Abstract
Objectives:
To evaluate the association of fresh and frozen embryo transfer with the development of ischemic placental disease (IPD), hypothesizing that differences in implantation environment affect placentation and thus pregnancy outcomes.
Design:
We performed a secondary analysis of a retrospective cohort study of deliveries linked to in vitro fertilization (IVF) cycles.
Setting:
Deliveries were performed over fifteen years at a single tertiary hospital. All IVF cycles were performed at a single infertility treatment center.
Patients:
We included all women who underwent an autologous IVF cycle and had a live born infant or an intrauterine fetal demise (IUFD). We excluded women less than 18 years of age.
Intervention:
We compared pregnancies resulting from frozen embryo transfer (frozen) cycles to those resulting from fresh embryo transfer (fresh) cycles.
Main Outcome Measures:
The primary outcome was a composite outcome of IPD or IUFD due to placental insufficiency. IPD included preeclampsia, placental abruption, and small for gestational age (SGA). We calculated risk ratios (RR) and 95% confidence intervals (CI).
Results:
Compared to fresh cycles, frozen cycles had a lower risk of IPD or IUFD from placental insufficiency (RR 0.75, 95% CI: 0.59-0.97). Frozen cycles also conferred a lower risk of SGA than fresh cycles (RR 0.58, 95% CI: 0.41-0.81). Risks of preeclampsia (RR 1.3, 95% CI: 0.84-1.9) and abruption (RR 1.2, 95% CI: 0.56-2.4) were similar.
Conclusion:
There was a lower risk of IPD among frozen cycles compared to fresh cycles. This association was largely driven by lower risk of SGA among frozen cycles.
Keywords: ischemic placental disease, frozen embryo transfer, in vitro fertilization, small for gestational age
Capsule:
Frozen embryo transfer cycles are associated with a lower risk of ischemic placental disease than fresh embryo transfer cycles, likely from a lower risk of small for gestational age.
Introduction
In vitro fertilization (IVF) has been associated with an increased risk of conditions of ischemic placental disease (IPD)—preeclampsia, placental abruption, and intrauterine growth restriction. The underlying pathophysiology has not yet been determined and could be due to the underlying infertility diagnosis (1), unmeasured confounding effects (2), or hormonal elevations associated with IVF stimulation that impact placental implantation (3). In particular, high levels of estradiol associated with controlled ovarian hyperstimulation have been reported to impede the normal development of spiral arteries, leading to placental insufficiency and alteration of angiogenesis (4,5). However, the literature has not universally supported the association of high estradiol with adverse pregnancy outcomes (6,7).
With the advent of vitrification for embryo cryopreservation and the identification of subgroups of patients that may benefit from a “freeze all” approach (8,9), the utilization and success of frozen embryo transfer has increased (10). Improvements in perinatal outcomes have been attributed to selection of the highest quality embryos for freezing and the more physiologic hormonal milieu that generally accompanies frozen embryo transfer cycles (11). While frozen embryo transfer generally has been associated with improved perinatal outcomes, including lower risk of small for gestational age (SGA) and low birthweight (12,13), it is paradoxically associated with a higher risk of preeclampsia (14–17), compared to fresh embryo transfer. In donor recipients, who undergo endometrial preparation but not controlled ovarian hyperstimulation, there is no difference in SGA between fresh and frozen cycles (18). This finding suggests that the disparate outcomes may be more related to the endometrial hormonal milieu rather than the effect of cryopreservation itself.
Given the apparent discrepancy that frozen embryo transfer is associated with a lower risk of SGA but a higher risk of preeclampsia, it remains unclear whether slow freezing or vitrifying an embryo changes the risk of IPD. Whether differences in embryo preparation or endometrial hormonal milieu lead to placental insufficiency and thus IPD and its associated disorders of preeclampsia, abruption, and SGA is of great interest in optimizing pregnancy outcomes from IVF pregnancies.
We evaluated the association of frozen embryo transfer with IPD, hypothesizing that differences in the characteristics of frozen cycles may affect placentation and thus pregnancy outcomes. Secondary objectives included identifying whether components of IPD were associated with the type of embryo transfer and whether estradiol levels during fresh embryo transfer cycles were associated with variations in the risk of IPD.
Materials and methods
We performed a secondary analysis of a retrospective cohort study of deliveries in order to identify the risk of IPD in pregnancies conceived with frozen embryos compared to fresh embryos. The cohort included all deliveries greater than or equal to 20 weeks of gestation from January 1, 2000 to June 1, 2015 at a tertiary hospital. These were linked with IVF cycles at a single institution and birth certificate data from the Massachusetts Department of Public Health. Both fresh embryo transfer and frozen embryo transfer cycles were included, with type of cycle identified as the one leading to the delivery. We excluded deliveries to women less than 18 years of age. The institutional review boards at our institution and the Massachusetts Department of Public Health approved this study.
