Nucleation errors at four-cell stage reduces chances of live births in single euploid-blastocyst transfers

Abstract Study question What is the impact of nucleation errors (NE) and morphokinetic parameters on live birth rates in single euploid blastocyst transfers? Summary answer Nucleation errors (NE) at the four-cell stage was the only parameter associated with decreased live birth (LB) rates. What is known already Morphokinetic studies using time-lapse monitoring (TLM) systems have proposed algorithms to select embryos with the highest implantation potential. However, no single biomarker can reliably predict live-birth so far. Morphokinetics can deselect embryos for Preimplantation-Genetic-Testing for Aneuploidies (PGT-A). However, it is unclear whether morphokinetic markers can be used for selecting embryos with highest chance of LB among the euploid blastocysts. Performing morphokinetics enable detection of features such as nuclear errors (NE) which could be underdiagnosed with intermittent evaluation. It remains to be determined to which extend NE, next to morphokinetic parameters, are associated with LB potential of euploid blastocysts. Study design, size, duration Retrospective observational study including 383 single euploid blastocyst transfers in FET cycles between October 2017 to June 2021. These blastocysts were graded ≥BL3CC (Gardner criteria) and underwent TE biopsy for PGT-A by Next Generation Sequencing (NGS). Only ICSI with fresh autologous cycles using ejaculated sperm were considered. Outcomes of euploid blastocyst transfers were recorded and as pregnant (P): biochemical pregnancy (BP) (n = 17), clinical miscarriage (CM) (n = 35), live birth (LB) (n = 205); or not pregnant (NP) (n = 126). Participants/materials, setting, methods Patient baseline characteristics (Age, AMH, BMI) and infertility duration were considered. Morphokinetics were annotated with time-lapse-imaging (KID-score) for time (t) of PB2 (tPB2), tPNa (PN-appearance), tPNf (PN-fading), t2-t9, CP (compaction), tM (Morula), tSB (start blastulation), tB (blastocyst) and timings between each developmental event. Nuclear errors (NE) such as micronucleation, binucleation, multinucleation were annotated when present in at least one blastomere at two- and four-cell stages between 19.4-51.4 hours and 22.1-63.1 hours, respectively, post-insemination. Main results and the role of chance Median (IQR) values of Age and AMH of female partner, and infertility duration were similar among the P and NP groups. BMI was significantly higher in patients who miscarried with Median (IQR) values of 28.4kg/m2 (26.1-33.4) vs < 27.9kg/m2 in all other groups (P < 0.01). Significantly more blastocysts with type A inner cell mass quality resulted in LB (24.9%) than to CM (14.3%), BP (5.9%) and NP (13.5%) (P < 0.001). Concerning TLM morphokinetics evaluation, none of the timed parameters evaluated, neither the timings between the different morphokinetics events resulted in significant differences among the LB and the other P and NP groups. However, LB group had significantly less euploid embryos presenting NE at four-cell stage (7.8%) compared to CM (17.0%), BP (23.5%) and NP (17.5%) (P = 0.02). In a logistic regression model, NE when present at four-cell stage but not at two-cell stage, had a negative impact in LB rates (OR:0.39, CI: 0.2-0.72; P = 0.003). Limitations, reasons for caution The retrospective nature of the study and the fact that NE evaluation was performed manually based on TLM videos by a single observer. Wider implications of the findings NE-constituted embryos can give rise to euploid blastocysts and have an implantation potential. Since these embryos are associated with lower LB, it is relevant to include NE as an evaluation parameter to deprioritize multinucleated four-cell embryos for transfer, even when they result in euploid blastocysts. Trial registration number not applicable