Portuguese original (keep this link): https://revistapahnorama.com.br/2026/09/18/a-industria-da-fiv/
Isabel de Fatima Alvim Braga, MD phD
In vitro fertilisation is the technical name of a market. On one side, women and couples who tried and did not succeed — or who postponed pregnancy and now buy time in the freezer. On the other, clinics, high-cost pharmacies, nitrogen tanks and a language that mixes medicine and shop window. In the middle, a body that takes an injection every day and signs a form almost no one reads to the end.
No one needs to deny that the technique has given children back. What needs refusing is the brochure. Three warnings for the reader. Association is not cause. Relative risk frightens; the absolute is sometimes one extra case in a thousand. The IVF patient is, on average, older and more medicalised: part of the “effect of the technique” is the effect of age, of twins, and of the woman who was already not conceiving.
What a woman buys when she buys a cycle
A private cycle in Brazil usually costs in the region of R$15,000 to R$30,000 without the drugs — the South-East generally above the North-East. The hormone that makes the ovary ripen several follicles in the same month is not a detail. A 900 IU Gonal-F runs around R$3,200 to R$3,340; a 1,200 IU Menopur starts near R$2,500. Add the trigger (Ovidrel and the like), the laboratory, sometimes genetic testing of the embryo, time-lapse, the “artificial intelligence” that picks a blastocyst. Every extra has a scientific name and a price. Not every extra has proof that more children are born because of it. In the United Kingdom there is a traffic light for those extras. In Brazil there is not. The person who is supposed to declare ties to industry is the doctor — CFM Resolution 2.386/2024. In practice, the person who indicates the cycle bills the cycle, the freezer and the extra.
The law treats this contract, as a rule, as an obligation of means. The clinic owes technique and information. It does not owe the baby. Advertising works with a cumulative rate and not with the rate of that woman, at that age, in that started cycle. The difference between the two sentences is the difference between consent and illusion.
The leaflet the patient leafs through after the first injection
The stimulation drugs are not vitamins. They are gonadotrophins — copies of hormones the pituitary would use sparingly. The clinic hands over a pen. The leaflet, if read, describes another country.
Gonal-F (follitropin alfa, recombinant FSH). Very common or common effects on the leaflet: ovarian cyst, headache, pain and swelling in the belly, nausea, reaction at the injection site. The warning that matters has a name: ovarian hyperstimulation syndrome. The follicles swell too far, the vessel becomes leaky, fluid goes into the belly and the chest. The severe form is uncommon; the mild and moderate forms are common. Rare complications of the severe form: ovarian torsion, clot, embolism, stroke, infarction. The leaflet also says that thromboembolic events can occur even without hyperstimulation, especially in someone who has already had thrombosis, is obese or has thrombophilia in the family. Severe allergic reaction, including anaphylaxis, is very rare — and it is written down.
Menopur (menotrophin: FSH and LH extracted from menopausal urine). The same territory: hyperstimulation, multiple pregnancy if the protocol is careless, abdominal pain, headache, reactions at the injection point. The extra LH changes the profile of the stimulus; it does not remove the risk that the ovary will go past the point.
Ovidrel (choriogonadotropin alfa, recombinant hCG) is the trigger that tells the follicle to release the egg. On the leaflet, hyperstimulation appeared in about 4 percent of trial patients; the severe form, in less than 0.5 percent. Nausea, vomiting and abdominal pain travel with it and rise with the dose. Thromboembolism: very rare. Hypersensitivity, including shock: very rare. The clinical detail the leaflet insists on: severe hyperstimulation almost only explodes when the hCG goes in. That is why modern protocols sometimes swap the trigger for a GnRH agonist. Not every clinic swaps.
The patient hears “normal swelling.” The leaflet describes shortness of breath and urine that disappears. The distance between the two sentences is real consent.