Oocyte source was identified through records at the infertility treatment center. Donor oocyte cycles were excluded due to known increased risk of IPD observed in other studies (19–21). All other data, including age, race, gravidity, parity, the type of cycle (fresh or frozen embryo transfer), peak estradiol and infertility diagnosis, was abstracted through electronic medical records at the tertiary hospital or the infertility treatment center. We did not have data on medical comorbidities, except for pre-gestational diabetes and smoking status, which were ascertained from birth certificate data.
Women underwent standard ovarian stimulation protocols, monitoring, and oocyte retrieval. In general, the fresh embryo transfer took place 3 or 5 days after the oocyte retrieval. The number of embryos transferred reflected national guidelines, with some variation according to treating physician’s specification, embryo quality and relevant patient history. Cryopreservation was generally performed 3 or 5 days after oocyte retrieval and included only embryos that were deemed viable by morphologic criteria. During the study period, vitrification was introduced in August 2011, with all freezing performed at the blastocyst stage by July 2013. Cycles using frozen embryos typically were performed after priming the uterus with estrogen and used progesterone.The primary outcome was IPD, which included preeclampsia, placental abruption, and SGA, or IUFD due to placental insufficiency. Preeclampsia was defined as the presence of elevated blood pressure (>140/90) during the delivery admission and either symptoms of preeclampsia (headache, visual changes, severe abdominal pain), eclampsia, or abnormal laboratory values (proteinuria, alanine aminotransferase, aspartate aminotransferase >80 units per liter, or platelets <100,000) before delivery. Placental abruption was defined as evidence of abruption or blood clot during delivery; evidence of abruption on placental pathology; or a very strong clinical suspicion that required hospitalization and intervention. Preeclampsia, placental abruption, and IUFD due to placental insufficiency were identified with ICD-9 codes.
Preeclampsia, abruption and IUFD diagnoses were then verified by medical record review. Due to institutional changes regarding scanned paper records, data sufficient for validation of preeclampsia and placental abruption were available only after July 1, 2008; therefore, we reviewed 59.2% of preeclampsia diagnoses and 53.6% of placental abruption diagnoses. Among the reviewed records, we confirmed 89.3% of the potential preeclampsia cases and 89.7% of the potential abruption cases. Pregnancies with an ICD-9 code for preeclampsia or abruption prior to July 1, 2008 were considered to have the diagnosis. To determine if placental insufficiency was a possible cause of an IUFD, we reviewed autopsy, pathology, and clinician notes for documented evidence. SGA was defined as <10th percentile for gestational age and sex using U.S. growth curves (22). Secondary outcomes included each of the components of IPD separately, as well as birthweight <3rd percentile for gestational age and sex to isolate a group of infants who were more likely to be pathologically small (23).
Peak estradiol was obtained from the infertility center records. This was defined as the highest estradiol level achieved during the cycle. This level was typically obtained on the day of human chorionic gonadotropin or gonadotropin-releasing hormone analog trigger. High estradiol, defined as >90th percentile for the population, was 3515 (pg/mL), which is consistent with prior published literature (6,11).
We used generalized estimating equations with an independent correlation matrix, accounting for multiple deliveries for the same woman, to estimate risk ratios (RR) and 95% confidence intervals (CI). All models were adjusted for maternal age. Covariates were tested individually for inclusion in the model. The covariate that had the strongest effect on the RR was maintained in the model and this process was repeated until no covariate changed the RR by more than 10%. All covariates that were included in the individual components of IPD were also included in the composite outcome. The covariates that we tested included race/ethnicity, gravidity, parity, smoking prior to pregnancy, diabetes prior to pregnancy, and year of delivery. Observations with missing data were dropped from the regression analyses given that the amount of missingness was less than 1%. Models restricted to singleton gestations were adjusted for the same covariates as the models for the full cohort.
We conducted several sensitivity analyses in this study. Given change in cryopreservation technique during the study period, we evaluated the outcomes before (2006-2012) and after (2013-2015) this practice change. In addition, given that we were only able to verify diagnosis of preeclampsia and placental abruption after 2008, we conducted a quantitative bias analysis(24) to assess the effect of misclassification of IPD would have on our results. The misclassification is assigned using sensitivities and specificities, allowing us to incorporate both over and under ascertainment of the outcome (IPD). We assumed differential misclassification for IPD in the fresh and frozen groups since there were more fresh cycles done in the time period when we were unable to verify the diagnoses. In the frozen group, we assumed a sensitivity and specificity of 80-95% and in the fresh group, we assumed a sensitivity and specificity of 70-90%. We ran 10,000 simulations to determine the “corrected” crude estimate of the risk of IPD in the frozen vs. fresh groups.
