Choosing an IVF plan is no longer just about starting stimulation and hoping for the best: as Dr Kristýna Frühaufová, PhD, Head Physician at Gynem Fertility Clinic in Prague, made clear, the most effective treatment is built around how a patient’s ovaries, hormones, medical history, and prior cycles actually behave.
Hosted by: Clara Toombs, Founder of Fertilia AI.
The central aim of a personalised IVF protocol is straightforward: get a strong ovarian response without pushing the body too far. On paper, the target is usually 8 to 15 oocytes. That range is high enough to support embryo creation, but not so high that stimulation becomes counterproductive.
That balance matters because IVF can go wrong in both directions. Too few eggs may leave a patient with limited chances in a cycle. Too much stimulation can raise the risk of OHSS (ovarian hyperstimulation syndrome) and may also compromise egg quality. In this framework, more than 20 eggs is generally seen as excessive, and avoiding that kind of overresponse is part of good protocol design.
Personalisation also affects more than egg numbers. Overstimulation can raise estradiol too much and lead to a premature rise in progesterone, which may reduce endometrial receptivity. A carefully chosen protocol is therefore trying to protect both the ovaries and the uterine lining at the same time. It can also reduce side effects and avoid unnecessary medication costs when a patient is likely to respond well to lower doses.
Before deciding on a stimulation plan, clinics look at how much functional ovarian capacity appears to be present. Two of the most important tools are AMH and antral follicle count. Together, they help estimate how the ovaries are likely to respond to stimulation.
AMH gives a sense of expected sensitivity. Low AMH is associated with fewer expected eggs, while high AMH raises concern for hyperresponse. Antral follicle count, ideally measured in the early follicular phase, provides an ultrasound-based view of the small follicles available at the start of the cycle.
These two markers are most useful when they tell the same story. High AMH should generally match a higher antral follicle count, and low AMH should match a lower one. When they do not correspond, a nonstandard response to stimulation may be more likely, which is exactly the kind of situation where a standard protocol can miss the mark.
Other baseline hormones help refine the picture. Follicle-stimulating hormone and LH are used to guide both dose and medication composition, and estradiol helps interpret those values correctly. A high FSH level may point toward diminishing ovarian reserve or possible premature ovarian insufficiency, and it may also mean higher gonadotropins are needed. But FSH cannot be read in isolation because estradiol naturally rises through the cycle while FSH falls.
A protocol is not chosen from bloodwork alone. Previous cycles, prior medication response, ultrasound findings, and broader health history all shape the plan. If a patient has already had a good response to a certain stimulation type, there may be no reason to change it in the next cycle.
Body composition also matters. The distribution of body fat may influence how stimulant drugs are absorbed, and in overweight or obese patients, higher doses are often needed because fatty tissue affects how the medication is handled. Smoking is another practical factor because it reduces ovarian capacity.
This is one reason IVF planning can look very different from one patient to another, even when they are the same age. A person’s hormone profile, ultrasound pattern, weight, prior cycle history, and diagnosis may all point in different directions. The protocol has to account for the whole picture, not just a single test result.
Patients considered poor responders often have low AMH, low antral follicle count, a previously low oocyte yield, or advanced age. In these cases, the challenge is not preventing overresponse but getting enough follicles to develop without using medication in a way that adds burden without benefit.
A common strategy is stimulation that combines FSH and LH activity. Some patients may also receive pretreatment such as testosterone or DHEA, and some may use estradiol priming. In selected cases, duo stimulation—two consecutive stimulations within one menstrual cycle—may also be considered.
What stands out is that more medication is not always better. Doses above 300 units have not been proven to improve fertilisation or oocyte yield. For that reason, very aggressive high-dose stimulation is not necessarily the preferred answer. Mild stimulation protocols are often favoured instead, especially when ovarian capacity is already limited.
That distinction can be especially relevant in conditions linked to reduced ovarian reserve. In a patient with a Fragile X premutation, for example, ovarian capacity may be limited, and multiple IVF rounds may be needed. Mild stimulation may be the more realistic approach, and low AMH on its own does not determine whether a natural pregnancy is still possible. IVF may also offer a separate advantage in this setting because it allows embryo genetic testing, whereas natural conception does not.
