Objectives: To compare the efficacy in embryo selection based on morphology alone compared to morphology and non-invasive preimplantation genetic testing for aneuploidy (niPGT-A) in women with recurrent pregnancy loss (RPL) undergoing in vitro fertilization (IVF).

Hypothesis to be tested: The embryo selection based on morphology and niPGT-A results in a lower miscarriage rate and a higher live birth rate in IVF as compared with that based on morphology alone.

Design and subjects: Randomized double-blind randomized controlled trial. Women with RPL undergoing IVF will be enrolled.

Interventions: Spent culture medium (SCM) of each blastocyst will be frozen individually. They will be randomly allocated into two groups: (1) the intervention group based on morphology and niPGT-A and (2) the control group based on morphology alone.

In the control group, blastocysts with the best quality morphology will be replaced first. In the intervention group, blastocysts with the best morphology and euploid result of SCM will be replaced first.

Main outcome measures: The primary outcome is the miscarriage rate per the first embryo transfer.

Data analysis: Comparison of quantitative variables will be performed using Student's t, while categorical variables will be compared using a Chi-square analysis. All statistical analyses will be performed with the intention to treat and per protocol, and a p-value <0.05 will be considered statistically significant.

Expected outcome results: The embryo selection based on morphology and niPGT-A results in a lower miscarriage rate and a higher live birth rate in IVF as compared with the control group based on morphology alone.

Trial Objectives and Purpose The primary objective of this randomized double blind controlled trial is to compare the efficacy in embryo selection based on morphology alone versus morphology and niPGT-A in women with RPL undergoing the first frozen embryo transfer. The secondary objectives are to evaluate the impact of eNK cells on the miscarriage and live birth rates and the prediction of live birth using spheroid/BAP-EB attachment assay.

Main hypotheses to be tested:

1. The embryo selection based on morphology and niPGT-A for aneuploidy results in a higher live birth rate of IVF in women with RPL as compared with the control group based on morphology alone.
2. The embryo selection based on morphology and niPGT-A for aneuploidy results in a lower miscarriage rate following IVF in women with RPL as compared with the control group based on morphology alone.

2. Trial Design This is a randomized double blind controlled trial. Eligible women will be recruited for the study and informed written consent will be obtained after counseling.

Endometrial assessment All women will have an endometrial biopsy using a Pipelle sampler 7 days after luteinizing hormone surge (LH+7) prior to the month of having IVF. Part of the endometrial samples will be fixed in paraformaldehyde for immunohistochemical staining of endometrial uNK cells with CD56 antibody [26,27]. The other part of the endometrial samples will be used for epithelial and stromal cell isolation for spheroid/BAP-EB attachment assay [28], to predict the embryo attachment rate in a laboratory setting. Due to ethical issues and practical difficulty, the use of human embryos for endometrial assessment is not feasible. BAP-EB model [28] will be used as the embryo (blastocyst) surrogate in this study. BAP-EB spheroid model are differentiated from human embryonic stem cells. The BAP-EB after 72h of differentiation (BAP-EB-72h) have molecular signature of Day 7 trophectoderm cells of blastocysts [28]. In this study, BAP-EB will be co-cultured with primary endometrial epithelial cells and the attachment rate will be determined according to our established protocol.

IVF protocol Women will undergo IVF with or without intracytoplasmic sperm injection (ICSI) as indicated. The women will receive ovarian stimulation using the progestin-primed protocol [29] or the antagonist protocol. Ultrasound scanning will be arranged on day 2-3 of menses for the antral follicle counts and to exclude the presence of ovarian cyst. Serum oestradiol and progesterone concentration will be checked and if they are basal, ovarian stimulation with gonadotropin injections (225-300 IU daily depending on the antral follicle count) for 10-12 day is given. Provera 10mg daily will be given from day 2-3 to day of trigger in the progestin-primed protocol to prevent premature LH surge while GnRH antagonist will be started on day 6 of ovarian stimulation to day of trigger. Regular ultrasound monitoring will be performed to monitor the growth of follicles. Adjustment of the gonadotropin dosage is corresponding to the number and size of follicles. When three follicles reach >17 mm in diameter, GnRH agonist (Decapeptyl 0.3mg) or human chorionic gonadotrophin (Ovidrel 0.25mg) will be administered. Oocyte retrieval will be scheduled 34-36 hours after the HCG or agonist trigger under transvaginal ultrasound guidance.