Accreta and pre-eclampsia: the price of freeze-all
Two outcomes where the protocol weighs. Placenta accreta is the placenta that will not detach because it has invaded the wall too far. In spontaneous pregnancy: about 0.1 percent of births. In IVF with a fresh embryo: 0.16 to 0.2 percent. With a frozen embryo returned in a cycle in which the woman ovulated, the risk looks like the fresh cycle. In a cycle of pharmacy oestrogen and progesterone only, with no ovulation, no corpus luteum, the incidence rises to 1.4 to 4 percent.
Japan published the cold cut. In 174,591 singleton births, the programmed hormone cycle had an adjusted odds ratio of 9.12 against fresh (interval 6.54 to 12.73). In numbers: 1,549 cases in 104,834 programmed deliveries (1.48 percent) against 39 in 23,827 fresh (0.16 percent). In vaginal birth after a single embryo, the hormone cycle made 1.4 percent against 0.11 percent in the natural cycle (aOR 11.4). Endometrium thinner than 6 mm: 16 percent accreta against 3.8 percent with a thicker lining. Two or more uterine operations: aOR 3.57. Blastocyst: aOR 2.89. Placenta praevia in ART versus spontaneous, in singleton pregnancy: OR near 3.8.
Pre-eclampsia: fresh transfer, without twins, does not show a consistent excess. Frozen transfer raises hypertensive disorders by 74 percent in the population analysis. In the sibling control — the same woman, one natural pregnancy and one FET — the risk almost doubled. It was not “the type of couple.” It was the protocol. The Society for Maternal-Fetal Medicine in the United States treats IVF as a moderate factor; low-dose aspirin comes in if there is another factor.
Without a corpus luteum the vascular profile of ovulation is missing. The placenta sticks. The pressure rises. Freeze-all, sold as a rested uterus, meant a uterus that did not ovulate.
Support — placenta accreta spectrum
| Conception | Incidence | Reading |
| Spontaneous | ~0.1% | Baseline |
| Fresh IVF | 0.16–0.2% | Modest excess |
| FET with ovulation | Near the fresh cycle | There is a corpus luteum |
| FET hormone-only | 1.4–4% (aOR 9.12) | Here the technique weighs |
Ovarian cancer: what can be said without putting the statistic on trial
Every time IVF and cancer are spoken in the same sentence, the sentence wants to be born whole: “the hormone causes the tumour.” The literature does not authorise that sentence. It authorises another, duller and more honest.
Women who go through ovarian stimulation are already, on average, a different group: nulliparous or with few children, endometriosis, unexplained infertility, higher age. Nulliparity and endometriosis already associate with ovarian cancer in the population that never saw an FSH pen. Separating the drug from the biography is the work almost no large Brazilian study has done.
Reviews and metas find, with some consistency, an excess of borderline ovarian tumour after ART, with odds ratios in the region of 1.5 to 2 in several cuts. For invasive carcinoma the signal oscillates: it vanishes in some adjustments, remains weak in others. The American Society for Reproductive Medicine, in a 2024 note, treats the absolute increase as small and causality as not established. That is not innocence. It is a refusal to turn an odds ratio into a destiny.
What ethics asked for, and the protocol rarely delivers, is to look at the ovary as an organ before using it as a shelf: an ultrasound that describes a mass, a family history, a mutation when the history asks. Stimulating an ovary that already had a complex cyst is not “an IVF side-effect.” It is a failure of work-up. Stories of women who found a tumour after a cycle exist and deserve a hearing. A case is not a cohort. The report can tell the case and refuse the sentence.
The harm that does not appear in the pregnancy rate
Infertility is already an ambiguous grief: a child is lost who never received a name. IVF organises that grief into a monthly invoice. Each cycle opens a window of hormonal hope and closes with a negative beta or an empty sac. Anxiety and depression rise during treatment. Quality of life falls. The literature describes a roller-coaster, isolation, guilt — above all the woman’s, on whom the culture still pours the “she cannot.”
