Placental pathology in live births conceived with in vitro fertilization after fresh and frozen embryo transfer

Journal Pre-proof Placental pathology in live births conceived with in vitro fertilization after fresh and frozen embryo transfer C.R. Sacha, A.L. Harris, K. James, K. Basnet, T.S. Freret, J. Yeh, A. Kaimal, I. Souter, D.J. Roberts PII:

S0002-9378(19)31213-X

DOI:

https://doi.org/10.1016/j.ajog.2019.09.047

Reference:

YMOB 12913

To appear in:

American Journal of Obstetrics and Gynecology

Received Date: 28 June 2019 Revised Date:

20 September 2019

Accepted Date: 27 September 2019

Please cite this article as: Sacha C, Harris A, James K, Basnet K, Freret T, Yeh J, Kaimal A, Souter I, Roberts D, Placental pathology in live births conceived with in vitro fertilization after fresh and frozen embryo transfer, American Journal of Obstetrics and Gynecology (2019), doi: https://doi.org/10.1016/ j.ajog.2019.09.047. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. 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. © 2019 Elsevier Inc. All rights reserved.

1 1

Title: Placental pathology in live births conceived with in vitro fertilization after fresh and frozen

2

embryo transfer

3 4 5

Sacha CR1*, Harris AL1, James K2, Basnet K3, Freret TS4, Yeh J1, Kaimal A5, Souter I1, Roberts DJ3

6

1

7

Outcomes Research, Department of OB/GYN, Massachusetts General Hospital and Harvard Medical

8

School, Boston, MA; 3Department of Pathology, Massachusetts General Hospital and Harvard Medical

9

School, Boston, MA; 4Department of OB/GYN, Massachusetts General Hospital and Harvard Medical

Massachusetts General Hospital Fertility Center and Harvard Medical School, Boston, MA; 2Center for

10

School, Boston, MA; 5Division of Maternal Fetal Medicine, Massachusetts General Hospital and Harvard

11

Medical School, Boston, MA

12 13

The authors report no conflicts of interest.

14

Funding: none

15

*Corresponding author:

16 17 18 19 20 21 22 23

Caitlin R. Sacha, MD Massachusetts General Hospital Fertility Center Yawkey 10A 55 Fruit Street Boston, MA 02114 617-724-8868 [email protected]

24 25 26 27 28 29 30 31

Word count: 6169

2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

Condensation: Frozen embryo transfer cycles may be associated with greater anatomic and vascular placental pathology than fresh embryo transfer cycles. Short Title: Placental pathology from fresh versus frozen embryos AJOG at a Glance: Applies to Original Research and Systematic Review submissions. This section is limited to no more than 130 words, 1-3 short sentences or phrases in bullet form • Given the rise in frozen embryo transfers, this study was performed to explore the differences in placental pathology in pregnancies arising from fresh versus frozen embryo transfers. • Compared to fresh embryo transfers, pregnancies arising from programmed frozen embryo transfers demonstrated more anatomic and vascular placental pathology. • This detailed placental pathology analysis of fresh versus frozen embryo transfers lays a foundation for understanding the basis of differences in obstetric risks between these two groups.

3 1

Abstract

2

BACKGROUND: The availability and use of frozen embryos after ovarian hyperstimulation for assisted

3

reproduction has increased with improvement in vitrification techniques and the rise of pre-implantation

4

genetic testing. However, there is conflicting data regarding whether obstetric outcomes differ between

5

fresh and frozen embryo transfer cycles.

6

OBJECTIVE: To compare placental pathology from live births arising from fresh and frozen embryo

7

transfer cycles.

8

STUDY DESIGN: A cohort of 1140 live births with placental pathology arising from autologous in vitro

9

fertilization cycles with fresh or frozen programmed transfer performed at MGH Fertility Center between

10

2004 and 2017 was retrospectively reviewed. An experienced placental pathologist categorized the

11

reported placental pathology as anatomic, infectious, inflammatory, or vascular/thrombotic. Our primary

12

outcomes were differences in these placental pathologies between the two groups. Patient demographic,

13

cycle, and birth outcomes were compared with chi square tests, Student’s t-test, or nonparametric tests, as

14

appropriate. Multivariate logistic regression models were used to compare placental pathology between

15

the fresh and frozen transfer groups.

16

RESULTS: Of the 1140 cycles included in our analysis, 929 arose from fresh embryo transfers (81.3%)

17

and 211 arose from programmed frozen embryo transfers (18.5%). For both transfer types, the average

18

age of the women at time of treatment was 35 years; mean BMIs were within the normal range (23.6

19

kg/m2 for fresh transfers and 23.2 kg/m2 for frozen transfers, p = 0.26); and mean day 3 FSH was 7.1 and

20

7.0 IU/L (p = 0.44), respectively. Deliveries occurred on average at 37.5 and 38.0 weeks gestational age

21

(p = 0.037) in the fresh vs. frozen transfer group, with similar rates of obstetric complications. However,

22

frozen transfers were more likely to be associated with marginal cord insertion [aOR 1.87 (CI 1.21, 2.91);

23

p = 0.01], accessory lobe formation [aOR 2.96 (CI 1.12, 7.79); p = 0.03], subchorionic thrombi [aOR 3.72

24

(CI 1.8, 7.71); p < 0.001], and fetal vascular malperfusion (FVM) characteristics with cord anomalies

25

[aOR 2.34 (CI 1.22, 4.46); p = 0.01]. These trends persisted when we analyzed day 5 transfers alone, and

4 1

single frozen embryo transfers remained associated with increased rates of subchorionic thrombi

2

compared to single fresh embryo transfers.

3

CONCLUSIONS: Pregnancies arising from frozen embryo transfers demonstrated more anatomic and

4

vascular placental pathology than those from fresh transfers in our cohort of patients, despite similar

5

maternal outcomes. More research is needed to explore how these differences in pathology may influence

6

obstetric and perinatal outcomes.

7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Key words: assisted reproduction, IVF, placenta, pathology, embryo, frozen transfer, fresh transfer

5 1 2

Introduction Freezing blastocysts after oocyte retrieval in in vitro fertilization (IVF) cycles has increased

3

greatly in recent years as a valuable tool to allow for pre-implantation genetic testing of blastocysts prior

4

to transfer, to reduce the risk of ovarian hyperstimulation syndrome (OHSS) in women with elevated peak

5

serum estradiol levels1, as well as to preserve additional (“supernumerary”) blastocysts available after

6

fresh embryo transfer. Some studies have shown that frozen embryo transfers (FETs) may have the

7

additional benefits of higher pregnancy rates and reduced adverse obstetric and perinatal outcomes, such

8

as preeclampsia or small-for-gestational-age (SGA) infants, potentially due to a more favorable uterine

9

environment.2-8 However, other data demonstrate greater risks of maternal and neonatal morbidities

10

associated with FETs.5,9-11 While higher clinical pregnancy rates and a lower incidence of OHSS were

11

reported in women with polycystic ovarian syndrome (PCOS) who undergo a FET, higher rates of LGA

12

and preeclampsia have also been demonstrated.12,13 A multicenter RCT by Wei et al. also showed that

13

single FETs may be associated with a higher risk of preeclampsia compared to single embryo fresh

14

transfers.14 Findings by Ernstad et al. suggest that programmed frozen cycles, performed using

15

supplemental progesterone and estrogen, may be associated with adverse outcomes such as postpartum

16

hemorrhage, macrosomia, and hypertensive disorders; natural and stimulated frozen cycles, in which

17

patients develop a corpus luteum that produces native progesterone, had similar outcomes in these

18

respects to fresh transfers in their study.15 It is thus clear from existing data that our knowledge of how

19

underlying maternal characteristics such as infertility diagnosis, uterine environment, and hormonal

20

milieu in the setting of a fresh vs. frozen transfer impact obstetric outcomes remains incomplete.

21

Because adverse obstetric outcomes such as hypertensive disorders and intrauterine growth

22

restriction are linked to abnormal placentation, understanding placentation in fresh versus frozen cycles is

23

a crucial step to understanding pregnancy complications associated with the different assisted

24

reproductive technology procedures.16-18 In this study, we aimed to assess the pathological profile of

25

placentas arising from fresh and programmed FETs to establish a foundation for understanding

26

differences in obstetric outcomes between these transfer cycle types.

6 1 2

Materials and Methods

3

Study Design

4

We conducted a retrospective cohort study of all live births resulting from autologous IVF

5

pregnancies at our academic institution from 2004-2017. The study was approved by the Institutional

6

Review Board (#2015P002672) of the Massachusetts General Hospital. Clinical data were derived from

7

our hospital electronic medical record system, IVF, and pathology databases. Donor oocyte and

8

gestational carrier cycles were excluded. Natural and stimulated FET cycles were also excluded, as our

9

center performed small numbers of these types of cycles during the study time period. Our primary

10

outcomes were anatomic, inflammatory, infectious, or vascular/thrombotic placental pathology.

