Comment planifier le début d’une stimulation ovarienne : différentes stratégies entre cycle naturel, pilule contraceptive ou double stimulation

Résumé libre de l'article publié dans Fertility & Sterility par Dominique de Ziegler (board scientifique de la Maison de la Fertilité), Chloé Tran (co-fondatrice de la Maison de la Fertilité), Sean Soktean et Pichetra Ou

La stimulation ovarienne est la mesure la plus efficace jamais mise en place en procréation médicalement assistée pour améliorer les résultats, en permettant de récolter plusieurs ovocytes, puis des embryons. Aujourd’hui, les protocoles de stimulation ovarienne consistent à administrer des gonadotrophines exogènes pour contourner les mécanismes naturels qui limitent normalement l’ovulation à un seul ovocyte chez l’humain.

Pour des raisons pratiques, de nombreuses tentatives ont été faites afin de contrôler, ou “programmer”, le moment du début de la stimulation ovarienne, dans le but d’améliorer son fonctionnement et son efficacité dans les programmes de PMA.

Les différentes options disponibles pour contrôler le déclenchement de la stimulation ovarienne sont abordées ici, ainsi que la nouvelle possibilité d’utiliser des progestatifs pour bloquer l’ovulation prématurée.

Texte Original en Anglais :

Programming the onset of ovarian stimulation: from early follicular phase start to oral contraceptive pill, to luteal phase E2, Duostim, and random start oral contraceptive protocols

Historically, the onset of menstrual cycle has been timed by reference to menses. Following this paradigm, the advancement of the follicular phase is judged in cycle days, with cycle day one being the first day of menses. This approach is anchored in practicality, as the onset of menses is aneasily identifiable event typically taking place between two menstrual cycles. Physiologically speaking, however, menses constitute the last event of the prior cycle, which f inishes with the onset of endometrial desquamation. Conversely, menstruation has no physiological connection with the onset of the new cycle. The functional initiation of the new cycle occurs a little before, during, or a little after the onset of menses but menstruation per se is an independent event from the onset of the next cycle. The physiological event that initiates the menstrual cycle, the reference for follicular phase advancement, is a slight follicle-stimulating hormone (FSH) elevation or ‘‘FSH signal.’’ This FSH signal, on average of 2.5 mIU/mL in amplitude, stimulates the FSH receptor-carrying antral follicles present in both ovaries. On average, there are 10–20 antral follicles between the two ovaries. These are characterized by their antral cavity– 2–9mm in diameter– which allows their identification on ultrasound and the presence of FSH receptors. In principle, the whole cohort of follicles responds by increasing the production of both E2 and inhibin B. This, in turn, lowers serum FSH levels by a negative feedback effect and thereby, terminates the FSH signal. The negative feedback that turns off the FSH signal is the mechanism through which the ovulatory quota is set at one– sometimes two– in humans.

The intercycle FSH signal cannot be easily identifiable due to its small amplitude (2.5 mIU/mL, on average), short of doing multiple samplings for several days (1). As we will discuss it later, the occurrence of the intercycle FSH signal can be controlled, however, or programmed by different means. Right from the inception of assisted reproductive technology (ART), over 4 decades ago, increasing the number of ovocytes retrieved, and thereby embryos obtained, was seen as instrumental for improving outcomes. Today, ovarian stimulation (OS) for the purpose of increasing the number of oocytes retrieved stands as the single most effective measure ever taken for improving ART outcomes. The objective of OS was therefore to induce multiple follicular development.