At the other end of the spectrum are patients expected to respond very strongly, including many with PCOS or PMOS. These patients often have a characteristic ultrasound pattern and stimulation response, and the main concern becomes avoiding OHSS and excessive follicle recruitment.
The usual approach is to start with a lower FSH dose and increase it stepwise only if needed. For many first cycles, the preferred starting framework is an antagonist protocol, and that is especially useful in patients with high AMH because it gives the clinical team more control if the ovaries begin responding too strongly.
This is where dose selection becomes critical. In a lean patient with PCOS, even a dose that might seem ordinary on paper can be too much. A step-up strategy with lower dosing may better protect egg quality and reduce the risk of overstimulation. Body composition matters here too: a low BMI and high ovarian reserve call for a different approach than obesity with the same diagnosis.
The goal is not to collect the highest possible number of eggs. In patients already prone to lower egg quality, an excessive response can make the tradeoff worse. When the ovaries are producing very large numbers, lowering the dose in a future cycle may improve the overall result rather than reduce it.
Some diagnoses do not mainly affect the stimulation dose, but they still change how IVF should be planned. With endometriosis, especially when endometriomas are present on the ovaries, one practical question is whether follicles will still be accessible at egg retrieval. That can influence cycle planning even before stimulation starts.
Routine endocrine screening also matters. TSH and thyroid antibodies are checked because thyroid dysfunction, including autoimmune thyroid disease, may affect implantation. Prolactin should also be tested routinely because hyperprolactinemia is common, often silent, and can interfere with ovulation and endometrial receptivity. It can usually be managed with medication, and thyroid dysfunction and hyperprolactinemia may occur together.
These are the kinds of issues that can quietly undermine a cycle if they are missed. They may not be the headline diagnosis, but correcting them can be part of making IVF more efficient and more rational.
Personalised IVF increasingly includes discussion of genetics, but not every available test has the same clinical value. Some clinics use FSH receptor polymorphisms to evaluate response at the receptor level, and LH polymorphism testing is used less often. Even so, the practical usefulness of this testing remains uncertain. A patient expected to be highly sensitive or at risk of overstimulation may still respond normally in real life.
Immune testing is another area drawing attention, especially in patients with repeated implantation problems. Research is emerging around CARE receptors on natural killer cells in the endometrium and antigens on the embryo, with the idea that maternal immune recognition of the embryo may influence implantation. This is a promising area, but still an early one.
That uncertainty matters for patients deciding how much testing to pursue. In someone with repeated implantation failure despite tested embryos, immunological testing and CARE-related testing may be relevant. But these are not presented as standard answers for every IVF patient. They are part of a more selective, case-by-case approach.
One of the most important decisions late in stimulation is how to trigger final oocyte maturation. The two main options are hCG and an agonist trigger, and the choice depends heavily on the patient’s response.
hCG is the most common trigger in cycles with an optimal response. It is generally safe and works well in most patients. Its drawback is a higher risk of OHSS, especially when the follicle number is borderline high.
When that risk rises, an agonist trigger is used instead. The tradeoff is that agonist triggering has a negative effect on endometrial quality. Because of that, it cannot be paired with a fresh embryo transfer. It is mainly used in freeze-all cycles, followed later by cryotransfer.
This is a good example of how personalisation in IVF extends beyond the stimulation dose itself. A patient may begin with one goal and end up with a different transfer plan because the ovaries responded too strongly to make a fresh transfer the safest or most effective choice.
In some cases, a dual trigger or double trigger is used. This combines agonist triggering with hCG and may help improve oocyte maturation when there is a mismatch between follicle number and prior oocyte yield.
Patients often think of personalisation as something that happens before egg retrieval, but the transfer strategy is just as individualised. If a patient is approaching hyperresponse, the safer path may be freeze-all with later cryotransfer rather than a fresh transfer in the same cycle.