11 12 13

Ovarian stimulation and transfer protocols All patients underwent a fertility evaluation, which routinely included a cycle day 3 follicular

14

stimulating hormone (FSH) level, as well as anti-Müllerian hormone (AMH) assessment from 2014

15

onwards. Patients then underwent controlled ovarian hyperstimulation, as previously described19, by

16

luteal-phase gonadotropin-releasing hormone (GnRH) agonist, GnRH-antagonist downregulation or

17

GnRH agonist flare protocol, as clinically indicated. Follicle synchronization and/or priming was

18

achieved by oral contraceptive pills (OCPs) (30 µg ethinyl estradiol/0.15 desogestrel, Apri, Teva

19

Pharmaceuticals, North Wales, PA) or ethinyl estradiol patch (0.1mg/day estradiol transdermal system,

20

Vivelle-Dot, Novartis Pharmaceuticals Corporation, East Hanover, NJ). Leuprolide acetate (Sandoz Inc.,

21

Princeton, NJ) or GnRH antagonist (0.25 mg ganirelix acetate, Antagon, Organon, Roseland, NJ; or 0.25

22

mg cetrorelix acetate, Cetrotide, EMD-Serono, Rockland, MA) were utilized for pituitary down-

23

regulation per protocol. Recombinant gonadotropins (follitropin beta, Follistim, Merck, Kenilworth, NJ,

24

or follitropin alpha, Gonal-F, EMD-Serono; and menotropins, Menopur or Repronex, Ferring

25

Pharmaceuticals, Parsippany, NJ) were used for stimulation; patients were serially monitored with

26

transvaginal ultrasound and serum estradiol to assess follicular measurements and endometrial thickness.

7 1

Intramuscular human chorionic gonadotropin (hCG) (10,000 IU, Novarel, Ferring Pharmaceuticals or

2

10,000 IU, Pregnyl, Merck) was administered to induce final oocyte maturation and 35-37 hours later the

3

patients underwent a transvaginal ultrasound-guided oocyte retrieval.19 Peak estradiol was measured on

4

the day of hCG trigger during the fresh ovarian hyperstimulation cycle that produced the embryo for

5

transfer.

6

Following egg retrieval, oocytes underwent conventional insemination or intracytoplasmic sperm

7

injection (ICSI) as clinically indicated. For those undergoing ICSI, the cumulus cells were stripped from

8

the oocytes 2-3 hours after retrieval. Embryos meant for fresh transfer were cultured until post-retrieval

9

day two, three or five, with day 5 transfer increasing in frequency over time. The decision for number of

10

embryos transferred was made based on institutional and ASRM guidelines.20-23 Cryopreservation of

11

embryos at the 2 pro-nuclei (2PN) stage, day 3, or day 5 was accomplished with slow oocyte freeze or

12

vitrification (starting 1/1/2016) per protocol.

13

Frozen programmed cycles for transfer of previously cryopreserved embryos were performed

14

with leuprolide acetate for pituitary down-regulation, supplemental estrogen (ethinyl estradiol patch, 0.1-

15

0.4mg/day, Vivelle-Dot estradiol transdermal system, Novartis Pharmaceuticals Corporation), and

16

progesterone replacement until 10 weeks gestational age (50mg/day Progesterone intramuscular injection,

17

West-Ward Pharmaceutical Corp., Eatontown, NJ; 100mg vaginal progesterone, Endometrin, Ferring

18

Pharmaceuticals, Parsippany, NJ).

19 20

Placental Pathology Assessment

21

All placentas from deliveries at MGH are received at the MGH pathology department and triaged

22

either to undergo complete (gross and histologic) pathologic examination or storage for two weeks before

23

incineration based on provided clinical history. Criteria for full gross and histopathologic examination

24

include, but are not limited to, a provided clinical history of ART and/or age ≥40 years. Singleton term

25

placentas delivered to women between the ages of 16 and 39 without medical or surgical complications,

26

obstetric complications or medical comorbidities, and with a live born infant with Apgar scores of 7 or

8 1

greater and 1 and 5 minutes with no resuscitation needed do not undergo a full pathologic examination

2

unless neonatal or maternal complications become apparent postpartum. The pathology reports of the

3

study patients were reviewed by one or two experienced placental pathologists (DJR and KB), and the

4

pathologic findings, if present, were categorized as follows: anatomic, inflammatory, infectious,

5

vascular/thrombotic, or other, using accepted diagnostic criteria from the Amsterdam Placental

6

Workshop24 (see Table 1 for all pathology category explanations).

7 8

Obstetric Outcomes

9

Live birth was defined as the live born delivery of a neonate at ≥ 24 weeks of gestation. Birth

10

weight and gestational age were collected for all patients. Birth information (including maternal or fetal

11

indication for delivery, route of delivery, estimated blood loss, delivery complications, and neonatal

12

outcomes), was collected from the electronic medical record. The diagnosis of preeclampsia was made by

13

the patients’ primary or delivering obstetrician and documented in the electronic medical record.

14 15

Statistical Analysis

16

Patient demographic, cycle, and birth outcomes were compared with chi square tests, Student’s t-

17

test, or nonparametric tests, as appropriate. Multiple multivariate logistic regression models were used to

18

estimate odds ratios (OR) and 95% confidence intervals (CIs) of placental pathology findings, adjusting

19

for maternal age, race, BMI, uterine vs. non-uterine factor diagnosis, number of embryos transferred,

20

number of fetuses born, and gestational age at delivery. We also controlled for date of egg retrieval before

21

or after January 1, 2016, when our laboratory transitioned from a slow freeze technique to vitrification for

22

embryo cryopreservation. Multiple births from the same woman were accounted for using mixed effects

23

logistic regression. All analyses were conducted using the Stata/IC package (StataCorp, College Station,

24

TX), and two-sided significance levels less than 0.05 were considered to be statistically significant.

25 26

Results

9 1 2

Patient demographic and cycle characteristics Autologous IVF cycles from 2004 to 2017 were reviewed for live births. Of 7803 fresh and

3

frozen autologous cycles with a transfer, 3169 cycles led to a live birth (40.6%), of which 1140 met our

4

inclusion criteria with placental pathology available for review from delivery within the Partners

5

Healthcare system (36.0%). Of those live births with pathology, 929 arose from fresh embryo transfers

6

(81.3%) and 211 arose from FETs (18.5%).

7

Demographic and clinical characteristics per transfer leading to live birth are shown for fresh and

8

programmed frozen cycles in Table 2. Women undergoing fresh transfers and FETs were similar in age

9

and predominantly white. Mean BMIs were within the normal range, and ovarian reserve testing by mean

10

day 3 FSH was normal on average (Table 2). Serum AMH tended to be lower in the fresh transfer group,

11

likely due to the practice of freezing all embryos in cases of high ovarian response (3.6 vs. 4.9 in FET

12

group, p = 0.007, Table 2). Most women undergoing fresh transfer were nulliparous (70.6%), while less

13

than half of the FET group were nulliparous (47.9%, p < 0.001). Peak estradiol (E2) level at time of

14

trigger was higher in cycles leading to a FET compared to fresh transfer cycles (2683 vs. 2100 pg/ml,

15

p=0.007; Table 2), as anticipated. Female factor infertility was the most common infertility diagnosis in

16

both groups. Transfers of 1-2 embryos were most commonly performed. Rates of conventional IVF and

17

ICSI fertilization methods were similar between the groups (Table 2). Demographic characteristics per

18

woman at the time of the first oocyte retrieval are additionally shown in Supplemental Table 1.

19 20 21

Obstetric characteristics of pregnancies from fresh and frozen transfers The mean gestational age at delivery was at term in both the fresh transfer and FET groups (37.5

22

vs. 38.0 weeks, p = 0.037; Table 2). There were more preterm deliveries (26.4% vs. 19.9%, p = 0.095)

23

and multifetal gestations (40.3% vs. 23.2%, p < 0.001) in the fresh transfer group compared to the FET

24

group, likely due to preterm deliveries of higher order multiple gestations. Singleton birth weights were

25

similar (3190g vs. 3230g, p = 0.36), while twin birth weights were lower in the fresh transfer group

26

(2485g vs. 2700g, p = 0.036). While there was no difference in the rate of SGA singletons, LGA infants

10 1

were more common in the FET group (5.7% vs. 3.0%, p = 0.013). There were no differences in rates of

2

cesarean delivery (58.8% in FET group vs. 52.9% in fresh transfer group, p = 0.12), or medical

3

comorbidities such as preeclampsia and gestational diabetes, between the groups. Although there was a

4

higher rate of retained placentas in vaginal deliveries in the FET group (13.8% vs. 6.2%, p = 0.013), the

5

rates of postpartum hemorrhage across all deliveries were similar (16.6% vs. 14.4%, respectively, p =

6

0.42).

