Storgaard. The Nordic Expert Group research work was unconditionally supported by Finox, Finland, Norway, Sweden and Denmark. Acknowledgements The authors wish to thank librarian Therese Svanberg for help with the literature search. Supporting Information Additional Supporting Information may be found in the online version of this article: Figure S1. Flow chart. Table S1. List of included studies. Table S2. Excluded articles. Table S3. Quality assessment of included studies. Table S4. HDP and PE in oocyte donation pregnancies and controls. Table S5. Gestational diabetes mellitus and Caesarean section in oocyte donation pregnancies and controls. Table S6. Post partum haemorrhage and preterm birth in oocyte donation pregnancies and controls. Table S7. LBW and SGA in oocyte donation pregnancies and controls. Table S8. Very preterm birth and very low birth weight in oocyte donation pregnancies and controls. Table S9. Placental abruption, placenta previa and blood transfusion in oocyte donation pregnancies and controls. Table S10. Stillbirth, perinatal or neonatal death and birth defects in oocyte donation pregnancies and controls. Appendix S1. Systematic search for evidence. Video S1. Author Insights.& References 1 Lutjen P, Trounson A, Leeton J, Findlay J, Wood C, Renou P. The establishment and maintenance of pregnancy using in vitro fertilization and embryo donation in a patient with primary ovarian failure. Nature 1984;307: Wiggins DA, Main E. Outcomes of pregnancies achieved by donor egg in vitro fertilization a comparison with standard in vitro fertilization pregnancies. Am J Obstet Gynecol 2005;192: Henne MB, Zhang M, Paroski S, Kelshikar B, Westphal LM. Comparison of obstetric outcomes in recipients of donor oocytes vs. women of advanced maternal age with autologous oocytes. J Reprod Med 2007;52: Laskov I, Michaan N, Cohen A, Tsafrir Z, Maslovitz S, Kupferminc M, et al. Outcome of twin pregnancy in women 45 years old: a retrospective cohort study. J Matern Fetal Neonatal Med 2013;26: Abdalla HI, Billett A, Kan AK, Baig S, Wren M, Korea L, et al. Obstetric outcome in 232 ovum donation pregnancies. Br J Obstet Gynaecol 1998;105: Stoop D, Baumgarten M, Haentjens P, Polyzos NP, De Vos M, Verheyen G, et al. Obstetric outcome in donor oocyte pregnancies: a matched-pair analysis. Reprod Biol Endocrinol 2012;10:42. 7 European IVF-Monitoring Consortium (EIM) European Society of Human Reproduction and Embryology (ESHRE), Kupka MS, D Hooghe T, Ferraretti AP, de Mouzon J, Erb K, Castilla JA, et al. Assisted reproductive technology in Europe, 2011: results generated from European registers by ESHRE. Hum Reprod 2016; 31: Bergmann S. Fertility tourism: circumventive routes that enable access to reproductive technologies and substances. Signs (Chic) 2011;36: Hudson N, Culley L, Blyth E, Norton W, Rapport F, Pacey A. Crossborder reproductive care: a review of the literature. Reprod Biomed Online 2011;22: Division of Reproductive Health NCfCDPaHP Assisted Reproductive Technology National Summary Report US Department of Health and Human Services. Ref Type: Report. 11 Serhal PF, Craft IL. Oocyte donation in 61 patients. Lancet 1989;1: Sauer MV, Paulson RJ, Lobo RA. Pregnancy after age 50: application of oocyte donation to women after natural menopause. Lancet 1993;341: Pados G, Camus M, Van Steirteghem A, Bonduelle M, Devroey P. The evolution and outcome of pregnancies from oocyte donation. Hum Reprod 1994;9: Soderstrom-Anttila V, Tiitinen A, Foudila T, Hovatta O. Obstetric and perinatal outcome after oocyte donation: comparison with in-vitro fertilization pregnancies. Hum Reprod 1998;13: Soderstrom-Anttila V. Pregnancy and child outcome after oocyte donation. Hum Reprod Update 2001;7: ª 2016 Royal College of Obstetricians and Gynaecologists 571

