Objective: To determine whether patients with borderline semen should be treated with conventional IVF or intracytoplasmic sperm injection (ICSI).

Design: Randomized study.

Setting: A university medical center in The Netherlands.

Patient(s): One hundred six couples with borderline semen who were undergoing IVF and ICSI on sibling oocytes.

Intervention(s): Performing IVF and ICSI on sibling oocytes.

Main outcome measure(s): Fertilization and pregnancy rates.

Result(s): One thousand five hundred eighteen oocytes were collected in 106 oocyte retrievals: 849 oocytes were randomly allocated to ICSI, of which 761 were microinjected, and 669 oocytes were randomly assigned to IVF. In 26 of the 106 patients, there was fertilization only after ICSI and not after IVF (IVF- group). The fertilization rate was 51% (92/182 oocytes). In 78 patients, there was fertilization after both IVF and ICSI (IVF+ group); the fertilization rate was 51% for both the IVF- and ICSI-treated oocytes (271/528 oocytes and 334/658 oocytes, respectively). In 2 patients, there was no fertilization after either IVF (0/6 oocytes) or ICSI (0/9 oocytes). Patients of the IVF+ group had a higher total motile sperm count after preparation than did those of the IVF- group. More high-quality embryos were obtained after ICSI in patients of the IVF+ group. In 101 patients, embryo transfer was performed: 26 in the IVF- group and 75 in the IVF+ group. No significant differences were found with regard to pregnancy rates between those two groups: pregnancy rates were 54% in the IVF- group and 48% in the IVF+ group.

Conclusion(s): Performing ICSI on at least some of the oocytes will avoid unnecessary fertilization failure in patients with borderline semen: in this study, 26 of 104 cycles (25%) were rescued by ICSI.