Recombinant Luteinizing Hormone (rLH) for controlled ovarian hyperstimulation in assisted reproductive cycles

BACKGROUND: During in vitro fertilization (IVF)/intracytoplasmic sperm injection (ICSI) treatment cycles, controlled ovarian hyperstimulation (COH) is performed with recombinant follicle stimulating hormone (rFSH) in combination with a gonadotrophin-releasing hormone (GnRH) analogue for the prevention of premature luteinizing hormone (LH) surges. The use of GnRH analogues however deprives the growing follicles of LH. The effectiveness of co-administrating rLH to rFSH for COH is at present unclear. OBJECTIVES: To compare the effectiveness and safety of a combination of recombinant LH and recombinant FSH with recombinant FSH alone in COH protocols in (IVF or ICSI followed by embryo transfer (ET). SEARCH STRATEGY: We searched the MDSG Group Specialised Register (searched up to Nov 2006) and CENTRAL, MEDLINE and EMBASE (1980 to November 2006) and reference lists of articles. SELECTION CRITERIA: Randomised controlled trials comparing COH with rFSH alone or in combination with rLH in IVF/ICSI were included. DATA COLLECTION AND ANALYSIS: Three review authors independently assessed trial quality and extracted data. We sought additional information if necessary. MAIN RESULTS: Fourteen trials involving 2612 women were included. Eleven trials involving 2396 women used a GnRH agonist .There was no evidence of a statistical difference in live birth rate reported in two trials (OR 1.51, 95% CI 0.79 to 2.87). There was no evidence of a statistical difference in clinical pregnancy rates reported in seven trials OR 1.15, 95% CI 0.91 to 1.45. There was no evidence of a statistical difference or in ongoing pregnancy rates seven trials OR 1.22, 95% CI 0.95 to 1.56. Three trials used a GnRH antagonist. No data on live birth rates was available. There was no evidence of a statistical difference in clinical pregnancy rates (one trial: OR 0.79, 95% CI 0.26 to 2.43) or in ongoing pregnancy rates (two trials: OR 0.83, 95% CI 0.39 to 1.80) comparing both groups. The pooled pregnancy estimates of trials including only poor responders showed significant increase in pregnancy rate, in favour of co-administrating rLH (three trials: OR 1.85, 95% CI 1.10 to 3.11) AUTHORS' CONCLUSIONS: There was no evidence of a statistical difference in pregnancy outcomes when rLH was used. Nevertheless, further large RCTs should be undertaken in long GnRH agonist down regulation protocols, since all pooled pregnancy estimates, although not statistically different probably due to the small numbers, point towards a beneficial effect of co-treatment with rLH, in particular with respect to pregnancy-loss and poor-responders