Assessment of the embryo flash position and migration with 3D ultrasound within 60 min of embryo transfer

Human Reproduction, Vol.31, No.3 pp. 591 596, 2016 Advanced Access publication on January 11, 2016 doi:10.1093/humrep/dev343 ORIGINAL ARTICLE Infertility Assessment of the embryo flash position and migration with 3D ultrasound within 60 min of embryo transfer Sotirios H. Saravelos, Alice Wai Yee Wong, Carol Pui Shan Chan, Grace Wing Shan Kong, Lai Ping Cheung, Cathy Hoi Sze Chung, Jacqueline Pui Wah Chung, and Tin-Chiu Li* Assisted Reproductive Technology Unit, Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong *Correspondence address. E-mail: tinchiu@cuhk.edu.hk Submitted on July 11, 2015; resubmitted on November 11, 2015; accepted on December 14, 2015 studyquestion: Does the air bubble (embryo flash) position and migration as visualized with 3D ultrasound (US) within 60 min of embryo transfer correlate with clinical outcome following fresh ART transfer cycles? summaryanswer: The location of the embryo flash and the direction of its movement at 60 min, but not at 1 or 5 min after transfer, are associated with clinical pregnancy. what is known already: Studies assessing the relation between the pregnancy rate and the position of the catheter tip and/or the position of the air bubbles following embryo transfer show conflicting results to date. study design, size and duration: This was a prospective cohort study including 277 infertile women undergoing ART between July 2011 and August 2013. participants/materials, setting and methods: Good prognosis patients undergoing fresh ART cycles within a single tertiary University unit were assessed by 3D US at 1, 5 and 60 min after embryo transfer. The distance of the embryo flash from the fundus was measured at these time points, along with the direction of the embryo flash movement within 60 min of transfer. main results and the role of chance: Within 60 min of embryo transfer, 76.4% (198/259) of the embryo flashes migrated towards the fundus, 12.4% (32/259) migrated towards the cervix and 11.2% (29/259) remained static. There was no significant association between the embryo position or movement and the pregnancy rate at 1 and 5 min. At 60 min, however, the pregnancy and implantation rates among subjects with embryo flashes located,15 mm from the fundus was significantly higher than those with embryo flashes located.15 mm from the fundus (46.5 and 32.8% versus 25.8 and 18.2%, respectively; P, 0.05). The pregnancy and implantation rates when the embryo flash was seen moving towards the cervix (25.0 and 15.0%) was significantly lower (P, 0.05 and P, 0.01, respectively) compared with those remaining static (55.2 and 37.7%) or moving towards the fundus (45.5 and 32.8%). limitations and reasons for caution: Although the air bubbles seen at the time of embryo transfer are thought to represent the position of the embryo, they are in fact a surrogate marker of the embryo itself, as this cannot be directly visualized by US. wider implications of the findings: Following embryo transfer the majority, but not all, embryo flashes undergo significant migration towards the fundus. The location of the embryo flash and the direction of its movement at 60 min, but not at 1 or 5 min after transfer, is associated with clinical outcome. These findings may challenge the traditional notion that the exact position of the embryo flash immediately following embryo transfer is related to clinical outcome. study funding/competing interest(s): The authors have no relevant funding or conflicts of interest to disclose. Key words: assisted reproductive technology / embryo transfer / embryo flash / implantation / 3D ultrasound & The Author 2016. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please email: journals.permissions@oup.com

592 Saravelos et al. Introduction Embryo transfer is arguably one of the most crucial steps in ART; however, for over 30 years there has been very little change or improvement in this procedure (Mains and Van Voorhis, 2010). Since the introduction of ultrasound (US) guided embryo transfer and the subsequent evidence favouring its use (Buckett, 2003; Eskandar et al., 2008), studies have emerged assessing the relation between distance of the catheter tip to the fundus and the chance of pregnancy. Most of these initial studies have suggested that placing the embryo in the lower part of the uterine cavity results in higher pregnancy rates (Coroleu et al., 2002; Frankfurter et al., 2004; Pacchiarotti et al., 2007), while others suggested a central location as being most favourable (Oliveira et al., 2004; Tiras et al., 2010). More recent publications have moved on to assess the position of the air bubbles that are deposited along with the embryo at the time of transfer, with the presumption that they may reflect the actual embryo position more accurately (Friedman et al., 2011). Gergely et al. (2005) in fact demonstrated that 3D ultrasound (3D US) could also be used to identify these air bubbles quite clearly