The Process of Controlled Ovarian Stimulation

Controlled ovarian stimulation (COS) is the hormone-driven process of coaxing a woman’s ovaries into maturing multiple eggs in a single menstrual cycle, rather than the one egg that would normally develop on its own. It forms the foundation of in vitro fertilization and is also used in egg freezing and some forms of fertility preservation. The process typically spans about ten to fourteen days and involves daily injections, regular monitoring visits, and a precisely timed “trigger” injection that prompts the final step of egg maturation before retrieval. While the basic idea is straightforward, the specific drugs, dosing strategies, and safety considerations vary enormously depending on the individual.

Why the Body Normally Releases Just One Egg

Each month, a group of small follicles in the ovaries begins to grow under the influence of follicle-stimulating hormone (FSH). Under natural conditions, one follicle becomes dominant, producing enough estrogen to signal the pituitary gland that it can stop pushing the others along. The remaining follicles shrivel. The dominant follicle then receives a surge of luteinizing hormone (LH), which triggers ovulation of its single mature egg.

COS overrides this selection process. By injecting FSH (and sometimes LH) at doses higher than the body would normally produce, doctors keep multiple follicles growing simultaneously instead of letting one winner take all. At the same time, a second class of drugs prevents the brain from firing that premature LH surge, which would trigger ovulation too early and ruin the cycle. That two-pronged approach, stimulating growth while suppressing the natural trigger, is the core logic behind every COS protocol.

Predicting Ovarian Response Before You Start

Before prescribing a single injection, clinicians try to estimate how aggressively the ovaries will respond. Two markers dominate that assessment: anti-Müllerian hormone (AMH), a blood test reflecting the pool of small follicles in reserve, and antral follicle count (AFC), the number of small follicles visible on ultrasound early in the cycle. A systematic review and meta-analysis found that both markers performed similarly overall in predicting whether someone would be a poor or high responder, with no significant difference between their summary curves for either category.¹

For the specific concern of hyper-response, where the ovaries produce a dangerously large number of follicles, a consensus study identified threshold values: an AMH of roughly 4.4 ng/mL or above and an AFC of 16 or more flagged increased risk in the overall cohort, with slightly different cutoffs for women under and over 35.² These numbers help doctors choose a starting dose, but they are guides, not guarantees. Age, body weight, the type of gonadotropin, and even the trigger strategy all contribute to how many eggs ultimately develop.

The Medications That Drive Stimulation

The workhorse drugs are gonadotropins, injectable forms of FSH sometimes combined with LH activity. They come in two broad families: recombinant products manufactured in cell lines, and urinary-derived products extracted and purified from postmenopausal urine (historically called human menopausal gonadotropin, or HMG). The practical question patients often ask is whether one type is better than the other.

A large Cochrane review covering 59 studies and more than 18,000 women found that live birth rates were modestly lower with recombinant FSH alone compared to HMG or highly purified HMG, and that the risk of ovarian hyperstimulation syndrome (OHSS) was higher with recombinant FSH.³ The differences were moderate in size, so many clinics still use recombinant FSH as a default and reserve HMG-based protocols for specific situations. The choice often comes down to a combination of cost, availability, and the clinician’s reading of the patient’s hormone profile.

Agonist Versus Antagonist Protocols

Preventing a premature LH surge is the second half of the COS equation, and two drug classes handle it differently. In a long GnRH agonist protocol, the patient starts a nasal spray or daily injection weeks before stimulation begins. The agonist initially stimulates the pituitary and then, after about ten days, shuts it down through a phenomenon called downregulation. Stimulation injections begin once blood tests confirm the pituitary is suppressed. In a GnRH antagonist protocol, the patient starts gonadotropin injections right away and adds the antagonist only once follicles reach a certain size, typically around day five or six. The antagonist blocks the pituitary’s LH release almost immediately, so there is no need for the long lead-in period.

