Hepatitis A Vaccine Clinical Trial Data Reviewed

How was the hepatitis A vaccine tested before approval? This page compiles the primary-source clinical trial data from every pre-licensure study behind HAVRIX and VAQTA.

Clinical trial data for the two hepatitis A vaccines on the US childhood schedule. Both were the first-ever hepatitis A vaccine of their kind, meaning no licensed comparator existed at the time of trial design — manufacturers were not constrained by an "ethical" requirement to use an active control. Both manufacturers nevertheless avoided placebo. Havrix used Engerix-B (a hepatitis B vaccine licensed with 4-day monitoring and no control). Vaqta used an injection of AAHS adjuvant + thimerosal — substances that are pharmacologically active and known to induce immunological effects.

Havrix (GSK)

Source: FDA-approved package insert; GSK licensure submission.

Control Chain Problem

Havrix's safety was benchmarked against Engerix-B — a vaccine that was:

• Licensed in 1989 with no control group
• Monitored for only 4 days post-injection in its own trial
• Subject to ICAN litigation in which the FDA could not produce safety data beyond a few days

The safety of the first hepatitis A vaccine in US history thus rests on a hepatitis B vaccine with a 4-day monitoring window and no baseline.

Engerix-B was chosen as the control because it was a GSK product — the same manufacturer. Havrix is "as safe as" Engerix-B; Engerix-B is "as safe as" nothing.

Vaqta (Merck)

Source: FDA-approved package insert; Merck licensure submission.

AAHS Control Problem

AAHS (Amorphous Aluminum Hydroxyphosphate Sulfate) is a proprietary Merck aluminum adjuvant. It is:

• Not inert — its function is immunological stimulation
• The same adjuvant used in Gardasil and Gardasil 9
• Known to induce autoimmunity in laboratory animals — used specifically for this purpose in autoimmune disease research
• Neurotoxic and cytotoxic at the cellular level

Thimerosal Control Problem

Thimerosal is an ethylmercury-containing preservative. It is:

• Not inert
• A mercury compound (mercury compounds are neurotoxic)
• Removed from most US childhood vaccines in 2001 due to public concern about cumulative mercury exposure (still in some flu vaccines)

Why This Is Not a Placebo

Vaqta's "control" was a combination of two biologically active, potentially harmful substances:

• If both groups had similar adverse event rates, this would be expected — both groups received potentially harmful injections
• If Vaqta caused additional harm beyond AAHS + thimerosal, the trial would detect it
• If AAHS + thimerosal caused harm in the "control" group, that harm would be invisible — attributed to "background rate"

The trial design ensures that any harm shared between the products would not be detectable.

Summary Table

Composition of the Vaccines and Their Controls

This section documents, for each product mentioned on this page, what was actually injected — the vaccine formulations on one side of the trial and the "control" formulations on the other. All figures are drawn from the FDA-approved package inserts for each product (linked in the Sources section) unless otherwise noted.

Havrix (GSK) — pediatric/adolescent formulation (0.5 mL dose)

The hepatitis A vaccine administered to the active arm of Havrix's pivotal trials.

Engerix-B (GSK) — the "control" used in Havrix's pre-licensure trial (0.5 mL dose)

Havrix's pre-licensure safety was benchmarked against Engerix-B, GSK's hepatitis B vaccine. Engerix-B was itself licensed (1989) without an inert-placebo control and with a 4-day post-dose monitoring window. Its ingredients are reproduced here because they were injected into the "control" arm of the Havrix trial. See Hepatitis B Vaccines (Pre-Licensure) for its own trial data.

Key observation: the Havrix "control" arm received a full dose of an aluminum-adjuvanted vaccine containing a recombinant hepatitis B surface antigen. It was not inert. Adverse events caused by aluminum-adjuvanted injection, yeast-derived recombinant protein, or hepatitis B vaccination would appear in both arms of the trial and would therefore not register as a Havrix-attributable signal.

Vaqta (Merck) — pediatric/adolescent formulation (0.5 mL dose)

The hepatitis A vaccine administered to the active arm of Vaqta's pivotal trials.

Vaqta's "placebo" control — composition of the control-arm injection

Vaqta's pivotal pediatric efficacy trial (Werzberger et al., NEJM 1992, the "Monroe trial") randomized children to receive either Vaqta or a "placebo" injection. The control-arm injection was documented as follows, drawn from Merck's licensure submission and from ICAN's consolidated "No Placebo Chart" (which reproduces the control-arm specifications from each FDA package insert).

Key observation: the Vaqta "placebo" was not an inert comparator. It was the vaccine's vehicle plus its aluminum adjuvant (and, per ICAN's documentation, the thimerosal preservative that was standard in the trial era), differing from the vaccine in only one variable — the viral antigen. This design answers the narrow question "does the HAV antigen cause additional events beyond what the adjuvant + vehicle cause?" It does not answer the broader question "is Vaqta safer or more reactogenic than an inert injection?"

If aluminum adjuvant or thimerosal contributed to an adverse event at some background rate, that event would appear in both arms and be classified as trial-wide "background," not as a safety signal. The trial cannot distinguish a harm caused by AAHS from a harm caused by AAHS-plus-HAV-antigen.

The "First-Ever" Pattern

Both Havrix and Vaqta were the first hepatitis A vaccines licensed in their categories. There was no licensed comparator that could ethically be used as an active control. Both manufacturers nonetheless designed their trials without a true placebo:

• GSK chose another of its own licensed vaccines (Engerix-B) — itself never validated against placebo
• Merck chose to inject the control group with the adjuvant alone, plus thimerosal

The deliberate choice of active comparators when no licensed comparator existed is direct evidence that placebo avoidance was intentional, not ethically required.

Schedule

Disease Context

Hepatitis A is a self-limiting viral infection spread through contaminated food and water (fecal-oral route). In healthy children, hepatitis A infection is typically mild or asymptomatic; severe illness and death occur primarily in adults with underlying liver disease. Before vaccine introduction, approximately 25,000–35,000 cases were reported annually in the US, with significant underreporting; deaths numbered fewer than 100 per year (overwhelmingly in adults).

See Also

Hepatitis A Vaccines (Post-Licensure), Pre-Licensure Safety Testing, Hepatitis B Vaccines (Pre-Licensure), HPV Vaccines (Pre-Licensure) (also uses AAHS), Childhood Vaccine Schedule, GSK, Merck

Sources

1. Havrix (GSK) — FDA-approved package insert (pediatric/adolescent formulation). Current label: Havrix prescribing information (GSK).

2. Vaqta (Merck) — FDA-approved package insert (pediatric formulation). Current label: Vaqta prescribing information (Merck).

3. Engerix-B (GSK) — FDA-approved package insert, referenced here because Engerix-B was the control arm in Havrix's pivotal trial. Current label: Engerix-B on DailyMed.

4. Werzberger A, et al. "A controlled trial of a formalin-inactivated hepatitis A vaccine in healthy children." New England Journal of Medicine. 1992;327(7):453–457. PMID: 1320736. (Vaqta pediatric efficacy trial — the Monroe, NY study.) PubMed record.

5. ICAN (ICAN), "No Placebo Chart." Consolidated documentation of control-arm formulations disclosed in FDA package inserts for every childhood-schedule vaccine: icandecide.org PDF.

Frequently Asked Questions

This is the plain-HTML rendering of vaccines/pre-licensure/Hepatitis A Vaccines.md, served for search engines, AI crawlers, and accessibility tools. The interactive 3D version of this wiki lives at vaccine-safety.org.