in the coronal plane, describing them as the embryo flash. Interestingly, the majority of the air bubble studies to date seem to suggest that pregnancy rates are highest when the air bubbles are seen closer to the fundus (Lambers et al., 2007; Friedman et al., 2011; Cenksoy et al., 2014), which is not consistent with the findings from the catheter tip studies. However, it yet remains unknown whether and how the embryo position may change following transfer into the uterine cavity. It is possible that the embryo undergoes significant migration, which could render the initial assessment of the air bubbles or catheter tip position following embryo transfer rather redundant. In this prospective study, we aimed to examine the position and movement of the air bubbles at 1, 5 and 60 min following embryo transfer and relate the findings to the implantation and clinical pregnancy rates. Materials and Methods Patient recruitment Two hundred and seventy-seven good prognosis infertility patients undergoing fresh cycle IVF/ICSI treatment at the Prince of Wales Hospital, Hong Kong, between July 2011 and August 2013, were recruited in this prospective cohort study. The inclusion criteria included all patients with infertility undergoing fresh cycle transfer of at least one good quality embryo. The exclusion criteria included: (i) women of 40 years or above, (ii) congenital or acquired uterine anomalies, (iii) presence of hydrosalpinx, (iv) repeated implantation failure and (v) cycle cancelation. Ethical approval The study was approved by the Institutional Review Board of our hospital (Joint Chinese University of Hong Kong New Territories East Cluster Clinical Research Ethics Committee registration number CRE 2011.303). All patients completed a written informed consent in line with the study protocol. Stimulation protocol For the long down-regulation protocol, buserelin nasal spray (Suprecur, Hoechst, Germany) 600 mg daily was administered for at least 14 days from the mid-luteal phase of the preceding cycle. For the short protocol, Ganirelix or Cetrorelix subcutaneous injection (Cetrotide, Merck Serono, Germany) 0.25 mg daily was administered from Day 6 of stimulation (or alternatively when estradiol (E2).800 pmol/l or the leading follicle was.14 mm). Ovarian stimulation was achieved using human menopausal gonadotrophins (hmg) (Pergonal, Serono, Aubonne/ Switzerland) or recombinant follicle stimulating hormone (rfsh) (Gonad-F, Serono, Switzerland; or Puregon, Organon, Holland) at doses ranging from 150 to 450 IU/day. When three or more follicles 18 mm in diameter were seen, 5000 IU of hcg (Profasi, Serono, Switzerland) was administered followed by transvaginal oocyte retrieval 36 h later and subsequent IVF or ICSI. Embryo transfer One or two embryos were transferred on Day 3. The majority of embryos transferred (.90%) were considered as good quality in this cohort ( 7 blastomeres of equal size, with,10% fragmentation), confirming that the study population was made of a uniformly good prognosis group of patients. Progesterone (Crinone gel 8% daily, Merck Serono, Germany or Endometrin 100 mg BD, Ferring, Switzerland) was given as luteal phase support (LPS). Patients were placed in a lithotomy position after confirming that they had a moderately full bladder. A bivalve speculum was inserted into the vagina, and the cervical mucus was cleared using a sterile cotton wool stick soaked in culture medium. Embryos were loaded into an atraumatic Cook Sydney embryo transfer catheter (Cook Medical, IN, USA) by using 20 ml of culture medium containing the embryos, followed by a drop of air, followed by a further drop of culture medium (total loading volume 25 ml). They were then transferred into the middle of the uterine cavity under US guidance, aiming for an inner catheter tip placement of 15 mm from the fundus, and an embryo/air bubble deposition 10 20 mm from the fundus. The catheter was then returned to the embryologist to confirm that there were no retained embryos. The patients were subsequently advised bed rest for 1 h following the transfer in a dedicated area adjacent to the procedure room. Ultrasound assessment 3D ultrasonography was performed by using a 3D/4D General Electric Voluson 730 Expert series US machine and a RAB2-5L, 3D/4D probe (GE Medical Systems Kretztechnik GmbH & Co, Austria). To reduce bias and variability all examinations were performed using the same machine and by a single operator (CPSC). 3D US was employed to allow for storing of the volume for subsequent review and analysis, as an alternative to storing a 2D US video. The 3D US volume contained a sweep of multiple 2D images, which allowed for longitudinal, sagittal and coronal views to be reviewed following the embryo transfer with no pressure of time. The scans were performed at 1, 5 and 60 min after embryo transfer in a standardized manner. The uterus was visualized in the sagittal plane, and an automated 60 80 degree sweep was then performed with the patient lying completely still. Each volume was stored and subsequently analysed in the Sectional Plane mode, where the coronal image (C plane) was reconstructed by placing the centre dot onto the embryo flash of the sagittal (A plane) and transverse (B plane) images. Adjustments were performed with the x y z axes where necessary to further clarify the relation of the embryo flash to the uterine cavity and fundal outline. The embryo flash position was assessed by measuring its perpendicular distance to the uterine fundus in the coronal image (example given in