PROGRAMMING OS: PROBLEM STATEMENT

Theoriginal objective of OSinARTwastooverridethecontrol of the ovulatory quota, set at one in humans. This has been achieved by preventing the FSH drop occurring in the early follicular phase after the initial antral follicle response to FSH. Not knowing exactly when the FSH signal occurs, original stimulation protocols were arbitrarily initiated on cycle days 2–4. In the early days of ART, OS– increasing the number of follicles developing and maturing– hasbeendoneusingmedications originally designed to induce ovulation in anovulatory women. Two options were available: increasing endogenous FSH by blocking the negative feedback loop with the anti-estrogen clomiphene citrate or now, aromatase inhibitors; or providing exogenous FSH. In the latter casethe most common approach used in ART today– exogenous FSH has been originally initiated on cycle days 2–4. This approach indeed amounts to maintain serum FSH at levels similar to those seen during the early follicular phase FSH signal throughout the whole follicular phase. Indeed, the daily administration of 150 IU of FSH–astandardstimulation dose for normal responders– elevates serum FSH levels by approximately 2.5 mIU/mL. This level is equivalent to an early follicular phase FSH signal. Hence, maintaining serum FSH at the levels of the intercycle signal throughout the entire follicular phase was the way to override the ovulatory quota set at one. Achieving multiple follicular development for harvesting multiple oocytes was called OS. This regimen, physiological in appearance, suffered from causing programming problems, however, as it implied starting OS at the beginning of menstrual cycles. On the contrary, ART teams, as their activity increased, became eager to organize or ‘‘program’’ their activity to aim for having an equal number of oocyte retrievals every day, as we will discuss it. This was seen as instrumental for optimizing the practical eff iciency of infertility treatments and in turn, quality control, both on the clinical side and in the laboratory. As discussed, several means have been developed for programming the onset of OS in ART. The advent of gonadotropin releasing hormone agonist (GnRH-a) used for preventing premature ovulation also allowed controlling the onset of OS, an approach that has been used for many years. Today new options are envisioned including even random start protocols, which we will review also (2). METHODS The most common databases, Embase, PubMed, and Cochrane, were searched for the following keywords. In vitro fertilization (IVF) oral contraceptive (OC) pill: 213 hits, 17 were found pertinent and retained in the article. IVF E2 treatment 1,330 and pretreatment 57. Ultimately, 24 were found most pertinent and retained in this review. IVF luteal phase stimulation: 36 hits, four were retained for this article. IVF progesterone primed OS (PPOS): 89 hits, 18 were retained for the article. DuoStim protocol: 28 hits, six pertinent articles were retained in the review. IVF random start: 248 results 10 years, five were retained and discussed in the present review.

EARLY FOLLICULAR STIMULATION START: HISTORICAL PERSPECTIVE The miserable results of early days ART– then called IVFreadily sparked interest in achieving multiple ovulations and retrieving several oocytes. One pioneer team in this endeavor has been Howard Jones’ group in Norfolk (3). These investigators adopted beginning in 1981 a novel stimulation protocol using human menopausal gonadotropin (hMG) started on days 2–4. Human menopausal gonadotropin had been developed in pre-ART times for inducing ovulation in women suffering from hypogonadotropic amenorrhea. It contained FSH and luteinizing hormone (LH) in equal proportions and was standardized in terms of bioactivity (4), as gonadotropin assays were not available when hMG was developed. A different article in this series discusses the relative advantages of LH bioactivity containing regimens in OS protocols and those constituted of pure FSH of recombinant origin. Moreover, exogenous human chorionic gonadotropin (hCG) was used to substitute for the midcycle LH surge (3). Laparoscopy for follicular aspiration was scheduled 36–38 hours after hCG administration (3,5). Soon hMG-based protocols established themselves as superior to clomiphene-based approaches (5). Concerns for possible deleterious effects of OS on the endometrium were raised from inception. This fear had led Steptoe and Edwards’ group to opt for natural cycle ART, used in the first successful ART attempt (6). Hence, attention was given to endometrial morphology when stimulation was initiated by the Jones’ group. In their reported series, 11 patients showed an ‘‘advanced’’ pattern and 10 an ‘‘in-phase’’ (7) endometrium. This was estimated according to the debatable Noyes criteria (8). A significant difference in serum progesterone levels on days 16 and 18 was found in these two groups. Serum progesterone levels were significantly higher by day 18, if pregnancy was established (7). In ART, the embryo is placed over the endometrium 24–48 hours earlier than it normally arrives in the endometrial cavity in natural conception. This led to the hypothesize that the ‘‘advanced’’ endometrium encountered in OS– at least in certain women– might be beneficial for embryo implantation (7). Of course, all the emphasis put on these issues of acceleration of endometrial changes in the early days of ART has become obsolete by the advances in cryopreservation and liberal access to deferred embryo transfers (9). The negative impact of OS on endometrial receptivity has now been established. This is notably the case in strong ovarian responders in whom ART outcome is improved by freezing all embryos and reverting to deferred embryo transfer (10).