7 8 9

Placental pathology in all fresh and frozen transfers The rates of specific placental pathologies in all transfers are shown in Table 3. Of the 929 fresh

10

transfer placentas, 97% contained some placental pathology. Of the 211 frozen transfer

11

placentas, 97% also contained some placental pathology. For all transfers, placental weight was

12

similar between the two groups (37.1 percentile in the FET group vs. 34.8 percentile in the fresh transfer

13

group by gestational age using published weight standards, p=0.35).25 There was a higher rate of several

14

anatomic features in the FET group: marginal cord insertion (18.0% vs. 10.8%, respectively, p = 0.004),

15

membranous vessels (6.6% vs. 3.1%, respectively, p = 0.016), and accessory placental lobes (10.4% vs.

16

5.2%, respectively, p = 0.004). Infectious pathology was similar between the two groups, as was

17

inflammatory pathology, though there was a trend towards increased low-grade villitis of unknown

18

etiology (13.7% vs. 9.5%, p = 0.065) in the FET group.

19

There was notably a high overall rate of vascular pathologies in both groups (73.0 vs. 69.5% in

20

the frozen vs. fresh groups, respectively; p = 0.32). FETs were associated with increased intervillous

21

thrombi (24.2% vs. 16.7%, respectively, p = 0.011), subchorionic thrombi (8.1% vs. 2.0%, respectively, p

22

< 0.001), and septal thrombi (2.4% vs. 0.1%, respectively, p < 0.001) compared to fresh cycles. Low-

23

grade fetal vascular malperfusion (FVM) was higher in frozen programmed cycles but the difference did

24

not reach statistical significance. Interestingly, placentas with both cord anomalies and maternal vascular

25

malperfusion (MVM) or FVM were more likely in the FET group compared to the fresh transfer group

11 1

(with MVM: 15.2% vs. 9.8%, respectively, p = 0.02; with FVM: 8.5% vs. 3.7%, respectively, p = 0.002).

2

If placentas from nulliparous women and singleton deliveries were analyzed separately, significantly

3

increased anatomic and vascular pathology persisted in the FET group (Supplemental Table 2).

4

In our adjusted model, placentas arising from programmed frozen cycles had almost twice the

5

odds of marginal cord insertion [aOR 1.87 (CI 1.21, 2.91); p = 0.01; Table 4] and three times the odds of

6

accessory lobe(s) [aOR 2.96 (CI 1.12, 7.79); p = 0.03; Table 4]. Subchorionic thrombi remained more

7

likely in FETs, as well [aOR 3.72 (CI 1.8, 7.71); p < 0.001]. Given the association of cord anomalies with

8

FVM, we also assessed this combination, which was more likely in the FET group [aOR 2.34 (CI 1.22,

9

4.46); p = 0.01].26 The remaining placental pathologies were no longer significantly different between the

10

groups in the adjusted model.

11 12 13

Comparison of placental pathology in day 5 and single embryo fresh and frozen transfers Given the increased prevalence of both day 5 transfers and single embryo transfers over time in

14

the fresh and frozen cycle context, we performed subgroup analyses of these transfers. Amongst day 5

15

transfers and single embryo transfers, anatomic pathology including marginal cord insertion, membranous

16

vessels, and accessory placental lobe remained more common in the FET group (Table 3).

17

Circummarginate or circumvallate membrane insertion occurred less frequently in frozen blastocyst

18

transfers compared to day 5 fresh transfers (6.0% vs. 2.2%, p = 0.042), but this comparison lacked

19

significance when single embryo transfers were assessed alone, likely reflecting the small numbers of

20

placentas with altered membrane insertion.

21

While there were no differences in infectious pathology between fresh and frozen day 5 or single

22

embryo transfers, there was a trend towards increased inflammatory pathology in frozen compared to

23

fresh day 5 transfers (22.3 vs. 15.9%, respectively, p = 0.054), driven by the significantly increased rate

24

of low-grade villitis of unknown etiology in the frozen group (14.1% vs. 8.2%, p = 0.021; Table 3).

25

Frozen single embryo transfers had increased rates of nonspecific inflammation compared to fresh single

26

embryo transfers (7.9% vs. 2.9%, p = 0.48).

12 1

Higher rates of vascular pathology were again seen in both the day 5 and single embryo FET

2

groups, including more subchorionic thrombi (day 5: 8.7% vs. 1.7%, respectively, p < 0.001; single: 7.9%

3

vs. 1.9%, respectively, p = 0.011) and septal thrombi (2.7% vs. 0.2%, respectively, p = 0.003; single:

4

2.0% vs. 0%, respectively, p = 0.043). Day 5 FETs also had more intervillous thrombi (24.5% vs. 16.1%,

5

p = 0.013). Placentas with both cord anomalies and evidence of MVM (day 5: 16.8% vs. 10.1%,

6

respectively, p = 0.017; single: 18.8% vs. 11.7%, respectively, p = 0.089) or FVM (day 5: 9.8% vs. 4.1%,

7

respectively, p = 0.005; single: 10.9% vs. 4.4%, respectively, p = 0.03) were seen more commonly in the

8

FET groups.

9

After controlling for the remaining potential confounders, marginal cord insertion and accessory

10

lobe(s) remained more likely to occur in the day 5 FET group. Interestingly, membranous cord insertion

11

was less likely to occur in the day 5 FET group [aOR 0.3 (CI 0.1, 0.89); p = 0.03; Table 4]. There was

12

also a trend in day 5 FETs towards increased odds of MVM [aOR 1.41 (CI 0.96, 2.05); p = 0.08] and

13

high-grade FVM [aOR 1.62 (CI 0.97, 7.14); p = 0.06]. The odds of subchorionic thrombi [aOR 5.38 (CI

14

2.15, 13.49); p < 0.001) and cord anomalies associated with FVM [aOR 2.39 (CI 1.16, 4.92); p = 0.02]

15

were significantly greater in the day 5 FET group, similar to the all transfers group. On the other hand,

16

when single embryo transfers were assessed alone, adjusting for day of transfer, only vascular pathology

17

(subchorionic thrombi, aOR 10.0 (CI 1.46, 69.1)) remained significantly greater in the FET group

18

compared to fresh transfers. There were similar overall odds of inflammatory and infectious pathology

19

after fresh and frozen day 5 and single embryo transfers.

20 21

Comment

22

Principal findings

23

Placental development is reflective of both embryo and maternal uterine factors during

24

implantation.18 Given the increasing prevalence of ART, as well as rates of supernumerary embryo

25

freezing, “freeze-all” policies, and embryo banking,27-29 understanding the potential impact of embryo

26

transfer cycles on placentation is important for adequate patient counseling and decision making,

13 1

knowledge of obstetric risk factors, and appropriate surveillance during pregnancy. Our study

2

demonstrated that programmed FETs give rise to placentas with a higher rate of anatomic and vascular

3

pathology compared to fresh transfers. After adjustment for potential confounding variables, the principal

4

findings were a significantly higher risk of marginal cord insertion, accessory lobes, cord anomalies with

5

associated FVM, and subchorionic thrombi after programmed FETs compared to fresh embryo transfers.

6

When day 5 fresh transfers were compared to frozen blastocyst transfers, these findings persisted. When

7

single embryo transfers alone were analyzed, subchorionic thrombi remained more common after

8

programmed FETs than fresh transfers.

9 10 11

Clinical and research implications of anatomic and vascular placental pathology findings ART has been previously recognized as a risk factor for anomalous cord insertion, including

12

marginal and membranous insertion, compared to natural conceptions.30,31 In our study, we found a rate of

13

18.0% and 10.8% for marginal cord insertion after programmed frozen vs. fresh transfer, respectively;

14

and a rate of 8.1% and 7.4% for membranous cord insertion after programmed frozen vs. fresh transfer,

15

respectively. Compared to a rate of 6.3% for marginal insertion and 1.5% for membranous amongst

16

623,478 singletons, as described by Ebbing et al. in their population-based registry study, our rates are

17

much higher regardless of transfer type, consistent with prior reports.30 The pathophysiology of this link is

18

unclear but some hypothesize that it could be related to gamete and embryo manipulations.30 Cord

19

insertion anomalies have also been associated with an increased risk of adverse obstetric and perinatal

20

outcomes, such as neonatal ICU admission, growth restriction or low birth weight, fetal malformations,

21

preterm delivery, premature rupture of membranes, placental abruption, preeclampsia, and cesarean

22

delivery.31-34 Given the similar risk factors and adverse outcomes with both membranous and marginal

23

cord insertion, anomalous cord insertion could be considered along a spectrum resulting from abnormal

24

implantation and placental development.31 How much of this process is impacted by the surrounding

25

hormonal milieu is unknown. Finally, ART has not previously been described as a risk factor for

26

accessory lobe formation, although uterine pathology, common in the infertility population, has.35 Despite

14 1

adjusting for uterine factor as a diagnosis, accessory lobes remained more common in the frozen group,

2

suggesting that there may be another mechanism related to abnormal implantation involved in its

3

pathogenesis. These anatomic pathologies did not occur more frequently in frozen single embryo transfers

4

after adjusting for confounders, suggesting that transferring multiple embryos, regardless of number of

5

infants born, may be an independent risk factor for altered anatomic pathology.