Results
Within the entire cohort of 69,084 pregnancies, we identified 2132 pregnancies resulting from autologous IVF cycles; 271 (12.7%) from frozen embryo transfer and 1861 (87.3%) from fresh embryo transfer. The groups were similar with respect to age, race, and marital status (Table 1). Almost all of the women in both groups had private insurance. Women who underwent frozen cycles were less likely to be nulliparous and more likely to have infertility due to diminished ovarian reserve.
Table 1:
Frozen n = 271 |
Fresh n = 1861 |
|
---|---|---|
Maternal age at conception (yrs) | 36.3 (33.0-39.8) | 35.8 (32.8-38.8) |
Race | ||
Caucasian | 220 (81.2) | 1544 (83.0) |
African American | 12 (4.4) | 74 (4.0) |
Hispanic | 5 (1.8) | 35 (1.9) |
Asian | 19 (7.0) | 137 (7.4) |
Other | 15 (5.5) | 69 (3.7) |
Not reported/unknown | 0 (0.0) | 2 (0.1) |
Marital status | ||
Married or partnered | 255 (94.1) | 1755 (94.3) |
Single | 13 (4.8) | 87 (4.7) |
Divorced, widowed, separated | 3 (1.1) | 16 (0.9) |
Unknown/missing | 0 (0.0) | 3 (0.2) |
Highest level of education achieved | ||
Less than high school, high school or GED | 13 (4.8) | 149 (8.0) |
College, associate’s degree, certificate | 118 (43.5) | 864 (46.4) |
Graduate degree | 129 (47.6) | 784 (42.1) |
Unknown | 11 (4.1) | 64 (3.4) |
Gravidity | ||
1 | 158 (58.3) | 1285 (69.0) |
2 | 58 (21.4) | 292 (15.7) |
3+ | 55 (20.3) | 284 (15.3) |
Parity | ||
0 | 140 (51.7) | 1259 (67.7) |
1 | 109 (40.2) | 520 (27.9) |
2+ | 22 (8.1) | 82 (4.4) |
Pre-pregnancy diabetes | 7 (2.6) | 31 (1.7) |
Pre-pregnancy smoking | 0 (0.0) | 27 (1.5) |
Primary infertility diagnosis | ||
Endometriosis | 11 (4.1) | 77 (4.1) |
Ovarian dysfunction | 25 (9.2) | 176 (9.5) |
Male factor | 50 (18.5) | 404 (21.7) |
Tubal | 13 (4.8) | 146 (7.8) |
Uterine | 7 (2.6) | 42 (2.3) |
Diminished ovarian reserve | 47 (17.3) | 91 (4.9) |
Unexplained | 69 (25.5) | 672 (36.1) |
Other | 29 (10.7) | 122 (6.6) |
Missing | 20 (7.4) | 131 (7.0) |
In vitro fertilization cycle parameters | ||
Peak serum estradiol (pg/mL) | - | 1682.5 (1091.02-553.0) |
Number of embryos transferred | 2.0 (1.0-2.0) | 2.0 (2.0-3.0) |
Note: Data presented as median (interquartile range) or n (%)
Mode of delivery and gestational age at delivery did not differ between pregnancies conceived after frozen embryo transfer and those conceived after fresh embryo transfer cycle (Table 2). Frozen embryo transfer cycles were less likely to result in multifetal gestations and had a slightly lower incidence of preterm delivery than fresh embryo transfer cycles (Table 2).
Table 2:
Outcomes | Frozen cycles n=271 |
Fresh cycles n=1861 |
---|---|---|
Gestational age at delivery | 38.1 (36.6-39.3) | 38.0 (36.0-39.0) |
Preterm delivery (<37 weeks) | 72 (26.6) | 594 (31.9) |
Intrauterine fetal demise | 1 (0.4) | 5 (0.3) |
Mode of delivery | ||
Vaginal delivery | 98 (36.2) | 678 (36.4) |
Cesarean delivery | 164 (60.5) | 1081 (58.1) |
Vaginal and cesarean deliverya | 8 (3.0) | 101 (5.4) |
Dilation and evacuationb | 0 (0.0) | 0 (0.0) |
Missing | 1 (0.4) | 1 (0.05) |
Gestations | ||
Singleton | 212 (78.2) | 1251 (67.2) |
Multifetal | 59 (21.8) | 610 (32.8) |
Birthweight (grams) | ||
Singleton | 3397.5 (3005-3663.5) | 3265 (2910-3600) |
Multifetal | 2385 (1835-2862.5) | 2322.5 (1847.5-2710) |
Note: Data presented as median (interquartile range) or n (%).
