Pfizer and Moderna Vaccines Face Off Again

While risk of COVID-19 infection was lower among veterans vaccinated with Moderna compared with Pfizer, both vaccines were associated with low risk, researchers found.

During the period of the Alpha variant, the estimated risk of documented infection was 5.75 events per 1,000 persons (95% CI 5.39-6.23) in the Pfizer group and 4.52 events per 1,000 (95% CI 4.17-4.84) in the Moderna group, reported Barbra Dickerman, PhD, of Harvard T.H. Chan School of Public Health in Boston, and colleagues.

The excess number of events per 1,000 persons in the Pfizer group versus Moderna group was 1.23 (95% CI 0.72-1.81) for documented infection, 0.44 (95% CI 0.25-0.70) for symptomatic disease, and 0.55 (95% CI 0.36-0.83) for hospitalization, the authors wrote in the New England Journal of Medicine (NEJM).

For more severe outcomes during the Alpha period, however, the differences appeared negligible: an excess of 0.10 (95% CI 0.00-0.26) for ICU admission and 0.02 (95% CI -0.06 to 0.12) for the Pfizer group.

Prior studies have attempted to compare mRNA vaccines, but Dickerman and colleagues noted their limitations.

“There has been a need for studies that compare the vaccines head-to-head, are large enough to provide precise risk estimates for severe COVID-19 outcomes, have adequate adjustment for confounding and sound study design, include racially diverse groups, and separately address periods that have predominance of different SARS-CoV-2 variants,” they wrote.

A second, far-smaller analysis — conducted as part of the current study — showed a larger excess risk of documented infection for a group receiving the Pfizer vaccine during the Delta period, but the difference wasn’t significant.

Dickerman’s group added that while both Pfizer and Moderna’s vaccines are mRNA-based, they differ in dose, interval between priming and boosting, and “the lipid composition of the nanoparticles used for packaging the mRNA content.”

“It has been suggested that this observed difference in antibody levels translates into a difference in the risk of COVID-19 outcomes,” the authors noted.

However, an accompanying editorial from NEJM editor-in-chief Eric Rubin, MD, PhD, and deputy editor Dan Longo, MD, saw this as less of an indictment of Pfizer and more of a reassurance for how well both vaccines work.

“For any given person, the difference in vaccine efficacy between [the Pfizer and Moderna vaccines] is unmeasurable,” they wrote. “So the lesson we take away is not about differences — it’s about similarities. … The message is that the best vaccine is the one that’s available.”

Dickerman’s group examined data from the Department of Veterans Affairs (VA) for adults who received an mRNA COVID vaccine from January to May (for Alpha analysis) or July to September (for Delta analysis) with no prior COVID infection, who lived outside a long-term care facility, and had smoking and body mass index status available in their electronic health record. Participants also used the VA health system in the prior year.

Primary outcomes were documented infection, symptomatic disease, hospitalization, ICU admission, and death during the Alpha-predominant period (January to July), and only documented infection for the Delta period due to the low number of serious events.

For the Alpha period, 219,842 veterans who received the Pfizer shot were matched to an equal number of veterans who received Moderna. Median age was 69, nearly all (93%) were men, and three-quarters were white. Two-thirds had hypertension, and 47% had obesity. Median follow-up was 126 days.

Over 24 weeks, 559 participants had symptomatic disease, 411 were hospitalized, 125 were admitted to an ICU, and 81 died.

Dickerman’s group estimated the number needed to vaccinate with Moderna over Pfizer was 813 to prevent one infection, 2,273 to prevent one case of symptomatic disease, and 1,818 to prevent one hospitalization.

For the Delta period, 3,580 Pfizer and Moderna recipients were matched. Here, the excess risk of documented infection was 6.54 events per 1,000 persons in the Pfizer group, albeit with wide confidence intervals (95% CI -2.58 to 11.82), for a risk ratio of 1.58 (95% CI 0.85-2.33).

Dickerman’s group enumerated the usual caveats of an observational study, including potential for misclassification, confounding, and limited generalizability given its population of older male veterans.

Rubin and Longo noted that these data do not take the effect of boosters into account. They also pointed out that these data only compared efficacy, not safety, of the two vaccines.

“We are lucky to have such good options,” they wrote. “Vaccination with any vaccine is far better than remaining unprotected.”