Today, very large programs that implement a 7-day-aweekactivity are often comfortable with earlyfollicular phase stimulations. Smaller programs however aim to even out their activity with different types of pre-ART treatments, including the OC pill, which we will review here.

PROGRAMMING STIMULATION WITH OC PILL

The option of program stimulation cycles with the OC pill dates back to before the use of GnRH analogs in ART and notably, GnRH-a. Gonen et al. (11) studied IVF cycles in 14 women whowere randomized to receive either leuprolide acetate (0.5 mg) or 5000 IU hCG to rigger ovulation at midcycle. Interestingly, however, all these patients had been pretreated byOCpills. Although the response to agonist trigger is not the issue discussed here, it is remarkable that no patient in Gonen’s cohort presented a premature LH elevation (11). This sparked a new study in which 181 stimulation cycles were pretreated with OC pill (study group) and their results comparedwith113othercyclesnormallyinitiatedintheearly follicular phase (control group) (12). The mean length of ovarian suppression with OC pill in the study group was 35.3 0.9 days. Remarkably, no spontaneous LH surges occurred when the use of OC preceded ovarian hyperstimulation, whereas in the control group, the incidence of LH surges was 19.5%. The mean amount of human menopausal gonadotropin required was significantly lower in the study group than in the control group (8.9 0.4 and 10.9 0.4 ampules, respectively). Significantly more follicles R1.5 cm in diameter were seen on the day before oocyte retrieval and significantly more oocytes were retrieved per attempt in the study group– pretreated with OC (12). This led the investigators to proclaim that OC pretreatment was useful in ART not only to facilitate the scheduling of cycles but also to prevent spontaneous LH surges (12). These data fell into oblivion, however, as soon later GnRH-a preparations became widely, if not universally, used for blocking premature ovulation. Interestingly pretreatment with the OC pill was shown to dampen the agonist phase of the response to the agonist. This wasseenasapractical advantage when agonists became used in OS. Keltz et al. (13) reported that pretreatment with the OC pill showed a similar FSH response to the agonist, as compared with nontreated controls. Conversely, the LH response was significantly quenched. In a different trial, Biljan et al. (14) showed pretreatment with OC abolished ovarian cyst formation, and shortened the time required to achieve pituitary suppression. Furthermore, gonadotropin requirements were decreased without any negative effect on pregnancy rates (14). A cyst developed in 27 patients in the control group (52.9%) but in no patients pretreated by OC (odds ratio [OR], 115; 95% confidence interval [CI], 10–617) (14). Patients in the study group achieved pituitary suppression faster (median difference, 7 days; 95% CI, 4–14) and required fewer ampules of gonadotropin (median difference, 10; 95% CI, 6–14) (14). Together, these data led to a wide use of OC pill pretreatment in combination with GnRH-a stimulation protocols primarily, for reducing the risk of cyst formation. The advent of GnRH antagonist progressively replaced the agonist for OS, although the debate about their equal efficacy lasted for quite a while. Finally, in a large Cochrane Review, Al-Inany et al. (15) reported data comparing the use of GnRH-aandthenewantagonistinARTprimarilylooking at live birth rates (LBRs). Their review based on data of moderate quality evidences how edno evidence of a difference in LBR between GnRH antagonist and long-course GnRH-a (OR, 1.02; 95% CI, 0.85–1.23; 12 randomized controlled trials [RCTs], n ¼ 2303) (15). The risk of ovarian hyperstimulation syndrome(OHSS) was however decreased in women receiving the antagonist. Later however, Lambalk et al. (16) reported that in the general IVF population, GnRH antagonist protocols were associated with lower ongoing pregnancy rates but confirmed the lower incidence of OHSS. Kolibianakis et al. (17) objected, however, indicating that there were flaws in that analysis. This was notably due to the presence in the antagonist group of women who had received OC pill pretreatment but with a far too short interval before starting OS, which hampered results (17). Today there is a general consensus to consider that agonist and antagonist protocols are equivalent. The medical community,however,largelyprefers antagonist stimulation protocols because of the decreased risk of OHSS. When assessing the outcome of frozen transfer cycles, both protocols have been found equivalent (18). Certain administer