Editorial on: Protection against influenza hospitalizations from enhanced influenza vaccines among older adults: A systematic review and network meta-analysis

Abstract

Influenza is associated with important impacts on health and wellbeing for older adults. Its impact on outcomes over the short term (e.g., hospitalization, ICU admission, death) are better understood than those over longer-term time horizons (e.g., persistent functional decline, need for assistance in Activities of Daily Living, or nursing home admission). The acute care burden alone means that it is important to prevent and mitigate the severity of illness to the greatest degree possible in those most in need of protection.¹ Influenza vaccination is the best protection we currently have at our disposal. However, its benefit is for older adults is currently limited for two very different reasons: (1) immune dysregulation experienced with aging (aka immunosenescense)reduces response to vaccines,² and (2) uptake of influenza vaccination remains suboptimal, even among older adults.³

The reduced protection that standard dose influenza vaccines (SD-IV) offer older adults has resulted in work to develop vaccine products that will be more effective in older adults. These include high-dose inactivated vaccine (HD-IV), adjuvanted inactivated vaccine (adj-IV), and recombinant influenza vaccine (RIV), collectively known as enhanced vaccines.⁴ Each of these attempts to increase the effectiveness of the vaccine for older adults either by increasing the dose of antigen (HD-IV, RIV), or adding an adjuvant (a local immune stimulant). Numerous studies of the effectiveness of these enhanced vaccines are available, with results suggesting that they are generally superior to standard dose influenza vaccines.⁵

However, the evidence landscape is very complex for several reasons. Vaccine effectiveness can be impacted by host, virus, and vaccine-related factors (Figure 1). Host factors may include age, sex/gender, immunosenescence, frailty, comorbidities, immunosuppression, prior history of immune exposures, and imprinting by first infection.⁶ There is substantial season-to-season variability in influenza activity, and in the circulating strains of influenza causing disease. Vaccines also differ in their composition (e.g., hemagglutinin antigen only versus other components, seed strains grown in eggs versus other media) in ways that may result in differences in relative vaccine effectiveness in different years. Individual-level randomized controlled trials (RCTs) are extraordinarily expensive and cannot be used for the necessary on-going assessment of vaccine effectiveness as viruses change. However, even the best observational trials are at risk of bias and un-measured confounders. Cluster- and pragmatic RCTs are being increasingly explored, but ensuring and assessing quality in these trials is more difficult than in individual-level RCTs.^{7, 8} More importantly, older adults are a heterogeneous population who have relatively high rates of serious health events. This means that they are often excluded from RCTs because of difficulties ensuring that serious events are unrelated to study vaccines. The combination of usual exclusions along with barriers such as frailty and mobility experienced by many older adults has resulted in very little recruitment of adults, especially over 75 or 80 years of age, in RCTs.⁹ For example, this was even seen in recent trials of respiratory syncytial virus (RSV) vaccines, where recruitment of this population was a priority.^10-12

Given these trial limitations, the multiple possible comparisons, different study designs, different definition of outcomes considered, and lack of head-to-head comparisons among the enhanced vaccines, the evidence landscape is very complex. The complexity and limitations of the evidence base have presented challenges for national immunization technical advisory groups (NITAGs) charged with developing guidance on the use of these vaccine products and for decision-makers and jurisdictions wishing to provide the best influenza vaccination programs to protect their populations.

In their present article published in JAGS, Ferdinands et al. have conducted a well-designed systematic review and network meta-analysis examining relative vaccine effectiveness of different enhanced influenza vaccines for older adults, compared among themselves and with standard dose vaccine.¹³

Their review was rigorously conducted and reported. They searched appropriate databases for studies published between 1990 and 2023 reporting relative vaccine effectiveness for high dose, adjuvanted or recombinant influenza vaccines (RIV), either compared with one another or to standard dose vaccine. They included not only individual-level RCTs but also cluster-randomized trials and observational studies, and identified 32 relevant studies for inclusion. The inclusion of observational studies is a strength not only because vaccine effectiveness varies significantly between seasons, but also because older and frail adults are often excluded from RCTs.¹⁴ They found that each of the enhanced vaccines was approximately 11%–18% more effective in preventing influenza hospitalization versus SD vaccines. No differences between the effectiveness of the different vaccines were identified, although there were limited data permitting head-to-head comparisons.

