P-308 Abnormal Oocytes are more likely to lead to abnormal embryo divisions (Direct Unequal Cleavage,DUC), but do not compromise embryo quality as assessed using CHLOE-EQ score

Abstract

Abstract Study question Do oocyte dysmorphisms lead to abnormal embryo divisions and compromised embryo quality? Summary answer Oocyte cytoplasmic abnormalities (granularity, Smooth Endoplasmic Reticulum,SER) did not affect CHLOE-EQ score; whilst zona abnormalities (thickness and unevenness) and SER tend to lead to DUCs. What is known already Oocyte dysmorphisms include extracytoplasmic [zona pellucida (ZP) evenness and thickness] and cytoplasmic abnormalities [SERs, inclusions, darkness, granularity]. The impact of these abnormalities on embryo development and viability as reported in the literature is contradictory. CHLOE-EQ score is an Artificial Intelligence (AI) based algorithm designed to support embryologists in assessing embryo viability, and has previously been demonstrated to automatically detect embryo development anomalies (such as DUCs), to be predictive of blastulation, utilisation, selection for transfer, ploidy, implantation and live birth. Therefore, CHLOE-EQ is a metric of embryo viability. The impact of oocyte dysmorphisms on CHLOE-EQ and DUCs is poorly understood. Study design, size, duration Retrospective cohort analysis of 742 embryo time-lapse videos, cultured at a private fertility clinic between June and July 2022. Participants/materials, setting, methods The clinic provided annotations on extracytoplasmic abnormalities (ZP thickness and uniformity) and cytoplasmic abnormalities [SERs, inclusions, darkness, granularity]. CHLOE-EQ (Fairtility) automatically annotated morphokinetics and DUCs and further quantified embryo viability scores (CHLOE-EQ and Blast Score). Main results and the role of chance CHLOE-EQ score was not affected by the oocyte having cytoplasmic abnormalities (no vs yes: 4.2 ± 4, n = 122vs4.1 ± 4, n = 359, NS): dark (4.1 ± 4, n = 476vs3.1 ± 4, n = 5, NS), granular (4.1 ± 4, n = 179vs4.1 ± 4, n = 302, NS), SER (4.1 ± 4, n = 455 vs 3.8 ± 4, n = 26, NS), inclusion (4.1 ± 4, n = 430vs4.1 ± 4, n = 51, NS); or ZP abnormalities [overall (4.2 ± 4, n = 379vs3.7 ± 4, n = 102, NS), non-uniformity (4.1 ± 4, n = 409vs4.2 ± 4, n = 72, NS), thick ZP (4.2 ± 4, n = 457vs2.8 ± 4, n = 24, NS), thin ZP(4.1 ± 4, n = 475vs2.1 ± 4, n = 6, NS]. DUC embryos were two times more likely to be derived from oocytes with thick ZP (9/98, 9%oocytes) than oocytes without thick ZP (15/383, 3.9%, p = 0.03). DUCs were more likely to have a non-uniform ZP compared to non-DUCs [DUCs: 7%(7/98) vs Non-DUCs: 17%(65/383), p = 0.015]. DUCs were not associated with the following oocyte cytoplasmic dysmorphias: SER (DUC vs Non-DUCs: 6/98,6% vs 20/388,5%,NS), dark (0/98 vs 5/383, NS), granular (61/98 vs 241vs383, NS), inclusions (11/98 vs 40/383, NS). DUCs had lower blastulation rate than non-DUCs [DUC:1.8%(2/113) vs Non-DUCs:77%(298/389), p < 0.001]. DUCs were 4-fold less likely to be multinucleated at the 2 cell stage than non-DUCs [DUC: 7%(2/29)vsNon-DUCs: 30%(91/305), p = 0.03]. DUCs were 7-fold more likely to be multinucleated at the 4 cell stage than non-DUCs [DUC: 7%(2/29)vs Non-DUCs:1%(3/302), p = 0.06]. Patient age was not associated with DUCs (DUCs: 36.9 ± 4 vs non-DUCs:37.1 ± 4, NS). Limitations, reasons for caution This was a retrospective-single clinic study. Causality is not determined. Wider implications of the findings Given the growing evidence that DUCs have compromised viability, it is important to understand the biology of how DUCs are connected to oocyte quality. Using AI to detect DUCs to avoid critical information being missed during embryo assessment can assist embryologists in maximising their efficacy of embryo selection. Trial registration number NA