A systematic review accounting for patient type found that in general IVF patients, ongoing pregnancy rates were modestly lower with the antagonist protocol compared to the long agonist protocol. However, in women with polycystic ovary syndrome (PCOS) and in women with poor ovarian response, no meaningful difference in pregnancy rates emerged between the two approaches.⁴ The trade-off is safety: antagonist protocols consistently produce lower rates of OHSS, both in general IVF populations and in women with PCOS.⁴ A separate meta-analysis focused on PCOS patients confirmed this, reporting a roughly 40% reduction in OHSS risk with the antagonist approach.⁵

Because of this safety advantage, antagonist protocols have become the dominant choice worldwide, particularly for patients at elevated OHSS risk. The long agonist protocol still has advocates who value its more complete suppression of premature LH surges and its slightly higher pregnancy rates in certain patient profiles, but the trend in practice has clearly shifted toward the shorter, safer antagonist approach.

Monitoring During Stimulation

Once injections begin, you will visit the clinic every two to three days for transvaginal ultrasound and often a blood draw. The ultrasound measures the number and size of growing follicles, while the blood test tracks estradiol, the estrogen produced by those follicles. As follicles grow, estradiol rises in a predictable relationship: for every additional millimeter of average follicle diameter, serum estradiol climbs by roughly 8 ng/L.⁶

Whether those blood draws are strictly necessary has been debated for years. A Cochrane review noted that monitoring with ultrasound alone, without blood estradiol levels, has been proposed as a simpler and less burdensome approach, and the evidence has not clearly shown that adding blood tests improves outcomes for most patients.⁷ Still, many clinics continue to use combined monitoring, especially for patients at higher risk of OHSS or those with unusual hormone patterns. In practice, the estradiol levels serve as an early warning system: if they climb too fast relative to the follicle count, the team may reduce the gonadotropin dose or plan for a safer trigger strategy.

The Trigger Shot and Final Egg Maturation

When lead follicles reach about 17 to 18 millimeters and the estradiol level corresponds to the number of mature-sized follicles, the doctor schedules the trigger injection. This replaces the natural LH surge and kicks off the final maturation process that loosens the egg from the follicle wall so it can be retrieved roughly 34 to 36 hours later.

The traditional trigger is human chorionic gonadotropin (hCG), usually given as a single injection of 5,000 to 10,000 IU.⁸ hCG mimics LH because it binds the same receptor, but its half-life is much longer, exceeding 24 hours compared to about one to two hours for natural LH. That prolonged activity is a double-edged sword. It ensures thorough egg maturation but also sustains the corpus luteum aggressively, which can drive the ovaries into hyperstimulation in susceptible patients.

The alternative, available only to patients on an antagonist protocol, is a GnRH agonist trigger. A single dose of GnRH agonist causes the patient’s own pituitary to release a burst of both LH and FSH, more closely mimicking the natural mid-cycle surge. Because this burst is brief and self-limiting, OHSS risk drops dramatically. A study in PCOS patients undergoing IVF found that the agonist trigger group also had better embryo quality, with fewer poor-quality embryos and a higher number of top-quality embryos on day three, likely because the shorter LH exposure avoided over-stimulating the cells that surround each egg.⁹ The catch is that the same brief surge that protects against OHSS also weakens luteal phase support, so a fresh embryo transfer after an agonist trigger typically requires either intensive progesterone supplementation or a freeze-all strategy, where all embryos are frozen for transfer in a later cycle.

Ovarian Hyperstimulation Syndrome

OHSS is the most feared complication of COS. It occurs when the ovaries over-respond and become swollen and leaky, allowing fluid to shift out of the bloodstream into the abdomen and sometimes the chest cavity. The underlying driver is vascular endothelial growth factor (VEGF), a protein found in high concentrations in the follicular fluid of hyperstimulated ovaries. Research has shown that follicular fluid from patients who develop severe OHSS contains roughly three times the VEGF concentration of fluid from patients who respond normally, and that blocking VEGF with an antibody reverses the vascular leakiness almost entirely.¹⁰