the antagonist on a fixed day– day6although other investigators follow a flexible approach based on follicular size and E2 levels, with report of better results in the fixedstrategy (19). Although the consensus in favor of the antagonist protocol is generally admitted, there are still reports suggesting that certain poor responders might do better with the agonist (20). For us, the issue at stake here– programming the ART cycle for better efficacy– the general adoption of antagonist protocols has revived the problem. Indeed antagonist ART cycles are normally started on cycle days 2–4. Hence, clinicians were found at a loss without the programming possibilities that agonist protocols offered. Hence efforts have been deployed for adapting pre-ART treatments to program ART cycles, for the sake of clinical efficacy and patient preference. Logically, the first interest has been for OC pill pretreatment, whichwouldallowustoprograminadvancethetiming of ART treatments. A first systematic review and metaanalysis has however slowed the initial enthusiasm, as it reported that OC pretreatment was associated with lower ART outcomes (21). In this review by Griesinger et al. (21), two of the four studies retained for the meta-analysis had however an unreasonably short pill-OS interval of only 2 days. Importantly, these two studies carried all the burden of the whole negative impact of OC pill pretreatment calculated in the meta-analysis of these first four studies (21). Of note too, all four studies retained in the meta-analysis used recombinant FSH exclusively during the stimulation (21). The possible negative impact of OC pill pretreatment appeared corroborated by a large Cochrane systematic review and metaanalysis (22). Of note, this meta-analysis did not sort results according to the stop OC pill– ART interval nor the type of gonadotropin used, with or without LH bioactivity. Subsequently, other studies provided results going against the Cochrane study (22). Montoya-Botero et al. (23) studied the effects of OC pill administration for 12–30 days for the purpose of programming the ART cycle, respecting a washout interval of 5 days before starting ART. The study included 3,517 women who were pretreated by the pill and 599 who did not and served as controls. The use of oral contraceptive pill (OCP) was independently associated with a small increase in the number of oocytes retrieved after adjusting for age, body mass index, use of OCP, cause of infertility, initial dose (IU), type of gonadotropin, stimulation days, and total stimulation units (23). The study revealed that OC pill hadasimilar probability of achieving a live birth as compared with patients not using OCPs after fresh embryo transfer (OR, 0.89; 95% CI, 0.69–1.15) and similar cumulated live births after the use of fresh and frozen embryos (OR, 0.94; 95% CI, 0.73–1.21) (23). These findings were supported by other studies such as those of Bellver et al. (24) showing no impact of OC pill pretreatment on the risk of miscarriage and Kim et al. (25) no alteration in poor responders. Pereira et al. (26) showedthatpretreatment withOCpillprovided equivalent results obtained after pretreatment with transdermal E2. Adifferent study– RCT–respecting a 5-day time interval between stopping OC and the onset of stimulation likewise showednodifference in ARToutcome(23). Antagonist protocols may cause an excessive reduction in endogenous LH levels which could be exacerbated by OC pill and increase needs in LH bioactivity, as discussed in a different article (27). Interestingly, a very strong argument in favor of OC pretreatment in antagonist protocols came from US ART centers, which generally have the best ART results in the world. Indeed, a recent article looking at the common practices among the best-performing centers in the United States indicated that nearly all used OC pretreatment and a combination of FSHandhMGintheirOSregimen(28). It is likely therefore that the profound LH suppression induced by OC pill benefits from OS regimen using LH bioactivity containing hMG preparation (28). In summary,therefore, the OCpill pretreatment is a valid way of programming ART treatments for the sake of the connivance and efficacy of ART teams. The proper pillstimulation interval of 5 days should be observed and the use of stimulation protocols containing LH bioactivity considered. Despite the accumulated evidence that OC pill treatment has no adverse consequence on ART and is handy for the scheduling cycle, the European Society Human Reproduction’s guidelines talk against it (https://www.eshre. eu/Guidelines-and-Legal/Guidelines/Ovarian-Stimulationin-IVF-ICSI).