Assessment of embryo flash position and migration with 3D ultrasound 593 Figure 1 Example of an embryo flash assessment following 3D ultrasound acquisition. The left image is the sagittal plane, the middle image is the transverse plane and the right image is the reconstructed coronal plane. In this case, the embryo flash is seen 20 mm from the fundus. Fig. 1). When more than one embryo flashes were seen, as in the case of bubble splitting (Confino et al., 2007), the largest one was used for the measurements. The embryo flash movement/migration was assessed by measuring the change of the embryo flash position 60 min after embryo transfer. If the embryo flash had remained within +1.5 mm from its original position it was classed as static, if it had migrated.1.5 mm towards the fundus or.1.5 mm towards the cervix it was classified as fundal and cervical, respectively. Statistical analysis All data were entered and analysed in SPSS for Windows (Version 22, SPSS, Inc., IL, USA). Continuous data were expressed in mean + standard deviation (SD) and were compared using Student s T-test. Categorical data were expressed as percentages (%) and compared using Fisher s exact test or 2 2 x 2 test (according to the minimum expected count), and 2 3 and 3 3 x 2 tests for higher order contingency tables. Binary logistic regression analysis was used to determine the association of all relevant variables with the achievement of clinical pregnancy (yes versus no). A two-sided value of P, 0.05 was considered as statistically significant. Results A total of 277 women were included in the analysis after exclusion of 14 women (9 requiring freezing of all embryos, 2 with no viable embryos, 2 with symptoms of genital tract infection and 1 because of withdrawal from the trial for personal reasons). Three cases with ectopic pregnancies and six cases with biochemical pregnancies were also excluded from the primary analysis. The patient characteristics for women with different embryo flash positions at 60 min following embryo transfer are presented in Table I. Embryo flash visualization Of 277 women the embryo flash was visualized in 99% after 1 and 5 min (275/277 and 273/277, respectively) and in 94% (259/277) after 60 min. In two subjects, the embryo flash was seen dropping steadily down to the cervical canal within 5 min of transfer. In these cases, as expected, the embryo flashes were no longer visualized in the uterine cavity at 60 min. The difference in pregnancy rate between women in whom the embryo flash was visualized (1 min, 43.3% [119/275]; 5 min 43.6% [119/273] and 60 min 44.0% [114/259]) and those in whom the embryo flash was not visualized (1 min, 0% [0/2]; 5 min 0% [0/4] and 60 min 27.8% [5/18]), did not reach statistical significance. Embryo flash location at different time points The mean embryo flash distance from the fundus at 1, 5 and 60 min following embryo transfer was 13.8 + 5.4, 11.6 + 6.0 and 8.1 + 6.5 mm, respectively, with significant differences (all P, 0.001) between groups. At 60 min, the pregnancy and implantation rates in women with embryo flash located,15 mm from the fundus (46.5 and 32.8%, respectively) were significantly higher (P, 0.05) than those of women in whom the embryo flash was located.15 mm from the fundus (25.8 and 18.2%, respectively). However, there was no significant difference between the two groups at 1 or 5 min (Table II). Embryo flash migration Overall, the mean distance of migration at 60 min was 5.9 + 7.8 mm towards the fundus. At this particular time point, 76.4% (198/259) had migrated towards the fundus, 11.2% (29/259) of embryo flashes had remained static, whereas the remaining 12.4% (32/259) had migrated towards the cervix, The respective pregnancy and implantation rates of these three groups were: 45.5 and 32.5% (fundal group); 55.2 and 37.7% (static group); 25.0 and 15.0% (cervical group) (P, 0.05 and P, 0.01 for pregnancy and implantation rates, respectively) (Table III). Regression analysis Binary logistic regression analysis was carried out for clinical pregnancy (yes versus no) against all relevant parameters listed in Table I. Only two variables were found to be significant in the regression model, namely, the progesterone level on the day of embryo transfer (P ¼ 0.012) followed by the uterine contraction frequency following embryo transfer (P ¼ 0.032). Further analysis, revealed that neither of these two factors were related to the position or migration of the embryo flash at any of the different time points.