Clinically, OHSS prevention has improved substantially. For patients identified as high risk, the combination of an antagonist protocol with a GnRH agonist trigger followed by a freeze-all policy has virtually eliminated severe OHSS in some centers. One retrospective study using this approach reported zero cases of OHSS, either early or late onset.¹¹ A broader review of the strategy confirmed that live birth rates remain high in the subsequent frozen embryo transfer cycle, meaning safety does not come at the expense of success.¹²

Progestin-Primed Ovarian Stimulation

A newer approach called progestin-primed ovarian stimulation (PPOS) sidesteps both the GnRH agonist and antagonist entirely. Instead, the patient takes an oral progestin, such as medroxyprogesterone acetate or dydrogesterone, alongside the gonadotropin injections from the start of the cycle. The progestin reliably blocks the LH surge through direct action on the pituitary, making the cycle simpler and cheaper by eliminating the need for a separate suppression drug.¹³ The mechanism behind this suppression is timing-dependent: progesterone administered during the follicular phase, before the brain has committed to generating an LH surge, inhibits the surge rather than facilitating it.¹⁴

A study comparing PPOS to the traditional long agonist protocol in women with normal ovarian reserve found that PPOS required fewer days of stimulation and a lower total gonadotropin dose, with no significant difference in clinical pregnancy rates, early miscarriage, or ectopic pregnancy. No cases of severe OHSS occurred in the PPOS group, compared to 11 cases in the long agonist group.¹⁵ The major limitation is that the progestin makes the uterine lining unreceptive during the stimulation cycle, so a fresh transfer is not an option. All embryos must be frozen. For clinics that have already moved toward a routine freeze-all approach, this limitation matters little.

When Ovarian Response Is Poor

Not everyone responds robustly to stimulation. Poor ovarian responders, typically older women or younger women with diminished ovarian reserve, may produce only a handful of eggs. Choosing the right protocol for these patients is one of the more frustrating challenges in reproductive medicine, because the differences between protocols tend to be small and inconsistent.

A study examining different protocols in poor responders classified by standardized criteria found that an early-follicular-phase long-acting agonist protocol was associated with better pregnancy and live birth rates in younger poor responders, but that no protocol showed a clear advantage in older patients.¹⁶ Another study found that the antagonist protocol offered some advantage in one subgroup of younger poor responders, while the agonist protocol performed better in the oldest and most depleted subgroup, though live birth rates did not differ significantly overall between the two approaches.¹⁷

For patients with very low egg yields, a strategy called DuoStim offers a creative workaround. Instead of stimulating once per cycle, the patient undergoes a second round of stimulation in the luteal phase of the same menstrual cycle. A study in women with reduced ovarian reserve found that the number of retrieved eggs and euploid blastocyst rates were similar between the follicular-phase and luteal-phase stimulations, meaning the second round effectively doubled the number of embryos available for transfer within a single month.¹⁸

Does the Stimulation Regimen Affect Embryo Quality?

A common worry is that pushing the ovaries harder with higher drug doses produces lower-quality eggs. The evidence here is more reassuring than many patients expect. A study examining aneuploidy and blastocyst formation found no difference in aneuploidy rates or cumulative live birth rates between women who received lower versus higher total gonadotropin doses. In fact, among younger women, a higher proportion of HMG in the protocol was associated with a lower aneuploidy rate.¹⁹ However, supraphysiological estradiol levels on the day of the trigger can impair uterine receptivity, creating a paradox where the ovaries produce plenty of good embryos but the uterine lining is less ready to accept one.²⁰ This is one reason freeze-all cycles, which separate egg retrieval from embryo transfer by at least one menstrual cycle, have become more popular: transferring into a naturally prepared uterus avoids the receptivity problem entirely.

