The objective of our study was to assess obstetric outcomes in pregnancies complicated by OHSS in early course following IVF. We compared OHSS cases to controls following both fresh and FET IVF cycles, and demonstrated a lower birthweight with OHSS and uncomplicated fresh transfers as compared to FET, and a significantly higher risk of placental abruption following OHSS, as compared to both fresh and FET cycles. This increased risk was independent of gestational age and smoking.
Controlled ovarian hyperstimulation is associated with an increase in estradiol levels, correlated to the amount and size of developing follicles. While the recruitment of more follicles generally increases patients’ chances of conception, the often supraphysiological levels of estradiol may affect implantation and subsequent placentation following transfer. Past in vivo studies have studied the effect of assisted reproductive procedures on placental development, and found overgrowth of the placental junctional zone with superovulation [14]. Superovulation has also been associated with altered expression of significant genes involved in in endometrial remodeling during early implantation [15]. These pathogeneses may explain the adverse correlation between birthweight and peak estradiol serum levels repeatedly demonstrated in past studies [16,17,18,19,20,21,22]. Furthermore, high estradiol levels have been linked to additional placental related adverse outcomes, including hypertensive disorders of pregnancy, premature preterm rupture of membranes and placenta previa [21, 22]. Although OHSS represents an “endpoint” to the continuum of peak serum estradiol levels attained following stimulation, this is probably not the sole pathogenesis of adverse pregnancy outcomes associated with it. OHSS entails a hypercoagulable state, and is associated with a systemic inflammatory response, and altered cytokine milieu [23, 24], which may in turn negatively affect the embryo-endometrial interaction.
Several past studies have explored the course of pregnancies complicated by OHSS following IVF, but few have focused on obstetric outcomes in singleton pregnancies. Hu et al compared patients hospitalized for moderate to severe OHSS to controls matched by age, ovarian reserve markers and multiple gestations [7]. They noted earlier gestational age at delivery, a higher rate of gestational diabetes mellitus and more neonatal complications with OHSS. Notably, almost half of pregnancies were multiple gestations, with no sub analysis for singletons, and information was not provided for zygosity and parity, which may affect outcomes. Haas and colleagues compared severe OHSS pregnancies following IVF or ovulation induction with IVF controls, matched by age and infertility etiology [2]. They found that patients with singletons following severe OHSS delivered at an earlier gestational age and had smaller babies, but multiple gestations were not at increased risk following OHSS. Three additional past studies have noted a higher rate of low birthweight, hypertensive disorders of pregnancy, gestational diabetes and placental abruption following OHSS [4,5,6], while two other studies did not demonstrate any differences from matched controls [8, 9].
The major finding of our study, in line with previous reports [5], is an increased rate of placental abruption following severe OHSS, as confirmed by logistic regression analysis. We demonstrated an incidence of 6.8% in the OHSS group, a significantly higher rate than in uncomplicated IVF pregnancies and following FET, and significantly higher than the previously reported 1% for all deliveries at our institution [25]. This finding is of utmost importance since placental abruption is one of the gravest complications of pregnancy with both maternal and neonatal consequences. It should be further explored whether this group of patients can benefit from closer pregnancy surveillance and/or from therapeutic interventions that may help reduce the risk. Of interest is that we did not demonstrate a difference in additional placental related complications, mainly birthweight following OHSS and fresh transfers, although this may relate to the overall low number of severe OHSS cases in our cohort. We did confirm a lower birth weight with stimulated IVF cycles as compared to FET, as previously demonstrated [26,27,28].
Study limitations refer first and foremost to the size of the study groups. The number of OHSS cases at our institution was relatively high for this now rare complication, most probably in light of the high rate of IVF cycles in our country (due to complete state funding) and higher number of fresh transfers earlier in the cohort period when freezing techniques were poor. Yet, we chose to include only singleton live births following IVF in our analysis, which limited the number of cases analyzed. A larger number of cases would have been powered for smaller differences in outcomes, such as in birthweight. Thus, a lack of power calculation must be taken into account when interpreting the results, as sample size was based on maximal sample obtained. We also assumed that programmed FET cycles entail a low level of serum estradiol, and thus included both programmed and natural FETs as controls. A larger number of cases may have allowed for a more suitably powered sub analysis of natural and programed FET cycles, to better account for the different obstetric outcomes associated with the two in past studies [29]. Second, as not all IVF cycles were performed at our institution, certain variables of interest were missing for analysis, such as infertility etiology, peak estradiol level during stimulation and number of embryos transferred. In a prospective study design, all patient data from IVF cycles would be available for analysis, and a uniform protocol for the screening of patients for OHSS would be applied (such as according to number of oocytes retrieved, peak estradiol levels, etc).
Study strengths refer to its design, which included matched groups of controls of both fresh and FET IVF pregnancies. The study also uniquely focuses solely on singleton IVF pregnancies, and accounts for age and parity for controls, both significant influencers of obstetric outcomes.