E2-BASED PROGRAMMING: A PHYSIOLOGICAL APPROACH

During the intercycle interval, two hormonal signals– a drop in luteal E2 and progesterone– take place and two distinct events ensue. These are the onset of menses; and intercycle FSHelevation (Fig. 1). In a prospective trial, we demonstrated that the drop in progesterone is solely responsible for triggering the onset of menses independently of E2 levels (29). Conversely, the drop in E2 on corpus luteum demise controls the timing of the intercycle FSH elevation, independently of progesterone levels (29) (Fig. 1). Applications of this lesson have allowed us to control the timing of the intercycle FSH elevation by timely administration of E2, with FSH elevation occurring on the third day after stopping E2 (30). This therefore amounts to block the onset of the new follicular phase until E2 treatment is discontinued. Prior work has demonstrated that E2 treatment (2 mg of oral E2 twice adayor0.1mgoftransdermalE2/d) can reliably postpone FSH elevation and ensuing follicular recruitment for %2weeks (31). The ability to control FSH elevation by E2 treatment has allowed us to effectively program the onset of stimulations. Some have claimed E2 pretreatment improved results notably because FSH suppression was less profound than with OC pill. Inaddition, the hypothesis has been made that E2 pretreatment in the luteal phase of the preceding cycle provided a more homogeneous cohort of follicles and in turn better ART outcome (32). According to these investigators, luteal E2 treatment was meant to prevent early follicular recruitment due to a premature intercycle FSH signal (32). Similar effects have also been reported by the same investigators with the timely use of an antagonist during the luteal phase for the same purpose (33). These results were challenged by Elassar et al. (34) who reported no difference in either normal or poor responders. Guivarc’h-Lev^eque et al. (35) reported that luteal E2 could be prescribed for programming ART cycles and avoiding retrievals on weekends without any deleterious effects. The superiority of E2-based programming claimed by some (32), notably in poor responders, has been challenged in many publications. For example, an RCT showed that pretreatment with oral E2 from day 7 after ovulation to day 2 of the next cycle did not increase oocyte yield in patients with a low ovarian response compared with no pretreatment (36). Chang et al. (37) used a luteal E2 treatment protocol at the dose of 4 mg/d initiated on luteal day 21 and showed no difference in outcome. Similarly, Ye et al. (38) observed no significant effect on implantation, clinical pregnancy, live birth, and early pregnancy loss rates between E2 pretreated and untreated patients. A different retrospective study showed no difference in poor responders between E2 pretreated– transdermal patch– antagonist and the microdose GnRH-a protocol (39). Garcia Velasco’s group compared ART priming with luteal E2 and OC pill in a randomized controlled trial (40). One hundred consecutive patients undergoing IVF with the GnRH antagonist protocol were randomized to either the OCP or E2 pretreatment arms. No differences were observed in the fertilization rates (68.1% vs. 64.8%), the total number of embryos obtained (5.9 vs. 6.2), the mean number of embryos transferred (1.8 vs. 1.8), implantation rate (36% vs. 39%), miscarriage rate (8.9% vs. 17%), ongoing pregnancy rate (47.8% vs. 53.9%), or LBR (44.3% vs. 47%) (40). Recently, Cedrin-Durnerin et al. (41) reported a multicentric trial on 324 women who received pretreatment with micronized E2 (2 mg twice a day). Results were compared with 160 nontreated patients. There were no differences in the number of retrieved oocytes (8.4 [6.1] vs. 9.1 [6.0]), in the number of metaphase 2 oocytes (7 [5.5] vs. 7.3 [5.2]) nor the number of obtained embryos (5 [4.6] vs. 5.2 [4.2]) in E2 pretreated patients compared with non-pretreated controls. The LBR after fresh transfer (16.2%vs.18.5%, respectively), and the cumulative LBR per patient (17.7%vs.22.9%, respectively) were similar in both groups. In a subgroup of poor responders (antimullerianhormone<1.2ng/mLand/orantral folliclecount<5); however, there was a larger number of oocytes retrieved–1.7mean– but no difference in cumulative LBR(41). In all, programming ART cycles with E2 initiated in the luteal phase (or on cycle day one) can be safely executed for the benefit of ‘‘programming’’ ART cycles according to the group’s needs. Luteal E2 is effective for that purpose, but the expected advantage of luteal E2 programming for poor responders has not been proven despite early hopes. Programming ART cycles with either luteal E2 or other preparations such as the OC pill notably, are equivalent (42,43).