The marriage does not come out unharmed. It is not true that “people who do IVF divorce more” as a law of nature. An analysis of the American national survey of families found a lower risk of divorce in those who used treatment — probably because the couple that reaches the clinic has already survived a selection: they have money, they have a project, they have the minimum of alliance to bank the protocol. Another design shows the opposite in the cut that hurts: among 47,515 Danish women evaluated for infertility, those who had no child after the evaluation had up to 3.13 times more chance of ending the union in the first year, and the excess lasted twelve years. Ten years after IVF, in a classic follow-up, 56 percent of the women had thought of divorce and 17 percent had ended the relationship. In a treatment cohort, 14 percent separated and formed a new union along the way; those who separated already arrived with more stress.
The honest reading: treatment does not invent the conflict; treatment without a child exposes it and accelerates it. The embryo in the tank, when the couple splits, becomes a lawsuit. The consent signed under hormone becomes the clause the court reads.
The inconvenient question
Before the clinic there is the Sunday table. “And the baby?” “Have you thought about it?” “You are not going to leave it for later.” The question falls on a childless couple as if the only possible story were carelessness or selfishness. No one asks whether the tube closed, whether the semen does not walk, whether there were three losses, whether the woman buried an eight-week sac last year. Infertility is an intimate disease. Other people’s curiosity is public.
That inconvenience is not a detail of etiquette. It is social pressure that pushes people into high complexity before low investigation — and that shames whoever decides to stop. Industry benefits from the embarrassment. The child becomes a certificate of adult life. Whoever does not present one owes explanations to relatives who will not pay for the Gonal-F.
The report can return the question: why must the childless couple justify itself, and the clinic that sells a cycle without investigating chlamydia does not?
The tube is an immune organ — and IVF dismisses it
The Fallopian tube is not a pipe. It is living mucosa, with dozens of cell types, cilia, secretion, macrophage, lymphocyte and a paradox: it must tolerate allogeneic sperm and a half-strange embryo without failing to kill gonococcus and chlamydia. Some describe immune privilege in the lumen: TGF-beta and interleukin-10 all the time in the epithelium, a Th2-type response, few lymphoid follicles. The uterus, when semen arrives, summons the neutrophil. The healthy tube turns the volume down.
In that corridor the spermatozoon is chosen. Of tens of millions, a few hundred reach the ampulla. The tube tests shape, swimming, head, DNA. Sperm with fragmented DNA lights a Toll-like receptor in the epithelium. Sound sperm open pathways that prepare fertilisation and calm inflammation. The tube also stores, capacitates, pushes the egg, feeds the pre-embryo. None of that fits on a Petri dish.
When IVF punctures the ovary and joins gametes in the laboratory, the tube is retired. Sometimes that is necessary — closed tube, hydrosalpinx, ligation. Often high complexity begins because the clinic’s diary begins with it. Uninvestigated chlamydia, unmapped tubal factor, simple induction not tried. Dismissing an immune organ without knowing whether it was still working is not modernity. It is a shortcut.
The egg chooses. IVF chooses for it
Popular biology tells the heroic sperm and the motionless egg. The literature of recent years describes another scene. The cumulus–oocyte complex releases chemoattractants — picomolar progesterone, defensins, CCL20. Only a subpopulation of sperm responds. Follicular fluid from different women consistently attracts semen from different men: there is matching, not a single queue. It is cryptic female choice at the level of the gamete. It does not necessarily reinforce the couple the person chose on the app.
On the membrane the egg exposes Juno; the sperm, Izumo. Without Juno there is no fusion. After the first fusion, Juno is shed — a rapid block to polyspermy. The zona pellucida hardens afterwards. There is still the whole tract before that: mucus, swimming against the flow, reserve in the tube.
Classic IVF already shortens that track. ICSI cancels it: a needle pushes in a spermatozoon chosen by the embryologist — often the one that swims less, because the one that swims more has already been used in conventional IVF. Chemotaxis goes. The zona as judge goes. The tube goes. The laboratory gains a fertilisation rate. The species loses a filter of millions of years. That is not a mystical argument. It is a mechanism. The ethical question is whether the couple hears that this filter exists and that it is being switched off.