12 Storgaard et al. 16 Zegers-Hochschild F, Masoli D, Schwarze JE, Iaconelli A, Borges E, Pacheco IM. Reproductive performance in oocyte donors and their recipients: comparative analysis from implantation to birth and lactation. Fertil Steril 2010;93: Malchau SS, Loft A, Larsen EC, Aaris Henningsen AK, Rasmussen S, Andersen AN, et al. Perinatal outcomes in 375 children born after oocyte donation: a Danish national cohort study. Fertil Steril 2013;99: Adams DH, Clark RA, Davies MJ, de Lacey S. A meta-analysis of neonatal health outcomes from oocyte donation. J Dev Orig Health Dis 2015;1 16 [Epub ahead of print]. 19 Nejdet S, Bergh C, K allen K, Wennerholm U, Thurin-Kjellberg A. High risks of maternal and perinatal complications in singletons born after oocyte donation. Acta Obstet Gynecol Scand 2016;95: Sibai B, Dekker G, Kupferminc M. Pre-eclampsia. Lancet 2005;365: Backes CH, Markham K, Moorehead P, Cordero L, Nankervis CA, Giannone PJ. Maternal preeclampsia and neonatal outcomes. J Pregnancy 2011;2011: Pennington KA, Schlitt JM, Jackson DL, Schulz LC, Schust DJ. Preeclampsia: multiple approaches for a multifactorial disease. Dis Model Mech 2012;5: Steegers EA, von Dadelszen P, Duvekot JJ, Pijnenborg R. Preeclampsia. Lancet 2010;376: Liu S, Liston RM, Joseph KS, Heaman M, Sauve R, Kramer MS, Maternal Health Study Group of the Canadian Perinatal Surveillance System. Maternal mortality and severe morbidity associated with low-risk planned cesarean delivery versus planned vaginal delivery at term. CMAJ 2007;176: Harrison MS, Goldenberg RL. Global burden of prematurity. Semin Fetal Neonatal Med 2016;21: Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso- Coello P, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 2008;336: Borenstein M, Hedges L, Higgins J, Rothstein H. Introduction to Meta-Analysis. Chichester, UK: Wiley, Lumley T. rmeta:meta-analysis, University of Auckland, New Zealand Accessed 1 February R Foundation for Statistical Computing. R: A Language and Environment for Statistical Computing. Vienna, Austria.: R Foundation for Statistical Computing, Accessed 1 February van Dorp W, Rietveld AM, Laven JS, van den Heuvel-Eibrink MM, Hukkelhoven CW, Schipper I. Pregnancy outcome of nonanonymous oocyte donation: a case control study. Eur J Obstet Gynecol Reprod Biol 2014;182: Young BC, Wylie BJ. Effects of twin gestation on maternal morbidity. Semin Perinatol 2012;36: Skora D, Frankfurter D. Adverse perinatal events associated with ART. Semin Reprod Med 2012;30: Barrington KJ, Janvier A. The paediatric consequences of Assisted Reproductive Technologies, with special emphasis on multiple pregnancies. Acta Paediatr 2013;102: Pecks U, Maass N, Neulen J. Oocyte donation: a risk factor for pregnancy-induced hypertension: a meta-analysis and case series. Dtsch Arztebl Int 2011;108: Masoudian P, Nasr A, de Nanassy J, Fung-Kee-Fung K, Bainbridge SA, El Demellawy D. Oocyte donation pregnancies and the risk of preeclampsia or gestational hypertension: a systematic review and metaanalysis. Am J Obstet Gynecol 2016;214: Herstad L, Klungsoyr K, Skjaerven R, Tanbo T, Eidem I, Forsen L, et al. Maternal age and elective cesarean section in a low-risk population. Acta