6

Vascular pathologies, in particular subchorionic thrombi and FVM in placentas with coexisting

7

cord anomalies, also occurred more commonly in pregnancies conceived after FET compared to fresh

8

transfer after adjusting for potential confounders. Furthermore, there was a trend in day 5 transfers alone

9

towards a higher risk of MVM. Transferring a single embryo appears to reduce the vascular pathology

10

differences between fresh and frozen transfers, though subchorionic thrombi were still more common in

11

the latter after adjusting for confounders. The clinical implications of subchorionic hematomas, which are

12

common in both natural and assisted conceptions, remain poorly understood, and our finding of an

13

increased risk of subchorionic thrombi (residual subchorionic hematomas) after FETs conflicts with a

14

prior study showing that subchorionic hematomas are more common after fresh embryo transfers.36,37

15

FVM, associated with fetal morbidity such as growth restriction and stillbirth, is often a downstream

16

consequence of umbilical cord anomalies or obstruction, as we observed in our data.38,39 Importantly,

17

MVM is commonly linked to an increased risk of hypertensive disorders, even in low risk populations.40

18

We did not see increased preeclampsia in the FET group, in contrast to prior literature showing increases

19

in this outcome after FET.5,10-12,14 However, the small absolute differences in these placental findings

20

between the two groups suggests that a larger sample is needed to identify differences in clinical obstetric

21

outcomes as these disorders remain uncommon and are likely multifactorial.

22

The increased rates of anatomic and vascular pathology after FETs did not lead to lower infant

23

birth weights; in fact, there were more LGA infants in the FET group, consistent with prior

24

literature.13,15,41 Prior studies have found that birth order is unlikely to account for more LGA infants from

25

FET cycles.42 Although mean age, BMI and rates of gestational diabetes were similar between the fresh

26

and FET groups, our results may reflect a subgroup of younger, high-responding women with FETs who

15 1

are more likely to have an LGA infant due to other maternal factors and are less susceptible to the non-

2

physiologic hormonal milieu that predisposes other FET pregnancies to clinically significant anatomic

3

and vascular pathology.

4

Fresh transfers had a higher incidence of preterm deliveries (26.4% vs. 19.9%) and multifetal

5

gestations (40.3% vs. 23.2%) compared to the FET group, likely reflecting the shift in the field towards

6

more single FETs over time. Since singletons and twins had similar rates of preterm deliveries, preterm

7

delivery of higher order multiples likely explains the overall higher rate of preterm deliveries in the fresh

8

transfer group. Rates of cesarean delivery were similar between the fresh and FET groups in our

9

population. However, significantly more vaginal deliveries in the frozen group were complicated by

10

retained placenta (13.8% vs. 6.2%), though rates of postpartum hemorrhage were similar between the

11

groups. Interestingly, prior literature has suggested a possible higher risk of hemorrhage and placenta

12

accreta after FET.5,9,10,15 A potential link between anatomic or vascular pathology and development of an

13

abnormally adherent placenta has not yet been elucidated, though it is possible that tissue hypoxia

14

generated in the setting of this pathology could alter placental development.

15

Gamete and embryo manipulations alone do not explain the increased odds of anatomic and

16

vascular pathology after FETs, as both fresh and frozen transfers involve oocyte retrieval, sperm

17

preparation and embryo development in culture. Therefore, the impacts of embryo freezing or the

18

hormonal environment of the uterus are two other hypotheses to consider. There is some evidence that the

19

freezing and thawing process itself may induce epigenetic that could impact placental development.43

20

Alternatively, Senapati et al., and others, have found that the super-physiologic hormone levels achieved

21

during ovarian stimulation are associated with alterations in endometrial gene expression that may affect

22

remodeling and angiogenesis, perhaps leading to abnormal implantation and placentation.44,45 However,

23

these findings do not explain the increased placental pathology that we see in FETs in the current study;

24

while women undergoing programmed FET take supplemental estrogen, their serum estradiol levels do

25

not reach those of women undergoing ovarian hyperstimulation.46

16 1

Another key consideration is that while patients undergoing fresh transfer develop corpora lutea ,

2

those undergoing programmed FET rely entirely on exogenous intramuscular progesterone. There is

3

increasing evidence that absent or excessive corpora lutea lead to altered levels of progesterone, relaxin,

4

and other growth factors essential for maternal vascular health in early pregnancy, which could explain

5

the increased risk of preeclampsia associated with IVF pregnancies and particularly programmed frozen

6

transfers.15,47-49 Since the corpus luteum is the primary source of steroid hormones supporting early

7

pregnancy as the placenta develops, an altered hormonal milieu impacting uterine vascularization during

8

this time may influence decidualization and placental development, leading to increased vascular

9

pathology.50,51 More research is needed both on a molecular level, evaluating the effect of progesterone

10

supplementation vs. native progesterone on the endometrium and the impact of freezing on the

11

trophoblast cells, and on a clinical level, investigating placental pathology in programmed FETs

12

compared to natural or stimulated FETs.

13 14 15

Strengths and limitations Our study has several strengths. First, IVF was performed at a single large academic fertility

16

center, reducing variation in IVF practice among the patients included in the study. Baseline

17

characteristics such as age, race, BMI, and markers of ovarian reserve were similar between the study

18

groups. There was a relatively low rate of uterine factor, and we were able to control for this variable in

19

our adjusted models. In addition, the majority of patients (> 90%) delivered at our institution, where all

20

patients with a provided history of IVF undergo full placental evaluation, thus reducing bias due to

21

pathological examination solely in the cases of delivery complications. However, a bias towards full

22

examination of complicated pregnancy cannot be completely excluded. Finally, all placental reports and

23

the majority of placental pathology slides were read and reviewed by a placental pathology expert (DJR),

24

minimizing variation and making this potential source of bias consistent across study groups.

25 26

There are also several limitations to this study, notably its retrospective design and inherent variability in pathologic examination. Importantly, our analysis lacks a natural conception control group,

17 1

which we were not able to collect retrospectively in an unbiased manner due to our institution’s placental

2

examination triage criteria. Such a control group should be utilized in future studies assessing placental

3

outcomes from fresh, natural/stimulated frozen, and programmed frozen embryo transfers. Limited power

4

in many of the detailed placental pathology subcategories also restricted our ability to control for

5

confounding variables in our clinical subgroups. Although significantly more women were parous in the

6

FET group, likely reflecting our practice of transferring supernumerary frozen embryos after a prior

7

pregnancy, we do not feel that this skewed our results towards abnormal pathology in the frozen group as

8

prior live birth suggests a favorable uterine environment for implantation. Finally, our findings are not

9

generalizable to all FETs, as natural and stimulated FETs were performed at a low volume at our center

10

during the study period and were not included in the current analysis.

11 12 13

Conclusions In summary, our findings represent a detailed assessment of the pathological findings in placentas

14

arising from live births after fresh and programmed frozen embryo transfers, which contributes to the

15

ongoing discussion of which transfer type is optimal for each patient circumstance. We demonstrated that

16

anatomic and vascular placental pathologies are seen more frequently in pregnancies arising from

17

programmed FETs compared with fresh embryo transfers, including amongst day 5 transfers, though

18

transferring a single blastocyst may reduce abnormal placental development. Therefore, there is placental

19

pathology associated with programmed FETs, but the impact of the specific features on obstetric and

20

perinatal outcomes demands further investigation.

21 22 23 24 25 26

Acknowledgments The authors acknowledge the MGH IVF embryologists and the MGH Pathology assistants and residents for their contribution to the data presented in this study.

18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44

References

1.

2.

3.

4.

5.

6.

7.

8.

9. 10.

11. 12. 13.

14.

