Abstract
Study question
Is oocyte freezing a risk factor for monozygotic splitting?
Summary answer
There is a trend towards a higher monozygotic splitting incidence among frozen oocytes, but this did not reach statistical significance.
What is known already
Laboratory techniques which involve embryo manipulation such as ICSI, assisted hatching, embryo biopsy for pre-implantation genetic testing and extended culture to the blastocyst stage appear to increase the risk of monozygotic splitting. Whilst there is some data that embryo freezing does not appear to increase the risk of monozygotic splitting, there is no comparable analysis on whether oocyte freezing increases the risk of monozygotic splitting.
Study design, size, duration
This was a retrospective cohort study analysing 988 015 ART (assisted reproductive technique) cycles from the HFEA anonymised database from 1990 to 2016. As frozen oocytes require ICSI, only fresh oocytes with ICSI were taken for comparison and frozen embryo transfers were excluded. Only single embryo transfers were included.[CM1] [MM2] We also noted ages of the female partner at the time of treatment, stage of embryo transfer, and whether pre-implantation genetic testing had been performed.
Participants/materials, setting, methods
There were 84 085 ICSI cycles with single embryo transfers using fresh oocytes and 596 using frozen oocytes. Monozygotic splitting was defined as the presence of two foetal hearts [CM1] [MM2] on ultrasound. Live birth (LB)was defined as either a singleton or a twin LB resulting from a monozygotically split embryo. Preterm birth (PTB) was defined as birth prior to 37 weeks gestation and early PTB as birth prior to 32 weeks gestation.
Main results and the role of chance
The frozen oocyte group had fewer women in the under–35 age group (frozen oocytes 16.6% vs fresh oocytes 53.6%, p < 0.0001) and a higher proportion of blastocyst transfers ( frozen oocytes 55.1% vs fresh oocytes 48.8%, p = 0.002) There were only 10 PGT cycles amongst monozygotically split embryos from fresh oocytes in our analysis, and none in the frozen oocyte group. Hence, this was not included as a confounder. There was a non-significant trend toward a higher incidence of monozygotic splitting amongst frozen oocytes (4/596, 2.3%, all monozygotic twins) than amongst fresh oocytes (378/27 019, 1.4%, 372 monozygotic twins and 6 monozygotic triplets); OR 1.688, 95% CI 0.623 to 4.574 and aOR 1.506, 95% CI 0.531 to 4.274 (maternal age and stage of embryo transfer adjusted as confounders). Of the 378 monozygotically split embryos from fresh oocytes, 308 (81.5%) had a LB: of which 47 (15.3%) were singletons and the rest were twins; 241 (78.2%) were PTB and 56 (18.2%) were early PTB. Of the four monozygotic twins from frozen oocytes, all reached a LB; one was a singleton term LB (Birthweight 3–3.5kg) whilst three were twin preterm LBs at 35–36 weeks, with no early PTBs and twin median birthweight 2–2.5 kg.
Limitations, reasons for caution
Albeit a large national database, this cohort study was restricted due to absence of data on potential confounders such as age at oocyte freezing, method of cryopreservation and length of storage.[CM1] Data was also lacking on amnionicity, obstetric risks including pre-eclampsia, twin-to-twin-transfusion syndrome, intrapartum and late effects.
Wider implications of the findings: With rapid rise in egg freezing, our findings would help reassure women that eggs on ice does not predispose to significant risk of two-in-one monozygotic splitting. However, the marginal trend (from 1.4% in fresh to 2.3% in frozen oocytes), does indicate that this subject merits further research.
Trial registration number
Not applicable. A database based retrospective study