Research Summary

Fresh Versus Frozen Embryo Transfer in Low-Prognosis IVF

Miranda Manier, BA

Key Highlights

  • Live birth after first transfer was lower with freeze-all versus fresh (32% vs 40%; relative ratio [RR], 0.79; 95% CI, 0.65–0.94; P = .009).
  • Clinical pregnancy was lower with freeze-all (39% vs 47%; RR, 0.83; 95% CI, 0.71–0.97).
  • Cumulative live birth within 1 year was lower with freeze-all (44% vs 51%; RR, 0.86; 95% CI, 0.75–0.99; P = .04).
  • No between-group differences in birth weight, obstetric complications, or neonatal morbidities.

In a pragmatic, multicenter randomized controlled trial of women with a low prognosis for in vitro fertilization (IVF), a freeze-all strategy resulted in lower live birth rates than fresh embryo transfer. Clinical pregnancy and cumulative live birth within 1 year were also lower with freeze-all, while safety outcomes were similar between groups.

Women with a low prognosis—defined by nine or fewer oocytes retrieved or poor ovarian reserve—represent a substantial proportion of IVF cycles and face markedly reduced chances of live birth. Although elective freezing of all embryos can mitigate potential endometrial effects of ovarian stimulation and has shown comparable or better outcomes in women with normal/good prognosis, its value in low-prognosis patients has been uncertain.

This pragmatic trial enrolled 838 women at nine academic fertility centers in China and randomized them 1:1 on the day of oocyte retrieval to freeze-all (all embryos cryopreserved for later transfer) or fresh embryo transfer. The primary outcome was live birth, defined as delivery of a neonate with heartbeat and respiration at 28 or more weeks’ gestation after the first embryo transfer. Secondary outcomes included clinical pregnancy, pregnancy loss, multiple gestation, birth weight, maternal and neonatal complications, and cumulative live birth within one year. Analyses were intention-to-treat.

Live birth occurred in 32% (132/419) of the freeze-all group versus 40% (168/419) of the fresh group (RR, 0.79; 95% CI, 0.65–0.94; P = .009). Clinical pregnancy was 39% (164/419) with freeze-all versus 47% (197/419) with fresh (RR, 0.83; 95% CI, 0.71–0.97). The cumulative live birth rate within one year remained lower with freeze-all (44% [185/419]) than with fresh (51% [215/419]) (RR, 0.86; 95% CI, 0.75–0.99; P = .04). Twin live birth was less frequent with freeze-all (5% vs 9%; RR, 0.53; 95% CI, 0.31–0.89), and pregnancy loss was numerically higher (31% vs 23%; RR, 1.38; 95% CI, 1.00–1.90; P = .05). Birth weight, obstetric complications, neonatal complications, and congenital anomalies did not differ meaningfully between groups.

Study limitations include the pragmatic design with site-determined stimulation protocols, embryo stage/number for transfer, and endometrial preparation, which led to more single blastocyst transfers in the frozen group and more double cleavage-stage transfers in the fresh group. Protocol deviations and crossovers were more common in the frozen group, and some participants in that group delayed or did not complete transfer within 1 year—factors that, under intention-to-treat assumptions, count as no live birth.

“Fresh embryo transfer may be a better choice for women with low prognosis in terms of live birth rate compared with a freeze-all strategy,” Daimin Wei and colleagues concluded.

Reference:
Wei D, Sun Y, Zhao H, et al. Frozen versus fresh embryo transfer in women with low prognosis for in vitro fertilisation treatment: pragmatic, multicentre, randomised controlled trial. BMJ. 2025;388:e081474. doi:10.1136/bmj-2024-081474