Froze the egg, guaranteed the child? The failure rates
The marketing of elective freeze speaks of autonomy and career. The ledger speaks of stock that does not come back. Among American women who froze eggs in 2014–2016 for planning, 5.7 percent thawed in five to seven years. Of those, 28.9 percent had a live birth. Two children, more or less, for every hundred freezes in that window. In 47,000 patients of planned preservation, return was 2.5 to 3 percent. In Spain, mean age 37.7 years: 9.3 percent came back; 2.7 percent of those who froze had a child from those eggs. Cost-effectiveness models only close with use of 50 percent or more. Real medium-term use: 3 to 15 percent. With more years, some centres rise to 15–25 percent. It remains a minority.
Eight or nine in every ten vitrified eggs survive thaw. Surviving is not being born. Chance that one mature egg becomes a child: about 6 to 8 percent if the woman froze before 35; 2 to 3 percent at 41 or 42. At 38, the honest target may be 25 to 40 eggs — two, three stimulations — not “fifteen and sleep in peace.”
Failure has floors. She does not return to the tank. She returns and the egg does not survive. It survives and does not fertilise. It fertilises and does not become a blastocyst. It becomes one and does not stick. It sticks and miscarries. The career gained time. The laboratory gained an annual fee. The child was not included.
Support — the insurance that is not cashed
| Source | Who uses the egg | Child from those eggs |
| UCLA/SART, 5–7 years | 5.7% | about 2 in 100 freezes |
| 47,000 planned OC | 2.5–3% | — |
| Cobo et al. | 9.3% | 2.7% of those who froze |
| Longer follow-up | 15–25% | Still a minority |
Egg donation: allograft, pre-eclampsia and ethics
When the egg is another woman’s, the pregnancy ceases to be semi-allogeneic — half the child is the gestating woman’s — and becomes an almost complete allograft: genes of the donor and the father, none of the woman who carries. The pregnant woman’s immune system meets a guest it did not train for. Metas find pre-eclampsia four to five times more frequent than in natural conception and two to three times more than in IVF with own egg. Clinical translation used in the literature: about one in six donor-egg pregnancies develops pre-eclampsia (range 13.5 to 18 percent against about 5.9 percent in IVF with own egg). Placentas of egg donation show more chronic villitis, deciduitis, fibrin, infarction — the face of immune conflict. The greater the HLA mismatch, the worse the picture. Donation of both gametes worsens it further. A sister donor, in theory, gentles it.
There is a second body in the scene: the donor. Young, often paid, submitted to the same Gonal-F, the same puncture, the same risk of hyperstimulation — for a child that will not be hers. The classic ethics of organ donation asks for benefit to the donor or informed altruism without a market. Commercial egg donation operates at the limit: one woman’s ovary becomes the input of another woman’s clinic. Consent from someone in her early twenties with a college debt is not the same consent as a treatise of bioethics.
The chimeric question
In ordinary pregnancy, cells of the fetus enter the mother’s blood and may stay for years — microchimerism. In egg donation those cells do not carry half the mother. They carry the genome of a third person. Allogeneic male fetal cells have already been documented circulating up to nine years after birth in women who gestated a child from a donated egg. No one knows, with a long cohort, what that guest does in the long run: tolerance, autoimmunity, nothing. The doubt is already enough for the consent form to stop treating donation as “just an egg.”
Chimera, here, is not myth. It is a cell of another lineage living in a body that did not match HLA with it. Industry speaks of a “gift.” Immunology speaks of a gestational allograft. Both sentences can fit in the same paragraph. Only one of them usually fits on the clinic’s website.
