Recombinant luteinizing hormone supplementation to recombinant follicle-stimulating hormone induced ovarian hyperstimulation in the GnRH-antagonist multiple-dose protocol

Abstract

BACKGROUND: Suppression of endogenous LH production by mid-follicular phase GnRH-antagonist administration in controlled ovarian hyperstimulation protocol using recombinant (rec) FSH preparations void of LH activity may potentially affect ovarian response and the outcome of IVF treatment. The present study prospectively assessed the effect of using a combination of recFSH and recLH on ovarian stimulation parameters and treatment outcome in a fixed GnRH-antagonist multiple dose protocol. METHODS: 127 infertile patients with an indication for IVF or ICSI were recruited and randomized (using sealed envelopes) to receive a starting dose of either 150 IU recFSH (follitropin α) or 150 IU recFSH plus 75 IU recLH (lutropin α) for ovarian hyperstimulation. GnRH-antagonist (Cetrorelix) 0.25 mg was administered daily from stimulation day 6 onwards up to and including the day of the administration of recombinant HCG (chorion gonadotropin α). Gonadotropin dose adjustments were allowed from stimulation day 6 onwards, HCG was administered as soon as three follicles ≥18 mm were present. The primary outcome parameter was treatment duration until administration of HCG. RESULTS: Exogenous LH did not shorten the time necessary to reach ovulation induction criteria. Serum estradiol (E₂) and LH levels were significantly higher on the day of HCG administration in the recLH-supplemented group (1924.7 ± 1256.4 vs 1488.3 ± 824.0 pg/ml, P<0.03), and 2.1 ± 1.4 vs 1.4 ± 1.5 IU/l, P<0.01, respectively). CONCLUSIONS: Except for higher E₂ and LH levels on the day of HCG administration, no positive trend in favour of additional LH was found as defined by treatment outcome parameters.

cetrorelix, GnRH-antagonist, ovarian stimulation, recombinant FSH, recombinant LH

Introduction

The role of both endogenous and exogenous LH in controlled ovarian hyperstimulation (COH) is controversial and might vary depending on the stimulation protocol and GnRH-analogue employed. Although only FSH is required for growth of the ovarian follicle, a minimal amount of LH is necessary to achieve adequate follicular steroidogenesis and to develop the capacity of preovulatory follicles to ovulate and transform into corpora lutea (European Recombinant LH Study Group, 1998).

For the GnRH-agonist ‘long’ ovarian hyperstimulation protocol, the effect of a profound follicular phase endogenous LH level suppression under recombinant (rec) FSH stimulation has been associated with an adverse reproductive outcome in some studies (Westergaard et al., 2000; Esposito et al., 2001; Copola et al., 2003), however, this association has been scrutinized in others (Loumaye et al., 1997; Balasch et al., 2001; Penarrubia et al., 2003). The residual LH level below which follicular development and treatment outcome might be impaired is denominated as the LH ‘threshold’ level.

Additional LH activity might optimize COH when GnRH-analogues and gonadotropin preparations depleted in LH content are combined. More specifically, it was indicated that the duration necessary to achieve ovulation induction criteria might be shorter and the gonadotropin consumption lower (Filicori et al., 1998, 2003; Balasch et al., 2003), when HMG rather than highly purified HMG or recFSH are used in pituitary suppressed patients.

The effect of LH suppression in GnRH-antagonist ovarian hyperstimulation protocols is also an issue of concern. Midcycle antagonist administration leads to a marked decrease in LH levels (Albano et al., 1997), down to the lower detection limit of commercial immunoassays for LH in some patients (Finas et al., 2003). A dose finding study on the GnRH-antagonist ganirelix revealed very low pregnancy rates under high antagonist doses which in turn are associated with profound LH suppression (Ganirelix dose-finding study, 1998). Furthermore, in comparative phase III trials the odds of clinical pregnancy was comparatively (but not significantly) lower in GnRH-antagonist study arms in which recFSH void of any LH activity was utilized for stimulation as compared to trials that utilized human menopausal gonadotropin (HMG) (Al-Inany and Aboulghar, 2003).