594 Saravelos et al. Table I Patient characteristics according to embryo flash position 60 min following embryo transfer. Embryo flash at 60 min... P-Value <15 mm from fundus (n 5 228) >15 mm from fundus (n 5 31)... Age (years) 35.0 + 3.1 34.7 + 3.4 0.63 BMI (kg/m 2 ) 21.4 + 3.2 21.3 + 2.8 0.89 Smoker 8.8% (20/228) 6.5% (2/31) 0.93 Type of infertility Primary 55.3% (126/228) 64.5% (20/31) Secondary 44.7% (102/228) 35.5% (11/31) 0.43 Infertility duration (years) 5.4 + 3.2 4.9 + 2.3 0.39 Main cause of infertility Ovulatory problem 4.8% (11/228) 3.2% (1/31) 0.69 Tubo-peritoneal 55.3% (126/228) 45.2% (14/31) 0.39 Male 27.7% (63/228) 38.7% (12/31) 0.29 Endometriosis 7.5% (17/228) 6.5% (2/31) 0.84 Unexplained 8.8% (20/228) 6.5% (2/31) 0.93 Baseline FSH (IU/l) 7.3 + 2.1 7.0 + 1.7 0.39 Baseline LH (IU/l) 2.6 + 1.7 2.8 + 2.0 0.69 Baseline E2 (pmol/l) 78.0 + 64.3 89.1 + 62.8 0.37 Treatment protocol Agonist 81.1% (185/228) 83.9% (26/31) Antagonist 18.9% (43/228) 16.1 (5/31) 0.90 E2 on day of HCG (pmol/l) 11070.4 + 6562.0 11208.1 + 5915.5 0.91 Number of mature oocytes retrieved per patient 8.1 + 3.3 8.5 + 3.9 0.55 Number of embryos transferred 1 14.0% (32/228) 22.6% (7/31) 2 86.0% (196/228) 77.4% (24/31) 0.33 E2 on day of embryo transfer (pmol/l) 5675.9 + 3466.4 5248.6 + 2588.3 0.52 P4 on day of embryo transfer (nmol/l) 275.1 + 137.8 300.1 + 152.0 0.36 Endometrial thickness on day of embryo transfer (mm) 11.3 + 2.6 12.2 + 3.0 0.08 Contraction frequency following embryo transfer 2.0 + 1.2 1.9 + 1.1 0.91 Embryo flash migration over 60 min from transfer Fundal 84.6% (193/228) 16.1% (5/31) Static 10.1% (23/228) 19.4% (6/31) Cervical 5.3% (12/228) 64.5% (20/31),0.001 Data presented as mean + SD for continuous variables and % (no) for categorical variables. Table II Pregnancy and implantation rates according to the position of the embryo flash relative to the fundus at 1, 5 and 60 min after embryo transfer. Embryo flash <15 mm group... >15 mm group... Implantation rate Pregnancy rate Implantation rate Pregnancy rate... 1 min 31.7% (99/312) 43.2% (73/169) 30.1% (59/196) 43.4% (46/106) 5 min 30.8% (119/386) 43.5% (91/209) 32.8% (39/119) 43.8% (28/64) 60 min 32.8% (139/424) a 46.5% (106/228) b 18.2% (10/55) a 25.8% (8/31) b a,b P, 0.05 in comparison between the,15 mm group and the.15 mm group.

