Influenza vaccination: a one-shot deal to reduce cardiovascular events

This editorial refers to ‘Influenza vaccination and risk of hospitalization in patients with heart failure: a self-controlled case series study’†, by H. Mohseni et al., on page 326. Influenza contributes to significant morbidity and mortality across the globe every year. The estimated number of hospitalizations attributable to influenza infection are grossly underestimated among patients with cardiovascular (CV) disease, in whom a severe upper respiratory infection can initiate a cascade of physiological events that ultimately lead to acute worsening or exacerbation of underlying heart disease. Patients with heart failure in particular are at a high risk for influenza-related complications,1 including hospitalizations, which occur more frequently during the winter months, often in conjunction with influenza infection.2 A number of mechanisms support a causal association between influenza infection and CV events (Figure 1). Influenza infection can result in increased metabolic demand, adrenergic surge, endothelial dysfunction, hypercoagulability, and hypoxia.3 Systemic immune stimulation and inflammation in the setting of influenza infection can trigger acute plaque rupture. Influenza predisposes patients to develop other infections such as pneumonia, which may itself be associated with increased CV risk due to prolonged elevation in cytokine levels and a procoagulant state.4–6 Moreover, influenza is known to cause myocardial depression directly,7 possibly mediated through increases in proinflammatory cytokines,8,9 and histological evidence of myocardial injury, myocarditis, and myocyte necrosis has been demonstrated from patients following influenza-related deaths.10 Influenza vaccination has been accepted as an effective, well-established strategy to reduce influenza-related illness, and this potential benefit in high-risk CV populations is reflected in recommendations regarding vaccination in these individuals. The World Health Organization recommends that influenza vaccination efforts be prioritized among individuals at highest risk for virus-related complications, including those of older age or with certain chronic conditions.11 Cardiac societies have also encouraged annual influenza vaccination among patients with CV disease.12–14 The body of data showing that influenza vaccination could directly reduce CV events has been less compelling, but is growing. Several studies and observational analyses of clinical trials have demonstrated lower rates of CV outcomes in patients who received influenza vaccination.15–17 In one meta-analysis of nearly 300 000 patients in five randomized trials, vaccination was associated with a 27% reduction in myocardial infarction, and a 40% reduction in all-cause mortality.18 In another meta-analysis of 6735 patients from six randomized controlled trials, influenza vaccine was associated with a 36% reduced risk of major adverse CV events.19 This benefit was most robust in patients with recent acute coronary syndrome. In this issue of the journal, Mohseni and colleagues examined the association between influenza vaccination and the risk for CV hospitalization in patients with heart failure using a UK-based primary care data set linked with hospitalization records from Hospital Episode Statistics.20 This is the largest investigation of its kind in heart failure, for which efficacy data on influenza vaccination have been limited to smaller scale, observational data sets. In this elegant study, rates of CV, respiratory, and all-cause hospitalizations on an individual patient level were compared from two years—one year during which the patient was vaccinated, and an adjacent year during which the patient was not vaccinated. The authors found a lower risk for CV hospitalizations, including hospitalizations for heart failure, during the vaccinated year compared with the risk during the unvaccinated year, an effect most pronounced within 120 days after vaccination. When categorizing risk periods by 30-day intervals post-vaccination, the authors appropriately dismissed the first 30 days to account for the approximate length of time for the peak immune response to influenza vaccine to occur. Observational analyses of medical interventions are challenged by healthy user bias, which is difficult to control for completely in statistical models. Also, there may be fundamental differences between patients who receive influenza vaccination year in, year out, and those, included in this study, who do not. Nonetheless, the authors implemented several approaches for minimizing bias, including examining hospitalizations among patients during two consecutive unvaccinated years, comparing events before and after the end of the influenza season, and assessing the effect of vaccination on cancer-related hospitalizations, which would not be