Corifollitropin alfa compared to daily FSH in controlled ovarian stimulation for in vitro fertilization: a meta-analysis

The present study offers a meta-analysis of published randomized controlled trials (RCTs) evaluating the outcomes of in vitro fertilization (IVF) cycles using corifollitropin alfa for controlled ovarian stimulation (COS) in comparison with daily recombinant FSH (rFSH). The study examined seven RCTs including 2138 patients receiving corifollitropin alfa and 1788 women receiving daily rFSH for COS. As a novel aspect, this meta-analysis included two specific subpopulations of IVF patients, i.e. egg donors and poor responders. There were no significant differences between corifollitropin alfa and rFSH with respect to the majority of the clinical parameters considered, and comparable were the outcomes in terms of live birth rate, ongoing pregnancy rate, and clinical pregnancy rate. Women receiving corifollitropin alfa had a significantly higher number of metaphase II oocytes at ovum pick-up, and number of formed embryos, in comparison to rFSH. The risk of cycle cancellation due to overstimulation was significantly higher in the corifollitropin alfa group. Ovarian hyperstimulation syndrome (OHSS) incidence was statistically comparable between patients receiving long lasting or daily rFSH. Nevertheless, in view of the fact that corifollitropin alfa resulted in a higher number of metaphase II oocytes collected and a higher number of cycles cancelled due to overstimulation, corifollitropin alfa should be cautiously considered in women with the potential of being hyper responders.


Introduction
In recent years, attention has been increasingly paid -in the field of IVF -to the development of simplified treatment approaches with the aim of reducing treatment burden and to prevent drop-out rate after a failed IVF cycle [1][2][3].
Corifollitropin alfa is a novel recombinant fertility hormone with prolonged follicle-stimulating activity used for COS in IVF [4]. A single subcutaneous injection of corifollitropin alfa has the capacity to initiate and sustain multiple follicular growth for the first seven days of COS, thus reducing the number of injections required during one treatment cycle [4,5]. The long-acting properties of this new drug have been obtained by incorporating the carboxy-terminal peptide from the beta subunit of human chorionic gonadotropin (hCG) to the beta subunit of FSH [6,7]. As unique pharmacokinetic properties, corifollitropin alfa displays a slower absorption and a longer elimination half-time in comparison to rFSH. Corifollitropin alfa has indeed a time interval to peak serum levels that is almost 4-fold longer, and the elimination half-time (approximately 69 h) that is about two-fold longer than the conventional rFSH [4,5,8]. Corifollitropin alfa has identical pharmacodynamics properties to rFSH since it interacts exclusively with the FSH receptor and is devoid of LH activity [5]. Based on pharmacokinetics studies and modeling [7,9], the optimal corifollitropin doses have been identified to be: 100 μg in women who weigh less than or equal to 60 kg and who are 36 years of age or younger; 150 μg in women who weigh more than 60 kg regardless of age and women who weigh 50 kg or more and who are older than 36 years of age. The effectiveness and safety of corifollitropin alfa in comparison to rFSH have been the subject of several randomized controlled trials (RCTs). Previous meta-analyses [10,11] have performed analysis on studies published up to 2011. The present paper aims to provide an updated meta-analysis on the corifollitropin alfa in IVF.

Search strategy
A systematic search of the electronic literature through MEDLINE and EMBASE databases was performed to identify studies published between 2001 and December 2014. The following headings and text strings were used alone or in combination: corifollitropin alfa, long acting FSH, in vitro fertilization, assisted reproductive technologies, in vitro fertilization-embryo transfer, intracytoplasmic sperm injection (ICSI). In addition, the Google Scholar database was similarly searched for studies related to corifollitropin alfa. A reference list of all included manuscripts and reviewers related to corifollitropin alfa was manually searched for additionally potentially eligible studies. Relevant journals and symposia proceedings were also evaluated to identify additional data. There was no language restriction.

Selection criteria and data extraction
Two reviewers performed data extraction and evaluation of trial quality independently. Any disagreement concerning the extracted data was resolved by consensus and, if necessary, by involving a third reviewer. Only randomized controlled trials that compared the clinical effectiveness and safety of corifollitropin alfa with rFSH were deemed eligible for inclusion in the meta-analysis. Publications were scrutinized to identify study characteristics, randomization, allocation, blinding, and intentionto-treat analysis. Excluded studies included retrospective and uncontrolled studies, editorials and reviews. The search results were cross-checked against papers considered in previous meta-analyses [10,11]. The target population was composed of infertile couples with any infertility factor undergoing to IVF/ICSI or egg donors, with the therapeutic intervention being corifollitropin alfa versus rFSH. The considered outcome measures were ongoing pregnancy rate, live birth and clinical pregnancy rate, early miscarriage rate, number of metaphase II oocytes per oocyte pick-up, number of embryos obtained and fertilization rate per woman with intracytoplasmic sperm injection, incidence of OHSS, adverse events, cycle cancellation and total duration. All outcomes were defined before undertaking the literature search. If additional information was required, the corresponding authors were contacted.