LUTEAL PHASE, PROGESTERONE PRIMED STIMULATION, AND DUO STIM PROTOCOLS

Luteal phase OS

The first report of OS conducted during the luteal phase came from China (44). In this early report, the ART outcome was similar to that of stimulations conducted in the follicular phase (44). Of course, embryos had to be cryopreserved and their transfer was deferred because the endometrium was not receptive. These findings were confirmed by Qin et al. (45) who found no differences in the mean number of mature oocytes retrieved in the conventional group, late follicular phase group, and luteal phase group (5.7 ± 3.6, 5.2 ± 3.7, and 5.2 ± 3.9, respectively). Similarly, no significant differences were observed in the viable embryo rate per oocyte retrieved (37.9%, 38.5%, and 43.6%), clinical pregnancy rates (41.5%, 45.5%, and 38.9%), and implantation rates (30.7%, 30.2%, and 27.1%) in the three groups (45).

A more recent study confirmed and extended these results (46). The latter data by Martinez et al. (46) consisted in conducting two identical stimulations in oocyte donors in the follicular and luteal phase looking at possible differences in the same patients. Follicular phase stimulation resulted in a significantly shorter duration of OS and a lower total dose of FSH was needed compared with luteal phase stimulation. The number of oocytes and blastulation rates were similar, however. Most importantly, the mean number of euploid blastocysts was equivalent in follicular and luteal phase stimulations at 1.59 and 1.61, respectively (46).

Of note, pituitary suppression using the antagonist was not necessary for luteal phase stimulation due to the antigonadotropin properties of endogenous progesterone (44). This observation has led to the development of progestin OS protocols, in which the use of antagonists to prevent premature ovulation is replaced by progestin treatment. These progestin stimulation protocols are often, but incorrectly, called progestin-primed protocol (PPOS). The term “priming” is indeed incorrect, as progestins are used to block ovulation and do not prime anything.

Progestin stimulation protocols

Results obtained in luteal phase stimulations unveiled that luteal phase progesterone protects against premature LH elevation, rendering the use of antagonists unnecessary (44). This led to testing whether exogenous progestins could replace GnRH antagonists if given during the follicular phase (47, 48). Of course, progestin protocols imply that all embryos are cryopreserved (freeze-all) and embryo transfers deferred (48).

Several comparisons have been conducted between the efficacy of progestin and regular stimulation protocols, including large meta-analyses by Ata et al. (49, 50). Ten studies compared progestin protocols to antagonist protocols and six to long agonist protocols. Most studies were of poor quality, originating mainly from one center in China, and ART outcomes were reported per transfer, not per retrieval. Ideally, cumulative pregnancy rates should be assessed, but this information is lacking (49).