likely to be altered by influenza vaccination. Surprisingly, they noted a less pronounced, but still statistically significant, association between receipt of influenza vaccine and lower rates of respiratory hospitalizations. As several conditions and other types of infections contribute to this endpoint, this outcome may be too imprecise to allow capture of potential vaccine benefits. Despite both observational and meta-analyses of clinical trials, and the existence of worldwide guidelines supporting the use of vaccination in high-risk populations, the rates of influenza vaccination clinically vary widely in patients with heart failure. In a contemporary heart failure clinical trial conducted in 47 countries, the use of influenza vaccination ranged between 0 and 77% depending on the country, with only 21% of patients overall receiving influenza vaccination within a year of trial enrolment.21 Some would argue that the low rates of vaccination and the compelling data call for more clinical trials to demonstrate the benefit of vaccination on CV outcomes in high-risk populations. Indeed, clinical trials comparing influenza vaccination with placebo in underutilized areas are currently ongoing or planned. Yet others would agree that the plethora of data, both randomized and observational, supporting the use of influenza vaccination in CV patients, makes a compelling argument that effective public health campaigns, rather than clinical trials, could substantially reduce morbidity and mortality. The incomplete uptake of these recommendations argue for a more forceful advocacy within treatment guidelines. Nevertheless, the best dose and formulation for vaccination remain unclear. Influenza vaccine contains both A- and B-type viral lineages, and is commercially available as several formulations. These differ by the amount of each viral antigen included (high dose vs. standard dose), or by the total number of antigens (trivalent or quadrivalent). A high-dose formulation is approved in the USA and Canada for medically stable individuals over the age of 65, as older adults exhibit blunted immune responses to standard dose influenza vaccination. High-dose influenza vaccine has been shown to elicit a more pronounced immune response, measured by antibody titre levels post-vaccination, and in a large randomized controlled trial high-dose vaccine reduced laboratory-confirmed influenza compared with standard dose influenza vaccine.22,23 Patients with heart failure have reduced antibody-mediated and altered cell-mediated immune responses to influenza vaccine, which are both important for mounting adequate protection from influenza infection.24,25 In a small randomized trial, heart failure patients randomized to a double dose of influenza vaccine demonstrated higher antibody titres compared with those taking standard dose vaccine, suggesting that blunted immunity may be overcome with a more potent vaccine in these patients.26 The exact vaccine formulations used in the cohort studied by Mohseni et al. were not known, making it impossible to investigate the influence of vaccine type on reduction in hospitalizations, and observational data are unlikely to determine the most effective strategy for influenza immunization. To address these issues directly, we have just initiated enrolment in the ‘INfluenza Vaccine to Effectively Stop CardioThoracic Events and Decompensated heart failure (INVESTED) study (NCT02787044), an NIH-sponsored 9300 patient trial that will determine whether a high-dose strategy compared with a standard-dose strategy will be more effective in reducing cardiopulmonary hospitalization and mortality in heart failure and post-infarction patients. Regardless of the dose or formulation utilized, the data from Mohseni et al. add to the growing body of literature supporting the use of influenza vaccination to lower morbidity and mortality in high-risk CV patients, possibly by reducing CV events and hospitalizations directly, and further the argument for widespread yearly immunization in patients with advanced heart disease, especially those with heart failure. There are few interventions in all of medicine that are as low-cost, low-risk, well tolerated, or easy to administer and with such large potential clinical benefits. Don’t we owe it to our patients to offer them this one-shot deal? Conflict of interest: Drs. Solomon and Vardeny serve as Co-principal investigators on a NIH-sponsored vaccine trial for which Sanofi-Pasteur is providing vaccine free of charge. Dr. Solomon's institution has received grants from Sanofi for projects unrelated to the topic of this editorial. Influenza infection increases the risk for acute cardiovascular events (myocardial infarction and heart failure exacerbations) through several mechanisms involving pro-inflammatory mediators, neurohormonal activation, and activation of the coagulation cascade.