In response to these observations, a group comparative trial was undertaken based on the hypothesis that using a combination preparation of recFSH and recLH might optimize ovarian hyperstimulation parameters and infertility treatment outcome in GnRH-antagonist multiple dose cycles, possibly in part by alleviating the pituitary suppressive effect of antagonist administration at a critical point of follicular development.

Materials and methods

Study design

The present study was a prospective, randomized, controlled, open, single-centre, group-comparative clinical trial assessing a starting dose of 150 IU recFSH vs 150 IU recFSH plus 75 IU recLH (2:1) for controlled ovarian hyperstimulation in the GnRH-antagonist multiple-dose protocol. Our institutional review board approved the study protocol and all study participants provided written informed consent. The randomization process was conducted by drawing sealed envelopes and patients were free to start ovarian stimulation within the next three spontaneous menstrual cycles after randomization.

Patient population

Main inclusion criteria were: indication for treatment with IVF or ICSI; age between 20 and 39 years; body mass index between 18 and 35 kg/m²; regular menstrual cycle, ranging from 24 to 35 days, intra-individual cycle variability of ≤3 days; use of fresh as well as frozen-thawed sperm retrieved by testicular biopsy. Main exclusion criteria were: >3 previous unsuccessful assisted reproduction technique attempts; previous poor response to gonadotropin stimulation defined as <3 preovulatory follicles; history of ovarian hyperstimulation syndrome (OHSS) grade II–III; polycystic ovarian syndrome; any other endocrine disorder; no natural luteal phase prior to treatment cycle; abnormal uterine cavity as evaluated by ultrasonography; presence of a clinically significant systemic disease.

Protocol

Ovarian stimulation started on day 2 of the natural cycle with 150 IU recFSH (Gonal-F; Serono, Geneva Switzerland) in the control group (‘recFSH’ group), and 150 IU recFSH (Gonal-F) plus 75 IU recLH (Luveris; Serono, Geneva Switzerland) in the study group (‘recFSH/recLH’ group). All injections were given once daily s.c. in the morning by the patient. After 5 days of gonadotropin treatment, GnRH-antagonist cetrorelix 0.25 mg (Cetrotide; Serono, Geneva Switzerland) administration was started by once daily s.c. injection in the morning. Gonadotropin and antagonist treatment was continued up to and including the day of HCG administration. From day 6 onwards, the recFSH dosage could be increased at the discretion of the physician to 225 IU or 300 IU according to the ovarian response as assessed by ultrasonography and estradiol (E₂) values. In case of a dose increment to 300 IU recFSH in the recFSH/recLH group, the recLH dose was concomitantly adjusted to 150 IU.

RecHCG 250 μg (Ovitrelle; Serono, Geneva, Switzerland) was administered s.c. as soon as three follicles were ≥18 mm in diameter, and 34–36 h thereafter oocyte retrieval was performed. All follicles ≥13 mm were punctured. After IVF or ICSI according to standard procedures, no more than three embryos were to be replaced on day 2 after oocyte retrieval. Spare two-pronuclei oocytes were frozen. The outcome of cycles utilizing embryos generated from those oocytes is not presented further herein.

Luteal phase support started the morning after oocyte retrieval and was provided with daily 90 mg micronized progesterone (Crinone 8%; Serono, Geneva Switzerland). Additionally, 5000 IU urinary HCG (Choragon; Ferring Arzneimittel GmbH, Kiel, Germany) were administered once on the day of embryo transfer in case E₂ levels on the day of HCG were ≤2500 pg/ml. When pregnancy was achieved, luteal phase support in the form of daily vaginal or twice weekly i.m. progesterone administration was continued until fetal heartbeat was visualized by vaginal ultrasonography.

Compulsory criteria for cycle cancellation on any monitoring visit were: premature luteinization, defined as progesterone rise ≥1.7 ng/ml; premature LH rise, defined as ≥12.1 mIU/ml; E₂ drop, defined as a reduction of E₂ levels of at least 50% between two monitoring visits; imminent OHSS, defined as ≥15 intermediate follicles (12–16 mm) on stimulation day 8 or ≥20 large follicles (16–20 mm) on day 10 or later. An optional criterion for cycle cancellation was poor response, defined as mono- or bifollicular development, in which case patients were free to cancel the cycle or to continue HCG injection.