Normal fertilization will be assessed and confirmed by the presence of two pronuclei at 16-18 h after insemination. All cleavage stage embryos will be grown individually to the blastocyst stage, usually day 5 or 6 after oocyte retrieval, in a monophasic medium. On Day 3, the culture medium will be replenished, and culture will be continued at 37oC and 6% CO2 in reduced oxygen tension (5%). No fresh transfer of blastocysts will be performed in the stimulated cycle.

Grading of blastocyst by morphology

Blastocysts are graded according to Gardner's classification (29). The blastocyst grading system assigns 3 separate quality scores to each blastocyst embryo, based on the followings:

1. Blastocyst development stage: expansion and hatching status
2. Inner cell mass score
3. Trophectoderm score

Expansion grade Blastocyst development and stage status

1. Blastocoel cavity less than half the volume of the embryo
2. Blastocoel cavity more than half the volume of the embryo
3. Full blastocyst, cavity completely filling the embryo
4. Expanded blastocyst, cavity larger than the embryo, with thinning of the zona pellucida
5. Hatching out of the zona pellucida
6. Hatched out of the zona pellucida

Inner cell mass grade Inner cell mass quality A Many cells, tightly packed B Several cells, loosely grouped C Very few cells

Trophectoderm grade Trophectoderm quality A Many cells, forming a cohesive layer B Few cells, forming a loose epithelium C Very few large cells

Blastocysts are cryopreserved at developmental stage with expansion score 3 or above. If a blastocyst does not reach expansion stage 3 on day 5, it will be assessed on day 6 whether it is suitable for cryopreservation. Blastocysts with either inner cell mass or trophectoderm scored as B or above are regarded as utilizable blastocysts.

Each blastocyst will be frozen by vitrification individually and its SCM (~8 µl) will be frozen at -800C separately and individually. The embryologist will prepare a sequence of blastocyst transfer based on the best morphology by Gardner's criteria.

Then, on the day of blastocyst freezing, women will then be randomly assigned into two groups in 1:1 ratio using a randomization program by the laboratory staff in the PGT laboratory.

1. the intervention group using morphology and niPGT-A and
2. the control group based on morphology alone.

The women and clinicians will be blinded to the treatment groups they are assigned. Only the laboratory staff in the PGT laboratory will be aware of the group assignment.

niPGT-A of spent culture medium (SCM) In the intervention group, comprehensive chromosome screening using NGS will be performed according to the recommendations of the company in all SCM samples. In the control group, the measurement will be done retrospectively in those SCM samples of blastocysts that are replaced in the first transfer. All SCM samples will be saved for possible future research.

A commercially available NI-PGT kit (PG-Seq Rapid Non-Invasive PGT kit, PerkinElmer) will be used to analyse SCM samples. The protocol has been previously optimized with non-invasive samples from 15 laboratories around the world. The kit follows a single tube workflow, two-steps PCR to whole genome amplify the DNA in SCM and then attaches indexes and sequence-specific adapters to template DNA, resulting in sequencing ready samples.

After purification, concentration of each sample is adjusted into equal molar, pooled (96 samples) and then sequenced on a MiSeq system (Illumina) at 1x75 bp read length. On-board secondary analysis was performed automatically by the MiSeq Reporter (Illumina) followed by PG-Find Software (v 1.0, PerkinElmer). Reads aligning to anomalous, unstructured and highly repetitive sequences are filtered from the analysis. A target bin size of 1,000 kb is used, giving a minimum resolution of 10Mb. All genomic positions refer to the human genome build NCBI 37.

According to the default setting of PG-Find software, classification of aneuploidy is determined by CNV (copy umber variation) value. CNV value >2.7 is considered as gain while CNV value <1.3 is considered as loss. Sample will be concluded as non-euploid when one or more of the chromosomes shows gain/loss.