Statistical analysis
Concerning the measures of the treatment effect, the present study considered both dichotomous and continuous data. As for the first, the results in the control and intervention groups of each study were expressed as Peto odds ratios (OR) with 95 % confidence intervals (CI). Continuous outcomes were expressed as weighted mean differences (WMD) with 95 % confidence intervals as well. The data were treated using fixed or random effects model. Heterogeneity of treatment effects was statistically evaluated by Hotelling T-square (τ 2 ), Higgins (I 2 ), Birge's ratio (H 2 ), and Chi-square test (Chi 2 ). A P-value < 0.05 referred to the overall effect was considered statistically significant. Statistical analysis was carried out using the package Metafor, version 1.9-3, a free and open-source add-on for conducting metaanalyses with the statistical software environment R.

Results
The search yielded 51 studies of which 44 were excluded by screening through titles and abstracts (Fig. 1). Full manuscripts were retrieved for the remaining papers that included 6 full texts published in peer reviewed journals, and 1 abstract from the annual meeting of the American Society of Reproductive Medicine. Study characteristics are shown in Table 1. The seven trials enrolled a total of    Fig. 2, A-K. There were differences in patient characteristics and selection. In the Engage study [12,13] women aged 18-36 years, had a body mass index (BMI) of 18-32 kg/m 2 and weighed 61-90 kg. In the Ensure trial [14], participants aged 18-36 years with a body weight ≤ 60 kg. In both Engage and Ensure trials patients with a history of ovarian hyper response to ovarian stimulation (more than 30 follicles ≥ 11 m) or OHSS, polycystic ovarian syndrome (PCOS) or more than 20 basal antral follicles on ultrasound (<11 mm, both ovaries combined) were excluded from the study. Women with a history of poor ovarian response were also excluded in both studies. In the Corifollitropin Alfa Dose-finding Study Group [9], patients were women aged 20-39 years with a BMI of 17-31 kg/m 2 . In the Devroey et al. study [15], participants were between 18 and 39 year of age and a BMI 18-29 kg/m 2 ). Patients with a history of OHSS, PCOS or poor response, were excluded. In the Requena et al. study [3], the oocyte donors included in the study aged 18-35 years, had regular menses and had a body weight > 60 kg with a BMI up to 29 kg/m 2 with at least seven antral follicles at the beginning of the cycle. Donors having PCOS or multifollicular ovaries were excluded. The study of Kolibianakis et al. [16] included patients with previous poor ovarian response defined as retrieval of ≤4 COCs in a previous IVF cycle in which a starting dose of at least 450 IU per day was used, age <45 years, regular spontaneous menstrual cycle (24-35 days), body mass index (BMI) of 18-32 kg/m 2 and basal FSH ≤20 IU/l. Differences among studies may represent sources of biases and, as it will be discussed later, influenced the result in terms of heterogeneity.