That logic also appears in patients with medical conditions that complicate implantation or pregnancy management. In a patient with lupus, for example, cryotransfer may be preferred, and treatment may include baby aspirin at 75 or 100 milligrams, blood thinners throughout pregnancy, and low-dose prednisolone. But corticosteroids are not used routinely in everyday IVF anymore. They are reserved for situations where they are truly needed because they carry risks, including preeclampsia, and are generally used only for a limited period, usually up to 8 weeks.
Even when embryos have already been genetically tested, transfer success is not guaranteed. In patients with tested embryos and repeated failed implantation, the success rate still does not exceed 60%. That gap is one reason clinicians continue to look at endometrial, immune, and compatibility factors rather than assuming embryo testing alone has solved the problem.
Although IVF planning often centres on ovarian response, sperm factors can shape outcomes too. Sperm morphology matters because lower morphology tends to reflect lower DNA quality, and that can become one reason to consider embryo genetic testing, such as PGT-A.
When sperm concentration is sufficient, Zymot may be used to address DNA fragmentation. Testing for sperm DNA fragmentation or oxidative stress can also help identify whether supplementation might improve sperm quality. Andrological assessment matters for broader reasons as well, since low morphology can be seen with varicocele and may reflect hormone-related or general health issues.
This broader view is part of what personalisation really means. If the focus stays only on the woman, an important part of the fertility picture may be missed.
Personalisation can improve the fit between treatment and patient, but it does not erase the effect of age. IVF couples are getting older, with the average age now around 38 compared with the early thirties about 5 years ago. That shift affects egg quality and overall success.
In practical terms, this means protocol selection can optimise response, reduce risk, and avoid waste, but it cannot fully compensate for age-related changes in the eggs. That is also why some patients may be advised to address reversible issues before moving straight into IVF. Weight loss in overweight or obese patients may restore cycles and allow natural conception if no other factors are present. The same principle applies when an andrological factor may be improved first.
Personalised IVF is less about finding a perfect formula than about avoiding obvious mismatches. The right protocol takes into account AMH, antral follicle count, baseline hormones, body composition, diagnoses such as PCOS or endometriosis, endocrine problems like thyroid disease or hyperprolactinemia, prior cycle response, and sometimes selected genetic or immune factors. It also extends to trigger choice and whether a fresh transfer or freeze-all approach makes more sense.
That kind of tailoring cannot guarantee success, but it can make treatment safer, more efficient, and more realistic. For patients considering IVF, the most useful question is not whether there is a “best” protocol in general. It is whether the plan in front of you actually matches your ovaries, your medical history, and the way your body is likely to respond.
With Fragile X, ovarian capacity is limited, so the expected oocyte yield may be lower, and multiple IVF rounds are often needed. Mild stimulation would most likely be the preferred approach, and the antral follicle count is also important when planning treatment.
PRP might help boost follicle growth, and the time after PRP can be used for back-to-back stimulations. Estradiol priming mainly helps prevent asynchrony of follicular growth and would not increase ovarian response. Another option is androgen pretreatment, but that usually takes up to 3 months to show an effect.
Natural pregnancy may still happen, and low AMH alone does not determine whether pregnancy can occur naturally. If needed, more aggressive progesterone supplementation can be used. The key difference is that natural conception does not allow genetic testing of the embryo, which is one of the roles of IVF.
A premutation can expand into a full mutation over generations, but the infant risk is almost none in this situation. The female line may still face infertility issues, but the risk of mental retardation in male offspring is not as elevated as with the full mutation.
With a very high AMH and age 40, a low-dose FSH antagonist protocol is appropriate. A step-up approach with lower doses is preferred, especially with a BMI of 20.
A starting dose lower than 225 should be used, and even 150 to 175 should be considered maximum because the BMI is 20. A dose of 225 raises concern for overstimulation and lower egg quality.
PGT-A is strongly recommended at age 40 because of the risk of abnormalities in the eggs and embryos. For transfer, cryotransfer is preferred.
With lupus, treatment is usually baby aspirin, 75 or 100 milligrams, plus blood thinners throughout pregnancy. Low-dose prednisolone is also added.
