Christine Cotton Preservation archive
English Français

← All interviews · All research follow-ups →

Pfizer : scandaleuses révélations sur l'essai clinique

Décoder l'éco · YouTube · Watch ▶

📅 2023-06-15🎙️ Décoder l'éco (host) with co-guest Pierre (OpenVAET, anonymous data analyst)📺 Décoder l'éco — YouTube channel⏱ 59 min
“The quasi-totality of people who received a Pfizer vaccine got a product that was not the one of the trial. Of the 40,000 in the study, you found only 250 who received the same one.”

Key statements

  • Cotton, biostatistician for 23 years and former CEO of a CRO repeatedly audited against GCP, wrote her January 2022 expertise of the Pfizer C4591001 trial as a critique against Good Clinical Practice standards. ▶ 02:01
    Sources: Cotton — Expertise of the Pfizer C4591001 trial (January 2022); ICH-GCP guidelines
  • The first Pfizer C4591001 report — the one announcing 95% efficacy — dates from December 2020 and is the foundation document Cotton dissects. ▶ 03:55
    Sources: Pfizer C4591001 interim clinical study report (Dec 2020)
  • PCR-positive participants in Argentina (Buenos Aires military hospital, ~5,700 subjects) had their samples shipped to Pearl River, a Pfizer-controlled lab in the United States — only the central test was authoritative, even when it conflicted with a positive local test. ▶ 08:05
    Sources: Pfizer C4591001 Protocol
  • The trial's stated objective was to prevent symptomatic COVID — never to prevent transmission. The final efficacy analysis was triggered when 164 cases (vaccine + placebo) had accumulated. ▶ 12:19
    Sources: Pfizer C4591001 Protocol
  • Pregnant or breastfeeding women, immunocompromised patients, and patients with comorbidities or autoimmune disease were excluded or so under-represented that no subgroup analysis exists for them. Co-administration with other vaccines was not studied; transmission was not studied; asymptomatic cases were not studied. ▶ 13:15
    Sources: Pfizer C4591001 Protocol; Comirnaty Risk Management Plan (Dec 2020, EMA)
  • The FDA, in October 2020, allowed labs to submit results based on three months of follow-up maximum (with half of participants followed less than two months), whereas previous vaccine guidance required six months to a year. ▶ 14:36
    Sources: FDA October 2020 vaccine guidance amendment
  • Symptomatic-PCR-confirmed endpoint is biased: participants self-report symptoms, vaccine-group antipyretic use is ~3.5× higher than placebo, and antipyretics suppress fever — a COVID symptom. No symptoms, no test, no COVID — to the vaccine's benefit. ▶ 15:04
    Sources: Pfizer C4591001 interim clinical study report (Dec 2020) — antipyretic use tables
  • Preclinical monkey data and Phase 1 already showed a decline of neutralizing antibodies starting at two months. Pfizer deliberately did not measure antibody titres at three months post-dose-2 — the timepoint when the drop would have been visible. ▶ 19:02
    Sources: Pfizer C4591001 Protocol (immunogenicity schedule); Pfizer preclinical macaque study
  • Pfizer vaccinated the placebo group shortly after the first interim analysis, destroying the long-term control comparison. Cotton notes this is only ethically justified for cancer trials where the active product is clearly superior — not for a vaccine. ▶ 20:05
    Sources: Pfizer C4591001 Protocol (amendment unblinding/crossover)
  • Pierre (OpenVAET) summarizes the database-level red flags: the trial was not truly blinded; vaccine-group subjects were less likely to be tested when symptomatic; 301 subjects disappeared from the database (with Argentina as the major outlier); reports to regulators disagree with the database; and the marketed product is not the one tested ("bait and switch"). ▶ 26:26
    Sources: Pfizer C4591001 court-released database (PHMPT / FOIA)
  • Of ~44,000 trial participants, only about 250 received the Process-2 manufactured product that was later distributed to the public. The quasi-totality of vaccinated people thus received a product never validated at trial scale. ▶ 27:28
    Sources: Pfizer C4591001 Protocol (Process 2 amendment); Pfizer C4591001 trial database (court-released)
  • Russian hackers leaked Pfizer–EMA correspondence in which the EMA itself flagged a mRNA-integrity problem: intact-mRNA levels differed between trial-product and marketed product. MEP Michèle Rivasi has publicly highlighted this point. ▶ 28:15
    Sources: EMA-Pfizer leaked correspondence (Dec 2020, hacked)
  • Augusto Roux (Argentine lawyer, Pfizer adult trial, dose 2 on 9 Sept 2020) fainted and was hospitalized with pericarditis. BioNTech intervened to reclassify the pericarditis as suspected COVID. The Argentine site director Fernando Polack — principal investigator and signatory of the 95% efficacy paper — sought to have him classified as mentally ill. ▶ 38:53
    Sources: Augusto Roux medical file; Polack et al. NEJM 2020 (C4591001 publication)
  • Of the 301 missing subjects in the database, 111 (36.9%) were in Argentina alone — yet only 6 documented duplicates exist trial-wide. Subject deletions are normally documented via a specific form; the volume here is unexplained. ▶ 42:25
    Sources: Pfizer C4591001 court-released database (PHMPT / FOIA); ICON Firecrest data-management platform documentation
  • Cotton and OpenVAET wrote to FDA's Peter Marks (Director of CBER) about the database anomalies. They received an acknowledgement from a deputy but no substantive response — "about two and a half months ago, and we're still waiting." ▶ 43:09
    Sources: Correspondence with FDA / CBER (Peter Marks)
  • The 250 subjects who received Process 2 (from 19 October 2020 onward) had an adverse-event reporting rate 2.7× higher than subjects who received the higher-purity (Process 1) product — and Process 1 itself was "already not great." ▶ 47:35
    Sources: Pfizer C4591001 court-released database (PHMPT / FOIA)

Documents cited

  • Pfizer C4591001 Protocol — NEJM supplementary materials
  • Pfizer C4591001 interim clinical study report (December 2020) — FDA briefing documents, Dec 2020 VRBPAC
  • Pfizer C4591001 court-released trial database (XPT / SAS files) — PHMPT / FOIA disclosures (phmpt.org)
  • EMA-Pfizer leaked correspondence (Dec 2020, hacked) — Russian-hacker disclosure; coverage in BMJ and elsewhere
  • Polack et al. NEJM 2020 — C4591001 efficacy publication — New England Journal of Medicine
  • Correspondence with FDA / CBER (Peter Marks) — Cotton / OpenVAET personal correspondence
  • Brook Jackson — Ventavia whistleblower BMJ report (Thacker, 2021) — BMJ 2021;375:n2635
  • ICON Firecrest data-management platform documentation — ICON plc — clinical-systems product page

Research follow-ups

All research follow-ups →