Relative effectiveness of BNT162b2, mRNA-1273, and Ad26.COV2.S vaccines and homologous boosting in preventing COVID-19 in adults in the US

Văn Hùng Nguyễn, Cathérine Boileau, Alina Bogdanov, Meg Sredl, Mac Bonafede, Thiérry Ducruet, Scott Chavers, Andrew Rosen, David Martin, Philip O. Buck, Diana L. Esposito, Nicolas Van de Velde, James A. MansimedRxiv 2023 · DOI

This research paper, published as a preprint on medRxiv in February 2023, provides a large-scale, real-world comparison of the three major COVID-19 vaccines used in the United States. By following a massive cohort of adults from their primary vaccination series through their booster doses, the authors offer a rare "head-to-head" look at how these vaccines performed against the Delta and Omicron variants.

TL;DR

In a massive real-world study of US adults, the Moderna (mRNA-1273) vaccine consistently showed higher relative effectiveness than the Pfizer-BioNTech (BNT162b2) and J&J (Ad26.COV2.S) vaccines. This advantage persisted across both the primary vaccination series during the Delta wave and homologous booster doses during the Omicron surge, with the most significant benefits seen in adults aged 65 and older.

What Problem Does This Paper Address?

While initial clinical trials proved that COVID-19 vaccines were highly effective against the original strain of the virus, the real-world landscape changed rapidly with the emergence of the Delta and Omicron variants. Most existing research focused on "absolute effectiveness" (vaccine vs. no vaccine).

However, for public health officials and individuals, the more pressing question became "relative effectiveness" (rVE): how do these vaccines compare to one another in the wild? This paper addresses the lack of longitudinal, head-to-head comparisons that follow the same group of people from their initial shots through their boosters, specifically during periods dominated by different viral variants.

Methods

The researchers conducted a retrospective observational longitudinal study. They employed a 1:1 matching strategy, pairing individuals who received the Moderna vaccine with those who received Pfizer or J&J based on sex, geographic region, race, age (within 5-year brackets), and the specific date of vaccination.

To analyze the data, they used multivariable Cox regression models to calculate adjusted hazard ratios (HR). The relative vaccine effectiveness (rVE) was then derived using the formula: $rVE = (1 - HR_{adjusted}) \times 100$. This approach allowed them to isolate the effect of the vaccine brand while controlling for other potential confounding variables.

Datasets

The study utilized a robust, integrated real-world dataset:

  • Veradigm Health Insights: An Electronic Health Record (EHR) database containing healthcare interactions for over 55 million patients in the US.
  • Komodo Health: Linked pharmacy and medical claims data, which provided a "closed" view of patient insurance claims to ensure continuous tracking of medical encounters.
  • Study Size: The final analysis included over 4.4 million individuals for the primary series comparison and over 736,000 for the booster comparison.

Key Insights and Ideas

  • The Moderna Edge: Across almost every metric—medically attended cases, outpatient visits, and hospitalizations—the Moderna vaccine (mRNA-1273) outperformed its competitors.
  • Variant Resilience: The study effectively captured two distinct eras: the primary series analysis took place during the Delta-dominant period, while the booster analysis occurred during the Omicron-dominant period. Moderna's relative advantage held up in both.
  • The Senior "Booster Bonus": One of the most striking findings was that the Moderna booster's relative advantage over Pfizer was significantly higher in seniors (≥65 years) than in younger adults. This suggests that the higher dosage or specific formulation of the Moderna vaccine may better counteract "immunosenescence" (the natural weakening of the immune system with age).
  • Antibody Persistence: The authors point to external data suggesting Moderna elicits higher and more persistent antibody titers, which likely explains the clinical outcomes observed in this study.

Key Results

The quantitative findings highlight a clear hierarchy in relative effectiveness:

Primary Series (Moderna vs. Others): * Vs. Pfizer (BNT162b2): Moderna was 23% more effective at preventing medically attended COVID-19 and 19% more effective at preventing hospitalizations. * Vs. J&J (Ad26.COV2.S): Moderna was 50% more effective at preventing medically attended cases and 57% more effective at preventing hospitalizations.