Prednisone is used in many autoimmune conditions. In lupus, it is intended to help ensure there are no circulating antibodies by the time of placental formation. The dose is usually continued until week 8 and then stopped.
Corticosteroids are not used routinely for every IVF patient. They should be used only when truly needed and for a limited period of time, usually up to 8 weeks, because by then the placenta should be developed. If another condition requires corticosteroids, close specialist monitoring is needed. Otherwise, the risks of continuing the medication outweigh the benefits. Corticosteroid use is associated with risks such as preeclampsia during pregnancy.
Implantation depends on many factors, and there is still a large unknown component behind successful implantation. Even with PGT-A testing, the success rate does not exceed 60%.
If there has been one implantation that stopped at 6 weeks and 3 failed IVF cycles with tested embryos, despite a 10 millimetre smooth lining and clear ultrasound, this can fit repeated implantation failure. Immune and genetic compatibility between the embryo and the endometrium may play a role. CARE gene testing may be relevant because there may be incompatibility between the egg and sperm. Immunological testing is also recommended.
Sperm morphology matters because lower morphology tends to reflect lower DNA quality in the sperm. It can be one of the indications for embryo genetic testing, such as PGT-A.
If sperm concentration is sufficient, Zymot is recommended. It is a microfluidic chip, but it requires a certain sperm concentration to work properly. The XE chip uses the same basic methodology and works with a lower concentration.
Testing sperm DNA fragmentation or oxidative stress may help identify whether supplementation could improve sperm quality. Low sperm morphology can be seen with varicocele, and men should also have andrological testing because the issue may involve hormone production and broader health consequences.
That depends largely on how much time the couple has. Age remains one of the key factors for IVF success, and the average age of IVF couples has shifted to around 38 compared with the early thirties about 5 years ago.
If a couple is young and otherwise healthy, it makes sense to take precautions before IVF. The same principle applies when an andrological factor is suspected: first look for possible ways to improve the situation. In some cases, no specific cause is found and IVF remains necessary.
The same logic applies more broadly in fertility care. For example, in young overweight or obese women, weight loss may be the first recommendation instead of proceeding directly to IVF stimulation, because cycles may return to normal and natural conception may happen if no other factors are present.
A very high retrieval number can suggest slight overstimulation, so lower stimulation doses are recommended. If an antagonist protocol has already been used, some cases are switched to a long protocol, but with very cautious dosing.
If more than 60 eggs have been retrieved over multiple cycles but only one 4BB blastocyst and one day 3 transfer resulted, the stimulation dose is likely too high. Zymot is useful for DNA fragmentation and is the best tool available for the andrological factor. If only one blastocyst is available, there is no role for genetic testing in that cycle.
Similarly, if 37 eggs were retrieved, 100% were mature on 225, there was no OHSS, and 6 blastocysts developed from 28 fertilised eggs; that is still an impressive result. Even so, the high egg number suggests slight overstimulation. The preferred adjustment would be lower stimulation doses while still using the antagonist protocol. If there is no male factor to explain poor blastocyst development, a lower dose could potentially give more embryos.
Day 6 blastocysts generally have lower implantation chances than day 5 blastocysts. Statistically, the main focus should still be on dose selection, since that may help.
Artificial intelligence may help in both the laboratory and clinical parts of IVF, but at present, the experience of the IVF professional is still the main factor. Computerised systems that combine AMH, antral follicle count, and dose suggestions could be useful, especially for young doctors who are learning.
Dose calculators already exist for specific medications such as follitropin delta. They are most useful when there is limited information, such as only an ultrasound report, patient photos, basic body weight, and BMI, rather than a full in-person assessment. Even then, the dose still needs adjustment, because real-world experience suggests the calculated dose is usually lower than what is needed.
Over the next 10 to 20 years, the biggest challenge will still be delayed pregnancy planning, because age affects egg quality and IVF success. Research is ongoing to improve egg quality and possibly reverse egg ageing, but this is still experimental and has not been done in humans yet. Better fertility education and earlier awareness of the consequences of delaying pregnancy would do a great deal to improve IVF outcomes.
