Previous studies have shown that repeated intracytoplasmic sperm injection (ICSI) failures can be caused by a paternal effect. Other studies have suggested that ICSI results are compromised if morphologically abnormal spermatozoa are injected into oocytes. This study was undertaken to evaluate the usefulness of a high-magnification optical system to select spermatozoa to be used for ICSI (high-magnification ICSI) in couples with repeated conventional ICSI failures. Couples with two or more previous conventional ICSI failures underwent an additional conventional ICSI attempt, followed by a high-magnification ICSI attempt. The outcomes of the two sequential attempts were compared. In 72 of these patients, sperm DNA integrity was assessed. In the whole group of 125 couples with repeated ICSI failures, high-magnification ICSI improved clinical outcomes (pregnancy, implantation, delivery and birth rates) without affecting biological outcomes (fertilization and cleavage rates, embryo morphology). The improvement of clinical ICSI outcomes was evident both in patients with an elevated degree of sperm DNA fragmentation and in those with normal sperm DNA status. It is concluded that high-magnification ICSI improves clinical outcomes in couples with previous repeated conventional ICSI failures.

Embryos are conventionally selected for transfer based on the evaluation of the cleavage speed and extent of blastomere fragmentation. Here we examined whether the predictive value of these criteria, as indicators of the chance of embryo implantation, can be further potentiated by adding previously described criteria reflecting the regularity of pronuclear development. In a group of embryos selected for transfer in 380 fresh embryo transfer cycles according to the conventional criteria, the transfer of only those embryos that developed from zygotes judged normal at the pronuclear stage (pattern 0) gave significantly higher pregnancy (44.8%) and implantation (30.2%) rates compared with the pregnancy (22.1%; P < 0. 05) and implantation rates (11.2%; P < 0.001) for the transfers of only those embryos that developed from zygotes judged abnormal (non-pattern 0). The transfer of only one pattern 0 embryo was sufficient for the optimal chance of pregnancy (no differences in pregnancy rates after transfer of one, two or three pattern 0 embryos), whereas the transfer of two pattern 0 embryos mostly resulted in a twin pregnancy. The inclusion of the criteria based on pronuclear morphology can thus lead to the application of a single embryo transfer policy and optimize the selection of embryos for transfer and cryopreservation.

Human embryo cryopreservation represents an indispensable extension of in-vitro fertilization (IVF) programmes as long as they are based upon the recovery of a large number of oocytes. The most widely used procedures include the cryopreservation of human zygotes or embryos in early cleavage, using 1,2-propanediol and sucrose as cryoprotectants. Our results over a 10 year period (1986-1995) on 5032 thawed cycles involving 14 222 stored embryos make it possible to appraise the results and the contribution of embryo freezing to assisted reproduction. Embryos survived the freeze-thaw process in 73% of cases leading to 4590 transfers of 2.2 embryos (91% of thawed cycles). The clinical pregnancy rate per transfer was 16%, the live birth rate 12%, and the rate of babies born alive per transferred embryo was 6%. Embryo freezing monitored 10 years later produced an average of 8% of additional births. By then, 86% of stored embryos had been thawed for transfer to patients. Destruction or donation were required for only 8% of all frozen embryos and there was no news from the parental couple in relation to almost 6% of embryos. The fate of the vast majority of embryos was decided during the first 5 years of storage. Blastocyst cryopreservation is making new strides, thanks to co-culture systems and embryo selection. Micromanipulation procedures seem to have little impact on the outcome of embryo freezing. Human oocyte freezing is again clinically applied. Indeed, much of the concern about injuries to the oocyte structures through the freeze-thaw process do not seem to be justified, and the problems with frozen-thawed oocyte fertilization has been overcome using intracytoplasmic sperm injection (ICSI). As long as oocyte in-vitro maturation is not well controlled, better results will probably be obtained with mature oocyte cryopreservation. Emerging methods include the freezing of immature oocytes, follicles and ovarian tissue.

The expression of both components of the high-affinity leukaemia inhibitory factor receptor, LIFR beta and glycoprotein 130 (gp130), was investigated in human oocytes and individual in-vitro cultured preimplantation embryos by reverse transcription-polymerase chain reaction (RT-PCR). Messenger RNA of both LIFR beta and gp130 was detected in as little as 1/30 and 1/12 sample equivalents of cDNA respectively, in oocytes (n = 4), 4-cell and expanded, blastacyst stage embryos. LIFR beta but not gp130 transcripts were detected at the 2-, 8- and 10-cell stages, and in cavitating and hatched blastocysts. In order to exclude a simian origin of these PCR products resulting from the Vero cell line that was used as a feeder during culture to the blastocyst stage, they were digested with restriction endonucleases Taql (LIFR beta) or Kpnl (gp130). Their human origin was confirmed. The results support an earlier finding of LIFR beta mRNA expression in human blastocysts, and extend these results to earlier stages and oocytes. This is the first report of LIFR beta and gp130 transcription in human oocytes. Taken together these results demonstrate that transcription of LIFR beta and gp130 takes place throughout human preimplantation development, and suggest that functional LIF receptors might be present at these stages. These results further confirm the feasibility of performing mRNA phenotyping of multiple genes with RNA derived from a single preimplantation stage embryo.