Multicenter prospective non-selection study of blastocyst transfer with low-medium-grade mosaicism

To investigate the clinical outcome of blastocyst transfer with low-medium-grade mosaicism in a non-selection study Multicenter prospective non-selection study involving five IVF clinics in Italy. Trophectoderm biopsies showing intermediate chromosome copy number (CN) values consistent with low mosaicism (20-50%) were blindly reported as euploid. The presence of mosaicism did not influence the embryo selection process as unbiased comparison of transfer outcomes between the fully euploid and putative-mosaic groups was performed. Ethical committee approvals were obtained at each site. The study involved 783 patients (mean female age 37.50 + 3.3) undergoing homologous IVF cycles with blastocyst-stage PGT-A and single frozen euploid embryo transfer (SEET) between Sept 2018 and Dec 2019. A total of 845 cycles with at least one embryo available for transfer (euploid or putative mosaic) were included. Main exclusion condition was blastocyst of the worst morphological class. Miscarriage rate and live birth rate (LBR) per transfer were the main outcome measures. A total of 897 SEET of tested blastocysts were performed in the course of the study, 484 (54%) were euploid, 282 (31,4%) showed low putative-mosaicism (20%-30%), and 131 (14,6%) showed moderate putative-mosaicism (30%-50%). Miscarriage rate was 12% (n=29/241), 11% (n=15/136) and 12.7% (n=8/63) for the three groups respectively; LBR was 43.4% (n=210/484), 42.9% (n=121/282) and 42% (n=55/131), respectively. Logistic regression analysis confirmed the lack of association between the “PGT-A category” and miscarriage risk (P=0.71) and live-birth rate per transfer. Therefore, no differences were observed in the clinical parameters testing across the three categories tested. The only prognostic factor for LBR was blastocyst developmental timing (OR=0.67 for day 6 vs day 5 and OR=0.19 for day 7 vs day 5). Mean mosaicism rate was not increased in biochemical pregnancy losses (BPL) and embryos resulting in miscarriage vs ongoing pregnancies, and in embryos resulting in LB vs embryos failing to implant. Mean number of mosaic chromosomes was also similar between LBR and control. The number of chromosomes with intermediate CN values per embryo was not associated with BPL (OR=1.00; 95%CI:0.85-1.18), miscarriage (OR=1.05; 95%CI=0.89-1.23) or LBR outcome (OR=0.98; 95%CI 0.91-1.07). Our data shows that the exclusion from transfer of all putative mosaics above 20% variability, results in an overall relative reduction in expected cumulative LBR of -36%. Intermediate chromosome CN values consistent with low/moderate-grade mosaicism (30-50%) do not provide any clinically useful criteria for defining embryonic reproductive competence and should not be considered for aneuploidy categorization and embryo selection purposes in PGT-A cycles. The previously reported mild association can be explained by a selection bias involving the transfer of putative mosaic embryos to patients that failed previous euploid embryo transfers, thus of lower prognosis.