Argentina Logged 32 Violations in Pfizer's Maternal RSV Trial 20 Months Before the FDA Approved It for Pregnancy

Participant 040 was 16 when she was randomised into Pfizer’s Phase 3 maternal RSV vaccine trial at the Hospital Italiano de Rosario. In Argentina the protocol required women aged 18 or older. She signed an adult consent form and received the intervention, and the identity document on file carried modifications that did not match the hospital’s or the laboratory’s systems. All of this comes from the administrative file of Argentina’s drug regulator, ANMAT, as reviewed by Realpolitik.

Her case heads a list of 32 findings of non-compliance that ANMAT compiled in a final technical report dated December 15, 2021. Several of the findings concern newborns whose serious events were misclassified or never properly entered in the hospital chart, and the report recommended a sumario, the formal proceeding that can end in sanctions. On August 21, 2023, twenty months after that report, the FDA approved the vaccine, Abrysvo, for use in pregnancy, and the European Commission granted its marketing authorisation two days later. Argentina’s sanction against Pfizer became final on September 11, 2025.

Maryanne Demasi reported this week that documents she examined show ANMAT recording failures to report serious adverse events involving newborns. According to her report, the violations had affected the quality of clinical data and safety reporting collected at the research centre. She asks what regulators in the United States, the European Union, the United Kingdom, Canada and Australia knew when they approved the product. The Argentine file answers part of that question.

What three inspectors found in Rosario

Inspectors Emanuel Kawata, Nicolás Ciranna and Manuel Rey Funes were at the hospital from August 31 to September 2, 2021, examining a site where Clemente Fernando Miguel Aguilar was the authorised principal investigator. While he was absent for health reasons, Mario Aguilar ran the study operationally without formal delegation. An electronic query had already flagged participant 040’s birth date in early June 2021. It was treated as a transcription error, and the deeper check did not come until August. The defense argued that she altered the document herself, and Realpolitik’s answer is that this does not relieve the system of its duty to verify eligibility. ANMAT ordered her treatment assignment unblinded, the ethics committee informed, monitoring of mother and baby continued, and her data excluded from the efficacy and safety analyses. The file records that order but does not show how it was carried into the published results.

The newborn findings go to the core safety question for a vaccine given in pregnancy. One baby’s clinical history contained no recorded assessment of neonatal respiratory distress syndrome. In another, low birth weight and congenital cytomegalovirus infection were not recorded as a serious event. A third infant’s congenital heart disease was entered in the chart as “no serio,” non-serious, although it had been reported as a serious event. In a fourth case, prematurity and respiratory difficulty were communicated to Pfizer, the ethics committee and ANMAT but not properly recorded in the hospital’s own primary source. The dates for a systolic heart murmur did not agree.

The delivery room failed the protocol as well. For participants 014 and 020, the protocol-required maternal and cord blood draws at birth were never taken, and procedures were carried out by “personal no delegado en el estudio,” staff not delegated to the study. Realpolitik is careful to say this does not prove those staff were untrained. Some participants were not re-consented in time on the updated form, some consent dates were missing or contradicted the clinical history, and vulnerability assessments were missing in two cases. These lapses occurred in a trial enrolling pregnant women, one of whom was a minor. Investigational product reached the Rosario site on February 12, 2021, before its formal initiation visit on March 30. Finding 32 records the same sequence at a second centre, Hospital Militar, which received product on January 12 and was initiated on January 27.

Realpolitik draws a firm line around all of this: “el expediente no demuestra que estos desvíos hayan causado una lesión concreta a una madre o a un bebé.” The file does not show that the deviations injured any mother or baby. The outlet’s conclusion is narrower. When required samples go uncollected, data that should have been gathered identically from every participant is lost, and “la confiabilidad del dato también queda debilitada,” which means the reliability of the data is weakened as well.

The clause Pfizer could not delegate

The authorisation came first. ANMAT approved protocol C3671008 in Disposición DI-2020-7240, signed September 24, 2020, by then-administrator Manuel Limeres. Article 4 states that Pfizer, as sponsor, “conserva la responsabilidad legal que le compete por el cuidado de los participantes y la integridad de los datos”: it keeps legal responsibility for the care of participants and the integrity of the data. ANMAT’s good-clinical-practice rules allow a sponsor to hand trial work to a contract organisation, but the legal responsibility stays with the sponsor, so the failures in Rosario’s charts are Pfizer’s under the terms of its own permit.

The trial, MATISSE, was the pivotal evidence for approval, and its results appeared in the New England Journal of Medicine in 2023. Demasi describes it as the principal clinical evidence behind Abrysvo.