Booster Dose (Moderna vs. Pfizer): * Overall: The Moderna booster was 14% more effective than the Pfizer booster against medically attended COVID-19. * Age Breakdown: For those aged 18–64, the rVE was 13%, but for those ≥65 years, it jumped to 35%.

Limitations and Future Work

The authors acknowledge several limitations inherent in observational research:

  • Non-Randomized: Unlike a clinical trial, participants weren't randomly assigned vaccines, meaning unmeasured behavioral or health differences between groups could exist.
  • Testing Bias: As the pandemic progressed, many people moved to at-home testing, which isn't captured in EHR or claims data. This could lead to an undercounting of mild cases.
  • Short Follow-up: The booster phase of the study had a maximum follow-up of 90 days, so the long-term durability of the booster's relative advantage remains to be seen.
  • Future Work: The authors suggest that future studies should include more data on younger age groups and continue to monitor effectiveness as new variants and updated "bivalent" vaccine formulations are introduced.

Summary updated

Abstract

Abstract Background Few head-to-head comparisons have been performed on the real-world effectiveness of COVID-19 booster vaccines. We evaluated the relative effectiveness (rVE) of a primary series of mRNA-1273 versus BNT162b2 and Ad26.COV2.S and a homologous mRNA booster against medically-attended, outpatient, and hospitalized COVID-19. Methods A dataset linking primary care electronic medical records with medical claims data was used for this retrospective cohort study of US patients ≥18 years vaccinated with a primary series between February and October 2021 (Part 1) and a homologous mRNA booster between October 2021 and January 2022 (Part 2). Adjusted hazard ratios (HR) were derived from 1:1 matching adjusted across potential covariates. rVE was (1-HR adjusted ) x 100. Additional analysis was performed across regions and age groups. Results Following adjustment, Part 1 rVE for mRNA-1273 versus BNT162b2 was 23% (95% CI: 22%–25%), 23% (22%–25%), and 19% (14%–24%) whilst the rVE for mRNA-1273 versus Ad26.COV2.S was 50% (48%–51%), 50% (48%–52%), and 57% (53%–61%) against any medically-attended, outpatient, and hospitalized COVID-19, respectively. The adjusted rVE in Part 2 for mRNA-1273 versus BNT162b2 was 14% (10%–18%), 13% (8%– 17%), and 19% (1%–34%) against any medically-attended, outpatient, and hospitalized COVID-19, respectively. rVE against medically-attended COVID-19 was higher in adults ≥65 years (35%; 24%–47%) than those 18–64 years (13%; 9%–17%) after the booster. Conclusions In this study, mRNA-1273 was more effective than BNT162b2 or Ad26.COV2.S following primary series during a Delta-dominant period, and than BNT162b2 as a booster during an Omicron-dominant period. Key points mRNA-1273 was associated with a lower risk than BNT162b2 or Ad26.COV2.S of any medically-attended, outpatient, or hospitalized COVID-19 after primary series and of any medically-attended, outpatient, or hospitalized COVID-19 vs BNT162b2 after a homologous mRNA booster
BibTeX
@misc{nguyen2023relative,
  author = {Văn Hùng Nguyễn and Cathérine Boileau and Alina Bogdanov and Meg Sredl and Mac Bonafede and Thiérry Ducruet and Scott Chavers and Andrew Rosen and David Martin and Philip O. Buck and Diana L. Esposito and Nicolas Van de Velde and James A. Mansi},
  title = {{Relative effectiveness of BNT162b2, mRNA-1273, and Ad26.COV2.S vaccines and homologous boosting in preventing COVID-19 in adults in the US}},
  year = {2023},
  doi = {10.1101/2023.02.10.23285603},
  howpublished = {medRxiv},
}
Figure