Nastri CO, Ferriani RA, Rocha IA, Martins WP. Ovarian hyperstimulation syndrome: pathophysiology and prevention. Journal of assisted reproduction and genetics. 2010;27(2-3):121-128. Imudia AN, Awonuga AO, Kaimal AJ, Wright DL, Styer AK, Toth TL. Elective cryopreservation of all embryos with subsequent cryothaw embryo transfer in patients at risk for ovarian hyperstimulation syndrome reduces the risk of adverse obstetric outcomes: a preliminary study. Fertil Steril. 2013;99(1):168-173. Imudia AN, Awonuga AO, Doyle JO, 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. Fertil Steril. 2012;97(6):1374-1379. 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 e361-369. Sha T, Yin X, Cheng W, Massey IY. Pregnancy-related complications and perinatal outcomes resulting from transfer of cryopreserved versus fresh embryos in vitro fertilization: a meta-analysis. Fertil Steril. 2018;109(2):330-342 e339. Johnson KM, Hacker MR, Resetkova N, O'Brien B, Modest AM. Risk of ischemic placental disease in fresh and frozen embryo transfer cycles. Fertil Steril. 2019;111(4):714-721. Shapiro BS, Daneshmand ST, Garner FC, Aguirre M, Hudson C, Thomas S. Evidence of impaired endometrial receptivity after ovarian stimulation for in vitro fertilization: a prospective randomized trial comparing fresh and frozen-thawed embryo transfers in high responders. Fertil Steril. 2011;96(2):516-518. Kansal Kalra S, Ratcliffe SJ, Milman L, Gracia CR, Coutifaris C, Barnhart KT. Perinatal morbidity after in vitro fertilization is lower with frozen embryo transfer. Fertil Steril. 2011;95(2):548-553. Kaser DJ, Melamed A, Bormann CL, et al. Cryopreserved embryo transfer is an independent risk factor for placenta accreta. Fertil Steril. 2015;103(5):1176-1184. Roque M, Valle M, Sampaio M, Geber S. Obstetric outcomes after fresh versus frozenthawed embryo transfers: A systematic review and meta-analysis. JBRA assisted reproduction. 2018;22(3):253-260. Sites CK, Wilson D, Barsky M, et al. Embryo cryopreservation and preeclampsia risk. Fertil Steril. 2017;108(5):784-790. Chen ZJ, Shi Y, Sun Y, et al. Fresh versus Frozen Embryos for Infertility in the Polycystic Ovary Syndrome. N Engl J Med. 2016;375(6):523-533. Zhang B, Wei D, Legro RS, et al. Obstetric complications after frozen versus fresh embryo transfer in women with polycystic ovary syndrome: results from a randomized trial. Fertil Steril. 2018;109(2):324-329. Wei D, Liu JY, Sun Y, et al. Frozen versus fresh single blastocyst transfer in ovulatory women: a multicentre, randomised controlled trial. The Lancet. 2019;393(10178):1310-1318.

19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25. 26. 27.

28. 29. 30.

Ernstad EF, Wennerholm U-B, Khatibi A, Petzold M, Bergh C. Neonatal and maternal outcome after frozen embryo transfer: Increased risks in programmed cycles. American journal of obstetrics and gynecology. 2019. Romero R, Kusanovic JP, Chaiworapongsa T, Hassan SS. Placental bed disorders in preterm labor, preterm PROM, spontaneous abortion and abruptio placentae. Best Pract Res Clin Obstet Gynaecol. 2011;25(3):313-327. Khong TY, De Wolf F, Robertson WB, Brosens I. Inadequate maternal vascular response to placentation in pregnancies complicated by pre-eclampsia and by smallfor-gestational age infants. Br J Obstet Gynaecol. 1986;93(10):1049-1059. Knofler M, Haider S, Saleh L, Pollheimer J, Gamage T, James J. Human placenta and trophoblast development: key molecular mechanisms and model systems. Cell Mol Life Sci. 2019. Chavarro JE, Ehrlich S, Colaci DS, et al. Body mass index and short-term weight change in relation to treatment outcomes in women undergoing assisted reproduction. Fertility and sterility. 2012;98(1):109-116. Medicine ASfR. Practice Committee Opinion. Guidelines on number of embryos transferred. Birmingham, AL: American Society for Assisted Reproductive Medicine. 1998. Practice Committee of Society for Assisted Reproductive T, Practice Committee of American Society for Reproductive M. Guidelines on number of embryos transferred. Fertil Steril. 2008;90(5 Suppl):S163-164. Practice Committee of the American Society for Reproductive M, Practice Committee of the Society for Assisted Reproductive T. Guidelines on number of embryos transferred. Fertil Steril. 2009;92(5):1518-1519. Practice Committee of American Society for Reproductive M, Practice Committee of Society for Assisted Reproductive T. Criteria for number of embryos to transfer: a committee opinion. Fertil Steril. 2013;99(1):44-46. Khong TY, Mooney EE, Ariel I, et al. Sampling and definitions of placental lesions: Amsterdam placental workshop group consensus statement. Arch Pathol Lab Med. 2016;140(7):698-713. Pinar H, Sung CJ, Oyer CE, Singer DB. Reference values for singleton and twin placental weights. Pediatr Pathol Lab Med. 1996;16(6):901-907. Parast MM, Crum CP, Boyd TK. Placental histologic criteria for umbilical blood flow restriction in unexplained stillbirth. Human pathology. 2008;39(6):948-953. Evans J, Hannan NJ, Edgell TA, et al. Fresh versus frozen embryo transfer: backing clinical decisions with scientific and clinical evidence. Hum Reprod Update. 2014;20(6):808-821. Weinerman R, Mainigi M. Why we should transfer frozen instead of fresh embryos: the translational rationale. Fertil Steril. 2014;102(1):10-18. Roque M. Freeze-all policy: is it time for that? Journal of assisted reproduction and genetics. 2015;32(2):171-176. Yanaihara A, Hatakeyama S, Ohgi S, et al. Difference in the size of the placenta and umbilical cord between women with natural pregnancy and those with IVF pregnancy. Journal of assisted reproduction and genetics. 2018;35(3):431-434.

20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

31.

32.

33.

34.

35. 36.

37. 38. 39. 40.

41. 42.

43. 44.

45.

46. 47.

Ebbing C, Kiserud T, Johnsen SL, Albrechtsen S, Rasmussen S. Prevalence, risk factors and outcomes of velamentous and marginal cord insertions: a populationbased study of 634,741 pregnancies. PloS one. 2013;8(7):e70380. Räisänen S, Georgiadis L, Harju M, Keski-Nisula L, Heinonen S. Risk factors and adverse pregnancy outcomes among births affected by velamentous umbilical cord insertion: a retrospective population-based register study. Eur J Obstet Gynecol Reprod Biol. 2012;165(2):231-234. Nasiell J, Papdogiannakis N, Löf E, Elofsson F, Hallberg B. Hypoxic ischemic encephalopathy in newborns linked to placental and umbilical cord abnormalities. J Matern Fetal Neonatal Med. 2016;29(5):721-726. Ebbing C, Johnsen SL, Albrechtsen S, Sunde ID, Vekseth C, Rasmussen S. Velamentous or marginal cord insertion and the risk of spontaneous preterm birth, prelabor rupture of the membranes, and anomalous cord length, a population-based study. Acta Obstet Gynecol Scand. 2017;96(1):78-85. Benirschke K, Burton GJ, Baergen RN. Pathlogy of the Human Placenta. 6th ed. New York, New York: Springer; 2012. Zhou J, Wu M, Wang B, et al. The effect of first trimester subchorionic hematoma on pregnancy outcomes in patients underwent IVF/ICSI treatment. J Matern Fetal Neonatal Med. 2017;30(4):406-410. Xiang L, Wei Z, Cao Y. Symptoms of an intrauterine hematoma associated with pregnancy complications: a systematic review. PloS one. 2014;9(11):e111676. Heider A. Fetal vascular malperfusion. Arch Pathol Lab Med. 2017;141:1484-1489. Redline RW, Ravishankar S. Fetal vascular malperfusion, an update. APMIS : acta pathologica, microbiologica, et immunologica Scandinavica. 2018;126(7):561-569. Wright E, Audette MC, Ye XY, et al. Maternal vascular malperfusion and adverse perinatal outcomes in low-risk nulliparous women. Obstet Gynecol. 2017;130(5):1112-1120. Zhang J, Du M, Li Z, et al. Fresh versus frozen embryo transfer for full-term singleton birth: a retrospective cohort study. J Ovarian Res. 2018;11(1):59. 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? Human reproduction. 2014;29(3):618-627. Hiura H, Hattori H, Kobayashi N, et al. Genome-wide microRNA expression profiling in placentae from frozen-thawed blastocyst transfer. Clinical epigenetics. 2017;9:79. Senapati S, Wang F, Ord T, Coutifaris C, Feng R, Mainigi M. Superovulation alters the expression of endometrial genes critical to tissue remodeling and placentation. Journal of assisted reproduction and genetics. 2018;35(10):1799-1808. Horcajadas JA, Riesewijk A, Polman J, et al. Effect of controlled ovarian hyperstimulation in IVF on endometrial gene expression profiles. Mol Hum Reprod. 2005;11(3):195-205. Vivelle-Dot (Estrogen Transdermal System) [package insert]. East Hanover, NJ: Novartis Pharmaceuticals Corporation; 2019. Conrad KP, Graham GM, Chi YY, et al. Potential influence of the corpus luteum on circulating reproductive and volume regulatory hormones, angiogenic and immunoregulatory factors in pregnant women. Am J Physiol Endocrinol Metab. 2019.

