STUDY QUESTION: Are cumulative live birth rates (CLBRs) similar in GnRH-antagonist and GnRH-agonist protocols for the first ART cycle including all subsequent frozen-thaw cycles from the same oocyte retrieval?

SUMMARY ANSWER: The chances of at least one live birth following utilization of all fresh and frozen embryos after the first ART cycle are similar in GnRH-antagonist and GnRH-agonist protocols.

WHAT IS KNOWN ALREADY: Reproductive outcomes of ART treatment are traditionally reported as pregnancies per cycle or per embryo transfer. However, the primary concern is the overall chance of a live birth. After the first ART cycle with fresh embryo transfer, we found live birth rates (LBRs) of 22.8% and 23.8% (P = 0.70) for the GnRH-antagonist and GnRH-agonist protocols, respectively. But with CLBRs including both fresh and frozen embryos from the first oocyte retrieval, chances of at least one live birth increases. There are no previous randomized controlled trials (RCTs) comparing CLBRs in GnRH-antagonist versus GnRH-agonist protocols. Previous studies on CLBR are either retrospective cohort studies including multiple fresh cycles or RCTs comparing single embryo transfer (SET) with double embryo transfer (DET).

STUDY DESIGN, SIZE, DURATION: CLBR was a secondary outcome in a Phase IV, dual-center, open-label, RCT including 1050 women allocated to a short GnRH-antagonist or a long GnRH-agonist protocol in a 1:1 ratio over a 5-year period using a web-based concealed randomization code. The minimum follow-up time from the first IVF cycle was 2 years. The aim was to compare CLBR between the two groups following utilization of all fresh and frozen embryos from the first ART cycle.

PARTICIPANTS/MATERIALS, SETTING, METHODS: All women referred for their first ART cycle at two public fertility clinics, <40 years of age were approached. A total of 1050 subjects were allocated to treatment and 1023 women started standardized ART protocols with recombinant human follitropin-β (rFSH) stimulation. Day-2 SET was planned and additional embryos were frozen and used in subsequent frozen-thawed cycles. All pregnancies generated from oocyte retrieval during the first IVF cycle including fresh and frozen-thaw cycles were registered. Ongoing pregnancy was determined by ultrasonography at gestational week 7-9 and live birth was irrespective of the duration of gestation. CLBR was defined as at least one live birth per allocated woman after fresh and frozen cycles. Subjects were censored out after the first live birth. Cox proportional hazard model was used to evaluate the relative prognostic significance of female age, BMI, the number of retrieved oocytes and the diagnosis of infertility in relation to the CLBR.

MAIN RESULTS AND THE ROLE OF CHANCE: Baseline characteristics were similar and equal proportions of patients continued with frozen-thaw (frozen embryo transfer, FET) cycles after their fresh ART cycle in the GnRH-antagonist and GnRH-agonist arms. When combining all fresh and frozen-thaw embryo transfers from first oocyte retrieval with a minimum of 2-year follow-up, the CLBR was 34.1% (182/534) in the GnRH-antagonist group versus 31.2% (161/516) in the GnRH-agonist group (odds ratio (OR):1.14; 95% CI: 0.88-1.48, P = 0.32). Mean time to the first live birth was 11.0 months in the GnRH-antagonist group compared to 11.5 months in the GnRH-agonist group (P < 0.01). The total number of deliveries from all FET cycles where embryos were thawed were higher in the antagonist group 64/330 (19.4%) compared to the agonist group 43/355 (12.1%) ((OR): 1.74; 95% CI: 1.14-2.66, P = 0.01). The evaluation of prognostic factors showed that more retrieved oocytes were associated with a significantly higher CLBR in both treatment groups. For the subgroup of obese women (BMI >30 kg/m2), the CLBR was significantly higher in the GnRH-antagonist group (P = 0.02).

LIMITATIONS, REASONS FOR CAUTION: The duration of the trial is a possible limitation with introduction of new methods as 'Freeze all' and 'GnRH-agonist triggering', but as these treatments were used in only few women, a systematic bias is not likely. Blastocyst culture of surplus embryos for freezing was introduced to both groups simultaneously, thereby minimizing the risk of bias. Furthermore, with a minimum of 2-year follow-up, a minority (<1%) still had cryopreserved embryos and no live birth at the end of the trial. The post hoc prognostic covariate analyses with multiple strata should be interpreted with caution. Finally, the physicians were not blinded to GnRH treatment group after randomization.

WIDER IMPLICATIONS OF THE FINDINGS: With the improvement of embryo culture, freezing and thawing methods as well as a strategy of elective SET, CLBR until first live birth provides an all-inclusive success rate for ART. When comparing GnRH-antagonist and GnRH-agonist protocols, we find similar CLBRs, despite more oocytes being retrieved in the GnRH-agonist protocol.

STUDY FUNDING/COMPETING INTERESTS: An unrestricted research grant is funded by Merck Sharp & Dohme Corp., a subsidiary of Merck & Co., Inc., Kenilworth, NJ, USA (MSD). The funders had no influence on the data collection, analyses or conclusions of the study. No conflict of interests to declare.

TRIAL REGISTRATION NUMBER: EudraCT #: 2008-005452-24. ClinicalTrial.gov: NCT00756028.

TRIAL REGISTRATION DATE: 18 September 2008.

DATE OF FIRST PATIENT'S ENROLLMENT: 14 January 2009.