Overall, the duration of stimulation, gonadotropin consumption, and oocyte yield were similar with progestins and GnRH analogs. Sensitivity analyses suggested that progestins were associated with significantly lower gonadotropin consumption than the long GnRH-agonist protocol. Live birth, ongoing, and clinical pregnancy rates per transfer were similar with progestins and GnRH analogs, confirming results from another review (51). Miscarriage rates were also similar (51).

Deng et al. (52) conducted a systematic review in women with PCOS. Eight studies including 2,156 PCOS women undergoing ART with progestin protocols showed similar ART outcomes and a very low incidence of OHSS compared to historical controls. Xi et al. (54) found no difference in ART outcomes between progestin and long agonist protocols but confirmed a lower OHSS risk in the progestin group.

Various types and doses of progestins have been tested, including medroxyprogesterone acetate (MPA), dydrogesterone (20 mg/day), and micronized progesterone. No significant differences were found in ART outcomes or miscarriage rates across formulations (50, 55–57). La Marca et al. (58) reported similar euploid embryo percentages and numbers per patient when comparing MPA protocols to GnRH antagonist protocols.

Newer flexible progestin protocols start treatment on day 6 or 7 of stimulation rather than day 1, similar to GnRH antagonist timing. Early results suggest potentially higher cumulative live birth rates, but confirmation is needed (59).

Vidal et al. (60) provided strong evidence of safety and efficacy: 44 women underwent two stimulation cycles, first with antagonist and second with oral micronized progesterone (200 mg/day), with all blastocysts analyzed by PGT-A. Outcomes, including euploidy rates, were similar, confirming the protocol’s reliability (60).

The main advantages of progestin protocols are simpler administration (no injections of antagonists) and lower cost. Evans et al. (61) estimated a $2,079 saving per cycle in the US in cases of planned freeze-all and deferred transfer.

In summary, progestin protocols are safe and effective, especially with freeze-all approaches for genetic testing, endometriosis, or high ovarian response.

Duostim protocol

A variant of luteal phase stimulation is DuoStim: dual consecutive stimulation in the follicular then luteal phase within the same cycle. Originally proposed by Capalbo et al. (62) for poor ovarian responders to maximize oocyte collection. DuoStim achieved similar euploid blastocyst formation rates. Stimulation protocols were identical for follicular and luteal phases, producing similar blastocyst numbers in reduced ovarian response patients.

Fresh embryo transfer can sometimes be done from the follicular phase harvest. DuoStim is well tolerated by patients, often perceived as a single extended OS cycle rather than two distinct stimulations (64). DuoStim reduces time to obtain euploid blastocysts while maintaining similar ploidy rates and obstetric outcomes (65–67). It is particularly useful for fertility preservation and poor responders in PGT-A cycles. Patient surveys indicate high interest in DuoStim when offered (68).

Random start stimulation

Random start stimulation allows initiation of OS at any menstrual phase, useful in urgent cases like oncologic fertility preservation (69–71). It shortens cycle duration without affecting oocyte yield or maturity (69). Encouraging results in oncology led to trials in standard ART patients, showing no differences in oocyte yield or clinical outcomes compared with conventional OS (72–73).

Random start protocols offer flexibility but carry OHSS risk in case of inadvertent pregnancy during stimulation (74). They are appropriate when freeze-all and deferred transfer is planned.

Conclusion

The traditional principle of starting OS in the early follicular phase has been challenged. Various programming options now exist, including pretreatment with contraceptive pills, luteal phase or cycle day 1 E2, random start protocols, and progestin-based stimulations (PPOS). All offer similar outcomes and can be chosen based on patient needs and clinical context.

L’étude complète est disponible en anglais ici :

https://www.fertstert.org/article/S0015-0282(24)02224-6/abstract