Assessments

Hormonal and ultrasound assessment was performed on stimulation day 6 just before first antagonist administration and thereafter as necessary, as well as on the day of HCG administration. For determining the optimal time point for HCG administration, a follicular growth of 2 mm per day was assumed. Five medical doctors involved in this trial performed the ultrasound examinations. Ultrasound and hormonal assessment, as well as oocyte retrieval and embryo transfer was performed 7 days a week. Serum E₂, LH and progesterone levels were assayed with the electrochemiluminescence immunoassay ‘ECLIA’ (Roche Diagnostics Inc., Germany) on the Roche Elecsys 2010 automated immunoassay analyser by the local laboratory. Intra-assay and inter-assay coefficients of variation were <2.0% and <7% for E₂, <2.5% and <5% for LH, and <2.5% and <5% for progesterone, respectively.

Retrieved oocytes were classified as metaphase II oocytes, metaphase I oocytes, germinal vesicle stage oocytes or degenerate oocytes for ICSI patients. Embryos were classified as top quality, medium quality and low quality and the cumulative embryo score in a modified version is presented, as described previously (Ludwig et al., 2000).

Outcome measures

Primary endpoint: number of days of gonadotropin treatment. Secondary endpoints: gonadotropin consumption in international units; hormone (E₂, LH, progesterone) levels on day 6 and the day of HCG; proportion of metaphase II to total number of retrieved oocytes (in ICSI cases); proportion of two pronuclei oocytes to number of injected or inseminated oocytes (fertilization rate); cumulative modified embryo score; biochemical pregnancy, defined as HCG ≥10 mIU/ml 14 days after embryo transfer; clinical pregnancy, defined as an ongoing pregnancy at 12 weeks of gestation; implantation rate, defined as number of gestational sacs with fetal heartbeat divided by the number of embryos transferred.

Power analysis

The sample size was based on the primary outcome measure duration of gonadotropin treatment (days). The hypothesis was that the addition of recLH would enhance ovarian response and lead to a shorter time necessary to reach ovulation induction criteria. The study was powered to detect a difference of 1 day between the two treatment modalities. In a previous GnRH-antagonist multiple-dose trial (Ludwig et al., 2002) with a similar ovarian stimulation study protocol and similar patient population, the number of days of gonadotropin stimulation was 12.04 ± 1.7 (mean ± SD) in the GnRH-antagonist multiple-dose protocol study arm with recFSH for ovarian stimulation and a fixed start of the antagonist on day 6. Group sample sizes of 47 and 47 patients who reach HCG administration achieve 81% power to detect a difference of 1 day in the number of gonadotropin treatment days between the null hypothesis that both group means are 12.0 days and the alternative hypothesis that the mean of group 2 is 11.0 days with assumed group SD of 1.7 and 1.7 and with a significance level (alpha) of 0.05 using a two-sided two-sample t-test.

Statistical tests

Chi-square test was used to analyse nominal variables in the form of frequency tables. Normally distributed metric variables (Kolmogorov–Smirnov test) were tested by a two-sided t-test for independent samples, whereas non-normally distributed metric variables were analysed by Mann–Whitney U test. Alpha <0.05 was considered significant. Descriptive and efficacy analyses were performed on an intention-to-treat (ITT) basis as appropriate, taking into account all randomized patients. Computations were performed using SPSS statistical software version 12.0.1 for Windows (SPSS lac., 1989–2002).

Results

Recruitment for the study was pursued from June 2003 to December 2003; follow-up of study participants was pursued until May 2004. The number of patients considered eligible for inclusion in the trial who turned down participation was not recorded. According to the sample size calculation, further recruitment was stopped as soon as 47 patients in each of the two treatment groups had reached HCG injection. At that time point 127 patients had already been recruited into the study and were thus included in the analysis.

Causes of infertility in the patient population are depicted in Table I. Proportions of couples with solely male, solely female or combined infertility were similar between the groups, as was the proportion of women with primary and secondary infertility.

Table II depicts the study participants' flow through the trial. Reasons for discontinuation were similar between the two groups. No significant differences were found between the groups with regard to known confounders of ovarian response and treatment outcome, as summarized in Table III.