Assessment of embryo flash position and migration with 3D ultrasound 595 Table III Pregnancy and implantation rates according to embryo flash movement within 60 min of embryo transfer. Embryo flash Implantation Pregnancy movement rate rate... Fundal migration group 32.8% (120/366) 45.5% (90/198) Static group 37.7% (20/53) 55.2% (16/29) Cervical migration group 15.0% (9/60) a 25.0% (8/32) b a P, 0.01; b P, 0.05 for cervical migration group in comparison with the static group and the fundal migration group. Discussion For some time now, it is believed that the position of the embryo flash at the time of transfer is related to the clinical pregnancy rates (Lambers et al., 2007; Friedman et al., 2011; Cenksoy et al., 2014). Our findings, however, suggest that it is not the position of the embryo flash at the time of transfer, but rather, the position at least 60 min after the transfer which is related to clinical outcome. This observation does make biological sense as it is the final position of the embryo which determines the location and likelihood of successful implantation. In analysing embryo flash migration, we found that the majority migrated towards the fundus (76.4%), while a similar proportion remained static (12.4%) or migrated towards the cervix (11.2%). Interestingly, the pregnancy rate in the latter appeared to be significantly reduced. This is consistent with the finding of Tiras et al. (2012), although they only assessed the embryo flash migration immediately after transfer. It is also compatible with the findings of Zhu et al. (2014), who found significant contrast fluid migration over 30 min to the extent of expulsion outside the uterine cavity among subjects undergoing mock embryo trials. It could indeed be that embryo flashes migrating towards the cervix denote an increased risk of embryo expulsion from the uterine cavity. To the best of our knowledge, this it is the first study to assess the position and migration of embryo flashes at three different time points following embryo transfer. Strengths of the study involve the homogeneous nature of the population studied (i.e. only Day 3 embryo transfers and. 90% good quality embryos), reduced observer variability (as all measurements made by a single observer) and the use of 3D rather than 2D US to locate the embryo flash. The main limitation of the study is that it is assumed that the embryo flash reflects the actual embryo position. This may not necessarily be true, although an earlier study has demonstrated that over 80% of the embryo flash locations are associated with the eventual gestational sac position (Baba et al., 2000). Another limitation is that although the measurement between the embryo flash and the fundus allows for accurate determination of migration, it does not account for the potential impact of different sizes of uteri. Future studies, could attempt to employ different measurements to account for this, for example by measuring the distance from the fundus to the internal os as described previously by Frankfurter et al. (2003). In addition, the findings cannot be extrapolated easily to Day 5 embryos, given the potentially different uterine environment and timing of implantation that may occur compared with Day 3 embryos. It is also important to point out that in the present study, the assessment of embryo flash migration was performed without additional exogenous hormonal treatment (i.e. LPS was initiated after the embryo transfer). The purpose for using this regime in the study was to examine the uterine physiology without the effect of exogenous hormonal support. However, it is possible that the results may be affected by the type, dose and duration of luteal support used. Finally, the numbers are too small to draw conclusions with regards to the biochemical and ectopic pregnancies (mean embryo flash distances from the fundus at 1, 5 and 60 min were 13.3 + 3.9, 12.3 + 7.9, 4.8 + 2.6 and 10.2 + 3.4, 9.8 + 7.3, 11.7 + 11.5, respectively), although these outcomes are also potentially related to the transfer technique and embryo flash position. In conclusion, this study demonstrates that the majority of embryo flashes undergo significant migration over a period of 60 min from the time of embryo transfer, predominantly towards the uterine