21 1 2 3 4 5 6 7 8 9 10 11 12 13

48.

49. 50.

51.

von Versen-Höynck F, Narasimhan P, Selamet Tierney ES, et al. Absent or excessive corpus luteum number is associated with altered maternal vascular health in early pregnancy. Hypertension. 2019;73(3):680-690. Conrad KP. Emerging role of relaxin in the maternal adaptations to normal pregnancy: implications for preeclampsia. Semin Nephrol. 2011;31(1):15-32. Vasquez YM, Wang X, Wetendorf M, et al. FOXO1 regulates uterine epithelial integrity and progesterone receptor expression critical for embryo implantation. PLoS Genet. 2018;14(11):e1007787. Conrad KP, Baker VL. Corpus luteal contribution to maternal pregnancy physiology and outcomes in assisted reproductive technologies. Am J Physiol Regul Integr Comp Physiol. 2013;304(2):R69-72.

Table 1: Pathology Categorization Pathology Anatomic Small placenta by weight Large placenta by weight Marginal cord insertion Membranous cord insertion Coiling anomalies

Two vessel cord Membranous vessels Other cord findings Circummarginate or circumvallate membranes Accessory placental lobe Infectious ACHA-MIR, grade and/or stage ≥ 2 ACHA-FIR, grade and/or stage ≥ 2 Inflammatory Villitis of unknown etiology (VUE), low grade Villitis of unknown etiology (VUE), high grade Chronic deciduitis Chronic histiocytic intervillositis Non-specific inflammation Vascular Maternal vascular malperfusion (MVM)

Fetal vascular malperfusion (FVM), low grade Fetal vascular malperfusion (FVM), high grade

Intervillous thrombi (IVT) Subchorionic thrombi Septal thrombi Disruption

Explanation Less than the 10th percentile by weight for gestational age using standard weight tablea Greater than the 90th percentile by weight for gestational age using standard weight tablea Umbilical cord inserts less than or equal to 1 cm from the placental margin Umbilical cord inserts into the membranesb Hyper- or hypo coiled umbilical cords (4 or more- or 1 or less- coils in 10 cm on average) or rightward coiled umbilical cordc Single umbilical artery Chorionic vessels within the membranes not supported by placental parenchyma e.g. true knots, thin cords, surface nodules, discoloration. Abnormal insertion of the membranes central to the margin of the placenta Succinturiate lobe of the placenta separate from the main disk Acute chorioamnionitis with moderate to severe maternal inflammatory response Acute chorioamnionitis with moderate to severe fetal inflammatory responseb,d Lymphohistiocytic infiltrate within villi, less than 10 contiguous villi affected in any focus, more than 1 focus neededb Lymphohistiocytic infiltrate within villi with 10 or more contiguous villi involved, more than one focus and more than one slide involvedb 50 or more lymphocytes and/or 5 or more plasma cells per high power field in the deciduae A diffuse infiltrate of histiocytes in the maternal vascular spacef,g

Placenta involves two or more of the following: distal villous hypoplasia or accelerated villous maturation, placental infarct, decidual arteriopathy, evidence of acute or chronic abruption, or increased fibrin depositionb Single focus of less than 45 total avascular villi or villous stromal-vascular karyorrhexis or one large vascular thrombusb. Greater than one focus, more than 45 total avascular villi or villous-stromal vascular karyorrhexis, more than 1 large vascular thrombib. Laminated thrombus in maternal vascular space. Laminated thrombus in the subchorionic space in the disk. Thrombus in the septum or within a septal cyst Torn, fragmented, or incomplete disk.

a

Pinar H, Sung CJ, Oyer CE, Singer DB. Reference values for singleton and twin placental weights. Pediatr Pathol Lab Med. 1996;16(6):901-907. Khong TY, Mooney EE, Ariel I, et al. Sampling and definitions of placental lesions: Amsterdam placental workshop group consensus statement. Arch Pathol Lab Med. 2016;140(7):698-713. c van Dijk CC, Franx A, de Laat MW, Bruinse HW, Visser GH, Nikkels PG. The umbilical coiling index in normal pregnancy. J Matern Fetal Neonatal Med. 2002;11(4):280-283. d Redline RW, Faye-Petersen OM, Heller DS, et al. Amniotic infection syndrome: nosology and reproducibility of placental reaction patterns. Pediatr Dev Pathol. 2003;6(5):435-448. e Maroun LL, Mathiesen L, Hedegaard M, Knudsen LE, Larsen LG. Pathologic evaluation of normal and perfused term placental tissue. Pediatr Dev Pathol. 2014;17(5):330-338. f Boyd TK, Redline RW. Chronic histiocytic intervillositis: a placental lesion associated with recurrent reproductive loss. Human pathology. 2000;31(11):1389-1396. g Chen A, Roberts DJ. Placental pathologic lesions with a significant recurrence risk - what not to miss! APMIS : acta pathologica, microbiologica, et immunologica Scandinavica. 2018;126(7):589-601. b

Table 2. Demographic and clinical characteristics of women with live births following fresh vs. programmed frozen cycles Fresh transfers (n=929) Frozen transfers (n=211) P-valuec 35.0 (4.0) 35.2 (3.6) 0.52 Agea 0.62 Race White 727 (78.3%) 160 (75.8%) Black 32 (3.4%) 10 (4.7%) Asian 119 (12.8%) 25 (11.8%) 0.49 Ethnicity Hispanic 10 (1.1%) 3 (1.4%) Other 36 (3.9%) 8 (3.8%) Not specified 12 (1.3%) 4 (1.9%) 23.6 (21.35, 26.8) 23.2 (20.9, 26.3) 0.26 Body mass index (kg/m2)b 656 (70.6%) 101 (47.9%) <0.001 Nulliparity 7.1 (2.2) 7.0 (2.3) 0.44 Day 3 FSH (IU/L)a 3.6 (3.8) 4.9 (5.1) 0.007 AMH (ng/ml)a 2100 (768.9) 2683 (1007.6) <0.001 Peak estradiol on day of trigger (pg/ml)a Infertility Diagnosis Male factor 428 (46.1%) 82 (38.9%) Female factor 469 (50.5%) 134 (63.5%) Tubal factor 138 (14.9%) 35 (16.6%) Ovulatory dysfunction 193 (20.8%) 49 (23.2%) Decreased ovarian reserve 127 (13.7%) 21 (10.0%) Uterine factor 31 (3.3%) 7 (3.3%) Endometriosis 59 (6.4%) 15 (7.1%) Idiopathic 225 (24.2%) 53 (25.1%) Other/Non-infertility 34 (3.7%) 12 (5.7%) <0.001 IVF Stimulation protocol Lupron down-regulation 645 (69.4%) 151 (71.6%) Flare 150 (16.1%) 12 (5.7%) GnRH antagonist 119 (12.8%) 28 (13.3%) Patch 15 (1.6%) 1 (0.5%) 2.0 (0.7) 1.6 (0.57) <0.001 Number of embryos transferreda 0.096 Fertilization method IVF 408 (43.9%) 106 (50.2%) ICSI 521 (56.1%) 105 (49.8%) 37.5 (2.9) 38.0 (3.1) 0.037 Gestational age at deliverya 245 (26.4%) 42 (19.9%) 0.095 Preterm delivery (<37 weeks) 374 (40.3%) 49 (23.2%) <0.001 Multifetal gestation Median birth weightb 3224 (2790, 3600) 3210 (2800, 3610) 0.89 Singletons 2483 (2109, 2802.5) 2700 (2332.5, 2854) 0.040 Twins Small for gestational age infant 61 (11%) 20 (12.4%) 0.111 (singletons) 0.013 Large for gestational age infant 27 (3.0%) 12 (5.7%) (singletons) 0.12 Mode of delivery 437 (47.0%) 87 (41.2%) Vaginal 491 (52.9%) 124 (58.8%) Cesarean 100 (10.8%) 19 (9%) 0.45 Preeclampsia 134 (14.4%) 35 (16.6%) 0.42 Gestational diabetes Hemorrhage (EBL > 500cc for vaginal or 134 (14.4%) 35 (16.6%) 0.42 >1000cc for c-section) Retained placenta requiring manual 27 (6.2%) 12 (13.8%) 0.013 extraction or curettage in vaginal deliveries

a

Mean (standard deviation) Median (interquartile range) c P < 0.05 is significant. b

