Impact of letrozole co-treatment during ovarian stimulation with gonadotrophins for IVF: a multicentre, randomized, double-blinded placebo-controlled trial

Abstract

Does letrozole co-treatment during ovarian stimulation with gonadotrophins for IVF reduce the proportion of women with premature progesterone levels above 1.5 ng/ml at the time of triggering final oocyte maturation? The proportion of women with premature progesterone above 1.5 ng/ml was not significantly affected by letrozole co-treatment. IVF creates multiple follicles with supraphysiological levels of sex steroids interrupting the endocrine milieu and affects the window of implantation. Letrozole is an effective aromatase inhibitor, normalizing serum oestradiol, thereby ameliorating some of the detrimental effects of IVF treatment. A randomized, double-blinded placebo-controlled trial investigated letrozole intervention during stimulation for IVF with FSH. The trial was conducted at four fertility clinics at University Hospitals in Denmark from August 2016 to November 2018. A cohort of 129 women with expected normal ovarian reserve (anti-Müllerian hormone 8-32 nmol/l) completed an IVF cycle with fresh embryo transfer and received co-treatment with either 5 mg/day letrozole (n = 67) or placebo (n = 62), along with the FSH. Progesterone, oestradiol, FSH, LH and androgens were analysed in repeated serum samples collected from the start of the stimulation to the mid-luteal phase. In addition, the effect of letrozole on reproductive outcomes, total FSH consumption and adverse events were assessed. The proportion of women with premature progesterone >1.5 ng/ml was similar (6% vs 0% (OR 0.0, 95% CI [0.0; 1.6], P = 0.12) in the letrozole versus placebo groups, respectively), whereas the proportion of women with mid-luteal progesterone >30 ng/ml was significantly increased in the letrozole group: (59% vs 31% (OR 3.3, 95% CI [1.4; 7.1], P = 0.005)). Letrozole versus placebo decreased oestradiol levels on the ovulation trigger day by 68% (95% CI [60%; 75%], P < 0.0001). Other hormonal profiles, measured as AUC, showed the following results. The increase in LH in the letrozole group versus placebo group was 38% (95% CI [21%; 58%], P < 0.0001) and 34% (95% CI [11%; 61%], P = 0.006) in the follicular and luteal phases, respectively. In the letrozole group versus placebo group, testosterone increased by 79% (95% CI [55%; 105%], P < 0.0001) and 49% (95% CI [30%; 72%], P < 0.0001) in the follicular and luteal phases, respectively. In the letrozole group versus placebo group, the increase in androstenedione was by 85% (95% CI [59%; 114%], P < 0.0001) and 69% (95% CI [48%; 94%], P < 0.0001) in the follicular and luteal phases, respectively. The ongoing pregnancy rate was similar between the letrozole and placebo groups (31% vs 39% (risk-difference of 8%, 95% CI [-25%; 11%], P = 0.55)). No serious adverse reactions were recorded in either group. The total duration of exogenous FSH stimulation was 1 day shorter in the intervention group, significantly reducing total FSH consumption (mean difference -100 IU, 95% CI [-192; -21], P = 0.03). Late follicular progesterone samples were collected on the day before and day of ovulation triggering for patient logistic considerations, and the recently emerged knowledge about diurnal variation of progesterone was not taken into account. The study was powered to detect hormonal variations but not differences in pregnancy outcomes. Although the use of letrozole has no effect on the primary outcome, the number of women with a premature increase in progesterone on the day of ovulation triggering, the increased progesterone in the mid-luteal phase due to letrozole may contribute to optimizing the luteal phase endocrinology. The effect of letrozole on increasing androgens and reducing FSH consumption may be used in poor responders. However, the effect of letrozole on implantation and ongoing pregnancy rates should be evaluated in a meta-analysis or larger randomized controlled trial (RCT). Funding was received from EU Interreg for ReproUnion and Ferring Pharmaceuticals, and Roche Diagnostics contributed with assays. N.S.M. and A.P. have received grants from Ferring, Merck Serono, Anecova and Gedeon Richter, and/or personal fees from IBSA, Vivoplex, ArtPred and SPD, outside the submitted work. The remaining authors have no competing interests. NCT02939898 and NCT02946684. 15 August 2016. 22 August 2016.