Some women who had multifetal gestations delivered one infant vaginally and the second via cesarean
For intrauterine fetal demise only
Compared to fresh embryo transfer cycles, frozen embryo transfer cycles had a lower risk of IPD or IUFD from placental insufficiency (RR 0.75, 95% CI: 0.59-0.97, adjusted for age) (Table 3). Frozen embryo transfer cycles also conferred a lower risk of SGA than fresh embryo transfer cycles (RR 0.58, 95% CI: 0.41-0.81, adjusted for maternal age). The risks of preeclampsia (RR 1.3, 95% CI: 0.84-1.9, adjusted for age) and abruption (RR 1.2, 95% CI: 0.56-2.4, adjusted for age and year of delivery) were elevated, but did not achieve statistical significance. When using the 3rd percentile cutoff for SGA, the risk of IPD or IUFD from placental insufficiency was no longer significantly lower in frozen embryo transfer cycles relative to fresh embryo transfer cycles (Table 3), although the risk of SGA remained lower among frozen embryo transfer cycles (RR 0.36, 95% CI: 0.16-0.79, adjusted for age).
Table 3:
Full cohort | Singletons Only | |||
---|---|---|---|---|
Frozen transfers n=271 |
Fresh transfers n=1861 |
Frozen transfers n=212 |
Fresh transfers n=1251 |
|
Small for gestational age < 10th percentile | ||||
Ischemic placental disease or IUFD | 56 (20.7) | 512 (27.5) | 25 (11.8) | 196 (15.7) |
Crude RR (95% CI) | 0.75 (0.59-0.96) | 1.0 | 0.75 (0.50-1.1) | 1.0 |
Adjusted RR (95% CI)a | 0.72 (0.56-0.93) | 1.0 | 0.70 (0.46-1.1) | 1.0 |
Preeclampsia | 27 (10.0) | 153 (8.2) | 13 (6.1) | 54 (4.3) |
Crude RR (95% CI) | 1.2 (0.81-1.8) | 1.0 | 1.4 (0.76-2.6) | 1.0 |
Adjusted RR (95% CI)b | 1.3 (0.84-1.9) | 1.0 | 1.5 (0.78-2.7) | 1.0 |
Placental abruption | 8 (3.0) | 61 (3.3) | 4 (1.9) | 36 (2.9) |
Crude RR (95% CI) | 0.90 (0.44-1.8) | 1.0 | 0.66 (0.24-1.8) | 1.0 |
Adjusted RR (95% CI)a | 1.2 (0.56-2.4) | 1.0 | 0.79 (0.29-2.2) | 1.0 |
Small for gestational age | 31 (11.4) | 367 (19.7) | 11 (5.2) | 127 (10.2) |
Crude RR (95% CI) | 0.58 (0.41-0.82) | 1.0 | 0.51 (0.28-0.93) | 1.0 |
Adjusted RR (95% CI)b | 0.58 (0.41-0.81) | 1.0 | 0.50 (0.28-0.90) | 1.0 |
Small for gestational age < 3rd percentile | ||||
Ischemic placental disease or IUFD | 39 (14.4) | 303 (16.3) | 17 (8.0) | 115 (9.2) |
Crude RR (95% CI) | 0.88 (0.65-1.2) | 1.0 | 0.87 (0.53-1.4) | 1.0 |
Adjusted RR (95% CI)b | 0.89 (0.65-1.2) | 1.0 | 0.87 (0.52-1.5) | 1.0 |
Small for gestational age | 6 (2.2) | 112 (6.0) | 1 (0.5) | 30 (2.4) |
Crude RR (95% CI) | 0.37 (0.16-0.83) | 1.0 | 0.20 (0.03-1.4) | 1.0 |
Adjusted RR (95% CI)b | 0.36 (0.16-0.79) | 1.0 | 0.17 (0.02-1.3) | 1.0 |
IUFD: intrauterine fetal demise
Data presented as n (%) or risk ratio (RR) and 95% confidence interval (CI)
Adjusted for maternal age and year of delivery
Adjusted for maternal age
We evaluated the risk of IPD or IUFD from placental insufficiency among fresh cycles only, comparing high peak estradiol to peak estradiol ≤90th percentile. There was a higher risk of IPD or IUFD from placental insufficiency for women who had high peak estradiol compared to the rest of the population, although this difference did not achieve statistical significance (Table 4). While the risk of preeclampsia was lower among cycles with high peak estradiol and the risks of abruption and SGA were higher, none of the differences in the individual components of IPD reached statistical significance.
