RESUMEN
To analyze the value of adding Gn-RH antagonist to agonist stop protocol, and to compare agonist stop-antagonist protocol, in normal and low responders. 69 previously poor responders on long agonist protocol [group A], 63 normal responders who did not become pregnant after >/= 2 previous ICSI cycles on long agonist protocol [group B], and 14 women with premature progesterone rise on previous non-agonist cycles [group C]. Group A, and B patients were randomly allocated into either agonist stop protocol, agonist stop-antagonist, or HMG antagonist only protocol, while all group [ patients were stimulated using agonist-antagonist protocol. Duration of stimulation, number of ampoules, number of follicles, late follicular estradiol and progesterone, and cycle outcome. Agonist antagonist protocol was associated with significantly lower late follicular progesterone than the other protocols in groups A, and B, and than the previous attempts in group C. There was a trend toward lower cancellations and higher pregnancy rate using the new protocol. The use of agonist stop-antagonist protocol resulted in lower late follicular progesterone and trend toward a higher pregnancy rate