Cleavage stage versus blastocyst stage transfers in patients with a single zygote: an emulated target trial.

内容摘要

Is the live birth rate higher for cleavage stage embryos compared to blastocysts in patients with a single zygote following oocyte fertilization using IVF or ICSI? Among patients with only a single zygote, transfer at cleavage stage was observed to result in a higher live birth rate than transfer at blastocyst stage. Existing evidence suggests that blastocyst transfer is superior to cleavage stage in terms of live birth rate per embryo transfer, cumulative live birth rate, and time to pregnancy with three or more zygotes. However, whether these findings generalize to cohorts with less than three zygotes remains unclear. This target trial emulation, with live birth as the primary outcome, involved a retrospective analysis of 11 163 nulliparous patients who undertook ART in Australia and New Zealand between 2009 and 2022. Participants were included in the study if they were undergoing their first-ever stimulated ART cycle resulting in a single zygote following fertilization using IVF or ICSI. In this cohort, there were 6505 patients who received a cleavage stage transfer, 2216 who received a blastocyst stage transfer, and 2442 who had no embryo available for transfer following embryo culture, with the intended day of transfer unknown for these patients. We modelled a comparison of intended transfer of a cleavage or blastocyst using g-computation within the target trial emulation framework. This involved fitting models that estimated the chance that embryo would survive to cleavage or blastocyst stage based on patient characteristics, and if it did so, the chance of a live birth when it is transferred. These models were used to simulate an idealized randomized controlled trial (target trial) on our retrospective cohort. We found that compared to blastocyst transfers, cleavage stage transfers were associated with a higher live birth rate per couple in this cohort (12.5% vs 10.1%), with an adjusted relative risk of 1.24 (95% CI: 1.15-1.50). This effect increased with female age from 35 years, with a relative risk of 1.34 (95% CI: 1.15-1.57) in a 35-year-old and 1.51 (95% CI: 1.25-1.80) in a 40-year-old woman. This is likely due to the high rate of embryo attrition between the cleavage and blastocyst stages; our models predict that on average 92.0% of zygotes would survive to cleavage stage, compared to 58.9% and 49.9% surviving to the blastocyst stage at ages 35 and 40 years. The originally intended/planned day of embryo transfer, i.e. cleavage stage or blastocyst, is not recorded in the data source, only the actual day of transfer. This required utilization of a multinomial mixture model that estimates the cleavage and blastocyst embryo development rates as a sub-model, using the year of treatment as an external source of variation (in additional to patient factors) for predicting the intended treatment group. As with any causal analysis using retrospectively collected observational data, the results are dependent on the accuracy of our modelling assumptions which cannot be verified. Additionally, the potential confounder of embryo quality on the day of transfer was not available. These results highlight the role target trial emulation can play in filling evidence gaps for patient cohorts excluded from existing RCTs, and where the prospect for future RCTs is limited due to sample size constraints or ethical considerations. Further, the assessment of new ART technologies and procedures needs to be stratified by markers of patient prognosis (in this case, female age and number of available zygotes), and there should be caution in generalizing the findings to a different group. This study was funded by the 2024 Ferring Australia Reproductive Medicine Research Grant Scheme. O.F., G.M.C., and L.R. are the listed investigators on this grant paid to UNSW Sydney. Ferring had no role in designing, analysing, interpreting, or reviewing the study. W.L. declares that they are a Human Reproduction Deputy Editor. L.R. declares consulting fees from Besins, Merck, and Organon, speaker's fees from Besins, travel support from Gedeon Richter, and shares in Monash IVF Group (ASX: MVF). C.V. declares receipt of gift vouchers as honorarium for consumer input to the study as part of their membership of UNSW YourIVFSuccess/National Perinatal Epidemiology and the Statistics Unit Consumer Advisory Group. The remaining authors have no conflicts of interest to declare. N/A.