Tables IV and V show stimulation characteristics, oocyte and embryological data. The gonadotropin treatment duration was not shortened by LH administration. An additional mean of 915 (±213.4) IU recLH were administered in the recLH supplemented group. A higher dose of recFSH was given to patients in the recFSH/recLH group, which reflects the relatively (not significantly, P=0.07) higher number of patients with dose increments in this group. Progesterone values on day 6 of stimulation were significantly higher in the recFSH only group. Serum E₂ and LH levels on the day of HCG were significantly higher in the recLH-supplemented group. Without exogenous LH supplementation, serum LH values decreased from 2.7 (±2.2) mIU/ml on stimulation day 6 to 1.4 (±1.5) on the day of HCG administration in the control group. No differences were observed between groups as regards oocyte and embryo data. Table VI depicts IVF/ICSI treatment outcome in the two groups, which was similar.

Discussion

This study shows that adding recLH to a recFSH driven stimulation in GnRH-antagonist multiple dose cycles is associated with a significant increase in serum E₂ and LH levels on the day of HCG administration (referred to as ‘peak’ values for E2 hereafter) (Table IV). Since E₂ levels were similar on day 6 in the two populations, it can be assumed that the higher peak serum E₂ values in the study population can be attributed to recLH supplementation, which possibly only exerts an effect under the circumstances of relative LH depletion caused by mid-follicular antagonist administration. It is unlikely that this difference is caused by a different number of preovulatory follicles, since the mean number of retrieved oocytes per oocyte retrieval was highly similar between groups. The mean E₂/oocyte ratio revealed the same trend with relatively higher E₂ values per retrieved oocyte in the recLH group (data not shown). It must however also be noted that the higher peak E₂ serum levels in the study group might be also be partially caused by the comparatively higher recFSH dose administered (Table IV).

Furthermore, this is indicative that GnRH-antagonist administration leads to suppression in endogenous LH to such an extent that low dose recLH supplementation will significantly affect E₂ biosynthesis of the maturing follicles. Although the amount of LH activity necessary for optimal follicular development is unknown, it has been suggested that <1% of follicular LH receptors need to be occupied in order to allow normal steroidogenesis (Chappel and Howles, 1991). However, the concept of normal ovarian steroidogenesis is not well defined and it is unlikely that the lower E₂ values in the recFSH only group represent a correlative of impaired follicular development or impaired oocyte developmental competence. Regarding follicular development, we found no shortening of the time until HCG criteria were met under exogenous LH administration. Although SDs were higher than the assumed SDs in the power calculation, it can be assumed that additional low dose LH will not increase the follicular growth rate in this antagonist protocol setting, since no positive trend with the treatment group was observed [12.0 (±2.4) days in the study group vs 11.4 (±2.1) days in the control group]. Similarly, regarding oocyte developmental competence, we found no evidence of a positive effect of higher peak serum E₂ values as a consequence of LH supplementation, as evidenced from a similar cumulus oocyte number, similar proportions of mature oocytes, fertilization rates and mean embryo scores in the two groups.

The fact that early follicular phase LH supplementation and relatively higher LH levels measured on stimulation day 6 did not translate into higher E₂ levels on stimulation day 6 in the recLH supplemented group, contradicts the observation that serum E₂ levels and endogenous LH levels before the onset of antagonist administration are positively correlated (Kolibianakis et al., 2003). This finding might be attributed to the fact that granulosa cells (the main source of E₂) will not acquire their own LH receptors before the mid- to late follicular phase under the influence of FSH (Erickson et al., 1979), whereas endogenous background LH levels in the earlier follicular phase might already be sufficient to occupy most of the constitutionally expressed LH receptors of the theca cells. Concomitantly, multifollicular recruitment appeared slower in the recLH group, resulting in a lower number of follicles in the recLH supplemented group on stimulation day 6, which is likely to have contributed to this observation. The mean number of follicles ≥13 mm was 1.5 (±2.6) and 2.3 (±3.6) (not significantly different) in the recLH supplemented group and recFSH only group, respectively. This obviously prompted the physicians to increase the gonadotropin dose in the recLH supplemented group more often than in the non-supplemented group. It may therefore be speculated that early follicular phase LH supplementation might alter multifollicular recruitment and support dominant follicle selection, while suppressing smaller follicles. However, two points must be considered: allowing gonadotropin dose adjustments dependent on individual response and according to the physician's discretion unquestionably renders assessment of follicular growth dynamics and ovarian response more difficult, and introduced variability in the present study. Second, hormone values and ovarian sonomorphology in early follicular phase were not assessed in this work; therefore the above stated has to be considered hypothetical. The finding of higher progesterone levels in the recFSH only group on stimulation day 6 is of unclear origin, although it might be attributed to the relatively higher number of growing follicles in the recFSH only group. Values in both groups were in the lower third of the reference range for the follicular phase and the clinical significance of this finding is likely to be low.