Table 3. Incidence of placental pathology in live births resulting from fresh vs. programmed frozen cycles All transfers Day 5 transfers Single embryo transfers Fresh Fresh Day 5 d d Frozen (n=211) P-value Frozen blastocyst (n=184) P-value Fresh (n=206) Frozen (n=101) (n=929) (n=466) Mean singleton placental weight (g) 453.5 (115.6) 455.1 (122.6) 0.88 443.8 (111.2) 456.8 (125.4) 0.27 441.7 (110.9) 447 (113.9) Mean placental weight percentile 34.8 (29.3) 37.1 (31.4) 0.35 34.2 (29.5) 37.8 (32.4) 0.20 30.9 (30.6) 32.9 (30.8) 568 (61.1%) 133 (63.0%) 0.61 294 (63.1%) 122 (66.3%) 0.44 149 (72.3%) 69 (68.3%) Anatomic Small (<10%ile) 295 (31.8%) 64 (30.3%) 0.69 155 (33.3%) 57 (31.0%) 0.58 88 (42.7%) 36 (35.6%) Large (>90%ile) 71 (7.6%) 22 (10.4%) 0.18 37 (7.9%) 22 (12.0%) 0.11 17 (8.3%) 8 (7.9%) Marginal cord insertion 100 (10.8%) 38 (18.0%) 0.004 50 (10.7%) 36 (19.6%) 0.003 25 (12.1%) 24 (23.8%) Membranous cord insertion 69 (7.4%) 17 (8.1%) 0.75 39 (8.4%) 16 (8.7%) 0.89 13 (6.3%) 5 (5.0%) Twisting anomalies 43 (4.6%) 8 (3.8%) 0.60 8 (1.7%) 1 (0.5%) 0.25 11 (5.3%) 5 (5.0%) Single umbilical artery 15 (1.6%) 1 (0.5%) 0.20 21 (4.5%) 8 (4.3%) 0.93 4 (1.9%) 1 (1.0%) Membranous vessels 29 (3.1%) 14 (6.6%) 0.016 17 (3.6%) 14 (7.6%) 0.033 6 (2.9%) 8 (7.9%) Other cord findings 42 (4.5%) 4 (1.9%) 0.080 23 (4.9%) 4 (2.2%) 0.11 9 (4.4%) 2 (2.0%) Circummarginate or circumvallate membranes 56 (6.0%) 6 (2.8%) 0.066 28 (6.0%) 4 (2.2%) 0.042 14 (6.8%) 2 (2.0%) Accessory placental lobe 48 (5.2%) 22 (10.4%) 0.004 19 (4.1%) 20 (10.9%) 0.001 7 (3.4%) 10 (9.9%) 242 (26.0%) 62 (29.4%) 0.32 127 (27.3%) 53 (28.8%) 0.69 68 (33.0%) 27 (26.7%) Infectious Acute chorioamnionitis with maternal 0.94 89 (9.6%) 23 (10.9%) 0.56 26 (12.6%) 10 (9.9%) inflammatory response, Moderate to Severe 49 (10.5%) 19 (10.3%) Acute chorioamnionitis with fetal 0.69 37 (4.0%) 8 (3.8%) 0.90 10 (4.9%) 3 (3.0%) inflammatory response, Moderate to Severe 21 (4.5%) 7 (3.8%) 160 (17.2%) 45 (21.3%) 0.16 74 (15.9%) 41 (22.3%) 0.054 38 (18.4%) 26 (25.7%) Inflammatory Villitis of unknown etiology, low-grade 88 (9.5%) 29 (13.7%) 0.065 38 (8.2%) 26 (14.1%) 0.021 22 (10.7%) 16 (15.8%) Villitis of unknown etiology, high-grade 26 (2.8%) 5 (2.4%) 0.73 15 (3.2%) 4 (2.2%) 0.48 7 (3.4%) 3 (3.0%) Chronic deciduitis 40 (4.3%) 10 (4.7%) 0.78 25 (5.4%) 9 (4.9%) 0.81 11 (5.3%) 3 (3.0%) Chronic Histiocytic Intervillositis 4 (0.4%) 2 (0.9%) 0.35 1 (0.2%) 1 (0.5%) 0.50 1 (0.5%) 0 (0.0%) Non-specific inflammation 26 (2.8%) 9 (4.3%) 0.26 11 (2.4%) 9 (4.9%) 0.092 6 (2.9%) 8 (7.9%) 646 (69.5%) 154 (73.0%) 0.32 313 (67.2%) 134 (72.8%) 0.16 141 (68.4%) 72 (71.3%) Vascular MVMa 332 (35.7%) 78 (37.0%) 0.74 147 (31.5%) 71 (38.6%) 0.087 65 (31.6%) 41 (40.6%) FVMb, low grade 83 (8.9%) 27 (12.8%) 0.086 43 (9.2%) 24 (13.0%) 0.15 21 (10.2%) 15 (14.9%) FVMb, high grade 26 (2.8%) 8 (3.8%) 0.44 12 (2.6%) 9 (4.9%) 0.13 4 (1.9%) 3 (3.0%) Intervillous thrombi 155 (16.7%) 51 (24.2%) 0.011 75 (16.1%) 45 (24.5%) 0.013 42 (20.4%) 28 (27.7%) Subchorionic thrombi 19 (2.0%) 17 (8.1%) <0.001 8 (1.7%) 16 (8.7%) <0.001 4 (1.9%) 8 (7.9%) Septal thrombi 1 (0.1%) 5 (2.4%) <0.001 1 (0.2%) 5 (2.7%) 0.003 0 (0.0%) 2 (2.0%) Cord anomalyc with MVM 91 (9.8%) 32 (15.2%) 0.02 47 (10.1%) 31 (16.8%) 0.017 24 (11.7%) 19 (18.8%) Cord anomalyc with FVM 34 (3.7%) 18 (8.5%) 0.002 19 (4.1%) 18 (9.8%) 0.005 9 (4.4%) 11 (10.9%) Disruption 134 (14.4%) 23 (10.9%) 0.18 69 (14.8%) 21 (11.4%) 0.26 25 (12.1%) 9 (8.9%) a MVM refers to maternal vascular malperfusion. b FVM refers to fetal vascular malperfusion. c Cord anomaly includes marginal cord insertion, membranous cord insertion, two vessel cords, coiling anomalies, membranous vessels, and other cord findings (Table 1). d P < 0.05 is significant.

P-valued 0.68 0.61 0.47 0.74 0.92 0.009 0.63 0.89 0.54 0.048 0.29 0.074 0.019 0.26 0.49 0.44 0.14 0.20 0.84 0.35 0.48 0.048 0.61 0.12 0.23 0.57 0.15 0.011 0.043 0.089 0.03 0.40