Ferdinands et al. describe a logical analytical approach allowing all relevant comparisons to be considered. The study methodology is sound and the use of PRISMA reporting guidelines is a strength.

They performed relevant subgroup stratification by circulating influenza subtype, study funding source, and use of laboratory-confirmed outcomes. They also included sensitivity analyses; in the RCT analyses this was a “leave one out” approach to be able to ascertain the relative contribution of individual RCTs to the overall findings. In the case of the observational studies, sensitivity analysis included restricting analysis to studies with less risk of bias and including additional studies that reported a composite outcome of influenza hospitalization and emergency department visits.

There are a number of novel analyses and data visualization techniques that enhance the value of this analysis. The leave one out analysis for the RCTs highlighted that one particular study (Johansen 2023), the pilot for a large pragmatic RCT of HD versus SD vaccine in Denmark,¹⁵ was particularly influential in the pooled results (see supplementary Figures 3–7), in part because of the large reduction in all-cause mortality identified during this trial. Identifying studies with differing findings is important because the reasons for the difference may have implications both for policy and future trial design and implementation. In this case, the reasons for the difference favor HD vaccine remain unexplained, although the size of the reduction in all-cause mortality, especially when compared with that of other outcomes, and the fact that a difference is seen outside of influenza season, may suggest a problem in randomization. Readers who are worried about the validity of indirect comparisons in network meta-analyses comparing vaccine effectiveness in different years will also appreciate the figures provided in Ferdinands et al.'s article comparing direct and indirect comparisons of the effectiveness of different vaccines provided in the supplementary figures.¹³

Ferdinands et al. are to be congratulated on having presented a well-designed and well-communicated study of a very complex topic. They have explained their rationale and analyses and displayed results clearly and have presented a thorough discussion of limitations.

Their work is topical and highly relevant in informing both policy and practice. Notably, their finding that all three of these enhanced products are very similar in terms of their effectiveness lends support to recent recommendations by the United State's Advisory Committee on Immunization Practice (ACIP) and Canada's National Advisory Committee on Immunization (NACI), both of which now have a preferential recommendation for any of the three enhanced products, considered equal to one another, versus standard dose vaccine in adults age 65+.^{5, 16} Importantly, both ACIP and NACI are also careful to recommend that in the absence of availability of an enhanced vaccine product, any available approved SD vaccine should be used. This is an important reminder that it is vaccination rather than vaccines that saves lives, and a SD-IV given is much more effective than an enhanced vaccine missed.

Given that older adults are among those most in need of protection from vaccination, it is critical to continue research and evaluation of vaccine products and programs with a geriatric lens in mind. This includes consideration of age and frailty, along with their associated features of immunosenescence, propensity to present “atypically” when ill (which is relevant for case definitions and outcome monitoring in research studies), and vulnerability to unique adverse outcomes (such as persistent functional declines) beyond those usually considered in studies. All of these factors support inclusion of geriatrics expertise at all levels of vaccine product development, testing, membership on National Immunization Technical Advisory Groups (NITAGs, such as ACIP and NACI), program implementation, surveillance, and evaluation.^{17, 18}

Melissa K. Andrew and Allison McGeer cowrote the article.

MKA reports grants from Sanofi, GSK, Merck, Icosavax and Pfizer for research studies relating to vaccine preventable illnesses in older adults, and is a member of Canada's National Advisory Committee on Immunization. AM reports grants to her institution from Pfizer and Sanofi, and personal payments for advisory board and DSMB membership from AstraZeneca, Merck, GlaxoSmithKline, Moderna, Novavax, Pfizer, Sanofi, and Seqirus.

No funding was received for this article.