The niPGT-A report of the SCM sample can be euploid, non-euploid and non-informative. It is used only to prioritize the sequence of embryo transfer. Blastocysts with non-euploid result in the niPGT-A report will not be discarded.

Blinding The embryologist will grade the morphology of blastocysts according to Gardner's criteria stated above and enter the grading of blastocysts into an online database, which will be managed by an IT technician. The laboratory staff in the PGT laboratory will enter the PGT result into a local database when the NIPGT results are available. The IT technician will collect the database from the PGT laboratory and enter it into the online software to compile the sequence of embryo transfer according to a pre-determined algorithm which depends on the day of blastocyst development (day 5 better than day 6), blastocyst morphology and niPGT-A result. The IT technician will issue the sequence of embryo transfer which does not contain information on the grading of the blastocyst and the NIPGT result to the embryologists in the IVF laboratory. Therefore, the subjects recruited, the clinicians and the embryologists will be blinded to the group allocation.

Our preliminary results of niPGT-A Preliminary results of the ongoing NIPGT-A trial involved 1168 SCM. Media cultured in parallel but without contact with embryos were collected as controls (n=238). Amplification was successful in 1158 SCM (99.1%, 1158/1168) and 1141 SCM resulted in conclusive result (97.6%, 1141/1168). All controls showed no amplification.

Frozen embryo transfer (FET) Blastocysts can be replaced in the subsequent natural, letrozole or hormonal replacement cycles, depending on whether the women have regular menstrual cycles or not. Only one blastocyst will be transferred each time. In the control group, blastocysts with the best quality morphology will be replaced first and the sequence of blastocyst transfer is decided prior to randomization. In the intervention group, blastocysts with the best morphology and euploid result will be replaced first as the sequence of blastocyst transfer will be modified after the niPGT-A reports are available.

Pregnancy A urine pregnancy test will be performed 14 days after the transfer. If the pregnancy test is positive, transvaginal ultrasound will be performed two weeks later to locate the pregnancy and confirm foetal viability and the number of fetuses. Subsequent management will be the same as other women with early pregnancy. They will be referred for antenatal care when the ongoing pregnancy is 8-10 weeks.

Follow-up Written consent regarding retrieval of pregnancy and delivery data will be sought from the women at the time of study as in all patients coming for IVF for infertility. The women will be contacted after delivery by phone to retrieve the information of the pregnancy outcomes as a routine after IVF pregnancy. The outcome of the pregnancy (delivery, miscarriage), number of babies born, birth weights and obstetrics complications will be recorded.

Women who fail to get pregnant in the first FET will be unblinded and an euploid embryo will be replaced if available. If no euploid embryo, the women can opt to undergo the second IVF cycle.

Inclusion Criteria:

1. Women aged less than 40 years at the time of ovarian stimulation and
2. >=two spontaneous miscarriages in the first trimester
3. Unexplained recurrent pregnancy loss after standard investigation
4. At least one blastocyst available on day 5 or 6 after the retrieval.

Exclusion Criteria:

1. Women undergoing PGT for monogenic diseases or structural rearrangement of chromosomes;
2. Use of donor oocytes;
3. Hydrosalpinx shown on pelvic scanning and not surgically treated
4. Uterine anomalies distorting the uterine cavity in three dimensional ultrasound
5. No usable blastocysts on day 5 or 6 after the retrieval

The embryologist will grade the morphology of blastocysts according to Gardner's criteria stated above and enter the grading of blastocysts into an online database, which will be managed by an IT technician. The laboratory staff in the PGT laboratory will enter the PGT result into a local database when the NIPGT results are available. The IT technician will collect the database from the PGT laboratory and enter it into the online software to compile the sequence of embryo transfer according to a pre-determined algorithm which depends on the day of blastocyst development (day 5 better than day 6), blastocyst morphology and niPGT-A result. The IT technician will issue the sequence of embryo transfer which does not contain information on the grading of the blastocyst and the NIPGT result to the embryologists in the IVF laboratory. Therefore, the subjects recruited, the clinicians and the embryologists will be blinded to the group allocation.