Table 4:
High estradiol n=156 |
Normal estradiol n=1406 |
|
---|---|---|
Small for gestational age < 10th percentile | ||
Ischemic placental disease or IUFD | 52 (33.3) | 367 (26.1) |
Crude RR (95% CI) | 1.3 (1.0-1.6) | 1.0 |
Adjusted RR (95% CI)a | 1.3 (0.97-1.6) | 1.0 |
Preeclampsia | 11 (7.1) | 108 (7.7) |
Crude RR (95% CI) | 0.92 (0.51-1.7) | 1.0 |
Adjusted RR (95% CI)a | 0.73 (0.38-1.4) | 1.0 |
Placental abruption | 7 (4.5) | 42 (3.0) |
Crude RR (95% CI) | 1.5 (0.69-3.2) | 1.0 |
Adjusted RR (95% CI)a | 2.0 (0.86-4.5) | 1.0 |
Small for gestational age | 36 (23.1) | 271 (19.3) |
Crude RR (95% CI) | 1.2 (0.88-1.6) | 1.0 |
Adjusted RR (95% CI)b | 1.2 (0.88-1.6) | 1.0 |
Small for gestational age < 3rd percentile | ||
Ischemic placental disease or IUFD | 31 (19.9) | 214 (15.2) |
Crude RR (95% CI) | 1.3 (0.93-1.8) | 1.0 |
Adjusted RR (95% CI)b | 1.3 (0.91-1.8) | 1.0 |
Small for gestational age | 14 (9.0) | 79 (5.6) |
Crude RR (95% CI) | 1.6 (0.93-2.7) | 1.0 |
Adjusted RR (95% CI)b | 1.7 (0.96-2.9) | 1.0 |
IUFD: intrauterine fetal demise
Data presented as n (%) or risk ratio (RR) and 95% confidence interval (CI)
Adjusted for maternal age and year of delivery
Adjusted for maternal age
Due to the higher incidence of single embryo transfer among frozen embryo transfer cycles and the known association of SGA and preeclampsia with multifetal gestation(25), we conducted an analysis restricted to singleton gestations (Table 3). Similar to the primary analysis, the risk of IPD or IUFD from placental insufficiency was lower among frozen compared to fresh embryo transfer cycles, but the difference was no longer statistically significant. SGA <10th percentile remained significantly less likely in deliveries resulting from frozen embryo transfer cycles (RR 0.50, 95% CI 0.28-0.90, adjusted for age).
To address the change in cryopreservation due to the introduction of vitrification, we evaluated the risk of IPD before and after 2012. From 2006-2012, the risk of IPD was lower among frozen embryo transfer cycles, compared to fresh embryo transfer cycles (RR 0.69, 95% CI 0.48-0.98, adjusted for maternal age). The risk of SGA, both <10th and <3rd percentile, was lower in the frozen embryo transfer cycles compared to the fresh cycles, but risks of preeclampsia and abruption did not differ (Supplementary Table 1). From 2013 to 2015, we found similar risk ratios, although the results were no longer statistically significant (Supplementary Table 1), likely due to smaller numbers in the latter time period.
The quantitative bias analysis showed that there was little error in our risk ratio of IPD when comparing the frozen and fresh embryo transfer cycles. The “corrected” crude risk ratio was 0.81 (simulation interval 0.25-2.81). The simulation interval is similar to the confidence interval except it takes into account both random error and systematic (misclassification) error.
Discussion
We found a lower risk of IPD or IUFD from placental insufficiency among frozen embryo transfer cycles compared to fresh embryo transfer cycles in this cohort study of deliveries at a single institution over a fifteen-year time period, with similar findings before and after adoption of vitrification. The lower risk of SGA in the frozen embryo transfer group likely drives this association, given that a similar decrease in risk was not observed for preeclampsia or abruption. Similar to prior reports (15,17), we found that preeclampsia was more common in frozen embryo transfer cycles, though this difference was not statistically significant. Finally, a high peak estradiol level was associated with a higher risk of IPD among fresh cycles, although this difference did not achieve statistical significance in our cohort.
Most studies evaluating maternal and perinatal outcomes by type of embryo transfer have examined components of IPD separately (12,15,17,18,26), rather than as a whole. A recent paper among ovulatory women demonstrated no difference in birth outcomes between those who underwent fresh versus frozen embryo transfer, including no difference in incidence of preeclampsia (27). Risks of abruption and SGA were not evaluated. Our study, which evaluated IPD overall, allowed for a more complete exploration of risk of placental disease by cycle type. We found that frozen embryo transfer cycles have a lower risk of IPD or IUFD from placental insufficiency compared to fresh embryo transfer cycles. The size of the effect was modest and the confidence interval approached 1, likely because this risk is predominantly driven by the lower risk of SGA. The lower risk of SGA in frozen embryo transfer cycles was preserved even among singleton gestations alone. This suggests that the lower risk of SGA with frozen embryo transfer was not due simply to the lower risk of multifetal gestations.
Restricting the analysis to singletons alone resulted in loss of a statistically significant difference of the primary outcome between fresh and frozen embryo transfers. While no longer statistically significant, the risk of IPD was still lower among frozen embryo transfer cycles. There are a couple of possible reasons for the attenuated result. First, the numbers were smaller in the singleton only analysis and thus our ability to detect a statistically significant difference between the exposure groups was less; the wider confidence interval but similar point estimates and similar directionality of relationship among all outcomes would argue for this reason. Second, it is possible, that the association between the type of cycle is driven by multifetal gestation. Multifetal gestation is likely at least on the causal pathway between IVF and the outcomes, however, and thus a singleton only analysis would be expected to be attenuated. Given the complex interplay between IVF, multifetal gestation, and IPD, ideally additional analyses to better evaluate multifetal gestation as a mediator would be done, but insufficient numbers in this analysis limited this approach.