It has been suggested previously that there is no need in further exploring LH supplementation in ovarian hyperstimulation (Hull et al., 1994). The introduction of highly purified and recFSH and the widespread use of these compounds combined with GnRH-agonists without additional LH supplementation has not led to a drop in overall success rates, as evidenced from annual statistics from large national programmes, such as the German IVF Registry (Data collection 1999–2002). However, a number of retrospective studies have indicated that low levels of circulating follicular phase LH in GnRH-agonist long protocol cycles might be associated with impaired E₂ synthesis and/or a low oocyte yield, low fertilization rate, low pregnancy rate or high abortion rate, respectively (Fleming et al., 1998, 2000; Westergaard et al., 2000; Esposito et al., 2001; Humaidan et al., 2002). In contrast with these findings, two recent studies (Balasch et al., 2001; Penarrubia et al., 2003) failed to find a lower LH threshold value in agonist-suppressed patients discriminative between conception and non-conception cycles, or early pregnancy loss and ongoing pregnancy cycles. Furthermore, no association between LH levels and parameters of ovarian response was found. Part of this disagreement between studies has been attributed to different relative potencies of different agonists used. Two interventional trials (Humaidan et al., 2004; Marrs et al., 2004) failed to find a significant positive impact as regards treatment outcome of recLH supplemented ovarian stimulation in long agonist protocols. An interventional trial in the antagonist protocol in which increasing the dose of HMG (containing both FSH and LH) at the time of GnRH antagonist introduction was tested, also failed to demonstrate a positive effect of this measure regarding number of oocytes retrieved, embryos obtained, implantation rate or clinical pregnancy rate (Aboulghar et al., 2004).

Notably, little is known about the association of follicular phase LH levels with ovarian response and treatment outcome in antagonist cycles. With no intrinsic activity at the GnRH-receptor, GnRH-antagonists lead to a direct and sudden suppression of endogenous gonadotropins within 4–6 h of administration. In contrast with the long agonist protocol this suppression does not occur before stimulation day 6 in the fixed protocol and is undulatory rather than steady, with a nadir every 4–6 h after antagonist administration. This, together with the pulsatile fashion in which LH is secreted, makes correlation of a single LH measurement with treatment outcome difficult. In the present trial, endogenous LH levels on day of HCG in the non-LH supplemented group were on average above suggested LH threshold values (≤0.7 IU/l, Fleming et al., 2000; <0.5 IU/l, Westergaard et al., 2000). However, this finding most likely reflects the study protocol rather than the true maximum nadir values for LH after antagonist administration, because LH assessment was pursued ∼24 h after the last antagonist administration. Recently, it has been reported that profound suppression of LH on day 8 of stimulation in antagonist treated cycles is associated with a significantly higher chance of achieving an ongoing pregnancy (Kolibianakis et al., 2004). Similarily, another recent study reported a higher number of oocytes retrieved, embryos obtained, and embryos cryopreserved in patients with LH <0.5 mIU/ml on the day of HCG in antagonist cycles (Merviel et al., 2004).

Another recent publication suggested LH supplementation as an alternative way of indirectly evaluating the consequences of LH depletion caused by the antagonist (Cedrin-Durnerin et al., 2004). In short, findings by this group were for the most part concordant with our results, although a flexible single dose cetrorelix protocol was used and LH was only administered concomitantly with the antagonist.