Consumer law: what has already changed and what the brochure has not yet read
A fertility clinic is a supplier. A private patient is a consumer. The Consumer Defence Code applies. Clear information is a basic right (art. 6, III). Misleading advertising — false or omitted — is forbidden (art. 37). The offer binds: what Instagram asserts with a large number becomes part of the contract (art. 30). A clause that puts the consumer at an exaggerated disadvantage is void (art. 51, IV). A disproportionate service is revised (art. 6, V). The burden of proving that the advertising did not mislead falls on the supplier when the claim is plausible (art. 38).
What is changing is not the code — it is the use. Judges who once contented themselves with “obligation of means, therefore nothing to indemnify” now ask whether the advertised rate was the rate of that age’s cycle, whether the extra was sold without evidence, whether the form was signed under the effect of the protocol. The obligation remains one of means. The means, however, include informing the risk of accreta in programmed FET, the real chance of the frozen egg, the fact that the health plan may not pay. Advertising of “success” without a denominator is, at the limit, an omission capable of inducing error. Not every negative beta becomes moral damages. Every rate without age and without a started cycle is a candidate for a binding offer.
The plan pays the face and does not pay the puncture
Law 9.656/1998, art. 10, III, takes “artificial insemination” out of the mandatory cover of the reference plan. The Superior Court of Justice, in Theme 1.067, nailed it: unless there is an express clause, the plan is not obliged to fund IVF. ANS, by resolution, treats IVF as a cousin of insemination and leaves it out. What the plan owes under so-called family planning, on the roll, is conversation, IUD, education. It is not a laboratory.
In 2022 Law 14.454 said the ANS roll is a basic reference, not a blind ceiling: a procedure off the list may be covered if there is evidence and a therapeutic plan, or a recommendation from Conitec or a serious international agency. Part of the doctrine read in that the chance to reopen IVF. In practice Theme 1.067 remains the stamp the operator pastes on the refusal, and many judges follow the stamp. The new law changed the weather. It did not knock down the express exclusion of art. 10, III.
At the other counter, the same Court — Third Panel, opinion by Justice Nancy Andrighi, 2026 — ordered the plan to pay facial feminisation surgery in a gender-affirmation process. It is not “aesthetics,” the Court said: it is integral health, prevention of the suffering of incongruence, a procedure recognised by the Federal Medical Council, present in the public system, listed in TUSS and on the roll with no guideline that bars it. ANS, in technical opinion 26/2024, already said that mastectomy, hysterectomy and related items on the roll, without a specific restriction, belong in the transsexualising process when the doctor asks.
The comparison discomforts both sides, and that is why it serves the brief. It is not an argument against the care of a trans person. It is an X-ray of how the system classifies a body. Infertility is a disease recognised by WHO; plan law treats it as an optional item when the path is IVF. Gender incongruence is suffering recognised by WHO; the face, in this cut of the Court, becomes therapeutic and mandatory. Both sufferings are real. Only one of them the operator is being forced to bank at this moment. The question of the report is not “who deserves a child.” It is: who wrote art. 10, III, and who decides, in 2026, what is aesthetics and what is health.
Consulting-room ethics
Autonomy requires the rate of that age in that cycle, not the cumulative 74 percent. Non-maleficence asks that the extra without proof not travel as gold standard. Justice asks that it be said that ninety-odd percent of this market is private and that the public-system queue is measured in years. The conflict is structural: the person who indicates, performs and charges for the tank is on the same side of the invoice.
None of this forbids IVF. It forbids the tone of miracle.
During the cycle the ovary is not a metaphor. It is an organ. It receives too much hormone, swells, sometimes twists, sometimes hyperstimulates, sometimes hides a tumour nobody went to see because the puncture clock was already set. Industry learned to count follicles. Counting follicles is easier than looking at the organ that releases them. If there is a last sentence: first the ovary, then the stock.
Isabel de Fátima Alvim Braga
Physician and lawyer. Master’s in Collective Health. Doctorate in Public Health and Environment. Public servant. She went to the Senate, said what she thought, was sued, and is not afraid of cancellation.