Our finding that SGA was less likely in pregnancies conceived with frozen embryo transfer cycles than fresh embryo transfer cycles is consistent with most prior published reports (12,18,28). Controlled ovarian hyperstimulation may play a role in this, based on a study by Vidal et al (18). They found an increased risk of low birthweight with fresh cycles compared to frozen cycles, but only in non-donor cycles. Given donor cycles do not involve controlled ovarian hyperstimulation, the authors postulated that the association of fresh cycles with low birthweight likely was due to controlled ovarian hyperstimulation rather the crypreservation process itself. Other studies also have supported the idea that controlled ovarian hyperstimulation may impact placental implantation and thus incidence of IPD (11,26), although they did not specifically evaluate differences between fresh and frozen cycles.
In an attempt to account for the effect of controlled ovarian hyperstimulation, we evaluated the risk of IPD or IUFD from placental insufficiency relative to peak estradiol. This value was only available in fresh cycles. While there was an increased risk of IPD among fresh cycles with peak estradiol >90th percentile, the result was not statistically significant. Moreover, we did not find a difference in SGA between fresh cycles with peak estradiol levels >90th percentile and those with lower peak estradiol levels. Our dataset likely was underpowered to detect potentially meaningful differences in our population. Our findings reflect the discordance observed in the literature. Inat least one other study,peak estradiol levels were not associated with adverse outcomes, including hypertensive disease, stillbirth, and SGA (6) while in another they were (11). The effects of controlled ovarian hyperstimulation are complex and are likely not completely captured by the effects of peak estradiol alone. Assessment of progesterone (29), among other factors, may provide additional information.
The risk of preeclampsia was higher in frozen cycles than fresh cycles, although this difference was not statistically significant. The magnitude of the effect was lower than what was observed in other studies evaluating the association of preeclampsia with frozen embryo transfer cycle, however (16,17). This may be because we excluded donor oocytes, which are associated with an increased risk of preeclampsia (19–21), and also because we included a general IVF population, rather than a population at high risk for preeclampsia, such as polycystic ovarian syndrome (17).
Our findings suggest that frozen embryo transfer may be a safer approach than fresh transfer, although it remains unclear if this is due to the endometrial milieu, cryopreservation techniques or other features of the IVF cycle. Further investigation into the association of preeclampsia and cycle parameters, as well as the interaction with multifetal gestation, is needed to elucidate this relationship.
This study has some limitations. First, the study took place over a long period of time with multiple practice changes and with a higher proportion of fresh embryo transfer cycles in the earlier part of the study. While we could not account for all practice changes that might have influenced our results, particularly number of embryos transferred, we did perform a sensitivity analysis in the latter part of the study after vitrification had become standardized and also during an era of higher likelihood of single embryo transfer and our results were similar, albeit limited, due to sample size. Second, we did not have information on several potential confounding factors, such as body mass index, chronic hypertension, gestational diabetes, history of preeclampsia, or polycystic ovary syndrome (30). We also could not control for intracytoplasmic sperm injection due to lack of available data linking the frozen cycle with the corresponding retrieval cycle. Third, we were unable to directly verify the assigned ICD-9 codes for outcomes among deliveries prior to July 2008. Since there were proportionally more fresh cycles prior to 2008, compared to frozen embryo transfer cycles, it is possible that the diagnoses were incorrect more often for fresh embryo transfer cycles. To address this limitation, we performed a quantitative bias analysis, and we observed similar risk ratios for the primary outcome, suggesting that the likelihood of misclassication bias is low. Although the simulation interval was wider than the confidence interval for the observed risk ratio, this is expected due to the increased uncertainly imposed by the simulations. Furthermore, this study evaluated cycles from a single infertility center in Massachusetts, and thus the generalizability of the results may be called into question. The infertility treatment center uses evidence-based practices similar to most other large infertility treatment centers in the country, and the population who use IVF is similar to the population which accesses care in Massachusetts (31). Finally, our study was underpowered to detect potentially meaningful differences for some of the secondary outcomes. Based on the observed incidences of preeclampsia in frozen and fresh embryo transfer cycles, we would need more than 4000 cycles per group to have 80% power to detect a difference of this magnitude.
