18. THE INSIGHTS OF EMBRYO MOSAICISM IN IVF CYCLES

Introduction The development of new methodologies for PGT-A has led to the detection of a certain percentage of mosaic embryos. It has been postulated that this could be a normal feature in human embryos. To date, the real incidence of mosaicism in blastocysts is unknown as illustrated by the very variable incidences reported between groups. Certain factors could influence mosaicism occurrence such as patients’ characteristics, ovarian stimulation and in vitro culture conditions. Biopsy sample manipulation could also induce artefactual mosaicism. Moreover, different diagnostic platforms, algorithms and detection limits are used by different groups, thus making comparisons difficult. Material & methods The cytogenetic results of 1339 blastocysts obtained from PGT-A cycles performed in our centre from September 2017 to January 2019 were retrospectively analysed. Embryos were biopsied at the blastocyst stage between day 5 and 7 of development and genetic analysis was performed by next-generation sequencing (NGS) according to VeriSeq™-PGS protocols (Illumina). Embryos were diagnosed as euploid, aneuploid or mosaic. An embryo was classified as mosaic when the percentage of affected cells ranged from 30-70%. Both the incidence of mosaicism in this cohort and the specific chromosomes affected were studied. Mosaicism was studied versus patients’ factors (age, indication for PGT-A, ovarian reserve), embryonic factors (blastocyst development and quality), potential iatrogenic factors (ovarian stimulation treatment and the use of two different single-step culture media), and the biopsy operator as a proposed source of artefactual mosaicism. A logistic mixed multivariable model was applied to estimate the odds for every endpoint. Patient and biopsy procedure were treated as random factor to control the correlated observations effect. Results The overall incidence of mosaic embryos was 13.0% (7.5% of euploid-aneuploid mosaic embryos and 5.5% of aneuploid-aneuploid mosaic embryos). Most mosaic embryos showed just one chromosome affected (85.5%). No chromosomes were especially associated with mosaic aneuploidy. However, segmental mosaicism was more common in larger chromosomes. No association could be stablished between embryo mosaicism and the factors studied. Neither maternal nor paternal ages showed any association with mosaicism (ORa=0.96(0.88-1.04); ORa=1.04(0.99-1.09)). The antral follicle count did not show any correlation either (ORa=0.987(0.95-1.02)) nor the indication for PGT-A. In terms of embryo development, no associations were found either: ORa for blastocyst formation in day 5 vs day 7 was 0.72(0.31-1.68); ORa for blastocyst formation in day 6 vs day 7 was 0.53(0.23-1.22); ORa for excellent and good quality embryos vs fair and poor quality embryos was 0.86(0.54-1.36). Regarding the studied factors that were potentially associated with iatrogenic mosaicism, neither type of stimulation (agonists vs antagonists), nor the use of two different single-step media showed an association. No impact on the incidence on mosaicism has been related to a given operator. Conclusions There is no preferential distribution of mosaicism among the chromosomes although large chromosomes seem to be more sensitive to segmental mosaicism. None of the studied potential sources of intrinsic, iatrogenic, or artefactual mosaicism showed an association with such feature.