Obstet Gynecol Scand 2012;91: Levron Y, Dviri M, Segol I, Yerushalmi GM, Hourvitz A, Orvieto R, et al. The immunologic theory of preeclampsia revisited: a lesson from donor oocyte gestations. Am J Obstet Gynecol 2014;211: Younis JS, Laufer N. Oocyte donation is an independent risk factor for pregnancy complications: the implications for women of advanced age. J Womens Health (Larchmt) 2015;24: Gundogan F, Bianchi DW, Scherjon SA, Roberts DJ. Placental pathology in egg donor pregnancies. Fertil Steril 2010;93: van der Hoorn ML, Lashley EE, Bianchi DW, Claas FH, Schonkeren CM, Scherjon SA. Clinical and immunologic aspects of egg donation pregnancies: a systematic review. Hum Reprod Update 2010;16: Lashley LE, Haasnoot GW, Spruyt-Gerritse M, Claas FH. Selective advantage of HLA matching in successful uncomplicated oocyte donation pregnancies. J Reprod Immunol 2015;112: Henderson JT, Whitlock EP, O Conner E, Senger CA, Thompson JH, Rowland MG. Low-Dose Aspirin for the Prevention of Morbidity and Mortality From Preeclampsia: A Systematic Evidence Review for the U.S. Preventive Services Task Force. Rockville, MD: Agency for Healthcare Research and Quality (US), Report No.: EF-1. U.S. Preventive Services Task Force Evidence Syntheses, formerly Systematic Evidence Reviews. 43 Askie LM, Duley L, Henderson-Smart DJ, Stewart LA. Antiplatelet agents for prevention of pre-eclampsia: a meta-analysis of individual patient data. Lancet 2007;369: Visintin C, Mugglestone MA, Almerie MQ, Nherera LM, James D, Walkinshaw S. Management of hypertensive disorders during pregnancy: summary of NICE guidance. BMJ 2010;341:c Bujold E, Roberge S, Lacasse Y, Bureau M, Audibert F, Marcoux S, et al. Prevention of preeclampsia and intrauterine growth restriction with aspirin started in early pregnancy: a meta-analysis. Obstet Gynecol 2010;116(2 Pt 1): Lashley LE, Buurma A, Swings GM, Eikmans M, Anholts JD, Bakker JA, et al. Preeclampsia in autologous and oocyte donation pregnancy: is there a different pathophysiology? J Reprod Immunol 2015;109: Simeone S, Serena C, Rambaldi MP, Marchi L, Mello G, Mecacci F. Risk of preeclampsia and obstetric outcome in donor oocyte and autologous in vitro fertilization pregnancies. Minerva Ginecol 2016;68: Pinborg A, Henningsen AA, Loft A, Malchau SS, Forman J, Andersen AN. Large baby syndrome in singletons born after frozen embryo transfer (FET): is it due to maternal factors or the cryotechnique? Hum Reprod 2014;29: Jacobsson B, Ladfors L, Milsom I. Advanced maternal age and adverse perinatal outcome. Obstet Gynecol 2004;104: Carolan M. Maternal age 45 years and maternal and perinatal outcomes: a review of the evidence. Midwifery 2013;29: Jackson S, Hong C, Wang ET, Alexander C, Gregory KD, Pisarska MD. Pregnancy outcomes in very advanced maternal age pregnancies: the impact of assisted reproductive technology. Fertil Steril 2015;103: Keegan DA, Krey LC, Chang HC, Noyes N. Increased risk of pregnancy-induced hypertension in young recipients of donated oocytes. Fertil Steril 2007;87: Le Ray C, Scherier S, Anselem O, Marszalek A, Tsatsaris V, Cabrol D, et al. Association between oocyte donation and maternal and perinatal outcomes in women aged 43 years or older. Hum Reprod 2012;27: Shrim A, Levin I, Mallozzi A, Brown R, Salama K, Gamzu R, et al. Does very advanced maternal age, with or without egg donation, really increase obstetric risk in a large tertiary center? J Perinat Med 2010;38: ª 2016 Royal College of Obstetricians and Gynaecologists