Conclusions
In conclusion, we found a lower risk of IPD or IUFD from placental insufficiency among frozen embryo transfer cycles compared to fresh embryo transfer cycles. This association was largely driven by a lower risk of SGA among frozen embryo transfer cycles, a finding which persisted even among singletons alone. More interdisciplinary research is needed to elucidate the effect of IVF cycle parameters on pregnancy outcomes, particularly given practice changes and improvements in quality of frozen embryo transfer cycles. Ultimately, this research may be helpful in determining guidelines for embryo transfer in high-risk groups.
Supplementary Material
Acknowledgements
AMM received funding from NIH T32 HD052458 - Boston University Reproductive, Perinatal and Pediatric Epidemiology Training Program
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Romundstad LB, Romundstad PR, Sunde A, von Düring V, Skjaerven R, Gunnell D, et al. Effects of technology or maternal factors on perinatal outcome after assisted fertilisation: a population-based cohort study. Lancet 2008;372(9640):737–43. [DOI] [PubMed] [Google Scholar]
- 2.Watanabe N, Fujiwara T, Suzuki T, Jwa SC, Taniguchi K, Yamanobe Y, et al. Is in vitro fertilization associated with preeclampsia? A propensity score matched study. BMC Pregnancy and Childbirth [Internet] 2014. [cited 2018 April 2];14(1). Available from: http://bmcpregnancychildbirth.biomedcentral.com/articles/10.1186/1471-2393-14-69 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Choux C, Carmignac V, Bruno C, Sagot P, Vaiman D, Fauque P. The placenta: phenotypic and epigenetic modifications induced by Assisted Reproductive Technologies throughout pregnancy. Clin Epigenetics 2015;7:87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Albrecht ED, Bonagura TW, Burleigh DW, Enders AC, Aberdeen GW, Pepe GJ. Suppression of extravillous trophoblast invasion of uterine spiral arteries by estrogen during early baboon pregnancy. Placenta 2006;27(4–5):483–90. [DOI] [PubMed] [Google Scholar]
- 5.Aberdeen GW, Bonagura TW, Harman CR, Pepe GJ, Albrecht ED. Suppression of trophoblast uterine spiral artery remodeling by estrogen during baboon pregnancy: impact on uterine and fetal blood flow dynamics. Am J Physiol Heart Circ Physiol 2012;302(10):H1936–1944. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Dunne C, Cho K, Shan A, Hutcheon J, Durland US, Seethram K, et al. Peak Serum Estradiol Level During Controlled Ovarian Stimulation Is not Associated with Lower Levels of Pregnancy-Associated Plasma Protein-A or Small for Gestational Age Infants: A Cohort Study. J Obstet Gynaecol Can 2017;39(10):870–9. [DOI] [PubMed] [Google Scholar]
- 7.Rallis A, Tremellen K. Controlled ovarian hyper-stimulation during IVF treatment does not increase the risk of preterm delivery compared to the transfer of frozen-thawed embryos in a natural cycle. Aust N Z J Obstet Gynaecol 2013;53(2):165–9. [DOI] [PubMed] [Google Scholar]
- 8.Roque M, Valle M, Guimarães F, Sampaio M, Geber S. Freeze-all policy: fresh vs. frozen-thawed embryo transfer. Fertil Steril 2015;103(5):1190–3. [DOI] [PubMed] [Google Scholar]
- 9.Magdi Y, El-Damen A, Fathi AM, Abdelaziz AM, Abd-Elfatah Youssef M, AbdAllah AA-E, et al. Revisiting the management of recurrent implantation failure through freeze-all policy. Fertil Steril 2017;108(1):72–7. [DOI] [PubMed] [Google Scholar]
- 10.Roy TK, Bradley CK, Bowman MC, McArthur SJ. Single-embryo transfer of vitrified-warmed blastocysts yields equivalent live-birth rates and improved neonatal outcomes compared with fresh transfers. Fertil Steril 2014;101(5): 1294–301. [DOI] [PubMed] [Google Scholar]
- 11.Imudia AN, Awonuga AO, Doyle JO, Kaimal AJ, Wright DL, Toth TL, et al. Peak serum estradiol level during controlled ovarian hyperstimulation is associated with increased risk of small for gestational age and preeclampsia in singleton pregnancies after in vitro fertilization. Fertility and Sterility 2012;97(6):1374–9. [DOI] [PubMed] [Google Scholar]
- 12.Pelkonen S, Koivunen R, Gissler M, Nuojua-Huttunen S, Suikkari A- M, Hydén-Granskog C, et al. Perinatal outcome of children born after frozen and fresh embryo transfer: the Finnish cohort study 1995-2006. Hum Reprod 2010;25(4):914–23. [DOI] [PubMed] [Google Scholar]