Because there is no reliable or cost-effective way to detect which women might possibly need additional LH administration, it was suggested to systematically add LH to ovarian stimulation protocols (Levy et al., 2000). In line with this proposal and also addressing the recent withdrawal of urinary derived HMG preparations from some markets, the present trial aimed at evaluating the effect of a combination preparation of recombinant gonadotropins on ovarian response and IVF treatment outcome. The trial did not aim to specifically assess the effect of LH supplementation concomitantly to LH suppression by a GnRH-antagonist, because the study design (LH administration in early and late follicular phase, respectively) inevitably would imply multiplicity of hypothesis testing within the trial. In conclusion, our findings do not support systematic exogenous low dose LH throughout the follicular phase in GnRH-antagonist multiple-dose cycles in an unselected patient population.

Table I.

Causes of infertility, primary or secondary infertility, previous abortuses, and sterilization

Cause of infertility	Number (%) in indicated treatment groupa
	recFSH (n =65)	recFSH/recLH (n =62)
Male factor only	32 (49.2)	34 (54.8)
Tubal factor only	7 (10.8)	9 (14.5)
Male and tubal factor	10 (15.4)	8 (12.9)
Idiopathic/endometriosis	9 (13.8)	5 (8.0)
Male factor and endometriosis	2 (3.0)	2 (3.2)
Azoospermia	3 (4.6)	3 (4.8)
Vasectomy	1 (1.5)	1 (1.6)
Tubal sterilization	1 (1.5)	0 (0.0)
Primary infertilityb	51 (78.5)	42 (67.8)
Secondary infertilityc	14 (21.5)	20 (32.2)
Abortusesd	6 (9.2)	8 (12.9)

recFSH: recombinant FSH; recLH: recombinant LH.

a

Intention-to-treat group (n=127).

b

Pregnancy has never occurred or never been established (i.e. no previous early or late abortion and no ectopic pregnancy).

c

All other cases not covered by primary infertility.

d

Number of patients with history of early or late spontaneous abortion.

Table II.

Patient data by treatment group and stage of trial

Treatment stage and reason for discontinuation	Number (%) in indicated treatment group
	recFSH	recFSH/recLH	Total
Randomized	65 (100)b	62 (100)b	127 (100)a,b
Treated with gonadotropin	65 (100)	61 (98.3)c	126 (99.2)
Insufficient ovarian response	3(4.6)	2(3.2)	5(3.9)
Insufficient ovarian response+LH rise	1(1.5)	0(0.0)	1(0.8)
Premature luteinization	3(4.6)	1(1.6)	4(3.1)
Risk of hyperstimulation	2(3.0)	2(3.2)	4(3.1)
E2drop	2(3.0)	1(1.6)	3(2.4)
Total discontinued stimulation	11(16.9)	6(9.6)	17(13.4)
Treated with HCG	54 (83.0)	55 (88.7)	109 (85.8)
With oocyte pick-up	54 (83.0)	54 (87.0)d	108 (85.0)
No oocyte retrieved	0(0.0)	1(1.6)	1(0.8)
No MII oocyte retrieved	0(0.0)	1(1.6)	1(0.8)
Total discontinued after oocyte pick-up	0(0.0)	2(3.2)	2(1.6)
With sperm incubation/injection	54 (83.0)	52 (83.9)	106 (83.5)
Number of ICSI cases	41 (63.1)	46 (74.2)	87 (68.5)
Number of IVF cases	13 (20.0)	6 (9.7)	19 (15.0)
Fertilization failure	2(3.0)	2(3.2)	4(3.1)
With embryo transfer	52 (80.0)	50 (80.6)	102 (80.3)
Total discontinued treatment	13(20.0)	12(19.4)	25(19.7)

E₂ = estradiol; MII=metaphase II.

a

Intention-to-treat group=all patients randomized group.

b

All percentages relative to the number intended to treat.

c

One patient never received treatment and was lost for follow-up.

d

One patient received HCG but no oocyte retrieval was performed because of psychogenic anejaculation on the day of the scheduled oocyte aspiration.

Table III.

