Despite the progression in assisted reproductive technology (ART), the preferred protocol for poor responders is still controversial. The management of poor responders consists of 10% of ART cycles .

The response to controlled ovarian hyperstimulation (COH) is lower regarding estradiol level , number of obtained oocytes , and fertilization , implantation and pregnancy rates in patients with low ovarian reserve . Furthermore , bad quality embryos are observed in these women more than normoresponders and the increase of cancellation rate and doses of gonadotropin administration are remarkable results in poor responders . Several criteria have introduced for poor responders , the main defect in the management of them is lack of specific definition .Several strategies are available to improve ART cycles outcome in poor responders. These modalities include using : high FSH dose , stop GnRH-agonist protocol , addition of growth hormone , transdermal testosterone , aromatase inhibitor , GnRH-antagonist and recombinant FSH ( r-FSH) ; while the improvement of pregnancy rate has been quite low.

The most common used protocol for ovarian stimulation is microdose GnRH-agonist flare in poor responders .Some investigators concluded that the use of GnRH-agonist " even in lower doses , led to prolonged stimulation and increased the cost without improving IVF outcome. Furthermore this method increased LH , progesterone and androgen of serum in follicular phase , which caused deleterious effect on follicular growth and oocyte quality .

Clomiphene citrate co-treatment with gonadotropin and antagonist are one of the recommended protocol in poor responders . Clomiphene citrate increases endogenous FSH versus agonist in microdose protocol. Decreasing the doses of used gonadotropin and duration of stimulation are its beneficial effects in COH cycle .

The aim of this study was comparing CC/gonadotropin/antagonist and GnRH agonist flare protocols on IVF outcome in poor responders .

Despite the progression in assisted reproductive technology (ART) , the preferred protocol for poor responders is still controversial. The management of poor responders consists of 10% of ART cycles .

The response to controlled ovarian hyperstimulation (COH) is lower regarding estradiol (E₂) level , number of obtained oocytes , and fertilization , implantation and pregnancy rates in patients with low ovarian reserve . Furthermore , bad quality embryos are observed in these women more than normoresponders and the increase of cancellation rate and doses of gonadotropin administration are remarkable results in poor responders Several criteria have introduced for poor responders , the main defect in the management of them is lack of specific definition .Several strategies are available to improve ART cycles outcome in poor responders. These modalities include using : high FSH dose , stop GnRH-agonist protocol , addition of growth hormone , transdermal testosterone , aromatase inhibitor , GnRH-antagonist and recombinant FSH ( r-FSH) ; while the improvement of pregnancy rate has been quite low.

The most common used protocol for ovarian stimulation is microdose GnRH-agonist flare in poor responders .Some investigators concluded that the use of GnRH-agonist " even in lower doses " led to prolonged stimulation and increased the cost without improving IVF outcome. Furthermore this method increased LH , progesterone and androgen of serum in follicular phase , which caused deleterious effect on follicular growth and oocyte quality .

Clomiphene citrate co-treatment with gonadotropin and antagonist are one of the recommended protocol in poor responders . Clomiphene citrate increases endogenous FSH versus agonist in microdose protocol. Decreasing the doses of used gonadotropin and duration of stimulation are its beneficial effects in COH cycle .

The aim of this study was comparing CC/gonadotropin/antagonist and GnRH agonist flare protocols on IVF outcome in poor responders .

Materials and Methods Study design This study was a prospective randomized controlled trial including 159 poor responder patients who were candidate for IVF . Women with ≥38 years old who had one or more previous failed IVF cycles in which three or fewer oocyte were been retrieved and/or serum E2 level on the day of hCG administration was ≤500 pg/ml were enrolled in this study . Patients with BMI > 30 , endocrine or metabolic disorders , history of ovarian surgery , sever endometriosis and sever male factor ( azospermia ) were excluded from the study . Patients were divided into two groups , 79 women in group I received CC/gonadotropin/antagonist (mild protocol) and 80 women in group II received microdose GnRH-agonist flare (microdose protocol) . A method of computer-generated randomization was used .

Treatment Protocols All women received oral contraceptive for 21 days which started on the first day of previous cycle . In group I , stimulation were performed by administration of clomiphene citrate (Iran hormone, Tehran, Iran) 100 mg from day 3 of withdrawal bleeding until day 7 of cycle and gonadotropin stimulation with 225-300 IU daily , recombinant FSH (r-FSH) SC or hMG IM , were started from day 5 of cycle . In group II ovarian stimulation was initiated with GnRH-agonist , buserelin (Suprefact, Aventis Pharma, Frankfurt, Germany) 50 µg SC twice a day from cycle day 2 of withdrawal bleeding . After two days , 225-300 IU/day recombinant FSH (r-FSH) SC or hMG IM were administered.

Ovarian response was monitored by serial ultrasound examinations and evaluation of serum E₂ levels , then doses of gonadotropin were adjusted as required in both groups.

In group I , when at least one follicle ≥ 14 mm in mean diameter was observed , 0.25 mg GnRH antagonist (ganirelix , Organon, netherlands) SC daily was started and continued until hCG injection . Urinary Human chorionic gonadotropin 10000 IU was administered intramuscular when at least two follicles reached a mean diameter of 18 mm in both groups . Also , endometrial thickness and serum E₂ level were measured on the day of hCG injection .Oocyte retrieval was performed 34-36 hours after hCG injection and conventional IVF or intracytoplasmic sperm injection (ICSI) was done as appropriately . All embryos were scored by the number , size , shape , symmetry and cytoplasmic appearance of blastomers , and the presence of anucleate cytoplasmic fragmentation .

Based on the number and quality of available embryos and patient's age , one to five embryos were transferred on the day 2 or 3 after oocyte retrieval under ultrasound guidance with a CCD embryo transfer catheter (Laboratory C.C.D., Paris, France). Luteal support with progesterone 100 mg daily IM was started on the day of oocyte retrieval and was continued until the documentation of fetal heart activity on ultrasound.

Cycle cancellation was defined as three groups : [1] poor ovarian response : fewer than two growing follicles on transvaginal ultrasound, and an E₂ level < 200 pg/ml on the day 7 of stimulation ; [2] failed oocyte retrieval : no obtained oocyte on the day of ovarian puncture ; [3] failed fertilization : no fertilized oocyte after IVF/ICSI.

Inclusion Criteria:

• Women with ≥38 years old
• women who had one or more previous failed IVF cycles in which three or fewer oocyte were been retrieved and/or serum E2 level on the day of hCG administration was ≤500 pg/ml were enrolled in this study

Exclusion Criteria:

• BMI > 30
• endocrine disorders
• metabolic disorders
• history of ovarian surgery
• sever endometriosis
• sever male factor ( azospermia )