References (as in the Portuguese original)
1. Gonal-F leaflets (follitropin alfa): Farmaindex, DailyMed, EMD Serono monographs. OHSS; torsion and thromboembolism; anaphylaxis very rare.
2. Ovidrel professional leaflet (Merck Brazil) and EMD Serono IRI. OHSS ~4% in trials; severe <0.5%.
3. Japanese ART registry, 174,591 singletons (2017–2020). PAS programmed cycle aOR 9.12 vs fresh; 1.48% vs 0.16%.
4. Same registry, vaginal birth after single embryo. HRC aOR 11.4; 1.4% vs 0.11%. Thin endometrium 16% vs 3.8%. Uterine surgery aOR 3.57.
5. Jauniaux et al., BJOG. Praevia in ART, singleton, OR ~3.76.
6. Petersen et al., Hypertension/AHA, 2022. FET +74% hypertensive disorder; siblings ~2×.
7. SMFM Consult Series #60. IVF = moderate pre-eclampsia factor.
8. ASRM 2024 and metas of ART and ovarian/borderline cancer. Borderline OR often ~1.5–2; invasive inconsistent; small absolute increase; causality not established.
9. Greil, Verhaak and psychosocial reviews of infertility and IVF.
10. U.S. national survey of families: treatment associated with lower divorce risk (selection).
11. Kjaer et al., Acta Obstet Gynecol Scand, 2014. 47,515 Danish women; no child, OR of ending the union up to 3.13.
12. Sundby et al., 2007: 10 years after IVF, 56% thought of divorce, 17% ended it. Martins et al., Fertil Steril 2014: 14% separated during treatment.
13. Immune cells of the tube. PMC10955054, 2024. Immune privilege; TGFB1 and IL-10.
14. Neutrophils in the female genital tract. Mol Hum Reprod, 2025.
15. Scientific Reports, 2025: tube selects sperm. Single-cell atlas of the tube (Michigan).
16. Fitzpatrick et al., Proc. B, 2020. Follicular fluid attracts semen of specific men.
17. Human Reproduction Update, 2026; Kekäläinen & Evans, Proc. B, 2018. Gamete-mediated choice.
18. Bianchi et al.: Juno (egg) and Izumo (sperm); block to polyspermy.
19. Kroener/UCLA, AJOG 2025. 5.7% return in 5–7 years; CLBR 28.9% among those who return.
20. Planned OC, ~47,000, return 2.5–3%. Cobo et al. 2016: 9.3% return; 2.7% have a child.
21. RBMO 2018: use 3.1–9.3%; theoretical threshold ≥50%. Johnston/Monash: stock > use.
22. Per-egg chance models: 6–8% under 35; 2–3% at 41–42; survival 85–95%.
23. Egg-donation and pre-eclampsia meta: OR ~5 vs natural and ~3 vs autologous IVF; 13.5–18% vs ~5.9%.
24. Placental lesions in egg donation: villitis, deciduitis, fibrin. Front. Med. 2024; Placenta 2023.
25. Tregs and alloimmunity in egg donation. J Reprod Immunol 2024.
26. Williams et al., 2009: allogeneic fetal cells up to 9 years after egg donation.
27. Brazilian Consumer Defence Code: arts. 6, III and V; 30; 37; 38; 51, IV.
28. Law 9.656/1998, art. 10, III. STJ Theme 1.067: plan not obliged to fund IVF unless a clause says so.
29. Law 14.454/2022: roll as basic reference. Tension with Theme 1.067.
30. STJ, Third Panel, 2026, Andrighi. Facial feminisation: mandatory cover; not aesthetics.
31. ANS Technical Opinion 26/2024. Transsexualising process and procedures on the roll.
32. CFM 2.320/2022 and 2.386/2024. SisEmbrio/Anvisa. CRM-MG 9/2025. Anvisa NT 31/2025.