fundus. The location and movement of the embryo flash at 60 min but not at 1 or 5 min is most predictive of clinical outcome. Acknowledgements The authors would like to thank the patients attending the Assisted Reproductive Unit of the Prince of Wales Hospital, Hong Kong, for their willingness to participate in this study. Authors roles S.H.S. contributed to the examination of the 3D US volumes, data analysis and drafting of the manuscript. A.W.Y.W. and G.W.S.K. contributed to the conception and design of the study and the embryo transfer procedure. C.P.S.C. performed the acquisition of the 3D volumes and raw data collection. L.P.C., C.H.S.C. and J.P.W.C. contributed to patient recruitment and data collection. T.C.L. supervised the overall project, contributing also to the data analysis and drafting of the manuscript. All authors contributed to the final production of the manuscript. Funding No external funding was sought or received for the completion of this study. Conflict of interest None declared. References Baba K, Ishihara O, Hayashi N, Saitoh M, Taya J, Kinoshita K. Where does the embryo implant after embryo transfer in humans? Fertil Steril 2000; 73:123 125. Buckett WM. A meta-analysis of ultrasound-guided versus clinical touch embryo transfer. Fertil Steril 2003;80:1037 1041. Cenksoy PO, Ficicioglu C, Yesiladali M, Akcin OA, Kaspar C. The importance of the length of uterine cavity, the position of the tip of the inner catheter and the distance between the fundal endometrial surface and the air bubbles as determinants of the pregnancy rate in IVF cycles. Eur J Obstet Gynecol Reprod Biol 2014;172:46 50. Confino E, Zhang J, Risquez F. Air bubble migration is a random event post embryo transfer. J Assist Reprod Genet 2007;24:223 226. Coroleu B, Barri PN, Carreras O, Martinez F, Parriego M, Hereter L, Parera N, Veiga A, Balasch J. The influence of the depth of embryo

596 Saravelos et al. replacement into the uterine cavity on implantation rates after IVF: a controlled, ultrasound-guided study. Hum Reprod 2002;17:341 346. Eskandar M, Abou-Setta AM, Almushait MA, El-Amin M, Mohmad SE. Ultrasound guidance during embryo transfer: a prospective, single-operator, randomized, controlled trial. Fertil Steril 2008;90:1187 1190. Frankfurter D, Silva CP, Mota F, Trimarchi JB, Keefe DL. The transfer point is a novel measure of embryo placement. Fertil Steril 2003;79:1416 1421. Frankfurter D, Trimarchi JB, Silva CP, Keefe DL. Middle to lower uterine segment embryo transfer improves implantation and pregnancy rates compared with fundal embryo transfer. Fertil Steril 2004;81:1273 1277. Friedman BE, Lathi RB, Henne MB, Fisher SL, Milki AA. The effect of air bubble position after blastocyst transfer on pregnancy rates in IVF cycles. Fertil Steril 2011;95:944 947. Gergely RZ, DeUgarte CM, Danzer H, Surrey M, Hill D, DeCherney AH. Three dimensional/four dimensional ultrasound-guided embryo transfer using the maximal implantation potential point. Fertil Steril 2005;84:500 503. Lambers MJ, Dogan E, Lens JW, Schats R, Hompes PG. The position of transferred air bubbles after embryo transfer is related to pregnancy rate. Fertil Steril 2007;88:68 73. Mains L, Van Voorhis BJ. Optimizing the technique of embryo transfer. Fertil Steril 2010;94:785 790. Oliveira JB, Martins AM, Baruffi RL, Mauri AL, Petersen CG, Felipe V, Contart P, Pontes A, Franco Junior JG. Increased implantation and pregnancy rates obtained by placing the tip of the transfer catheter in the central area of the endometrial cavity. Reprod Biomed Online 2004; 9:435 441. Pacchiarotti A, Mohamed MA, Micara G, Tranquilli D, Linari A, Espinola SM, Aragona C. The impact of the depth of embryo replacement on IVF outcome. J Assist Reprod Genet 2007;24:189 193. Tiras B, Polat M, Korucuoglu U, Zeyneloglu HB, Yarali H. Impact of embryo replacement depth on in vitro fertilization and embryo transfer outcomes. Fertil Steril 2010;94:1341 1345. Tiras B, Korucuoglu U, Polat M, Saltik A, Zeyneloglu HB, Yarali H. Effect of air bubble localization after transfer on embryo transfer outcomes. Eur J Obstet Gynecol Reprod Biol 2012;164:52 54. Zhu L, Xiao L, Che HS, Li YP, Liao JT. Uterine peristalsis exerts control over fluid migration after mock embryo transfer. Hum Reprod 2014; 29:279 285.