- 13.Maas K, Galkina E, Thornton K, Penzias AS, Sakkas D. No change in live birthweight of IVF singleton deliveries over an 18-year period despite significant clinical and laboratory changes. Hum Reprod 2016;31(9):1987–96. [DOI] [PubMed] [Google Scholar]
- 14.Klatsky PC, Delaney SS, Caughey AB, Tran ND, Schattman GL, Rosenwaks Z. The Role of Embryonic Origin in Preeclampsia: A Comparison of Autologous In Vitro Fertilization and Ovum Donor Pregnancies. Obstetrics & Gynecology 2010;116(6):1387–92. [DOI] [PubMed] [Google Scholar]
- 15.Opdahl S, Henningsen AA, Tiitinen A, Bergh C, Pinborg A, Romundstad PR, et al. Risk of hypertensive disorders in pregnancies following assisted reproductive technology: a cohort study from the CoNARTaS group. Hum Reprod 2015;30(7):1724–31. [DOI] [PubMed] [Google Scholar]
- 16.Barsky M, St Marie P, Rahil T, Markenson GR, Sites CK. Are perinatal outcomes affected by blastocyst vitrification and warming? Am J Obstet Gynecol 2016;215(5):603.e1–603.e5. [DOI] [PubMed] [Google Scholar]
- 17.Chen Z-J, Shi Y, Sun Y, Zhang B, Liang X, Cao Y, et al. Fresh versus Frozen Embryos for Infertility in the Polycystic Ovary Syndrome. N Engl J Med 2016;375(6):523–33. [DOI] [PubMed] [Google Scholar]
- 18.Vidal M, Vellvé K, González-Comadran M, Robles A, Prat M, Torné M, et al. Perinatal outcomes in children born after fresh or frozen embryo transfer: a Catalan cohort study based on 14,262 newborns. Fertil Steril 2017;107(4):940–7. [DOI] [PubMed] [Google Scholar]
- 19.Jeve YB, Potdar N, Opoku A, Khare M. Three-arm age-matched retrospective cohort study of obstetric outcomes of donor oocyte pregnancies. International Journal of Gynecology & Obstetrics 2016;133(2):156–8. [DOI] [PubMed] [Google Scholar]
- 20.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(1):9–14. [PubMed] [Google Scholar]
- 21.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. American Journal of Obstetrics and Gynecology 2014;211(4):383.e1–383.e5. [DOI] [PubMed] [Google Scholar]
- 22.Oken E, Kleinman KP, Rich-Edwards J, Gillman MW. A nearly continuous measure of birth weight for gestational age using a United States national reference. BMC Pediatr 2003;3:6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Ananth CV, Smulian JC, Vintzileos AM. Ischemic placental disease: maternal versus fetal clinical presentations by gestational age. J Matern Fetal Neonatal Med 2010;23(8):887–93. [DOI] [PubMed] [Google Scholar]
- 24.Lash TL, Fox MP, Fink AK. Applying Quantitative Bias Analysis to Epidemiologic Data. New York, NY: Springer Science & Business Media; 2010. [Google Scholar]
- 25.Committee on Practice Bulletins—Obstetrics, Society for Maternal-Fetal Medicine. Practice Bulletin No. 169: Multifetal Gestations: Twin, Triplet, and Higher-Order Multifetal Pregnancies. Obstet Gynecol 2016;128(4):e131–146. [DOI] [PubMed] [Google Scholar]
- 26.Kalra SK, Ratcliffe SJ, Coutifaris C, Molinaro T, Barnhart KT. Ovarian stimulation and low birth weight in newborns conceived through in vitro fertilization. Obstet Gynecol 2011;118(4):863–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Shi Y, Sun Y, Hao C, Zhang H, Wei D, Zhang Y, et al. Transfer of Fresh versus Frozen Embryos in Ovulatory Women. N Engl J Med 2018;378(2): 126–36. [DOI] [PubMed] [Google Scholar]
- 28.Maheshwari A, Pandey S, Shetty A, Hamilton M, Bhattacharya S. Obstetric and perinatal outcomes in singleton pregnancies resulting from the transfer of frozen thawed versus fresh embryos generated through in vitro fertilization treatment: a systematic review and meta-analysis. Fertil Steril 2012;98(2):368–377.e1-9. [DOI] [PubMed] [Google Scholar]
- 29.Ibrahim Y, Haviland MJ, Hacker MR, Penzias AS, Thornton KL, Sakkas D. Elevated progesterone and its impact on birth weight after fresh embryo transfers. J Assist Reprod Genet 2017;34(6):759–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Dayan N, Pilote L, Opatrny L, Basso O, Messerlian C, El-Messidi A, et al. Combined impact of high body mass index and in vitro fertilization on preeclampsia risk: A hospital-based cohort study: IVF, High BMI, and Preeclampsia. Obesity 2015;23(1):200–6. [DOI] [PubMed] [Google Scholar]
- 31.Stern JE, Liu C-L, Cabral HJ, Richards EG, Coddington CC, Missmer SA, et al. Factors associated with increased odds of cesarean delivery in ART pregnancies. Fertil Steril 2018;110(3):429–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.