Discussion
Controlled ovarian stimulation for IVF has relied, since its introduction, on daily administration of various formulations of gonadotropins. A single injection of the long lasting FSH corifollitropin alfa can replace the first standard seven daily injections of gonadotropins. This may result, in line with current expectations in IVF, in a simplification of the treatment regimen and reduced treatment burden [1][2][3]. Corifollitropin alfa is administered as a single injection during menstrual cycle days 2 or 3 (stimulation day 1). Daily injections of FSH are started on stimulation day 8 if needed. A gonadotropin releasing hormone antagonist is associated starting on stimulation day 5 to prevent premature LH surge. In the current study, available RCTs were included in a metaanalysis comparing corifollitropin alfa with conventional daily injections of FSH in IVF patients. A substantial amount of data derived from two large phase III randomized, double dummy, double blind trials designed by the manufacturer, and including patient less than 36 years that were considered not at risk of developing OHSS (antral follicle count ≤ 20). There were 756 women weighting > 60 kg that received corifollitropin alfa at 150 μg [12,13], and 268 with a body weight ≤ 60 kg that received 100 μg [14]. Results were compared with reference groups receiving daily injections of rFSH at 150 IU [15] or 200 IU [12,13]. The Corifollitropin Alfa Dose Finding Study Group [8], a phase II study including 325 women, was undertaken to investigate the dose-response relationship with respect to the number of oocytes retrieved after administration of corifollitropin alfa at 60, 120 or 180 μg. Devroey et al. [15], in their  : tau^2=0.1035, Chi^2=12.8948, df=5 (P=0.0244), H^2=3.05 : tau^2=0.9283, Chi^2=41.4411, df=5 (P<0.0001), H^2=6.23 tau^2=0.6996, Chi^2=19.3828, df=6 (P=0.0036), H^2=4.28 : tau^2=0.4224, Chi^2=15.8575, df=3 (P=0.0012), H^2=6.86 [17]. 1390 women were randomized to a single injection of 150 μg corifollitropin alfa (n = 694) or daily 300 IU rFSH during the first seen days of ovarian stimulation. A GnRH antagonist was started on stimulation day 5. When required, ovarian stimulation was continued with daily injections of rFSH at the maximal dose of 300 IU. As result, corifollitropin alfa was found to be non-inferior to rFSH in terms of effectiveness. Requena and co-workers [3] evaluated the degree of satisfaction in oocyte donors undergoing treatment with corifollitropin alfa (n = 48) compared with women receiving rFSH (n = 40). The study involved patients weighting more than 60 kg and thus receiving corifollitropin alfa at 150 μg. Assignment to each treatment group was made according to a quasi-experimental design which included consecutive opportunity sampling. Patients were pre-treated with an oral contraceptive pill before starting COS. Potential high responders were excluded. Moreover, to further reduce the risk of OHSS, GnRH agonist (instead of hCG) was used to trigger final oocyte maturation. The study carried out by Kolibianakis by et al. [16], included seventy-nine women and aimed to evaluate the potential benefit of corifollitropin alfa treatment compared to daily rFSH in poor responder patients. In this prospective, randomized, open label, non-inferiority clinical trial, women were considered poor responders if they had a previous poor response to ovarian stimulation. Other inclusion criteria were: age less than 45; regular menses; body mass index (BMI) of 18-32 kg/m2; and basal FSH ≤20 IU/l. On day 2 of menses women received either a single dose of corifollitropin alfa at 150 μg. When needed, 450 IU of rFSH were given from day 8 of stimulation up the hCG day. Results were compared with control groups receiving daily injections of rFSH at 450 IU. As a major finding, the number of cumulus oocyte complex that were recovered was not statistically different between the corifollitropin alfa and rFSH groups.
Pooling data from these studies showed no significant differences in the majority of clinical parameters considered between corifollitropin alfa and rFSH. This metaanalysis revealed comparable outcomes in terms of live birth rate, ongoing pregnancy rate and clinical pregnancy rate. Pregnancies obtained in women undergoing COS with corifollitropin alfa had the same risk of spontaneous abortion as pregnancies from rFSH. When the ovarian response to COS was considered, corifollitropin alfa yielded a higher number of oocytes than rFSH. In line with this finding, the number of embryos obtained per woman randomized in the corifollitropin alfa group was also higher. These outcomes, along with the evidence that a higher number of women treated with this drug experienced cycle cancellation due to overstimulation, underscore the need for a careful selection of the patients that are eligible for corifollitropin alfa in order to prevent OHSS. However, the major ovarian response associated to COS with corifollitropin alfa did not result in a statistically significant increased risk of OHSS. When discussing the risk of OHSS, it appears important to consider the criteria for cycle cancellation used. These are available for the Ensure and Engage RTCs. In both studies, the investigator was allowed to withhold rFSH administration for a maximum of 3 days. When the investigator considered the ovarian response too high, he was allowed to cancel the cycle at any time. In the case of a risk for OHSS (more than 30 follicles > 11 m on ultrasound scan) hCG was to be withheld and the treatment cycle would be cancelled. The Kolibianakis et al. [16] study compared corifollitropin alfa with rFSH in poor responder patients. Thus, there was no risk of OHSS. There were no significant differences with respect of length of stimulation associated to corifollitropin alfa in comparison to rFSH. Corifollitropin alfa has been designed to relieve the treatment burden experienced by IVF women. Results offered by the pooled analysis of this meta-analysis are compatible with the notion that proper exposure to corifollitropin alfa results in IVF outcome comparable to that observed with rFSH. This concept is in line with conclusions provided by two previously published meta-analyses [10,11]. As a novel aspect, this meta-analysis included RCTs on two specific subpopulations of IVF patients, i.e. egg donors [3] and poor responders [16]. Conclusions of our meta-analysis could be slightly different when separately considering these subpopulations. Unfortunately, the number of this type of studies is not sufficient to perform a populationspecific meta-analysis. This circumstance is reflected by moderately higher levels of heterogeneity, as seen in the previous section. To satisfyingly detect it, we need to consider that H 2 indexes in the above experiments are always significantly bigger than one, and this is sign of strong diversity between groups. The resulting analysis may be thus more reflective of the everyday clinical practice. Also consider that at the time of writing only a limited number of RCTs was available. This represented a major limitation of this meta-analysis. For instance we were unable to analyze publication bias using funnel plots, since it is known that at least ten RCTs are required to get statistically significant results [18]. A concept that must be born in mind when using corifollitropin alfa is that its unique pharmacokinetic profile can result in the recruitment of an increased cohort of developing follicles, and this contraindicates its use in high responder patients that are at risk of OHSS. Adequate definition of the ovarian reserve and thus exclusion of patients with high ovarian reserve appears to remain a crucial step before considering long lasting FSH for COS.

Conclusions
Corifollitropin alfa was designed to simplify the treatment regimen in patients undergoing ovarian stimulation for IVF. This systematic review and meta-analysis suggests that corifollitropin alfa is effective as rFSH in terms of live birth rate, ongoing pregnancy rate and clinical pregnancy rate. The increased number of eggs retrieved under corifollitropin alfa regimen reflects the elevated effectiveness of this novel FSH formulation, but warns at the same time against the possible increased risk of OHSS in high responder women.