P-744 Reporting chromosomal mosaicism reduces the overall accuracy of preimplantation genetic testing for aneuploidies: results from the extended in vitro culture of 230 human embryos

Abstract

Abstract Study question What is the accuracy of reporting chromosomal mosaicism in the context of next generation sequencing (NGS)-based preimplantation genetic testing for aneuploidies (PGT-A)? Summary answer Extended in vitro culture of human embryos demonstrated that reporting mosaicism increased the risk of false positives, reducing the overall accuracy of PGT-A to 70%. What is known already PGT-A has been introduced into IVF practice to improve clinical outcomes. However, diagnostic accuracy remains controversial, particularly regarding chromosomal mosaicism. While NGS allows the detection and reporting of mosaicism from a trophectoderm (TE) biopsy, the inability to discern biological variability from technical artefacts continues to limit the interpretation of PGT-A results. As reports of mosaicism often have a major impact on patient management, achieving an appropriate level of standardization will be vital to avoid overestimation. Extended in vitro culture of human embryos up to 12 days post fertilization (D12) delivers an important, clinically relevant platform for the validation of PGT-A. Study design, size, duration A total of 230 clinically unsuitable blastocysts were included in the study. Embryos were selected based on their original diagnosis following PGT-A as: uniformly abnormal (n = 158), aneuploid and mosaic (n = 21), mosaic only (n = 36) and euploid (n = 15) but carrying a monogenic disease mutation in homozygosis. Blastocysts were originally classified as mosaic if they contained 30-70% abnormal cells, with <50% mosaicism defined as low level and >50% considered high level mosaicism. Participants/materials, setting, methods Vitrified blastocysts were warmed and cultured to D12 using established protocols to generate embryo outgrowths. A total of 90 outgrowths were selected for NGS. Outgrowths were collected whole (n = 64) or separated into 2-4 portions (n = 26), which were then analyzed individually. We correlated the original PGT-A diagnoses to developmental outcomes and the chromosomal status of the embryo outgrowths. Fisher’s Exact test (two-sided) was used for evaluating associations. All p-values <0.05 were considered significant. Main results and the role of chance Following extended culture, 49.1% (113/230) of embryos remained viable and attached, while 50.9% (117/230) degenerated and detached. As previously observed in our model, euploid blastocysts and embryos diagnosed with trisomies, duplications or chromosomal mosaicism were significantly more likely to attach throughout the extended culture, while monosomies, deletions and complex chromosomal constitutions impaired in vitro development (total attached, 78.3% vs. 24.8%, p < 0.0001). When compared to the original biopsy, we established 100% concordance when reporting euploidy and uniform whole chromosome aneuploidies, demonstrating an overall high sensitivity of PGT-A. Of the blastocysts diagnosed as mosaic only, 80.6% (29/36) remained viable throughout the extended culture. These were sequenced either whole (n = 11) or in parts (n = 18). At D12, 69.0% (20/29) of these embryos were uniformly euploid, while 27.6% (8/29) were uniformly aneuploid across all outgrowth samples. Notably, the latter were all originally diagnosed as high level mosaics. One embryo (3.4%) revealed a true mosaic configuration, with evidence of a reciprocal event across the different outgrowth samples. Ultimately, the proportion of embryos accurately diagnosed as euploid was 25.0% (8/35), with an overall false positive error rate of 77.1%. False positives were largely attributed to the diagnosis of mosaicism, 85.2% (23/27) or segmental aberrations, 14.8% (4/27). Limitations, reasons for caution This study was performed using a single PGT-A assay and cannot be extrapolated to other platforms. To evaluate the predictive value of chromosomal mosaicism, we selected embryos based on their chromosomal profiles. This may underestimate the accuracy of PGT-A, as sampling of embryos was not performed randomly. Wider implications of the findings We confirm the high sensitivity of PGT-A, as euploidy and uniform whole chromosome abnormalities were detected with high accuracy. However, predictive value is considerably reduced when diagnosing mosaicism. Given the high false positive rate, reporting mosaicism remains difficult to justify, as potentially viable embryos continue to be excluded for transfer. Trial registration number NA