The safety question for this product has always been the babies. In a review its own staff published in npj Vaccines, the FDA reported that 5.7 percent of infants born to vaccinated mothers in the Phase 3 trial (202 of 3,568) arrived preterm, against 4.7 percent on placebo (169 of 3,558). Before 34 weeks the count was 21 against 12. The agency’s reviewers wrote that the imbalance “might represent a true vaccine-associated risk of preterm birth.” The FDA restricted use to 32 through 36 weeks of gestation, a window in which the gap persisted at 4.2 percent against 3.7 percent, and required four postmarketing safety studies covering preterm birth and hypertensive disorders of pregnancy. The same paper records that a competing maternal RSV candidate, RSVPreF3-Mat, was halted in 2022 after an imbalance in preterm births and neonatal deaths. The FDA’s approval announcement adds that low birth weight and jaundice ran higher in infants of vaccinated mothers, and that pre-eclampsia occurred in 1.8 percent of vaccinated women against 1.4 percent on placebo.

Europe accepted more of that risk. The EMA’s scientific committee issued a positive opinion on July 20, 2023, and the authorisation covers vaccination between weeks 24 and 36, a window opening eight weeks earlier than the FDA would allow.

Across 7,126 infants, the preterm difference comes down to a few dozen births, so the reliability of each site’s newborn record carries weight. The Rosario documents do not show that the trial’s preterm totals were wrong, and only one of the newborn findings, the prematurity case that was reported to Pfizer and the regulator but not properly charted, involves gestational age at all. The others concern how serious neonatal events were assessed and classified, which is the evidence a reviewer uses to decide whether a safety imbalance is noise. The file establishes a narrower point. At one site in the pivotal trial, the source record for newborn serious events failed a regulator’s inspection, and under Article 4 Pfizer was answerable for that record.

From report to referral: 21 months

Nélida Agustina Bisio co-signed the inspection plan on August 30, 2021, and the final report on December 15. ANMAT applied an Indicación de Acción Oficial, a formal official-action notice, to Pfizer and the investigator, kept recruitment closed and demanded information on every Argentine centre. In January 2022 the file began moving toward the sumarios office, with the proceedings addressed expressly to Bisio, and the formal opening did not follow for almost two years. In October 2023, as director of the DERM directorate, she signed the referral to the sumarios coordinator. Subadministrator Valeria Teresa Garay signed the formal opening that December, four months after the FDA and European approvals.

Bisio became head of ANMAT on December 26, 2023. She set the fines on August 25, 2025, and on September 11 signed disposition DI-2025-6789 making them final, one million pesos for Pfizer and one million for Aguilar. “Ese millón era el máximo de la escala pecuniaria aplicada,” Realpolitik notes: the million was the ceiling of the penalty scale applied, and at that month’s official exchange rate it was well under a thousand US dollars per party. The decision does not explain how either amount was reached, although the sponsor and the investigator carried different obligations. A generic charge under point C.9 of ANMAT Disposition 6677/10 was dropped without stated reasons. Pfizer and Aguilar paid in October 2025.

Even that sanction exceeds what ANMAT did in a comparable case involving the same sponsor. In 2020 the agency logged 20 findings at the site Fernando Polack led in Pfizer’s COVID-19 vaccine trial and served the same official-action notice on Polack and Pfizer. Realpolitik found no equivalent sumario order in that case, where ANMAT opted for reinforced surveillance and corrections.

Nothing in the file, as Realpolitik describes it, shows whether the December 2021 report reached the FDA, the EMA or any other agency that approved Abrysvo, and the FDA’s public approval materials cited here do not mention it. The public record also does not show whether Pfizer’s regulatory submissions disclosed the Rosario findings, or what ANMAT’s demand for data on every Argentine centre produced. The file reviewed contains no criminal referral over the minor’s altered identity document, though Realpolitik notes that this absence does not rule out communication by other routes. In April 2026 the file was circulating through ANMAT’s legal affairs office for “prosecución del trámite,” continuation of proceedings, and the record does not explain why.

The copy of the trial material that ANMAT audited was stamped “Pfizer Confidential,” and the participant data remain with the company. The FDA holds the biologics license file and ordered the postmarketing preterm-birth studies. Neither has made public whether Argentina’s 32 findings were in front of the reviewers who approved this vaccine for pregnant women in August 2023.

Sources

1. Realpolitik – “Pfizer y ANMAT: una menor embarazada, 32 hallazgos y una sanción tardía”
2. ANMAT – Disposición DI-2020-7240 authorising protocol C3671008 (September 24, 2020)
3. Maryanne Demasi, MD Reports – “EXCLUSIVE: Serious violations in Pfizer’s maternal RSV vaccine trial” (November 2, 2026)
4. npj Vaccines – FDA reviewers’ account of the Abrysvo maternal approval (2024)
5. FDA – “FDA Approves First Vaccine for Pregnant Individuals to Prevent RSV in Infants” (August 21, 2023)
6. European Medicines Agency – Abrysvo EPAR
7. New England Journal of Medicine – Kampmann et al., “Bivalent Prefusion F Vaccine in Pregnancy to Prevent RSV Illness in Infants” (2023)