Individualized Dosing

The idea of tailoring the gonadotropin dose to each patient’s ovarian reserve markers sounds intuitive, but the evidence for its impact on the outcome that matters most, live births, is surprisingly weak. A large multicenter trial comparing individualized FSH dosing based on ovarian reserve testing to a standard-dose approach found no difference in live birth rates between the two strategies. Where individualized dosing did help was in safety: it produced more normal responses and fewer hyper-responses, reducing mild and moderate OHSS events.²¹ A smaller randomized trial using an AMH-based dosing algorithm in an antagonist protocol reached a similar conclusion: individualized dosing did not increase the proportion of patients with the target number of eggs.²²

In practice, this means clinicians still use AMH and AFC to set a starting dose, and they absolutely should because it reduces complications, but they should not promise patients that a personalized dose will translate into more babies. The body’s response to gonadotropins involves dozens of variables beyond what a blood test can capture.

Luteal Phase Support After Retrieval

After egg retrieval, the structures left behind by the emptied follicles (the corpora lutea) would normally produce progesterone to prepare the uterine lining for implantation. In a stimulated cycle, particularly one using a GnRH antagonist or agonist trigger, the luteal phase is disrupted. The supraphysiological estrogen levels generated during stimulation suppress the pituitary in a way that weakens the corpora lutea’s natural progesterone output.²³

Supplemental progesterone is therefore a near-universal part of any fresh transfer cycle. Both vaginal and intramuscular progesterone have comparable success rates, and both outperform placebo. hCG injections can also support the luteal phase and produce even higher pregnancy rates, but they reintroduce OHSS risk, so progesterone is the standard choice.²³ Patients doing a freeze-all cycle skip luteal support entirely during the retrieval cycle, since no embryo is being transferred; they begin progesterone before the future frozen embryo transfer instead.

Random-Start Stimulation for Fertility Preservation

For cancer patients or others who need to freeze eggs urgently, waiting for the next menstrual period to begin stimulation can feel like an unacceptable delay. Random-start protocols allow stimulation to begin at any point in the cycle, including the luteal phase. A study of breast cancer patients found that the average time spent in ovarian stimulation was about 13 days and that the time from cancer diagnosis to the start of chemotherapy was virtually identical between patients who underwent egg freezing and those who did not.²⁴ This has been a meaningful shift in oncofertility, removing one of the main objections oncologists used to raise about referring patients for egg freezing before treatment.

What It Feels Like

The physical and emotional burden of COS is not trivial. A survey of patients undergoing stimulation found that more than half said the treatment affected their daily life, and about a third felt that the daily injections limited their everyday activities. Over half of patients reported concerns about the injection process itself, and the most common questions directed at medical staff involved side effects from the fertility drugs.²⁵ The experience is typically described as a mix of bloating, mood swings, injection-site discomfort, and the logistical inconvenience of frequent clinic visits, all layered on top of the emotional weight of a high-stakes medical process.

A validated patient-reported outcome tool designed specifically for COS identified four main domains of impact: interference with daily life at work and home, the burden of injections, psychological health, and worry about whether you are doing the injections correctly.²⁶ That last domain, compliance worry, is worth noting. Fertility experts in the same survey expressed concern about whether nearly half their patients were injecting correctly. In a process where timing and dosing precision matter, the anxiety that patients feel about making a mistake is both understandable and, to some degree, clinically relevant.

A Brief History of How We Got Here

The drugs and protocols patients use today represent nearly a century of incremental advances. The story begins in 1927 with the isolation of hCG from the urine of pregnant women. By the 1930s, scientists had figured out how to superovulate laboratory mice using sequential gonadotropin injections, establishing the basic principle. The first report of ovulation induction in a human patient using pituitary-derived gonadotropins came in 1958, and by 1964, gonadotropins extracted from human menopausal urine were available in clinically useful quantities.²⁷ The first pregnancy achieved with recombinant FSH, the lab-manufactured version that avoided the messiness of urine extraction, was reported in 1992.²⁷ The shift from urinary to recombinant products, and from agonist to antagonist suppression protocols, happened over the 1990s and 2000s. The newest addition, PPOS, has gained traction primarily in the past decade and reflects a broader trend toward simpler, cheaper, and safer stimulation strategies that accept a freeze-all approach as the default.