Influenza Vaccine's Efficacy Far From Perfect

View on the News

There's a Lot of Research on Flu Vaccines Right Now

The findings of the meta-analysis and the CIDRAP report are generally correct, but they aren’t new. We all know that influenza vaccine is a good vaccine, but far from a perfect vaccine. We’ve always known about the limitations of flu vaccine and the limitations of the data in various age groups, particularly in older persons. But I think there are ample data to show some efficacy of the vaccine. Some people don’t get sick, some get milder illness, and we do prevent deaths and hospitalizations (which was not addressed in the CIDRAP paper). Do we prevent them all? No. But preventing some is a very good thing.

I don’t agree with the exclusion of studies that used serologic end points. I think that there are many people who would not simply throw out all those studies, and would say that although they’re not perfect, they have merit if you add them all together, and they all point in the same direction, and they provide information that’s consistent with other studies. The objective should be to interpret the entirety of the evidence.

The recent pace of vaccine advances should also be acknowledged. The last 5 years have seen more research aimed at producing better influenza vaccines than there had been in the previous 40 years. Indeed, Fluzone High-Dose vaccine, which is much more immunogenic than the older TIV, has recently been approved for adults aged 65 and older. Efficacy studies of that vaccine are ongoing. In addition, both adjuvanted vaccines and cell-based influenza vaccines have already been approved in Europe and are expected to receive U.S. licensure soon. The last few years have just opened up the spigot on flu vaccine research. There’s a lot of it going on.

William Schaffner, M.D., is professor and chair of preventive medicine and professor of medicine at Vanderbilt University in Nashville, Tenn., serves on the data safety monitoring board for Sanofi-Pasteur and Merck, and is an occasional consultant for Novartis, GlaxoSmithKline, and Pfizer.



There are substantial gaps in the evidence supporting the effectiveness of influenza vaccines, particularly in the elderly, according to the findings of large systematic review and meta-analysis published online Oct. 25 in the Lancet Infectious Diseases.

Although the published report highlights the dearth of strong and consistent efficacy and effectiveness data for influenza vaccine in studies that met very strict criteria, it should not be interpreted as a suggestion to stop vaccinating, according Michael T. Osterholm, Ph.D, the study’s lead author and director of the center for infectious disease research and policy at the University of Minnesota, Minneapolis.

Dr. Michael T. Osterholm

Dr. Osterholm said his intent in conducting and publishing this analysis was not to cast doubt on current influenza immunization efforts, but rather to influence the pace of new vaccine development.

"There is a major barrier to entry right now for venture capital and start-up companies to bring new novel technologies forward. When you have a vaccine that’s universally recommended, said by public health to be effective, and is quite cheap, why would anybody spend a billion dollars to try to make a new vaccine that’s going to be much more expensive?"

Of a total 5,707 studies published from 1967 through Feb. 15, 2011, just 31 (17 randomized and 14 observational studies) met a list of strict criteria, the most salient being the use of influenza confirmed by culture or real-time polymerase chain reaction as an outcome (Lancet Infect. Dis. 2011 Oct. 25 [doi:10.1016/S1473-3099(11)70295-X]).

"Efficacy" was defined as the reduction in influenza risk assessed from randomized clinical trials, whereas vaccine "effectiveness" was determined in observational, "real-world" trials. Studies using serology end points to diagnose influenza were excluded because of limitations that were first identified more than 50 years ago: Increased antibody titers following vaccination with an inactivated vaccine make it difficult to document a fourfold rise in hemagglutinin antibodies necessary to confirm an influenza infection, thereby leading to a high number of false negatives. Nonetheless, a large number of published studies on the inactivated vaccine continue to rely on serology as an end point, the authors pointed out.

The analysis included studies of both the trivalent inactivated vaccine (TIV) and the live attenuated (intranasal) influenza vaccine (LAIV). Among the 10 randomized, controlled trials of TIV over 12 influenza seasons, analyses for 8 of the seasons showed significant efficacy, whereas 4 did not. Of eight studies that were conducted in healthy adults aged 18-64 years over a total of nine flu seasons, the pooled efficacy was 59%. One study conducted in children aged 6-24 months over two flu seasons produced dramatically different efficacy results: 66% in the first year, –7% in the next (JAMA 2003;290:1608-16). The "minus" essentially means zero, rather than suggesting an increased risk from the vaccine, Dr. Osterholm said in the interview.

No randomized, controlled trials met the criteria for children aged 2-17 years, or for adults aged 65 years and older. Indeed, conducting placebo-controlled trials in adults aged 65 and older would be considered unethical because influenza vaccine has been recommended for that age group since 1960, the authors noted.

The picture for LAIV was different: Of the 10 randomized controlled trials assessing LAIV efficacy during 12 flu seasons, 9 showed significant efficacy. All of these were done in healthy individuals. In children aged 6 months to 7 years, there were six studies covering eight influenza seasons. The vaccine was effective in all eight, with a pooled efficacy of 83%.

But LAIV data in other age groups were less impressive. One study of LAIV in adults aged 60 and older showed significant overall efficacy (42%), but – oddly – efficacy was lower in those aged 60-69 years and higher in those aged 70 and older (Vaccine 2009;28:228-34).

Of three randomized, controlled trials of LAIV in adults aged 18-49, none showed significant protection, and no such trials met the inclusion criteria for children aged 8-17 years, Dr. Osterholm and his associates reported.

Vaccine effectiveness varied in the nine observational trials of seasonal flu vaccine, with 6 of 17 embedded analyses showing significant protection against medically-attended, laboratory-confirmed influenza. The proportion of patients receiving TIV or LAIV was not explicitly stated for these studies, but based on the age and licensed use, most estimates were for TIV. Of the five observational studies that assessed effectiveness of the 2009 pandemic H1N1 vaccine, median efficacy was 69% (range, 60%-93%).

Except for the LAIV studies in children aged 7 years or younger, the data showed substantial variability by influenza season and by age group. In some influenza seasons, the level of protection was low or not evident. In contrast to the 70%-90% overall effectiveness that is often cited for the vaccine in seasons when the vaccine is well matched to circulating strains, "we noted this magnitude of effectiveness only for LAIV use in children aged 7 or younger," Dr. Osterholm and his associates wrote.