Table 4. Unadjusted and adjusted odds of clinically significant placental pathologies in programmed frozen transfers compared to fresh transfers All transfers (n = 1140)a Unadjusted ORd Adjusted ORe (aOR) 1.08 (0.8, 1.48) 0.93 (0.67, 1.31) Anatomic Marginal cord insertion 1.82 (1.21, 2.74) 1.87 (1.21, 2.91) Membranous cord insertion 0.45 (0.19, 1.09) 0.32 (0.09, 1.14) Accessory lobe 2.5 (1.08, 5.76) 2.97 (1.1, 7.96) Cord anomaly with small placentab 1.68 (1.03, 2.72) 1.6 (0.95, 2.7) 1.19 (0.84, 1.67) 1.13 (0.79, 1.61) Infectious 1.39 (0.84, 2.29) 1.29 (0.84, 2) Inflammatory Villitis of unknown etiology, high-grade 0.84 (0.3, 2.32) 0.49 (0.11, 2.15) 1.19 (0.84, 1.71) 1.24 (0.84, 1.81) Vascular/thrombotic MVM 1.05 (0.77, 1.44) 1.12 (0.8, 1.57) FVM, high grade 1.37 (0.61, 3.07) 1.47 (0.61, 3.5) Subchorionic thrombi 4.2 (2.14, 8.22) 3.72 (1.8, 7.71) 2.45 (1.36, 4.44) 2.34 (1.22, 4.46) Cord anomaly with FVM Day 5 transfers (n = 650)c 1.15 (0.8, 1.65) 1.06 (0.72, 1.55) Anatomic Marginal cord insertion 2.58 (0.99, 6.76) 2.88 (1.1, 7.57) Membranous cord insertion 0.35 (0.12, 1.01) 0.3 (0.1, 0.89) Accessory lobe 6.72 (1.62, 27.77) 5.32 (1.54, 18.38) Cord anomaly with small placentab 1.77 (1.04, 3.03) 2.03 (0.67, 6.15) 1.08 (0.74, 1.58) 1.05 (0.7, 1.57) Infectious 1.88 (0.02, 152.89) 1.51 (0.73, 3.1) Inflammatory Villitis of unknown etiology, high-grade 0.64 (0.17, 2.31) 0.46 (0.13, 1.64) 1.31 (0.9, 1.91) 1.31 (0.88, 1.95) Vascular/thrombotic MVM 1.36 (0.96, 1.95) 1.41 (0.96, 2.05) FVM, high grade 1.95 (0.81, 4.70) 2.62 (0.97, 7.14) Subchorionic thrombi 5.45 (2.29, 12.97) 5.38 (2.15, 13.49) 2.55 (1.31, 4.98) 2.39 (1.16, 4.92) Cord anomaly with FVM Single embryo transfers (n = 597)d 0.82 (0.49, 1.39) 0.68 (0.34, 1.32) Anatomic Marginal cord insertion 4.71 (1.19, 18.64) 2.35 (0.42, 13.07) Membranous cord insertion 0.28 (0.06, 1.24) 0.21 (0.02, 2.69) Accessory lobe 5.07 (0.91, 28.2) 12.31 (0.65, 234.66) Cord anomaly with small placentab 2.84 (0.67, 12.1) 2.29 (0.36, 14.6) 0.74 (0.44, 1.26) 0.75 (0.38, 1.49) Infectious 1.58 (0.78, 3.21) 1.32 (0.48, 3.67) Inflammatory Villitis of unknown etiology, high-grade 0.8 (0.1, 6.68) 0.62 (0.08, 4.58) 1.14 (0.68, 1.93) 1.41 (0.72, 2.76) Vascular/thrombotic MVM 1.48 (0.9, 2.43) 1.74 (0.9, 3.37) FVM, high grade 1.55 (0.34, 7.04) --Subchorionic thrombi 4.34 (1.28, 14.8) 10.0 (1.46, 69.1) 2.68 (1.07, 6.68) --Cord anomaly with FVM a Adjusted for age, race, BMI, uterine factor infertility, number of embryos transferred, day of embryo transfer, date of retrieval, gestational age at delivery, multiple gestation. b Cord anomaly includes marginal insertion, membranous insertion, SUA, twisting, membranous vessels, and other cord anomalies. c Adjusted for age, race, BMI, uterine factor infertility, number of embryos transferred, date of retrieval, gestational age at delivery, multiple gestation. d Adjusted for age, race, BMI, uterine factor infertility, day of embryo transfer, date of retrieval, gestational age at delivery, multiple gestation. e Presented as odds ratio (95% CI).

Supplemental Table 1. Maternal demographic and cycle characteristics at time of first oocyte retrieval Patients (n=1079) 34.9 (3.9) Age Race White 839 (77.8%) Black 42 (3.9%) Asian 134 (12.4%) Ethnicity Hispanic 12 (1.1%) Non-Hispanic 782 (72.5%) Other 43 (4.0%) Not specified (7) 14 (1.3%) 23.6 (21.3, 26.8) BMI (median, IQR) 758 (70.3%) Nulliparous 7.1 (2.2) Day 3 FSH 4.0 (4.2) AMH (2014-2017) Infertility Diagnosis Male factor 481 (44.6%) Female factor 569 (52.7%) Tubal factor 164 (15.2%) Ovulatory dysfunction 69 (6.4%) PCOS 120 (11.1%) Decreased ovarian reserve 136 (12.6%) Uterine factor 34 (3.2%) Endometriosis 66 (6.1%) Idiopathic 269 (24.9%) Other/Non-infertility 45 (4.2%)

Supplemental Table 2. Clinical subgroup analysis of placental pathology in fresh and programmed frozen transfers All transfers, nulliparous women All transfers, singleton births Fresh (n=699) Frozen (n=127) P-valued Fresh (n=555) Frozen (n=162) P-valued Mean singleton weight (g) 448.9 (115.3) 461.1 (125.6) 0.35 453.5 (115.6) 455.1 (122.6) 0.88 Mean weight %ile 33.6 (29.3) 35.2 (30.7) 0.59 33.5 (31.0) 34.5 (32.2) 0.71 437 (62.5%) 79 (62.2%) 0.95 358 (64.5%) 106 (65.4%) 0.83 Anatomic Small (<10%ile) 236 (33.8%) 42 (33.1%) 0.88 207 (37.3%) 58 (35.8%) 0.73 Large (>90%ile) 53 (7.6%) 10 (7.9%) 0.91 51 (9.2%) 16 (9.9%) 0.79 Marginal cord insertion 79 (11.3%) 25 (19.7%) 0.009 51 (9.2%) 30 (18.5%) <0.001 Membranous cord insertion 55 (7.9%) 7 (5.5%) 0.35 29 (5.2%) 10 (6.2%) 0.64 Twisting anomalies 23 (4.1%) 7 (4.3%) 0.92 11 (5.3%) 4 (3.1%) 0.89 Single umbilical artery 8 (1.4%) 7 (4.3%) 0.41 8 (1.1%) 0 (0%) 0.54 Membranous vessels 23 (3.3%) 11 (8.7%) 0.005 9 (1.6%) 11 (6.8%) <0.001 Other cord findings 19 (2.7%) 2 (1.6%) 0.45 18 (3.2%) 2 (1.2%) 0.17 Circummarginate or circumvallate membranes 41 (5.9%) 2 (1.6%) 0.045 32 (5.8%) 3 (1.9%) 0.042 Accessory placental lobe 37 (5.3%) 17 (13.4%) <0.001 25 (4.5%) 16 (9.9%) 0.010 212 (30.3%) 47 (37.0%) 0.14 179 (32.3%) 51 (31.5%) 0.85 Infectious Acute chorioamnionitis with maternal inflammatory 76 (10.9%) 18 (14.2%) 0.28 62 (11.2%) 18 (11.1%) 0.98 response, Moderate to Severe Acute chorioamnionitis with fetal inflammatory response, 31 (4.4%) 8 (6.3%) 0.36 24 (4.3%) 8 (4.9%) 0.74 Moderate to Severe 119 (17.0%) 25 (19.7%) 0.47 102 (18.4%) 33 (20.4%) 0.57 Inflammatory Villitis of unknown etiology, low-grade 69 (9.9%) 16 (12.6%) 0.35 60 (10.8%) 20 (12.3%) 0.59 Villitis of unknown etiology, high-grade 16 (2.3%) 3 (2.4%) 0.96 15 (2.7%) 3 (1.9%) 0.54 Chronic deciduitis 26 (3.7%) 6 (4.7%) 0.59 24 (4.3%) 5 (3.1%) 0.48 Chronic Histiocytic Intervillositis 4 (0.6%) 1 (0.8%) 0.77 3 (0.5%) 1 (0.6%) 0.91 Non-specific inflammation 21 (3.0%) 6 (4.7%) 0.32 17 (3.1%) 8 (4.9%) 0.25 480 (68.7%) 95 (74.8%) 0.17 373 (67.2%) 117 (72.2%) 0.23 Vascular a MVM 249 (35.6%) 49 (38.6%) 0.52 178 (32.1%) 59 (36.4%) 0.30 FVMb, low grade 54 (7.7%) 18 (14.2%) 0.018 47 (8.5%) 21 (13.0%) 0.086 FVMb, high grade 21 (3.0%) 4 (3.1%) 0.93 11 (2.0%) 4 (2.5%) 0.70 Intervillous thrombi 114 (16.3%) 32 (25.2%) 0.016 91 (16.4%) 36 (22.2%) 0.087 Subchorionic thrombi 12 (1.7%) 10 (7.9%) <0.001 11 (2.0%) 14 (8.6%) <0.001 Septal thrombi 1 (0.1%) 2 (1.6%) 0.014 1 (0.2%) 2 (1.2%) 0.067 Cord anomalyc with MVM 69 (9.9%) 19 (15.0%) 0.087 39 (7.0%) 24 (14.8%) 0.002 Cord anomalyc with FVM 26 (3.7%) 10 (7.9%) 0.035 14 (2.5%) 14 (8.6%) <0.001 Disruption 102 (14.6%) 13 (10.2%) 0.19 81 (14.6%) 16 (9.9%) 0.12 a MVM refers to maternal vascular malperfusion. b FVM refers to fetal vascular malperfusion. c Cord anomaly includes marginal cord insertion, membranous cord insertion, two vessel cords, coiling anomalies, membranous vessels, and other cord findings (Table 1). d P < 0.05 is significant.