Abstract Study question Does the choice of PGT-A provider have any impact on clinical results, such as rates of aneuploidy, pregnancy and miscarriage? Summary answer The laboratory providing PGT can have a significant impact on various important clinical outcomes. PGT-A providers are not equivalent and should be chosen with care. What is known already A growing number of IVF cycles include PGT-A to assist in the identification of euploid embryos for transfer to the uterus. All modern PGT laboratories utilise a method called next generation sequencing (NGS) to predict the copy number of each chromosome in embryo biopsy samples. Given this technical convergence, it is often supposed that genetic results will be similar regardless which PGT laboratory is used, and consequently a clinic choosing a genetics provider need only consider convenience, quality of support and price. Here we consider whether the choice of PGT-A provider might have more profound affects, potentially impacting clinical results. Study design, size, duration A large network of IVF clinics switched from PGT-A provider ‘A’ to provider ‘B’. The final 6 months of clinical data using provider A was compared to 6 months of data after the switch to B. There were no changes in any aspect of patient population, treatment protocols, or embryological practice occurred between the two time periods evaluated. Genetic results (aneuploidy/euploidy rates) and clinical outcomes (implantation, miscarriage, ongoing pregnancy/birth) were compared to identify any differences. Participants/materials, setting, methods Within the two time periods considered, 9,091 embryos underwent PGT-A with provider A and 10,281 using B. The average female age was 39.0 and 39.3, respectively. Embryos were biopsied at the blastocyst stage and the cells were shipped to the genetics laboratories. Provider A was in the same country as the IVF clinics, while provider B was located overseas. Both PGT-A laboratories used NGS, but employed different methods for DNA amplification, sequencing and data analysis. Main results and the role of chance A higher proportion of embryos were classified euploid by provider B in comparison to A (49.3% versus 39.8%; p < 0.0001). Differences in reported aneuploidy were particularly striking for younger patients (<38 years; 67.5% euploid according to B versus 52.5% reported by A; p < 0.0001). For patients over 37 years, the lower aneuploidy frequency reported by B resulted in fewer cycles with all embryos classified ‘abnormal’ and more cycles with a transfer (52.2% of 2115 cycles in this age group had a transfer, compared to 48.6% of 1915 cycles for A; p = 0.02). Having more embryos classified euploid also benefitted younger patients, providing opportunities to combine morphological selection with genetic evaluation. Improved selection may explain a higher ongoing pregnancy rate observed for patients <38 years (63.2% ongoing after the first embryo transfer for B versus 55.3% for A). Differing miscarriage rates were also observed (18.1% when using A versus 15.1% for B; p = 0.047). We speculate that higher rates of euploidy and pregnancy when using provider B might be a consequence of viable embryos being excluded due to incorrect classification as ‘abnormal’ by A. Additionally, failure to detect some aneuploidies, leading to transfer of abnormal embryos, might explain the higher miscarriage rate associated with A. Limitations, reasons for caution There were no detectable differences in the composition of patient populations receiving PGT-A from the two providers, and no apparent changes to protocols used during the two time periods assessed. Nevertheless, it must be acknowledged that the conclusions are based on a retrospective review of data, not a controlled study. Wider implications of the findings PGT-A and NGS are umbrella terms encompassing multiple methods with widely varying levels of validation and accuracy. Our results suggest that the choice of PGT-A provider has important implications for clinical results, affecting the number of transferrable embryos, pregnancy, and miscarriage rates. The observed differences likely reflect differing PGT-A accuracies. Trial registration number Not applicable
Read full abstract