Summary statistics of patients characteristics. Values are Mean ± SD, except stated otherwise

Variables	Treatment groupa
	recFSH	recFSH/recLH	P
Age (years)	30.5 (±4.2)	30.3 (±4.7)	NS
Body height (cm)	167.2 (±6.1)	168.3 (±5.4)	NS
Body weight (kg)	67.7 (±13.2)	68.1(±12.3)	NS
Body mass index (kg/m2)	24.1 (±4.3)	24.0 (±4.1)	NS
Menstrual cycle duration (days)	28.3 (±3.0)	28.4 (±2.0)	NS
Duration of infertility (years)	4.43 (±2.8)	3.92 (±2.7)	NS
Number of patients in first treatment cycle (%)	48 (73.8)	41 (66.1)	NS
Number of patients with previous ET (%)b	17 (26.2)	21 (33.8)	NS
Number of cigarette consumers (%)c	24 (37.0)	20 (32.2)	NS

ET = embryo transfer.

a

Intention-to-treat group (n=127).

b

This includes up to a maximum of three previous unsuccessful embryo transfers.

c

defined as consumption of more than five cigarettes per day.

Table IV.

Stimulation outcome in the two groups of patients. Values are mean ± SD, except stated otherwise

Variables	Treatment groupa
	recFSH (n =54)	recFSH/recLH (n =55)	P
Number of patients with HCG on day of ET (%)	39 (75.0)	34 (68.0)	NS
Number of patients with dose increment (%)	12 (22.2)	22 (40.0)	NS
Stimulation length (days)	11.4 (±2.1)	12.0 (±2.4)	NS
Number of follicles ≥13 mm on day 6	2.3 (±3.6)	1.5 (±2.6)	NS
Recombinant FSH (IU)	1875.4 (±646.4)	2082.8 (±695.7)	NS
Recombinant LH (IU)	NA	914.7 (±213.4)	NA
Stimulation day 6
E2 (pg/ml)	427.7 (±274.5)b	352.0 (±288.9)c	NS
LH (mIU/ml)	2.7 (±2.2)b	3.6 (±6.3)c	NS
Progesterone (ng/ml)	0.6 (±0.2)b	0.5 (±0.2)c	0.04
Day of hCG administration
E2(pg/ml)	1488.3 (±824.0)	1924.7 (±1256.4)d	0.03
LH (mIU/ml)	1.4 (±1.5)	2.1 (±1.4)d	0.01
Progesterone (ng/ml)	0.8 (±0.3)	0.9 (±0.9)d	NS

IU = international units; NA = not applicable.

a

Patients reaching HCG injection (n=109)

b

Values from four patients are missing due to patient non-compliance.

c

Values from two patients are missing due to patient non-compliance.

d

Value from one patient is missing due to patient non-compliance.

Table V.

Fertilization and cleavage outcome in the two groups of patients. Values are mean ± SD, except where stated otherwise

	Treatment groupa
Variables	recFSH (n =54)	recFSH/recLH (n =54)	P
Total number of COC retrieved	416	426	NA
Mean COC number
per ITT	6.4 (±5.5)	6.9 (±5.7)	NS
per OPU	7.7 (±5.1)	7.9 (±5.3)	NS
MII oocytes/number of  oocytes (%)b	85.3 (±18.8)	82.9 (±17.6)	NS
Fertilization rate (%)
IVF	43.2 (±31.3)	51.4 (±31.8)	NS
ICSI	64.1 (±25.3)	53.7 (±28.3)	NS
Mean embryo scorec	27.92 (±10.11)	24.94 (±11.34)	NS

COC=cumulus oocyte complex; ITT=intention-to-treat; OPU=oocyte pick-up.

a

All patients with ovarian puncture.

b

Only ICSI cases.

c

All embryos generated were transferred.

Table VI.

Implantation, pregnancy and miscarriage in the two groups of patients. Values are mean ± SD, except where stated otherwise

	Treatment group
Variables	recFSH	recFSH/recLH	P
Total number transferred embryos	109	99	NA
Mean number of transferred embryos
per ITT	1.68 (±0.9)	1.6 (±0.9)	NS
per ET	2.1 (±0.5)	1.98 (±0.6)	NS
Positive pregnancy tests (n)	15	16
% per ITT	23.1	25.8	NS
% per ET	28.8	32.0	NS
Early pregnancy lossesa (n)	3	8b	NS
Clinical pregnancies (n)	12	8	NS
% per ITT	18.4	12.9	NS
% per ET	23.0	16.0	NS
Implantation rate (%)	13.8	8.1	NS

a

Miscarriages before 12 weeks of gestation.

b

Including one extrauterine gravidity (treated by laparoscopy).